OBSERVARE
Universidade Autónoma de Lisboa
e-ISSN: 1647-7251
Vol. 12, Nº. 2 (November 2021-April 2022)
175
A KUZNETS ADAPTIVE APPROACH TO LIFE EXPECTANCY AT BIRTH: AN
APPLICATION ON RISING POWERS
HÜSEYIN ÜNAL
huseyin.unal@ktu.edu.tr
Karadeniz Technical University, Department of Econometrics
HÜLYA KINIK
hulya.ercan@ktu.edu.tr
Karadeniz Technical University, Department of International Relations
Abstract
This study aims to test the validness of Kuznets’ hypothesis in major rising powers between
the years of 2000 and 2018 within the scope of the relationships between life expectancy at
birth (throughout the paper-life expectancy-LE) and economic growth. Using panel data
analysis method, we investigate if there is a curve such as Health Kuznets Curve (HKC) for
life expectancy. The empirical findings indicate that the validity of HKC hypothesis could not
be obtained for Brazil, Mexico, Russian Federation, South Africa and Turkey. A U-shaped
relationship exists between these two variables for these countries. In other respects, we
found empirical evidence of a Kuznets’curve and inverted U-shaped relations between
economic growth and life expactancy for Australia, China, Indonesia and Korea. Empirical
evidence also suggests that there is not any relationship between economic growth and life
expectancy for India.
Keywords
Life Expectancy, Rising Powers, Economic Growth, Panel Data Analysis, Kuznets’s Hypothesis
How to cite this article
Ünal, Hüseyin; Kinik, Hülya (2021). A Kuznets adaptive approach to life expectancy at birth:
an application on rising powers. In Janus.net, e-journal of international relations. Vol12, Nº.
2, November 2021-April 2022. Text… Consulted [online] on the date of the last visit,
https://doi.org/10.26619/1647-7251.12.2.11
Article received on October 17, 2020 and accepted for publication on March 19, 2021
JANUS.NET, e-journal of International Relations
e-ISSN: 1647-7251
Vol. 12, Nº. 2 (November 2021-April 2022), pp. 175-188
A Kuznets adaptive approach to life expectancy at birth: an application on rising powers
Hüseyin Ünal, Hülya Kınık
176
A KUZNETS ADAPTIVE APPROACH TO LIFE EXPECTANCY AT
BIRTH: AN APPLICATION ON RISING POWERS
HÜSEYIN ÜNAL
HÜLYA KINIK
Introduction
Life expectancy is among the most significant indicators of health and public welfare
widely used to measure the general health status of a population. In practice it is a
reasonable proxy for population health (Canning 2012: 1784) and a measure that
summarizes the mortality level of a given population in a given year. It provides us with
key information about the development level of a country’s welfare state (Bayın, 2016:
94). Health issues have become essential as countries with higher life expectancy have
a tendency to show a better level of development and achieve long-term economic
development (Hassan et al, 2016: 105).
In this context, we analyze the relationship and causality between life expectancy and
economic growth and different control variables under the Simon Kuznets's “inverted U-
curve hypothesis” for 10 rising powers named BRICS group (Brazil, Russia, India, China,
South Africa) and MIKTA countries (Mexico, Indonesia, Korea, Turkey and Australia)
during the period 2000-2018, using panel data method. Although rising powers
phenomenon is a new concept, it has been a subject to many studies but there has been
very little published on their status in health. These countries not only prioritize economic
development, but they also put emphasis on the cooperation in the field of global health.
They have been recognised for their capacity and potential to influence global health. On
the other hand, BRICS and MIKTA countries together represent nearly 50% of global
population. Accordingly, it is crucial to analyze their situation in terms of life expectancy
as a key representation of a population’s general state of health.
Over the past few decades, new rising powers have achieved notable success with regard
to their life expectancy. These improvements have been result of several factors such as
growing incomes and increasing education as well as governments’ attempts to develop
their citizens’ health status. Global life expectancy at birth in 2018 was 72.5 years ranging
from the lowest as 63.9 years for South Africa to the highest as 82.7 years for Australia
among selected countries. As the Figure 1. shows that during the given period, there has
been an increase in life expectany of South Africa but it is stil below the world average.
On the other hand, Australia and Korea rank among ten nations with the highest life
expectancy. The life expectancy at birth is 77.4 years for the total population in Turkey
which ranked 52 in the world in 2018.
