Policy Research Working Paper 9809
Understanding Drivers of Decoupling of Global
Transport CO2 Emissions from Economic Growth
Evidence from 145 Countries
Vivien Foster
Jennifer Uju Dim
Sebastian Vollmer
Fan Zhang
Infrastructure Chief Economist Office
October 2021
Policy Research Working Paper 9809
Abstract
This paper examines the extent to which countries have suc- countries have not achieved decoupling; their emissions are
ceeded in decoupling transport emissions from economic growing as fast as or faster than gross domestic product. To
growth, and how changes in emissions intensity, economic understand the driving factors of transport-related carbon
growth, and population growth have contributed to changes emissions, the paper conducts index-decomposition and
in transportation-related emissions. The paper employs an econometric analysis. The results reveal that while trans-
a modified version of the Tapio decoupling model, and portation emission intensity has declined in most countries,
demonstrates that over the 1990–2018 study period only economic growth and population growth have offset these
12 of 145 countries achieved “absolute decoupling,” defined declines. If these patterns continue, achieving the goals of
as reducing emissions while growing gross domestic product. the Paris Agreement with improvements in efficiency alone
The majority of the top emitters remain in a “relative decou- seems unrealistic. The paper also shows evidence that higher
pling” state, with emissions growing more slowly than gross energy prices are associated with strong emissions reduction.
domestic product. Many of the middle- and low-income
This paper is a product of the Infrastructure Chief Economist Office. It is part of a larger effort by the World Bank to
provide open access to its research and make a contribution to development policy discussions around the world. Policy
Research Working Papers are also posted on the Web at http://www.worldbank.org/prwp. The authors may be contacted
at fzhang1@worldbank.org.
The Policy Research Working Paper Series disseminates the findings of work in progress to encourage the exchange of ideas about development
issues. An objective of the series is to get the findings out quickly, even if the presentations are less than fully polished. The papers carry the
names of the authors and should be cited accordingly. The findings, interpretations, and conclusions expressed in this paper are entirely those
of the authors. They do not necessarily represent the views of the International Bank for Reconstruction and Development/World Bank and
its affiliated organizations, or those of the Executive Directors of the World Bank or the governments they represent.
Produced by the Research Support Team
Understanding Drivers of Decoupling of Global Transport CO2 Emissions from Economic
Growth: Evidence from 145 Countries
Vivien Foster, Jennifer Uju Dim, Sebastian Vollmer, Fan Zhang
Keywords: Decoupling, Transport carbon emissions, Economic growth, Population
JEL Classification: O18, O44, Q54
1. Introduction
The transportation sector, which relies heavily on fossil fuels, is a major contributor to total global
greenhouse gas emissions. In 2018, emissions from transport activities reached a level equivalent to 8.2
gigatons of CO2, representing roughly 24 percent of all global CO2 emissions. Carbon emissions from
transport have more than doubled since 1970. This rapid increase in emissions from transportation is
expected to continue, largely because the demand for transportation is likely to increase as global
population grows and incomes rise. According to the International Energy Agency (IEA) Energy Technology
Perspectives report, between 2019 and 2070, global transport demand is expected to double, and car
ownership rates are projected to grow by 60 percent (IEA, 2020). According to the International Panel on
Climate Change (IPCC), emissions linked to transportation will likely increase faster than those from any
other sector unless the link between transportation and economic growth can be severed (IPCC et al.,
2014).
Reducing transport-related carbon emissions without undermining economic growth is pivotal to
combating climate change and maintaining living standards. An effective transport emissions-reduction
strategy requires a deep understanding of the degree to which countries have broken or kept the
connection between the growth of transportation emissions and the growth of their economies. Such
strategy also demands a clear understanding of the drivers of transportation related emissions of CO2.
Thus, this paper aims to answer the following questions: To what extent have countries succeeded in
decoupling economic growth from transport carbon emissions? What policies can facilitate decoupling?
To answer these questions, first, we explore the issue of decoupling of economic growth from transport-
related carbon emissions at the country level. We use a modified version of the Tapio decoupling model
(Tapio, 2005). We define three decoupling states: absolute decoupling, relative decoupling, and no
decoupling. “Absolute decoupling” occurs when CO2 emissions are decreasing or stabilized even though
GDP is increasing. “Relative decoupling” occurs when both GDP and CO2 emissions are increasing, but
CO2 emissions are increasing at a slower rate. “No decoupling” occurs when a country is in neither of
these two states.
Only 12 countries achieved absolute decoupling over the study period. The remaining countries
experienced either relative decoupling or no decoupling. This implies that transport emissions in most
countries, including the top-emitting countries, are still increasing, and that the growth of their emissions
2
remains strongly linked to their economic growth. Hence, to achieve decarbonization in the transport
sector without prejudice to economic growth, there is a dire need to break this link.
The long-term decoupling results show considerable differences in the decoupling status among countries
by income levels. While many advanced economies have managed to achieve absolute or relative
decoupling, in many developing countries per capita transport emissions are increasing at a much faster
rate than per capita GDP growth. Increases in transport CO2 emissions were particularly pronounced in
the two largest middle-income countries: China and India. Both are still contributing less to global
transport CO2 emissions than their population shares, unlike the United States or the EU countries, but
they are rapidly catching up and are to a large extent driving current increases in global transport CO2
emissions.
