See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/283180578

Environment and air pollution like gun and bullet for low income countries:
war for better health and wealth

Article  in  Environmental Science and Pollution Research · October 2015

DOI: 10.1007/s11356-015-5591-3

CITATIONS READS

15 149

4 authors, including:

Muhammad Azam Khan Khalid Zaman

Abdul Wali Khan University Mardan University of Wah
95 PUBLICATIONS   1,004 CITATIONS    221 PUBLICATIONS   2,514 CITATIONS   

SEE PROFILE SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Malaysia Renewable Energy Potential View project

Sustainablity View project

All content following this page was uploaded by Muhammad Azam Khan on 29 October 2015.

The user has requested enhancement of the downloaded file.

Environ Sci Pollut Res
DOI 10.1007/s11356-015-5591-3
RESEARCH ARTICLE
Environment and air pollution like gun and bullet for low-income
countries: war for better health and wealth
Xiang Zou 1 & Muhammad Azam 2 & Talat Islam 3 & Khalid Zaman 4
Received: 20 January 2015 / Accepted: 8 October 2015
# Springer-Verlag Berlin Heidelberg 2015
Abstract The objective of the study is to examine the impact
of environmental indicators and air pollution on “health” and
“wealth” for the low-income countries. The study used a num-
ber of promising variables including arable land, fossil fuel
energy consumption, population density, and carbon dioxide
emissions that simultaneously affect the health (i.e., health
expenditures per capita) and wealth (i.e., GDP per capita) of
the low-income countries. The general representation for low-
income countries has shown by aggregate data that consist of
39 observations from the period of 1975–2013. The study
decomposes the data set from different econometric tests for
managing robust inferences. The study uses temporal forecast-
ing for the health and wealth model by a vector error correc-
tion model (VECM) and an innovation accounting technique.
The results show that environment and air pollution is the
menace for low-income countries’ health and wealth. Among
environmental indicators, arable land has the largest variance
to affect health and wealth for the next 10-year period, while
air pollution exerts the least contribution to change health and
Responsible editor: Philippe Garrigues
* Khalid Zaman
Khalid_zaman786@yahoo.com
Talat Islam
talatislam@yahoo.com
1
Institute of Graduate Studies, University of Malaya, 50603 Kuala
Lumpur, Malaysia
2
School of Economics, Finance and Banking, College of Business,
Universiti Utara Malaysia, Sintok, Malaysia
3
Hailey College of Banking & Finance, University of the Punjab,
Lahore, Pakistan
4
Department of Economics, University of Sargodha, Canal Campus,
Lahore, Pakistan
wealth of low-income countries. These results indicate the
prevalence of war situation, where environment and air pollu-
tion become visible like “gun” and “bullet” for low-income
countries. There are required sound and effective macroeco-
nomic policies to combat with the environmental evils that
affect the health and wealth of the low-income countries.
Keywords Environment . Air pollution . Health . Wealth .
Low-income countries
Introduction
The ultimate ambition of every state is to achieve the desired
level of sustainable economic growth and development and to
sustain it for a long term. Usually, low-income countries face
numerous cultural, socio-economic, institutional, and political
problems and difficulties during the process of economic
growth and development. When talking about sustainable eco-
nomic growth and development, the environmental degradation
and environmental pollution cannot be ignored. Environmental
pollution has many bad consequences because it directly atten-
uates aggregate output and largely reduces productivity of cap-
ital and labor. The ultimate adverse results and the long-term
harmful impact of environmental effluence can add to damag-
ing the social welfare and the economy as well. Consequently,
it will lead to enlargement in health and therefore social costs.
Somehow, the term sustainable development or green
economy or green growth or low carbon development is used
more or less similar. In this context, by theWCED (1987: 57)
sustainable development means “development that meets the
needs of the present without compromising the ability of fu-
ture generations to meet their own needs.” The UNEP (2010:
29) explains it as “one that results in improved human well-
being and social equity, while significantly reducing

