GDP AND PUBLIC EXPENDITURE ON HEALTH-CARE: THE CASE OF GREECE

Transcription

1 GDP AND PUBLIC EXPENDITURE ON HEALTH-CARE: THE CASE OF GREECE Nicolaos Dritsakis 1 Ekaterina Sarri 2 Abstract In this paper we examine the time trend of public expenditure on health-care in Greece and the relationship between Gross Domestic Product (G.D.P) and Public Expenditure, for the decades , , and Expenditures on health have been classified in three categories: Expenditures for hospital-care, for non-hospital care and for medication. Wei nvestigated the evolution of each one of the above categories, by estimating their average annual increase. The year 1980 was used as base year. Finally the effect of the income elasticity upon the expenditures for health is estimated. 1 Department of Applied Informatics, University of Macedonia, Thessaloniki, (Greece). 2 Business Department, American College of Thessaloniki, Thessaloniki, (Greece).

2 Introduction It has been observed that in all European countries public expenditure on health-care has been increasing over time by more than their GDP. This happens because the increase in GDP of the various countries is accompanied by the increase of the citizens desire and demand for better quality of health-care services. Although the aim of each European state is to improve the health of its citizens, different policies have been exercised by different states. Not all policies have succeeded. As the literature suggests, even countries like Great Britain have not managed to attain their aim despites all the reform and re-organization attempts at the National Health System (N.H.S). (Cooper and Culyer 1970, Maynard 1975 and 1977, West 1981, Steele 1981, Healy,Yule and Reid 1994). The Theoretical Model In order to investigate the time trend of public expenditure on health, we estimated the average rate of increase of the three categories of expenditures for health and the income elasticity by using the functions of the public expenditure and of the GDP. For estimation purposes the exponential functions are logarithmically transformed in order to become linear is used as base year. 2

3 The function of the expenditures for health: The function of public expenditure on health is: Y t = Y 0 e ët The logarithmic transformation is: lny t = lny 0 + ët where: Y t = Public expenditure for health. Y 0 = Expenditures for health in the starting period. ë = The requested trend of the expenditure for health on an annual base t = Values for the period : , , The function of the public expenditure for health and gross domestic product: Y t = a + bgdp t + U the logarithmic transformation is: lny t = lna + blngdp t + lnu 3

4 where: Y t = Public expenditure for health. GDP t = Gross domestic product. a, b = parameters to be estimated. U = stochastic variable. From the above equation we estimated the income elasticity (b) which shows us the percent (%) effect of the economic growth on the expenditure on health. This means that if GDP increases by 10% then the elasticity of income shows us the corresponding expected increase of the expenditure for health. The elasticity s mathematic expression is: e = DY Y DGDP GDP = % change exp enditureforhealth % changegdp When e = 1, then if GDP increases, the expenditures on health also increase by the same proportion. When e > 1, the expenditures on health increase more rapidly than GDP. This happens because economic development creates additional social needs. When e < 1, the expenditure for health does not increase in direct proportion to increases in GDP. This is mainly because economic policy has been shifted towards other aims (than health). 4

5 Time trends of the public expenditure for health in Greece. Due to the inconsistencies of the different statistical resources with regard to the three categories of public expenditure on health in Greece, our research was restricted to a macroeconomic-level investigation. A number of statistical models were used in order to estimate the time trends and the distribution of the public expenditure on health in Greece. Table 1 presents results obtained through the use of the equation lny t = lny 0 + ët which was estimated using the OLS method. where: Y t = variable (expenditure for health) t = time Y 0, ë = parameters to be estimated. These parameters show the average rate of change on an annual basis. The numbers in the parentheses and under the estimated coefficients are the so called t - ratios. The empirical results in table 1 suggest the following: For the period the highest trend is observed in the expenditures on medication (average annual rate of increase 12,71%). The lowest increase in public expenditure on health is observed in expenditures on hospital care (average annual rate of increase 7,92%). The expenditure on non-hospital care lays somewhere in between expenditures on the other two categories (average annual rate of increase 9,85%). For the period the highest trend is observed in the expenditure on non - hospital care (average annual rate of increase 9,36%). The lowest rate of increase is observed in the expenditure on medication 5

6 (average annual rate of increase 3,13%), while expenditure on hospital care increased by an average annual rate of 6,34%). For the period the highest trend is observed in the expenditure on hospital care (average annual rate of increase 3,77%) followed by expenditures on medication (average annual rate of increase 2,18%), while expenditures on non - hospital care increased by an annual average rate of only 1.11%. TABLE 1 Means and average annual trends of the public expenditure for health. Variable Mean Constant Time Trend R 2 Hospital Care ,1076 9,1932 0,0589 0,9833 (40,5657) ,4975 9,0617 0,0792 0,9927 (33,1143) ,1507 9,1675 0,0634 0,9600 (13,8571) ,6745 9,7124 0,0377 0,8885 (7,9833) Non - hospital Care ,8214 7,5981 0,0789 0,9086 (16,6834) ,9723 7,4302 0,0985 0,9327 (10,5281) ,8953 7,4433 0,0936 0,9745 (17,4829) ,5966 9,8791-0,0111 0,3249 (-1,3941) Medication ,7365 8,8581 0,0566 0,8005 (10,5848) ,0364 8,3374 0,1271 0,9763 6

