CJFA_2_2017_DRUK.pdf e- 65 -1 Date of submission: August 8, 2017; date of acceptance: September 18, 2017. * - - ** - pernican Journal of Finance & Accounting, 6(2), 33–43. http://dx.doi.org/10.12775/CJFA.2017.009 * ** IN EU 12 AND V4 STATES Keywords: convergence, European Union, V4 states, government expenditure, health- care. J E L Classification: F0, I1. Abstract: The subject of this paper is an attempt to assess the level of convergence of government health spending in the European Union countries. The ever-increasing le- vel of life expectancy, demographic changes, including the aging of society in many Eu- ropean countries, and the development of medical technologies are a growing challen- ge for modern nations. The similarities in the changes taking place in the EU states lead to considering whether there are similar transformations in health spending? Ensuring access to primary healthcare is one of the tasks of the welfare state and it can be assu- med to be a public good. Accepting these reasons, the authors set out to investigate the phenomenon of sigma and beta unconditional convergence for the level of government spending on health care in the EU countries. The answer to the question has been so- ught: does participation in the EU leads to convergence in government spending on he- 34 alth care? The study uses selected dispersion measures for sigma convergence calcula- tions and econometric modelling for beta convergence. Health care expenditure is one of the most important types of state expend- iture. Due to the prolongation of life expectancy, the aging process of Euro- pean societies, the use of modern medical technologies, their importance is constantly growing. The similarity in these phenomena makes it possible to suppose that changes in health policy, especially in health care expenditure, will show some form of convergence in EU countries. Previous research results show the disproportion in this spending in EU countries, and the presented re- sults of research on convergence are not clear. Due to the spatial and temporal extent of analyses, the presentations of many authors’ conclusions differ. The authors sought to answer the question whether the presence of states in the European Union leads to convergence in government spending on health care. The discussed issue may be useful for further ref lections on the issue of inte- gration in the European Union. The dependent variable in the study was the level of government spending on health care to GDP, as a result of the disaggre- gation of total government spending on individual sub-spending. The analysis uses statistical and econometric methods. The method of convergence analysis is known and applied to a wide spectrum of research not only in economic sciences, especially with regard to changes in economic growth. Initially, convergence analysis was mainly used to study the occurrence of real convergence between countries or regions. For this pur- pose, the main variable was real GDP per capita. Such research has been con- ducted by many authors for different groups of states. Over time, the idea of convergence began to be used to assess the transformations of other areas of refer only to GDP per capita analysis. An example of this is the use of conver- gence in the analysis of health care expenditure in the EU countries (Hnaty- szyn-Dzikowska & Wyszkowska, 2015, p. 127–135), in the analysis of govern- CONVERGENCE OF HEALTH EXPENDITURE IN EU 12 AND V4 STATES 35 2016, pp. 247–255), when analysing expenditures on the agricultural sector - often it is possible to find studies using this method in research on the size of state expenditure and their spatial convergence. Such research are also oc- curring as regards the convergence of national health spending, although their numbers are relatively small. Referring to this area of convergence assessment, on the one hand, in most countries the aging process of societies is observed, albeit with varying dynamics, which is linked to the growing demand for care for the elderly. The existence of a European model of the welfare state (though in various forms) leads to posing a hypothesis of convergence in the amount of health care expenditure. On the other hand, European societies differ in their social and family structures, as well as in the characteristics of welfare states. (Zaidi et al., 2017, p. 139). It is also pointed out that unequal patterns of health care expenditure, which are a clear feature of late capitalist society (Lau, Fung & Pugalis, 2014, p. 137). The measurement of convergence of health expendi- tures alone, as suggested by Strzelecka (2011, p. 214), is justified by the provi- sion of information on