The Effects of Economic Growth and Public Support of Health Services on Longevity--A Panel Data Analysis

Size: px

Start display at page:

Download "The Effects of Economic Growth and Public Support of Health Services on Longevity--A Panel Data Analysis"

Milton Norton
6 years ago
Views:

Utah State University DigitalCommons@USU Economic Research Institute Study Papers Economics and Finance 22 The Effects of Economic Growth and Public Support of Health Services on Longevity--A Panel

1 Utah State University Economic Research Institute Study Papers Economics and Finance 22 The Effects of Economic Growth and Public Support of Health Services on Longevity--A Panel Data Analysis Basudeb Biswas Utah State University Follow this and additional works at: Recommended Citation Biswas, Basudeb, "The Effects of Economic Growth and Public Support of Health Services on Longevity--A Panel Data Analysis" (22). Economic Research Institute Study Papers. Paper This Article is brought to you for free and open access by the Economics and Finance at It has been accepted for inclusion in Economic Research Institute Study Papers by an authorized administrator of For more information, please contact

2 Economic Research Institute Study Paper ERI#22-1 THE EFFECTS OF ECONOMIC GROWTH AND PUBLIC SUPPORT OF HEALTH SERVICES ON LONGEVITY-A PANEL DATA ANALYSIS by BASUDEB BISW AS Department of Economics Utah State University 353 Old Main Hill Logan, UT July 22

3 11 THE EFFECTS OF ECONOMIC GROWTH AND PUBLIC SUPPORT OF HEALTH SERVICES ON LONGEVITY-A PANEL DATA ANALYSIS Basudeb Biswas, Professor Department of Economics Utah State University 353 Old Main Hill Logan, UT The analyses and views reported in this paper are those of the author(s). They are not necessarily endorsed by the Department of Economics or by Utah State University. Utah State University is conunitted to the policy that all persons shall have equal access to its progran1s and employment without regard to race, color, creed, religion, national origin, sex, age, marital status, disability, public assistance status, veteran status, or sexual orientation. Information on other titles in this series may be obtained from: Department of Economics, Utah State University, 353 Old Main Hill, Logan, Utah Copyright 22 by Basudeb Biswas. All rights reserved. Readers may make verbatim copies of this document for noncommercial purposes by any means, provided that this copyright notice appears on all such copies.

4 111 THE EFFECTS OF ECONOMIC GROWTH AND PUBLIC SUPPORT OF HEALTH SERVICES ON LONGEVITY-A PANEL DATA ANALYSIS Basudeb Biswas ABSTRACT It is now recognized that the growth rate of per capita real income is not the only indicator of development. Development is a multidilnensional concept that focuses on development ofhuman beings through removal of poverty, ill health, illiteracy, and other social deprivations. The observed positive correlation between the growth of GNP per head and the average quality of life may not be as strong as it appears to be. It is argued that the causal relation between income and life expectancy, which is taken as a proxy for the index of quality of life works mainly through the impact of GNP on public expenditure in health care and poverty reduction. The objective of this paper is to examine this issue by using panel data. The dependent variable is life expectancy at birth while explanatory variables are real GDP per capita and per capita public expenditure on health. Using the data set for 29 countries over the period we find that public provisioning of health services is Inore important than the growth of real income in expanding the most basic capability for human development. JEL codes: 11,15,5

