b i o m a s s a n d b i o e n e r g y 5 0 ( 2 0 1 3 ) 1 9 e2 4

Available online at www.sciencedirect.com

http://www.elsevier.com/locate/biombioe

Economic growth and biomass energy

Melike E. Bildirici*
Yildiz Technical University, Yildiz Campus, Istanbul, Turkey

article info abstract

Article history: This paper investigates the short-run and long-run causality analysis between biomass
Received 15 December 2011 energy consumption and economic growth in the selected 10 developing and emerging
Received in revised form countries by using the Autoregressive Distributed Lag bounds testing (ARDL) approach of
14 September 2012 cointegration and error correction models. It covers annual data from 1980 to 2009. The
Accepted 18 September 2012 cointegration test results show that there is cointegration between the biomass energy
Available online 8 December 2012 consumption and the economic growth in nine of the ten countries (Argentina, Bolivia,
Cuba, Costa Rica, El Salvador, Jamaica, Nicaragua, Panama, Paraguay, Peru). The cointe-
Keywords: gration test results show that there is no cointegration between the biomass energy
Biomass energy consumption consumption and the economic growth in one of the ten countries (Paraguay).
Economic growth ª 2012 Elsevier Ltd. All rights reserved.
ARDL
Short-run causality
Long-run causality

1. Introduction energy is attracting growing interest around the world [3].

Biomass energy may be consumed either directly or indirectly.
Direct consumption is the traditional consumption of biomass
energy and involves the process of combustion as cooking,
space heating, and industrial processes. Indirect consumption
and/or modern consumption are the more advanced processes
of converting biomass into secondary energy [1]. As the
economy progresses, commercial energy consumption and
electricity consumption become predominant and traditional
biomass energy consumption decreases. During the period of
rapid urbanization combined with industrialization and
economic development, the energy transition from traditional
biomass energy consumption to commercial fossil fuels
energy consumption accelerated and a higher penetration of
commercial fossil fuels in countries caused a decline in the
share of traditional biomass energy consumption [2]. In addi-
tion to the effects of industrialization, the 1974 and 1979 oil
crises, led to an “energy transition” away from modern fossil
fuels toward modern biomass energy. Such modern biomass
Modern biomass energy is an alternative for reducing foreign
oil dependency because it is renewable, abundant and can be
produced everywhere [4]. Moreover, biomass energy can be
converted to useful thermal energy, electricity and fuels for
power by means of transferring [5]. In developing countries,
modern biomass energy can provide a basis for rural employ-
ment and income. Since biomass production is labor intensive,
feedstock production could be an important source of both
primary employment and supplemental income in rural areas
[1]. Socio-economic benefits of biomass energy consumption
can be identified as a significant driving force in increasing the
share of bioenergy in the total energy supply.
The economic impacts of biomass energy production and
consumption are generally analyzed for three socio-economic
indicators: gross output, value-added and employment [6]. In
evaluating the economic impacts, inputeoutput (IeO) models
are generally used. IeO models developed multipliers that
estimate the relationship between the initial effect of
a change in demand and the total effects of that change. In

* Tel.: þ90 212 3832527.

