.


International Journal of Energy Economics and Policy | Vol 7 • Issue 2 • 201734

International Journal of Energy Economics and 
Policy

ISSN: 2146-4553

available at http: www.econjournals.com

International Journal of Energy Economics and Policy, 2017, 7(2), 34-43.

Does Too Much Energy Consumption Harm Economic Growth 
for Turkish Republics in The Transition Process? New Evidence 
on Threshold Effects#1

Celil Aydın1, Ömer Esen2*

1Department of Economics, Faculty of Economics and Administrative Sciences, Mus Alparslan University, 49250, Muş, 
Turkey, 2Department of Economics, Faculty of Economics and Administrative Sciences, Mus Alparslan University, 49250, 
Muş, Turkey. *Email: o.esen@alparslan.edu.tr

ABSTRACT

This paper examines whether the effect of energy consumption on economic growth is dependent on the level of energy intensity for 5 Turkish republics 
in the transition process over the period 1991-2012. These economies are divided into two groups in the context of their energy balance. The first group 
comprises all of the Turkish republics in the transition process while the second group is composed of only the net energy exporter countries among 
them. Using an innovative dynamic panel threshold technique, the estimated threshold of the energy intensity for the first group is 0.68%, while for 
the second group countries the threshold is 0.44%. The empirical results indicate that the energy consumption rate above the threshold energy intensity 
level adversely affects the economic growth, but this negative relationship becomes positive one when the energy consumption is below the threshold 
level. These findings reveal that the energy consumption is beneficial to economic growth only up to a certain threshold of the energy intensity; beyond 
the threshold level further the consumption tends to adversely affect the growth. In this regard, policy makers in the transition economies should not 
ignore threshold levels within the context of energy intensity while determining energy policies.

Keywords: Dynamic Panel Threshold Analysis, Energy Consumption, Economic Growth, Energy Intensity 
JEL Classifications: C24, Q43, O11

# A preliminary version of this paper was presented at the 5th World Conference on Business, Economics and Management (WCBEM 2016) held in Antalya, Turkey, 
May 12-14, 2016.

1. INTRODUCTION

Energy is an important resource in the classification of natural 
resources and used in all phases of production. It is considered key 
input and is consumed as output for increasing the welfare level. 
The important role of energy as a production input was disregarded 
until the oil crisis of the 1970s. By that time, even the theories 
of economic growth paid no attention to the role of energy and 
its effect on growth (Stern and Cleveland, 2004). Following this 
period, energy was perceived as a factor of production, together 
with labor and capital. Neoclassical economists like Hamilton 
(1983) and Burbidge and Harrison (1984) reported some findings 
supporting that energy has an important place in the development 
of economies. Attributed with great importance in today’s world, 

energy is considered not only as a production input, but also as 
one of the most important determinants of the economic, social, 
and geographic order of the world as a strategic commodity that 
constitutes the basis for international relations and shapes the world 
economy and politics (Esen, 2016). Countries’ desire to obtain 
energy resources, gain control over them, or have a say in them 
indicates that energy is an indispensable commodity in terms of 
politics. Given that it has such a mission, potential problems in 
the supply of energy used in production raise serious concerns 
about the sustainability of economic growth. Having the necessary 
energy resources, which constitute an extremely important input for 
sustaining economic growth and ensuring modern living standards, 
or at least supplying energy from other countries is of critical 
importance for a country to ensure sustainable economic growth.



Aydın and Esen: Does Too Much Energy Consumption Harm Economic Growth for Turkish Republics in The Transition Process? New Evidence on 
Threshold Effects

International Journal of Energy Economics and Policy | Vol 7 • Issue 2 • 2017 35

The resources used in the generation of energy are limited in 
both amount and diversity. On the other hand, the distribution of 
energy resources around the world is not balanced. Some regions 
have more reserves compared to others in terms of the amount 
and diversity. Such uneven distribution of energy resources is 
true not only for the reserves, but also for the consumption levels 
(Bayrak and Esen, 2014). This leads to the emergence of great 
competition for the energy supply under reasonable, reliable, and 
sustainable conditions among the countries in need of energy 
resources (Öztürkler, 2009). In the world economy, increased 
production means increased energy consumption and increasing 
countries’ dependence on energy import. Therefore, the issues of 
ensuring and sustaining stability in energy production and supply 
constitute the framework of all strategies and plans toward the 
objective of sustainable growth.

During the era of the former Union of Soviet Socialist Republics, 
the means of production in the republics of Kazakhstan, Azerbaijan, 
Turkmenistan, Kyrgyzstan and Uzbekistan—altogether known as 
the Turkic republics—were structured in a way that they were 
heavily dependent on each other and Russia. The production 
process, from raw materials to the end-product, could not be 
handled in a single country as a whole. The raw materials produced 
in one country were made into intermediate goods and, most of 
the time, the end product was produced in Ukraine or Russia 
(Demirkan, 1993). For the Turkic Republics that took over such 
a production structure, it became impossible to reach a production 
capacity sufficient to meet their domestic demands with their 
current economy during the transition period due to, for example, 
a lack of existing infrastructure, the use of outdated technology, 
and undercapitalization. Therefore, these Turkic republics in the 
transition period resorted to operating their energy resources to 
transition to a market economy and ease the problems associated 
with the transition period. During this period, the energy sector 
was the locomotive of economic growth. The energy sector had a 
considerable share in the public revenues as well. Energy resources 
also had a large share in the export revenues. However, due to 
the existing factor endowment and technological infrastructure, 
the Turkic republics focused their attention mostly on mineral 
and raw energy production and their exports, and they had to 
import most industrial goods from the industrialized countries 
of the West. During this period, poverty could not be prevented 
in these republics despite the rapid increase in welfare. Their 
exports concentrated primarily in energy products, leading to 
huge instabilities in energy prices and foreign exchange earnings. 
Therefore, their economic development was adversely affected by 
the fluctuations in energy prices.

