ESTUDIOS SOBRE LA ECONOMÍA ESPAÑOLA

Decentralisation and health care outcomes: An empirical analysis within the European Union David Cantarero Prieto Marta Pascual Saez

EEE 220

March 2006

ISSN 1696-6384

Las opiniones contenidas en los Documentos de la Serie EEE, reflejan exclusivamente las de los autores y no necesariamente las de FEDEA.

The opinions in the EEE Series are the responsibility of the authors an therefore, do not necessarily coincide with those of the FEDEA.

DECENTRALISATION AND HEALTH CARE OUTCOMES: AN EMPIRICAL

ANALYSIS WITHIN THE EUROPEAN UNION

David Cantarero and Marta Pascual

Department of Economics. University of Cantabria (Spain)

Corresponding authors

David Cantarero. Department of Economics. Facultad de CCEE y EE. University of

Cantabria. Avda de los Castros s/n. Santander (Cantabria) 39005. SPAIN. Tel: 34-942-

201625. Fax: 34-942-201603. E-mail: david.cantarero@unican.es

Marta Pascual. Department of Economics. Facultad de CCEE y EE. University of

Cantabria. Avda de los Castros s/n. Santander (Cantabria) 39005. SPAIN. Tel: 34-942-

201628. Fax: 34-942-201603. E-mail: marta.pascual@unican.es

Abstract

The aim of this paper is to explore the impact of decentralisation on health care outcomes in the European Union. We investigate the hypothesis that shifts towards greater decentralisation would be accompanied by improvements in population health by using infant mortality and life expectancy as dependent variables. The results of the empirical analysis suggest that income, decentralisation, health care resources and lifestyles in European Union did have an influence on infant mortality and life expectancy. This paper adds a new empirical perspective to the evaluation of the economic gains arising from greater decentralisation in health care.

JEL Codes: I12, H77

Key words: Fiscal Federalism, decentralisation, health care outcomes, infant mortality, life expectancy, European Union.

Introduction

Health and its measurement is a multidimensional phenomenon, so it should be explained through multiple indicators. One of them could be decentralisation of health care resources. In this sense and as maintained by the fiscal federalism theory, fiscal decentralisation of the public sector is likely to produce allocative and productive efficiency gains (Oates, 1999). In health care, a trend towards decentralisation is becoming evident in most European Union countries (Banting and Corbett, 2002) (for example, federal countries as Belgium or quasi-federations such as the United Kingdom, Spain or Italy).

The main objective of this paper is to analyse the effects of different socio-economic factors and decentralisation on health care outcomes. We will focus on European Union countries from 1990 to 2003. The results of the empirical analysis suggest that income, decentralisation, health care resources and lifestyle in European Union did have an influence on the effectiveness of public policies on infant mortality and life expectancy.

This paper is organized as follows. Section 2 describes health care decentralisation from the perspective of fiscal federalism literature. In Section 3, we present the model and methodology that we use to test the influence of different socio-economic factors and health care decentralisation on health care outcomes (infant mortality and life expectancy). The results from the empirical study are summarized in Section 4. Finally, the concluding remarks are provided in Section 5.

Decentralisation of health care: Theory and evidence

Health care is an example of goods of a mixed nature among local and national public goods. Thus, externalities in health care do not necessarily imply centralised provision as a superior alternative, since there might still be welfare gains from decentralised provision relative to a centrally determined level of health care services. Providing local governments with subsidies may encourage efficient levels of health services to the point where the marginal social benefits for society as a whole from the provision of health care equals marginal costs.

On the other hand, economies of scale may need central intervention to prevent inefficient location of facilities such as hospitals by local decision makers accountable to local electors. Another argument frequently adduced for central intervention in health care is the more efficient pricing of inputs by a single purchaser of health care (Jiménez and Smith, 2005)).

Following the decentralisation theorem, the main argument for decentralisation in health care is that local decision maker have greater knowledge of their population health needs than national policy makers. However, although decentralisation can result in greater total health gains, it may also lead to increased inequalities in access or health care expenditure and not in health outcomes as compared with a centralised solution.

Nevertheless, there is little evidence that countries with a more decentralised health care system have better health outcomes. On the whole these studies find a positive association between decentralization and some indicators of health outcomes.

Mahal et al. (2000) use data from rural villages in India for 1994 to test the hypothesis that decentralisation is positively associated with infant mortality once the effect of socio-economic factors, civil society organisations, and so on, are accounted. While the estimated coefficients for decentralised states have the expected positive signs, the election frequency variable is statistically insignificant. Asfaw et al. (2004) corroborates these previous results using an index of fiscal decentralisation obtained by factor analysis on the basis of three variables for the period 1990-1997. Their results also show that the effectiveness of fiscal decentralisation increases with the level of political decentralisation.