E-mail address: melikebildirici@gmail.com.
0961-9534/$ e see front matter ª 2012 Elsevier Ltd. All rights reserved.
http://dx.doi.org/10.1016/j.biombioe.2012.09.055
20 b i o m a s s a n d b i o e n e r g y 5 0 ( 2 0 1 3 ) 1 9 e2 4
these models, output is the total value of production and
value-added is measured as total output minus variable costs
and employment, which includes the number of full- and
part-time jobs in the sector [7]. Hjerpe [8], Gan and Smith [6],
and Grebner et.al. [9] show that models have demonstrated
the multiplicative effects of bioenergy in various regions of the
U.S.
The aim of this study is to estimate the relationship
between biomass energy consumption and economic growth
by ARDL method in the following countries: Argentina,
Bolivia, Cuba, Costa Rica, El Salvador, Jamaica, Nicaragua,
Panama, Paraguay and Peru. This study can be defined as
complementary to the previous papers in the context of
energy economics. However, it differs from the existing liter-
ature of energy economics in some aspects. First, to distin-
guish it from the previous works, it employs biomass energy
consumption. Second, it is the first study in the literature that
analyzes the causal relationship between biomass energy
consumption and economic growth for the analyzed
countries.
In perspectives of this paper, knowledge of the direction of
causality between biomass energy consumption and
economic growth is important to determine appropriate
energy policies.
In the second section of the study, causality literature will
be presented. Econometric theory is identified in the third
section. The fourth section consists of the empirical results,
while the last section includes conclusions and policy
implications.
2. The literature of energy consumption and
economic growth
Rasche and Tatom [10], Kraft and Kraft [11], Berndt [12],
Akarca and Long [13] and even Proops [14], Yu and Hwang [15],
Yu and Choi [16], and Erol and Yu [17] were among the first
researchers to examine the relation between GDP and energy
consumption in the framework of energy economy. Many
papers analyzed the relationship between energy consump-
tion and economic growth in pursuit of these pioneering
studies. Depending on developments in econometrics tech-
niques, the Granger causality test and Tado Yamamoto test
were applied following Vector Autoregressive (VAR) and
cointegration methods: the JohanseneJuselius procedure, the
Engle-Granger procedure and the ARDL model. Different
results on direction of causality allowed for four hypotheses:
1) the “neutrality hypothesis”; 2) the “conservation hypoth-
esis”; 3) the “growth hypothesis” and 4) the “feedback
hypothesis”.
The neutrality hypothesis postulates that causality
between GDP and energy consumption does not exist. Energy
consumption is a relatively small component of overall output
and thus will have little or no impact on economic growth.
Second, the conservation hypothesis determines the unidi-
rectional causality running from GDP to energy consumption.
Energy conservation policies, such as efficiency improvement
measures and demand management policies, designed to
reduce energy consumption and waste may not have an
adverse impact on economic growth. These hypotheses are
supported in cases when an increase in economic growth
causes an increase in energy consumption. Third, the growth
hypothesis suggests the unidirectional causality running from
energy consumption to GDP. The growth hypothesis postu-
lates that energy consumption has played a vital role in
economic growth both directly and indirectly. Since an
increase in energy consumption has a positive impact on
economic growth, energy conservation-oriented policies that
reduce energy consumption can impact economic growth.
Fourth, the feedback hypothesis accepts the existence of a bi-
directional causality between GDP and energy consumption.
The feedback hypothesis determines the interdependent
relationship between energy consumption and economic
growth, whereby each serves as complement to the other [18].
For America, with the exception of the United States, the
research on the causal relationship between energy
consumption and GDP is relatively sparse. Nachane et al. [19]
find support for the growth hypothesis in Guatemala. Murray
and Nan [20] identify unidirectional causality from real GDP to
electricity consumption for Colombia. Cheng [21] shows
evidence of unidirectional causality from energy consumption
to real GDP for Brazil and the neutrality hypothesis, absence of
causality between energy consumption and real GDP for
Venezuela. Huang B.N. et.al. [22], in an 82-country panel that
includes Argentina, Bolivia, Brazil, Chile, Colombia, Ecuador,
Paraguay, Peru, Uruguay and Venezuela, determined for the
low-income panel the absence of causality between energy
consumption and real GDP, whereas unidirectional causality
from real GDP to energy consumption was found for the
middle- and high-income panels. Chontanawat et al. [23]
found unidirectional causality from energy consumption to
real GDP for Chile, Colombia and Uruguay, and unidirectional
causality from real GDP to energy consumption for Boliva,
Paraguay, Peru and Venezuela. Squalli [24] reveals evidence of
unidirectional causality from electricity consumption to real
GDP in Venezuela. Apergis and Payne [25] determined both
short-run and long-run causality from energy consumption to
economic growth in Central America. They found evidence of
bi-directional causality between energy consumption and real
output. Apergis and Payne [26] examined the relationship
between energy consumption and economic growth for
a panel of nine South American countries over the period 1980
to 2005 within a multivariate framework. The Granger
causality results indicate both short-run and long-run
causality from energy consumption to economic growth,
which supports the growth hypothesis. Seung-Hoon Y. and
So-Yoon K. [27] investigated the causal relationship between
electricity consumption and economic growth among seven
South American countries, namely Argentina, Brazil, Chile,
Columbia, Ecuador, Peru and Venezuela, for the period
1975e2006. Their results indicated that the causal nexus
between electricity consumption and economic growth varies