Even today, these countries do not have the necessary economic 
structure to operate the oil and natural gas reserves with their 
own equity capital due to, for example, a lack of technology and 
undercapitalization. Therefore, they have to cooperate with large 
Western energy companies to exploit their energy resources. 
However, in order for these Turkic republics to sustain their 
economic growth, it is of crucial importance to reduce their 
dependency on a production structure based on the export of 
labor-intensive goods and original raw materials (especially oil and 
natural gas) to developed countries; they must also turn to other 

industries that require intensive capital and advanced technology. 
This process highlights the importance of energy consumption 
based on increased production and the relationship between energy 
consumption and growth.

The effective and efficient use of energy resources is as important 
as energy consumption for economic development, because the 
efficient use of energy is one of the ways to create new energy 
resources. The literature review shows that energy intensity is 
considered one of the indicators of energy efficiency. A country’s 
energy intensity ratios depend on certain factors, such as the 
production structure of the economy, scale of the industries, 
capacity utilization levels, consumption habits of the societies, 
population, climatic conditions, technological developments, 
energy policies, and the amount of energy resources. Energy 
intensity represents the amount of energy used by an economy 
for each unit value added in a given production process that 
indicates the intensity of energy usage during the production 
process (Gomulka, 1990; Fisher-Vanden et al., 2004; Liao et al., 
2007). Thus, energy intensity measures the efficiency of energy 
use by an economy during the production process (Fisher-Vanden 
et al., 2004). Low energy intensity indicates that existing energy 
resources are being used in production in a more efficient and 
effective way by that economy, thereby also showing that the 
amount of energy consumed to produce a unit of GNP is also 
low (Kavak, 2005). In this sense, meeting the energy required 
for growth by consuming less energy would be beneficial for 
an economy in many ways, such as decreasing dependence on 
external sources of energy and increasing resistance to external 
shocks. In such an environment, it becomes necessary to channel 
the resources to investments with high added value and create 
an environment that will allow efficient energy use as much 
as possible. Otherwise, if the contribution of each factor to 
production (growth) is low, the income level will remain low, 
thereby restricting new investment opportunities and hampering 
long-term and steady economic growth.

The relationship between energy consumption and economic 
growth has become one of the most frequently discussed issues 
in the economics literature, especially following the oil crises in 
the 1970s. However, the issue has not been addressed in terms of 
the transition economies trying to change from centrally planned 
economies to free markets. This study aims to empirically reveal 
the effect of energy consumption on economic growth in five 
Turkic republics undergoing a transition from a centrally planned 
economy to a free market economy following the collapse of the 
socialist regime at the end of the 1980s. In line with this purpose, 
we believe that this study will make a significant contribution to 
the literature on the foundations of economic growth in transition 
economies by examining the issue in terms of the transition 
economies where the states take active roles.

This paper is expected to provide three significant contributions 
to the literature. (i) The sample of this study includes five Turkish 
republics that had a socialist economic system in the past, but are 
currently trying to create their market economies. Although these 
economies are going through different experiences of market 
economy, their common ground is the problems caused by poverty 



Aydın and Esen: Does Too Much Energy Consumption Harm Economic Growth for Turkish Republics in The Transition Process? New Evidence on 
Threshold Effects

International Journal of Energy Economics and Policy | Vol 7 • Issue 2 • 201736

and a public sector that is growing and becoming cumbersome. 
Besides the cumbersome structure inherited from the socialist 
system, deficiency of existing infrastructures, old technologies, 
and undercapitalization directly affect productivity and economic 
growth performance. Hence, the examination of these countries, 
which are resorting to institutional and structural regulations to get 
rid of poverty, will make an important contribution to the literature, 
(ii) previous studies have focused on the energy consumption-
economic growth nexus but less on the effects of energy efficiency 
on growth. Unlike previous studies, this paper investigates the 
effects of energy consumption on growth on the basis of energy 
intensity. Level of energy intensity is one of the important factors 
determining the relationship between energy consumption and 
growth, (iii) previous studies focusing on the effects of energy 
consumption on economic growth have developed models based 
on the assumption that the relationship is linear. However, this 
paper assumes that the effects of energy consumption on economic 
growth will be asymmetrical (i.e., there will be a difference between 
the effect of energy consumption on growth when energy intensity 
is high and the effect of energy consumption on growth when 
energy intensity is low). With this assumption, the study employs 
dynamic panel threshold regression analysis, which is an innovative 
non-linear model, thereby offering an alternative to the literature. 
Accordingly, the study provides significant new findings regarding 
the effects of energy efficiency on the relationship between energy 
consumption and economic growth.

The literature review shows that most of the studies examining 
the relationship between energy consumption and economic 
growth focus on the models and analysis techniques developed 
with a linear hypothesis based on a causality relationship. The 
analyses and interpretations based on the assumption that the 
relationship between energy consumption and economic growth 
is linear point out the existence of a symmetrical relationship 
between the increases and decreases in energy consumption and 
economic growth. These studies view the energy consumption 
level as an indicator of development and consider that a country 
produces based on the amount of energy it consumes and achieves 
a high level of welfare. However, development should be based 
on the ability to create much more economic value with less 
energy (i.e., deliver the same production with a lower amount 
of energy), rather than the energy consumption rates. Energy 
consumption is important for economic growth as well; however, 
the consumption must be balanced. Such asymmetry points to a 
non-linear relationship. The empirical studies conducted thus far 
provide very little information about whether the relationship is 
linear—in other words, whether the energy consumption has a 
threshold level or not.

This study aims to examine the relationship between economic 
growth and energy consumption, which is considered an indicator 
for development, in the Turkic republics. In the study, the 
threshold regression model was used to find the threshold level 
above which energy consumption adversely affects economic 
growth. Moreover, the study aims to put the energy consumption-
growth debates—built most often on an incorrect theoretical and 
empirical basis—on solid ground. In the first stage of the study, 
we will address the relationship in terms of the effect of energy 

consumption on economic growth and the efficient use of energy. 
In the subsequent sections, we will first review the theoretical 
and empirical literature focusing on the relationship between 
energy consumption and economic growth and then examine the 
interaction between energy consumption and economic growth in 
the Turkic republics for the 1990-2012 period in order to determine 
whether any threshold level exists above which the inefficient 
use of energy reaches unsustainable levels. In the final stage, the 
findings will be compared with the findings of other studies in 
the literature.