environmental risks and ecological scarcities.” According to
OECD (2011: 17) “Green growth is about fostering economic
growth and development while assuring that natural assets
continue to provide the resources and environmental services
on which our well-being relies. To do this it must catalyze
investment and innovation which will underpin sustained
growth and give rise to new economic opportunities.” Simi-
larly, low carbon growth and development is refers to when
economic growth is achieved without industries’ emissions by
adopting lower carbon technologies. It also alleviates poverty
and strengthens equal social justice (Sweeney 2008; Lutken
et al. 2011; Urban and Nordensvard 2013).
In fact, energy consumption is considered one of the deter-
minants of economic growth like other factors, namely capital,
labor, etc., because there is a close association between energy
use and economic growth. At the same time, the extensive
usage of energy results in augmenting greenhouse gas
(GHG) emissions. On the other hand, environmental deterio-
ration acts as a visible hazard to human health globally. Det-
rimental results of environmental degradation to human health
have been observed, and it can even grow extensively worse
over the coming 50 years (Millennium Ecosystem Assessment
2005). The harmful consequences of pollution also generate
when the industrial sector produces dangerous chemicals and
toxic waste into water where all these severely affect human
health. Similarly, in the agricultural sector, the use of insecti-
cides, pesticides, and fertilizers can also have an adverse im-
pact on human health through many channels including drink-
ing water becoming contaminated and polluted. Environmen-
tal degradation is a consequence of many economic and social
activities including urbanization, population growth, econom-
ic growth, extension of agriculture, increase in energy con-
sumption, and upsurge in transportation.
It has been observed that there are several factors explaining
environmental degradation, land degradation, and water pollu-
tion including crowding and traffic, industrialization, over cut-
ting of wood, flooding, overgrazing, transportation, fast grow-
ing population, poverty, soil erosion, and misuse or extensive
use of open-access available scarce resources.1 All these gen-
erate many problems including air pollution, noise, population
density, urban disease, and water pollution. Also, there is an
extensive belief that environmental quality is a main factor of
health and well-being, implying that these factors affect health
and health-associated behavior.2 Cohen et al. (1979) notes that
population density and environmental noise have negative in-
fluence on behavior and health. Moreover, switch over of cul-
tivable land and timberlands to urban development decreases
the amount of lands offered for foodstuff and timber produc-
tion. Desertification, soil erosion, salinization, and extra soil
dilapidations related with thorough agriculture and cutting of
1
See also (Khan et al. 2009)
2
Moss (1976) and Cohen et al.(1979)
Environ Sci Pollut Res
wood adversely affect the quality of land resources and upcom-
ing agricultural yield (Lubowski et al. 2006; Wu 2008). Esti-
mates reveal that almost 75 % of all carbon dioxide (CO2)
emissions are produced by developed economies. Nevertheless,
economic growth depends on energy consumption that gener-
ates environmental pollution. On the other hand, it has been
observed that at the initial stage of economic growth it may
deteriorate the environment, but later on, with the implementa-
tion of improved abatement technologies, it might bring some
development to the environmental quality (Grossman and
Krueger 1995; Hitz and Smith 2004).
The study of Shanmugam and Hertelendy (2011) reveals
that the developed countries contaminate the air by their too
much usage of fossil fuels and coal energies in the industrial
sector, as well as oil refineries and auto emissions. Developing
countries desire to catch up with the developed countries’ po-
sition, therefore contributing to air pollution by their augmented
struggles to consume exceptionally the fossil fuels and coal
energies. The usage of saltpeter, urea, ammonia, and nitric ox-
ides in agricultural sectors/activities may add up largely to air
pollution. The cultivation and farming activities along with
fertilizers, insecticides, pesticides, and soil surface erosions in
the developing countries are also visible sources of air pollu-
tion. Similarly, CO2 from coal consumption is a strong source
of air pollution. Therefore, more preventive worldwide air pol-
lution policies would largely make better global public health.
Generally, it is believed that global warming is often ac-
knowledged by scientists across the world as a stern public
health and environmental apprehension. The most important
impact of air pollution is the climate change. The fundamental
mechanism of global warming is that where GHGs are in-
volved, which consist of CO2, methane, nitrous oxide, and
ozone (Houghton et al. 1992). Air pollution severely affects
human health (QUARG 1993). In a study, Banister (1998)
shows that environmental destruction and damage in China
is due to China’s population growth rate. The study of Pope
(2007) reveals that the air we inhale comprises emissions from
several different sources including heating and commercial
sources, household fuel industry, motor vehicles, and smoking
of tobacco. The adverse effects of air pollution has been stud-
ied by prior studies, where the findings of those studies reveal
that air pollution damages human health and mostly is dam-
aging for children and older people or those suffering from
health problems. Air pollution directly affects the respiratory
system because air pollution in the form of dust pollutants3
plays a visible role in the environmental health risk. The
World Health Organization expounds that the air pollution
leads to enlarged mortality rate and condensed life expectancy
(Samet and Krewski 2007). The severe impact of climate
3
More visible pollutants are nitrogen oxide, sulfur dioxide, carbon mon-
oxide, carbon dioxide, ammonia, methane, particulate matter, and non-
methane volatile organic compounds.
Environ Sci Pollut Res
change on health can be seen from many ways, for example,
as a result of enlarged frequency and intensity of heat waves,
decrease in cold-associated deaths, more floods and droughts,
and impacts on the risk of adversities and malnourishment.
The inclusive climate change effect on health is probably go-
ing to be harmful, and populations in poor countries are ex-
pected to be mostly susceptible to the negative effects. The
study suggests that easing of climate change by decreasing the
consumption of fossil fuels and increasing the adoption of the
renewable energy technologies is supposed to improve health
shortly by diminishing air pollution (Haines et al. 2006).
It is evident that nitrous oxide (from meadow land and
arable land usage to grow feed crops) and methane (from the
digestive activities of ruminant animals like cows and sheep)
account for almost 80 % of overall agricultural GHG emis-
sions. However, per unit of livestock goods, emissions vary by
animal type and seem to be relatively greater in beef, sheep,
and dairy farming compared to pig and poultry farming
(Williams et al. 2006). In a study, McMichael et al. (2007)
notes that globally, the agricultural sector, particularly live-
stock production, is responsible for around a fifth of total
GHG emissions, therefore contributing to climate change
and badly affecting health, as well as imposing a threat to food
produces in several regions. It is estimated that GHG emis-
sions from the agriculture activity account for approximately
22 % of worldwide total emissions; this share is parallel to that
of the industrial sector and bigger than that of the transport
sector. Furthermore, estimates indicate that livestock produc-
tion (comprising transport of livestock and feed) accounts for
about 80 % of the sector’s releases. The study of Friel et al.
(2009) shows that agrarian food output and agriculturally rel-
evant change in land usage extensively contribute to GHG
emissions globally; it has been observed that four-fifths of
agricultural emissions result from the animal farm/livestock
sector. Even though livestock goods are a source of some
indispensable nutrients, they provide huge amounts of
drenched fat, which is a considered risk factor for cardiovas-
cular illness. Approximately half of all food-related GHG
emissions are arising during agriculture activities. Emissions
during agriculture activities consist of nitrous oxide and meth-
ane from livestock sector, and carbon dioxide from agricultur-
ally stimulated change in land usage, particularly deforestation
(UN Food and Agriculture Organization 2006; McAlpine
et al. 2009).
The OECD (2011) notes that everyone knows that the
world economy is gradually and unfairly coming out of the
severe crisis. However, there are several problems including
high unemployment, inflation, or fiscal deficits; therefore, ap-
propriate policies are required that confirm the desired level of
growth and progress. On the other hand, reappearance to
“business as usual” would certainly be imprudent and
eventually unjustifiable, including risks that could enforce
human costs and restraints on the process of economic
growth and development. Ultimately, it enlarged water
shortage, resource bottlenecks, air and water pollution,
climate change, and biodiversity damage which would be
permanent. Therefore, prudent plans are required in order to
achieve greener growth. The study of Omoju (2014) expounds
that pollution is the predominant environmental challenge the
world is facing. The effect of pollution is relatively more se-
rious in developing nations, leading to bad health, death, and
disabilities of millions of people per annum. On the other
hand, advanced countries have abundance resources and
technologies for controlling pollution and have little effect
on health compared to developing countries. Bozkurt and
Akan (2014) show that for every country, factors including
upsurge in GHG and CO2 emissions, global warming, and
climate changes are considered the main threats, because
these factors badly affect economic growth. Therefore, CO2
emissions should be condensed in order to avoid this huge
environmental degradation and pollution.
The World Energy Outlook (2014) indicates that about 1.3
billion people have yet no access to electricity and 2.7 billion
people depend on the traditional usage of biomass for cooking
(i.e., dung, crop residues, wood, and charcoal) which causes
destructive indoor air pollution. The ratios of these people are
mostly in developing Asia or sub-Saharan Africa and in
countrysides.
Therefore, the central goal of the present study is to examine
empirically the impact of various factors contributing to envi-
ronmental degradation, affecting human health, and conse-
quently attenuating aggregate output relatively more in low-
income countries. Nevertheless, there are many environmental
degradation factors, but this study confines to factors such as
arable land, population density, carbon dioxide emissions, and
fossil fuel energy consumption. We hypothesized in this study,
as it is also evident from the title of the study that environment
and air pollution are like gun and bullet for the development
process of low-income countries. To the best of our knowledge,
of course, this is the first inclusive study considering a different
set of response and explanatory variables. Therefore, the out-
comes are expected to help the decision makers of low-income
countries in reducing pollutants maximally in order to mitigate
health risk and bolster sustainable growth and development.
The reminder of the study is design as follows. The section
“Literature review” deals with previous studies. “Data source
and methodology” section presents data description and em-
pirical methodology. The section “Results” interprets the re-
sults. The section “Conclusion” concludes the study with brief
policy implications.
Literature review
Several prior studies theoretically as well as empirically eval-
uates the effect of environmental degradation on human health

and economic growth using primary and secondary data,
different empirical methodologies, time period, and single or
panel of countries. For example, on the basis of a study of 75
community areas in Chicago, in a study, Galle et al. (1972)
observe small but statistically significant relationships be-
tween population density and five social pathology
variables, namely deaths, fertility, public support to minors,
juvenile crime, and mental hospital admissions. Using data
from 125 geographic areas in the Netherlands, the study of
Levy and Herzog (1974) analyzes the influences of population
impaction such as areal density and crowding upon many
indices of human health and social adaptation. The
multivariate regression results strongly support the positive
impact of density on increased pathology, while the impact
of crowding is less and, excluding mental hospital
admissions, has an inverse weight in the estimated
regression equations. The study of Simon (1979) reveals that
higher population density suggests rapid economic growth in
developing countries, where the empirical results obtained are
economically and statistically significant for cross-sections of
less developed countries over the periods of 1960–1970,
1950–1960, and 1950–1970. However, population size shows
no visible effect. The study extends that an empirical finding
of positive impact of population density indicates that in the
long run, population growth has a positive impact on econom-
ic growth (per capita income).
In a study, Udofia and Udom (2011) examine the nature
of the linkage between environmental degradation and the
socio-economic position of an oil-bearing community in the
Niger Delta Coast of Akwa Ibom State, Nigeria. The empir-
ical findings indicate that soil degradation has no significant
impact on the socio-economic life, whereas, among the wa-
ter factors, just the Minor Nutrient factor (potassium) influ-
ences considerably the socio-economic life of the people
through effects on fishing which is a major means of liveli-
hood of the Nigerian society. Zaman et al. (2011) study
indicates that high population growth has a detrimental ef-
fect on environment quality in three selected South Asian
Association for Regional Cooperation (SAARC) countries,
namely India, Pakistan, and Sri Lanka during 1985–2009.
The study further adds that due to rapid population growth,
demand to produce further increases, so consequently, the
use of arable land also increases which has some bad envi-
ronmental consequences. The study of Borhan et al. (2012)
reveals that CO2 shows significantly negative linkage with
economic growth in eight countries from ASEAN during
1965–2006. Further, the study finds that the estimated coef-
ficient of air pollution indicator indicates a statistically sig-
nificant impact on population density, whereas CO2 emis-
sion has an inverse linkage with population density. This
implies that when air pollution upsurges, population density
declines which causes death of human beings. The empirical
findings of Chen (2001) reveals that Taiwan’s GDP will fall
Environ Sci Pollut Res
34 % off the projected GDP growth rate for the year 2000,
where Taiwan’s economy will be extremely deteriorated if,
per annum, CO2 emissions become constant at the 1990
level.
The studies of Greiner (2004; 2005) reveal that an up-
surge in GHG emissions has negative effect on the aggre-
gate production and the marginal capital productivity.
However, greater abatement goings-on could mitigate
GHG emissions and stimulate desirable level of
economic growth. However, in a study, Say and Yucel
(2006) find a close positive linkage between the total
CO2 emissions and national output during 1970–2002.
Erbaykal (2008) observes that in the short run, both oil
and electricity have significantly positive impacts on the
economic growth (GDP) of Turkey during1970–2008. Tol
(2009) also mentions that GHG emissions largely and
harshly affect the process of economic development.
Gosh (2009) examines the causal association between
income and the CO2 emissions, but the study finds no
causality between income and the CO2 emissions in the
case of India. Liu and Lin (2009) observe empirically
using data for the period from 1990–2005 that the carbon
tax on coal will be the optimum strategy for China be-
cause high CO 2 causes a great loss on GDP growth.
Bhusal (2010) study suggests that in the short and long
run, there exists two-way Granger causality between oil
usage and economic growth in Nepal during 1975–2009.
Tiwari (2011) finds that CO2 emissions have a significant-
ly positive effect on energy consumption and capital but a
negative effect on population and economic growth in
India during 1971–2007.Akram (2012) finds that econom-
ic growth is inversely affected by changes in growth rate
in population, while urbanization growth encourages eco-
nomic growth of selected Asian countries during 1972–
2009. Alkhathlan et al. (2012) find that the estimated co-
efficients of CO2 emissions along with some other control
variables have significantly positive effects on GDP in the
long run for Saudi Arabia during 1980–2008.Farhani and
Rejeb (2012) find that there exists a long-run equilibrium
connection between energy use, real income or GDP, and
the CO2 emissions in 15 selected Middle East and North
African (MENA) countries during 1973–2008.
In a similar way, the study of Ong and Sek (2013) mentions
that results of both approaches, i.e., panel and non-panel, in-
dicates no or low relation between environmental quality mea-
sured by CO2 emissions and economic growth in three income
groups, namely low-, middle-, and high-income countries dur-
ing 1970–2008.The empirical results of Bozkurt and Akan
(2014) study show that the effect of CO2 emissions is signif-
icantly negative on economic growth in Turkey during 1960–
2010. Alam et al. (2014) expounds that environmental efflu-
ence is one of the crucial issues in the sustainable economic
growth process everywhere. The empirical findings of the
Environ Sci Pollut Res