across countries. There is a unidirectional, short-run causality
from electricity consumption to real GDP for Argentina, Brazil,
Chile, Columbia and Ecuador. In Venezuela, there is a bi-
directional causality between electricity consumption and
economic growth. However, no causal relationships exist
in Peru.
In recent years, the causal relationship between renewable
energy consumption and economic growth was investigated
 b i o m a s s a n d b i o e n e r g y 5 0 ( 2 0 1 3 ) 1 9 e2 4
for Central America. Apergis and Payne [25] studied the rela-
tionship between renewable energy consumption and GDP for
a panel of six Central American countries over the period
1980e2006. They indicated bi-directional causality between
renewable energy consumption and GDP in both the short-run
and long-run. Apergis et al. [28] analyzed the causal relation-
ship between CO2 emissions, nuclear energy consumption,
renewable energy consumption and GDP in a group of devel-
oped and developing countries for the period 1984e2007 by
a panel error correction model.
This paper analyzes the causal relationship between
biomass energy and economic growth.
3. Data and methodology
3.1. Data
In this study, the relationship between biomass energy
consumption and economic growth was analyzed using the
ARDL method in Argentina, Bolivia, Cuba, Costa Rica, El Sal-
vador, Jamaica, Nicaragua, Panama, Paraguay and Peru. bc
represents the biomass energy consumption log (bct), and py
represents the logarithm of real GDP. Data were taken from
World Bank, the International Energy Agency. Data covers the
1980e2009 period, with the exception of the data for
Argentina, which covers 1980e2005.
3.2. Methodology
Engle and Granger’s [29] two-step cointegration methods,
which are based on the analysis of residuals, and Johansen’s
[30] maximum likelihood reduced rank methods have been
used intensively to determine the relationship between
energy consumption and economic growth. However, criti-
cisms of these method make the results questionable. Harris
[31] and Rahbek and Mosconi [32] emphasized the statistical
deficiencies in those models. Huang and Yang [33] determined
that the Johansen procedure is very sensitive to assumptions.
Ahking [34] and Hjelm and Johansson [35] showed that the
results of Johansen tests are sensitive to the choice of deter-
ministic components.
These problems are eliminated with the ARDL method
developed by Pesaran et al. [36]. The ARDL method allows for
the simultaneous estimation of the short-run and long-run
components within a error correction model. Johansen coin-
tegration techniques require that all the variables in the
system must be of equal order of integration. The ARDL is
applied irrespective of whether underlying regressors are
purely I(0), purely I(1) or mutually cointegrated and thus
requiring no unit root pre-testing. On the other hand, the
ARDL procedure is a statistically more significant method to
determine the cointegration relation in small samples. The
ARDL also allow for the variables to have different optimal
lags. Finally, the ARDL procedure employs only a single
reduced-form equation; however, the conventional cointe-
gration technicals estimate the long-run relationship within
a context of system equations [37]. The ARDL model for the
standard log-linear functional specification of long-run
21
relationships between biomass energy consumption and real
GDP may follows as:
X
m X
n
Dpy ¼ A10 þ B1i Dpyti þ f1i Dbctj þ d11 pyt1 þ d12 bct1 þ ε1t
i¼1 j¼0
(1)
X
m X
n
Dbc ¼ A20 þ B2i Dbcti þ f2i Dpytj þ u21 bct1 þ u22 pyt1 þ ε2t
i¼1 j¼0
(2)
where D and ε1t are the first difference operator and the white
noise term.
The null hypothesis of no cointegration among the vari-
ables in Eq. (1) are H0:d11 ¼ d12 ¼ 0 against the alternative
hypothesis H1:d11 s d12 s 0. In Eq. (2) the null hypothesis of no
cointegration is H0:u21 ¼ u22 ¼ 0 against the alternative
hypothesis H1:u21 s u22 s 0. The bounds testing procedure is
based on the joint F-statistic or Wald statistic that tests the
null hypothesis of no cointegration. If the calculated
F-statistics are below the upper CV, we cannot reject the null
hypothesis of no cointegration. One set of critical values
assumes that all variables in the ARDL model are I(0), while
the other is calculated on the assumption that the variables
are I(1). Thus, rather than using standard critical F-statistic
values, the upper (for I(1)) and lower (for I(0)) bounds of the
F-statistics presented by Pesaran et al. [37] are used. If the
F-statistic is lower than the lower bounds value, then the null
hypothesis of no cointegration cannot be rejected. If the
computed test statistic exceeds the upper critical bounds
value, then the H0 hypothesis is rejected. The appropriate lag
structure is determined using Akaike Information Criterion
(AIC), Schwarz Bayesian Criterion (SBC) and/or Bayesian
Information Criterion (BIC). Estimations are completed using
an ordinary least squares procedure with White’s test for
cross-sectional heteroskedasticity-consistent standard errors
and covariance matrix.
ARDL tests provide information about a long-run rela-
tionship between variables, but no information about the
direction of causality between variables. It is important to
assess how biomass energy consumption and economic
growth affect one another. This can be achieved with the
ARDL procedure by completing separate estimations of Dbct
and Dpyt as dependent variables.
The following error correction model based on the two-
step procedure for Granger causality is used. The advantage of
using an error correction term to test for causality is that it
allows testing for short-run causality through the lagged
difference explanatory variables and for long-run causality
through the lagged ECMt1 term. A statistically significant
ECMt-1 determines long-run causality running from indepen-
dent variables toward the dependent variable [38]. According
to Narayan and Smyth [39], after estimating the long-run
model in order to obtain the estimated residuals, the next
step is to employ the following error correction based on
Granger causality model.
The Error Correction model used to analyze relationships
between the variables was constructed as follows:
22 b i o m a s s a n d b i o e n e r g y 5 0 ( 2 0 1 3 ) 1 9 e2 4