Next, to explore the underlying drivers of CO2 emissions, we conduct both index decomposition and
econometric analyses. For the decomposition analysis we specify two identities. The first decomposition
breaks down emissions into three factors: emission intensity measured by transportation emissions per
unit of GDP, economic growth (in terms of GDP per capita), and population growth. In many developed
countries, declining emission intensity contributes to the decoupling of transportation emissions from
economic growth. In most developing countries, although emission intensity also fell, population and GDP
growth outstripped these efficiency-related transport emission savings.
In the second decomposition analysis, we further break down emission intensity (measured as CO2
emissions per unit of GDP) into carbon intensity of fuel consumption, energy intensity of transportation,
and transport demand. We focus on the inland transportation sector, including rail and road sectors for
which disaggregated data are available on travel volume measured by kilometers traveled by passengers
and freight. Inland transportation is the largest source of global transport emissions, accounting for 84
percent of total emissions in 2018. The results again show that efficiency improvements have played an
important role in flattening the emissions curve. Most countries have managed to reduce the energy
intensity of inland transport. By contrast, there has so far been minimal progress in reducing the carbon
intensity of energy consumption, and that concentrated in a handful of countries. Gasoline and diesel are
still the dominant fuels for transportation; renewable fuels and electricity accounted for about 10 percent
of the energy mix as of 2018.
To further explore the economic drivers of transport carbon emissions, we conduct country-level
regression analysis to understand the associations between CO2 emissions and GDP, urbanization,
3
structural changes, energy prices and public transit. Urbanization is first associated with an increase then
a decrease in emissions – possibly because urbanization initially coincides with a rise in vehicle ownership
and then later allows for economies of scale through improved public transit. The results also show that
per capita CO2 emissions are negatively correlated with agriculture value-added share. Higher energy
prices and the presence of transportation policies (such as fuel economy standards and tailpipe emissions
control) are both significantly correlated with lower per capita CO2 emissions.
The remainder of the paper is structured as follows: Section 2 reviews the existing literature. Section 3
describes the methodology. Section 4 presents the data. Section 5 discusses results from index
decomposition and the econometric analyses. Section 6 concludes with a summary of the main findings
and a discussion of policy implications.
2. Related literature
This study contributes to two strands of the literature: the decoupling literature and the literature on the
drivers of carbon emissions from the transportation sector. Decoupling analyses have been conducted in
different contexts – at the global level (Shuai et al., 2019), national level (Chen et al., 2018; de Freitas &
Kaneko, 2011; Jiang et al., 2016; Q. Wang et al., 2018; Zhang et al., 2020), provincial level (FY Fan, 2016;
JJ Jiang, 2017; A. Li et al., 2017; Q. Lu et al., 2015; Wu et al., 2019), and sectoral levels, such as agriculture,
construction, and transportation (Bai et al., 2019; Hang et al., 2019; Huo et al., 2021; M & C, 2019).
Decoupling analysis examines the relationship between environmental pressure and economic prosperity.
The relationship between economic growth and carbon emissions has been widely studied using the
decoupling indices created by the Organisation of Economic Development and Co-operation (OECD) and
Tapio (OECD, 2002; Tapio, 2005). Although the OECD decoupling index is easier to calculate, it is sensitive
to the choice of the base period, leading to poor stability in calculated results. In addition, the OECD
decoupling index is associated with emission intensity reductions only. It does not specifically address the
percentage change of emissions in relation to GDP growth (Grand 2016). The Tapio decoupling index
addresses these shortcomings. In the Tapio model, decoupling states are divided into three categories
and eight subcategories. We adopt the Tapio model but slightly modify and simplify the categorization of
the decoupling states.
Many studies have explored the determinants and drivers of carbon emissions from the transportation
sector. For instance, (Lakshmanan & Han, 1997) employ a decomposition analysis to identify the
magnitude and the relative effects of the various factors in U.S. transportation energy use and carbon
4
emissions between 1970 and 1991. They show that the growth in the propensity to travel, population,
and GDP were the three most important factors driving U.S. transportation energy use and CO2 emissions.
Timilsina & Shrestha (2009a, 2009b) analyze the factors influencing the growth of carbon dioxide (CO2)
emissions from the transportation sector in selected Asian and Latin American and Caribbean countries.
Using the Log Mean Divisia Index (LMDI) approach, they decomposed annual emissions growth into
several factors; they find that changes in GDP per capita and transportation energy intensity are the main
factors driving transport-sector CO2 emission growth.
Several decomposition methods can be used to understand the drivers of the changes in transport carbon
emissions including index decomposition, variance decomposition and structural decomposition.