JANUS.NET, e-journal of International Relations
e-ISSN: 1647-7251
Vol. 12, Nº. 2 (November 2021-April 2022), pp. 175-188
A Kuznets adaptive approach to life expectancy at birth: an application on rising powers
Hüseyin Ünal, Hülya Kınık
177
Figure 1. Life Expectancy At Birth (both sexes combined, world rank, 2018)
Source: World Bank, World Development Indicators
Within this framework, this study is organized as follows: the first part summarizes the
existing literature on the determinants of life expectancy; section 2 reviews Kuznets
Hypothesis as theoretical background and describes the method of data collection and
methology of the study; section 3 examines the results of the study and the latest section
reports the conclusions.
1. Literature Review on the Determinants of Life Expectancy
Numerous previous studies devoted to investigate different life expectancy determinants
have taken into consideration several factors like income, inflation, education, health
care spending, improved water coverage and sanitation, employment rate, urbanization,
and many others. However, there is a lack of consensus about the variables that
determine life expectancy in empirical evaluation. The only consensus is that income
affects life expectancy positively.
In his cross-sectional study, Grossman (1972) investigated that inflation negatively
affects life expectancy, and household welfare was generally damaged due to increasing
prices. Preston (1976) evaluated the relative importance of income and variations in
income in determining the levels and fluctuations in the level of life expectancy. The main
result of his study is that life expectancy was correlated with per capita income, but
50
55
60
65
70
75
80
85
2000 2001 2002 2003 2004 2005 20062007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018
Brazil (74) Russian Federation (105) India (125)
China (51) South Africa (153) Mexico (48)
Indonesia (124) Korea, Rep. (9) Turkey (52)
Australia (6) World, Total
JANUS.NET, e-journal of International Relations
e-ISSN: 1647-7251
Vol. 12, Nº. 2 (November 2021-April 2022), pp. 175-188
A Kuznets adaptive approach to life expectancy at birth: an application on rising powers
Hüseyin Ünal, Hülya Kınık
178
eventually changes in income have been quite unimportant since World War II to affect
changes in life expectancy.
In their study, Rogers and Wofford (1989) found that urbanization, agrarian population,
illiteracy rate, safe drinking water, average calorie per person and doctor per population
had a significant role on life expectancy for developing states. Gulis (2000) found that
income per capita, public spending on health, access to water, calorie intake and rate of
literacy are highly effective in determining life expectancy for 156 countries of the world.
Kalediene and Petrauskiene (2000) indicated that urbanization is among the main life
expectancy determinants for both developed and developing countries as they are able
to reach better medical aid, more opportunities for education and advanced social and
economic background which positively affected the health.
Hussain (2002) have also studied factors that affect life expectancy based on the cross-
sectional data using multiple OLS. The result of his study suggested that life expectancies
in developing countries could be significantly developed if close attention was given to
fertility decrease and raising calorie intakes.
Yavari and Mehrnoosh (2006) examined how socio- economic factors affect life
expectancy based on multiple regression analysis. The results of their study suggest a
positive and strong interaction between life expectancy and per capita income, health
expenses, literacy rate and daily caloric intake. Their study also shows that the number
of people per doctor negatively affects life expectancy in African countries. Erdogan and
Bozkurt (2008) analyzed the correlation between life expectancy and economic
development in Turkey between 1980-2005 basen on ARDL boundary test model. They
asserted that economic growth positively affects life expectancy in Turkey.
Kabir (2008) examined the socio-economic factors that have effect on life expectacy with
ten widely used variables for 91 developing countries by applying multiple regression
probit models. The findings suggest that almost all explanatory variables turned out to
be unimportant, showing that socio-economic conditions can not be regarded as
influential on the life expectancy of developing nations all the time.
Lei et al. (2009) explored the socioeconomic determinants of life expectancy in Beijing
by using the linear stepwise regression model. The results show that floor space available
per rural resident and GDP per capita have a positive relationship with life expectancy,
while there is a negative relationship between life expectancy and the rural population
proportion and illiteracy rate.
Balan and Jaba (2014) investigated the factors that determine life expectancy in Romania
between 1970 and 2008. The results of their study reveal that a positive relationship
exists between life expactancy and wages, the number of beds in hospitals, the number
of doctors, and the number of readers subscribed to libraries. In addition, the proportion
of the Roma population and the illiterate population ratio have negative effects on life
expectancy.
Bilas et al. (2014) examined life expectancy for 28 European Union countries during
2001-2011 using panel data analysis method. They pointed out that both GDP per capita
and level of education expained between 72.6% and 82.6% of differences in life
expectancy.