study suggests that population density has a negative associa-
tion with the financial indicators in Malaysia during 1975–
2013. In a study of 54 countries, Omri et al. (2014) find the
existence of one-way causality running from CO2 emissions
to economic growth, with the exclusion of the Middle East,
North Africa, and Sub-Saharan, where two-way causality be-
tween CO2 emissions to economic growth cannot be refused
during 1990–2011. The empirical findings of Alshehry and
Belloumi (2015) study indicate a long-run two-way causality
between CO2 emissions and economic growth in Saudi Arabia
over the period of 1971–2010. In a similar study, Saidi and
Hammami (2015) find the CO2 emissions have significantly
negative effect on economic growth in a panel of 58 countries
over the period 1990–2012. The study maintains that the ef-
fect of economic growth on the environment depends on en-
ergy emission type; for instance, carbon monoxide, CO2 emis-
sions and sulfur dioxide have unfavorable effects on human
health and the environment.
Data source and methodology
The data of environmental indicators, air pollution, health
expenditures, and economic growth were taken from
World Development Indicators published by World Bank
(2014). According to Armstrong (2001), the disaggre-
gated data are noisier than aggregates constructed from
them. Consequently, disaggregated series appear harder
to forecast. Therefore, the study prefers to use aggregate
data of low-income countries for evaluating the conjunc-
tion between environment, air pollution, health, and
wealth over the period of 1975–2013. There are a number
of proxies used for the variables; i.e., environmental indi-
cators include arable land in hectares per person, fossil
fuel energy consumption in total percentage, and popula-
tion density in people per square kilometers of land area.
Air pollution is represented by carbon dioxide emissions
in kiloton, health variable is presented by health expendi-
ture per capita in current US dollars, and wealth is shown
by GDP per capita in current US dollars.
The study adopted the following sequential orders for the
methodology:
1. Descriptive statistics: It is used to assess the trend of the
individual variables that includes mean value, maximum
value, minimum value, standard deviation, and skewness
and kurtosis of the variables.
2. Covariance, correlation, and sum of squares and cross
products (SSCP) matrix: These matrices are used to de-
scribe the covariance between the variables, while correla-
tion matrix is used to measure the magnitude and directions
between the variables. An SSCP matrix is used to evaluate
deviation score sums of squares between the variables.
3. Centered long-run covariance test by Bartlett Kernel,
Newey-West fixed bandwidth=4.0000: This test is the
modified version of the simple covariance, in which
Bandwidth is used up to 4.0000.
4. Test for equality of means between series by ANOVA and
Welch F test: These tests assumed that when the variances
across groups are heterogeneous, the usual analysis of
variance assumptions are not satisfied and the ANOVA
F test is not valid. The study adopted the Welch F test
which satisfied the likelihood-ratio test under normal
distributions.
5. Engle-Granger cointegration test (Single equation
cointegration test): Single cointegration test is used to
evaluate whether the series exhibit the long-run relation-
ship between them or there is a need to evaluate the time
series cointegration test, i.e., Johansen’s cointegration test
for multiple equations.
6. Unit root test: The study employed Augmented Dickey-
Fuller (ADF) unit root test (Dickey and Fuller 1979,
1981) to check the stationary series of the candidate
variables.

Table 1 Descriptive statistics

Statistics ALAND CO2 FFUEL GDPPC HEXP PD

(hectares per person) (kt) (% of total energy) (current US$) (current US$) (people per sq. km of land area)

Mean 0.231 231,441.1 33.020 321.660 12.822 37.730

Maximum 0.319 345,948.0 40.216 721.967 31.589 57.118
Minimum 0.180 155,103.2 28.458 192.552 7.125 22.707
Std. Dev. 0.043 63,697.35 3.853 134.109 7.307 10.305
Skewness 0.570 0.609 0.312 1.767 1.556 0.225
Kurtosis 1.971 1.880 1.455 4.890 3.986 1.833
Observations 39 39 39 39 39 39

Source: Authors’ estimation

ALAND arable land, CO2 carbon dioxide emissions, FFUEL fossil fuel energy consumption, GDPPC GDP per capita, HEXP health expenditures, PD
population density
 Environ Sci Pollut Res

Table 2 Covariance, correlation, and SSCP matrix

Covariance
Correlation
SSCP
t Statistic
Probability
Cases
Observations
Weights ALAND CO2 FFUEL GDPPC HEXP PD

ALAND 0.002
1
0.073
–
–
39
39
39
CO2 −638.85 3,950,000,000.00
−0.23 1.00
−24,914.99 154,000,000,000.00
−1.47 –
0.15 –
39.00 39.00
39.00 39.00
39.00 39.00
FFUEL 0.120 41,632.5 14.467
0.732 0.174 1.000
4.695 1,623,668.000 564.210
6.533 1.075 –
0 0.289 –
39 39 39
39 39 39
39 39 39
GDPPC −3.224 389,685.400 −218.680 17,524.070
−0.563 0.047 −0.434 1.000
−125.727 15,197,729.000 −8528.508 683,438.900
−4.145 0.285 −2.933 –
0.000 0.777 0.006 –
39 39 39 39
39 39 39 39
39 39 39 39
HEXP −0.207 19,000.080 −15.358 936.630 52.031
−0.662 0.042 −0.560 0.981 1.000
−8.057 741,003.100 −598.969 36,528.570 2029.196
−5.377 0.255 −4.109 30.667 –
0 0.8001 0.0002 0 –
39 39 39 39 39
39 39 39 39 39
39 39 39 39 39
PD −0.414 70,726.320 −29.993 1055.609 63.206 103.471
−0.942 0.111 −0.775 0.784 0.861 1.000
−16.157 2,758,326.000 −1169.741 41,168.740 2465.028 4035.386
−17.037 0.677 −7.465 7.680 10.317 –
0.000 0.503 0.000 0.000 0.000 –
39 39 39 39 39 39
39 39 39 39 39 39
39 39 39 39 39 39