5 THE EFFECTS OF ECONOMIC GROWTH AND PUBLIC SUPPORT OF HEALTH SERVICES ON LONGEVITY-A PANEL DATA ANALYSIS It is now recognized that the growth rate of per capita real income is not the only criterion of development. Development is a multidimensional concept that focuses on human development through removal of poverty, ill health, illiteracy, and other social deprivations. Sen (1999) has summarized his views on human development as an expansion of freedom from all deprivations. Sen's influential ideas are reflected in the goals set in the first chapter of the first (199) report on human development published by United Nations Development Program. Sen's intellectual influence on policy makers has been increasing over time. The recently published Human Development Report 21 states, "Development is thus about expanding the choices people have to lead lives that they value. And it is thus much more than economic growth, which is only a means- if a very ilnportant one-of enlarging people's choices" (p. 9). Those who concentrate on the growth of real income as the main indicator of development do so because of the presumed positive causal relation between growth of real income and such indicators of development as life expectancy at birth, literacy rate, and freedom from other social deprivations. Lucas (1988) stresses the importance of the growth of per capita real income as the primary determinant of human welfare. Lucas's paper begins, "By the problem of economic development I mean simply the problem of accounting for the observed pattern across countries and across time, in levels and rates of growth of per capita income" (p. 3). Lucas admits that this definition of economic development is narrow, but when he considers the implications of diverse rates of growth of real per capita GNP over sustained periods, he finds stark differences in the average well-being

6 2 of the people of a country like India experiencing 1.4 percent growth for and the average well-being of he people of South Korea having 7 percent growth over the same period. "Indian incomes will double every 5 years; Korean every 1" (p. 4). But to say "An Indian will, on average, be twice as well as his grandfather; a Korean 32 times" is to presume a strong positive causal relation between the growth of income per capita and the consequent increase in well-being. The observed positive correlation between the growth of GNP per head and the average quality of life may not be as strong as it appears to be. Sen argues "... it would indicate that the connection tends to work particularly through public expenditure on health care, and through success of poverty removal. The basic point is that the impact of economic growth depends much on how fruits of economic growth are used" (p. 44). Anand and Ravallion (1993) have made some statistical analysis by using cross-country data and have found that there is a positive correlation between life expectancy and GNP per head. Their empirical work suggests that the causal relation between income and life expectancy works mainly through the impact of GNP on public expenditure in health care and poverty reduction. This work intends to examine the same issue by using panel data. In this study, the dependent variable is life expectancy at birth, which is taken as a proxy for quality of life. An increase in life expectancy indicates a better quality of life or expansion of freedom. Two explanatory variables are used. One is real GDP per capita and the other is per capita public expenditure on health. We would like to see the contribution of these two factors toward improving the quality of life. Econometric Specification of the Model Anand and Ravallion (1993) have made a cross-country analysis based on 1992 data. In this work we take a sample of 29 countries over the eleven-year period This panel data

7 3 approach provides multiple observations on each country in the sample. The major advantage over cross-sectional data sets is that it gives a large number of data points increasing the degrees of freedom and improving the efficiency of econometric estimates. The two factors affecting the quality of life and, hence, longevity are identified as growth of real income and public provisioning of health services. The former is designated by Dreze and Sen (1989) as the strategy of "growth mediated security" for improving the quality of life. The idea is that economic growth will enhance the capability of the people in general to buy medical services on the one hand and economic growth also will provide an improved basis for public support. It is this aspect that is emphasized by Lucas. The other strategy is to provide social assistance directly in the domain of health care, education, and other social arrangements. This process is called by Dreze and Sen as the strategy of "support-led security." Theoretically, the strategy of growth-mediated security and the strategy of support-led security are distinct. But empirically they are correlated. About the degree of correlation between these strategies, there is remarkable heterogeneity among the countries. Examples of two high economic growth economies given by Sen (1999) are those of Brazil and South Korea. Growth of income in Brazil has not contributed significantly to improvement of the quality of life. In South Korea we observe a strategy of growth-mediated security and also support-led security. In South Korea the rapid growth of GNP per capita over the last few decades and the positive involvement of the state in public provisioning of health services has resulted in the significant improvement of the quality of life. It is likely that there is some collinearity among the two strategies. In a panel data model with a large number of data points, the collinearity will be reduced and the efficiency of parameter estimates will be improved. The main motivation of this panel data study is to test empirically the relative effectiveness of these strategies and examine the robustness of results of the earlier study by Anand and Ravallion.