JANUS.NET, e-journal of International Relations
e-ISSN: 1647-7251
Vol. 12, Nº. 2 (November 2021-April 2022), pp. 175-188
A Kuznets adaptive approach to life expectancy at birth: an application on rising powers
Hüseyin Ünal, Hülya Kınık
179
Based on data from 1970 to 2012, Ali and Ahmad (2014) also studied determinants of
life expectancy for Oman by using ARDL boundary test method. In their analysis, they
included the following determinant factors: per capita income, food production, schooling
rate, population growth, inflation and CO2 emissions. According to the results, food
production and schooling rate positively affect life expectancy and have statistically
significant effects on life expectancy while inflation and per capita income had negative
and unreasonable impacts on life expectancy. The results also suggest that population
growth had a negative and significant effect on life expectancy while CO2 emissions had
positive and statistically insignificant impact on the life expectancy in the long-term and
a negative and statistically significant effect in the short-term.
Jaba et al. (2014) studied the correlation of health expenditures with life expectancy in
selected 175 countries between 1995 and 2010 by using panel data method. There is a
significant correlation between these two variables.
Memarian (2015) analyzed the relationship among health spending, life expectancy and
economic growth in Iran from 1989 to 2011 deploying the ARDL econometric model. He
found that as life expectancy and health care expenditure increased economic growth
increased as well.
Based on A Vector Autoregression (VAR) Analysis method, Sede and Ohemeng (2015)
analyzed the socio-economic determinants of life expectancy in Nigeria between 1980
and 2011. They mesured the effects of different independent variables as follows: per
capita income, secondary school enrolment, public expenditures on health,
unemployment rate and the Naira exchange rate. Schooling rate in secondary education,
per capita income and government expenditure on health were not significant in
determining the life expectancy in Nigeria. However, unemployment and exchange rate
had a significant effect on life expectancy.
Şahbudak and Şahin (2015) studied the relationship between health indicators and
economic growth in BRIC countries between 1995 and 2013 by using panel data method.
They used GDP as dependent variable and included the share of health expenditures in
GDP, life expectancy at birth and child mortality rates as independent variables. Results
showed that there is a positive relationship among the share of health expenditures in
GDP, life expectancy at birth and economic growth; but negative correlation exists
between economic growth and child mortality rates.
Monsef and Mehrjardi (2015) studied the determinants of life expectancy in 136 countries
during 2002-2010 based on panel data analysis method. Their study shows that
unemployment and inflation have a significant negative effects on life expectancy.
However, a positive relationship exists among the gross capital formation, national
income, and life expectancy.
Hassan et al. (2016) researched the relationship between life expectancy rate and
expenses on health, GDP, education index, improved water coverage, and improved
sanitation in 108 developing countries during 2006-2010 based on panel data analysis.
The empirical results indicate that there is a positive correlation between life expectancy
rate and all selected indicators.
Within these framework, this paper tries to answer what factors determine life
expectancy as a key elemet for the nation’s health status for BRICS and MIKTA groups
JANUS.NET, e-journal of International Relations
e-ISSN: 1647-7251
Vol. 12, Nº. 2 (November 2021-April 2022), pp. 175-188
A Kuznets adaptive approach to life expectancy at birth: an application on rising powers
Hüseyin Ünal, Hülya Kınık
180
based on panal data analysis as a parallel of Kuznetz’ theoretical model. Several authors
have examined Kuznets’ hypothesis with cross-country data, but these studies have
generally tested the validity of this hypothesis analyzing relationship between income
growth and environmental pollution. Therefore, this study will make a significant
contribution to the scarce literature which have tested the validity of Kuznets’ hypothesis
in health. Nevertheless, country selection is another substantial contribution for existing
literature on rising major powers. Based on the literature review, the data used in this
study were listed below and they were all obtained from World Bank’s website.
− LE = Life expectancy at birth, total (years)
− GDP = Real GDP per capita (2010 constant)
− INF = Inflation rate (annual %)
− POPD = Population density (people per sq. km)
− HE = Current health expenditure per capita(% of GDP)
− FR = Fertility rate (births per woman)
2. Testing For a Kuznets Curve: Econometric Methodology
Simon Kuznets’ “inverted U-curve hypothesis” is among the most enduring and
significant arguments in the social sciences history. Kuznets’s central purpose was to
question if inequality in income distribution increase or decrease during a country's
economic growth (Kuznets, 2015: 1). At early stage of development little inequality is
seen in a poor country. Later, inequality worsens as income increase, but after reaching
a peak, inequality begins to decrease with more increase in growth. Present studies which
have referred to and tested a health Kuznets’ hypothesis are quite rare. While some of
the studies confirmed a Kuznets’ curve others failed to find any evidence. For instance,
Sahn and Younger (2009) examined the relation between level of well-being and
inequality at inter-country and intrahousehold levels by applying individuals’ body mass
index (BMI) as the proxy of well-being. They did not find an evidence of a quadratic curve
for BMI-inequality. Molini et al. (2010) have also explored a relationship between the
Human Development Index (HDI) and the index of concentration of BMI in developing
countries applying quadratic specifications. They found a U-shaped relation between
inequalities in BMI and HDI for Vietnam.