7 (18,1647) ,0117 9,5264 0,0313 0,8610 (7,2246) ,1613 9,6030 0,0218 0,3956 (2,2886) Total expenditure ,7922 9,8448 0,0611 0,9414 (21,2167) ,1144 9,5743 0,0982 0,9928 (33,4170) , ,0774 0,0253 0,9818 (20,7793) , ,6941 0,0253 0,5639 (3,2165) Time trends of the public expenditure for health in relation to the GDP. This part emphasizes the relationship between expenditures for health and GDP. The effect of economic development upon the public expenditure for health is also reported. A further analysis of the various values of public expenditure on health in relation to GDP (as reported in table 2), suggests that in the first period under consideration ( ) the average annual rate of increase in GDP is 10% while the average annual increase in expenditure on medication is expected to be 17,7%, for hospital - care 11,0% and on non - hospital care 13,8%. In the period expenditures for non - hospital care increase more rapidly than expenditures on hospital care and medication. More specifically, an increase of 10% in the GDP is associated with 19,8% increase in the expenditure on non - hospital care and an increase of 13,1% and 6,7% in the expenditures on hospital care and medication respectively. 7

8 In the last period under consideration ( ) expenditures on hospital care increase more rapidly than expenditure on the other two categories. A 10% increase in the GDP is associated with a 24,2% increase in hospital care expenditure and a 15,9% increase in medication expenditure, but an 11,7% decrease in non - hospital expenditure. Finally, in the last period being examined ( ) total public expenditure on health is higher than in any of the other two periods ( and ). TABLE 2 Estimations of income elasticity for various categories of public expenditure for health. Depended variable Constant term Income elasticity Hospital care ,1289 1,2306 0,9651 (27,8710) ,3927 1,1042 0,9915 (30,6627) ,4562 1,3169 0,9126 (9,1439) ,2073 2,4218 0,8693 (7,2960) Non - hospital care ,7153 1,6805 0,9290 (19,1520) ,6584 1,3816 0,9429 (11,4972) ,2119 1,9893 0,9693 (15,9102) ,5666-1,1782 0,5118 (-0,6569) Medication ,4475 1,2983 0,9472 R 2 8

9 (22,4152) ,8956 1,7747 0,9795 (19,5493) ,4265 0,6783 0,8982 (8,4026) ,7538 1,5910 0,4961 (2,8069) Total expenditure ,6985 1,3202 0,9905 (54,2060) ,3654 1,3704 0,9948 (39,4351) ,2780 1,1465 0,9612 (14,0900) ,8649 1,6112 0,5396 (3,0622) Conclusions Evidence from the above analysis of the time trends in public expenditures on health in relation to GDP supports the following: In the period the highest increase in public expenditure on health was observed in expenditures on medication while in the second period ( ) expenditure on non - hospital care had the highest increase and in the third period ( ) expenditure on hospital care had increased more rapidly than expenditure on any of the other two categories. Correlating the price indexes produces an estimated coefficient between the medication and hospital care indexes of r 1 = 0,8949, between the medication and non - hospital care of r 2 = 0,8036 and between hospital and non - hospital care of r 3 = The above estimated coefficient suggests that there is a close relationship between the price indexes of medication and hospital care as well as between hospital and non - hospital care. 9

10 It should be noted that a further investigation of the substitution or complementary relation between the hospital care, non - hospital care and medication indexes could provide important insights in order to implement a more rational economic policy regarding the expenditure on health-care (Dritsakis 1996). In the last years different European countries developed various policies in order to control public expenditure on health-care. One of the policies implemented was to decrease the number of beds in hospitals and also decrease the in-patient period in an attempt to restrict expenditures on hospital care. In Greece there is a small number of hospital beds in relation to its inhabitants and also the average in-patient period has been decreasing (Seretis 1994). However, the in-patient expenses and the expenses for lab-tests have been increasing dramatically. According to the data, in the last five years of the study, the average in-patient period has decreased from 6 to 5,5 days, and the cost of lab-tests has increased by 16%. The question raised from the above results is whether the increase in the public expenditure on health in Greece will result in an increase in the productivity of the health services provided or it will only increase prices without any change in the quantity and quality of the services offered. 10

11 References Cooper, M. and Culyer, A. (1970). An Economic Assessment of Some Aspects of the operation of the National Health Service. Health Service Financing British Medical Association pp Dritsakis, Í. (1996). Factors Affecting the Increase of Public Expenditure for Health in Greece. Scientific Bulletin, Medical School, The Aristotelian University of Thessaloniki, vol. 23, No 1, p.65-68, Healey, A. Yule, B. and Reid, J. (1994). Variations in General Practice Prescribing Costs and Implications for Budget Setting. Health Economics, Vol. 3: Maynard, A. (1975). Health Care in the European Community. Groom Helm, London. Maynard, A. (1977). "An Economic Analysis of Medical Care in Western Europe", in Habberstadt, V. and Culyet, A. (1977), Public Economics and Human Resources. Editions Cujas, Paris. Seretis, D. (1994).Why do Hospitals have Deficits. Greek daily Vima. 23rd November

12 Steele, R. (1981). Marginal met need and geographical equity in health care. Scottish Journal of Political Economy. 28: West, P. (1981). Theoretical and practical equity in the National Health Service in England. Social Science and Medicine. 15C: