PUBLIC AND PRIVATE HEALTH EXPENDITURE IN CENTRAL ASIA: NEW EVIDENCE FROM PANEL DATA ANALYSIS Текст научной статьи по специальности «Клиническая медицина»

УДК 61:001.83(100)

I PUBLIC AND PRIVATE HEALTH EXPENDITURE IN CENTRAL ASIA: NEW EVIDENCE FROM PANEL DATA ANALYSIS

by

Saltanat Akhmetova

Data scientist at GARCH consulting, Cambridge, UK

Abstract

Health care expenditure has been low over the years in developing regions of the world. The objectives of this study are to determine the effect of health care expenditure on population health status and to examine the effect by public and private expenditure sources. The study used panel data from 1995 to 2015 covering 6 countries in Central Asia. Fixed and random effects panel data models were fitted to determine the effects of health care expenditure on health outcomes. The results show that public health care expenditure, unlike private, negatively influences health status through deteriorating life expectancy at birth, increasing death and infant mortality rates. Public and private health care spending showed opposite strong association with health status while no significant relationship was observed toward infant mortality rate. The findings imply that public health care expenditure is ineffective in enhancing health status in Central Asia countries. Increasing efficiency and better allocation ofpublic health expenditures will be a significant step in the future health system reforms.

Keywords: public and private health care expenditure, fixed and random effects, Central Asia, Kazakhstan, Uzbekistan, Kirgizstan, Turkmenistan, Tajikistan, Azerbaijan.

Тужырым

Денсаулыц сацтау саласындагы шыгындар жылдар бойы элемнщ дамыган аймацтарында бэсендеу ба-рысында. Осы зерттеудщ мацсаты денсаулыц сацтау шыгындарыныц халыц денсаулыгына тиг1зетщ эсергн аныцтау болып табылады. Зерттеу 1995 жылдан 2015 жылга дейгнгг аралыцта 6 Орта Азия мемлекеттгргн цамтыды жэне турацты жэне кездейсоц эффекттж панельдж деректер Yлгiлерi эдга пайдаланды. Зерттеу нэтижелерi денсаулыц сацтау саласында жумсалатын мемелекеттiк каражат, жекеменшiкке цараганда, вмiр CYPУ узацтыгын, влiм мен влiм-жiтiм децгейте терк эсерт кврсеттi. Зерттеу, денсаулыц сацтауга жумсалатын шыгындар Орталыц Азия елдерiндегi денсаулыц жагдайын жацсартуда тишаз болып табылатындыгын дэлелдедi. Денсаулыц сацтау саласыныц жэне денсаулыц сацтауга жумсалатын шыгындарды тиiмдi жумсау болашац денсаулыц сацтау ЖYйесiн реформалауда мацызды цадамы тиiстi болу цажет.

Кiлттi свздер: мемлекеттт жэне жеке денсаулыц сацтау шыгындары, тiркелген жэне кездейсоц эсерлер, Орталыц Азия, Казацстан, взбекстан, Кыргызстан, ТYрiкменстан, Тэжiкстан, Эзiрбайжан.

Резюме

Расходы на здравоохранение за последние годы были низкими в развивающихся регионах мира. Цели этого исследования состоят в том, чтобы определить влияние государственных и частных расходов на здравоохранение, состояние здоровья населения. В исследовании использовались панельные данные за период с 1995 по 2015 год, охватывающие 6 стран в Центральной Азии. Результаты показывают, что расходы на здравоохранение, в отличие от частных, отрицательно влияют на состояние здоровья за счет ухудшения ожидаемой продолжительности жизни, увеличения смертности и младенческой смертности. Государственные и частные расходы на здравоохранение демонстрировали противоположную сильную связь со статусом здоровья, в то время как существенных отношений не наблюдалось в отношении младенческой смертности. Полученные данные свидетельствуют о том, что расходы на здравоохранение неэффективны в улучшении состояния здоровья в странах Центральной Азии. Повышение эффективности и лучшее распределение расходов на здравоохранение должно быть важным шагом в будущих реформах системы здравоохранения.