In the study, the Kuznet hypothesis adapted for 10 rising powers between 2000 and 2018
is tested with the following model:
(1)
where, i denotes rising powers, t denotes year 2000-2018 under observation,
denotes
constant term,
,
,
,
,
and
denote the effects of the regressors on the life
expectancy, and
denotes the error term. Besides these,
is the log-transformed
life expectancy,
is the log-transformed per-capita real GDP,
is the log-
transformed current health expenditure,
is the log-transformed population
JANUS.NET, e-journal of International Relations
e-ISSN: 1647-7251
Vol. 12, Nº. 2 (November 2021-April 2022), pp. 175-188
A Kuznets adaptive approach to life expectancy at birth: an application on rising powers
Hüseyin Ünal, Hülya Kınık
181
density,
is the log-transformed fertility rate and
is inflation rate. The data on
life expectancy (year), real GDP per capita (2010 constant), fertility rate (births per
woman), current health expenditure (% of GDP), population density (people per sq. km)
and inflation rate (annual %) are achived from the World Development Indicators.
Equation (1) indicates the effects of economic growth, inflation rate, health expenditure,
population density and fertility rate on the life expectancy. We also present an
experimental model to examine whether there is a KC for life expectancy by adding
to the model. If
and
, there will be a U-shaped relationship between
life expectancy and real GDP growth rate, but If
and
, there will be a inverse
U-shaped relationship between real GDP growth rate and life expectancy (HKC valid).
2.1. Cross-sectional dependence test
It is important to test the Cross-Sectional Dependence (CD) in estimating panel data
models. If CD is checked, the estimator results can be unbiased and consistent (Pesaran,
2004; Breusch and Pagan, 1980). Therefore, cross section dependence should be
determined in panel data. LM test proposed by Breusch and Pagan (1980) is used for
panel data whose cross-section (N) dimension is smaller than the time dimension (T).
The LM test statistic is calculated as follows:
in which
denotes the correlation coefficients and calculated as follows:
For LM test, the null hypothesis is
(cross-sections are independent)
and the alternative hypothesis is
(cross-sections are dependent).
2.2. Panel unit root test
In the existing literature, according to the cross-sectional dependence, panel unit root
tests are examined under two groups as first generation and second generation. First
generation unit root tests give unreliable results in the occurrence of CD. Second
generation unit root tests are tests that are robust to CD (Pesaran, 2007; Phillips and
Sul, 2003). In this study, we use second-generation CADF (cross-sectionally augmented
ADF) and CIPS (cross-sectionally augmented IPS) unit root tests to examine the
stationarity of the series (Pesaran, 2007). The CADF regression is identfied in Eq.(4).
JANUS.NET, e-journal of International Relations
e-ISSN: 1647-7251
Vol. 12, Nº. 2 (November 2021-April 2022), pp. 175-188
A Kuznets adaptive approach to life expectancy at birth: an application on rising powers
Hüseyin Ünal, Hülya Kınık
182
Firstly, the CADF statistics are calculated for each cross-section in the panel data from
the t statistcs ratios of
in shown Eq.(4). Then CIPS statistics are computed for the
entire panel by taking the average of the CADF test statistics.
In Eq.(5) CIPS statistics values are compared with the table Critical Values (CV)
calculated by Pesaran’s Monte Carlo simulation, which tests the stationary hypotheses.
If the calculated CIPS statistic values are smaller than the table CV, the null hypothesis
that assumes the existence of the unit root is refused. If not, the null hypothesis are
accepted and the series are said not to be stationary (Pesaran, 2007: 277-278).