the degree and convergence of countries in the case of budget allocations to health care. Due to the small number of studies conduct- ed in this field, there are no clear results indicating the convergence (or lack thereof ) of health expenditure. Additional diagnostic problems and the compa- rability of the results are based on the method adopted by the authors of vari- ous papers, the time range included in the econometric models and finally the selection of states. Each of these factors can affect the final outcome of conver- gence indicators. Long-term studies demonstrated the convergence of health expenditure (sigma and beta) for 1960–1995 for the EU15 countries (Nixon, 1999, pp. 1–37; Nixon, 2000, pp. 1–27). In the Hitiris study (1997, pp. 1–6), how- ever, the occurrence of such convergence for the period 1960–1991 for the EU- 12 countries was undermined. An analysis of the various EU Member States in the period 1992–2004 confirmed the existence of convergence, although the tendency for convergence was not maintained throughout the analysed time range. The main period of convergence was 2002–2004 for sigma convergence in the EU 15. (Kerem, Püss, Viies & Maldre 2008, pp. 29–43). The studies on the convergence of health expenditures were also made for 19 OECD countries (1972–2006) where convergence was demonstrated, although differences in US and Norway were indicated (Panopoulou & Pantelidis, 2012, pp. 3909–3920). Relatively new research on the convergence of health expenditure were also 36 conducted, among the others, in Africa (Odhiambo, Wambug & Kiriti-Ng'ang'a, 2015, pp. 185–205) or in China (Zhang, Zhang, Wu, Xia & Lu, 2016, pp. 1–11). Obviously, each country has the opportunity to develop its own health care management system, although similar historical conditions and health policy goals caused that these systems have some common features (Kujawska, 2015, p. 113). Despite those similarities, particular countries responded different- ly to external shocks, including the financial crisis that affected health care spending. These shocks were therefore asymmetrical. Some countries also used the crisis to reduce their spending level, for example through its abso- lute reduction (Cyprus, Greece, Ireland, Lithuania, Portugal, Romania) or freez- ing salaries of healthcare workers (Great Britain, Slovenia) or lowering salaries (e.g. Denmark) (Karanikolos, Mladovsky et al., pp. 1323–1331). The adopted study interval took into account the period of crisis that could have had an im- pact on disparities in health care expenditure. In the analysed group of coun- tries in the time period from 2004 to 2015, the expenditure on health care was clearly differentiated. Relatively low values were observed in Luxembourg, in Poland and the Czech Republic, while the highest level of expenditure was ob- served in Denmark. The aim of the study is to investigate the phenomenon of sigma and beta un- conditional (absolute) convergence of the level of government spending on healthcare for the EU12 (Old EU) and Visegrad Group (V4) countries. For this purpose, two analyses were conducted – sigma and beta convergence for 16 countries in the period 2002–2015. The research material came from the Eurostat database. GRETL 2016d program was used for the estimation of re- gression parameters. Discussing the convergence considerations, a number of methods must be taken into account to confirm its existence. In the literature we can find differ- - beta analyses, which most often appear in classical literature (Sala-i-Martin, 1995, p. 3). Considering the accepted purpose of the paper, the authors analysed the convergence of health expenditure using the sigma convergence analysis and the beat unconditional (absolute) convergence. For this purpose, Eurostat CONVERGENCE OF HEALTH EXPENDITURE IN EU 12 AND V4 STATES 37 databases were used and decomposition of total government expenditure on individual sub-spending was made, on the basis of which both analyses were carried out. Such an approach to obtaining particular state spending was also widely used in a number of studies, for example, selected components of gov- by initially disaggregating total government expenditure for government ad- ministration expenditure, education, health, agriculture, construction, public transport and communications, and government spending on social security. Conducting the sigma convergence analysis, a standard deviation of the loga- rithm of natural government expenditures on health care was used. The ana- lysed equation for each time unit tadopted the form (Malaga, 2004, p. 57)1: (1) where: i – country index, HE it – the level of government spending on health care in the country i and at time t, – the average level of government spending on health care in the con- sidered group of countries. When there is a decrease in the dispersion of the examined feature over time, we refer to the occurrence of sigma convergence (table 1). It is also con- venient to determine the rate of change of the sigma parameter under study or to make a visual analysis of the distribution of sigma convergence points in the 2014, p. 99): t 0 1t, (2) where: t– standard deviation of the natural logarithm of the examined feature be- tween countries at time t. 1 - ferent subject of the study. 