2. LITERATURE REVIEW

Despite some problems and bottlenecks in the energy field 
experienced by the countries in the past, energy became an issue 
of debate only after the great oil crises of the 1970s, which deeply 
affected the international markets. As a result, the number of studies 
examining the relationship between energy and economic growth 
gradually increased. However, most of these studies focused 
on whether a relationship exists between energy consumption 
and economic growth and, if it does, what the direction of this 
relationship is. The literature review shows that the causality 
relationship between energy consumption and economic growth 
is handled under four hypotheses. The studies in the first group 
reported a unidirectional relationship from energy consumption 
to economic growth (Stern, 2000; Shiu and Lam, 2004; Altınay 
and Karagöl, 2005; Odhiambo, 2009; Apergis and Payne, 2010; 
Iyke, 2015). This hypothesis—called the growth hypothesis in 
the literature—claims that energy is one of the key indicators of 
economic growth. Such a causality relationship indicates that the 
examined countries have energy-dependent production structures 
and they may have serious problems with the sustainability of 
economic growth in cases where energy demand cannot be met. 
The studies in the second group found a unidirectional causality 
from economic growth to energy consumption (Kraft and Kraft, 
1978; Cheng and Lai, 1997; Ghosh, 2002; Lise and Montfort, 
2007; Mozumder and Marathe, 2007; Herrerias et al., 2013). 
Their hypothesis is known as the conservation hypothesis and 
asserts that the countries are not energy-dependent; moreover, 
energy consumption occurs as a result of increases in economic 
growth. The studies in the third group suggest that a mutual 
causal relationship exists between these two variables (Paul and 
Bhattacharya, 2004; Shahbaz and Lean, 2012; Dagher and Talar, 
2012; Wesseh and Zoumara, 2012; Esseghir and Khouni, 2014). 
Known as the feedback hypothesis, this hypothesis asserts that 
energy consumption and economic growth mutually affect each 
other. Pirlogea and Cicea (2012) argue that, for countries with 
bidirectional energy dependence, it is of crucial importance to pay 
attention to energy efficiency, reduce energy consumption based 
on conventional energy resources, move to renewable energy 
resources, and reduce greenhouse gas emissions. Finally, the fourth 
group includes those studies that reported no causal relationship 
between these two variables (Akarca and Long, 1980; Eden and 
Hwang, 1984; Cheng, 1995; Hondroyiannis et al., 2002; Altınay 
and Karagöl, 2004; Acaravci and Ozturk, 2010a). Known as the 
neutrality hypothesis in the literature, this hypothesis asserts that 
policies toward reducing energy consumption do not have any 
effect on economic growth.



Aydın and Esen: Does Too Much Energy Consumption Harm Economic Growth for Turkish Republics in The Transition Process? New Evidence on 
Threshold Effects

International Journal of Energy Economics and Policy | Vol 7 • Issue 2 • 2017 37

Although many studies examine the relationship between energy 
consumption and economic growth, the debates are still ongoing 
about the direction of the causal relationship between these 
two variables. In other words, no agreement exists on whether 
economic growth dictates energy consumption or whether energy 
consumption is the locomotive of economic growth or no causal 
relationship exists between these two variables. Many studies 
in the literature show that the existence and direction of such 
a relationship varies from one country to another (Erol and Yu, 
1987; Soytas and Sari, 2003; Asafu-Adjaye, 2000; Glasure and 
Lee, 1998; Masih and Masih, 1996; Akinlo, 2008; Wolde-Rufael, 
2006; Acaravci and Ozturk, 2010b; Ozturk et al., 2010; Ouedraogo, 
2013; Salahuddin et al., 2015).

Most studies on the relationship between energy consumption and 
economic growth are based on the assumption that a linear causal 
relationship exists between these two variables. Econometrical 
models are selected, and analyses are carried out in line with this 
assumption. Selecting models and carrying out analyses based 
on the assumption that a linear relationship exists between these 
two variables automatically means accepting the presupposition 
that the variables demonstrate similar behaviors independent of 
the country’s production structure. However, energy consumption 
might have a non-linear effect on economic growth not only in 
the net energy importing countries, but also in the net exporters of 
energy. Although energy consumption is considered an indicator 
of development in the growth and development literature, recent 
debates have mostly focused on the necessity of treating the ability 
of creating more economic value with less energy as a development 
indicator, rather than the energy consumption levels. In parallel 
with these debates, the number of studies on the possibility of a 
non-linear relationship between energy consumption and economic 
growth and the potential asymmetric effect of energy consumption 
on economic growth has increased.

The empirical studies in the literature on energy and growth provide 
very little information about whether energy consumption has a 
non-linear effect on economic growth—in other words, whether 
the energy consumption has a threshold level or not. Lee and 
Chang’s (2007) study is one of the most important studies based on 
the assumption of a non-linear relationship; they examine whether 
energy consumption has linear and non-linear effects on economic 
growth in Taiwan for the 1955-2003 period by using one-sector 
and two-sector growth models. Their findings suggest that a non-
linear relationship exists between these two variables in the case 
of Taiwan and this relationship is characterized by a “U-shape” 
(i.e., first decreasing and then increasing) pattern. A similar study 
by Huang et al. (2008) separates data extending from 1971 to 2002 
into two periods for 82 countries and examines the relationship 
between energy consumption and economic growth during 
these periods. The results of their analysis indicate a non-linear 
relationship between these two variables. Aydin and Esen (2016) 
investigate whether level of energy intensity is important in the 
effect of energy consumption on economic growth of Turkey for 
the period 1975-2013. The findings indicate that there is a non-
linear relationship between energy consumption and economic 
growth in the Turkey economy and the threshold level of energy 
intensity below which energy consumption significantly promotes 

economic growth is estimated at 0.191. An energy consumption 
above this threshold has statistically significant negative effect 
on economic growth.