Source: Authors’ estimation

ALAND arable land, CO2 carbon dioxide emissions, FFUEL fossil fuel energy consumption, GDPPC GDP per capita, HEXP health expenditures, PD
population density
Environ Sci Pollut Res

Table 3 Centered long-run covariance

ALAND CO2 FFUEL GDPPC HEXP PD

ALAND 0.007 −2021.533 0.468 −11.282 −0.741 −1.512

CO2 −2021.533 12,964,539,073.015 163,413.239 −207,295.942 −10,140.500 162,686.768
FFUEL 0.468 163,413.239 53.722 −812.363 −57.444 −115.088
GDPPC −11.282 −207,295.942 −812.363 57,202.355 3150.336 3607.357
HEXP −0.741 −10,140.500 −57.444 3150.336 178.799 222.150
PD −1.512 162,686.768 −115.088 3607.357 222.150 373.393

Source: Authors’ estimation

ALAND arable land, CO2 carbon dioxide emissions, FFUEL fossil fuel energy consumption, GDPPC GDP per capita, HEXP health expenditures, PD
population density

7. Johansen cointegrationt: Johansen cointegration test is Model 2 Wealth Equation

used to examine the long-run relationship between the
variables. For this purpose, eigenvalue and trace statistics
are used for examining the cointegration equations be-
tween the variables. The test for cointegration used is
the likelihood ratio put forward by Johansen and Juselies
(1990), signifying that the maximum likelihood method is
more suitable in a multivariate structure.
8. Vector error correction model: Vector error correction
models (VECM) offered the convenient framework for
separating the long-run and the short- run components
of the data generation process. Therefore, the study
employed the VECM model for reliable parameter
estimates.
9. Innovation accounting time series techniques: Impulse re-
sponse function and variance decomposition technique is
used for analyzing the responses of variables to unexpect-
ed changes (i.e., shocks) in other variables, over the next
10-year time period.
By using the above methodological order, the study
estimated two different models—one is for “health” and
the other is for “wealth.” These two equations are as
follows:
Model 1 Health Equation
HEXP ¼ β 0 þ β1 ALAND þ β2 FFUEL þ β3 CO2
þ β 4 PD þ ε
Table 4 Test for equality of means between series
Method df Value
ANOVA F test (5, 228) 514.5141
Welch F test (5, 88.6732) 814.4386
GDPPC ¼ β0 þ β1 ALAND þ β2 FFUEL þ β3 CO2
þ β 4 PD þ ε ð2Þ
where HEXP represents health expenditures per capita,
GDPPC represents GDP per capita, ALAND is arable land
in hectares per person, FFUEL represents fossil fuel energy
consumption in total percentage, CO2 represents carbon diox-
ide emissions in kiloton, PD shows population density in peo-
ple per square kilometer of land area, and ε shows error term.
These two equations serve as health and wealth for low-
income countries.
Results
The result comprises descriptive statistics, parametric and
non-parametric tests, unit root, multivariate cointegration test,
and forecasting. Table 1 shows the descriptive statistics of the
variables.
The results show that arable land has a minimum value of
0.180 to maximum 0.319 hectares per person, while on aver-
age there are 0.231 hectares occupied per person with standard
deviation of 0.043. Carbon dioxide emissions have a mean
value of 231,441.1 kt with standard deviation of 63,697.35.
ð1Þ Fossil fuel energy consumption which is the percentage of
total energy consumed by minimum is 28.458 % while the
maximum value is 40.216 %. Population density shows that
on average, 37.730 people per square kilometer of land area
occupied with a standard deviation of 10.305. Economic con-
dition and health performance is worst in the low-income
Probability
countries, as the minimum value of GDP per capita is
0.0000 US$ 192 .55 2 (c urre nt), while ma ximu m valu e is
0.0000 US$721.967. On average, the economic performance of
low-income countries is about US$321.660 per capita.
 Environ Sci Pollut Res

Table 5 Engle-Granger cointegration test (single equation cointegration test)

Series: ALAND CO2 FFUEL GDPPC HEXP PD

Sample: 1 39
Included observations: 39
Null hypothesis: Series are not cointegrated
Cointegrating equation deterministics: C
Automatic lags specification based on Schwarz criterion (maxlag=9)

Dependent tau-statistic Prob.a z-statistic Prob.a

ALAND −4.131738 0.2688 −354.7172 0.0000
CO2 −1.863600 0.9846 −7.597931 0.9831
FFUEL −2.845655 0.8122 −12.60229 0.8643
GDPPC −2.891681 0.7959 −15.41873 0.7295
HEXP −4.362985 0.1958 −1327.389 0.0000
PD −4.112882 0.2754 −765.1884 0.0000
Intermediate results:
ALAND CO2 FFUEL GDPPC HEXP PD
Rho - 1 −0.598 −0.199 −0.331 −0.405 −0.688 −0.606
Rho S.E. 0.144 0.107 0.116 0.140 0.157 0.147
Residual variance 9.61E−06 9.66E+08 1.069 184.866 0.254 0.232
Long-run residual variance 0.002 9.66E+08 1.069 184.866 729.177 285.571
Number of lags 2 0 0 0 2 2
Number of observations 36 38 38 38 36 36
Number of stochastic trendsb 6 6 6 6 6 6

Source: Authors’ estimation

ALAND arable land, CO2 carbon dioxide emissions, FFUEL fossil fuel energy consumption, GDPPC GDP per capita, HEXP health expenditures, PD
population density
a
MacKinnon (1996) p values
b
Number of stochastic trends in asymptotic distribution

Subsequently, the health facility is also too low in low-income performance of health and wealth in low-income countries.
countries, where the average value of health expenditure per The study at final recommends some policy implications for
capita is US$12.822 per person. These statistics show the low escaping of low-income countries from the tsunami of

Table 6 ADF unit root results

Variables Level First difference Decision

Constant, linear trend Constant, linear trend

GDPPC 1.813 −3.794 Non-stationary at level, however, stationary at first difference

(1.000) (0.028)
HEXP −1.739 −2.106 Non-stationary at level, however, stationary at second difference
(0.713) (0.525)
ALAND 1.951 −5.358 Non-stationary at level, however, stationary at first difference
(1.000) (0.000)
FFUEL −2.133 −4.356 Non-stationary at level, however, stationary at first difference
(0.511) (0.007)
CO2 −1.602 −5.113 Non-stationary at level, however, stationary at first difference
(0.773) (0.001)
PD 0.006 −5.220 Non-stationary at level, however, stationary at first difference
(0.994) (0.000)

Source: Authors’ estimation

ALAND arable land, CO2 carbon dioxide emissions, FFUEL fossil fuel energy consumption, GDPPC GDP per capita, HEXP health expenditures, PD
population density
Note: Small bracket shows MacKinnon (1996) one-sided p values
Environ Sci Pollut Res

Table 7 Johansen cointegration

test for the “wealth” model Trend assumption: Linear deterministic trend
Series: GDPPC FFUEL ALAND CO2 PD
Lags interval (in first differences): 1 to 1
Unrestricted cointegration rank test (trace)

Hypothesized Trace 0.05

No. of CE(s) Eigenvalue Statistic Critical value Prob.b
Nonea 0.620 80.709 69.818 0.005
At most 1 0.376 44.863 47.856 0.093
At most 2 0.333 27.393 29.797 0.092
At most 3 0.232 12.380 15.494 0.139
At most 4 0.068 2.613 3.841 0.106
Unrestricted cointegration rank test (maximum eigenvalue)
Hypothesized Max-Eigen 0.05
No. of CE(s) Eigenvalue Statistic Critical value Prob.b
Nonea 0.620 35.846 33.876 0.028
At most 1 0.376 17.469 27.584 0.539
At most 2 0.333 15.013 21.131 0.288
At most 3 0.232 9.767 14.264 0.227
At most 4 0.068 2.613 3.8414 0.106

