CASE Network Report 77 - Development of Scenarios for Health Expenditure in the New EU Member States: Bulgaria, Estonia, Hungary, Poland and Slovakia

The views and opinions expressed here reflect the authors' point of view and not
necessarily those of CASE Network.
The report was prepared as a result of the Ageing, Health Status and Determinants of the
Health Expenditure (AHEAD) project carried out under the European Commission 6th
framework programme by the ENEPRI network. The contents of this publication is the
sole responsibility of the authors and can in no way be taken to reflect the views of the
European Union, CASE or other institutions the author may be affiliated to.
The publication was financed from a institutional grant extended by Rabobank Polska S.A.
Keywords: health care system, demographic projection, health care system revenues and
expenditures projection, CEE, NMS
CASE Network Reports No. 77/2008
JEL codes: H51 Government Expenditures and Health
Graphic Design: Agnieszka Natalia Bury
© CASE – Center for Social and Economic Research, Warsaw, 2008
ISBN 978-83-7178-453-8
EAN 9788371784538
Publisher:
CASE – Center for Social and Economic Research on behalf of CASE Network
12 Sienkiewicza, 00-010 Warsaw, Poland
tel.: (48 22) 622 66 27, 828 61 33, fax: (48 22) 828 60 69
e-mail: case@case-research.eu
http://www.case-research.eu

Contents
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6
Chapter 1. Applied model of health revenues and expenditures . . . . . . . . . . . . . . . . . . 7
1.1. Models used for health care expenditure projection . . . . . . . . . . . . . . . . . . . . . . . . 8
1.1.1. Actuarial models . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8
1.1.2. Macroeconomic models. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11
1.1.3. Econometric models . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11
1.2. ILO social budget model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12
Chapter 2. Health care system in analyzed countries. . . . . . . . . . . . . . . . . . . . . . . . . . 15
Chapter 3. Data sources . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17
Chapter 4. Assumptions and development of variables . . . . . . . . . . . . . . . . . . . . . . . . 20
4.1. Development of demographic variables . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20
4.2. Economic assumptions and projections . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29
4.3. Assumption regarding income elasticity of health expenditures . . . . . . . . . . . . . . 30
4.4. Assumption on health status and health service utilization . . . . . . . . . . . . . . . . . . 31
Chapter 5. Projection results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 35
5.1. Baseline scenario . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 36
5.2. Scenario of increased expenditures in the last year of life. . . . . . . . . . . . . . . . . . . 39
5.3. Diverse longevity scenarios . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 45
5.4. Diversified wage growth scenario . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 46
5.5. Labor market indicators diversified development scenario . . . . . . . . . . . . . . . . . . 48
5.6. Sensitivity analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 50
Chapter 6. Conclusions, discussion and policy recommendations. . . . . . . . . . . . . . . . 53
Annex 1 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 57
Annex 2 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 63
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 66
CASE Network Reports No. 77/2008 3

The Authors
Stanisława Golinowska, Professor of Economics, is a co-founder of CASE – Center for
Social and Economic Research and Vice Chairwoman of the CASE Supervisory
Council, where she conducts her research. She is also a Director at the Institute of
Public Health, Jagiellonian University Medical College (Krakow, Poland). Stanisława
Golinowska graduated from the Faculty of Economics (Warsaw University) and studied
at the Mannheim University as a scholarship- holder of the Humboldt Foundation. She
is the author of numerous articles and books on social aspects of economics and social
policy reforms
Agnieszka Sowa is a researcher at CASE - Center for Social and Economic Research,
and holds an M.A. in the field of public policy from Warsaw University, Department
of Sociology as well as an MsC in Social Protection Financing from Maastricht
University, Faculty of Economics and Business Administration. She also completed a
public policy course at the University of Pittsburgh, Department of Russian and
Eastern European Studies. Her interests include healthcare system and modelling of
health expenditures, poverty, unemployment and social assistance programmes.
Since 2004 she has been teaching social policy and social security at the Institute of
Public Health, Collegium Medicum, Jagiellonian University.
Ewa Kocot is an assistant in the Institute of Public Health at Jagiellonian University
Medical College (Krakow, Poland). She graduated with a Masters in mathematics from
Jagiellonian University. She also completed the postgraduate study of management
and administration of public health at the School of Public Health in Krakow. She is
especially interested in health indicators and their potential application in healthcare
sector financing, modelling and forecasting. These issues were the subject of her
recently completed doctoral thesis.
4
Stanisława Golinowska, Ewa Kocot, Agnieszka Sowa

Abstract
The report is a result of the Ageing, Health Status and Determinants of Health
Expenditure (AHEAD) project within the EC 6th Framework programme. The objective of
the research was to present the model of future health care system revenues and
expenditures in selected Central and Eastern European countries (CEE) which are now
the New EU Member States, and to discuss projection assumptions and results. Selected
countries include Bulgaria, Estonia, Hungary, Poland and Slovakia. The projections are
based on methodology adopted in the International Labour Organization (ILO) Social
Budget model. The projection examines impact of demographic changes and changes in
health status on future (up to 2050) health expenditures. Next to it, future changes in the
labour market participation and their imact on the health care system revenues are
examined. Results indicate that due to demographic pressures health expenditures will
increase in the next 40 years and health care systems in the NMS will face deficit.
Moreover, health revenues, expenditures and deficit/surplus are slightly sensitive to
possible labour market changes. Health care system reforms are required in order to
balance the disequilibrium of revenues and expenditures caused by external factors
(demographic and economic), and decrease the premium needed to cover expenditures.
Such reforms should lead, on the one hand, to the rationing of medical services covered
by public resources, and on the other, to more effective governance and management of
the sector and within the sector.
5
DEVELOPMENT OF SCENARIOS FOR HEALTH EXPENDITURE...

The objective of this report is to present the model of future health care system
revenues and expenditures in selected Central and Eastern European countries (CEE)
which are now the New EU Member States1, and to discuss projection assumptions
and results.
The report was prepared in result of the Ageing, Health Status and Determinants of
the Health Expenditure (AHEAD) project. The project was prepared within the European
Commission 6th framework programme. The report discusses results of health budget
models for Bulgaria, Estonia, Hungary, Poland and Slovakia that were prepared by
experts working within the AHEAD. Detailed results are presented in reports
concentrated on country studies (Gabos, Gal 2007, Golinowska, Kocot, Sowa 2007,
Kvetan, Palenik, Mlynek, Radvansky 2007, Rangelova, Sariiski 2007, Roovali 2007).
Health expenditure analysis and projections are based on the ILO social budget
model, a part of which is the health budget model. The model covers health care
system revenues and expenditures. It is suitable for the analysis of impact exerted by
demography (especially ageing) on health care system revenues and expenditures. Up
to date, data and information sources in New Member States that could be used for
the long-term comparative projections have been limited. A structure of the report is
as follows. After the presentation of applied methodology and model, data and
information sources are described. Next, model assumptions are discussed. In the
projection an attempt was made to standardize the assumptions, although the
assumptions for each country are also responsive to country policy processes, latest
policy proposals and reforms. The development of specific model variables during
projection period is described in detail. Finally, the report presents projection results
together with sensitivity analysis and conclusions with regard to future development
of health care system revenues, expenditures and financial balance. The report
concludes with policy recommendations.
6
Introduction
1 Estonia represents Baltic countries, Bulgaria - Balkan countries, and Slovakia represents countries from
Central Europe. Poland is a unique country in that it has demographic and epidemiological characteristics
similar to Slovakia, but is much larger, with a high share of rural population. In Hungary, demographic
processes related to second demographical transition started much earlier and are still dynamic, thus being
a reference for other CEE countries.

Chapter 1.
Applied model of health revenues
and expenditures
A design of health care system expenditure projection is a complicated and risky
task. Christiansen at al. (2006) have argued that theoretical, especially
macroeconomic background for health expenditure determinants is rather modest, if
not weak. Research in the area pointed at income as the dominant determinant of
health expenditures (Newhouse 1977), while demographic variables were not taken
into account or – when included in the analysis – impact of demography was not
significant from the standpoint of expenditure development. A number of analyses
emphasized the importance of policy processes, political affairs and quality of
management in the health sector that affect the supply side of expenditures (i.e.
Getzen 1992.) Only in late 1990’s did the research begin to concentrate on impact of
demography on aggregated health expenditures.
Analyses pertaining to health care revenues and expenditures are complicated in
that they must take into account numerous data from various fields, such as
demography and labor market, macroeconomics, health status and quality of
management in the health sector, and finally, policy and institutional implications.
Furthermore, it is necessary to perform additional analysis to estimate the inter-relations
between factors that have impact on the level of revenues and expenditures.
In view of the reasons referred to above, most projection models are not as
comprehensive and complex, and take into account only some of the variables that
affect the level of aggregated revenues / expenditures.
The choice of projection method and determinants strongly depends on the time
horizon of projection, i.e. the period included in projection. For example, changes of
age structure of the population are so slow that they will not affect the outcome of
short- or even medium-term projections. On the other hand, factors such as health
care sector reforms or changes in medical procedures may have strong impact on
results of short-, medium-, and long-term projections (Lee 2002). These factors are the
CASE Network Reports No. 77/2008 7

most difficult to foresee and project due to potential rapid changes in the policy
process. In consequence, they may introduce high level of uncertainty in projections.
Length of projection period affects not only the variables, but also the unit of
expenditure measurement (Getzen 2000.) According to Getzen, nominal monetary
unit is the best measure for short-term projections, and the variables covering
insurance, labor market and inflation represent main determinants of health
expenditure level. Real monetary unit of per capita expenditures is useful for mid-term
projections, as these are mostly driven by income. In the case of long-term
projections, the best unit to present the level of expenditures is the percentage of GDP
reflected in expenditures (Getzen op.cit., p.56-57.)
1.1. Models used for health care expenditure projection
The most common projection methods can be summarized in three groups:
(Mahal, Berman 2001):
• Projections based on actuarial models,
• Macroeconomic projections,
• Projections based on econometric models.
1.1.1. Actuarial models
Aggregated actuarial models are most common in health care expenditure
projections. They are based on an equation that quantifies the level of expenditures in
the year t :
where:
( ) , , , , =ΣΣ ⋅
HEt N HE
a
t a g t a g
g
HEt – health expenditures in the year t,
g – gender,
a– age cohort (most typically, 5-year age cohorts),
Nt,a,g – headcount of age cohort ain the year t by gender g ,
HEt,a.g – average per capita expenditures in the age cohort a, by gender g in the
year t.
8

Among the most popular actuarial models, two approaches can be identified: the
“traditional” approach, where the share of population in the last year of life (in each
age cohort) is not taken into consideration, and the “generalized” projection
approach, where the level of per capita expenditures is differentiated depending on
the time period left till death (Kildemoes et al. 2006).
Projection prepared by L. Mayhew (Mayhew 2000)2 is a typical example of
“traditional” approach to health care expenditure modeling and projection. It looks at
the level of health expenditures in two regions of the world, namely: the more developed
countries (MDC) and less developed countries (LDC), and covers the period up to
2050.
The analysis of expected development of health expenditure growth takes two
types of factors into account: demographic changes (changes in the size and structure
of the population), and other changes aggregated, including technological changes,
changes in per capita medical service utilization, and cost factors. Inter-relations
between the level of expenditures in the base year and the level of expenditures in year
t are quantified in the equation:
where:
( ) ( ) (0) t rp ru H t H e + = ⋅
t – year under analysis,
H(t) – expenditures in the year t,
rp – demographic change coefficient,
ru – other factors’ change coefficient.
The indicator responsible for demographic changes takes into account changes in
health needs depending on population size and costs of care for the elderly. An index
of relative level of expenditures3 depending on age cohort and year is used. Index
stability was assumed for the MDC (different values of the index are only assumed for
different age cohorts based on data for the period of 1980-1990), while in the case of
LDC index value is linked to mortality by age. In the MDC, index value is slightly below
historical data. Due to lack of current data, the same values were assumed in the LDC.
A similar approach with demographic variables taken into account is represented,
9
2 Projection is an extention of projections based on economic and demographic model of IIASA (International
Institute for Applied Systems Analysis.) The model analyses the impact of ageing on economic development
and, with respect to social insurance, concentrates on pension expenditures.
3 Index has a value 1 for the age kohort 0-4, for other age cohorts its value is quantified in relation to expenditures
for the youngest cohort.

among others, by Denton et al. (2001), Antolin et al. (2001) and by Fuchs (1998.)
Projection of short-term expenditures in Medicare and Medicaid schemes in the
USA, prepared by Cutler and Sheiner (1998), is an example of a generalized approach
to analysis. Two factors that influence changes in per capita expenditures were taken
into account: age of death and disability indicator among survivors. The age of death
is shifting towards older cohorts due to increasing life expectancy and decreasing
mortality. At the same time, the number of persons for whom a given year is the last
year of life is decreasing for younger cohorts.
Two models recently developed by the Ageing Working Group (European
Commission 2006) and the OECD experts (OECD 2006), also belong to the group of
models that apply the factor of increasing health expenditures in the last stage of life
in the projection.
Actuarial models are used to build financial balance models as well, covering the
revenue and the expenditure side of health care sector. These models are applied to
the analysis of public and private health insurance schemes. The main purpose behind
the application of financial balance models is to evaluate financial sustainability of
health insurance system and estimate the amount of health insurance premium
needed to assure system sustainability4. Total health insurance revenues are
calculated as the sum of health insurance premiums paid by (or on behalf of) the
insured. Additional revenues of insurance institutions can also be included in
projections, depending on the source of revenue and type of activity of the insurer (i.e.
interest rates from the capital or capital investments.) Most balance models use
similar methodology in calculations, but some models are restricted to variables
pertaining to health care sector only (health insurance, mostly), while others include
system variables and are well set in the macroeconomic context.
Projection prepared by the Management Sciences for Health (MSH)5 in the USA is an
example of a financial balance model applied to health insurance projections.
Projection is used to evaluate health insurance sustainability, with diverse scenarios
of future development of actual contribution rate and insurance premium needed to
cover the expenditures.
Another example of a model based on similar methodology is the WHO model
applied to regional/local insurance systems. The model is suitable for projecting the
impact of the introduction of regional/local insurance systems on the national health
care system (Carrin, Ron 1993). As with other models, the level of revenues,
10
4 Financial sustainability of insurance system is understood as the level of revenues higher or equal to the level
of expenditures.
5 Management Sciences for Health – a private, non-profit organization established in 1971, which works for health
managers, policy-makers, medical service providers and often patients; and assists in solving problems in the
health care sector by providing information and expertise.