2.3. Slope homogeneity test
It is important to check the slope homogenity of the cross-section units in the panel data,
in the occurrence of CD. This is because, the units in the panel data can interact with
each other and slope heterogeneity may occur. Therefore, it is necessary to check slope
homogenity in order to make reliable estimation (Breitung, 2005). The first known
studies in the literature on heterogeneity with panel data were conducted by Swamy
(1970). The next the standardized dispersion statistic
and the biased-adjusted one
was proposed by Pesaran and Yamagata (2008). This statistics, which utilizes
and
, are described in the following equations
where
denotes Swamy test statistic. In the heterogeneity test, the null hypothesis is
defined as the slope coefficients are homogeneous.
2.4. The AMG estimator
This article make use of Augmented Mean Group (AMG) estimator that is immune to
slope heterogeneity and CD. The AMG estimator was proposed by Eberhardt and Teal
(2010) and Eberhardt and Bond (2009). The procedure for the AMG test is shown in
Eq.(8) and Eq.(9).
JANUS.NET, e-journal of International Relations
e-ISSN: 1647-7251
Vol. 12, Nº. 2 (November 2021-April 2022), pp. 175-188
A Kuznets adaptive approach to life expectancy at birth: an application on rising powers
Hüseyin Ünal, Hülya Kınık
183
in Eq.(8) express an OLS regression at first difference, θ and Δ represent the coefficient
of dummy variable and the first order difference operator, respectively, in Eq.(9) which
indicates the estimates of
.
3. Empirical Results and Discussion
In this article, the impacts of economic growth, inflation rate, health expenditures,
population density and fertility rate on life expectancy at birth were analyzed for 10 rising
powers using panel data method during the period from 2000 to 2018. Table 1. shows
the sum of the selected countries’ statistics on the basis of these variables. Using these
data, first of all, the cross-sectional dependency of the series (since T> N) was examined
with the Breusch and Pagan (1980) LM test. According to the results of CD, the stability
of the variables was tested with the CIPS test, one of the second generation unit root
tests, and the test results were presented in Table 2. In the second step, the
heterogeneity of the slope parameters was checked with the Pesaran and Yamagata
(2008) test and the results were summarized in Table 3. In the final phase, the
relationship between the series was estimated by using the AMG estimator which is
resistant to the CD and the heterogeneity of the slope coefficient, and the results were
given in Table 4.
Table 1- Summary statistics of BRICS and MIKTA countries
Country
LE
GDP
HE
POPD
FR
INF
Australia
81.360
51209.913
8.449
2.838
1.837
2.694
Brazil
73.130
10468.200
8.393
23.122
1.901
6.495
China
74.127
4361.510
4.443
141.673
1.632
2.196
India
66.157
1334.378
3.678
407.740
2.697
6.363
Indonesia
68.742
3073.312
2.691
132.004
2.460
6.788
Korea, Rep.
79.656
22021.771
5.865
508.278
1.195
2.525
Mexico
75.005
9545.083
5.618
57.856
2.390
4.638
Russian Federation
68.465
9996.949
5.100
8.790
1.498
10.727
South Africa
57.844
7048.591
7.355
41.959
2.568
5.360
Turkey
74.010
11091.942
4.769
93.621
2.215
16.364
Descriptive Statistics
Mean
71.850
13015.170
5.636
141.788
2.039
6.415
Median
72.760
9139.397
5.237
73.542
2.099
4.920
Maximum
82.749
56864.330
9.467
529.359
3.311
54.915
Minimum
53.444
826.593
1.909
2.493
0.977
-0.732
Standard Deviation
6.858
14041.350
1.916
166.825
0.506
6.923
Skewness
-0.679
1.995
0.243
1.314
-0.048
4.668
Kurtosis
3.285
5.999
2.030
3.211
2.293
30.717
JANUS.NET, e-journal of International Relations
e-ISSN: 1647-7251
Vol. 12, Nº. 2 (November 2021-April 2022), pp. 175-188
A Kuznets adaptive approach to life expectancy at birth: an application on rising powers
Hüseyin Ünal, Hülya Kınık
184
Table 1. illustrates that Australia has the highest value in terms of life expectancy, per
capita income and health expenditure, and has the lowest value in terms of population
density. Although Korea has the lowest fertility rate, it is the country with the highest
population density. Turkey and then Russia differ greatly from other countries in terms
of high inflation rates. Except South Korea, Australia and China, all countries in the table
have inflation rates above the world average. While India draws attention as the country
with the highest fertility rate, it generaly has the latest place among the given countries
in terms of other variables. India’s life expectancy is less than other countries in the
table.