Source: Authors’ estimation. Trace test indicates 1 cointegrating eqn(s) at the 0.05 level. Max-eigenvalue test
indicates 1 cointegrating eqn(s) at the 0.05 level
ALAND arable land, CO2 carbon dioxide emissions, FFUEL fossil fuel energy consumption, GDPPC GDP per
capita, HEXP health expenditures, PD population density
a
Denotes rejection of the hypothesis at the 0.05 level (MacKinnon (1996) one-sided p values)
b
MacKinnon et al. (1999) p values

environmental hazards in the region. Table 2 shows the co-
variance, correlation, and the SSCP matrix.
The table shows three matrices, i.e., covariance, correla-
tion, and SSCP. The covariance and SSCP results were shown
just for the ready reference for the readers; however, the cor-
relation coefficient results were presented accordingly, i.e.,
there is a positive but weaker correlation between GDP per
capita and carbon emissions in low-income countries, while
fossil fuel energy consumption and arable land significantly
decrease economic growth as the correlation coefficient is
about −0.434 and −0.563, respectively. The same results ap-
peared in the case of health expenditures, where carbon emis-
sions have a positive but low correlation with health expendi-
tures, while arable land and fossil fuel energy consumption
significantly decrease health expenditures in a region. These
results somehow indicate the vulnerable situation in the low-
income countries where the health and wealth badly affected
with the environmental and pollution indicators. Table 3
shows the long-run covariance with kernel, Newey-West
Fixed Bandwidth 4.0000.
The results show that there is a negative covariation be-
tween GDP per capita and rest of the variables, except popu-
lation density that shows the positive covariance, i.e.,
3607.357 with the GDP per capita. On the other hand, health
expenditures have a positive covariation with the population
density; however, remaining variables show negative covari-
ance. The important result indicates that the magnitude of
covariance is greater in GDP per capita compared to health
expenditures per capita in a region. The results emphasize the
need of better health facilities in a region which may uplift the
economic performance of low-income countries. Table 4
shows the equality of means by Welch F test and ANOVA F
test.
Both the test statistics confirmed the equality of means
between the series as ANOVA F test and Welch F test have
significant F values. The results justify the surging inundation
of environmental degradation with the battle of health and
wealth of low-income countries. Before going to estimation
using the unit root test, the study estimates the single equation
cointegration test including all variables. Table 5 shows the
Engle-Granger cointegration test for single equation.
The results of tau statistics show that there is no
cointegration relationship between the variables; however, an-
other statistics, i.e., z statistics, indicates the significance of
ALAND, HEXP, and PD in the given model; therefore, we
may safely conclude that in a single equation cointegration
test, a long-run relationship between the variables has been
established. The rest of the statistical results summary is
 Environ Sci Pollut Res

Table 8 Johansen’s cointegration

test for the “health” model Trend assumption: Linear deterministic trend
Series: HEXP FFUEL CO2 PD ALAND
Lags interval (in first differences): 1 to 1
Unrestricted cointegration rank test (trace)

Hypothesized Trace 0.05

No. of CE(s) Eigenvalue Statistic Critical value Prob.b
Nonea 0.597 67.255 69.818 0.078
At most 1 0.343 33.574 47.856 0.525
At most 2 0.225 18.019 29.797 0.564
At most 3 0.204 8.544 15.494 0.409
At most 4 0.002 0.100 3.841 0.750
Unrestricted cointegration rank test (maximum eigenvalue)
Hypothesized Max-Eigen 0.05
No. of CE(s) Eigenvalue Statistic Critical value Prob.b
Nonea 0.597 33.680 33.876 0.052
At most 1 0.343 15.555 27.584 0.703
At most 2 0.225 9.475 21.131 0.792
At most 3 0.204 8.443 14.264 0.335
At most 4 0.002 0.100 3.8414 0.750

Source: Authors’ estimation. Trace test indicates no cointegration at the 0.05 level. Max-eigenvalue test indicates
no cointegration at the 0.05 level
ALAND arable land, CO2 is carbon dioxide emissions, FFUEL fossil fuel energy consumption, GDPPC GDP per
capita, HEXP health expenditures, PD population density
a
Denotes rejection of the hypothesis at the 0.05 level (MacKinnon (1996) one-sided p values)
b
MacKinnon et al. (1999) p values

presented in the “intermediate results” panel for the ready
reference. Next, the study estimated the stationary properties
of the candidate variables by ADF unit root test in Table 6.
The results show that except health expenditures per capita
(HEXP), all the remaining variables are non-stationary at lev-
el; however, after taking first difference, these variables be-
come stationary. The results indicate that GDP per capita,
arable land, fossil fuel energy consumption, carbon dioxide
emissions, and population density, drastically changed over a
period of time and have a wide fluctuation in the data series.
Similarly, health expenditures exhibit more fluctuation in the
data series which do not even counter with the first difference.
These results suggest that low-income countries should have
to formulate sustainable policies which aid to stabilize

Table 9 VECM estimates—

“wealth” model dependent Variables Coefficient Standard error t statistics Probability value
variable: Dlog(GDPPC)t
Constant 0.359 0.098 3.657 0.001
Dlog(ALAND)t−1 2.007 1.100 1.825 0.080
Dlog(CO2)t−1 0.013 0.060 0.225 0.823
Dlog(FFUEL)t−1 −0.049 0.299 −0.166 0.868
Dlog(PD)t−1 −11.167 4.336 −2.692 0.012
log(ALAND)t−1 1.164 0.550 2.115 0.043
log(CO2)t−1 0.014 0.050 0.291 0.772
log(FFUEL)t−1 −0.331 0.213 −1.550 0.132
log(PD)t−1 0.762 0.339 2.248 0.032
ρ −0.185 0.081 2.268 0.034

Statistical tests: R2 0.550, adjusted R2 0.348, F statistics 2.721, F prob. value 0.029. Diagnostic tests: Jarque-
Bera 1.539, LM test 2.443, ARCH tests 1.872, Ramsey RESET test 0.467
ALAND arable land, CO2 carbon dioxide emissions, FFUEL fossil fuel energy consumption, GDPPC GDP per
capita, HEXP health expenditures, PD population density
a
MacKinnon (1996) one-sided p values
Environ Sci Pollut Res

Table 10 VECM estimates—

“health” model dependent Variables Coefficient Standard error t statistics Probability valuea
variable: Dlog(HEXPD)t
Constant −0.618 0.644 −0.959 0.345
Dlog(ALAND)t−1 2.741 1.156 2.369 0.039
Dlog(CO2)t−1 0.144 0.092 1.562 0.149
Dlog(FFUEL)t−1 0.073 0.023 3.178 0.007
Dlog(PD)t−1 −19.917 8.669 −2.297 0.044
log(ALAND)t−1 0.664 0.378 1.753 0.090
log(CO2)t−1 0.063 0.034 1.836 0.077
log(FFUEL)t−1 0.073 0.146 0.502 0.619
log(PD)t−1 0.606 0.233 2.598 0.014
ρ −0.919 0.281 −3.555 0.005

Statistical tests: R2 0.524, adjusted R2 0.456, F statistics 7.707, F prob. value 0.000. Diagnostic tests: Jarque-
Bera 1.624, LM test 1.998, ARCH tests 1.527, Ramsey RESET test 0.589
ALAND arable land, CO2 carbon dioxide emissions, FFUEL fossil fuel energy consumption, GDPPC GDP per
capita, HEXP health expenditures, PD population density
a
MacKinnon (1996) one-sided p values

Response of GDPPC to Cholesky Response of ALAND to Cholesky

One S.D. Innovations One S.D. Innovations
20 .0025
16 .0020

12 .0015

8 .0010

4 .0005
0 .0000

-4 -.0005
1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 10
GDPPC ALAND CO2 GDPPC ALAND CO2
FFUEL PD FFUEL PD

Response of CO2 to Cholesky Response of FFUEL to Cholesky

One S.D. Innovations One S.D. Innovations
40,000 1.2
1.0
30,000
0.8
20,000 0.6

10,000 0.4
0.2
0
0.0
-10,000 -0.2
1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 10
GDPPC ALAND CO2 GDPPC ALAND CO2
FFUEL PD FFUEL PD