expenditures and insurance financial balance are estimated. Likewise, financial
structure of health care system, the number of insured and projected level of medical
service utilization are described.
1.1.2. Macroeconomic models
The second category of models that can be distinguished are macroeconomic
models that present health care system in a broad macroeconomic context. This
approach allows for a broader analysis of the reasons underlying the increase in
health expenditures and its long term economic implications.
A two-sector model of health care system and social insurance by M. Warshawsky
(1994, 1999). The model is based on simple equilibrium between two production
sectors: health care sector and the sector covering the rest of the market with two
production factors: namely, labor and capital. Health care sector development is
driven by the Leontiew production function6, while the rest of the market is driven by
the Cobb-Douglas production function7. Aggregated supply of labor is a function of
demographic, social and institutional factors (i.e. high school and university
enrolment, women’s labour market participation, retirement age, etc.) Projection
covers 40-year period. The results include not only the estimation of health and
pension expenditures in relation to GDP, but also the relations between labor and
capital in each of the two sectors, the level of consumption per capita, productivity in
health care sector, and impact of various insurance systems on the level of
expenditures. The model also includes assumptions on technological changes: better
utilization of the capital from technical perspective (capital deepening), decreasing
productivity in the health care sector, and technological development on the market.
1.1.3. Econometric models
The work by Getzen and Poullier (1992) is an example of econometric projections.
It includes health expenditure projections for 19 OECD countries. Projections were
made using various econometric methods: naive method, assuming the same growth
rate of expenditures as in the previous year, moving averages method, exponential
smoothing and regressions on historical data, projecting each country’s growth rate
as the average growth rate for all 19 countries combined, econometric model, and
econometric models with additional macroeconomic information.
In addition, combination of methods listed above was also applied in the model. In
11
6 q = min{K,L}, where q – production size, K,L – factors of production (capital and labor.)
7 q = a Kb Lc, where a,b,c – constant.

general, projections based on econometric models are rarely used to estimate the level
of long-term expenditures, rather, they are applied to short-term projections.
However, if the projection aims at the evaluation of impact of demographic changes
on health expenditure, projection period should be extended, because changes in the
size and age structure of the population are clearly visible only in the long term.
1.2. ILO social budget model
Projections of health care system financing and financial balance were made, as
was already mentioned, based on the actuarial model prepared by the Social Security
Department (SECSOC) of International Labor Organization (ILO).
The ILO social budget model has been broadly applied in the countries which
cooperate with ILO, including CEE countries. In the past, the model was used to
assess financial balance of the whole social welfare system in these countries. It is
suitable for middle- and long-term projection of revenues and expenditures of major
social welfare schemes, and for the simulation of impact of demographic, economic
and legislative changes on social welfare financing. The model can be used for short-term
evaluation of already existing schemes, as well as the simulation of future
reforms of some parts of the system (i.e. the introduction of a defined contribution
pension system or insurance-based health care system.) The social budget model
consists of four sub-models (Scheme 1):
• Pension system sub-model;
• Health care system sub-model;
• Other sub-models, including unemployment benefits, family benefits, social
assistance scheme.
The baseline model of social budget is restricted to health care budget, which in
the CEE countries is mostly represented by the institution of health insurance budget.
The baseline model has been further adjusted to country situation, health care system
institutional framework, country legal regulations and policy proposals.
The ILO health budget model has certain advantages compared to other potential
models that could have been applied for health care financing projections in accession
economies.
• This model reflects the character of health care sector, as it comprises many
different determinants that influence financial situation in the sector.
• It constitutes a basic platform of a kind, allowing wide adaptation to specific
12

Scheme 1. ILO social budget model structure
Pension model
Projection of sector-specific
variables, revenue and
expenditure
SOCIAL BUDGET MODEL SCHEME
Source: Scholz W., Cichon M., Hagemejer K. (2000), Social budgeting, International Labor Office, Geneva.
country health system situation. There is great variety in health care systems,
especially from the standpoint of their financing structure, so no single model
can be applied to each country.
• The ILO model allows for a comparison of the expenditure and the revenue side
and, in consequence, for a deficit or surplus calculation. Such feature is
absolutely essential, especially in financial projections. Therefore, it is useful not
only for international comparisons of health care system expenditures – as was
the case with the Ageing Working Group projection – but also for country-based
policy makers who are interested in financial sustainability of health care system
in the long term.
• This model, unlike some others, takes demographic changes into account (ageing
of the population.)
Following the structure of the ILO social budget model, four types of factors are
used for the purpose of projecting health care system revenues and expenditures
(Scheme 2).
The model applied in the research is slightly different from the original social
budget model presented above (Schemes 1 and 2). Main groups of variables used in
the analysis are classified under four headings:
• Demographic: TFR, LE, population structure (share of 65+ population) ;
13
Health model
variables, revenue and
expenditure
Other social protection
variables,
revenue and expend.
Demographic,
labour market,
macro-economic
and governmental
model
Social budget model

Scheme 2. Key factors that influence financial balance of the health care system
Demographic
and labor market
factors
Source: Scholz W., Cichon M., Hagemejer K. (2000), Social budgeting, International Labor Office, Geneva.
• Labor market: participation and employment rate, wages;
• Macroeconomic: GDP, inflation and labor productivity;
• Health status and utilization patterns.
Policy variables were not included in the analysis, with the assumption that the
legal framework and financial mechanisms in health care system remain unchanged
during projection period.
One of the most significant modifications to the health module of the social budget
model (in case of projections for Hungary and Poland) is the introduction of per
capita expenditures instead of average medical service utilization level (number of per
capita primary care visits and hospital days.) This modification was made due to
greater availability of data on expenditure levels per capita, by gender and by age than
the data pertaining to average medical service utilization in these two countries.
Moreover, the introduction of average per capita expenditures allowed for death-related
costs scenario.
The health care budget model applied in projections is not limited to health
insurance in a given country, but covers the whole public financing of the health care
sector. This implies that not only insurance revenues and expenditures, but also
budgetary revenues and expenditures are included.
14
Financial balance
of the health care
system: revenues
and expenditures
Health status of the
population, medical
service utilization,
medical technologies
Economic
factors
Governmental
factors

Chapter 2.
Health care system
in analyzed countries
A vast majority of transition countries (CEE and CIS) achieved the first objective
of health care reform in that they introduced health insurance with an earmarked
fund established from payroll tax, instead of integrated budgetary funding financed
from general taxation. Health insurance was introduced gradually: in Hungary social
insurance fund, including health care, was separated from budgetary funding in 1990,
and then (in 1992) health insurance fund was further isolated from social insurance
fund8. Over the same period of time, the Baltics designed health insurance approach
within the framework of reforms prepared for the new independent states (1991). In
Slovakia, health insurance was introduced in 1994, in the Czech Republic in 1997, in
Romania in 1998 and, finally, in Bulgaria and Poland in 1999.
Table 1. Introduction of health insurance in analyzed countries (2003)
Country
Year of HI
introduction Contribution (in %)
Source: Golinowska, Sowa (2006), Kvetan, Palenik (2006), Rangelova (2006), Remak, Gal, Nemeth (2006),
Roovali (2006)
Medical circles strongly supported the introduction of health insurance. It was
believed that the inflow of funds to the system would be more secure and stable when
the funds are contingent upon economic results (payroll tax) rather than upon
political decisions made every year while setting the priorities in budgetary
15
8 In Hungary, health insurance premium was earmarked as late as in 1996. Before that, one premium for all
insurable social benefits was calculated. The total premium equaled 52.5 % gross compensation, and the
health insurance premium isolated therein – 22% (Gaal 2004, p. 37.).
Share of HI in the total
HC funding (in %)
Bulgaria 1999 6.0 32.0
Estonia 1991 13.0 66.0
1990 as SI
Hungary
1996 - separate HI 14.0 71.6
Poland 1999 9.0 62.7
Slovakia 1994 14.0 86.0

expenditures. Another advantage was related to the possibility of defining a package
of insurance benefits for the purpose of separating the public component, and thus
creating the space for private health insurance.
Graph 1. Composition of health expenditures in analyzed NMS countries by source (% GDP)
8,0
7,0
6,0
5,0
4,0
3,0
2,0
1,0
0,0
Bulgaria Estonia Hungary Poland Slovakia
Source: Golinowska, Sowa (2006), Kvetan, Palenik (2006), Rangelova (2006), Remak, Gal, Nemeth (2006),
The introduction of health insurance generated tension between increasing health
needs on the one hand, and relatively low health care funding and requirements of
low taxes and low non-wage labor costs needed to stimulate employment and
economic growth, on the other. In consequence of these tensions, health care sector
has been systematically falling in debt. Debts are incurred either by insurance
institutions or by medical service providers, depending on institutional framework of
the health care sector. High level of indebtedness is stable or even growing
(Golinowska, Sowada, Woźniak 2007), which has adverse impact on the effectiveness
of health care management. Future, long-term financial decisions should be made
with the awareness of these tendencies, as well as external determinants of health care
sector in the future.
16
Roovali (2006)
soc.sec publ.gov others

Chapter 3.
Data sources
Health budget model is based upon national data supplemented (or adjusted) with
the data collected and used by the international institutions in their comparative
analyses. These include: United Nations, Organization for Economic Co-operation
and Development, International Labor Organization, Eurostat, and European
Commission. Domestic data includes the data collected by national statistical offices,
national health insurance and social insurance institutions, national banks, ministries
of finances, ministries of labor and social policy, and ministries of health (Table 2.)
Due to rapidly changing economic environment in the countries of political and
economic transition, such as CEE countries, historical data used in the model is not
always the data valid for the base year, but rather the average data from the longer
period. By the same token, in some cases long-term historical data was not available
or not fully reliable as the background for projections.
17

18
Table 2. Sources of data for projections
Country National statistical offices Social insurance Health insurance Governmental agencies Independent organizations National banks
Bulgaria
NSI
(Nazionalen statisticheski
institut – National Statistical
Office): demographic data
– population figures, including
population by gender,
population age structure,
total fertility rate, mortality
rate, birth rate, rate of natural
increase, life expectancy
at birth, etc.; macroeconomic
data – GDP volume and growth,
CPI (inflation), GDP deflator,
income of population; labor
statistics taken from National
Survey on Labor Force,
organized periodically.
NOI
(Nazionalen osiguritelen
institut – National
Insurance Institute): data
on the socio-insurance
system, number
and structure of insured,
insurance payments.
NZOK (Nazionalna
zdravno-osiguritelna kasa
– National Health
Insurance Fund (NHIF):
health revenues
and expenditures,
financial balance
of the health sector.
MZ
(Ministerstvo na zdraveopazvaneto
– Ministry of Health Care) NCZI
(Nazionalen centur po zdravna
informazia pri Ministerstvoto
na zdraveopazvaneto – National
Center for Health Information
at the Ministry of Health): papers
on health care policy and strategy.
MF
(Ministerstvo na finansite
– Ministry of Finance):
state budget data
AZ
(Agenzia po zaetostta pri
Ministerstvoto na truda I sozialnata
politika – Employment Agency
at the Ministry of Labor and Social
Policy): economically active
population, employed persons
both total number and by gender,
employment rate, unemployment
rate.
II na BAN
Ikonomichwski institut
na Bulgarska academia na
naukite (Institute
of Economics at the Bulgarian
Academy of Sciences
– IE-BAS): GDP, labour
productivity projections.
CIN (Centur za izsledvane
na naselenieto – Centre
for Population Studies
at the Bulgarian Academy
of Sciences): ageing
population projections.
BNB
(Bulgarska
nazionalna
banka – Bulgarian
National Bank):
macroeconomic
data – interest rates.
Estonia
Statistikaamet – SA (Statistics
Estonia): population,
employment, economic data,
national population prognosis.
NA Eesti Haigekassa – EHK
(Estonian Health
Insurance Fund): number
and structure of insured,
health insurance
expenditures and revenues,
utilisation of hospital
and outpatient care by age.
Sotsiaalministeerium – SoM
(Ministry of Social Affairs
of Estonia): total utilisation
of health care, expenditures
of health care providers, national
health accounts, consultation
on health expenditure prognosis
Rahandusministeerium – RM
(Ministry of Finance of Estonia):
official financial prognosis,
consultation on health
expenditure prognosis
UN (United Nations): UN
population prognosis data.
Poliitikauuringute Keskus
PRAXIS (Center for Policy
Studies): macroeconomic
assumptions used in various
projections. Tartu Ülikool
(TÜ) University of Tartu:
macroeconomic projection
NA
Hungary
KSH (Hungarian Central
Statistical Office)
KSH NKI (Population Research
Institute): target values
for TFR and life expectancy,
forecast on migration.
NA OEP (National Health
Insurance Fund):
aggregate data on health
care utilisation,
expenditures and revenues;
micro data on utilisation
and health expenditures
by age.
Pénzügyminisztérium (Ministry
of Finance): target values for
macroeconomic and labour market
variables from the Convergence
ESKI (National Institute for
Strategic Health Research,
Ministry of Health): National
Health Accounts 2001, 2002.
Kopint-TÁRKI: forecast
on real wage growth.
MNB (Hungarian
National Bank):
interest rates.

19
Table 2. continued. Sources of data for projections
Country National statistical offices Social insurance Health insurance Governmental agencies Independent organizations National banks
Poland
GUS (Główny Urząd
Statystyczny – Central
Statistical Office): population
data, labor market data,
economic data for the base
year, health expenditures
(National Health Account)
ZUS (Zakład Ubezpieczeń
Społecznych – Social
Insurance Institution),
KRUS (Kasa Rolniczego
Ubezpieczenia Społecznego
– Agricultural Social
Insurance Fund): number
of insured, farmer
insurance data.
NFZ (Narodowy Fundusz
Zdrowia – National Health
Fund): health expenditures
by age groups,
for survivors and deceased
(base year), insurance
revenues in the base year.
MZ (Ministerstwo Zdrowia
– Ministry of Health): budget
health expenditures;
MF (Ministerstwo Finansów
– Ministry of Finance), MPiPS
(Ministerstwo Pracy i Polityki
Społecznej – Ministry of Labor
and Social Policy):
macroeconomic projection.
CASE (Center for Social
and Economic Research):
macroeconomic projection
consultation.
Source: Gabos, Gal (2007), Golinowska, Kocot, Sowa (2007), Kvetan, Palenik, Mlynek, Radvansky (2007), Rangelova, Sariiski (2007), Roovali (2007)
NA
Slovakia
SU SR (Štatistický úrad
Slovenskej republiky) Statistical
Office of Slovak Republic
Data used: national accounts,
prices, labor market.
SOCIÁLNA POIS OVÒA
Social Insurance
Data used: number
of insured, population
shares in the groups
of insured.
NCZI (Národné centrum
zdravotníckych informácií)
National Centre for Health
Information
Data used: health data,
health care utilisation.
MZ SR (Ministerstvo
zdravotníctva Slovenskej
Republiky) Ministry of Heatlh
of the Slovak Republic
Data used: contribution rates,
data regarding health care reform.
VDC (Výskumné demografické
centrum) Demographic
Research Centre
Data used: demographic
trends forecast (assumptions
for life expectancy,
migration and fertility rates);
EU SAV (Ekonomický ústav
Slovenskej Akadémie vied)
Institute of Economic
Research, Slovak Academy
of Sciences
Data used: forecast
of economic trends – GDP,
labor market, prices.
NBS (Národná
banka Slovenska)
National Bank
of Slovakia
Data used:
exchange
rates, prices (CPI.)