8 4 The Basic Model With Country-Specific Effects ~t = a j + p] X j(, ] + P 2 Xit, 2 + Cit i=1,2,...,25 (1) t = 1,2,...,1 The sample data are represented by observations on 29 countries over eleven periods of time. The explanatory variables are Xl and X 2, which stand for real GDP per head and public expenditures on health per head, respectively. Y it is life expectancy of the ith country in year t. There is no constant term in the model. The individual effect is ~, which is taken to be constant over time, t, and specific to the individual cross-sectional unit, i. There are differences in countries with regard to distribution in income. These differences across countries can be captured by differences in the constant term. Thus, each ~ is an unknown parameter to be estimated. The Data Set The data set includes observations on 29 countries over the period. The variables in this analysis are annual time series data. The dependent variable is the life expectancy at birth and the explanatory variables are PPP adjusted per capita real GDP and per capita real public expenditures on health. The sample size has been curtailed by the availability of data. Figure 1 is the scatter plot with GDP per capita on the x-axis and life expectancy in years on the y-axis. Figure 2 is a scatter plot of per capita health expenditures on the x-axis and life expectancy in years on the y-axis. Figures 3 and 4 are scatter plots of the logarithms of the variables.

9 5 8 ~ ~ o o o o OJ ~ :::r 4-3 ~ '-I ~I ~ o Real GOP per capita Figure 1. Annual time-series data for life expectancy at birth and real GDP per capita for 29 countries over the period

10 6 8 cp oo!t r 7 "o~ '" ~ CD CQ gp a m ~ ~ X -. CD 6 o&f~ () r-+ o ru ~ ~o ::::l, () < '< ru 5 r-+ OJ ;::l.. ::::r 4 a 3~----~----~------~----~----~ o Public Health Expenditure per capita Figure 2. Annual time-series data for life expectancy at birth and public health expenditure per capita for 29 countries over the period

11 7 4.4, ~ r o co 4.2- ep o O o o o 3.6; ,, ,----~ ~,------~ Log (Real GOP per capita) Figure 3. Annual time-series data for log (life expectancy at birth) and log (real GDP per capita) for 29 countries over the period

12 8 4.4 r <C C$19 o...-- o Q) fp r o CD o~o~ ~ (/ CD 4.2- m : e#oj ~ >< ocp#o ge "' 8 c9 r:j CD o o o~ cp ~OO :::J 4.- ci- of:.'e o ttj8 ~o> CO <9 '< CPo ~IO - \ r-+ J ~ ru 9 o~ go. ru co ~ r-+ OJ o ~ CQ ", (/ JO- lj o o~ q, ~~ OlD ::+ o o ad 8 ::T 3.8- '-'" o <9=> > 3.6 I I I Log (Public Health Expenditure per capita) Figure 4. Annual time-series data for log (life expectancy at birth) and log (public health expenditure per capita) for 29 countries over the period

13 9 Choice a/the Functional Form and the Estimating Equation Visual inspection of the scatter diagrams in Figures 1 through 4 suggest the following log linear functional form with dummy variables to capture the country-specific characteristics: where Yjt is life expectancy of country i in year t, X;t is real GDP per capita (purchasing power adjusted), X;t, 2 is the per capita expenditure on health, ~ represents country-specific fixed effects, and f3s are slope coefficients. For the disturbance term, ViI' it is assumed that V it,..., NID(O, 2 ). We also used a specification test called the PE test to test for linear versus log linear specification. Test results (not reported here) indicate rejection of the linear model in favor of the log linear form. The estimates reported in Table 1 indicate that the effect of public provisioning of health services on life expectation is positive and significant. A 1 percent increase in per capita public expenditure on health services will result in a 1.74 percent increase in life expectancies. However, the effect of an increase in income on life expectancy is positive but not significant. This result is comparable to results of Anand and Ravallion who did not find any significant relation between income and life expectancy in a regression where poverty incidence, public health spending, and average income were the explanatory variables. In our model we did not include the variable, poveliy reduction, because annual time-series data for this variable were not available for all countries over the period But to capture the cross-country variation in life expectancy due to difference in measures of poverty reduction and other country-specific characteristics dummy variables were included in the model (2)