3. MODEL, DATA SET AND ECONOMETRIC 
METHOD

This study investigated the relationship between energy 
consumption and economic growth for the 1991-2012 period 
in five Turkish republics (Azerbaijan, Kazakhstan, Kyrgyzstan, 
Uzbekistan, and Turkmenistan) in the transition process using 
dynamic panel data analysis that takes into account the energy 
intensity threshold level. Such a relationship was examined 
based on the neoclassical production function used by Ghali and 
Sakka (2004), Lee and Chang (2007), and Huang et al. (2008). 
Equation 1 shows the production function, which includes energy 
consumption as a factor of production like capital stock and labor:

 Y = + initial + EC + x +
it 0 1 it 2 it it it

α α α β ε  (1)

Where Y  indicates the real gross domestic product (GDP) growth 
at time t in country i; indicates the initial level of income; EC  is 
the energy consumption rate; x represents other macroeconomic 
variables that might have an impact on economic growth; and ε 
denotes the white noise error term.

The literature review shows that the real GDP is taken as a basis 
in the international comparisons as it gives important information 
about the growth performances of economies; however, in 
comparing the living standards of different countries or examining 
the changes in the welfare level of a country over time, the real 
GDP per capita is preferred as it takes into account the number 
of people living in a country. In this sense, we used the annual 
growth rate of real GDP per capita (dgdp) as an indicator of living 
standards. We also used the growth rate of energy consumption 
per capita (dec) as the independent variable of the model. The 
energy consumption growth rate was calculated based on the 
logarithmic difference of energy consumption per capita (tons of 
oil equivalent). The graphs for the variables are shown in Figure 1.

To control the effects of other macroeconomic variables related to 
energy consumption on economic growth, we used the following 
control variables based on the studies of Khan and Senhadji 
(2001), Drukker et al. (2005), Huang et al. (2008), and Kremer 
et al. (2013): Percentage of GDP dedicated to investment (igdp), 
growth of labor force participation rate (dlab), inflation (π), initial 
income level (initial), openness (open), foreign trade rate (dtot), 
standard deviation of openness (sdopen), and standard deviation 
of foreign trade rate (sdtot). We also used energy intensity (tpes/Y) 
as the threshold variable. We obtained the data about inflation 
rates from the International Financial Statistics. The data about 
the energy intensity rates were obtained from the International 
Energy Agency, and the data about the other variables were 
acquired from the database of the World Development Indicators. 
Table 1 shows the basic data about the variables; Table 2 shows 
the descriptive statistics.



Aydın and Esen: Does Too Much Energy Consumption Harm Economic Growth for Turkish Republics in The Transition Process? New Evidence on 
Threshold Effects

International Journal of Energy Economics and Policy | Vol 7 • Issue 2 • 201738

We included the lagged value of GDP per capita in the model as 
the explanatory variable based on the assumption that production 
level and structure of an economy are not entirely independent 
of the previous periods; thus, production levels in the previous 
period are also reflected in the subsequent periods (Ramirez and 
Rondán, 2013).

The use of the lagged values of the dependent variable as the 
explanatory variable in the fixed effects and random effects models 
used in the static panel data analysis causes the emergence of a 
relationship between the lagged values of the dependent variable 
and the error terms. Such a relationship causes the estimation 
made by fixed and random effects models and the estimators 
to be inconsistent (Greene, 2000). In such cases, the use of 
the dynamic panel data method eliminates such a relationship 
between the lagged values of the dependent variable and the error 
terms, thereby increasing the reliability of the estimation and the 
consistency of the estimators.

In this study, we used the dynamic panel threshold model 
developed by Kremer et al. (2013) by extending Hansen’s (1999) 
static model for endogenous regressors. We chose the initial level 
of income as the endogenous regressor (initial=dgdpt-1). Our 
panel threshold model was built on Caner and Hansen’s (2004) 
cross-sectional threshold model, where generalized method of 
moments (GMM) type estimators are used in order to allow for 
endogeneity. Equation 2 shows the model.

y = + z I(q £ )+ z I(q > )+
it i it it it it it

µ β γ β γ ε′ ′
1 2

 (2)

Where i represents the units within the scope of the cross-section 
(i = 1,…,n); t indicates the dimension of the time series for each 
unit (t = 1,…,T); yit is the dependent variable; µi is the country-
specific fixed effect; εit≈(0,σ

2) is the independently and identically 
distributed error term; I(.) is the indicator function indicating the 
regime; qit is the threshold variable; and ϒ is the threshold value. 
In addition, zit, indicates an m-dimensional vector of explanatory 
regressors that may include lagged values of the dependent 
variable and other endogenous variables. The vector of explanatory 
variables is partitioned into a subset z1it of exogenous variables 
uncorrelated with eit and a subset of endogenous variables z2it 
correlated with eit (Kremer et al., 2013).

In the first step of the estimation of the model in equation 2, the 
individual effects (µi) have to be eliminated via a fixed-effects 
transformation. Therefore, we apply the forward orthogonal 
deviation method suggested by Arellano and Bover (1995). 
Equation 3 shows the method.

ε ε ε ε
it

*

it i(t+1) iT
=

T t

T t+1
[

1

T t
( +...+ )]

−
−

−
−

 (3)

The most distinguishing feature of this method is that it can avoid 
the serial correlation of the transformed error terms. According to 

Figure 1: Per capita gross domestic product and energy consumption growth rates across countries

Table 1: Basic information about the variables
Abbreviation Explanation
dgdp Annual growth rate of real GDP per capita ($)
dec Logarithmic difference of energy consumption per 

capita (kg of oil equivalent) 
tpes/Y Total primary energy consumption as a percentage 

of GDP measured at purchasing power parity (tons 
of oil equivalent/$1,000)

π Annual percent changes in the CPI
igdp Percentage of GDP dedicated to investment ($)
dlab Employment/population growth rate
initial GDP per capita of the previous term
open Logarithm of export and import as percentage of 

GDP
sdopen Standard deviation of openness
dtot Import/export growth rate
sdtot Standard deviation of foreign trade rate
CPI: Consumer price index, GDP: Gross domestic product



Aydın and Esen: Does Too Much Energy Consumption Harm Economic Growth for Turkish Republics in The Transition Process? New Evidence on 
Threshold Effects

International Journal of Energy Economics and Policy | Vol 7 • Issue 2 • 2017 39

Kremer et al. (2013), this feature allows for the application of the 
estimation procedure derived by Caner and Hansen (2004) for a 
cross-sectional model to the dynamic panel data models.