X
m X
n cointegration at 1% and 5% significance levels or better
Dpy ¼ a0 þ bi Dpyti þ fi Dbctj þ 21 ECMt1 þ e1t
i¼1 j¼0
X
p X
q
Dbc ¼ a0 wi Dpyti þ qi Dbctj þ 22 ECMt1 þ e2t
i¼0 j¼1
where residuals et are independently and normally
distributed with zero mean and constant variance and
ECMt1 is the error correction term resulting from the long-
run equilibrium relationship, and b, f, q, w are parameters
to be estimated. 2 is a parameter indicating the speed
of adjustment to the equilibrium level after a shock. It
shows how quickly variables converge to equilibrium and it
must have a statistically significant coefficient with
a negative sign.
The Error Correction model should be a starting point for
the causality analysis. Granger causality can be examined in
three ways [40]: (1) Short-run or weak Granger causalities are
detected by testing H0:fi ¼ 0 and H0:wi ¼ 0 in Eqs. (3) and (4). (2)
Long-run Granger causalities are represented by the ECMs in
equations. The coefficients of the ECMs represent how fast
deviations from the long-run equilibrium are eliminated
following changes in each variable. Long-run causalities are
examined by testing H0: 21¼ 0 and H0: 22¼ 0. (3) Strong Granger
causalities are detected by testing H0:fi ¼ 21¼ 0 and H0:wi¼ 22¼ 0
in Eqs. (3) and (4).
4. Empirical results
Lag length supplying the smallest critical value is determined
as the lag length of the model by using Akaike Information
Criteria. Models were determined after applying the LM test to
all possible models.
According to the computed F-statistics for Paraguay, we do
not have enough evidence to reject the null hypothesis of no
cointegration at 1% and 5% significance levels between
biomass energy consumption and economic growth. We have
enough evidence to reject the null hypothesis of no
(3) between biomass energy consumption and economic growth
for nine (9) countries. This simply means that the computed F-
statistics for these models are above the upper bound critical
(4) value. It was determined to Fy ( pyjbc) for all countries:
Argentina, Bolivia, Cuba, Costa Rica, El Salvador, Jamaica,
Nicaragua, Panama and Peru.
The long-run elasticities and some diagnostic checks for
the underlying ARDL model are displayed in Table 1.
4.1. Granger causality results
According to the ARDL results, a long-run relationship exists
between biomass energy consumption and GDP for countries.
A causality relationship must exist by definition in at least
one direction, it is used Granger causality test by incorpo-
rating the error correction term It is estimated to examine the
possible short-run and long-run causality between these
variables. The causality relationships between biomass
energy consumption and economic growth are summarized
in Table 2.
The ECMs in Table 2 indicate that there exists a mechanism
to correct the disequilibrium between economic growth and
biomass energy consumption. The error correction term was
negatively and statistically significant, showing a speed of
adjustment ranging from 16 to 65% within one year of any
disequilibrium toward a long-run equilibrium state. The error
correction term is statistically significant with a relatively
slow speed of adjustment toward equilibrium. If the coeffi-
cient of the error correction term is smaller than 1, then the
system is equilibrating by fluctuating, and this fluctuating will
decrease in each term before providing the transition to
equilibrium.
The short-run and long-run Granger causality tests reveal
several interesting results. The causality test results (see Table
2) for Argentina, Bolivia, Cuba, Costa Rica, El Salvador,
Jamaica, Nicaragua, Panama and Peru are as follows:
There is evidence of unidirectional Granger causality from
biomass energy consumption to GDP in Argentina, Bolivia,
Cuba, Costa Rica, Jamaica, Nicaragua, Panama and Peru, and