Ключевые слова: государственные и частные расходы в здравоохранении, фиксированные и случайные эффекты, Центральная Азия, Казахстан, Узбекистан, Кыргызстан, Туркменистан, Таджикистан, Азербайджан.

менеджер! ^ 2017

r

9

INTRODUCTION

The causal relationship between health expenditures and health outcomes continues to attract the attention of many [13;15;16;17]. The empirically estimation of this relationship may serve as an important monitoring indicator of health system progress. Improving the health care system in Central Asia (CA) has been a hot topic of discussion and important questions are how far public expenditure has been instrumental in bringing about the progress in health status, and what programs have been particularly effective. However, little attention has been placed on the relationship between spending on health and health outcomes in the region.

Previous studies that investigated the relationship between health resources and health outcomes are diverse in what concerns the analysed indicators, the model approach, and the countries studied [17;16;8;2;3]. Some recent studies have found a positive relationship between spending on health and health outcomes [8;11;12;13], but others did not find a significant relationship between the two variables [5;6]. The increase in the health status is different from one country to another; some of them have been very successful in changing the health status, while some have not.

At the regional level, Novignon et al [16], Anand and Ravallion [15] using panel data from 1995 to 2010 covering 44 countries in SSA and cross-country data for 22 countries, respectively, found that healthcare spending, particularly public spending on health, promotes health status. The current study is motivated by the inconclusive debate on the relationship between health expenditure and health outcomes with particular attention on CA. The purpose of the study is, therefore, to investigate the impact of total health care spending on various health outcomes after controlling for country-specific demographic structures and economic conditions. Secondly, a differential analysis of public and private health care spending is performed.

MATERIALS AND METHODS

The study pooled cross-section and

annual time series data from 1995 to 2010 for 6 countries in Central Asia. The data used in the empirical analysis were sourced from the World Bank, World Development Indicators (WDI) [1]. The study uses life expectancy at birth, infant mortality rate and death rate as health status indicators. Life expectancy at birth reflects how many years a person might be expected to live given prevailing mortality rates. Using the three indicators of health outcome will, therefore, allow for robustness in the analysis. While total, public and private health care expenditures are measured as percentage of GDP, Income per head is measured as GDP per capita at constant 2000 US$. Higher health care expenditure is expected to relate to higher life expectancy at birth and lower infant mortality and death rates [12;13;16;17]. Finally, different population age groups namely, age below 14 years, 15 to 64 yearsand above 65 years were measured as percenCag6 of total population. Twese were included to condil for tee dieptrebp country demographic structures. Relative eo the younger population ehe populationagp g60pp above 65 years is exp ecteo to reduce .pc1! outeemes by increasin6 death rates [16;9].

For the purposes of this study, the followinf motM fpefifiyatloac were ewtim ated;

HIit = at + p1PubExit + p2PrExit + (!3Yit + p4P 1 u + /35 P 2 it+p6P 3 u + ?it ( 1 )

Wh^e Hc: 6ppresents tfiree It em^ri outcomes (Life expec6pacy ah birth, Infant mortelity rate anR Death rate) in countiy i in periop t. Ex is tiie total health exhepdi6ure as ^rcentage o6 renl national income. Y is for nap1Sareeiincome which acts as a coniaol varioWs for tfie demand for healtw servicesmd 6thcr economia dactore.

Thu variable ^ 2 owd 3 represente populati<tn age 6c°uPS °e bblow 14, 15-64 and above Pv yearn reephctivbly eeprec sed a6 a percentage of total! potation, so is time; invarianthhd c apIurescountty-spacificehSect that was notmc lueed in the mwdel. Th e error terns

were ascumad to be normaHy diitribuced. e_ih

the errorteim Total ilií^lrh c^p^i^^ expnndi ture i s further rlouped into private bad publia; health

№2, 2017 MeH^Pi

care expenditure. This was to allow for the analysis of the individual impact of each of these components, Where PubEx and PrEx represent

Table 1 Descriptive Statistics

Health expenditure and life expectancy at

birth

Results from the fixed and random effects models are reported in Tables 2, 3 and 4 for life expectancy at birth, death rate and infant mortality rate respectively. Table 2 shows that increase in

public and private health care expenditure, respectively.

total health expenditure (as% of GDP) was more likely to increase life expectancy at birth at 10% significance level. A 1% increase in total health expenditure leads to an improvement in life expectancy at birth by approximately 0.4 years in both models (Table 4).