The assumptions on future development of health care sector revenues and
expenditures which lie at the background of long-term projections are positive. This
implies that in almost every case the trend for future development is positive. At first
glance, in the case of CEE countries which went through rapid changes during
1990’s, it would seem rather risky to assume only positive trends. Over the 17 years
of transition, several indicators which are important for the model did not have a
positive trend: TFR was declining, employment was declining, and GDP growth
strongly fluctuated. Therefore, it is necessary to keep it in mind that projections
assume recovery from the crisis and negative trends of the 80’s and 90’s.
Assumption of positive economic growth is an extrapolation of ambitions of the
New Member States in convergence to the EU-15 economic level. Similar
assumptions are adopted in the National Action Plans in the framework of OMC and
National Convergence Programs in the framework of the accession process.
According to these documents, economic growth in the NMS should guarantee
growth in employment and productivity at the same time. Previous experience shows
that this objective is difficult to reach – Poland, for example, experienced (still does?)
the situation of “jobless growth” (Kwiatkowski 2002).
4.1. Development of demographic variables
Assumptions on future demographic changes in the countries under analysis are
based on country sources, which include analytical research and projections prepared
by national statistical offices and institutes of demography. These are confronted with
the Eurostat projections (EUROPOP 2004), especially with aggregated data on the
NMS and the EU15 country comparisons.
20
Chapter 4.
Assumptions and development
of variables

According to demographic projections, in the oncoming decades countries
included in the analysis will face the same trends as those observed earlier in the EU-
15: fertility decline, life expectancy increase, and increase in the share of the elderly
(65+) in the population. The dynamics of those changes, however, will be higher than
in the EU-15. Furthermore, one must remember that they will take place in relatively
poorer countries9.
The scale of population decrease varies from 2% in Slovakia to about 20% in
Bulgaria, during the next 25 years.
Graph 2 shows that, in the long run, the fastest decrease of population will be
observed in Bulgaria. Negative trend will slow down after 2030. In Poland and Slovakia
in the period of 2005 – 2010, a slight increase of the population is foreseen, while in the
following 5 years (2010 – 2015) the population will start to decrease, but replacement
will still be positive. After 2015, the population of these countries is going to shrink.
Graph 2. Size of population changes (5-year period)
1,0%
0,5%
0,0%
-0,5%
-1,0%
-1,5%
-2,0%
-2,5%
-3,0%
-3,5%
-4,0%
2005 2010 2015 2020 2025 2030 2035 2040 2045 2050
Bulgaria Hungary Poland Slovakia Estonia
Source: Gabos, Gal (2007), Golinowska, Kocot, Sowa (2007), Kvetan, Palenik, Mlynek, Radvansky (2007),
Rangelova, Sariiski (2007), Roovali (2007).
According to country-based projections and the Eurostat, the decrease in
population size is significant. On average, NMS population will decrease by 12%, while
in EU-15 the population will increase by 1%. These changes are also influenced by
migration processes. Eurostat projection assumes immigrants to the EU-15, also from
the NMS, though the scale of immigration from the NMS will be lower than in the base
21
9 International analyses, i.e. Eberstatd 2005, World Bank 2007, indicate that CEE countries face unusual
phenomenon of co-existence of demographic transformation and lower development level.

projection year. The NMS face emigration related to EU accession. Emigration
increase leading to negative migration balance can be observed especially in Poland. It
is assumed that the dynamics of migrations will slow down after 2010, and positive
migration balance (more immigrants than emigrants) is projected after 2025.
National demographic projection for Hungary assumes steady increase of
immigrants (about 1 200 individuals, annually), and positive migration balance is
achieved faster.
Graph 3. Changes in population size, 2004-2050
50
40
30
20
Latvia %
10
0
-10
-20
Luxemberg
Irland
Cypres
Malta
Sweden
France
Netherlands
United Kingdom
Belgium
Denmark
Spain
Austria
In the EU-15, average (1%) increase of the population is a result of population changes
in France, Great Britain, the Netherlands, Sweden and Ireland. Likewise, the smallest EU
countries experience significant increase in the population size (see: Graph 3.)
Total fertility rate
During the last 15 years one could observe a decrease in fertility in analyzed
countries. It results from modernization in life style (the so-called “westernization”),
accompanied by changes in family structure (marriages in older ages, more common
cohabitation, birth in older ages, one-child families.) The cost related to raising a child
has increased dramatically during transition period. The relations between goods for
children and other goods and consumption goals have changed, accessibility of childcare
has decreased, and family policy has changed from universal to targeted benefits.
22
Source: EPC and European Commission 2006.
EU15
Finland
EU25
Greece
Portugal
Slovenia
Germany
Italy
Hungary
Poland
Slovakia
EU10
Czech Republic
Lithuania
Estonia

Despite negative tendencies in the last 15-20 years, in all countries under analysis
the assumption of fertility increase is adopted (Table 3.) In the next 25 years, the scale
of foreseen TFR improvement is smaller than in the EU-15, while in the following
period the value of fertility indicator is closer to the projected EU-15 level. This is due
to increasing similarities in family formation process, i.e. the age of giving birth to the
first child (30.)
Notwithstanding the positive trend in fertility development projection, fertility
level is not sufficiently high to offset systematic decrease in the size of the population
in analyzed countries10.
Table 3. TFR development
Country
2002
(2003) 2005 2010 2015 2020 2025 2030 2035 2040 2045 2050
Bulgaria 1.23 1.27 1,38 1.49 1.59 1.67 1.75 1.81 1.85 1.88 1.89
Estonia 1.37 1.39 1.44 1.50 1.55 1.60 1.65 1.70 1.76 1.81 1.85
AWG data
for Estonia 1.4 1.45 1,5 1.54 1,60 1.60 1.60 1.60 1.60 1.60
Hungary 1.31 1.35 1.41 1.47 1.54 1.60 1.60 1.60 1.60 1.60 1.60
AWG data
for Hungary 1,3 1.33 1.4 1.51 1.58 1.59 1.60 1.60 1.60 1.60
Poland 1.22 1.26 1.35 1.43 1.51 1.58 1.64 1.69 1.73 1.75 1.76
AWG data
1.2 1.19 1.3 1.42 1.5 1.58 1.60 1.60 1.60 1.60
for Poland
Slovakia 1.29 1.25 1.19 1.22 1.29 1.36 1.43 1.5 1.57 1.64 1.71
AWG data
for Slovakia 1.2 1.18 1,2 1.33 1.4 1.52 1.60 1.59 1.60 1.60
NMS (AWG data) 1.23 1.24 1.3 1.44 1.5 1.56 1.6 1.58 1.6 1.58
EU 15 (AWG data) 1.53 1.57 1.60 1.60 1.60 1.60 1.60 1.60 1.60 1.61
Rangelova, Sariiski (2007), Roovali (2007), EPC and European Commission 2006
Life expectancy
The assumption of systematic growth in life expectancy is based on historical
tendencies observed in EU-15 countries, and positive impact of factors that determine
longevity. The latter, on the one hand, include changes in dietary habits (more fruit
and vegetables, higher vegetable fat consumption), decrease in smoking, changes in
the structure of consumed alcohols, intensification of physical activity, which all lead
to decreasing mortality due to cardiovascular diseases, the main reason of high
mortality in the CEEC (Zatoński, Willert 2005). On the other hand, access to
pharmaceuticals and new technologies has increased, which allows for faster and
more effective reaction when the disease occurs (Drygas 2005).
23
10 Simple replacement level is assured when TFR equals at least 2.1.

Back in 1980’s, CEE countries experienced stagnation or even slight decrease in
life expectancy, especially for labor market active male population (Okólski 2004,
Golinowska, Sowa, Topór-Mądry 2006). This unfavorable trend is reversed slowly, but
24
Graph 4. Life expectancy – males
90,0
85,0
80,0
75,0
70,0
65,0
60,0
2002 (03) 2005 2010 2015 2020 2025 2030 2035 2040 2045 2050
2002 (03) 2005 2010 2015 2020 2025 2030 2035 2040 2045 2050
Rangelova, Sariiski (2007), Roovali (2007)
Graph 5. Life expectancy – females
90,0
85,0
80,0
75,0
70,0
65,0
60,0

its improvement is observed in Hungary and in Poland, while in Slovakia male life
expectancy has not improved yet. In Estonia, the improvement is still unsatisfactory.
Changes in population structure – ageing
Changes in demographic structure according to age are clearly reflected in the shape
of age pyramid, which even at the beginning of projection period looks like a Christmas
tree, and at the end of projection is shaped as a mushroom with a narrow mushroom leg.
The graphs presented in Annex 1 describe age structure of the population in the
base year and at the end of projection period. In the base year in Hungary, Poland and
Slovakia two periods of population growth can be distinguished: the first one is the
baby boom following World War II, while the other one is its echo after 20-25 years.
The highest share of population is observed for the cohort of 55-75 years of age. In
Poland, the peak is at the age of 67-68, in Bulgaria about 60-70, in Estonia 60-65,
while in Hungary and Slovakia the peak is at the age of 70-75 and 68-72, respectively.
A decrease in the share of children in the population of analyzed countries can be
observed until 2010. Following that period, this decreasing trend remains at the level
of about 15%. The share of children slowly decreases again after the year 2020.
The lowest shares of young population are observed in Bulgaria in the first
projection period (up to 2025), and in Hungary in the second period (up to 2050.)
25
Graph 6. Share of population 0-14
30,0%
25,0%
20,0%
15,0%
10,0%
5,0%
0,0%
2002 (03) 2005 2010 2015 2020 2025 2030 2035 2040 2045 2050

A decrease in the share of labor market active age population can be observed after
2010, when baby boomers begin to retire (at the age of 65.) In the NMS the drop in
the share of active age population is rapid (about 27% compared to 13% in the EU-
15), and it is projected to continue dynamically up to 2025. In years 2025 – 2030 the
decrease will slow down, but in the last decade of projection the share of labor market
active population will rapidly shrink again (Graph 7).
Labor market activity development is similar in Eastern and Western Europe,
although the dynamics is higher in the NMS. Decrease in labor market activity rate is
the main reason for policies of employment promotion and new labor market
programs at the EU-level. It also leads to changes in immigration policy in the
enlarged EU. In the future, the potential of Eastern Europe as a migration source for
Western Europe countries will diminish.
Graph 7. Share of population 15-64
75,0%
70,0%
65,0%
60,0%
55,0%
50,0%
45,0%
40,0%
2002
(03)
2005 2010 2015 2020 2025 2030 2035 2040 2045 2050
An increase in the share of the elderly has already been described by age
pyramids at the beginning and at the end of projection period. Still, it is worth
noting that the dynamics of increase in the share of the elderly (from 15% to almost
30% at the end of projection period) will be higher in the period of 2015-2025, when
baby boomers get older.
As a result of presented demographic scenario, the old age dependency ratio, which
is calculated as a proportion of the number of the elderly (65+) to the number of labor
26

market active age population, doubles – from 20% to around 40%. In Hungary, Poland
and Bulgaria the indicator dramatically increases to over 50% (Graph 9.)
27
Graph 8. Share of population 65+
50,0%
45,0%
40,0%
35,0%
30,0%
25,0%
20,0%
15,0%
10,0%
5,0%
0,0%
2002
(03)
2005 2010 2015 2020 2025 2030 2035 2040 2045 2050
Graph 9. Old age dependency ratio
60
50
40
30
20
10
0
2002
(03)
2005 2010 2015 2020 2025 2030 2035 2040 2045 2050

Employment rate is one of the indicators that deteriorated during transition. In
Bulgaria and Poland, only half of the population in labor market active age is
currently employed. Estonia has the highest employment rate among all countries
under research; moreover, it had the highest rate of economic growth in recent years.
*Eurostat data prepared for the purposes of Ageing Working Group (EPC and European Com mission 2006)
Rangelova, Sariiski (2007), Roovali (2007), EPC and European Com mission 2006
Assumptions regarding the development of employment rate in all the countries
are quite optimistic, but they are coherent with the national action plans on
employment and convergence strategies adopted by selected New Member States
(NMS) after EU accession. Projected level of employment rate is necessary in the long
run to keep the economy growing (even if productivity growth is high) in view of
unfavorable demographic changes.
When comparing these assumptions to the Economic Policy Committee (EPC) and
European Commission projection, one can see some discrepancies, especially in
phase one of projection period (until 2025.) Presented projections for Poland,
Hungary and Slovakia assume higher employment rate than the EPC projection. In
general, labor market indicators projected by the EPC and European Commission are
less favorable for nearly all NMS in comparison to the EU-15. Most likely, this is a
result of demographic situation in the NMS and its impact on labor market activity
level. Population ageing dynamics in the NMS is higher than in the EU-15, so the size
of labor market active cohort is smaller. Significant improvements in the economic
situation in the NMS in recent years seem to indicate that there are grounds for more
optimistic employment assumptions than those included in EPC prognosis. Following
Lisbon Strategy assumptions, the NMS have released some employment improvement
programs which could bring about improvements in labor market situation.
Employment rate assumptions in national and European projections in the second
part of projection period (2025-2050) are similar. One should emphasize that the
28
Table 4. Employment rate development
2005 2010 2015 2020 2025 2030 2035 2040 2045 2050
Bulgaria 51.4 54.8 64.0 64.8 63.8 62.1 60.1 58.9 58.9 59.3
Estonia 63.8 69.4 71.6 73.0 73.4 73.6 73.5 73.7 73.5 73.4
AWG* data for Estonia 64.4 68.4 71.8 72.3 71.9 71.6 71.1 71.2 71.1 70.8
Hungary 56.4 58.7 60.8 64.3 64.0 62.6 61.9 62.0 63.4 63.2
AWG data for Hungary 58.4 60.8 62.4 64.3 65.3 64.6 63.5 62.4 63.1 63.2
Poland 51.2 55.3 59.4 62.9 66.5 70.0 71.0 72.0 73.0 74.0
AWG data for Poland 52.7 57.0 61.0 64.9 68.4 68.6 67.4 66.2 65.6 66.1
Slovakia 57.7 62.0 65.2 68.6 69.4 69.3 68.3 68.0 68.4 68.9
AWG data for Slovakia 59.2 62.1 66.7 70.2 72.7 72.6 71.2 69.5 69.0 68.7
NMS (AWG data) 57.2 60.7 64.2 67.2 69.4 69.2 68.1 67.0 66.8 67.1
EU 15 (AWG data) 65.6 68.1 70.1 70.5 70.5 70.7 71.1 71.5 71.5 71.5

projections do not assume that the NMS will have reached the goals of the Lisbon
Strategy, and national projections are even more skeptical. In Hungary and Bulgaria,
estimated employment level is lower than the average for the NMS in 2050. It is assumed
that by 2020-2025 employment level will decrease by 20-25%, compared to 2003.
Table 5. Unemployment rate development
2005 2010 2015 2020 2025 2030 2035 2040 2045 2050
Bulgaria 13.2 8.4 6.0 5.6 5.2 4.9 4.7 4.5 4.3 4.1
Estonia 9.4 5.5 5.7 5.0 5.3 5.3 5.3 5.3 5.3 5.3
AWG data for Estonia 9.1 7.8 7.0 7.0 7.0 7.0 7.0 7.0 7.0 7.0
Hungary 8.0 7.2 6.4 4.8 6.2 6.6 5.6 4.2 3.9 4.8
Poland 18.2 14.3 11.4 9.9 8.5 7.0 7.0 7.0 7.0 7.0
Slovakia 17.9 11.7 8.9 6.0 6.0 6.0 6.0 6.0 6.0 6.0
NMS (AWG data) 13.8 12.0 10.0 8.3 6.6 6.6 6.6 6.6 6.6 6.6
EU 15 (AWG data) 7.7 7.0 6.1 6.1 6.1 6.1 6.0 6.0 6.0 6.0
European Commission projections on the level of unemployment assume that the
indicator will decrease up to the NAIRU level of 6-7%. In the case of Poland and
Slovakia, where unemployment rate is the highest, it will take even 20 years to
approach this estimated level (EPC and European Commission 2006.)
Background for projections on unemployment level prepared by each country is
similar to the assumptions made by the European Commission, although the indicators
regarding the first period of projection are more optimistic. Only in Hungary is the
slope of unemployment rate decrease less steep (overall decrease of unemployment is
at the level of 40%), due to the already low unemployment rate in the base year (about
8.0%.) In other countries unemployment rate decreases by 60-70% during projection
period, with the lowest unemployment rate estimated in 2050 in Bulgaria.
4.2. Economic assumptions and projections
Projection of economic growth in analyzed countries is built upon convergence
assumptions, which implies that these countries will catch up with the average
European economic level relatively fast. It is assumed that labor productivity in the
NMS will be substantially higher than in the EU-15, particularly with lower
employment rate.
29