14 Table 1. Estimates of Equation (2) by LSDV Method (dependent variable = log of life expectancy at birth) 1 Dependent Variable: Log (Life Coefficient Standard Expectancy at Birth) Estimates Error t-statistic Prob. Dummies: D 1 for Bangladesh D2 for Benin D3 for Bolivia D4 for Burundi D5 for Cote d'ivoire D6 for Egypt D7 for Ethiopia D8 for Ghana D9 for Guinea DI for India D 11 for Indonesia D12 for Iran, Islamic Rep D13 for Jordan D 14 for Madagascar D15 for Mali D16 for Nepal D 17 for Niger D 18 for Pakistan D19 for Papua New Guinea D2 for Paraguay D21 for Peru D22 for Senegal D23 for Sri Lanka D24 for Togo D25 for Turkey D26 for Zambia D27 for Zimbabwe D28 for Brazil D29 for Korea, Rep Major Independent Variables: Lgdp Lhea Other Statistics Reference: R-squared Mean dependent var Adjusted R-squared.9244 S.D. dependent var S.E. of regression Akaike info. criterion Sum squared resid Schwarz criterion Log likelihood F -statistic Durbin-Watson stat Prob (F-statistic)

15 11 Conclusion The main motivation of this work was to empirically test the relative strength of "growth-mediated approach to and the support-led security" approach to an improvement in the quality of life proxied by life expectation at birth. The findings of this study indicate that public provisioning of health services is more important than the growth of real income in expanding the most basic capability for human development, that is, the capability to lead a long and health life. Although the results should be taken with caution, the main conclusion is consistent with earlier results reported in Anand and Ravallion (1993). References Anand, Sudhir, and Martin Ravallion "Human Development in Poor Countries: On the Role of Private Incomes and Public Services." Journal of Economic Perspectives 7(1, Winter): Dreze, Jean P., and Amartya Sen Hunger and Public Action, WIDER Studies in Development Economics. Oxford: Clarendon Press. Greene, William H. 2. Econometric Analysis. New York: Macmillan Publishing Company. Knlenta, Jan Elements of Econometrics, 2nd edition. New York: Macmillan Publishing Company. Lucas, R.E "On the Mechanics of Economic Development." Journal of Monetary Economics 22:3-42. Sen, Amartya K Development as Freedom. New York: Alfred A. Knopf. UNDP Human Development Report New York: Oxford University Press.

16 12 Appendix Variable Definitions and Data Sources Variable Definition LE: Stands for Life Expectancy at Birth, the number of years a newborn infant would live if prevailing patten1s of mortality at the time of its birth were to stay the same throughout its life. GDP: Stands for Real (PPP $ adjusted) GDP per capita. HEA: Stands for Public Health Expenditure per capita, which is calculated as the product of the GDP per capita and the Public Health Expenditure as percentage of GDP. Data Sources Variable LE GDP HEA Data Source Human Development Report (199-2) Human Development Report (199-2) Human Development Report (199-2) There are 29 countries for which balanced data were available. These countries are: Bangladesh,** Benin,** Bolivia,* Brazil,* Burundi,** Cote d'ivoire,** Egypt,* Ethiopia,** Ghana,** Guinea,** India,** Indonesia,** Iran,* Islamic Rep.,* Korea,* Jordan,* Madagascar,** Mali,** Nepal,** Niger,** Pakistan,** Papua New Guinea,* Paraguay,* Peru,* Senegal,** Sri Lanka,* Togo,** Turkey,* Zambia.** *Middle income countries (GNP per capita of$725-$9,265 in 1999). **Low income countries (GNP per capita of$755 or less in 1999).

ANALYSIS OF THE LINKAGE BETWEEN DOMESTIC REVENUE MOBILIZATION AND SOCIAL SECTOR SPENDING

ANALYSIS OF THE LINKAGE BETWEEN DOMESTIC REVENUE MOBILIZATION AND SOCIAL SECTOR SPENDING NATHAN ASSOCIATES INC. Leadership in Public Financial Management II (LPFM II) 1 MOTIVATION Strengthening domestic