The next step of the estimation procedures involves the use of 
two-stage least squares (2SLS) to estimate the energy intensity 
threshold level. To this end, following Caner and Hansen (2004), 
we first estimate a reduced form regression for the endogenous 
variables (z2it) as a function of the instruments (Xit). The 
endogenous variables (z2it) are then replaced in the structural 
equation by the predicted values 

2itẑ . Finally, the model in 
equation 2 is estimated via least squares for a fixed threshold (ϒ). 
This step is repeated for the subsets of the threshold variable q. 
Among the threshold values, the threshold value with the smallest 
sum of squared residuals (S[γ]) is selected as the most appropriate 
threshold value ( ŷ ). Equation 4 shows this procedure (Hansen, 
2000. p. 578).

nŷ=arg min S ( )γ  (4)

In accordance with the studies by Hansen (1999), Caner and 
Hansen (2004), and Kremer et al. (2013), the critical values are 
estimated to determine the 95% confidence interval of the energy 
intensity threshold value. Equation 5 is used to estimate these 
critical values:

Г={γ:LR(γ)≤C(α)} (5)

In equation 5, C(α) is the 95% percentile of the asymptotic 
distribution of the likelihood ratio statistic LR(γ). According to 
Hansen (1999), the underlying likelihood ratio is adjusted to 
account for the number of time periods used for each cross-section. 
In the dynamic panel model, after the appropriate threshold value 
( ŷ ) is determined, the slope coefficients are estimated by the 
GMM for the previously determined instruments and the previous 
estimated threshold. Equation 6 shows the dynamic panel threshold 
model formed with GMM to examine the effect of energy intensity 
threshold value on the relationship between energy consumption 
and economic growth.

dgdpit=µI + β1decit I[(tpes/Y)it≤γ] + δ1 I[(tpes/Y)it≤γ] + β2decit 
I[(tpes/Y)it>γ] + ∅zit + εit (6)

Where decit represents the growth rate of energy consumption 
per capita for both regime types and zit represents the vector 
of control variables. Tpes/Y denotes the threshold variable. 
β1 and β2 indicate the regime-dependent slope coefficients, and 
δ1 indicates the fixed regime coefficient. Following Bick (2010) 
and Kremer et al. (2013), we used the initial income level (z2it) as 
the endogenous variable.

According to Roodman (2009), the use of all lags of the dependent 
variable as instruments in the dynamic panel analysis makes the 
coefficient estimation both unbiased and consistent. Therefore, 
based on the study by Arellano and Bover (1995), we used all the 
lags of the dependent variable as instruments. Besides, in dynamic 
panel model, estimator is consistent when T/N→c for 0<c≤2 
(Alvarez and Arellano, 2003). In cases where c is <2, the bootstrap 
method should be used in order to make the estimator consistent.

4. FINDINGS

In Table 3, Model 1 shows the results of the dynamic panel 
threshold model applied to examine the relationship between 
economic growth and energy consumption in 5 Turkic republics 
in transition. Among these five countries, Azerbaijan, Kazakhstan, 
Uzbekistan, and Turkmenistan were found to be net energy 
exporting countries with important oil and natural gas reserves, 
while Kyrgyzstan was found to import some of its energy 
despite its high hydropower potential. Because it is a net energy 
importing country, Kyrgyzstan was left out of the analysis, and the 
relationship between energy consumption and economic growth 
was examined again together with the role of the energy intensity 
threshold value. The findings are shown in Model 2.

The upper part of Table 3 shows the estimated energy intensity 
threshold level and the corresponding 95% confidence interval. 
The middle part of the table shows the effect of energy consumption 
per capita on economic growth for both regime types. 

1β̂  denotes 
the marginal effect of energy consumption per capita on economic 
growth in the low energy intensity regime, while 

2β̂  indicates the 
marginal effect of energy consumption per capita on economic 
growth in the high-energy intensity regime. Low-energy intensity 
regime indicates the case when the energy intensity level is below 
the estimated threshold value, while high-energy intensity regime 
indicates the case when the energy intensity level is above the 
estimated threshold value.

As shown in Table 3, the estimated threshold values for energy 
intensity were found to be 0.68 and 0.44 for Models 1 and 2, 
respectively (via 2SLS). Figure 2 shows the sum of the squared 
errors of the threshold values for both models. In Model 1, the 
lower limit for the threshold value at the 95% confidence level 
is 0.26 and the upper limit is 0.71. In Model 2, the lower limit 
is 0.24 and the upper limit is 0.72. The findings show that, in 
general, energy intensity increases in the first stages of economic 
development; however, it shows a declining tendency in the 
developed economies due to the increasing share of the service 
(tertiary) sector in economic growth during the development 
process. In the underdeveloped or developing economies, increased 
energy intensity is an expected consequence of industrialization. 

Table 2: Descriptive statistics for all variables
tpes/Y dgdp dec π igdp dlab open dtot

Mean 0.467 6.779 −1.562 203.421 6.879 0.175 4.449 3.182
SD 0.245 17.100 9.747 464.919 25.899 1.430 0.530 25.057
Minimum 0.090 −58.133 −40.981 −18.930 −65.217 −6.840 0.000 −52.132
Maximum 1.020 44.904 22.248 3100.000 111.400 5.609 5.173 77.334
GDP: Gross domestic product, SD: Standard deviation



Aydın and Esen: Does Too Much Energy Consumption Harm Economic Growth for Turkish Republics in The Transition Process? New Evidence on 
Threshold Effects

International Journal of Energy Economics and Policy | Vol 7 • Issue 2 • 201740

In countries that have achieved a certain level of industrialization, 
the economic structures turning toward more low energy-intensive 
industries together with technological developments that allow 
for the efficient use of energy may cause a decrease in the energy 
intensity (Leach et al., 1986).

According to Model 1, the regime-dependent coefficients are 
statistically significant (

1β̂ =0.816 and 2β̂ =−0.588), meaning 
that energy consumption per capita has a positive marginal 
effect on economic growth in the low energy intensity regime, 
but a negative marginal effect in the high energy intensity 
regime. In other words, the rate of energy consumption per 
capita below the threshold level affects economic growth 
positively. If the rate is above the threshold value, then it affects 
economic growth negatively. If we examine the regime-
dependent coefficients, we can see that the effect of energy 
consumption per capita on economic growth is higher in the 
low energy intensity regime.

