TÜRKİYE’DE ENERJİ TÜKETİMİ GSYH İLİŞKİSİ: DİNAMİK BİR ANALİZ
ENERGY CONSUMPTION GDP RELATIONSHIP FOR TURKEY: A DYNAMIC ANALYSIS
Hasan Murat ERTUĞRUL
Enerji tüketiminin GSYH üzerindeki etkileri Türkiye gibi enerji tüketiminde dışa bağımlılığı yüksek olan ülkelerde özellikle politika yapıcıları için önem arz eden bir konudur. Çalışmada enerji tüketimi GSYH ilişkisi 1970-2011 dönemleri arasında Türkiye ekonomisi için literatürden farklı olarak dinamik olarak incelenmiştir. Çalışmada önce, GSYH ve enerji tüketimi serileri arasındaki eş bütünleşme ilişkisi incelenmiş ve seriler arasında eş bütünleşme ilişkisi bulunmuş, daha sonra seriler arasındaki dinamik ilişki zamana göre değişen parametre yaklaşımı olan Kalman Filtresi modeliyle analiz edilmiştir. Çalışma bulgularına göre, enerji tüketimindeki %1’lik bir artışın GSYH üzerinde yol açacağı yüzdelik artışı ifade eden GSYH’nin enerji tüketimi esnekliğinin 1980-2003 döneminde genel olarak bir azalma eğiliminde olduğu, 2003 yılından itibaren ise söz konusu esneklik değerinin artmaya başladığı görülmektedir. Bu bulgular, Türkiye ekonomisinde özellikle 2003 yılından itibaren sanayide gözlenen, yüksek katma değerli ve enerji tüketiminin yoğun olduğu sektörlere doğru gerçekleşen dönüşümle uyumludur. Söz konusu esneklik değerinin önümüzdeki dönemde artış eğilimini devam ettirmesi halinde, GSYH artışı için enerji tüketiminin öneminin artacağı görülmektedir. Bu durumda Türkiye’nin enerjide yüksek dışa bağımlılığı da düşünüldüğünde enerji arz güvenliği sorunu politika yapıcılar için önemli bir husus olarak ön plana çıkmaktadır. Bu sebeple enerji dengesi içerisinde yerli kaynakların payının mümkün olduğunca arttırılması ve ithal kaynaklarda ise mümkün olduğu kadar ithalatçı ülke çeşitliliğine gidilmesi enerji arz güvenliği için son derece önemlidir.
Energy is an indispensable input of production and its importance was neglected until
the oil crisis in the 1970s. Especially two oil crises in the 1970s showed the importance of
energy which was started to be added to production function as a production factor
(Ertugrul, 2011). Especially with the widespread use of econometrics package programs,
the relationship between energy consumption and GDP growth was started to be tested
empirically and a broad literature was formed in this area.
Import dependency of Turkey is high in energy. Primary energy demand of Turkey in
2011 was 115 million TEP (tones equivalent of petroleum). The share of natural gas in
primary energy demand is 32 percent and the share of coal, hydro and renewable is 31, 27
and 4 percent respectively. As of 2011, import dependency of Turkey is 90% in oil and
98% in natural gas. Import dependency ratio of Turkey is 73% in energy (ETKB, 2012-