In consequence of convergence, GDP growth is expected to be nearly two times
higher than in the EU-15 in the first period of projection. National projections are
even more optimistic for the next 20 years. Nearly all the countries (except for
Hungary in 2025) assume higher GDP growth than that included in European
Commission projection.
In phase two of projection period, economic growth dynamics in the EU-15 and in
the NMS is assumed to be similar, and by 2050 economic growth in some of EU-15
countries will be higher than in the NMS. National projections assume lower decrease
in GDP growth in the second part of projection period, especially in Bulgaria. This is
a result of high labour productivity (especially productivity capital deepening)
accompanied with adverse tendency in demography and labour utilization.
4.3. Assumption regarding income elasticity of health expenditures
Presented projections are based upon an assumption that income elasticity of
health expenditures is equal to 1.0. This supposition is arbitrary, drawn upon partial
analysis of the relation between income and health expenditures (MZ 2004). It should
be noted that there are numerous discussions on the value of income elasticity
indicator. The outcome of one of the first analyses dedicated to that subject were
published in 1977 (Newhouse 1977), and indicated that income elasticity of health
expenditures is above 1, which implies that health expenditures are, in fact, luxury
expenditures. Over the next 30 years various analyses were published, but their
outcome was inconclusive. According to Getzen (2000), different results of those
analyses should be attributed to the differences in the definitions of the subject under
30
Table 6. GDP growth rate
2005 2010 2015 2020 2025 2030 2035 2040 2045 2050
Bulgaria 4.4 4.3 4.2 4.0 3.9 3.7 3.6 3.5 3.4 3.2
Estonia 10.5 7.2 4.3 4.1 4.1 3.9 3.8 3.6 3.4 3.3
AWG data for Estonia 6.3 5.6 3.7 2.7 2.4 2.3 1.5 1.3 0.9 0.6
Hungary 4.1 3.6 3.0 2.5 2.3 2.0 1.8 1.6 1.3 1.1
Poland 5.3 5.1 4.8 4.2 3.7 3.2 2.6 2.1 1.5 1.0
Slovakia 6.0 6.2 5.1 4.1 3.2 3.1 2.9 2.8 2.6 2.5
NMS (AWG data) 4.4 4.6 3.5 2.9 2.7 2.1 1.1 0.7 0.6 0.6
EU 15 (AWG data) 2.0 2.5 2.2 1.7 1.5 1.2 1.2 1.3 1.3 1.3

analysis. Income elasticity varies depending on the subject under analysis: individual,
group (i.e. insurance company clients), or the whole population covered by national
health insurance. Income elasticity measured at the individual level is typically lower
than 1 (health is understood as a basic, necessary asset), while income elasticity
measure at the national level is often higher than 1 (which implies that health is a
luxury good) (Getzen 2000.) Table below presents some of the results of income
elasticity analysis at the macro (national) level.
Table 7. Selected results of the analysis of impact of income changes on the level of health
expenditures
Source
Newhouse (1977) 1.30 13 developed countries
Gertler i van der Gaag (1990) 1.30 25 countries
Getzen (1990) 1.60 USA
Getzen i Poullier (1992) 1.40 19 OECD countries
Murthy i Ukpolo (1994) 0.77 USA
Gerdtham et al. (1998) 0.66-0.82 22 OECD countries
Fogel (1999) 1.60 USA
Okunade i Murthy (2002) 1.56-1.64 USA
Ariste i Carr (2003) 0.88 Canada
Clemente et al. (2004) 1.47-3.65 22 OECD countries: results depend on the country
Okunade et al. (2004) 0.87 17 OECD countries
Dreger i Reimers (2005) 0.68-0.84
Source: Mot (2006).
Income
elasticity Comments
4.4. Assumption on health status and health service utilization
Medical service utilization curve is the basic element of health expenditures
projection. It presents diversification of medical services utilization pattern by age,
what implies changes in morbidity in the population age cohorts. Utilization curve has
a well-known and broadly described J-curve shape; it varies, however, from country
to country. Lately (ECP and European Commission 2006) it was shown that the
spending peak occurs earlier in life cycle in the NMS that in the EU-15: at the ages of
75-80 and 85-90, respectively. Te main reason of such diversification is the difference
31
21 OECD countries: results depend on medical
development indicator (LE, infant mortality,
share of 65+ population)
OECD (2006) 0.87-1.5
30 OECD countries: results depend on the method
applied in the analysis
0.79-0.97
EU14+Iceland+Norway+Switzerland:
results depend on the model adopted
0.81 EU11+Iceland+Norway+USA+Canada
CPB (2006)*
0.69 OECD countries, excl. Turkey

in life expectancy (LE). At the same time, the shape of J-curve is flatter for some of the
NMS (especially Baltic states and Poland) than for other European countries. This
implies that medical service utilization and health expenditures incurred by the
elderly are not as high as in the other countries. As some Polish researches ascertain,
lower costs and utilization levels for the elderly are related to social and economic
structure of the society, as well as some cultural factors. The milder slope of the J-curve
can be explained, most likely, by a high share of rural population that tends to
use medical services less commonly.
A significant impact of hospital expenditures on the shape of J-curve is implied by
Hungarian data on health expenditures by age (research coordinated by the National
Health Insurance Fund – random survey sample of 5% of the population), and by
Polish health expenditure data (the National Health Insurance.) The share of hospital
expenditures in total health expenditures is relatively high (closer to 50% than to
40%.) Graphs below present this relation in case of Estonia and Hungary.
Graph 10. Medical services utilization by age group in Estonia
18
16
14
12
10
8
6
4
2
0,40
0,35
0,30
0,25
0,20
0,15
0,10
0,05
The following graph presents the assumptions on the shape of the J-curve in
Poland. Its shape varies significantly from the shape typical of EU-15 countries.
Namely, higher expenditures in Poland start for younger cohorts than in the old EU
countries; respectively: above 40 and above 50 years of age. However, the peak of
expenditures occurs in similar age: about 75-80 for men and 90 for women, but the
level of spending in Poland can be even 7 times lower.
Per capita health expenditures in Hungary also have a typical J-curve shape.
Analogically to Poland, a rapid increase in per capita health expenditures can be
observed for younger cohorts than is the case in the EU-15. The peak of expenditures
32
Source: Roovali (2007)
Utilisation by Age Groups
0
0-4
5-9
10-14
15-19
20-24
25-29
30-34
35-39
40-44
45-49
50-54
55-59
60-64
65-69
70-74
75-79
80-84
85-89
90-94
95-99
Outpatient Visits per Capita per Year
0,00
Hospital Admissions per Year
Outpatient - visits per year Inpatient - admissions per year

Graph 11. Medical services utilization by age group in Hungary
25
20
15
10
5
Source: Gabos, Gal (2007)
0,9
0,8
0,7
0,6
0,5
0,4
0,3
0,2
0,1
occurs at the age of 70-79, irrespective of gender. The same male age cohort faced the
peak of health expenditures in Poland, while for females the peak of health
expenditures is observed for a younger cohort.
33
Graph 12. Average insurance expenditures per capita in Poland as percentage of GDP per
capita (base year)
12,00%
10,00%
8,00%
6,00%
4,00%
2,00%
0,00%
0 - 4
5 - 9
10 - 14
15 - 19
Utilisation by Age Groups
0
0-4 5-9 10-14 15-19 20-24 25-29 30-34 35-39 40-44 45-49 50-54 55-59 60-64 65-69 70-74 75-79 80-84 85-89 90-94 95-99
Outpatient Visits per Capita per Year
0,0
Hospital Admissions per Year
Outpatient - visits per year Inpatient - admissions per year
20 - 24
25 - 29
30 - 34
35 - 39
40 - 44
45 - 49
50 - 54
55 - 59
60 - 64
65 - 69
70 - 74
75 - 79
80 - 84
85 - 89
90 - 94
95 - 99
males females

34
Graph 13. Average health cost per capita in Hungary as percentage of GDP per capita
(base year)
0-9 10-19 20-29 30-39 40-49 50-59 60-69 70-79 80-89 90+
2,5%
2,0%
1,5%
1,0%
0,5%
0,0%
Males Females

Chapter 5.
Projection results
Projections pertaining to public funding in health care sector (revenues,
expenditures and financial balance) based on the ILO social budget model were
prepared according to several scenarios. The first one, identical for all the countries
under research, is the (1) baseline scenario. It is based on some assumptions,
established for all the countries:
• the base year is 2003,
• life expectancy improvement is medium, based on national demographic
projections, taking gender and age-specific mortality rate into account,
• total fertility rate development is the medium variant of national demographic
prognosis, with the assumption of no changes / very minor changes in the
second part of projection (2025-2050),
• yearly increase in health expenditures per capita has the same rate as GDP
per capita (Scenario II in ILO model),
• economic and labor market variables are assumed to be the most probable
version of national projections, for Hungary, Poland and Slovakia based on
updated country projections used in Governmental Convergence Programs
prepared for the European Committee last year (2006), and for Bulgaria
based on national Bulgarian considerations from different institutions.
Additional scenarios are not identical across countries. (2) for Poland and
Hungary the death-related costs scenario was prepared, where expenditures in the
last year of life are taken into account. This scenario was possible thanks to
disaggregation of health insurance expenditure data. (3) For Bulgaria, Hungary,
Poland and Slovakia different prognoses contingent upon future LE developments
were prepared. (4) Additionally, for Bulgaria, Hungary and Slovakia variants with
different assumptions concerning wage development were analyzed. (5) Lastly,
different development of labor market indicators: activity rate (Hungary),
employment rate (Poland and Slovakia), and unemployment rate (Bulgaria), was
taken into account. (6) Variants with some combinations of different variable
35

development assumption were also prepared, for instance, wages and employment
rate (Slovakia), and different wage development (Bulgaria, Hungary and Slovakia.)
[RTF bookmark start: _Toc169873599]
5.1. Baseline scenario
Revenue side
Public health care sector projected revenues depend on adopted scenarios on the
development of demographic and economic variables, including employment,
unemployment and economic growth. Institutional framework in the health care
sector and insurance premium are assumed to be constant during the whole
projection period, according to the ceteris paribus principle.
The picture of health care sector revenues in relation to GDP is strongly
differentiated. In projection phase one we can observe the trend of increasing
revenues, although the trend is very strong only in Bulgaria, in Poland it is weaker, in
Hungary the trend is non-linear (decreasing in the first 5 years), and in Slovakia it
begins to decrease as early as 2010. In projection phase two, that increasing trend is
reversed and we observe a drop in the level of health care revenues in relation to GDP.
Only in Poland do the revenues grow almost continuously.
Table 8. Projected health revenues (public) as a share of GDP
2005 2010 2015 2020 2025 2030 2035 2040 2045 2050
Bulgaria 4.6 5.6 6.2 6.5 7.0 7.0 6.9 6.7 6.4 6.1
Estonia 4.2 4.2 4.2 4.1 4.1 4.0 3.8 3.7 3.5 3.3
Hungary 4.0 3.9 4.1 4.3 4.1 3.9 3.8 3.6 3.5 3.4
Poland 4.5 4.7 4.8 5.0 5.2 5.2 5.3 5.4 5.6 5.8
Slovakia 6.3 6.4 6.0 5.8 5.6 5.4 5.2 5.0 4.7 4.5
Expenditure side
At the beginning of projection period, countries under research are characterized
by a low share of public expenditures on health care sector in relation to GDP. Only
in Hungary the share of expenditures exceeds 5% of GDP. In the first projection
period, relative increase in expenditures is slow in each country except Bulgaria,
where the level of 6% of GDP is reached as early as in 202011.
36

Graph 14. Public health care revenue growth in relation to the year 2005 (2005 = 100%)
1800,0%
1600,0%
1400,0%
1200,0%
1000,0%
800,0%
600,0%
400,0%
200,0%
0,0%
2005 2010 2015 2020 2025 2030 2035 2040 2045 2050
Table 9. Projected public expenditures on health as a share of GDP
2005 2010 2015 2020 2025 2030 2035 2040 2045 2050
Bulgaria 5.0 5.4 5.4 6.1 6.4 6.6 6.8 7.0 7.1 7.1
Estonia 4.0 4.3 4.5 4.7 5.0 5.3 5.6 6.0 6.3 6.7
Hungary 5.5 5.5 5.6 5.8 5.9 6.1 6.2 6.4 6.6 6.7
Poland 4.6 4.9 5.2 5.5 5.9 6.2 6.6 6.8 7.0 7.1
Slovakia 4.8 5.0 5.1 5.4 5.6 5.8 6.1 6.4 6.6 6.9
In the second projection period, public expenditures on health systematically
increase in relative terms, up to the level of 7% of GDP. The lowest dynamics is
projected in Slovakia and Hungary.
The highest share of public expenditures on health in relation to GDP in the year
2050 is projected in Bulgaria and Poland. In Poland, the growth rate of public
expenditures on health is the highest during the whole projection period. Level of
expenditures, the lowest as compared to other countries in the base year, rapidly
increases to the level of 7.1% of GDP in 2050.
37
11 In many CEE countries the level of public expenditures on health equal to 6% of GDP is perceived as a
standard of appropriate financing in the sector, in relation to other expenditure objectives of the state. In
Poland, medical associations are lobbying for priority for health care sector and stipulate for fast increase of
expenditures to the level of 6% of GDP.

Graph 15. Public health expenditures growth in relation to the year 2005 (2005 = 100%)
4000,0%
3500,0%
3000,0%
2500,0%
2000,0%
1500,0%
1000,0%
500,0%
0,0%
2005 2010 2015 2020 2025 2030 2035 2040 2045 2050
Balance: surplus/deficit
Public funding surplus / deficit indicators in health care sector in relation to GDP
illustrate the scale of financial balance / imbalance and reflect financial tensions in the
sector. In the base year, health care sector is in balance only in Slovakia, in
consequence of reforms introduced several years ago12. The reforms included the
introduction of co-payment for services provided in the public sector.
Table 10. Deficit / surplus as a share of GDP
2005 2010 2015 2020 2025 2030 2035 2040 2045 2050
Bulgaria -0.5 0.1 0.4 0.4 0.6 0.4 0.0 -0.3 -0.7 -1.0
Estonia 0,2 0,0 -0,3 -0,6 -0,9 -1,3 -1,8 -2,3 -2,8 -3,4
Hungary -1.5 -1.6 -1.5 -1.5 -1.8 -2.2 -2.5 -2.8 -3.1 -3.4
Poland -0.2 -0.3 -0.4 -0.6 -0.7 -1.0 -1.3 -1.4 -1.4 -1.3
Slovakia 1.5 1.4 0.8 0.4 0.0 -0.4 -0.9 -1.4 -1.9 -2.3
In Bulgaria, despite the presence of financial deficit in the base year, financial
balance is projected in the first period and over several subsequent years. Bulgarian
health care system experiences financial deficit after 2035, but until 2050 the deficit
does not exceed 1% of GDP, staying well above the level of deficit in other countries.
38
12 It should be noted that the current government withdraws some of the reforms introduced earlier.