In Robalino et al. (2001), a panel data of low and high income countries is used to test how a measure of fiscal decentralisation (the proportion of sub national government spending over central government spending), affects infant mortality rates over the period 1970-1995. After controlling by a set of structural variables, one of the main results of the fixed effects estimation is that decentralisation is associated with lower infant mortality rates.

Yee (2001) uses a panel data of 29 Chinese provinces for the period 1980-1993 in order to examine the relationship between several indicators of health care perfomance and various measures of decentralisation. The results shows that decentralisation has been beneficial to the health sector in terms of reducing mortality rates and increasing local health expenditure.

Habibi et al. (2001) shows that the percentage of revenue raised locally and the proportion of controlled revenue over the total have a negative and significant association with infant mortality rates for a panel data of Argentinean provinces over the period 1970-1994. In addition, they find that decentralisation reforms lead to a significant reduction in regional inequalities.

Finally, Jimenez and Smith (2005) use ten provinces of Canada as a case study and the results of the empirical analysis suggest that decentralisation in Canada did have a positive and substantial influence on the effectiveness of public policy in improving the population´s health.

A model of decentralisation of health care: Data and methodology

This empirical study evaluates, after controlling by a set of socio-economic factors, the relationship between decentralisation and health outcomes as a whole on health perfomance.

To examine the model that we propose, we use a panel of the fifteen countries of the European Union for the period 1990-2003. We have used infant mortality rates and life expectancy from the OECD Health Data (2004) as measures of health status. Infant mortality has been considered as the most important indicator of health outomes in a society and it has been used in different and previous studies on this issue (Mahal et al., 2000; Robalino et al., 2001; Habibi et al., 2001; Jiménez and Smith, 2005). It reflects infant health and pregnant women´s health, in addition to the state of health development within the society. Moreover, infant mortality could be better health indicator than life expectancy, another alternative measure of health status, for two main reasons. Firstly, because infant mortality is more reliably measured than life expectancy and is based on actual data. Secondly, because disparities in the risk of infant death are higher than disparities in life expectancy in European Union over this period. Nevertheless, we have used life expectancy as an alternative health measure and in order to test our previous results.

Figure 1 shows the scatter plot of infant mortality against per capita income (GDP per capita-$ Purchasing Power Parity) in the European Union countries from 1990 to 2003. Log specification in infant mortality and real income is used because inspection of scatter plots suggests an approximately log-linear relationship between GDP per capita and infant mortality.

Take in Figure 1

In Figure 2 the scatter plot of life expectancy against per capita income (GDP per capita-$ Purchasing Power Parity) in the European Union countries from 1990 to 2003 is showed. Again we will consider log specification in life expectancy and real income.

Take in Figure 2

The first socio-economic determinant of infant mortality and life expectancy could be income. In this sense, GDP may also be inversely related to key health indicators like infant mortality and positively related to life expectancy (Kanavos and Mossialos, 1996). Nevertheless, this possible relationship fails to explain why infant mortality in a country like the United States (one of the wealthiest countries in the world in terms of per capita GDP) is higher than in other OECD countries with similar or even lower per capita income levels (Starfield, 2000). So, there are additional variables, other than GDP, which affect and explain health indicators like infant mortality and life expectancy.

Secondly, a precise measure of health care decentralisation is difficult to find. Fiscal decentralisation applied to health care reflects how responsibilities for tax revenues and public expenditures are distributed among different tiers of government. The complexity of vertical government structure makes this notion difficult to quantify. A reliable measure of fiscal decentralisation needs to effectively quantify the activities of subcentral governments arising from their autonomous decisions. So, the standard approach in international analysis is to make use of accounting measures of revenue and expenditure shares for sub-central government relative to general government as a proxy for fiscal decentralisation.

Up to now the most reliable quantitative measure of health care decentralisation is a fiscal one: the ratio of sub national health care spending to the total health spending for all the levels of government. The main source of the fiscal data is the International Monetary Fund´s (IMF) Government Finance Statistics (GFS).

Global fiscal decentralisation indicators for health care spending can be computed net of intergovernmental transfers although some authors have pointed out that fiscal decentralisation indicators based on IMF data may overestimate real sub national autonomy and our results could be biased (Ebel and Yilmaz, 2003; Fiva, 2005). We have not used revenue measures of decentralisation constructed on the basis of IMF data because they have different problems and sub national revenue autonomy could be overestimated.