Table 1 e The results of ARDL test.

Bounds testing for cointegration Long-run coefficients

Fy(pyjbc) Fbc(bcjpy) c bc R2(LM)

Argentina 5.02516 0.1258 0.1178 (1.907) 0.2358 (2.003) 0.76

Bolivia 16.460 0.19519 0.35571(3.57) 0.7860
Cuba 4.9749 1.7102 1.1822 (2.512) 1.23128 (2.318) 0.51
Costa Rica 8.905 2.7841 2.353 (2.0825) 0.5548
El Salvador 14.0660 0.32939 0.59516 (3.8270) 1.85252 (2.4563) 0.61105
Jamaica 4.8820 1.7905 0.28962 (0.83384) 0.169921 (4.4258) 0.54008
Nicaragua 5.3282 0.22813 0.15038 (2.96) 0.1308 (1.7896) 0.5789
Panama 5.9838 0.94999 0.5878 (2.9231) 0.391248 0.5048
Paraguay 0.03298 1.9239
Peru 5.9164 0.96622 0.11238 (3.6818) 0.77145 (19.1853) 0.50248

It is given t value and probabilities in paranthesis, The F-statistic is non-standard and is tabulated in Pesaran et al. (2001). Due to the small size,
a maximum lag structure of one (m ¼ 1) was considered for the ARDL in Eq. (1). The appropriate lag structures were selected according to the
Schwartz criteria (see Pesaran and Pesaran (1997) and Enders (2004)). The results indicate one lag was appropriate for most countries.
 b i o m a s s a n d b i o e n e r g y 5 0 ( 2 0 1 3 ) 1 9 e2 4 23

Table 2 e Granger causality and error correction term.