38 1 signals sigma convergence occurrence (table 2). The second common measure of convergence is beta convergence. We can distinguish beta absolute (unconditional) convergence and beta conditional convergence. Beta absolute convergence occurs when poorer economies devel- op faster than the richer ones (Sala-i-Martin, 1995, p. 3). In the case under con- sideration, countries with lower levels of health expenditures increase them faster than those with relatively high levels of health expenditure. Then the dif- ferent countries are striving for the same steady state of long-term equilibri- um. We assume that all economies, at the same percentage, reduce the distance to long-term equilibrium and at the same time reach that state. In turn, the con- ditional convergence beta analyses the phenomenon when a country's economy converges to its own steady-state equilibrium (Malaga, 2004, p. 66). The coeffi- in the analysed group of states. For this purpose, an estimate of the regression 2: (3) where: HE r – level of expenditure on health in the final period, HE 0 – level of expenditure on health in the end period. 1 the formula: (4) the less developed economies (or the relative weakness of the surveyed fea- ture) will develop rapidly enough to surpass the more developed economy 2 the ones accepted by the authors due to another subject of the study. CONVERGENCE OF HEALTH EXPENDITURE IN EU 12 AND V4 STATES 39 - - In the analysed time range for the 12 EU countries and the V4 group, there is a clear divergence of health care expenditure. In the period in study there was a continuous increase in the divergence of the examined feature (table 1). The same conclusions can be drawn by analysing a regression equation, in which 1 has a statistically significant higher value than zero. Conse- quently, individual countries, although operating within one integration group, did not show one path of adjustment to public health spending. This process has been systematically increasing disproportion over time since 2005, due to the substantial enlargement of the EU by 10 new countries. This can be represent- ed as a time-dependent regression function (table 2). Table 1. Measuring the sigma convergence of health care spending in the EU12 states and V4 group Research period Sigma convergence Change rate 2002 0.568314 – 2003 0.580515 1.021469 2004 0.599286 1.032335 2005 0.51917 0.866315 2006 0.52827 1.017528 2007 0.622461 1.178301 2008 0.623386 1.001485 2009 0.662759 1.063161 2010 0.676734 1.021086 2011 0.70729 1.045153 2012 0.74164 1.048565 2013 0.752369 1.014466 2014 0.825579 1.097306 2015 0.807344 0.977913 S o u r c e : own study on Eurostat data. 40 Table 2. Regression function for sigma analysis Coefficient Standard error t-Student p value R2 Const 0.4961 0.0223719 22.1751 <0.0001*** 0.849408 Time 0.0216163 0.00262745 8.2271 <0.0001*** S o u r c e : own study based on Eurostat data using GRETL 2016d program. Table 3. Measurement of beta absolute convergence for health expenditure in EU12 states and V4 group Research period Estimated values R2 2002–2006 0.078036 -0.0369546 1.811 -1.526 0.0916* 0.1493 0.142587 0.040858 2002–2007 0.030116 -0.0124092 0.6959 -0.5101 0.4979 0.6179 0.018247 0.012895 2002–2008 0.054976 -0.0234355 1.2491 -0.9472 0.2321 0.3596 0.060225 0.025598 2002–2009 0.058022 -0.0198219 1.3361 -0.812 0.2028 0.4304 0.044973 0.021582 2002–2010 0.045125 -0.0151188 1.1424 -0.6809 0.2724 0.5071 0.032053 0.016251 2002–2011 0.029286 -0.00851732 0.7935 -0.4105 0.4408 0.6877 0.011892 0.008902 2002–2012 0.027027 -0.0080678 0.7178 -0.3812 0.4847 0.7088 0.010272 0.008448 2002–2013 0.031181 -0.0114644 0.8302 -0.543 0.4203 0.5957 0.020626 0.012334 2002–2014 0.019065 -0.00554119 0.4891 -0.2529 0.6324 0.8041 0.004546 0.005751 2002–2015 0.023336 -0.00909286 0.6334 -0.439 0.5367 0.6673 0.01358 0.009726 S o u r c e : own study based on Eurostat data using GRETL 2016d program. - tive, which would indicate a convergence of health care expenditure. It should be noted, however, that none of the structural parameters of the model reached satisfactory statistical significance, and