Table 2- Cross-section dependence and panel unit root tests results
Breusch-Pagan LM [p-value]
CIPS-stat. (level)
InLE
646.60*** [0,000]
-2.599***
InGDP
725.02*** [0,000]
-2.671***
InGDP
2
723.02*** [0,000]
-2.597***
InHE
266.68*** [0,000]
-2.575***
InPOPD
702.39*** [0,000]
-3.089***
InFR
464.53*** [0,000]
-2.316**
INF
107.81*** [0,000]
-2.898***
Notes: ** and *** denote at the 5% and 1% significance levels, respectively. Critical values
for the CIPS test are -2.560, -2.290 and -2.150 at 1,5, and 10 percent at level, respectively.
According to the Breusch-Pagan LM outcomes presented in Table 2., the null hypothesis
is refused and alternative hypothesis, which states that there is CD was accepted.
Therefore, it was decided that there is a CD between units. The CIPS statistics, used in
the occurrence of the CD, presented on the right of Table 2. demonstrated that all
variables are stationarity at levels.
Table 3- Slope heterogeneity test results
Slope homogeneity
Test statistics
p-value
5.556***
0.000
7.301***
0.000
Note: *** indicates 1% significiance level
The homogeneity tests of the slope coefficients were checked by Pesaran and Yamagata
(2008) test. According to all standardized dispersion
and biased-adjusted
statistics given in Table 3., the null hypothesis which assumes that the slope
coefficients are homogeneous is rejected at the 1% significance level. Accordingly, we
can conclude that the slope coefficients of the panel data used in the study are
heterogeneous.
JANUS.NET, e-journal of International Relations
e-ISSN: 1647-7251
Vol. 12, Nº. 2 (November 2021-April 2022), pp. 175-188
A Kuznets adaptive approach to life expectancy at birth: an application on rising powers
Hüseyin Ünal, Hülya Kınık
185
Table 4-Panel AMG parameter estimation results for Life Expectancy
Country
InGDP
InGDP
2
InHE
InPOPD
InFR
INF
HKC
Australia
17.666***
[0.000]
-0.813***
[0.001]
-0.048
[0.342]
-0.204***
[0.008]
-0.058**
[0.013]
-0.001
[0.680]
✓
Brazil
-3.428**
[0.031]
0.180**
[0.034]
-0.012
[0.272]
-0.266***
[0.000]
-0.242***
[0.000]
-0.001
[0.339]
U-
Shaped
China
0.173**
[0.014]
-0.015***
[0.000]
-0.005
[0.257]
1.060***
[0.000]
0.015
[0.917]
-0.001
[0.781]
✓
India
0.177
[0.270]
-0.015
[0.146]
0.006
[0.396]
0.412***
[0.000]
0.113***
[0.000]
-0.001
[0.257]
X
Indonesia
2.194***
[0.008]
-0.146***
[0.004]
0.002
[0.674]
0.308***
[0.008]
-0.391**
[0.014]
-0.001
[0.385]
✓
Korea, Rep.
2.144***
[0.000]
-0.108***
[0.000]
0.012
[0.361]
-0.164
[0.391]
-0.022***
[0.000]
-0.001
[0.610]
✓
Mexico
-7.041***
[0.006]
0.385***
[0.006]
0.031***
[0.002]
-0.917***
[0.000]
-0.559***
[0.000]
-0.001***
[0.006]
U-
Shaped
Russian
Federation
-3.071***
[0.006]
0.172***
[0.005]
0.049**
[0.029]
0.151
[0.665]
-0.025
[0.234]
-0.001
[0.643]
U-
Shaped
South
Africa
-51.547***
[0.000]
2.906***
[0.000]
0.082
[0.170]
0.769***
[0.000]
0.491*
[0.053]
-0.001
[0.498]
U-
Shaped
Turkey
-1.994***
[0.002]
0.108***
[0.002]
0.009
[0.358]
-0.118
[0.134]
-0.144*
[0.078]
-0.001***
[0.000]
U-
Shaped
Panel
-4.473
[0.427]
0.266
[0.392]
0.013
[0.258]
0.103
[0.565]
-0.082
[0.363]
-0.001***
[0.000]
X
Notes: ***, ** and * denote at the 1, 5 and 10 percent at levels, respectively. INF coefficients
is taken as -0.001 because the parameters are less than -0.001 in models.
Table 4. shows that HKC hypothesis is valid in Australia, China, Indonesia and Korea
suggesting that life expectancy increases with economic development up to a turning
point, while economic growth continues to increase, life expectancy begins to drop after
this turning point. On the other hand, there is a U-shaped relationship between life
expectancy and economic growth for Brazil, Mexico, Russian Federation, South Africa and
Turkey. In other words, for these countries, as economic growth increases life expectancy
decreases up to a turning point and then these variables begin to increase together.