ECM term Dbc / Dpy ECM / Dpy Dbc, ECM / Dpy
Dpy / Dbcc ECM / Dbcb Dpy, ECM / Dbca

Argentina 0.2451(2.77) 9.24365 55.4856 123.456

1.14256 71.9559 245.789
Bolivia 0.1550(2.47) 12.6509 109.4885 210.896
2.058 103.2788 195.567
Cuba 0.41632(3.001) 89.8820 90.0843 93.2003
0.64210 0.40444 1.6365
Costa Rica 0.32113 (2.1728) 8.7248 23.5907 389.0349
1.7901 0.4578 377.2396
El Salvador 0.4480 (4.5046) 6.3087 8.3130 6.5474
6.3266 3.9728 0.44144
Jamaica 0.52350 (2.7838) 6.6991 15.9110 5.8268
0.051691 3.2599 0.78556
Nicaragua 0.5599 (3.05) 44.6447 43.9365 43.5086
2.2771 44.9593 24.9613
Panama 0.65124 (2.896) 39.0888 15.8420 38.5813
1.0150 0.8513 41.5013
Peru 0.63613 (3.00128) 58.862 2.3825 41.0947
2.2730 2.8966 61.9195

a Strong causality.
b Long-run causality.
c Short-run causality.

there is bi-directional causality from biomass energy
consumption to GDP for El Salvador. Furthermore, the long-
run causality results indicate that there is bi-directional
causality for Argentina, Bolivia and Nicaragua, and unidirec-
tional causality for Cuba, Costa Rica, El Salvador Jamaica and
Panama. According to long-run causality results, any causal
relationships within the dynamic error correction model for
Peru cannot be estimated. Finally, a strong causality result
shows bi-directional causality for Argentina, Bolivia, Costa
Rica, Nicaragua, Panama and Peru. The strong causality test
results for Jamaica and El Salvador are disputable.
As a result, no cointegration is found in one out of the ten
countries studied in this paper. A cointegration and causal
relationship is found in nine countries. Thus, we may
conclude that there is a relationship between biomass energy
consumption and economic growth in most of the analyzed
countries.
demand. In the countries studied, there is a unidirectional
relationship from biomass energy consumption to GDP, which
means that biomass energy consumption acts as a stimulus to
economic growth. With these findings, energy policies aimed
at improving energy infrastructure and increasing energy
supply are the appropriate options for these countries since
biomass energy consumption increases the income level. The
unidirectional causality runs from economic growth to
biomass energy consumption. It suggests that the policy of
conserving biomass energy consumption may be imple-
mented with little or no adverse effect on economic growth.
Biomass energy delivers energy in all forms that people
need. Biomass energy is a stimulus for economic growth and
contributes to poverty reduction in developing countries
because it meets energy needs at all times and for all coun-
tries, without expensive conversion devices.

references
5. Conclusion

This paper investigates the causal relationship between [1] http://www.eria.org/research/images/pdf/PDF%20No.6-3/No.