2014 Strategic Plan).
Since import dependency is high in energy, an important part of Turkey’s current
account deficit arises from energy imports. When the current account balance of Turkey
for 2011 was 77 Billion USD Dollars, the energy import in the same year was 54 Billion
USD Dollars. While the share of current account deficit to GNP was 9.9 percent in 2011,
current account without energy which is obtained by subtracting energy import from
current account balance is 2.9 percent only. High import dependency in energy causes
current account balance to worsen and risk premium of Turkey to increase. Therefore,
proper analysis of energy consumption and GDP relationship in Turkey is crucial for
Various studies in the applied literature analyze the relationship between energy
consumption and economic growth. First papers in the empirical literature focus on
causality analysis. The pioneering paper in the literature was Kraft and Kraft (1978)
paper. In Kraft and Kraft (1978) paper, they investigated energy consumption and GNP
relationship for USA for 1947-1974 periods and found unidirectional granger causality
from GNP to energy consumption. Akarca and Long (1980) replicated Kraft and Kraft
(1978) paper by only shortening the sample periods 2 years and they couldn’t find
causality relationship between energy consumption and GNP. Erol and Yu (1987)
investigated causality relationship between energy consumption and GNP for United
Kingdom, France, Italy, Germany, Canada and Japan. They found bidirectional causality
relationship for Japan, unidirectional causality relationship from energy consumption to
GNP for Canada, unidirectional causality relationship from GNP to energy consumption
for Germany and Italy and no causality for France and United Kingdom (Erol and Yu,
There are huge literateure which focus on causality relationship between energy
consumption and GDP relationship for different countries and different sample periods.
Some of these papers are Stern (1993 and 2000), Masih and Masih (1997), Yang (2000),
Soytaş, Sarı and Özdemir (2001), Ghosh (2002), Soytaş and Sarı (2004), Altınay and Karagöl (2004), Shiu and Lam (2004) and Jumbe (2004) papers. Moreover, there are
papers in the literature which investigate causality relationship by using electricity
consumption instead of energy consumption as Ghosh (2002), Jumbe (2004), Mozumder
and Marathe (2007), Rufael (2006), Altınay and Karagöl (2005), Karagöl, Erbaykal and
Ertuğrul (2007) and Ertuğrul (2011) papers.
The other methods employed in empirical literature which focused on energy
consumption and GNP relationship are co- integration models, panel data models and
VAR models. Some papers used these models are as Soytaş and Sarı (2004), Rufael
(2006) and Altınay and Karagöl (2004) papers.
There is no consensus in the literature for the energy consumption and GDP
relationships. There are mixed results for different countries and even same countries for
different time periods.
There is a lack of studies which analyze the relationship between energy consumption
and economic growth dynamically in the literature. Papers in the literature use static
models and focus on average trends.
In this study, different from literature, energy consumption and GDP relationship is
investigated dynamically by employing Kalman Filter method for Turkey.
In empirical model, we employ real GDP and energy consumption data for Turkey for
the 1970-2011 periods. Real GDP series were obtained from TURKSTAT (Turkish
Statistical Institute) and energy consumption series are obtained from Ministry of Energy
and Natural Resources of Turkey. Real GDP and energy consumption series are measured
in natural logarithms similar to the empirical literature and natural logarithms of real GDP
and energy consumption were denoted as LY and LE respectively.
In the emprical analysis, firstly stationarity properties of variables are checked. For
stationarity check, we employed ADF, PP and Ng-Perron unit root tests which are
commonly used in the empirical literature. According to ADF, PP and Ng-Perron unit root
test results, energy consumption and GDP series are found stationary after differencing
that means integrated of order one (which is denoted as I(1)).
After analyzing the stationarity properties of the series, co-integration relationship
between energy consumption and GDP is examined. Since both of the series are stationary
in the same order, co-integration relationship is analyzed employing Johansen’s method.
With respect to Johansen model results we found co-integration relationship between
energy consumption and GDP. According to the co-integration equation, energy
consumption elasticity of GDP is found as 0.98. This implies that 1 percent increase in
energy consumption will lead to an increase in the GDP by 0.98 percent on average in the
Lastly, we analyzed energy consumption and GDP relationship dynamically by using
time varying parameter method employing Kalman Filter model.
Our results show that energy consumption elasticity of GDP, which indicates the
percentage change in GDP in response to one percent change in energy consumption, has
a decreasing trend between 1980-2003 periods and begin to increase after 2003.
These findings are compatible with the transformation of Turkish industry towards
higher value added and energy intensive sectors since 2003. If the upward trend in this
elasticity continues in the forthcoming period, the importance of energy consumption for
GDP growth will increase. In this case, energy supply security is one of the most
important issues for policymakers considered together with Turkey’s high energy dependency. Therefore, it is very important to increase the share of domestic resources in
energy mix and diversify import countries for energy supply security of Turkey.