Population density and fertility rate positively affect life expectancy at birth for India, and
there is no relationship between economic growth and life expectancy at birth.
According to the results in Table 4., health expenditures positively affect the life
expectancy only in Mexico and the Russian Federation, and there is no relationship
between the given variables in other countries. The results also demonstrate that the
fertility rate has a negative effect on life expectancy in general. The inflation rate
coefficient is negative for each country, but this variable does not affect life expectancy
in the given period except Turkey and Mexico. While the population density negatively
affects life expectancy in Australia, Brazil and Mexico, it has a positive effect on life
expactancy in China, India, Indonesia and South Africa.
Conclusion
Life expectancy is one of the most important indicators of health and community well-
being used to measure the general health status of the population. It is an identifiable
measure of the overall mortality level of a given population over a given period of time
and it is often used to compare health status disparities between countries. Life
expectancy is also an indicator of a country’s economic and social development. There
JANUS.NET, e-journal of International Relations
e-ISSN: 1647-7251
Vol. 12, Nº. 2 (November 2021-April 2022), pp. 175-188
A Kuznets adaptive approach to life expectancy at birth: an application on rising powers
Hüseyin Ünal, Hülya Kınık
186
are several studies aimed at revealing the level of life expectancy and the variables
affecting it. Some of these studies aimed to analyze the trends in life expectancy over
time; some of them aimed to compare the health status of countries; others aimed to
examine the relationship between life expectancy and variables that affect life
expectancy. The identification of the factors that affect LE is expected to contribute to
the planning of future health resources and services. In addition, learning more about
the relationships between these two variables is significant for policy implementations
for governments to cope with challenges resulted from increasing life expectancy.
In this study, we examine the existence of a quadratic relationship between economic
growth and life expectancy at birth for BRICS and MIKTA countries and test for a health
Kuznets’s curve and which has been widely overlooked in the literature by using panel
data method. The results of the AMG model applied in the study suggest that the
relationship between economic growth and life expectancy seems to fit a Kuznets’ curve
for Australia, China, Indonesia and Korea. On the other hand, the validity of HKC
hypothesis could not be obtained for Brazil, Mexico, Russian Federation, South Africa and
Turkey. There is a U-shaped relation between economic growth and life expectancy at
birth for these countries. We did not find any evidence of a quadratic curve for India,
which means there is no quadratic relationship between economic growth and life
expectancy at birth. In this case, it is thought that there may be a linear relationship
between economic growth and life expectancy at birth for India in the examined period,
and further studies are recommended to reveal what extent and how exactly these or
other factors affect life expactancy in major rising powers.
References
Ali, A., & Ahmad, K. (2014). The impact of socio-economic factors on life expectancy for
sultanate of Oman: An empirical analysis. Middle-East Journal of Scientific Research, 22
(2): 218-224.
Balan, C. and Jaba, E. (2011). Statistical analysis of the determinants of life expectancy
in Romania. Romanian Journal of Regional Science, 5 (2): 26-38.
Bayın, G. (2016). Determination of factors affecting life expectancy at birth and at age
65. Turkish Journal of Family Practice, 20 (3): 93-103.
Bilas, V., Franc, S. and Bosnjak, M.(2014). Determinant factors of life expectancy at birth
in the european union countries. Coll. Antropol. 38 (1): 1–9.
Breitung, J. (2005). A parametric approach to the estimation of cointegration vectors in
panel data. Econometric Reviews, 24 (2): 151-173.
Breusch, T. S., & Pagan, A. R. (1980). The Lagrange multiplier test and its applications
to model specification in econometrics. The review of economic studies, 47(1): 239-253.
Canning, D. (2012). Progress in Health around the World. The Journal of Development
Studies, 48 (12): 1784-1798.
Eberhardt, M., and Bond, S.( 2009). Cross-section dependence in nonstationary panel
models: A novel estimator. MPRA Paper 17180:1–28.
JANUS.NET, e-journal of International Relations
e-ISSN: 1647-7251
Vol. 12, Nº. 2 (November 2021-April 2022), pp. 175-188
A Kuznets adaptive approach to life expectancy at birth: an application on rising powers
Hüseyin Ünal, Hülya Kınık
187
Eberhardt, M., Teal, F. (2010). Productivity Analysis in Global Manufacturing Production.