biomass energy consumption and GDP for Argentina, Bolivia,
Cuba, Costa Rica, El Salvador, Jamaica, Nicaragua, Panama,
Paraguay and Peru from 1980 to 2009. To examine this rela-
tionship, we used a two-step procedure. The first step
explored the long-run relationship between the variables by
using the ARDL bounds testing approach of cointegration. The
second step employed a dynamic ECM model to test causal
relationships between biomass energy consumption and GDP.
The results determine that there is weak evidence of causal
relationships between biomass energy consumption and GDP
in some countries.
The implication of this finding is that a high level of
economic growth leads to a high level of biomass energy
6-3-4%20Chap%203%20Bioamss.pdf.
[2] Yemane Wolde R. Energy demand and economic growth: the
African experience. Journal of Policy Modeling 2005;27:
891e903;
Toman T, Jemelkova B. Energy and economic development:
an assessment of the state of knowledge. Energy Journal
2009;24:93e112.
[3] Domaca J, Richardsb K, Risovicc S. Socio-economic drivers in
implementing bioenergy projects. Biomass and Bioenergy
2005;28:97e106.
[4] Balat M, Balat H. Recent trends in global production and
utilization of bio-ethanol fuel. Appl Energy 2009;86:2273e82.
[5] Mengjie W, Suzhen D. A potential renewable energy resource
development and utilization of biomass energy. FAO
Corporate Document Repository; 1994. p. 9408.
24 b i o m a s s a n d b i o e n e r g y 5 0 ( 2 0 1 3 ) 1 9 e2 4
[6] Gan J, Smith CT. Co-benefits of utilizing logging residues for
bioenergy production: the case for East Texas, USA. Biomass
and Bioenergy 2007;31(9):623e30.
[7] Schaffer WA. Regional impact models. Regional Research
Institute, West Virginia University, 1999. In: Perez-Verdin G,
Grebner DL, Munn IA, Sun C, Grado SC, editors. Economic
impacts of woody biomass utilization for bioenergy in
Mississippi, http://www.fwrc.msstate.edu/pubs/10487.pdf;
2008.
[8] Hjerpe, EE. Economics of forest restoration and fuels
reduction programs in the Southwest. Dissertation. Flagstaff,
AZ: Northern Arizona University, School of Forestry; 2006.
[9] Grebner D, Perez-Verdin G, Henderson J, Londo AJ. Bioenergy
from woody biomass, potential for economic development
and the need for extension. J Extension 2009;47(6).
[10] Rasche RH, Tatom JA. The effects of the new energy regime
on economic capacity, production, and prices. Federal
Reserve Bank of St. Louis; 1977. 2e12.
[11] Kraft J, Kraft A. On the relationship between energy and GNP.
J Energy Dev Spring 1978:401e3.
[12] Berndt ER. The demand for electricity: comment and further
results. MIT energy laboratory working paper; 1978. MIT-
EL78e021WP.
[13] Akarca AT, Long TV. On the relationship between energy and
GNP: a re-examination. J Energy Dev 1980;5:326e31.
[14] Proops JLR. Modelling the energyeoutput ratio. Energy Econ
1984;6(1):47e51.
[15] Yu ESH, Hwang B. The relationship between energy and GNP:
further results. Energy Econ 1984;6:186e90.
[16] Yu SH, Choi JY. The causal relationship between energy and
GNP: an international comparison. J Energy Dev 1985;10(2):
249e72.
[17] Erol U, Yu ESH. On the relationship between electricity and
income for industrialized countries. J Electricity Emp 1987;13:
113e22.
[18] Apergis, Payne JE. The renewable energy
consumptionegrowth nexus in Central America. Appl
Energy 2011;88:343e7;
Bildirici ME. The relationship between economic growth and
biomass energy consumption. JRSE 2012;4(2), http://dx.doi.
org/10.1063/1.3699617.
[19] Nachane DM, Nadkarni RM, Karnik AV. Cointegration and
causality testing of the energy-GDP relationship: a cross-
country study. Appl Econ 1988;20:1511e31.
[20] Murray DA, Nan GD. A definition of the gross domestic
product-electrification interrelationship. J Energy Dev 1996;