Variable GLS-fixed effects model GLS-random effects model

(1) (2) (1) (2)

Constant 103.2(31.98) *** 104.6(32.07) *** -169.82(-4.98) *** 35 (14.20) ***

Health expenditure, total 0.45(2.68) * 0.42(2.49) *

Health expenditure, public (% of GDP) -,044 (-0.15) -,79(--3.12) ***

Health expenditure, private (% of GDP) ,78 (3.31) *** 0.4 (2.83) **

Variable Mean Minimum Maximum Std.Dev.

Health expenditure, total 4,9 1,9 7,9 1,3

Health expenditure per capita 98,4 3,0 582,1 123,6

Out-of-pocket health expenditure (% of total expenditure on health) 52,2 18,5 82,4 14,7

Out-of-pocket health expenditure (% of private expenditure on health) 94,5 77,7 99,8 5,7

Health expenditure, public (% of total health expenditure) 44,2 11,2 81,5 17,4

Health expenditure, public (% of GDP) 2,1 0,8 4,2 0,9

Life expectancy at birth, total (years) 65,1 56,1 72,0 3,5

Death rate, crude (per 1,000 people) 8,8 5,6 15,3 2,3

Mortality rate, infant (per 1,000 live births) 50,0 12,6 87,9 19,7

Health expenditure, private (% of GDP) 2,9 0,7 7,0 1,3

GDP per capita (constant 2010 US$) 2747,5 364,6 10645,5 2725,1

Population ages 0-14 (% of total) 33,2 21,9 45,0 6,3

Population ages 15-64 (% of total) 60,9 49,5 72,5 6,0

Population ages 65 and above (% of total) 5,9 4,2 7,7 0,8

Table 2 Effects of health care expenditure on life expectancy at birth (years)

IHiig1 №2, 2017

Y

11

Note: ***significant at 1%; "significant at 5%; *significant at 10%. t-statistics are reported in parenthesis. (1) is model with aggregate health care expenditure. (2) is model with aggregate health expenditure decomposed into public and private.

GDP per capita (constant 2010 US$) 0.0002(2.84) ** 0.002 (2.15) 0.0002(2.12) 0.000143(0.063)

Population ages 0-14 (% of total) -0.72(-19.23) *** -0.72 (-19.37)*** 2.003(5.69) *** 2.2 (6.19) ***

Population ages 1564 (% of total) 1.04(3.33) *** 1,294966 (4,411) *** 0.69(18.03) *** 0.56(13.12) ***

Population ages 65 and above (% of -2.86(-7.43) *** -3.08 (-7.78)*** -1.78(-5.02) *** -0.73(-2.61) *

total)

R-squared 0.78 0.79

Durbin-Watson 0.173 0.161

F-Stat. 29.33*** 22 9***

Observations 120 120

Cross section 6 6

included

Disintegrating the effect of total health expenditure shows that an increase in public health care expenditure decreases life expectancy at birth by about 0.8 years and in private increase 0.4 years, respectively in the random effects model and about 0.04 and 0.78 years, respectively, in the fixed effects model (Table 2).