For Model 2, 
1β̂  was found to be statistically significant (1.298), 

while 
2β̂  was not (−0.217). Thus, energy consumption per capita 

has a positive marginal effect on economic growth in the low 
energy intensity regime. Although the relationship between 
energy consumption and economic growth is not statistically 
significant in the high energy intensity regime, the effect has 
been found to be negative. In other words, the rate of energy 
consumption per capita below the threshold level affects 
economic growth positively. If the rate is above the threshold 
value, then it affects economic growth negatively. If we examine 
the regime-dependent coefficients, we can see that the effect of 
energy consumption per capita on economic growth is in the low 
energy intensity regime.

A comparison of Models 1 and 2 shows that the energy intensity 
threshold level is 0.68 in the five countries studied and 0.44 in the 
net energy exporting countries. According to the energy intensity 
levels, the coefficient of the effect of energy consumption per 
capita on economic growth is 1.298 in the net energy exporting 
countries in the low energy intensity regime. This rate was 0.816 
when all five countries in transition were included in the analysis. 
On the other hand, in the high energy intensity regime, the 
coefficient of this effect is 0.588 in all five countries and −0.217 in 
the energy exporting countries. Thus, the marginal positive effect 
of energy consumption per capita on economic growth is higher in 
the energy exporting countries compared to the five countries in 
transition in the low intensity regime. In the high-energy intensity 
regime, the marginal negative effect of energy consumption per 
capita on economic growth is higher in the five countries in 
transition than in the energy exporting countries.

Table 3: Energy consumption and economic growth
Estimated threshold 
value (tpes/Y)

Model 1 Model 2

ŷ 0.68*** 0.44***

95% confidence interval [0.26, 0.71] [0.24, 0.72]
Effect of energy consumption per 
capita (dec)

1β̂
0.816***
(0.136)

1.298***
(0.292)

2β̂
−0.588*
(0.357)

−0.217
(0.222)

Effect of control variables
initialit −1.774

(4.901)
7.843*
(4.426)

igdpit 0.076**
(0.047)

0.085**
(0.057)

dlabit 0.073
(0.773)

−0.574
(0.652)

totit 0.157***
(0.044)

0.167***
(0.051)

sdtotit −60.031***
(22.085)

−20.872
(19.983)

openit 3.728***
(1.103)

2.772**
(1.260)

sdopenit 24.467
(27.175)

19.320
(51.087)

πit −0.010***
(0.003)

−0.015***
(0.004)

1δ̂
14.969***

(4.291)
5.220
(3.88)

Number of observations
ˆtpes/Y y≤ 86 37

ˆtpes/Y y≤ 24 51

Number of countries 5 4
The table shows the results for the dynamic panel threshold model. All available lags of 
the dependent variable were used as instruments in the analysis. Standard errors are given 
in parentheses. *,**,*** indicate significance at 10%, 5% and 1% levels, respectively. 
P values are obtained using 1000 bootstrap replications, GDP: Gross domestic product

Figure 2: Sum of squared of errors of the threshold values (a) Model 1, (b) Model 2

a b



Aydın and Esen: Does Too Much Energy Consumption Harm Economic Growth for Turkish Republics in The Transition Process? New Evidence on 
Threshold Effects

International Journal of Energy Economics and Policy | Vol 7 • Issue 2 • 2017 41

5. CONCLUSION AND DISCUSSION

Energy is one of the basic requirements for economic and social 
development of a country. However, the uneven distribution of 
energy resources around the world has caused many countries 
to rely heavily on outside sources to meet their energy needs. 
Energy demands of such countries increase depending on their 
developing economies and changing socio-economic structures. 
Because the Turkic republics are rich in energy reserves, the 
world’s increasing energy demand turns out to be advantageous 
for these countries. Therefore, the energy industry constitutes 
the basis of the economies of these countries. The revenues 
from the energy industry have a large share in both public and 
export revenues. Moreover, such energy potential provides these 
countries with a great advantage in achieving economic growth and 
expanding into international markets. The high energy potential 
in the region provides the opportunity to attract foreign capital to 
these countries.

After gaining independency, the Turkic republics started to use 
their high energy potential to provide the necessary resources 
for the transition process and deal with the problems associated 
with this process of change. Since then, the energy trade has 
been an “engine of growth” for these countries. However, such 
energy potential cannot be viewed as a resource that will make a 
contribution only to the export potential. The resources obtained 
from the energy industry must be channeled to the production 
activities and investments required for economic development, 
and production must be moved to industries with high added 
value that require high technology. In this way, energy demand 
will show an increase in these countries as the economy grows 
and living standards improve. This process brings attention to the 
importance of energy consumption based on economic growth 
and the relationship between energy consumption and economic 
growth.

In this sense, in the present study, the relationship between 
energy consumption and economic growth and the effects 
of energy intensity threshold value on this relationship were 
examined using annual data for the 1991-2012 period for five 
Turkic republics undergoing a transition from a centrally planned 
economy to a free market economy following the collapse 
of the Soviet Union at the end of the 1980s. To this end, we 
used the dynamic panel threshold model developed by Kremer 
et al. (2013) by extending Hansen’s (1999) static model for 
endogenous estimators.

The findings prove that a non-linear relationship exists between 
energy consumption and economic growth in the transition 
economies in the long term. They further reveal that an energy 
consumption rate above a certain energy intensity level will affect 
economic growth adversely in these transition economies. The 
estimated threshold value was found to be 0.68 for all transition 
economies examined in this study and 0.44 in the net energy 
exporting countries. The findings support the assumption that 
energy consumption below the threshold value has an increasing 
effect on economic growth. Moreover, we believe that this study 
will guide future studies by revealing the importance of the 

energy intensity threshold level in the relationship between energy 
consumption and economic growth.

