14-10-2016, 02:53 PM
1459042041-dco2italy.pdf (Size: 4.46 MB / Downloads: 15)
Abstract
This paper investigates major determinants of CO2 emissions in a small open economy
such as Italy over the period 1960-2012 using Granger causality and cointegration
methods to ascertain short-run and long-run relationships between emissions, trade
openness and energy consumption. The research findings do not support a possible
decoupling between economic growth and energy consumption, so that energy
conservation policies are expected to have a negative impact on economic growth.
Therefore, the use of environmentally friendly and renewable energy sources, such as
solar, hydro and wind power, should be further encouraged instead of fossil fuels ones.
Highlights
►CO2 emissions, economic growth, trade openness and energy consumption are
cointegrated
►Economic growth is a strong and positive driver of emissions in the short-run
►Support for feedback hypothesis between economic growth and energy consumption in
both the short-run and long-run
►Granger causality running from emissions to economic growth and energy consumption,
but no evidence of reverse causality
►Energy conservation policy will reduce economic growth
1. Introduction
The environmental impact of economic activities has received increasing attention
from academics and researchers, politicians and the society all together in recent decades.
The wide use of fossil fuels has been one of the most important stimuli of economic
growth. The nexus between pollution and economic development and the use of natural
resources has been explained by the environmental Kuznets curve which hypothesizes an
inverted-U relationship between pollution and economic development (Kuznets, 1955).
Initially, when a country’s per capita income is low environmental degradation will
increase, but may decline with higher per capita income over time. Or, in other words,
environmental pressure increases faster at early stages of development and then slows
down relative to economic growth at higher levels of development. Environmental
degradation might even be reduced in absolute terms.
The literature argues from an empirical point of view that there are three streams
of research looking at the link between economic growth and environmental pollution.
The first strand explores the relationship between economic growth and environment
degradation by testing the validity of the environmental Kuznets curve hypothesis.
Empirical evidence has not yet reached a consensus (Agras and Chapman, 1999; Dinda,
2004; Friedl and Getzner, 2003; Grossman and Krueger, 1995; Kearsley and Riddel, 2010;
Liu, 2005; Selden and Song, 1994; Stern et al., 1996; Suri and Chapman, 1998). The second
stream of research explores the relationship between economic growth and energy
consumption (Akarca and Long, 1980;Kraft, 1978; Yu and Hwang, 1984). To infer the
relationship between economic growth and environmental pollution, empirical studies
make out that economic growth and energy consumption are in close relation to each
other. Granger causality analysis with cointegrated variables applied to bivariate
regression models (Bentzen and Engsted, 1993; Ghali and El-Sakka, 2004) and multivariate
analysis (Apergis and Payne, 2009b; Lee, 2005; Soytas and Sari, 2003) appear to dominate
this literature that aims at identifying the direction of both short-run and long-run
causality in the relationship between the two variables. Overall, the specifications of
econometric models have suffered from omitted variable bias yielding mixed results
(Ozturk, 2010; Payne, 2010a, b). A third stream of research has emerged which combines
the previous two strands by examining dynamic relationships between economic growth,
energy consumption and pollution emissions (Apergis and Payne, 2009a, 2010; MartínezZarzoso
and Maruotti, 2011; Omri, 2013; Poumanyvong and Kaneko, 2010; Saboori et al.,
2012; Sari and Soytas, 2007; Shahbaz et al., 2013; Wang et al., 2011). Growing concern
over climate change has given rise to a new literature, mainly panel-based research,
devoted to investigate linkages between economic growth, energy consumption and
pollutant emissions. Many empirical studies posit a nonlinear quadratic relationship
according to the environmental Kuznets hypothesis (Ang, 2007; Halicioglu, 2009; Ozturk
and Acaravci, 2013). The empirical studies typically determine Granger causality in the
short-run and long-run sense and somehow do not pay attention to the measurement of
the size and direction of short-term and long-term parameters among the variables of
interest. As the literature stands, the research provides significant evidence on the drivers
of CO2 emissions for a larger set of countries such as industrialized and newly
industrialized countries, emerging economies and less regarding small open economies
within a single-country setting (Ang, 2008; Apergis and Payne, 2009a; Chandran and Tang,
2013; Ozturk and Acaravci, 2010; Shahbaz et al., 2011; Sharma, 2011; Soytas et al., 2007;
Zhang and Cheng, 2009).
As far as Italy is concerned, the empirical evidence is firmly based on multi-country
studies applying panel unit root, panel cointegration, and panel causality techniques. Total
energy consumption has a statistically significant impact on economic growth (Huang et
al., 2008; Narayan et al., 2010). One study finds a unidirectional long-run causality running
from GDP per capita to energy consumption per capita (Lee and Chang, 2007), whereas
another a reversed relationship (Lee, 2006). Another study that exclusively examines the
long-run relationship between energy consumption and real GDP finds a bidirectional
causal relationship between these two variables (Belke et al., 2011). In contrast, there is
only bidirectional short-run causality and unidirectional long-run causality from energy
consumption to economic growth (Acaravci and Ozturk, 2010). Moreover, a study finds a
reciprocal causal relationship among real income, real energy price, and total energy
consumption, and a unidirectional causality running from income and electricity price to
electricity consumption (Lee and De Lee, 2010). The results for the panel as a whole
suggest that the demand for total energy and electricity in the OECD countries is driven
largely by strong economic growth, while consumers are largely insensitive to price
changes. On top of that, further empirical results suggest bidirectional causality between
primary energy consumption and real GDP in both the long-run and short-run, supporting
the feedback hypothesis (Fuinhas and Marques, 2012). Focusing on electricity
consumption, some scholars find evidence in favour of electricity consumption causing
real GDP in Italy without being able to identify any causal relationship (Narayan and
Prasad, 2008).
The aforementioned studies have primarily based their findings on cointegration
analysis and mainly on multi-country evidence. It is somehow surprising to observe that these papers report separate results for Italy. Although the Italian economy has a
relatively small energy market and limited domestic energy resources, the rapid increase
in the service-based sectors have placed significant pressure on energy consumption in
the past years. Italy has a strong industrial basis and is highly dependent on fossil fuels so
that the reduction of CO2 emissions represents a serious environmental challenge for this
economy. Therefore, the question on how energy conservation may be viable without
being detrimental to economic growth might be re-examined with time-series data to
discuss differences in results for the case of Italy. Moreover, it is noticeable that the
primarily goal of the published literature has not been on examining the drivers of
pollutant emissions, and therefore estimating the size and direction of short-run and longrun
parameters is of interest. This paper is a contribution attempting to partly fill these
empirical and policy related gaps.
The remainder of the paper is structured as follows. Section 2 presents the
econometric model, along with the data and the methods of estimation. Section 3 reviews
and discusses the main empirical findings. Section 4 concludes and suggests further
research directions.