Poland faces the deficit starting from the base year; but further increase of the deficit
is milder than in other countries, and in final years of projection the deficit even
slightly decreases, reaching the level close to the level of the deficit in Bulgaria (1.3%
of GDP.) The least favorable situation is expected in Hungary, where the deficit is the
highest during the whole projection period. After the year 2020 the deficit increases
with the highest slope, reaching the level of 3.4% GDP in 2050.
5.2. Scenario of increased expenditures in the last year of life
Death-related costs – theoretical background
From the beginning of 1990’s, numerous analyses were conducted to examine the
relation between per capita expenditures on health and the time left until death.
Research findings point at two main conclusions:
• Majority of per capita expenditures on health is incurred in the last years
(months) of individual’s life (death-related costs), and the increase in per capita
expenditures on the elderly is caused not only by the impact of age, but by the
fact that the elderly face higher probability of death (Raitano 2006, Brockmann
2002, Lubitz et al. 1995.)
• The indicator that describes the relation of costs generated by an individual in
the final stage of life to costs generated by an individual in the same age group
but not in the final stage of life is very high; though it has a tendency to decrease
with age (Raitano 2006, Kildemoes et al. 2006 (expenditures on drugs), Gabriele
et al. 2005, Busse 2002, McGrail et al. 2000, Zweifel et al. 1999. Lubitz et al.
1995, Lubitz, Riley 1993.)
Research shows that per capita expenditures in the last years of life decrease with
age (Gabriele et al. 2005, Polder, Achterberg 2004, Brockmann 2002, Hoover et al.
2002, Serup-Hansen et al. 2002, Madsen 2002, McGrail et al. 2000, Lubitz 1993.) In
contrast to individual expenditures in the group of survivors, which increase with age
(except for the oldest age groups, where in some cases a decrease of expenditures is
observed) (Gabriele et al. 2005, Hoover et al. 2002, Serup-Hansen et al. 2002, McGrail
et al. 2000.) McGrail et al. argue that the expenditures on nursery and social care
increase with age, interdependently of the status of the patient and proximity to death.
Similar relation is confirmed in the research on utilization of health care services,
especially day-care in hospitals in Germany (including psychiatric care, excluding long-term
care) (Busse et al. 2002.) Among the groups included in the analysis, in the third,
second and last year before death the number of days of hospital care decreases with
39

age.
High share of aggregated health expenditures is targeted towards care over
individuals in their last year of life. According to Medicare expenditures in 1978
(Lubitz, Prihoda 1984), health expenditures on individuals who died in that year
amounted to 28% of total health expenditures in Medicare. These expenditures were
only incurred by the care over 5.9% of the population covered by the analysis. In the
following ten years, percentage of the deceased covered by Medicare was between 5.1-
5.4%, and the cost of their care was at the level of 26.9-30.6% of total health
expenditures annually (Lubitz, Riley 1993). In 1992-1996 the share of costs incurred
by individuals in their last year of life amounted to 26% of total Medicare costs
(Hoover et al. 2002), so only slight differences in costs can be observed in the course
of years. Similar findings were described by Seshamani and Gray (2004), who found
that 28% of hospital expenditures in England were generated by the deceased.
When death-related costs are included in the analysis of the impact of ageing on
health expenditures, it may lead not only to the constraint of health expenditure
growth, but even a decrease in projected health expenditures. This is a result of the
fact that expenditures in the last year of life of younger population are significantly
higher than those incurred in the case of the elderly (due to more common utilization
of intensive and costly treatment for younger population, types of diseases related to
age, and the fact that the costs of care for the elderly are shifted outside of health care
system – to long-term and home care) (Brockmann 2002, Kramer 1995.)
Assuming increasing life expectancy and decreasing mortality rates, the point of
death and higher health expenditures are shifted towards older age. Average
expenditures by age change over the years, and in every age cohort the share of
persons generating higher costs (costs related to death) decreases (Ahn et al. 2005,
Batljan 2004, Seshamani, Gray 2004.)
Lack of necessary data on average health expenditures on the deceased and
average health expenditures on survivors represents a common problem in the
estimation of increased expenditures related to death. One way to resolve that issue it
to use DRG data (Ahn et al. 2005); but in the countries where DRG payment system
has not been introduced, the problem remains unresolved.
Owing to lack of necessary data, most analyses are based on the division of
population into two groups: the deceased and the survivors in a given year (or shorter
period of time) (Batljan, Lagergren 2004 – 1 year for expenditures on ambulatory care,
Hoover et al. 2002, Serup-Hansen et al. 2002, McGrail et al. 2000 – 6 months, Lubitz,
Riley 1993). Health expenditures are generated especially over the last 2-3 months of
life (Hoover et al. 2002 – 51% in the last 3 months of life, Lubitz, Riley 1993 – up to
50% in the last two months.) Only in some research are the expenditures examined
depending on the time left until death in a period longer than 1 year (Kildemoes et al.
40

2006 – 2 years before death for pharmaceutical expenditures, Batljan, Lagergren 2004
– 8 groups of population for hospital expenditures and death in a given calendar year
(0 years until death), 6 groups depending on proximity to death (from 1 to 6 years), and
the last group – over 6 years till death), Dixon et al. 2004 – 3 groups of population – 1,
2, 3 years before death, analyzed for hospital expenditures, Busse et al. 2002.)
Indicator of proportion of per capita expenditures on a person in the last of life to per
capita expenditures on a survivor (k indicator) – comparison of diverse research.
It is difficult to compare projections prepared in different countries mainly due to
diversified baskets of benefits included in the analysis, different periods before death
41
Table 11. Selected results of the analysis of expenditures per deceased in proportion
to expenditures per survivor (k indicator.)
Analysis
prepared by: Country Type of care
Age of population
under research
Indicator of proportion of per capita
expenditures per deceased to per capita
expenditures per survivor
Kildemoes et al.
(2006) Denmark Pharmaceuticals
Age groups
0-24, 25-49,
50-74, 75+,
by gender
Age
0-24
25-49
50-74
75+
Males
12.36
7.87
4.00
1.80
Females
1.37
9.91
3.68
1.63
Gabriele et al.
(2005)
Italy
(four
regions)
Hospital care
5-year age
cohorts
For all age groups:
Lombardy: 13.8
Tuscany: 14.0
Apulia: 11.9
Abruzzo: 10.2
Ahn et al.
(2005) Spain Hospital care
5-year age
cohorts
Age
0
1-5
6-10
11-15
16-20
21-25
26-30
31-35
36-40
41-45
Indicator
7.6
71.1
82.1
92.7
96.5
75.6
48.9
40.7
43.7
43.5
Age
46-50
51-55
56-60
61-65
66-70
71-75
76-80
81-85
86+
Total
Indicator
35.0
26.9
21.7
15.8
11.9
9.4
7.4
6.3
5.0
24.1
Batljan
& Lagergren
(2004)
Sweden
(region
Skane)
Hospital and
outpatient care,
excluding
long-term care
Indicators with
no distinction
by age
5.96 for outpatient care,
21.5 for hospital care.
National Health
Insurance
gency for
Wage Earners
(2003)
France
Medical
treatment
Six age groups:
35-44, 45-54,
55-64, 65-74,
75-84, 85+
Age
35-44
45-54
55-64
65-74
75-84
85+
Indicator
6.5
8.8
3.3
2.6
2.8
1.8
Busse et al.
(2002)
Germany
Hospital
days per year
Up to 24 and
above 85 years
of life, between
24 i 85, 10-year
age groups
Age
<24
25-34
35-44
45-54
55-64
65-74
75-84
85+
Indicator
29.2
30.8
31.0
21.1
17.6
12.0
6.6
4.3

Table 11. continued Selected results of the analysis of expenditures per deceased in proportion
to expenditures per survivor (k indicator.)
Analysis
prepared by: Country Type of care
that are taken into account, and discrepancies in age groups under analysis.
Compilation presented above illustrates the differences in the results of k-indicator
analyses depending on a country, age and type of services under research.
Irrespective of the values of k-indicator shown, one can identify decreasing trend in
indicator’s value by age (except for the youngest age cohort, which relates to intensive
treatment of newborns.)
Results of public health expenditure projection, including the analysis of increased
expenditures in the last year of life – case of Hungary and Poland
Due to limited availability of adequate data, the scenario of health expenditures
adjusted for death-related costs is conducted only for Hungary and Poland.
Methodology referred to above is presented in Annex 1. Minor differences between the
two estimations include different age groups (5-year cohorts for Poland and 10-year
cohorts for Hungary), and the base year for calculations (2004 for Poland and 2002
for Hungary.) In the case of Hungary, estimations are based on total health
expenditures, while in the case of Poland they are based on insurance expenditures.
Population split into two groups - the deceased and the survivors - does not affect
projected revenues, but only impacts projected expenditures and, as a result, the size
of financial deficit.
In both countries, together with the introduction of diversified level of average per
capita expenditures depending on individual status (deceased or survivor), projected
42
Source: Own compilation
Age of population
under research
Indicator of proportion of per capita
expenditures per deceased to per capita
expenditures per survivor
Hoover et al.
(2002) USA Medicare
Three age groups
over 65 years
of life: 65-74,
75-84, 85+
Age
65-74
75-84
85+
Indicator
8.57
6.28
3.61
McGrail et al.
(2000) Canada
Medical treatment
including nursing
care (half a year
before death)
Four age groups:
65, 75-76, 85-87,
90-93
Age
65
75-76
85-87
90-93
Year
1986
1993
1986
1993
1986
1993
1986
1993
Indicator
16.6
16.7
8.4
8.6
3.4
3.8
2.5
2.5
Lubitz & Riley
(1993) USA Medicare
Over 65 years
of life
6.92

43
Table 12. Indicator showing the proportion of expenditures per deceased to expenditures per survivor
Belgium Czech Republic Denmark Spain Italy Netherlands Austria Poland EU average
Age group
M* F* M F M F M F M F M F M F M F M F
0-4 12.1 20.1 34.5 43.5 4.5 4.0 3.4 3.4 68.0 79.5 31.7 79.1 27.0 39.1 25.7 39.7 25.9 38.5
5-9 33.3 33.0 55.3 48.2 77.4 58.4 6.4 6.9 79.5 163.0 39.6 60.0 104.8 153.0 47.0 50.3 55.4 71.6
10-14 27.7 9.5 74.0 42.5 8.7 14.5 6.9 6.3 73.1 101.4 26.9 43.3 121.6 120.4 40.7 49.3 47.4 48.4
15-19 10.7 21.1 31.0 26.2 1.1 1.3 4.1 7.0 38.7 46.7 21.6 24.7 64.7 69.1 29.5 37.3 25.2 29.2
20-14 8.9 11.7 17.1 26.2 0.3 0.3 3.3 7.1 26.0 32.5 47.4 33.2 41.7 87.3 23.0 26.1 21.0 28.0
25-29 9.4 13.1 19.1 28.7 12.0 12.1 3.9 5.9 29.0 25.5 38.0 10.4 57.7 41.3 27.4 24.5 24.6 20.2
30-34 13.6 11.4 23.1 32.0 11.4 12.7 3.2 6.2 30.4 28.4 25.3 18.9 48.1 33.4 21.2 25.6 22.0 21.1
35-39 14.3 11.7 20.2 25.7 7.1 6.0 2.8 4.6 40.5 37.2 26.7 23.5 42.9 29.6 18.3 23.0 21.6 20.2
40-44 12.4 13.8 19.2 20.4 6.3 5.9 2.6 3.2 35.3 40.7 17.0 18.1 34.6 33.9 13.6 20.5 17.6 19.6
45-49 11.0 14.3 16.8 17.1 8.2 7.2 2.3 2.8 30.9 31.5 15.1 17.2 31.4 28.0 11.1 15.1 15.9 16.6
50-54 10.1 12.1 11.0 13.6 7.5 7.0 2.3 2.6 21.1 26.9 14.2 15.5 21.4 25.7 8.9 12.3 12.1 14.5
55-59 9.5 10.4 8.1 10.7 7.5 6.8 2.2 2.4 17.1 23.7 8.8 12.9 18.9 22.0 7.8 10.9 10.0 12.5
60-64 7.4 9.6 7.2 10.0 6.2 6.0 2.0 2.3 12.1 16.8 8.3 12.4 16.3 20.6 6.6 9.3 8.3 10.9
65-69 5.5 6.8 5.4 6.8 5.0 5.0 1.8 2.1 8.5 11.9 6.4 8.3 13.2 15.0 5.6 7.4 6.4 7.9
70-74 4.5 5.0 4.3 5.1 4.4 4.3 1.7 1.8 6.2 8.2 5.1 6.4 11.6 11.0 4.5 5.6 5.3 5.9
75-79 3.3 3.5 3.5 3.7 2.8 2.9 1.6 1.6 4.5 5.4 4.1 4.6 8.9 8.9 3.9 4.4 4.1 4.4
80-84 2.8 2.5 2.8 2.9 2.0 2.1 1.3 1.3 3.3 3.8 3.4 3.1 8.0 7.1 3.3 3.7 3.4 3.3
85-89 2.1 1.8 2.3 2.2 1.7 1.7 1.3 1.3 2.5 2.6 3.0 2.5 7.3 6.5 3.0 3.3 2.9 2.7
90-94 1.7 1.4 2.3 2.2 1.4 1.4 1.3 1.3 1.7 1.7 2.5 2.0 7.3 6.5 2.9 2.8 2.6 2.4
95-99 1.4 1.1 2.3 2.2 1.6 1.8 1.3 1.3 1.7 1.7 2.0 1.7 7.3 6.5 3.0 2.6 2.6 2.4
100+ 0.7 0.9 2.3 2.2 1.6 1.8 1.3 1.3 1.7 1.7 2.0 1.7 7.3 6.5 3.0 2.6 2.5 2.3
* M - males, F - females
Source: European Commission (2006) – ECFIN calculations based on country sources

Table 13. Projected public expenditures on health and the deficit with/without death-related costs
2010 2015 2020 2025 2030 2035 2040 2045 2050
POLAND
Total public health expenditures
– A (PLN mln) 56 654 76 414 100 683 129 842 162 756 196 793 228 340 255 028 275 730
– B (PLN mln) 56 131 75 211 98 500 126 199 157 216 189 222 218 565 242 699 260 821
Reduction in projected health
0.92 1.57 2.17 2.81 3.40 3.85 4.28 4.83 5.41
expenditures (%)
Deficit – A (PLN mln) 3 015 6 470 10 255 15 890 27 159 38 336 46 848 50 777 49 777
Deficit – B (PLN mln) 2 493 5 267 8 072 12 246 21 619 30 764 37 072 38 448 34 869
Reduction in projected
17.34 18.59 21.29 22.93 20.40 19.75 20.87 24.28 29.95
deficit (%)
HUNGARY
– A (HUF bln) 1 800 2 585 3 551 4 721 6 100 7 686 9 451 11 313 13 242
1 780 2 538 3 457 4 551 5 851 7 295 8 896 10 545 12 153
– B (HUF bln)
Reduction in projected health
expenditures (%)
1.14 1.82 2.64 3.60 4.07 5.09 5.88 6.79 8.23
Deficit – A (HUF bln) 518 686 906 1 456 2 159 3 040 4 100 5 297 6 642
Deficit – B (HUF bln) 498 639 812 1 286 1 911 2 649 3 545 4 530 5 553
Reduction in projected
deficit (%) 3.96 6.85 10.35 11.67 11.51 12.87 13.54 14.49 16.40
expenditure increase is slowed down and, as a result, projected financial deficit in the
health care system is reduced.
Table 12 presents total projected public expenditures on health in Poland and
Hungary. When differences in average per capita expenditures for the survivors and
the deceased are taken into account, projected health expenditures in Poland decrease
by 0.92% (in comparison to the baseline scenario), and in Hungary by 1.14% already
in 2010. The gap between baseline scenario and death-related costs scenario broadens
in years, amounting to 5.41% for Poland and 8.23% for Hungary at the end of
projection period. A decrease in estimated level of expenditures is reflected in the
decrease in estimated deficit. In Poland, deficit reduction amounts to 17% in 2010, and
almost 30% in 2050. In Hungary, deficit reduction is not that impressive, but quite
substantial still – close to 4% in 2010, and over 16% in 2050. Less extensive deficit
reduction in Hungary is a result of the size of the deficit – already in the year 2025 it
amounts to 30% of total projected health expenditures, and in 2050 it is equivalent to
50.2%. In Poland, projected financial balance looks a bit better since the peak of the
deficit falls for 2040 and amounts to 21% of total public expenditures on health care.
The gap in deficit size with and without the assumption regarding death-related
costs is quite significant and widens during projection period for both countries,
notwithstanding the differences in the size of the deficit between the two countries
(measured as percentage of GDP). By the end of projection period, projected deficit
reduction (scenario ‘including’ death-related costs compared to the scenario
44
A – baseline scenario
B – death-related costs scenario
Source: Gabos, Gal (2007), Golinowska, Kocot, Sowa (2007)