Response of PD to Cholesky
One S.D. Innovations
.16

.12

.08

.04

.00

-.04
1 2 3 4 5 6 7 8 9 10
GDPPC ALAND CO2
FFUEL PD

Fig. 1 Impulse response function for the “wealth” model

 Environ Sci Pollut Res

macroeconomic factors in their region. Subsequently, the
study proceeds for Johansen’s cointegration test for both the
models separately in Table 7.
The results of Johansen’s cointegration test indicate that
there is only one cointegration equation, as shown in both
the trace statistics and the maximum eigenvalue test. The re-
sults confirm that the wealth model exhibit the long-run rela-
tionship between the variables. The null hypothesis of no
cointegration is rejected at 5 % level of significance. Eigen-
value is 0.620 containing a trace statistic value of 80.709
which is greater than the critical value of 69.818. As the trace
statistics value is greater than the critical value, it indicates that
there is a probability of rejecting the null hypothesis at 5 %
level of significance. Similarly, the same result has been
depicted in the case of the maximum eigenvalue statistic, as
these statistics also show that one cointegration equation
exists in the wealth model. The overall results show that the
wealth model has one cointegration equation; therefore, we
may safely conclude that the wealth model exists in the
long-run relationship. Table 8 in continuation of Table 7
shows another model statistics, i.e., the health model.
Table 8 shows the results of the health model where we
may see that the null hypothesis of no cointegration is accept-
ed, as both trace statistics value and maximum eigenvalue are
less than the critical value at 0.05 %. These results indicate
that there is no long-run relationship between the variables
and it has no cointegration equations in the model. After a
careful analysis of the unit root test where HEXP is not even
stationary at the first difference and the Johansen cointegration
test of the health model contains no cointegration equation,
the study switches to the VECM that offered both the short-
run and long-run parameter estimates presented in Table 9.

Response of HEXP to Cholesky Response of ALAND to Cholesky

One S.D. Innovations One S.D. Innovations
.8 .0024
.0020
.6
.0016
.4 .0012

.2 .0008
.0004
.0
.0000
-.2 -.0004
1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 10

HEXP ALAND CO2 HEXP ALAND CO2

FFUEL PD FFUEL PD

Response of CO2 to Cholesky Response of FFUEL to Cholesky

One S.D. Innovations One S.D. Innovations
40,000 1.2

30,000
0.8
20,000
0.4
10,000
0.0
0

-10,000 -0.4
1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 10
HEXP ALAND CO2 HEXP ALAND CO2
FFUEL PD FFUEL PD

Response of PD to Cholesky
One S.D. Innovations
.16

.12

.08

.04

.00

-.04
1 2 3 4 5 6 7 8 9 10
HEXP ALAND CO2
FFUEL PD

Fig. 2 Impulse response function for the “health” model

Environ Sci Pollut Res

The results show that in the short run, arable land has a
positive relationship with the economic growth while popu-
lation density exerted the negative influence on economic
growth that tends to decrease around 11.167 % if there is
1 % increase in population density. However, this result is
evaporated in the long run, when population density signif-
icantly increases economic growth around 0.762 %. On the
other side, arable land has a more elastic relationship with
the economic growth, as the coefficient value exceeds the
unity value. The results are somehow contradictory with the
previous estimation of the variables by correlation matrix;
however, it might be possible due to a one-to-one relation-
ship and a multivariate case. The VECM estimates are more
robust in nature, and inferences drawn from this technique
give the conclusive results. Overall, the results deduce that
along with the increase in arable land and population densi-
ty, wealth significantly increases; however, the magnitude of
increasing wealth depends upon the individual nature of the
regressors. The results indicate that as arable land increases,
it has a direct impact on the economic performance of the
low-income countries. The results are somehow contradicto-
ry with the previous studies that based on middle-income
and high-income countries; i.e., Gar-O Yeh and Li (1999)
argued that economic development could be achieved at the
expense of loss of agricultural land in China, while Tan
et al. (2005) further extended that due to massive urban
land, China considerably loses arable land. Harrison et al.
(2014) conclude that across Europe, there is an increase
in food production which on the cost of forest area and
unmanaged land satisfies increasing food demand. Popula-
tion density has a positive association with the GDP per
capita of low-income countries. Usually, this relationship is
indirect, i.e., as population density increases, it decreases
economic growth; however, in low-income countries, where
the people fight for their survival with not only environmen-
tal hazards but with food scarcity, this battle provides an
opportunity to find job opportunities in the urban cities on
the cost of environmental pollution. The adjustment param-
eter, i.e., ρ, shows the negative value of the coefficient pa-
rameter, i.e., −0.185, that imply the long-run convergence of
the model. The remaining statistics, i.e., adjusted R-square,
show that around 98 % of regressors response to the wealth
model, while the F statistics show that empirically there is
no such problem in the wealth stability model. The remain-
ing statistics are shown for the ready reference of the com-
mon reader that indicate the model fitness and robust

Variance Decomposition of GDPPC Variance Decomposition of ALAND

100 100

80 80

60 60

40 40

20 20

0 0
1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 10
GDPPC ALAND CO2 GDPPC ALAND CO2
FFUEL PD FFUEL PD

Variance Decomposition of CO2 Variance Decomposition of FFUEL

100 100

80 80

60 60

40 40

20 20

0 0
1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 10
GDPPC ALAND CO2 GDPPC ALAND CO2
FFUEL PD FFUEL PD

Variance Decomposition of PD
100

80

60

40

20

0
1 2 3 4 5 6 7 8 9 10
GDPPC ALAND CO2
FFUEL PD

Fig. 3 Variance decomposition for the “wealth” model

 Environ Sci Pollut Res

inferential statistics. Table 10 evaluated the health model in
low-income countries by using VECM estimates.
The results of the health model show that arable land and
fossil fuel increases along with the increase in health expen-
ditures of low-income countries, while population density de-
creases along with the increase in health expenditures in the
short run. The long-run results indicate that arable land, car-
bon dioxide emissions, and population density increase health
expenditures in low-income countries. The results imply that
as the arable land increases, it affects the GDP per capita,
particularly the share of health expenditures as percentage of
GDP increases that help to facilitate massive population for
health-related problems. In addition, fossil fuel energy con-
sumption is one of the major contributors for greenhouse gas
emissions, for prevention of emissions; there is a severe need
of health-care expenditures to the low-income countries’ res-
idents. Carbon dioxide emission is the strong regressor that
represented air pollution in most of the previous studies (i.e.,
Alam et al. 2014; Sajjad et al. 2014; Rafindadi et al. 2014;
Qureshi et al. 2014; Akhmat et al. 2014, etc.). The results
show that carbon dioxide emission increases with health ex-
penditures; therefore, there is a significant need to device a
carbon-free policy for the region. Fossil fuel energy consump-
tion exerts the positive impact on low-income countries’
growth reforms, as this is an undeniable fact which states that
energy consumption in the form of fossil fuel energy has con-
tributed to economic growth in the countries. However, it is
also true that fossil fuel combustion accounts for about 90 %
of total global carbon dioxide emissions; i.e., Bölük, and Mert
(2014) concluded for European Union countries, Apergis and
Payne (2014) concluded for Central America, and Magné
et al. (2014) concluded for global scenario etc. The overall
results summarize that both health and wealth are a necessity
for the low-income countries for devising the sustainable po-
lices in their region. The caution is to develop a free-carbon
economy while devising health reforms in the region.
These results enforce to further analyze the relationship
between wealth and health models for the next 10-year time