This paper show that energy efficiency efforts have significant 
potential savings associated with them and are a cost-efficient way 
to reduce the emissions. The effective and efficient use of energy in 
all phases, from energy generation to consumption, as well as the 
necessary steps to ensure its effective and efficient use, are a must 
for achieving a sustainable competitive economic structure in all 
countries, even if they are net energy exporters rich in resources 
or net energy importers lack of resources. Therefore, even if the 
majority of the Turkic republics in transition do not have problems 
in terms of energy reserves, they should still realize strategies 
and policies that include measures to increase the efficiency and 
effectiveness of the use of energy resources in order to speed up 
their economic development, achieve their sustainable growth 
targets, and improve their living standards.

REFERENCES

Acaravci, A., Ozturk, I. (2010a), Electricity consumption-growth nexus: 
Evidence from panel data for transition countries. Energy Economics, 
32(3), 604-608.

Acaravci, A., Ozturk, I. (2010b), On the relationship between energy 
consumption, CO 2 emissions and economic growth in Europe. 
Energy, 35(12), 5412-5420.

Akarca, A.T., Long, T.V.IInd. (1980), Notes and Comments on the 
relationship between energy and GNP: A reexamination. The Journal 
of Energy and Development, 5(2), 326-331.

Akinlo, A.E. (2008), Energy consumption and economic growth: Evidence 
from 11 Sub-Sahara African countries. Energy Economics, 30(5), 
2391-2400.

Altınay, G., Karagöl, E. (2004), Structural break, unit root, and the 
causality between energy consumption and GDP in Turkey. Energy 
Economics, 26(6), 985-994.

Altınay, G., Karagöl, E. (2005), Electricity consumption and economic 
growth: Evidence from Turkey. Energy Economics, 27(6), 849-856.

Alvarez, J., Arellano, M. (2003), The time series and cross-section 
asymptotics of dynamic panel data estimators. Econometrica, 71(4), 
1121-1159.

Apergis, N., Payne, J.E. (2010), Energy consumption and growth in South 
America: Evidence from a panel error correction model. Energy 
Economics, 32(6), 1421-1426.

Arellano, M., Bover, O. (1995), Another look at the instrumental variable 
estimation of error-components models. Journal of Econometrics, 
68(1), 29-51.

Asafu-Adjaye, J. (2000), The relationship between energy consumption, 
energy prices and economic growth: Time series evidence from Asian 
developing countries. Energy Economics, 22(6), 615-625.

Aydin, C., Esen, Ö. (2016), Threshold effects of energy consumption on 
economic growth in Turkey. Journal of Environmental Management 
and Tourism, 3(15), 370-382.

Bayrak, M., Esen, Ö. (2014), Türkiye’nin enerji açiği sorunu ve 
çözümüne yönelik arayişlar. [Turkey’s problem of energy deficit and 
recommendations for its solution]. Atatürk Üniversitesi İktisadi ve 
İdari Bilimler Dergisi, 28(3), 139-158.

Bick, A. (2010), Threshold effects of inflation on economic growth in 
developing countries. Economics Letters, 108(2), 126-129.

Burbidge, J., Harrison, A. (1984), Testing for the effects of oil-price 
rises using vector autoregressions. International Economic Review, 
25(2), 459-484.



Aydın and Esen: Does Too Much Energy Consumption Harm Economic Growth for Turkish Republics in The Transition Process? New Evidence on 
Threshold Effects

International Journal of Energy Economics and Policy | Vol 7 • Issue 2 • 201742

Caner, M., Hansen, B.E. (2004), Instrumental variable estimation of a 
threshold model. Econometric Theory, 20(5), 813-843.

Cheng, B.S. (1995), An investigation of cointegration and causality 
between energy consumption and economic growth. The Journal of 
Energy and Development, 21(1), 73-84.

Cheng, B.S., Lai, T.W. (1997), An investigation of co-integration and 
causality between energy consumption and economic activity in 
Taiwan. Energy Economics, 19(4), 435-444.

Dagher, L., Talar, Y. (2012), The causal relationship between energy 
consumption and economic growth in Lebanon. Energy Policy, 50, 
795-801.

Demirkan, H. (1993), Orta Asya Türk Cumhuriyetleri Yer Altı Kaynakları 
Potansiyeline Yaklaşım [An Outlook on the Underground Sources 
of the Central Asian Turkic Republics]. The 13th Mining Congress 
of Turkey. p385-395.

Drukker, D., Gomis-Porqueras, P., Hernandez-Verme, P. (2005), Threshold 
Effects in the Relationship Between Inflation and Growth: A New 
Panel-Data Approach. MPRA Working Papers, WP. No. 38225.

Eden, S.H., Hwang, B.K. (1984), The relationship between energy and 
GNP: Further results. Energy Economics, 6(3), 186-190.

Erol, U., Yu, E.S. (1987), On the causal relationship between energy 
and income for industrialized countries. The Journal of Energy and 
Development, 13(1), 113-122.

Esen, O. (2016), Security of the energy supply in Turkey: Prospects, 
challenges and opportunities. International Journal of Energy 
Economics and Policy, 6(2), 281-289.

Esseghir, A., Khouni, L.H. (2014), Economic growth, energy consumption 
and sustainable development: The case of the union for the 
Mediterranean countries. Energy, 71, 218-225.

Fisher-Vanden, K., Jefferson, G.H., Liu, H., Tao, Q. (2004), What is 
driving China’s decline in energy intensity? Resource and Energy 
Economics, 26(1), 77-97.

Ghali, K.H., El-Sakka, M.I.T. (2004), Energy use and output growth in 
Canada: A multivariate cointegration analysis. Energy Economics, 
26(2), 225-238.

Ghosh, S. (2002), Electricity consumption and economic growth in India. 
Energy Policy, 30(2), 125-129.

Glasure, Y.U., Lee, A.R. (1998), Cointegration, error-correction, and the 
relationship between GDP and energy: The case of South Korea and 
Singapore. Resource and Energy Economics, 20(1), 17-25.

Gomulka, S. (1990), The Theory of Technological Change and Economic 
Growth. London: Routledge.

Greene, W.H. (2000), Econometric Analysis. 5th ed. USA: Prentice Hall.
Hamilton, J.D. (1983), Oil and the macroeconomy since World War II. 

Journal of Political Economy, 91(2), 228-248.
Hansen, B.E. (1999), Threshold effects in non-dynamic panels: 

Estimation, testing, and inference. Journal of Econometrics, 93(2), 
345-368. Available from: http://www.dx.doi.org/10.1016/S0304-
4076(99)00025-1.