the coefficient R2 in each case is low. initial period and amounts to 14.93%, but for this period it is unsatisfactory (the variable can be considered significant if the index does not exceed 10% should be carefully evaluated, especially with such unambiguous divergence CONVERGENCE OF HEALTH EXPENDITURE IN EU 12 AND V4 STATES 41 of beta convergence. The analysis of public health expenditure has shown a varied course of these paths over time. Due to the differences between the social and family struc- tures as well as the specific characteristics of welfare states, convergence in the expenditure group under consideration cannot be stated. The conducted sigma convergence analysis negated the existence of health care spending f luc- tuations for EU12 and V4 countries. The assessment of this convergence is in- dicative of a reversal phenomenon – divergence. There was disagreement in the assessment between beta and sigma convergence (although in the first case the necessary significance was not achieved). It seems that in the case of this sec- ond convergence, weaknesses in this approach arose, indicating that the rela- tionship between the rate of change in health care expenditure and its initial level does not imply a reduction in the dispersion structure and there may even the high variability between countries in the rate of change (effect of extreme magnitude). On the other hand, sigma divergence was identified, and therefore the increase in discrepancies over time. This may indicate the phenomenon of polarization of these expenses within the EU states. wsparcia sektora rolnego? Zeszyty Naukowe SGGW w Warszawie – Problemy Rol- Dada, M.A. (2013). Composition Effects of Government Expenditure on Private Con- sumption and Output Growth in Nigeria: a Single-Equation Error Correction Model- ling. Romanian Journal of Fiscal Policy, Volume 4, Issue 2(7). - y, nr 24. Hitiris, T. (1997). Health care expenditure and integration in the countries of the Euro- pean Union. Applied Economics, 29(1). - http://appsso.eurostat.ec.europa.eu/nui/submitViewTableAction.do (accessed: 10.05.2017). 42 Tom 6(42). Karanikolos, M., Mladovsky, P., Cylus J., Thomson, S., Basu, S., Stuckler, D., Macken- Lancet, 383(9874), http://dx.doi.org/10.1016/S0140-6736(13)60102-6. Kerem, K., Püss, T., Viies, M., & Maldre, R. (2008). Health And Convergence Of Health Care Expenditure In EU. International Business & Economics Research Journal, 7(3). - dx.doi.org/10.7172/1644-9584.53.7. - - nomicznych Uniwersytetu Warszawskiego, Warszawa. Lau, C.K., Fung, K. W., & Pugalis, L. (2014). Is healthcare expenditure across Europe con- verging? Findings from the application of a nonlinear panel unit root test. Eurasian Business Review, 4(2). http://dx.doi.org/ 10.1007/s40821-014-0014-9. Malaga, K. (2004), Konwergencja gospodarcza w krajach OECD w latach 1960–1999. - wo Akademii Ekonomicznej w Poznaniu, Zeszyt 41. Nixon, J. (1999). Convergence Analysis of Health Care Expenditure in the EU Countries Using Two Approaches. Department of Economics and Related Studies, University of York, Heslington, Discussion Papers in Economics, No 1999/03. Nixon, J. (2000). Convergence of Health Care Spending and Health Outcomes in the Eu- ropean Union, 1960–95. Centre for Health Economics, Discussion Paper: The Uni- versity of York, No. 183. Odhiambo, S. A., Wambugu A., & Kiriti-Ng’ang’a, T. (2015). Convergence of Health Ex- penditure in Sub-Saharan Africa: Evidence from a Dynamic Panel. Journal of Eco- nomics and Sustainable Development, 6(6). Panopoulou, E., & Pantelidis, T. (2012). Convergence in Per Capita Health Expenditures and Health Outcomes in the OECD Countries. Applied Economics, Volume 44. http:// dx.doi.org/10.1080/00036846.2011.583222. empirycznych. Zeszyty Naukowe SGH, nr 20. - istycznych w latach 1990–2005. Bank i Kredyt, nr 8–9. Sala-i-Martin, X. (1996). The Classical Approach to Convergence Analysis. Economic Journal, 106(437). krajach w latach 1990–2008. Acta Universitatis Nicolai Copernici, Ekonomia XLII – CONVERGENCE OF HEALTH EXPENDITURE IN EU 12 AND V4 STATES 43 Wan, G. H. (2005). Convergence in Food Consumption in Rural China: Evidence from Household Survey Data. China Economic Review, 16(1). http://dx.doi.org/10.1016/j. chieco.2004.09.002. Zaidi, A., Gasior, K., Zolyomi, E., Schmidt, A., Rodrigues, R., & Marin, B. (2017). Measur- ing active and healthy ageing in Europe. Journal of European Social Policy, 27(2). http://dx.doi.org/10.1177/0958928716676550. Zhang, G., Zhang, L., Wu, S., Xia, X., & Lu, L. (2016). The convergence of Chinese county government health expenditures: capitation and contribution. BMC Health Services Research. http://dx.doi.org/10.1186/s12913-016-1635-8.