Variance Decomposition of HEXP Variance Decomposition of ALAND

100 100

80 80

60 60

40 40

20 20

0 0
1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 10
HEXP ALAND CO2 HEXP ALAND CO2
FFUEL PD FFUEL PD

Variance Decomposition of CO2 Variance Decomposition of FFUEL

100 100

80 80

60 60

40 40

20 20

0 0
1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 10
HEXP ALAND CO2 HEXP ALAND CO2
FFUEL PD FFUEL PD

Variance Decomposition of PD
100

80

60

40

20

0
1 2 3 4 5 6 7 8 9 10
HEXP ALAND CO2
FFUEL PD

Fig. 4 Variance decomposition for the “health” model

Environ Sci Pollut Res
period. Figure 1 shows the impulse response function for the
wealth model.
Figure 1 depicts that arable land, carbon emissions, and
population density would increase with the GDP per capita
for the next 10-year period; however, fossil fuel energy con-
sumption will decline after the eighth period. Arable land grad-
ually increases from period 1 to period 10, and its frequency is
far greater than other regressors in the wealth model. Figure 2
shows the impulse response function for the health model.
Figure 2 shows that along with the increase in health ex-
penditures, arable land and population density subsequently
increase, while fossil fuel energy consumption has periodical
ups and downs. Carbon dioxide emissions, however, have a
negative association with the health expenditures over the next
10-year period. Carbon dioxide emissions exhibit the increas-
ing trend up to the seventh period, and then gradually decrease
for the next subsequent years. The results further emphasize
the most contributing factor for the next ten year period that
observes with the variance decomposition technique present-
ed in Figs. 3 and 4, respectively. Figure 3 shows the variance
decomposition for the wealth model.
Figure 3 shows that arable land has the major contribution
to economic growth in low-income countries during the next
10-year period, followed by population density, fossil fuel
energy consumption, and carbon emissions, respectively.
The variance was explained by arable land by 37 %, popula-
tion density by 6.25 %, fossil fuel energy consumption by
0.78 %, and carbon emissions by a 0.543 %. Similarly,
Fig. 4 shows the health model for low-income countries.
Figure 4 shows that arable land explained a 24.373 % var-
iance on health expenditures per capita, population density
explained by 5.151 %, fossil fuel energy consumption by
1.104 %, and carbon emissions by 0.504 %, for the next 10-
year time period. The overall health and wealth results indi-
cate that arable land is the most influential variable in the
future while carbon emissions will exert the least contribution
to influence both health and wealth for low-income countries.
Conclusion
The study focused on the health and wealth issues for low-
income countries. For this purpose, the study brings an ag-
gregate data of low-income countries for decomposing the
time series-consistent data, over the period of 1975–2013.
The study employed a number of econometric tools for
decomposing aggregate time series data and estimated dif-
ferent robust inferences on number of variables, i.e., GDP
per capita, health expenditures per capita, arable land, fossil
fuel energy consumption, carbon dioxide emissions, and
population density. The study divided the variables in two
models which represented the health and wealth models.
Both models strictly passed on through sophisticated
decomposing techniques and finally drew the following
conclusions:
& The candidate variables have a wide fluctuation in their
data series; therefore, we safely conclude that the variables
have a unit root problem.
& Correlation results have been evaporated while using the
VECM approach.
& Single equation cointegration test estimated by Engle-
Granger test reveals that the model contains long-run rela-
tionship between them; however, Johansen cointegration
test for the health model denied this fact and conclude that
the model has no cointegration equations.
& Engel-Granger single equation cointegration test some-
how confirmed the wealth model, where the null hypoth-
esis of no cointegration is rejected at 5 %, and Johansen
cointegration test indicates one cointegration equation in
the model.
& VECM estimates for the health model shows that health
expenditures increase on the cost of environmental and air
pollution indicators, as arable land and fossil fuel energy
consumption increase health expenditures while population
density decreases health expenditures in the short run. In
the long run, however, arable land, carbon dioxide emis-
sions, and population density increase along with the in-
crease of health expenditures in the low-income countries.
& VECM estimates for the wealth model show that arable
land and population density both have a positive associa-
tion with the GDP per capita of low-income countries in
the long run.
& These results further encompass with the “innovation ac-
counting technique” which comprises impulse response
function and variance decomposition.
& Both the results show that arable land is the most impor-
tant variable that increases health and wealth for low-
income countries, while carbon dioxide emissions have a
weak and less contribution to the response variables, over
the next 10-year period.
These results endorse the statement of Tilman et al. (2002,
p.671), i.e.,
New incentives and policies for ensuring the sustain-
ability of agriculture and ecosystem services will be
crucial if we are to meet the demands of improving
yields without compromising environmental integrity
or public health.
There are different types of environmental concerns of low-
income countries, the foremost is environmental degradation
that is caused by the massive population and anti-poor reforms
by the federal government of the respective countries, while
another one is related with the climatic factors that drastically