19:275e83.
[21] Cheng BS. Energy consumption and economic growth in
Brazil, Mexico, and Venezuela: a time series analysis. Appl
Econ Lett 1997;4:671e4.
[22] Huang BN, Hwang MJ, Yang CW. Causal relationship between
energy consumption and GDP growth revisited: a dynamic
panel data approach. Ecol Econ 2008;67:41e54.
[23] Chontanawat J, Hunt LC, Pierse R. Causality between energy
consumption and GDP: evidence from 30 OECD and 78 non-
OECD countries. Surrey Energy Economics Discussion Paper
Series, vol. 113; 2006;
Chontanawat J, Hunt LC, Pierse R. Does energy consumption
cause economic growth? Evidence from a systematic study
of over 100 countries. J Policy Model 2008;30:209e20.
[24] Squalli J. Electricity consumption and economic growth:
bounds and causality analyses of OPEC countries. Energy
Econ 2007;29:1192e205.
[25] Apergis N, Payne JE. Energy consumption and economic
growth in Central America: evidence from a panel
cointegration and error correction model. Energy Econ 2009;
31:211e6.
[26] Apergis N, Payne JE. Energy consumption and growth in
South America: evidence from a panel error correction
model. Energy Econ 2010;32:1421e6.
[27] Seung-Hoon Y, So-Yoon K. Electricity consumption and
economic growth in seven South American countries. Energy
Pol 2010;38:181e8.
[28] Apergis N, Payne JE, Menyah K, Rufael YW. On the
causal dynamics between emissions, nuclear energy,
renewable energy and economic growth. Ecol Econ 2010;
69:2255e60.
[29] Engle RF, Granger CWJ. Cointegration and error correction:
representation, estimation and testing. Econometrica 1987;
55:251e76.
[30] Johansen S. Statistical analysis of cointegration vectors.
J Econ Dynamic Control 1988;12:231e54.
[31] Harris R. Using cointegration analysis in econometric
modelling. London, New York: Prentice Hall; 1995.
[32] Rahbek A, Mosconi R. Cointegration rank inference with
stationary regressors in VAR models. Econometrics J 1999;2:
76e91.
[33] Huang BN, Yang CW. Long-run purchasing power parity
revisited: a Monte Carlo simulation. Appl Econ 1996;28(8):
967e74.
[34] Ahking FW. Model mis-specification and Johansen’s co-
integration analysis: an application to the US money
demand. J Macroeconomics 2002;24:51e66.
[35] Hjelm, Johansson MW. A Monte Carlo study on the pitfalls in
determining deterministic components in cointegrating
models. J Macroeconomics 2005;27:691e703.
[36] Pesaran MH, Shin Y, Smith RJ. Bounds testing approaches to
the analysis of level relationships. J App Economet 2001;16(3):
289e326.
[37] Pesaran MH, Shin Y, Smith RJ. Bounds testing approaches to
the analysis of level relationships. J App Economet 2001;16(3):
289e326;
Pesaran MH, Shin Y. An autoregressive distributed lag
modelling approach to cointegration analysis. In: Strøm S,
editor. Econometrics and economic theory in the twentieth
century: the Ragnar Frisch centennial symposium.
Cambridge: Cambridge University Press; 1999;
Acaravci A. The causal relationship between electricity
consumption and GDP in Turkey: evidence from ARDL
bounds testing approach. Ekonomska istraz ivanja 2010;23(2):
34e43.
[38] Dergiades T, Tsoulfidis L. Revisiting residential demand for
electricity in Greece: new evidence from the ARDL approach to
cointegration analysis. Empirical Econ 2011. http://dx.doi.org/
10.1007/s00181-010-0381-8.
[39] Narayan P, Smyth R. The effect of inflation and real wages on
productivity: new evidence from a panel of G7 countries.
Appl Econ 2009;41(10):1285e91.
[40] Ozturk I, Acaravci A. Electricity consumption and real GDP
causality nexus: evidence from ARDL bounds testing
approach for 11 MENA countries. Appl Energy 2011;88(8):
2885e92;
Lee CC, Chang CP. Energy consumption and economic
growth in Asian economies: a more comprehensive
analysis using panel data. Resour Energy Econ 2008;30:
50e65.