Graph 16. Projected deficit in Hungary and in Poland, as GDP percentage
4,00%
3,50%
3,00%
2,50%
2,00%
1,50%
1,00%
0,50%
0,00%
2005 2010 2015 2020 2025 2030 2035 2040 2045 2050
Poland-baseline scenario Poland-with death-related costs
Hungary-baseline scenario Hungary-with death-related costs
Source: Gabos, Gal (2007), Golinowska, Kocot, Sowa (2007)
‘excluding’ death-related costs) equals 29.95% for Poland and 16.4% for Hungary.
Overall, these results confirm the findings presented in other publications (some of
them quoted above), and indicate that when we take increased health expenditures in
the last year of life into account, it has significant impact on projection results and
decreases projected level of expenditures.
5.3. Diverse longevity scenarios
Scenarios allowing for different longevity development patterns were prepared for
each country. There are three possibilities with regard to the improvement of life
expectancy in the model applied: fast, medium and slow. In the baseline scenario, the
medium one was adopted, and the two residuals were considered as alternative scenarios.
There are no significant differences between the results from various scenarios. In
general, faster life expectancy improvement can lead to deeper deficit, but the extent
of that effect is rather insignificant (no differences in the deficit in Bulgaria, 0.1
percentage points in Hungary and 0.07 percentage points in Poland.) Faster increase
45

Table 14. Projection results for the scenarios of different life expectancy improvement:
slow, medium (baseline scenario), and fast (as a share of GDP)
LE – slow growth
in LE affects both sides of health care funding, expenditures as well as revenues. On
the one hand, there are more elderly individuals to take care of due to the unique
pattern of mortality and morbidity among the elderly, but on the other, increased LE
has positive influence on employment rate, most likely. Overall, the expenditures
grow faster than the revenues, so the deficit is higher. Slower life expectancy
improvement has the opposite effect.
5.4. Diversified wage growth scenario
The main source of revenues for social and health insurance system is the pay-roll
tax or insurance contributions. Therefore, wage development has significant influence
on the revenue side of health care sector financing. In the baseline scenario it was
assumed that wage growth is a derivative of assumed labor productivity, which variable
46
LE – medium growth
(baseline scenario) LE – fast growth
2005 2025 2050 2005 2025 2050 2005 2025 2050
BULGARIA
Health expenditures (% of G
DP) 5.0 6.4 7.0 5.0 6.4 7.1 5.0 6.4 7.2
Health revenues (% of GDP) 4.6 7.0 6.0 4.6 7.0 6.1 4.6 7.1 6.1
Surplus/deficit (% of GDP) -0.5 0.6 -1.0 -0.5 0.6 -1.0 -0.5 0.6 -1.0
ESTONIA
Health expenditures
(% of GDP) 4.0 4.9 6.7 4.0 5.0 6.7 4.0 5.0 6.8
Surplus/deficit (% of GDP) 0.2 -0.9 -3.3 0.2 -0.9 -3.4 0.2 -0.9 -3.4
HUNGARY
Health expenditures
5.5 5.9 6.7 5.5 5.9 6.7 5.5 5.9 6.8
(% of GDP)
Surplus/deficit (% of GDP) -1.5 -1.8 -3.3 -1.5 -1.8 -3.4 -1.5 -1.8 -3.5
POLAND
Health expenditures
(% of GDP) 4.63 5.84 6.92 4.63 5.88 7.08 4.63 5.92 7.25
Surplus/deficit (% of GDP) -0.18 -0.71 -1.21 -0.18 -0.72 -1.28 -0.18 -0.73 -1.35
SLOVAKIA
Health expenditures
(% of GDP) 4.83 5.57 6.78 4.83 5.59 6.85 4.83 5.61 6.95
Surplus/deficit (% of GDP) 1.50 -0.02 -2.32 1.50 -0.03 -2.35 1.50 -0.04 -2.44

is subjected to convergence strategy13. In alternative scenarios it was assumed,
additionally, that these variables (productivity and wages) will vary in plus and in minus
from the baseline scenario. In the case of higher dynamics of real wages growth (in
Bulgaria, 0.6 percentage points higher in 2025 and the same value in 2050, 1.5
percentage points higher in Slovakia, and in Hungary 3% higher over the whole period)
during projection period, health care sector deficit will be lower by at least 1 percentage
point of GDP share Although higher average wage growth has impact on the health
expenditure increase, by expanding labor costs, this effect is smaller than the effect of
47
Table 15. Scenarios of different wage development patterns (revenues, expenditures and deficit as
a share of GDP)
2005 2010 2015 2020 2025 2030 2035 2040 2045 2050
BULGARIA
Revenues 4.3 5.0 7.1 7.6 8.3 7.7 7.8 5.9 9.0 7.3
Expenditures 4.7 4.8 6.5 6.9 7.2 6.9 7.3 5.8 9.3 7.9
Faster
wage
growth
Surplus/deficit -0.4 0.2 0.6 0.7 1.1 0.8 0.5 0.1 -0.3 -0.6
Revenues 4.6 5.6 6.2 6.5 7.0 7.0 6.9 6.7 6.4 6.1
Expenditures 5.0 5.4 5.4 6.1 6.4 6.6 6.8 7.0 7.1 7.1
Baseline
scenario
Surplus/deficit -0.5 0.1 0.4 0.4 0.6 0.4 0.0 -0.3 -0.7 -1.0
Revenues 4.7 4.7 6.5 5.0 6.6 5.9 5.4 5.8 5.3 5.1
Expenditures 5.2 4.6 6.3 4.9 6.3 5.9 5.7 6.5 6.3 6.4
Slower
wage
growth
Surplus/deficit -0.5 0 0.2 0.1 0.3 0 -0.3 -0.7 -1 -1.3
HUNGARY
Revenues 4.0 3.9 4.2 4.4 4.4 4.3 4.2 4.2 4.2 4.1
Expenditures 5.5 5.5 5.6 5.7 5.9 6.0 6.2 6.4 6.5 6.7
Faster
wage
growth
Surplus/deficit -1.5 -1.6 -1.4 -1.3 -1.6 -1.8 -2.0 -2.2 -2.4 -2.5
Revenues 4.0 3.9 4.1 4.3 4.1 3.9 3.8 3.6 3.5 3.4
Expenditures 5.5 5.5 5.6 5.8 5.9 6.1 6.2 6.4 6.6 6.7
Baseline
scenario
Surplus/deficit -1.5 -1.6 -1.5 -1.5 -1.8 -2.2 -2.5 -2.8 -3.1 -3.4
Revenues 4.0 3.9 4.1 4.1 3.9 3.6 3.4 3.2 3.1 2.9
Expenditures 5.5 5.5 5.6 5.8 6.0 6.1 6.3 6.5 6.7 6.8
Slower
wage
growth
Surplus/deficit -1.5 -1.6 -1.6 -1.7 -2.1 -2.5 -2.9 -3.3 -3.6 -3.9
SLOVAKIA
Revenues 6.33 6.40 5.99 5.86 5.87 5.93 5.95 5.85 5.69 5.54
Expenditures 4.83 4.98 5.15 5.35 5.59 5.84 6.10 6.36 6.61 6.85
Faster
wage
growth
Surplus/deficit 1.50 1.42 0.84 0.51 0.28 0.09 -0.15 -0.51 -0.92 -1.31
Revenues 6.33 6.40 5.98 5.75 5.56 5.41 5.25 5.00 4.73 4.50
Expenditures 4.83 4.98 5.15 5.35 5.59 5.84 6.10 6.36 6.61 6.85
Baseline
scenario
Surplus/deficit 1.50 1.42 0.83 0.40 -0.03 -0.43 -0.86 -1.36 -1.87 -2.35
Revenues 6.33 6.40 5.97 5.66 5.32 5.03 4.75 4.42 4.10 3.84
Expenditures 4.83 4.98 5.15 5.35 5.59 5.84 6.10 6.36 6.61 6.85
Slower
wage
growth
Surplus/deficit 1.50 1.42 0.82 0.31 -0.27 -0.82 -1.35 -1.94 -2.50 -3.01
Source: Gabos, Gal (2007), Kvetan, Palenik, Mlynek, Radvansky (2007), Rangelova, Sariiski (2007)
13 In the baseline scenarios of convergence strategy prepared in the national plans and AWG report (EPC and
European Commission 2006), it was assumed that welfare in NMS will be close to an average EU level in the
final part of projection period, which means that labor productivity rate should be three times as high in the
following years, and twice as high at the end of the first part of projection period (up to 2025.)

revenue increase. Hence, the impact on health sector financial balance is positive.
Slower wage growth scenario (in Bulgaria, 0.6 percentage points lower in 2025
and 1 percentage point in 2050, in Hungary 0.5% from 2007) yields the opposite result
– higher deficit compared to the baseline scenario, and twice as high in comparison
to the high wage growth scenario.
5.5. Labor market indicators diversified development scenario
Different values of indicators pertaining to the labor market (employment and
unemployment rate) do impact projection results, but the scale of the change is
smaller than in the case of wage growth. By the same token, employment rate
diversification in projections made for Poland and Slovakia exerts only slight
48
Table 16. Scenarios of different activity rate development (expenditures, revenues and deficit
as GDP share)
2005 2010 2015 2020 2025 2030 2035 2040 2045 2050
HUNGARY
Revenues 4.1 4.1 4.4 4.6 4.5 4.3 4.2 4.1 4.0 4.0
Expenditures 5.5 5.5 5.6 5.8 6.0 6.1 6.3 6.4 6.6 6.8
Faster
activity
rate
growth
Surplus/deficit -1.4 -1.4 -1.2 -1.2 -1.5 -1.8 -2.1 -2.3 -2.6 -2.8
Revenues 4.0 3.9 4.1 4.3 4.1 3.9 3.8 3.6 3.5 3.4
Expenditures 5.5 5.5 5.6 5.8 5.9 6.1 6.2 6.4 6.6 6.7
Baseline
scenario
Surplus/deficit -1.5 -1.6 -1.5 -1.5 -1.8 -2.2 -2.5 -2.8 -3.1 -3.4
Revenues 4.0 3.8 3.8 3.9 3.7 3.5 3.4 3.3 3.2 3.1
Expenditures 5.5 5.5 5.6 5.7 5.9 6.0 6.2 6.4 6.5 6.7
Slower
activity
rate
growth
Surplus/deficit -1.5 -1.7 -1.7 -1.9 -2.2 -2.5 -2.8 -3.1 -3.3 -3.6
Table 17. Scenarios of different employment rate development (expenditures, revenues and deficit
as GDP share)
2005 2010 2015 2020 2025 2030 2035 2040 2045 2050
POLAND
Revenues 4.45 4.73 4.85 5.04 5.24 5.28 5.36 5.48 5.65 5.88
Expenditures 4.63 4.92 5.22 5.53 5.88 6.24 6.56 6.80 6.96 7.08
employment
Faster
rate
growth
Surplus/deficit -0.18 -0.20 -0.37 -0.48 -0.64 -0.96 -1.20 -1.32 -1.31 -1.20
Revenues 4.45 4.66 4.77 4.96 5.16 5.20 5.28 5.40 5.57 5.80
Expenditures 4.63 4.92 5.22 5.53 5.88 6.24 6.56 6.80 6.96 7.08
Baseline
scenario
Surplus/deficit -0.18 -0.26 -0.44 -0.56 -0.72 -1.04 -1.28 -1.39 -1.39 -1.28
Revenues 4.45 4.63 4.70 4.88 5.08 5.12 5.21 5.33 5.50 5.73
Expenditures 4.63 4.92 5.22 5.53 5.88 6.24 6.56 6.80 6.96 7.08 Slower
employment
rate
growth
Surplus/deficit -0.18 -0.30 -0.52 -0.64 -0.80 -1.12 -1.36 -1.47 -1.46 -1.35