Health expenditure and death rate Table 3 shows that an increase in total health expenditure reduces death rate (per 1000

people) by approximately 0.23 in the fixed effects model and 0.19 in the random effects model. While public health care's expenditure is increasing the death rate by about 0.09 in fixed and 0.5 in random effects models. The private health care expenditure reduced death rate by approximately 0.45 and 0.22 per 1000 people in the fixed and random effects models, respectively, at 5% and 1%

Variable GLS-fixed effects model GLS-random effects model

(1) (2) (1) (2)

Constant -13(-6.6) *** -14,5 (-7.04) *** 20(-13.8) *** 18(14.74) ***

Health expenditure, total -,23(-,2.17) -,189(-1.89)

Health expenditure, public (% of GDP) ,09(0.5) ,48 (3.92) ***

Health expenditure, private (% of GDP) -,45(-3) *** -,22 (-2.76) **

GDP per capita (constant 2010 US$) -,0002(--4.75) *** -,000144 (-3,757) *** -,000215(-4,675) *** -,000168 (-3,824) ***

Population ages 0-14 (% of total) 0.35(14.84) *** ,322612(15,774) *** ,343530(15,177) *** ,343112 (15,899) ***

Population ages 1564 (% of total) -0.33(-14.73) *** -0.32(-15.38) ***

Population ages 65 and above (% of total) 2.09(8.61) *** 2,24 (8.96) *** 1.69(8.19) *** 1.76 (12.85) ***

R-squared 0.69 0.70

Durbin-Watson 0.173

F-Stat. 10.34*** 5.10***

Observations 120 120

Cross section included 6 6

Table 3 Effects of health care expenditure on death rate (per 1000 people)

№2, 2017

Note: ***significant at 1%; "significant at 5%; *significant at 10%. t-statistics are reported in parenthesis. (1) is model with aggregate health care expenditure. (2) is model with aggregate health expenditure decomposed into public and private.

Health expenditure and infant mortality

rate

Estimation results provides evidence of other studies that have found that the influence of health expenditure on health status is either small or statistically insignificant [2;3;4;7;14;17].

Meanwhile, total health care expenditure was more likely to reduce infant mortality rate (per 1000 live births) with 1% level of significance

(Table 6).A 1% increase in total health expenditure reduced infant mortality rate by approximately 2 infants per 1000 live births in fixed effects model. While public health care expenditure reduced infant mortality rate by approximately 6 infants in both models, an increase in private health care expenditure by 1% reduced infant mortality rate by 2 infants per 1000 live births in both models (Table 6).

Table 4Effects of health care expenditure on infant mortality rate (per 1000 live births)

Variable GLS-fixed effects model GLS-random effects model

(1) (2) (1) (2)

54.7(2.2) 41.9 (24.7) 36.11(1.75) 98.09(0.4)

Health expenditure, total -1.95 (-3.2) *** 0.58(0.9)

Health expenditure, public (% of GDP) 6.5 2.86) ** 5.6 (2.64)

Health expenditure, private (% of GDP) -1.2 (-0.69) -0,87 (-0.59)

GDP per capita (constant 2010 US$) 0.001 (1.74) 0.0007 (1.23) 0.002 (0.5) 0.0006 (1.13)

Population ages 0-14 (% of total) -2.8(-1.13) -2.75(-1.1) 0.95 (0.36) -0.75(-0.27)

Population ages 1564 (% of total) -3.68 (-1.39) -3.59(-1.34) 1.13 (0.68) -0.57(-0.19)

Population ages 65 and above (% of total) 0.7(0.24)

R-squared 0.07 0.01

Durbin-Watson 0.173 0.161

F-Stat. 21.705*** 20.952***

Observations 120 120

Cross section included 6 6

Note: ***significant at 1%; "significant at 5%; *significant at 10%. t-statistics are reported in parenthesis. (1) is model with aggregate health care expenditure. (2) is model with aggregate health expenditure decomposed into public and private.

DISCUSSION

The findings of the study suggest that efficient management of health care spending is an important step in improving health outcomes in Central Asia. The results show that the public health expenditure was negatively related to life expectancy, infant mortality and death rates. However, a positive relationship was found between the private health expenditures and all three measures of health outcomes.

mmrn ^ 2017

However, our findings contradict earlier results of other studies that found significant positive relationship between public expenditure on health and health outcomes [13;14;16;17]. It must be noted that the pace of health expenditure growth is different for countries at different levels of economic development, while the findings of the current study provides evidence in ineffective and inefficient allocation of health care expenditures. It is possible for population

f

13

health to worsen even as health care expenditure increase in the face of misallocation and poor management.

The study is limited in the sense that most of the health outcome variables for CA did not have enough time series observation which would have improved a panel data study as this one.

Including three different health outcome measures, however, allows for robustness of results. While these limitations may be the bases for future research, they do not invalidate the results of the current study.

Conclusion

The study sought to determine the impact of health care expenditure on health status measured by life expectancy at birth, crude death rate and infant mortality rate in CA. The results provided evidence that the public health care expenditure, opposite to private, was found to have negative impact on life expectancy and death rates. But the relationship with infant mortality for both public and private health expenditures was statistically insignificant.

The findings imply that, health care expenditures effective allocation and management are essential components in improving health status in CA. These results have important implications for attaining the targets envisioned by the future reforms and to decision-making in many different contexts.

The following countries were included in the study: Azerbaijan, Kazakhstan, Kyrgyz Republic, Tajikistan and Turkmenistan.

REFERENCES

1. WB: The World development indicators. In Book The World development indicators.: The World Bank; 2012.

2. Akinkugbe O, Mohanoe M: Public health expenditure as a determinant of health status in Lesotho. Soc Work Public Health 2009, 24:131-147.

3. Akinkugbe O, Afeikhena J: Public health care spending as a determinant of health status: a panel data analysis for SSA and MENA. In Applied macroeconomics and economic development. Edited by Adenikinju A, Olaniyan O. Ibadan: Ibadan University Press; 2006.

4. Musgrove P: Public and private roles in health. In Technical report 339. Washington DC: World Bank; 1996.

5. Filmer D, Pritchett L: Child mortality and public spending on health: how much does money matter. In World Bank Policy Research Working Paper No 1864. Washington DC: World Bank; 1997.

6. Filmer D, Pritchett L: The impact of public spending on health: Does money matter? Soc Sci Med 1999, 49:1309-1323.

7. Wagstaff A, Cleason M: The millennium development goals for health: rising to the challenge. Washington DC: The World Bank; 2004.

8. Berger MC, Messer J: Public financing of health expenditures, insurance and health outcomes. Applied Economics 2002, 34:21052113.

9. Hernandez de Cos P, Moral-Benito E: Health care expenditure in the OECD countries: efficiency and regulation. In The occassional paper series No 1107. Madrid: Bank of Spain; 2011.

10. Baltagi BH, Song SH, Jung BC, Koh W: Testing for serial correlation, special autocorrelation and random effects. Journal of Econometrics 2007, 140:5-51.

11. Baldacci E, Guin-Siu MT, de Mello L: More on the effectiveness of public spending on health care and education: a covariance structure model. Journal of International Development 2002, 15:709-725.

12. Or Z: Determinants of health outcomes in industrialised countries: a pooled, crosscountry, time series analysis. In OECD Economic Studies No 30, 2000/1. Paris: Organization for Economic Cooperation and Development; 2000b.

13. Or Z: Exploring the effects of health care on mortality across OECD countries. In Labour Market and Social Policy Occasional-Papers No 46. Paris: Organisation for Economic Cooperation and Development; 2000a.

14. WHOSIS. Per capita health expenditures in an average exchange rates. Geneva: World Health Organization; 2009

15. Anand, S. and Ravallion, M. (1993), "Human Development in Poor Countries: On the Role of Private Incomes and Public Services",

№2, 2017 Maiew.

The Journal of Economic Perspectives, 7(1), 133-150.

16. Novignon, J., Olakojo, S.A., Novignon, J. (2013). The Effects of Public and Private Health Care Expenditure on Health Status in Sub-Saharan Africa: New Evidence from Panel

Data Analysis. Health Economics Review 2: 22

17. Nixon, J., Ulmann, P., 2006. The relationship between health care expenditure and health outcomes. Evidence and caveats for a causal link. European Journal of Health Economics 7, p. 7.

№2, 2017