Also, health care resources and lifestyle (Contoyannis and Jones, 2004) could explain the relationship between socio-economic characteristics and health indicators like infant mortality and life expectancy.

Our results are based on the following general model:

\[H = H (E, H C, L),\tag{1}\]

where H denotes some health indicator (in particular we have considered infant mortality and life expectancy); E represents a vector of economic references; HC, represents health care resources and L denotes lifestyle and behaviour.

In particular, we have considered the following variables in log terms (see Table I) for the European Union countries since 1990 to 2003: Gross Domestic Product (per capita US$ Purchasing Power Parity), decentralisation of health care expenditure, acute care beds (density per 1000 population), general practitioners (density per 1000 population)

and alcohol consumption (litres per capita). We have used these health and economic indicators taken from the Organization for Economic Development and Cooperation (OECD) Health Data and Government Finance Statistics of the International Monetary Fund (IMF). It allows for the comparison and the analysis of international health care and fiscal federalism systems. Results of summary statistics are shown in Table II.

Take in Tables I and II

Results

In this section, estimates of the determinants of infant mortality and life expectancy in the European Union countries are presented. We provide results derived from the estimation using STATA 8.0. For the econometric estimation, the standard panel technique was used (Jones, 2000; Greene, 2003) and the fundamental advantage of this panel data set over a cross section is that it allows us great flexibility in modelling differences across European countries and the attenuation of the problem of omitted variables. Panel data models allow to control for individual heterogeneity (Greene, 2003; Baltagi, 1995). Fixed effects and random effects are the most usual panel data methods.

The basic framework is a regression model of the form:

\[I N F A N T M \log_ {i t} = \alpha_ {i} + \beta_ {1} G D P \log_ {i t} + \beta_ {2} D E C E X P _ {i t} + \beta_ {3} A C B \log_ {i t} + \beta_ {4} G P \log_ {i t} + \beta_ {5} A L C \log_ {i t} + \varepsilon_ {i t}\tag{2}\]

\[L I F E X \log_ {i t} = \alpha_ {i} + \beta_ {1} G D P \log_ {i t} + \beta_ {2} D E C E X P _ {i t} + \beta_ {3} A C B \log_ {i t} + \beta_ {4} G P \log_ {i t} + \beta_ {5} A L C \log_ {i t} + \varepsilon_ {i t}\tag{3}\]

where i refers to the country (i=1,..., 15 member states) and t is the year 2003).

Also, we have used Hausman’s specification test for the random effects model. This specification, which was devised by Hausman (1978), is used to test for orthogonality of the random effects and the regressors. Finally, a F test is included to evaluate the joint significance of the variables. We can use the fixed-effects approach or the randomeffects approach. The Hausman test value shows that the first one should be used. The results of the estimation are given in Tables III and IV.

Take in Tables III and IV

It is very interesting to point out that the results obtained considering European Union-15 countries show that GDP per capita, health care decentralisation and the number of general practitioners per 1000 population are negatively related to infant mortality. On the other hand, acute care beds per 1000 population and alcohol consumption are positively related to infant mortality.

Also, GDP per capita, health care decentralisation and number of general practitioners per 1000 population are positively related to life expectancy. On the other hand, acute care beds per 1000 population and alcohol consumption are negatively related to life expectancy.

Finally, the level of explanation, as measured by , is acceptable, signs of variables are those to be expected and their statistical significance is accepted.

Conclusions

The theoretical literature on fiscal federalism applied to health care services predicts potential efficiency gains from placing responsibilities of local public goods at the local level, these being manifested in an improvement in the population´s health. However, in the empirical literature few attention has been paid to the evaluation of the outcomes of decentralisation in health care public sector.

In this paper, we have explored the relationship between, among other socio-economic factors, the most common measure of health care decentralisation –the proportion of local health spending on the total health spending for all the levels of government- and two traditional indicators of health outcomes –infant mortality and life expectancy- in European Union Countries. To formalise the linkages between health care decentralisation and health outcomes, we have developed a simple model that we have then estimated on the basis of panel data for the period 1990-2003.

The results of the econometric estimations for European Union Countires suggest that infant mortality is negatively related to GDP per capita, health care decentralisation and the relative number of general practitioners. Also, infant mortality is positively related to the relative number of acute care beds and alcohol consumption. Decentralisation in these countries is associated with the effectiveness of public policy on improving population´s health (in terms of lower infant mortality rates).

On the other hand, results obtained considering European Union-15 countries show that GDP per capita, health care decentralisation and the number of general practitioners per 1000 population are positively related to life expectancy. Finally, acute care beds per 1000 population and alcohol consumption are negatively related to life expectancy. In this case, decentralisation is associated with greater life expectancy.

These results should be taken into account in order to make adequate health care policies in the European Union countries. However, some caution is required on interpreting these results. First of all, the indicator of health care decentralisation used captures only one of the multiple dimensions of the health care decentralisation process (the fiscal one). Secondly, the two measures of health outcomes employed do not fully reflect the underlying level of health in a society. In spite of these problems, this paper adds a new empirical perspective to the evaluation of the economic gains arising from greater decentralisation in health care.

References

Asfaw, A., Frohberg, K., James, K.S., Juting, J. (2004), “Modelling the impact of fiscal

decentralisation on health outcomes: empirical evidence from India”, ZEF Discussion

Paper 87, Bonn.

Baltagi, B. (1995), Econometric Analysis of Panel Data. New York: Ed. Wiley.

Banting, K.G., Corbett, S. (2002), Health Policy and Federalism. A Comparative

Perspective on Multi-Level Governance. Kingston, Ont.: Ed. Queen´s University.

Contoyannis, P. & Jones, A. (2004), “Socio-economic status, health and lifestyle”,

Journal of Health Economics, 23, pp. 965-995.

Ebel, R.D., Yilmaz, S. (2003), “On the measurement and impact of fiscal

decentralization”, In Martinez-Vazquez, J. and Alm, J. (eds.). Public finance in

developing and transitional countries, Cheltenham, UK: Edward Elgar.

Fiva, J.H. (2005), “New evidence on fiscal decentralization and the size of

government”, CESifo Working Papers, Nº 1615, December 2005.

Greene, W.H. (2003), Econometric Analysis. New York: 5th Edition, Prentice Hall,.

Habibi, N., Huang, C., Miranda, D., Murillo, V., Ranis, G., Sarkar, M., Stewart, F.

(2001), “Decentralization in Argentina”, Center Discussion Paper No. 825, Economic

Growth Centre, Yale University.

Hausman, J.A. (1978), “Specification tests in econometrics”, Econometrica, 46, pp.

1013-1029.

Jimenez, D., Smith, P.C. (2005), “Decentralisation of health care and its impact on

health outcomes”, Discussion Papers in Economics No. 2005/10, The University of

York.

Jones, A.M. (2000), Health Econometrics, In Culyer, A.J. and Newhouse, J.P. (eds.):

Handbook of Health Economics, Elsevier, Amsterdan.

Kanavos, P. & Mossialos, E. (1996), “The methodology of international comparisons of

health care expenditures: any lessons for health policy?” LSE Health Discussion Paper,

No. 3, London.

Mahal, A., Srivastava, V., Sanan, D. (2000), Decentralization and its impact on public

service provsion on health and education sectors: the case of India. In: Dethier, J. (eds.):

Governance, Decentralization and Reform in China, India and Russia, London: Kluwer

Academic Pubishers and ZEF.

Oates, W.E. (1999), “An Essay on Fiscal Federalism”, Journal of Economic Literature, 37, pp. 1120-49.

Robalino, D.A., Picazo, O.F., Voetberg, A. (2001), “Does fiscal decentralization improve health outcomes? Evidence from a cross-county analysis”, Policy Research Working Paper 2565, World Bank, Washington D.C.

Starfield, B. (2000), “Is US health really the best in the world?”, Journal of American

Medical Association, Vol. 284, pp. 483-485.

Yee. E. (2001), The Effects of Fiscal Decentralisation on Health Care in China. Princeton University.

FIGURES AND TABLES

Figure 1 Infant mortality and GDP per capita ($ PPP). European Union countries (1990-2003) Source of data: OECD Health Data (2004).

Life expectancy and GDP per capita ($ PPP). European Union countries (1990-2003) Source of data: OECD Health Data (2004). Figure 2

Table I Variables and definitions

(*) It has been calculated taking into account Purchasing Power Parity (PPP). Source: Authors’ elaboration.

Table II Summary Statistics of selected variables used in estimations

(*) It has been calculated taking into account Purchasing Power Parity (PPP). Source: Authors´ calculations from ECHP and IMF.

Table III Results. Panel Data approach. Dependent variable: Infant Mortality

Source: Authors´ calculations from OECD Health Data and Government Finance Statistics, IMF.

Table IV Results. Panel Data approach. Dependent variable: Life Expectancy

Source: Authors´ calculations from OECD Health Data and Government Finance Statistics, IMF.