49
Table 17. continued. Scenarios of different employment rate development (expenditures, revenues
and deficit as GDP share)
2005 2010 2015 2020 2025 2030 2035 2040 2045 2050
SLOVAKIA
Revenues 6.33 6.38 5.93 5.74 5.70 5.54 5.37 5.11 4.83 4.59
Expenditures 4.83 4.98 5.15 5.35 5.59 5.84 6.10 6.36 6.61 6.85
employment
Faster
rate
growth
Surplus/deficit 1.50 1.40 0.78 0.38 0.11 -0.30 -0.73 -1.25 -1.78 -2.27
Revenues 6.33 6.40 5.98 5.75 5.56 5.41 5.25 5.00 4.73 4.50
Expenditures 4.83 4.98 5.15 5.35 5.59 5.84 6.10 6.36 6.61 6.85
Baseline
scenario
Surplus/deficit 1.50 1.42 0.83 0.40 -0.03 -0.43 -0.86 -1.36 -1.87 -2.35
Revenues 6.33 6.43 5.96 5.59 5.38 5.24 5.09 4.85 4.61 4.40
Expenditures 4.83 4.98 5.15 5.35 5.59 5.84 6.10 6.36 6.61 6.85
employment
Slower
rate
growth
Surplus/deficit 1.50 1.45 0.82 0.23 -0.21 -0.60 -1.02 -1.50 -2.00 -2.46
Source: Golinowska, Kocot, Sowa (2007), Kvetan, Palenik, Mlynek, Radvansky (2007)
Table 18. Scenarios of different labor force participation and employment rate development
(expenditures, revenues and deficit as GDP share)
ESTONIA
Revenues 4.4 4.4 4.4 4.3 4.3 4.2 4.1 3.9 3.8 3.6
Expenditures 4.0 4.3 4.5 4.7 4.9 5.3 5.6 6.0 6.3 6.7
Faster labor
participation
force
employment
rate growth
and
Surplus/deficit 0.4 0.1 -0.1 -0.4 -0.7 -1.1 -1.6 -2.0 -2.6 -3.1
Revenues 4.2 4.2 4.2 4.1 4.1 4.0 3.8 3.7 3.5 3.3
Expenditures 4.0 4.3 4.5 4.7 5.0 5.3 5.6 6.0 6.3 6.7
Baseline
scenario
Surplus/deficit 0.2 0.0 -0.3 -0.6 -0.9 -1.3 -1.8 -2.3 -2.8 -3.4
Revenues 4.3 4.3 4.2 4.1 4.0 3.9 3.8 3.7 3.5 3.3
Expenditures 4.0 4.3 4.5 4.7 5.0 5.3 5.6 6.0 6.4 6.7
Slower labor
participation
force
employment
rate growth
and
Surplus/deficit 0.3 0.0 -0.3 -0.6 -1.0 -1.4 -1.8 -2.3 -2.9 -3.4
2005 2010 2015 2020 2025 2030 2035 2040 2045 2050
Table 19. Scenarios of different unemployment rate (expenditures, revenues and deficit as GDP
share)
2005 2010 2015 2020 2025 2030 2035 2040 2045 2050
BULGARIA
Revenues 5.0 8.0 6.4 6.2 7.1 7.9 6.2 5.8 6.3 6.0
Expenditures 5.5 7.9 6.1 5.9 6.5 7.5 6.2 6.1 7.0 7.0
unemployment
rate growth*
Different
Surplus/deficit -0.5 0.1 0.3 0.3 0.6 0.4 0 -0.3 -0.7 -1.0
Revenues 4.6 5.6 6.2 6.5 7.0 7.0 6.9 6.7 6.4 6.1
Expenditures 5.0 5.4 5.4 6.1 6.4 6.6 6.8 7.0 7.1 7.1
Baseline
scenario
Surplus/deficit -0.5 0.1 0.4 0.4 0.6 0.4 0.0 -0.3 -0.7 -1.0
* Changes in the minimum unemployment rate: 14% (for the base year, instead of 7% in the baseline
scenario); 7% (in 2025, instead of 5% in the baseline scenario); 4% (in 2050 the value is the same as in
the baseline scenario)
Source: Rangelova, Sariiski (2007)

influence on financial balance of health sector, and none at all on health care
expenditures (Gabos, Gal (2007), Golinowska, Kocot, Sowa (2007), Kvetan, Palenik,
Mlynek, Radvansky (2007), Rangelova, Sariiski (2007), Roovali (2007)).
Results for Bulgaria, presented in Table 19, do not corroborate the impact of
unemployment rate variable on health care sector finances. It seems that
unemployment rate indicator is not sufficient to illustrate problems connected with
the labor market in the context of health care sector budget projection.
5.6. Sensitivity analysis
For the purpose of sensitivity analysis of projection results, a new indicator was
established – the relative change of revenues that takes place in consequence of the
changes in the parameters under analysis. The analyzed year is 2050, the last year of
projection period. The indicator has got a percentage value. The following sensitivity
scale was adopted:
• insensitivity: up to 10%;
• moderate sensitivity: 10% - 40%;
• sensitivity: above 40%.
In sensitivity analysis, relative change in parameter value was not taken into
account – it was recognized simply as a probable for the analyzed country. Death-
50
'-' indicates that no such scenario was prepared
* activity rate for Hungary, activity and employment rate for Estonia
Source: own calculations based on Gabos, Gal (2007), Golinowska, Kocot, Sowa (2007), Kvetan, Palenik,
Mlynek, Radvansky (2007), Rangelova, Sariiski (2007), Roovali (2007)
Table 20. Revenue side sensitivity
Country Bulgaria Estonia Hungary Poland Slovakia
Sensitivity degree
Scenario
insensitive
moderate
sensitive
insensitive
moderate
sensitive
insensitive
moderate
sensitive
insensitive
moderate
sensitive
insensitive
moderate
sensitive
Death-related costs scenario - - X X -
Fast X X X X X
LE improvement
Slow X X X X X
Faster X - X - X
Wage growth
Slower X - X - X
Employment Faster - X X X X -
rate growth* Slower - X X X X
Unemployment
rate decline Slower X - - - -

Table 21. Expenditure side sensitivity
Sensitivity degree
Scenario
insensitive
moderate
sensitive
insensitive
moderate
sensitive
insensitive
moderate
sensitive
insensitive
moderate
sensitive
insensitive
moderate
sensitive
FastX X X X X
LE improvement
SlowX X X X X
Faster X - X - X
Wage growth
Slower X - X - X
Unemployment
Table 22. Deficit/surpluse sensitivity
Sensitivity degree
Scenario
insensitive
moderate
sensitive
insensitive
moderate
sensitive
insensitive
moderate
sensitive
insensitive
moderate
sensitive
insensitive
moderate
sensitive
FastX X X X X
LE improvement
SlowX X X X X
Faster X - X - X
Wage growth
Slower X - X - X
Unemployment
costs scenario is of different nature than the other scenarios, in that the revenue
(expenditures, deficit/surplus) change is the result of a new variable used rather than
a change of the variable used previously.
51

The main conclusion is that health revenues, expenditures and deficit/surplus are
slightly sensitive to variable changes in general. There are only two cases of real
sensitivity – in the scenario of faster wage growth in Hungary and Slovakia.
The consideration of higher health expenditures in the last year of life has no
impact on the revenue side and a slight impact on expenditures, but it strongly affects
the deficit in health care sector.
In the case of different life expectancy scenarios, both the revenues, the
expenditures, and the deficit are insensitive, even to the point of 0% value.
As was mentioned earlier, a stronger impact can be observed for changes in wage
growth rate, especially on the revenue side.
Projection results are not sensitive to the changes in labor market variables, except
for the impact of higher activity rate in Hungary on the revenue side and,
consequently, on the deficit as well.
52

Chapter 6.
Conclusions, discussion and policy
recommendations
Health care sector balance projection based on the ILO social budget model shows
basic trends in the development of future revenues and expenditures shaped by the
impact of the so-called external factors, i.e. those that do not belong to the domain of
health care system: demography and economy, as well as current medical service
utilization pattern depending on age (J curve.) The variables pertaining to population
ageing are especially important for the results of the analysis. These factors include:
increasing share of the elderly in the population, decreasing share of labor market
active population, and – as a result - decreasing labor supply and changing trends in
labor productivity. The impact of those factors was also reflected in adopted scenarios
of future economic growth development. Projected revenues, expenditures and public
health care sector balance are presented in relation to economic growth.
The results of health care sector financial projections enable us to analyze
simultaneously the two sides of public health care sector: revenues, expenditures and
financial balance (surplus/deficit) with interactions between sides, as well as impact
on balance effect. That is the advantage of ILO social budget model. As a result,
conclusions and recommendations can be well-targeted, depending on the element of
the health care sector that each of the variables addresses.
If labor market activity rate strongly affects health care sector revenues,
decreasing the deficit in public health care sector, recommendations address labor
market policy and economy. This policy includes high economic growth, taxable
income and employment.
If, on the other hand, longevity has strong impact on increasing expenditures, the
policy is targeted towards expenditure decrease, which is a less effective policy.
Rational demographic policy, required to shape health care system expenditures,
should create incentives towards changing values and behaviors in the life-cycle
related to procreation and health. One must add that there is a correlation between
longer economic activity and health of the population. Only healthy and more active
53

population can stay active on the labor market for the longer period, which will lead
to the increase in expenditures on health, especially public health.
Health care sector model also takes into account expenditures on health modified
by higher costs incurred in the last year of life. Restricted access to data was an
obstacle to the process of preparing similar projections in all the countries under
analysis. Projections for Hungary and Poland, however, confirm that death-related
cost is a significant factor. Longevity leads to an “extension in time” of higher health
care costs incurred in the last year of life. In consequence, aggregated health
expenditures in a given projection period are lower. Death-related costs scenario
substantially decreases estimated financial deficit in health care system, in
comparison to the baseline scenario. Still, the impact of death-related costs on
expenditures is restricted since it is contingent upon potential future longevity
developments. These are not infinite.
In the table below, key results of analyses and projections are compiled, together
with generalized relationships and tendencies. Direction of these changes is similar in
the analyzed group of countries. These countries, with higher dynamics of
demographic changes and, at the same time, lower level of economic development,
may face greater tensions, or even conflicts, between future development objectives.
From social and political perspective, decisions regarding health care financing
are particularly sensitive due to higher health needs. Changes are also necessary due
to the high costs of the health care system – costs incurred and potential costs related
to the ageing. Finally, one of the most crucial cost drivers in the health sector – new
medical technology – was not isolated and accounted for in the analysis.
The ILO social budget model not only lends itself to the projection of future
financial balance (in this case, a deficit at the end of the projection period), but it also
allows for the calculation of insurance premium (tax) needed to balance health care
system revenues and expenditures.
Health care system reforms are required in order to balance the disequilibrium of
revenues and expenditures caused by external factors (demographic and economic), and
decrease the premium needed to cover expenditures. Such reforms should lead, on the
one hand, to the rationing of medical services covered by public resources, and on the
other, to more effective governance and management of the sector and within the sector.
Finally, it should be added that the ageing process affects other types of
expenditures, i.e. those outside of health care sector. In the CEE, one issue is
financing income benefits (pensions), which is not discussed here, and another issue
is the need to increase the level of long term care (LTC) financing. Up to date, this type
of care was most commonly provided within the family rather than within the
framework of institutionalized social protection. In Poland, for example, it has also
been provided by religious organizations and, albeit only lately, by commercial
54

Table 23. Summary – factors that affect health care sector budget in projection exercise
Sides
of the budget Variables
Source: Own comparison based country reports
(private) long-term care institutions. Access to this type of services is restricted, and
long-term diseases related to old age are becoming one of the most urgent social
issues. Overall, the hypothesis that new social problems will arise, the demand for
social care will increase, and new financing mechanisms will be needed in the face of
ageing, has been confirmed. The other reasons include rapid economic development
and modernization processes in the NMS. Thus, rationalization of health care
expenditures becomes a necessity, as well as building the space for new taxes and – if
socially accepted – increased private financing.
55
Channel trough which
budget side is affected Impact on projection results – evidence
Demographic:
population size,
total fertility rate
(TFR) and longevity
Population size and age structure
determine the number of persons
who can potentially contribute
to public health care budget
Positive impact for variants of improved
demographic development: especially
total fertility rate (TFR) and healthy
longevity, negative – for population
decline and dynamic aging. Longevity
quantitative, not qualitative
Labor market
development:
participation rate
and employment rate
Higher employment rate
increases the number of tax
payers (contributors) to health
budget
Significant positive effect strengthened
by the increase of wages in Slovakia,
moderate positive effect in Hungary
Economic development:
wages
Wages are the main source
of taxable income
Positive effect, proven for
Hungary, Slovakia and Bulgaria
Revenues
Utilization pattern
Promotion of investments
in health; approval for increased
health payments
Assumption – not modeled
Demographic: longevity
and age structure
Increased aging and longevity
extend the time spent with
poorer health status
Positive effect: higher for Bulgaria,
moderate for Hungary
Death related costs
Morbidity picture of the elderly
tends to sharply increase
the expenditures at an older age,
but the costs are substantially
higher only in the last year of life,
so the ratio between the deceased
and the survivors declines with age
Slower increase; 1 percentage point
of GDP (Poland) and 0.7 percentage
points of GDP (Hungary)
Labor market
development:
employment
and unemployment
Unemployment leads to poorer
health status and more frequent
utilization of health services
No evidence
wages, labor
productivity and GDP
Income elasticity of demand
for health rather neutral
at an aggregate level; wage
increase entails higher labor cost
in the health sector as well
Moderate influence
Expenditures
Utilization pattern
Changes in line with ageing
and wage increase when incom
elasticity is greater than 1
Assumption
Demographics:
population size, age
structure, longevity
Strengthens the effect on both
sides: decrease in revenues
and increase in expenditures
Higher deficit
Death related costs On the expenditure side
Deficit is modified – smaller than
in the baseline scenario
Labor market
development:
employment
On the revenue side Smaller deficit – moderate effect
Financial
balance
wages and labor
productivity
Interactions between sides:
increase in revenues higher
than a decrease in expenditures
Smaller budget deficit – moderate effect
(Hungary), stronger effect (Bulgaria)

56
Table 24. Comparison: insurance premium required to balance public health care sector revenues
and expenditures
Country Base year 2025 2050
Bulgaria n/a n/a n/a
Estonia 12.4% 15.8% 25.8%
Hungary 17% 21% 28%
Poland 8.6% 10.7% 11.7%
Slovakia 9.4% 12.7% 21.5%

Annex 1
Population changes 2003-2050
57
Graph 17. Population histogram – Bulgaria 2003
90-94
80-84
70-74
60-64
50-54
40-44
30-34
20-24
10-14
Males Females
Year : 2003
350 250 150 50 -50 -150 -250 -350
0-4
Thousands

58
90
80
70
60
50
40
30
20
10
0
20 000 10 000 0 10 000 20 000
Graph 18. Population histogram – Bulgaria 2050
90-94
80-84
70-74
60-64
50-54
40-44
30-34
20-24
10-14
Year : 2050
350 250 150 50 -50 -150 -250 -350
0-4
Thousands
Males Females
Graph 19. Population histogram – Estonia 2003
Age
Males Females

59
Graph 20. Population histogram – Estonia 2050
90
80
70
60
50
40
30
20
10
0
Graph 21. Population histogram – Hungary 2002
100
95
90
85
80
75
70
65
60
55
50
45
40
35
30
25
20
15
10
5
0
-100 000 -80 000 -60 000 -40 000 -20 000 0 20 000 40 000 60 000 80 000 100 000
20
000
10
000
10 0
000
20
000
Age
Males Female
Population pyramid -
2002
Age
Males Females

60
10
96
92
88
84
80
76
72
68
64
60
56
52
48
44
40
36
32
28
24
20
16
12
8
4
0
400 300 200 100 0 100 200 300 400
Graph 22. Population histogram – Hungary 2050
100
95
90
85
80
75
70
65
60
55
50
45
40
35
30
25
20
15
10
5
-80
000
-60
000
Source: Gabos Gal (2007)
-40
000
-20
000
0 20 000 40 000 60 000 80 000
0
Population pyramid -
2050
Age
Males Females
Graph 23. Population histogram – Poland 2003
Age
Source: Golinowska, Kocot, Sowa (2007)
Females Males

61
Graph 24. Population histogram – Poland 2050
10
96
92
88
84
80
76
72
68
64
60
56
52
48
44
40
36
32
28
24
20
16
12
8
4
0
400 300 200 100 0 100 200 300 400
Age
Source: Golinowska, Kocot, Sowa (2007)
Graph 25. Population histogram – Slovakia 2005
96
90
84
78
72
66
60
54
48
42
36
30
24
18
12
6
0
-50 000 -40 000 -30 000 -20 000 -10 000 0 10 000 20 000 30 000 40 000 50 000
Age
Females Males
Source: Kvetan, Páleník, Mlýnek, M.Radvanský (2007)
Females Males

62
96
90
84
78
72
66
60
54
48
42
36
30
24
18
12
6
0
50000 40000 30000 20000 10000 0 10000 20000 30000 40000 50000
Graph 26. Population histogram – Slovakia 2050
Age
Source: Kvetan, Páleník, Mlýnek, M.Radvanský (2007)
Females Males