change from over the last decade (Bevan 2012). The problem
area for low-income countries needs to be resolved with the
proper implementation of resources that put forward the pre-
vention of environmental hazards which, if not treated, may
ultimately affect the massive population living in this region.
There are some encouraging facts related with the energy sec-
tor reforms, and improvements in the business environment in
low-income countries have been received recently(e.g., Ban-
gladesh, Kenya, Nigeria, and Senegal) which are a positive
sign to rebuilding peace and stability in macroeconomic fac-
tors including energy subsidy and structural reforms in a re-
gion. There is a need of broad-based policies for managing
health and wealth issues that could be settled down by improv-
ing institutional and regulatory environment, as well as mac-
roeconomic reforms that would help to keep the environment
from air pollution which is like a “gun and bullet” for low-
income countries.
References
Akhmat G, Zaman K, Shukui T, Malik IA, Begum S, Ahmed A (2014) A
contemporary landscape of air pollution and greenhouse gas emis-
sions leads to inevitable phenomena of low birthweight. Environ Sci
Pollut Res 21(15):9408–9414
Akram N (2012) Is climate change hindering economic growth of Asian
economies? Asia Pac Dev Journal 19(2):1–18
Alam A, Azam M, Abdullah A, Malik IA, Khan A, Hamzah TAA,
Faridullah, Khan MM, Zahoor H, Zaman K (2014) Environmental
quality indicators and financial development in Malaysia: unity in
diversity. Environ Sci Pollut Res. doi:10.1007/s11356-014-3982-5
Alkhathlan K, Alam MQ, Javid M (2012) Carbon dioxide emissions,
energy consumption and economic growth in Saudi Arabia: a mul-
tivariate cointegration analysis. Br J Econ Manage Trade 2(4):327–
339
Alshehry AS, Belloumi M (2015) Energy consumption, carbon dioxide
emissions and economic growth: the case of Saudi Arabia. Renew
Sust Energ Rev 41:237–247
Apergis N, Payne JE (2014) Renewable energy, output, CO<sub>2</
sub>emissions, and fossil fuel prices in Central America: evidence
from a nonlinear panel smooth transition vector error correction
model. Energy Econ 42:226–232
Armstrong JS (2001) Principles of forecasting: a handbook for re-
searchers and practitioners. Springer science+Business Media Inc,
USA
Banister J (1998) Population, public health and the environment in China.
China Q 156:986–1015
Bevan D (2012) Aid, fiscal policy, climate change, and growth. Working
paper No. 2012/77. United Nations University – UNU-Wider, UK
Bhusal TP (2010) Econometric analysis of oil consumption and economic
growth in Nepal. Econ J Dev 11&12(1–2):135–143
Bölük G, Mert M (2014) Fossil & renewable energy consumption, GHGs
(greenhouse gases) and economic growth: evidence from a panel of
EU (european union) countries. Energy 74:439–446
Borhan H, Musa EA, Hitam M (2012) The impact of CO2 on economic
growth in ASEAN 8. Proc Soc Behav Sci 35:389–397
Bozkurt C, Akan Y (2014) Economic growth, CO2 emissions and energy
consumption: the Turkish case. Int J Energy Econ Policy 4(3):484–
494
Environ Sci Pollut Res
Chen T (2001) The impact of mitigating CO2 emissions on Taiwan’s
economy. Energy Econ 23(2):141–151
Cohen S, Glass DC, Phillips S (1979) Environmental factors in health. In:
Freeman H, Levine S, Reeder LG (eds) Handbook of medical soci-
ology. Prentice-Hall, Englewood Cliffs
Dickey D, Fuller W (1979) Distribution of the estimators for
autoregressive time-series with a unit root. J Am Stat Assoc 74:
427–431
Dickey D, Fuller W (1981) Likelihood ratio statistics for autoregressive
TimeSeries with a unit root. Econometrica 49(1057):1072
Erbaykal E (2008) Disaggregate energy consumption and economic
growth: evidence from Turkey. Int Res J Financ Econ 20:172–179
Farhani S, Rejeb JB (2012) Energy consumption, economic growth and
CO2 emissions: evidence from panel data for MENA region. Int J
Energy Econ Policy 2(2):71–81
Friel S, Dangour AD, Garnett T, Lock K, Chalabi Z, Roberts I, Butler A,
Butler CD, Waage J, McMichael AJ, Haines A (2009) Public health
benefits of strategies to reduce greenhouse-gas emissions: food and
agriculture. Lancet 374(9706):2016–2025
Galle OR, Grove WR, McPherson JM (1972) Population density and
pathology: what are the relations for man? Science 176:23–30
Gar-O Yeh A, Li X (1999) Economic development and agricultural land
loss in the Pearl River Delta, China. Habitat Int 23(3):373–390
Gosh S (2009) Import demands of crude oil and economic growth: evi-
dence from India. Energy Policy 37:699–702
Greiner A (2004) Anthropogenic climate change in a descriptive growth
model. Environ Dev Econ 9(5):645–662
Greiner A (2005) Anthropogenic climate change and abatement in a
multi-region world with endogenous growth. Ecol Econ 55(2):
224–234
Grossman GM, Krueger AB (1995) Economic growth and environment.
Q J Econ 110(2):353–377
Haines A, Kovats RS, Campbell LD, Corvalan C (2006) Climate change
and human health: impacts, vulnerability and public health. Public
Health 120:585–596
Harrison PA, Dunford R, Savin C, Rounsevell MDA, Holman IP, Kebede
AS, Stuch B (2014) Cross-sectoral impacts of climate change and
socio-economic change for multiple, European land-and water-
based sectors. Climat Change 1–14
Hitz S, Smith JB (2004) Estimating global impacts from climate change.
Glob Environ Chang 14(3):201–218
Houghton JT, Callander BK, Varney SK (eds) (1992) Climate change
1992, the supplementary report to the IPCC scientific assessment.
Cambridge University Press, Cambridge
Johansen S, Juselies K (1990) Maximum likelihood estimation and infer-
ences on cointegration. Oxf Bull Econ Stat 52:169–210
Khan H, Inamullah E, Shams K (2009) Population, environment and
poverty in Pakistan: linkages and empirical evidence. Environ Dev
Sustain 11:375–392
Levy L, Herzog AN (1974) Effects of population density and crowding
on health and social adaptation in the Netherlands. J Health Soc
Behav 15(3):228–240
Liu Z, Lin J (2009) Macroeconomic effects of carbon dioxide emission
reduction: cost and benefits. J Camb Stud 4(3):86–94
Lubowski RN, Vesterby M, Bucholtz S, Baez A, Roberts MJ (2006)
Major uses of land in the United States, 2002. Economic
Information Bulletin No.14. Retrieved from http://www.ers.usda.
gov/media/250091/eib14_1_.pdf
Lutken S, Fenhann J, Hinostroza M, Sharma S, Olsen KH (2011) Low
carbon development strategies: a primer on framing nationally ap-
propriate mitigation actions (NAMAs) in developing countries.
UNEP Riso Centre, Technical University of Denmark
MacKinnon JG (1996) Numerical distribution functions for unit root and
cointegration tests. J Appl Econ 11:60–618
Environ Sci Pollut Res
MacKinnon JG, Haug AA, Michelis L (1999) Numerical distribution
functions of likelihood ratio tests for cointegration. J Appl Econ
14:563–577
Magné B, Chateau J, Dellink R (2014) Global implications of joint fossil
fuel subsidy reform and nuclear phase-out: an economic analysis.
Clim Chang 123(3–4):677–690
McAlpine CA, Etter A, Fearnside PM, Seabrook L, Laurance WF (2009)
Increasing world consumption of beef as a driver of regional and
global change: a call for policy action based on evidence from
Queensland (Australia), Colombia and Brazil. Glob Environ
Chang 19:21–33
McMichael AJ, Powles JW, Butler CD, Uauy R (2007) Food, livestock
production, energy, climate change, and health. Lancet 370(9594):
1253–1263
Millennium Ecosystem Assessment (2005) Ecosystems and human well-
being: synthesis. Island Press: Millennium Ecosystem Assessment
Series, Washington, DC
Moss RH (1976) The human context. Willey, New York
OECD (2011). Towards green growth: monitoring progress. Retrieved
from http://www.oecd.org/greengrowth/48224574.pdf
Omoju O (2014) Environmental pollution is inevitable in developing
countries. Retrieved from http://breakingenergy.com/2014/09/23/
environmental-pollution-is-inevitable-in-developing-countries/
Omri A, Nguyen DK, Rault C (2014) Causal interactions between CO2
emissions, FDI, and economic growth: evidence from dynamic si-
multaneous equation models. Working Paper 2014–542. IPAG
Business School 184, Boulevard Saint-Germain 75006 Paris France
Ong SM, Sek SK (2013) Interactions between economic growth and
environmental quality: panel and non-panel analyses. Appl Math
Sci 7(14):687–700
Pope CA III (2007) Mortality effects of longer-term exposures to fine
particulate air pollution: review of recent epidemiological evidence.
Inhal Toxicol 19(Suppl 1):33–38
QUARG (1993) United Kingdom quality of urban air review group,
diesel vehicle emissions and urban air quality. Department of the
Environment, UK
Qureshi MI, Rasli AM, Awan U, Ma J, Ali G, Alam A, Zaman K (2014)
Environment and air pollution: health services bequeath to gro-
tesque menace. Environ Sci Pollut Res, 1–10
Rafindadi AA, Yusof Z, Zaman K, Kyophilavong P, Akhmat G (2014)
The relationship between air pollution, fossil fuel energy consump-
tion, and water resources in the panel of selected Asia-Pacific coun-
tries. Environ Sci Pollut Res 21(19):11395–11400
Saidi K, Hammami S (2015) The impact of energy consumption and CO2
emissions on economic growth: fresh evidence from dynamic
simultaneous-equations models. Sustain Cities Soc 14:178–186
Sajjad F, Noreen U, Zaman K (2014) Climate change and air pollution
jointly creating nightmare for tourism industry. Environ Sci Pollut
Res 21(21):12403–12418
Samet J, Krewski D (2007) Health effects associated with exposure to
ambient air pollution. J Toxicol Environ Health 70(3–4):227–242
Say NP, Yucel M (2006) Energy consumption and CO2emissions in
turkey: empirical analysis and future projection based on economic
growth. Energy Policy 34:3870–3876
View publication stats
Shanmugam R, Hertelendy A (2011) Do developing or developed nations
pollute air more? An assessment of health consequences, The
Impact of Air Pollution on Health, Economy, Environment and
Agricultural Sources, Khallaf, M. (Ed.), ISBN: 978-953-307-528-
0, InTech
Simon JL, (1979) The relationship between population and economic
growth in LDC’S. Faculty Working Papers #538, College of
Commerce and Business Administration University of Illinois at
Urbana-Champaign
Sweeney JL (2008) Low carbon growth: our ethical responsibility-special
report. Development Outreach World Bank Institute, 11-14
Tan M, Li X, Xie H, Lu C (2005) Urban land expansion and arable land
loss in china—a case study of beijing–Tianjin–hebei region. Land
Use Policy 22(3):187–196
Tilman D, Cassman KG, Matson PA, Naylor R, Polasky S (2002)
Agricultural sustainability and intensive production practices.
Nature 418(6898):671–677
Tiwari AK (2011) Energy consumption, CO2 emissions and economic
growth: a revisit of the evidence fromIndia. Appl Econ Int Dev
11(2):165–189
Tol RSJ (2009) The economic effects of climate change. J Econ Perspect
23(2):29–51
Udofia EP, Udom ES (2011) The effects of environmental degradation on
socio-economic condition in IKO community. Sacha J Environ Stud
1(2):17–29
UN Food and Agriculture Organization (2006) Livestock’s long shadow:
environmental issues and options. UN Food and Agriculture
Organization, Rome
UNEP (2010) Driving a green economy through public finance and fiscal
policy reform, WORKING PAPER v. 1.0. United Nations
Environment Programme
Urban F, Nordensvard J (2013) Low carbon development: key issues (key
issues in environment and sustainability). Routledge, Oxford, UK.
ISBN 9780415538985
WCED (1987) Our common future. Oxford University Press, Oxford
Williams AG, Audsley E, Sandars DL (2006) Final report to defra on
project IS0205: determining the environmental burdens and re-
source use in the production of agricultural and horticultural com-
modities. Department for Environment, Food and Rural Affairs,
London
World Bank (2014). World Development Indicator, World Bank,
Washington D.C.
World Energy Outlook ( 2014) International energy agency, 2014
OECD/IEA
Wu JJ (2008) Land use changes: economic, social, and environmental
impacts. CHOICES Agric Appl Econ Assoc 23(4):6–10
Zaman K, Khan H, Khan MM, Saleem Z, Nawaz M (2011) The impact of
population on environmental degradation in south Asia: application
of seemingly unrelated regression equation model. Environ Econ
2(2):80–88