Hansen, B.E. (2000), Sample splitting and threshold estimation. 
Econometrica, 68(3), 575-603.

Herrerias, M.J., Joyeux, R., Girardin, E. (2013), Short-and long-run 
causality between energy consumption and economic growth: 
Evidence across regions in China. Applied Energy, 112, 1483-1492.

Hondroyiannis, G., Lolos, S., Papapetrou, E. (2002), Energy consumption 
and economic growth: Assessing the evidence from Greece. Energy 
Economics, 24(4), 319-336.

Huang, B.N., Hwang, M.J., Yang, C.W. (2008), Does more energy 
consumption bolster economic growth? An application of the 
nonlinear threshold regression model. Energy Policy, 36(2), 755-767.

Iyke, B.N. (2015), Electricity consumption and economic growth in 
Nigeria: A revisit of the energy-growth debate. Energy Economics, 
51, 166-176.

Kavak, K. (2005), Energy Efficiency in the World and Turkey and 
Investigation of Energy Efficiency in Turkish Industry. Ankara: 
State Planning Organization (Turkey). DPT-2689. Available from: 
http://www.dpt.gov.tr/DocObjects/Download/3226/enerji.pdf. [Last 
accessed on 2012 Oct 10].

Khan, M.S., Ssnhadji, A.S. (2001), Threshold effects in the relationship 
between inflation and growth. IMF Staff Papers, 48(1), 1-21.

Kraft, J., Kraft, A. (1978), On the relationship between energy and GNP. 
Journal of Energy and Development, 3(2), 401-403.

Kremer, S., Bick, A., Nautz, D. (2013), Inflation and growth: New 
evidence from a dynamic panel threshold analysis. Empirical 
Economics, 44(2), 861-878.

Leach, G., Obermair, J.L., Hoffmann, G.M.L. (1986), Energy and Growth. 
England: Butterworth Publishers.

Lee, C.C., Chang, C.P. (2007), The impact of energy consumption on 
economic growth: Evidence from linear and nonlinear models in 
Taiwan. Energy, 32(12), 2282-2294.

Liao, H., Fan, Y., Wei, Y.M. (2007), What induced China’s energy intensity 
to fluctuate: 1997-2006? Energy Policy, 35(9), 4640-4649.

Lise, W., Montfort, K.V. (2007), Energy consumption and GDP in 
Turkey: Is there a co-integration relationship? Energy Economics, 
29(6), 1166-1178.

Masih, A.M., Masih, R. (1996), Energy consumption, real income and 
temporal causality: Results from a multi-country study based on 
cointegration and error-correction modelling techniques. Energy 
Economics, 18(3), 165-183.

Mozumder, P., Marathe, A. (2007), Causality relationship between 
electricity consumption and GDP in Bangladesh. Energy Policy, 
35(1), 395-402.

Odhiambo, N.M. (2009), Energy consumption and economic growth 
nexus in Tanzania: An ARDL bounds testing approach. Energy 
Policy, 37(2), 617-622.

Ouedraogo, N.S. (2013), Energy consumption and economic growth: 
Evidence from the economic community of West African States 
(ECOWAS). Energy Economics, 36, 637-647.

Ozturk, I., Aslan, A., Kalyoncu, H. (2010), Energy consumption and 
economic growth relationship: Evidence from panel data for low 
and middle income countries. Energy Policy, 38(8), 4422-4428.

Öztürkler, H. (2009), ortadoğu ülkelerinin enerji kaynaklarinin öneminin 
ekonomi-politik bir değerlendirmesi. [A political economic 
evaluation on the importance of energy resources in Middle East]. 
Ortadoğu Analiz, 1(7-8), 74-79.

Paul, S., Bhattacharya, R.N. (2004), Causality between energy 
consumption and economic growth in India: A note on conflicting 
results. Energy Economics, 26(6), 977-983.

Pirlogea, C., Cicea, C. (2012), Econometric perspective of the energy 
consumption and economic growth relation in European Union. 
Renewable and Sustainable Energy Reviews, 16(8), 5718-5726.

Ramirez-Rondán, N.R. (2013), Essays on Dynamic Panel Threshold 
Models. PhD Thesis, University of Wisconsin-Madison.

Roodman, D. (2009), A note on the theme of too many instruments. Oxford 
Bulletin of Economics and Statistics, 71(1), 135-158.

Salahuddin, M., Gow, J., Ozturk, I. (2015), Is the long-run relationship 
between economic growth, electricity consumption, carbon dioxide 
emissions and financial development in Gulf Cooperation Council 
countries robust? Renewable and Sustainable Energy Review, 51, 
317-326.

Shahbaz, M., Lean, H.H. (2012), The dynamics of electricity consumption 
and economic growth: A revisit study of their causality in Pakistan. 
Energy, 39(1), 146-153.

Shiu, A., Lam, P.L. (2004), Electricity consumption and economic growth 
in China. Energy Policy, 32(1), 47-54.

Soytas, U., Sari, R. (2003), Energy consumption and GDP: Causality 



Aydın and Esen: Does Too Much Energy Consumption Harm Economic Growth for Turkish Republics in The Transition Process? New Evidence on 
Threshold Effects

International Journal of Energy Economics and Policy | Vol 7 • Issue 2 • 2017 43

relationship in G-7 countries and emerging markets. Energy 
Economics, 25(1), 33-37.

Stern, D.I. (2000), Multivariate cointegration analysis of the role of energy 
in the US macroeconomy. Energy Economics, 22(2), 267-283.

Stern, D.I., Cleveland, C.J. (2004), Energy and Economic Growth. 
Rensselaer Working Papers in Economics. WP. No: 0410. USA: 
Rensselaer Polytechnic Institute.

Wesseh, P.K., Zoumara, B. (2012), Causal independence between 
energy consumption and economic growth in Liberia: Evidence 
from a non-parametric bootstrapped causality test. Energy Policy, 
50, 518-527.

Wolde-Rufael, Y. (2006), Electricity consumption and economic growth: 
A time series experience for 17 African countries. Energy Policy, 
34(10), 1106-1114.