66
M. Conte Grand
in their measurement of “ a comprehensive measure of wealth.” Other alternatives
to measure progress are happiness or subjective satisfaction indices (Helliwell et al.
2016), or, as proposed by Jakob y Edenhofer (2014) a “dashboard of welfare indicators” as the Sustainable Development Indicators. The problem in this latter case is
that, as stated by Fleurbaey and Blanchet (2013), the difficulty in measuring welfare
is the multiplicity of indicators. There is no agreement on how social progress can
be measured.
Decoupling: Previous Evidence
Beyond any definition of words, there are studies dealing with indicators to actually
measure how GDP and carbon emissions decouple. To date, two of them are the
most employed. One is the decoupling factor introduced in OECD (2002), defined
by the rate of growth of emissions’ intensity (emissions/GDP). It states that there
is decoupling if emissions’ intensity decreases. Unfortunately, it has clear limitations. Decoupling is only associated to a reduction in emissions’ intensity, but that
scenario can coexist with emissions increasing while the economy is expanding and
with emissions decreasing but economic activity falling. The second indicator was
introduced by Tapio (2005) and is defined as an emissions-to-economic activity elasticity (rate of emissions’ change/rate of GDP change). Depending on the value of
this elasticity, there are several types of decoupling scenarios, whose description is
the main contribution of Tapio (2005).
2
Decoupling indicators have been used in several studies to analyze the link
between energy, environment and economy. For example, Lu et al. (2007) calculate
decoupling in Germany, Taiwan, South Korea and Japan on a yearly base between
1990 and 2003 using the OECD indicator. They find coupling between environmental
pressure (transportation CO 2 emissions and energy demand) and GDP except for
several years in the first two countries. Freitas and Kaneko (2011), using the same
indicator, examine the case of Brazil from 1980 to 2009 and uncover substantial
separation between economic activity and CO 2 emissions from energy consumption. Conrad and Cassar (2014) calculate the OECD indicator for several endpoints
in the small island of Malta and uncover relative decoupling for greenhouse gases
from 1995 to 2011. Gupta (2015) uses that same index to study decoupling for several
environmental (not only carbon emissions) endpoints in OECD countries.
Ren and Hu (2012) find different degrees of decoupling for the Chinese nonferrous metals industry in the period 1996–2008 using the Tapio (2005) decoupling
index. Zhang and Wang (2013) employ it for decoupling between CO 2 emissions of
the whole industry and primary, secondary and tertiary industries in a province of
2 A third measure of decoupling was introduced by Lu et al. (2011) and employed by Wang et al.
(2013). Its formula includes, in addition to GDP growth, the emissions’ intensity decreasing rate.
The three indices can be compared and, in fact, as shown in Conte (2016), Lu et al. (2011) and
Tapio (2005) indicators are one a linear transformation of the other.
M. Conte Grand
in their measurement of “ a comprehensive measure of wealth.” Other alternatives
to measure progress are happiness or subjective satisfaction indices (Helliwell et al.
2016), or, as proposed by Jakob y Edenhofer (2014) a “dashboard of welfare indicators” as the Sustainable Development Indicators. The problem in this latter case is
that, as stated by Fleurbaey and Blanchet (2013), the difficulty in measuring welfare
is the multiplicity of indicators. There is no agreement on how social progress can
be measured.
Decoupling: Previous Evidence
Beyond any definition of words, there are studies dealing with indicators to actually
measure how GDP and carbon emissions decouple. To date, two of them are the
most employed. One is the decoupling factor introduced in OECD (2002), defined
by the rate of growth of emissions’ intensity (emissions/GDP). It states that there
is decoupling if emissions’ intensity decreases. Unfortunately, it has clear limitations. Decoupling is only associated to a reduction in emissions’ intensity, but that
scenario can coexist with emissions increasing while the economy is expanding and
with emissions decreasing but economic activity falling. The second indicator was
introduced by Tapio (2005) and is defined as an emissions-to-economic activity elasticity (rate of emissions’ change/rate of GDP change). Depending on the value of
this elasticity, there are several types of decoupling scenarios, whose description is
the main contribution of Tapio (2005).
2
Decoupling indicators have been used in several studies to analyze the link
between energy, environment and economy. For example, Lu et al. (2007) calculate
decoupling in Germany, Taiwan, South Korea and Japan on a yearly base between
1990 and 2003 using the OECD indicator. They find coupling between environmental
pressure (transportation CO 2 emissions and energy demand) and GDP except for
several years in the first two countries. Freitas and Kaneko (2011), using the same
indicator, examine the case of Brazil from 1980 to 2009 and uncover substantial
separation between economic activity and CO 2 emissions from energy consumption. Conrad and Cassar (2014) calculate the OECD indicator for several endpoints
in the small island of Malta and uncover relative decoupling for greenhouse gases
from 1995 to 2011. Gupta (2015) uses that same index to study decoupling for several
environmental (not only carbon emissions) endpoints in OECD countries.
Ren and Hu (2012) find different degrees of decoupling for the Chinese nonferrous metals industry in the period 1996–2008 using the Tapio (2005) decoupling
index. Zhang and Wang (2013) employ it for decoupling between CO 2 emissions of
the whole industry and primary, secondary and tertiary industries in a province of
2 A third measure of decoupling was introduced by Lu et al. (2011) and employed by Wang et al.
(2013). Its formula includes, in addition to GDP growth, the emissions’ intensity decreasing rate.
The three indices can be compared and, in fact, as shown in Conte (2016), Lu et al. (2011) and
Tapio (2005) indicators are one a linear transformation of the other.