Discussion Paper 515. Department of Economics, University of Oxford. Available at:
https://ora.ox.ac.uk/objects/uuid:ea831625-9014-40ec-abc5-516ecfbd2118. (10
September, 2020)
Erdoğan, S. And Bozkurt, H. (2008). The Relation Between Life Expectation And Economic
Growth in Turkey: An Analyse With ARDL Model. The Journal of Knowledge Economy &
Knowledge Management, 3, 25-38
Grossman, M. (1972). On the concept of health capital and the demand for health.
Journal of Political economy, 80 (2): 223-255.
Gulis, G. (2000). Life expectancy as an indicator of environmental health. European
journal of epidemiology, 16 (2): 161-165.
Hassan, F., Minato, N., Ishida, S. And Nor, N. (2016). Social environment determinants
of life expectancy in developing countries: a panel data analysis. Global Journal of Health
Science, 9 (5): 105-117.
Husain, A. R. (2002). Life expectancy in developing countries: a cross-section analysis.
The Bangladesh Development Studies, 28 (1/2): 161-178.
Jaba, E., Balan, C. And Robu, I. (2014). The relationship between life expectancy at birth
and health expenditures estimated by a cross-country and time-series analysis. Procedia
Economics and Finance, 15, 108-114.
Kabir, M. (2008). Determinants of Life Expectancy in Developing Countries. The Journal
of Developing Areas. 41(2) :185-204.
Kalediene, R., and Petrauskiene, J. (2000). Regional life expectancy patterns in Lithuania.
The European Journal of Public Health, 10 (2): 101-104.
Kuznets, S. (1955). Economic growth and income inequality. American Economic Review,
45 (1): 1–28.
Memarian, E. (2015). The relationship between health care expenditure, life expectancy
and economic growth in Iran. J. Appl. Environ. Biol. Sci., 5 (10): 284-290.
Molini, V., Nube, M., and van den Boom, B. (2010). Adult BMI as a health and nutritional
inequality measure: Applications at macro and micro levels. World Development, 38 (7):
1012-1023.
Monsef, A. And Mehrjardi, A. (2015). Determinants of life expectancy: a panel data
approach. Asian Economic and Financial Review, 5 (11): 1251-1257.
OECD (Organisation for Economic Co-operation and Development) (2010). Health at a
Glance: Europe 2010. OECD Publishing
Pesaran, H. M. (2004). General diagnostic tests for cross-sectional dependence in
panels. University of Cambridge, Cambridge Working Papers in Economics, 435.
Pesaran, M. H. (2007). A simple panel unit root test in the presence of crosssection
dependence. Journal of applied econometrics, 22 (2): 265-312.
Pesaran, M. H., & Yamagata, T. (2008). Testing slope homogeneity in large
panels. Journal of econometrics, 142(1): 50-93.
JANUS.NET, e-journal of International Relations
e-ISSN: 1647-7251
Vol. 12, Nº. 2 (November 2021-April 2022), pp. 175-188
A Kuznets adaptive approach to life expectancy at birth: an application on rising powers
Hüseyin Ünal, Hülya Kınık
188
Phillips, P. C., & Sul, D. (2003). Dynamic panel estimation and homogeneity testing under
cross section dependence. The Econometrics Journal, 6 (1): 217-259.
Preston, S. H. (1980). Causes and Consequences of Mortality Declines in Less Developed
Countries during the 20th Century in Richard A. Easterlin (ed), Population and Economic
Changes in Less Developing Countries. University of Chicago Press, 289-360.
Rogers, R. G., and Wofford, S. (1989). Life expectancy in less developed countries:
socioeconomic development or public health?. Journal of Bio sociological Science, 21 (2):
245-252.
Sahn, D. E. And Younger, S. D. (2009). Measuring intra-household health inequality:
Explorations using the body mass index. Health Economics, 18 (1): 13–S36.
Sede, P. And Ohemeng, W. (2015). Socio-economic determinants of life expectancy in
Nigeria (1980 -2011). Health Economics Review, 5 (2): 1-11.
Swamy, P. A. (1970). Efficient inference in a random coefficient regression
model. Econometrica: Journal of the Econometric Society, 38 (2): 311-323.
Şahbudak, E. and Şahin, D. (2015). Analysis of Relationship Between Health and
Economic Growth: A Panel Regression Analysis on BRIC Countries. Journal of Business
and Economics Studies, 3 (4): 154-160
Yavari, K. and Mehrnoosh, M. (2006). Determinants of life expectancy: A cross – country.
Iranian Economic Review, 11 (15): 13-142.