Annex 2
Methodology for death related costs estimation for Hungary and Poland
1. Estimation based on National Health Fund data on average expenditures per capita
(per deceased and survivor, broken down according to gender and 5-year age
cohorts) for medical services for which appropriate data is available:
where:
l ∈{s,d}
l
AE TE
l a p
a p N
, ,2004
l
a p
, ,2004
, ,2004 =
, status of an individual: s – survivor, d– deceased
, 5-years age cohort
, sex: m– male, f – female
– average per capita expenditures for an individual of l status, age
cohort a and sex p in the year 2004
– total expenditures for l status, age cohort a and sex p (according to
a∈{1,..,20}
p∈{m, f }
l
AEa, p,2004
l
TEa, p,2004
National Health Fund data in the year 2004)
– number of individuals with l status (deceased in the given year or
l
a p N , ,2004
survivors) in age cohort a and sex p in the year 2004 (according to the
demographic projection used in the ILO model.)
2. Estimation of k-indicator, broken down according to gender and age:
d
a p
, ,2004
s
a p
k AE
, =
a p AE
, ,2004
K-indicator describes the ratio between average expenditures for given medical
services per deceased in the base year and average expenditures per survivor. The
63

value of the indicator is constant during the projection period and specified according
to the sex and age cohort.
3. Estimation based on National Health Fund data, pertaining to annual insurance
expenditures for given medical services. As the National Health Fund data includes
the data with an undefined status of a person (without information on sex and/or
age), it was necessary to divide these expenditures using the k-indicator between
groups, in order to receive expenditures coherent with National Health Data. These
estimations, covering year 2004, are further applied to the 2003 National Health
Data assuming constancy of the k-indicator and unchanged proportions of
expenditures between genders and age cohorts. Average expenditures per capita for
a person with s (survivor) and d(deceased) status were calculated as:
where:
, ,2003
−α + ⋅α
d
a p AE k AE , ,2003 , , ,2003 = ⋅
a - probability of death depending on sex and age cohort in a given year,
– average per capita expenditures depending on sex and age,
a, p,2003 AE
calculated based on National Health Data expenditures, applying sex and age
structure from National Health Fund data
4. Calculation of average per capita expenditures depending on sex, age and status for
the year n and assuming growth of expenditures in line with GDP per capita.
where:
AE s
= AE s
⋅ + r a p n a p n
−
(1 ) , , , , 1 n
AE d
= AE d
⋅ + r a p n a p n
−
(1 ) , , , , 1 n
n – following year of projection
r n – GDP per capita growth rate in year n
s
a p a p
=
k
AE
AEs a p
a p
(1 )
, ,2003
64

5. Calculation of total expenditures in year n:
where:
20
20
, , =Σ ⋅ +Σ ⋅
TE s
AE N AE N n , , , ,
, ,
a
= =
1
20
a
1
20
s
a f n
s
a m n
s
a m n
, , =Σ ⋅ +Σ ⋅
s
a f n
TE d
AE N AE N n , , , ,
, ,
d
n
d
a m n
= =
s
a
d
a m n
1
n n TE = TE +TE
d
a f n
d
a f n
a
1
– total expenditures in year n for survivors (s ) and deceased (d)
d
n
s
TEn ,TE
– total expenditures in year n
n TE
6. Due to lack of adequate data on pharmaceutical expenditures, additional indicator
defining the ratio between average pharmaceutical expenditures per deceased and
average pharmaceutical expenditures per survivor is used. It is assumed that the
value of this indicator equals the value of the indicator for other types of medical
services. As a result, total expenditures are summed up with expenditures on
pharmaceuticals, divided by sex, age cohort and status.
7. In view of the capitation system in primary care services financing, National Health
Fund expenditures on primary care are divided proportionally between the
insured, and added to total expenditures.
65

Ahn N., Garcia J.R., Herce J.A. (2005), Demographic Uncertainty and Health Care Expenditure in
Spain, Paper prepared for the DEMWEL project, FEDEA
Antolin P., Oxley H., Suyker W. (2001), How will ageing affect Finland?, OECD Economics Departme
Working Papers No.295
Ariste R., Carr J. (2003), New Considerations on the Empirical Analysis of Health Expenditures in
Canada: 1966-1998, Health Policy Research Working Paper Series, WP 02-06, Policy Research
Communication Unit, Canada
Batljan I., Lagergren M. (2004), Inpatient/outpatient health care costs and remaining years of life
– effect of decreasing mortality on future acute health care demand, Social Science and Medicine 59,
pp.2459-2466
Brockmann H. (2002), Why is less money spent on health care for the elderly than for the rest of the
population? Health care rationing in German hospitals, Social Science & Medicine 55, pp.593-
608
Busse R., Krauth C., Schwartz F.W. (2002), Use of acute hospital beds does not increase as the
population ages: results from a seven year cohort study in Germany, Journal of Epidemiology
and Community Health 56, pp.289-293
Carrin G., Ron A. (1993), Towards a framework for health insurance in Viet Nam, WHO, Geneva
Christiansen T., Bech M., Lauridsen J., Nielsen P. (2006), Demographic changes and aggregate health
care expenditure in Europe, paper for AHEAD project, WP VI
Cichon M., Newbrander W., Yamabana H., Weber A., Normand Ch., Dror D., Preker A. (1999),
Modelling in health care finance. A compendium of quantitative techniques for health care
financing, International Labour Office, Geneva
Clemente J., Marcuello C., Montanes A., Pueyo F. (2004), On the international stability of health
care expenditure functions: Are government and private functions similar ?, Journal of Health
Economics 23, pp. 589-613
Cutler D.M., Sheiner L. (1998), Demographics and Medical Care Spending: standard and non-standar
effects, Burch Working Paper No.B98-3, University of California, Berkeley, USA
Denton F.T., Gafni A., Spencer B.G. (2001), Exploring the effects of population change on the costs
of physician services, Journal of Health Economics 21, s.781-803
Dixon T., Shaw M., Frankel S., Ebrahim S. (2004), Hospital Admissions, age and death: retrospective
cohort study, British Medical Journal 328, pp.1288-1291
Dreger Ch., Reimenrs H-E. (2005), Health Care Expenditures in OECD Countries: A Panel Unit Root
66
References

and Cointegration Analysis, The Institute for the Study of Labour Discussion Paper No.1469, Bonn
Drygas, W. (2005), Usprawnienie wczesnej diagnostyki i czynnej opieki nad osobami z ryzykiem
rozwoju niedokrwiennej choroby serca, (Improvements in early diagnostics and active care over
myocardial ischemia risk patients), In: P. Goryński, B. Wojtyniak K. Kuszewski (ed.) Monitoring
oczekiwanych efektów realizacji Narodowego Programu Zdrowia 1996-2005. Suplement (Monitoring
of expected outcome of 1996-2005 National Health Plan), Ministerstwo Zdrowia i Państwowy
Zaklad Higieny (Ministry of Health and National Institute of Hygiene), Warszawa
Eberstadt N. (2005), Health and Demography: The Achilles’ Heel of Post – Socialist development in
Europe, presentation on the conference “Advancing Economic Growth: Investing in Health”,
Chatham House, June 22-23, London
EPC and European Commission (2006), The impact of ageing on public expenditure: projections for
the EU25 Member States on pensions, health care, long-term care, education and unemployment
transfers (2004-2050), Report No 1/2006
Fogel R.W. (1999), Catching Up with the Economy, The American Economic Review 89(1), pp.1-21
Fuchs V.R. (1998), Provide, provide: the economics of aging, Working Paper 6642, National Bureau
of Economic Research, Cambridge, USA
Gabos A., Gal R.I. (2007), Development of Scenarios for Health Expenditure in the Accession
Economies. Country report – Hungary, http://shop.ceps.eu/BooksList.php?category_id=17&
Gabriele S., Cislaghi C., Constantini F., Innocenti F., Lepore V., Raitano M., Tediosi M., Valerio M.,
Zocchetti C. (2005), Demographic factors and health expenditure profiles by age : the case of Italy,
XVII Conferenza Societa Italiana di Economia Pubblica paper, Universita di Pavia
Gerdtham U.G., Sögaard J., MacFarlan M., Oxley H. (1998), The determinants of health expenditure
in the OECD countries, w: Zweifel P. (red.), Health, the Medical Profession and Regulation,
Kluver, pp. 113-134
Gertler P., van der Gaag (1990), The willingness to pay for medical care, Johns Hopkins Press,
Baltimore
Getzen T. (1990), Macro forecasting of national health expenditures, Advances in health economics
and health services research 11, pp. 27-48
Getzen T., Poullier J.P. (1992), International health spending forecasts: Concepts and evaluation,
Social Science and Medicine 34(9), 1057-1068
Getzen T.E. (2000), Health care is an individual necessity and a national luxury: applying multilevel decision
models to the analysis of health care expenditures, Journal of Health Economics 19, pp. 259-270
Getzen T.E. (2000a), Forecasting Health Expenditures: Short, Medium and Long (Long) Term, Journal
of Health Care Finance, 26(3), 56-72
Golinowska S., Kocot E., Sowa A. (2006), Development of Scenarios for Health Expenditure in the
Accession Economies. Country report – Poland, http://shop.ceps.eu/BooksList.php?category_id=17&
Golinowska S., Sowa A., Topór-Mądry R. (2006), Health Status and Health Care Systems in Central
and Eastern European Countries: Bulgaria, Estonia, Poland, Slovakia and Hungary,
http://shop.ceps.be/BooksList.php?category_id=17&
Golinowska S., Sowada Ch., Wozniak M. (2007), Sources of inefficiency and financial deficits in
67

Poland’s health care system, WHO memo
Gornick M., McMillan A., Lubitz J.A. (1993), A longitudinal perspective on patterns of Medicare
payments, Health Affairs (Millwood) 12(2), pp.140-150
Hoover D.R., Crystal S., Kumar R., Sanbamoorthi U., Cantor J.C. (2002), Medical Expenditures
during the Last Year of Life: Findings from the 1992-1996 Medicare Current Beneficiary Survey,
Health Services Research 37 (6), pp.1625-1642
Kildemoes H.W., Christiansen T., Gyrd-Hansen D., Kristiansen I.S., Andersen M. (2006), The impact
of population ageing on future Danish drug expenditure, Health Policy 75, 298-311
Kildemoes H.W., Christiansen T., Gyrd-Hansen D., Kristiansen I.S., Andersen M. (2006), The impact
of population ageing on future Danish drug expenditure, Health Policy 75, pp.298-311
Kramer A.M. (1995), Health Care for Elderly Persons – Myths and Realities, New England Journal of
Medicine 332 (15), pp.1027-1029
Kvetan V., Palenik V., Mlynek M., Radvansky M. (2006), Health Expenditure Scenarios in Slovakia.
Country Report – Slovak Republic, http://shop.ceps.eu/BooksList.php?category_id=17&
Kwiatkowski E. (2002), Bezrobocie. Podstawy teoretyczne (Unemployment. Theoretical premises),
Wydawnictwo Naukowe PWN, Warszawa
Lee R., Miller T. (2002), An Approach to Forecasting Health Expenditures, with Application to the
U.S. Medicare System, Health Services Research 37(5), 1365-1386
Lubitz J., Prihoda R. (1984), The use and costs of Medicare services in the last 2 years of life, Health
Care Financing Review 5, pp. 117-131
Lubitz J.D., Riley G.F. (1993), Trends in Medicare payments in the last year of life, New England
Journal of Medicine 328, pp.1092-1096
Madsen J., Serup-Hansen N., Kragstrup J., Kristiansen I.S. (2002), Ageing may have limited impact
on future costs of primary health providers, Scandinavian Journal of Primary Care 20, pp.169-173
Mahal A., Berman P. (2001), Health Expenditures and the Elderly: A Survey of Issues in
Forecasting, Methods Used and Relevance for Developing Countries, Research Paper No.01.23,
Harvard Burden of Disease Unit, Cambridge
Mayhew L. (2000), Health and Elderly Care Expenditure in an Aging World, International Institute for
Applied Systems Analysis, Austria
McGrail K., Green B., Barer M.L., Evans R.G., Hertzman C. (2000), Age, costs of acute and long-term
care and proximity to death: evidence for 1987-88 and 1994-95 in British Columbia, Age and Ageing
29, pp. 249-253
Murthy N., Ukpolo V. (1994), Aggregate health care expenditure in the United States: evidence from
cointegration tests, Applied Economics 26, pp.797-802
MZ (Ministerstwo Zdrowia) (2004), Finansowanie ochrony zdrowia w Polsce – Zielona Księga [Health
Care Financing in Poland – the Green Book], Ministry of Health, Warsaw
National Health Insurance Agency for Wage Earners (2003), Des Tendances de Fond aux
Mouvements de Court Terme, Point de conjoncture No 15
Newhouse J.P. (1977), Medical care expenditure: a cross-national survey. Journal of Human
Resources 12, 115-125
68

OECD (2006), Projecting OECD health and long-term expenditures: what are the main drivers?,
ECO/WKP(2006)5, OECD
Okólski, M. (2004), Demografia zmiany społecznej (Social change demographics), Wydawnictwo
Naukowe Scholar, Warszawa
Okunade A.A., Murthy V.N.R. (2002), Technology as a “major driver” of health care costs: a
cointegration analysis of the Newhouse conjecture, Journal of Health Economics 21(1), pp.147-159
Polder J.J., Achterberg P.W. (2004), Cost of illness in the Netherlands, RIVM Report No.270751006,
The National Institute for Public Health and the Environment, Bilthoven, Holandia
Raitano M. (2006), The impact of Death-Related Costs on Health-Care Expenditure: A Survey,
ENEPRI Research Report No.17, AHEAD WP7
Rangelova R., Sariiski G. (2006), Development of Scenarios for Health Expenditure in the Accession
Economies. Country report for Bulgaria, http://shop.ceps.eu/BooksList.php?category_id=17&
Roovali L. (2007), Health Expenditure Scenarios in the New Member States: Country Report on
Estonia, http://shop.ceps.eu/BooksList.php?category_id=17&
Scholz W., Cichon M., Hagemejer K. (2000), Social budgeting, International Labour Office, Geneva
Serup-Hansen J., Wickstrom J., Kristiansen I.S. (2002), Future health care costs – do health care costs
during the last year of life matter?, Health Policy 62, pp.161-172
Seshamani M., Gray A. (2004), Time to death and health expenditure: an improved model for the
impact of demographic change on health care costs, Age and Ageing 33 (6), pp.556-561
The World Bank (2007), From red to gray. The “Third Transition” of Ageing Populations in Eastern Europe
and the Former Soviet Union, (authors: M.Chawla, G.Betcherman, A.Banerji), Washington D.C.
Warshawsky M. (1994), Projections of health care expenditures as a share of GDP: Actuarial and
macroeconomic approaches, Health Services Research 29, 293-313
Warshawsky M.(1999), An enhanced macroeconomic approach to long-range projections of health care
and social security expenditures as a share of GDP, Journal of Policy Modeling 21(4), 413-426
Zatonski W. A., Willert W. (2005), Changes in dietary fat and declining coronary heart disease in
Poland: population based study, BMJ Volume 331
Zweifel P., Felder S., Meiers M. (1999), Ageing of population and health care expenditure: a red
herring?, Health Economics nr 8, pp.485-496
69

CASE Network Report 77 - Development of Scenarios for Health Expenditure in the New EU Member States: Bulgaria, Estonia, Hungary, Poland and Slovakia

More Related Content

What's hot (19)

Similar to CASE Network Report 77 - Development of Scenarios for Health Expenditure in the New EU Member States: Bulgaria, Estonia, Hungary, Poland and Slovakia (20)

More from CASE Center for Social and Economic Research (20)

Recently uploaded (20)

CASE Network Report 77 - Development of Scenarios for Health Expenditure in the New EU Member States: Bulgaria, Estonia, Hungary, Poland and Slovakia