However, the index decomposition method is the most commonly used approach in the literature, with
the Laspeyres and Divisia being the most popular (Ang, 2004; F. Li et al., 2019; & Shrestha, 2009b;
Yasmeen et al., 2020). The index decomposition approach allows the change in emissions to be
decomposed into several factors, including GDP per capita, population, energy intensity, fuel mix, fuel
carbon intensity, modal structure, and car ownership. The approach (W. W. Wang et al., 2011). We
employ this approach to identify the driving forces of the past changes in transport CO2 emissions.
Some papers have explored the determinants of transport emissions using regression techniques
including, dynamic panel quantile regressions, dynamic nonparametric additive regressions, and fully
modified ordinary least squares regressions (Huang et al., 2020; Saboori et al., 2014; Xu & Lin, 2015).
Typically, these papers have employed either an index decomposition analysis or a regression approach,
but not both. We use both methods to obtain a robust understanding of the determinants of global
transport emissions. The fixed effects and generalized method of moments regressions expand the results
of the index decomposition by including additional economic and policy factors that influence the change
in global transport emissions.
Furthermore, previous decomposition analyses were conducted at either the country level (Georgatzi et
al., 2020; Kwon, 2005; Liang et al., 2017; I. J. Lu et al., 2007; Rasool et al., 2019; Q. Wang et al., 2018),
regional level (Amin et al., 2020; Andreoni & Galmarini, 2012; Timilsina & Shrestha, 2009a, 2009b), or city
level (Fan & Lei, 2016; Li et al., 2019; Wang et al., 2011). There have been mixed results in the literature.
While some studies find that population growth decreases carbon emissions (I. J. Lu et al., 2007; Xie et al.,
2017), others find a positive relationship between population growth and carbon emissions (Fan & Lei,
2016; Kim, 2019; Timilsina & Shrestha, 2009b). In addition, the majority of the studies focused on road
transportation, specifically passenger cars (González et al., 2019; Papagiannaki & Diakoulaki, 2009; Shiraki
5
et al., 2020). None of the existing papers has conducted a comprehensive study that looks at the main
factors affecting the growth of all types of transport emissions on a global scale.
In sum, our paper contributes to the literature in three ways: First, we provide the literature’s first
comprehensive study of transport-related emissions globally; our data allow us to examine the situation
in 145 countries over an extensive period, from 1990 to 2018. Such in-depth analysis of the trends,
patterns, and drivers of transport emissions of countries in different development stages is important for
the design of effective decarbonization policies for the transportation sector. Second, the study provides
evidence of the degree of decoupling between economic growth and emissions in the transport sector
that has taken place – or failed to occur – in a large number of countries with widely differing economic
conditions and income levels. Third, we estimate associations of transport carbon emissions with various
country level indicators. These analyses provide helpful insights for the design and implementation of
effective emission-reduction policies in the transportation sector, which are urgently needed to address
its role as a driver of climate change.
3. Methodology
In this section, we present the methodology used to conduct the decoupling analysis and to examine the
sources and drivers of CO2 emissions in the transport sector. First, we employ the Tapio decoupling model
to identify the decoupling status of countries. Second, we conduct an index decomposition analysis using
two identities. The first identity investigates how changes in CO2 emissions are driven by changes in
population, economic growth, and transportation carbon intensity (measured by transportation emissions
per unit of GDP). The second identity further examines how changes in transport emission intensity are
driven by changes in carbon intensity of transport energy consumption, transport energy intensity, and
transport demand. Third, we explore the economic determinants of CO2 emissions and emission intensity
based on a panel regression analysis.
3.1. Tapio decoupling model
The formula for the Tapio decoupling model is given by:
% ∆������������������������2 (������������������������2������������ − ������������������������20 )/������������������������20
������������ = =
% ∆������������������������������������ (������������������������������������������������ − ������������������������������������0 )/������������������������������������0
Where ∆������������������������2 and ∆������������������������������������ denote the change of transport carbon emissions and economic growth
between a base year 0 to a target year t respectively. Different from Tapio (2005) in which the decoupling
6
status is divided into eight categories depending on the magnitude and sign of the elasticity (������������), we define
the decoupling status into three categories – absolute decoupling, relative decoupling, and no decoupling.
1. Absolute decoupling: occurs when carbon emissions are decreasing, and GDP is increasing, i.e.
∆������������������������2 < 0 and ∆������������������������������������ > 0. This follows that ������������ < 0.
2. Relative decoupling: occurs when both carbon emissions and GDP are increasing, but GDP is
increasing at a faster rate. Here ∆������������������������2 > 0 and ∆������������������������������������ > 0. This follows that 0 < ������������ < 1.
3. No decoupling: is defined as residual category where no absolute or relative decoupling occurs.
3.2. Index decomposition analysis
To conduct the decomposition analysis of the driving forces of CO2 emissions, we specify two identities
starting from an IPAT equation, one in which environmental impact (I) is a product of population (P),
affluence (A), and technology (T) (Commoner et al., 1971; Ehrlich & Holdren, 1971). According to Kwon
(2005), the IPAT formula is a valuable starting point to understand the determinants of past changes in an
aggregate environmental indicator to inform future environmental policies. Our first IDA identity is given
by:
������������������������������������������������ ������������������������2������������������������
������������������������2������������ = � ������������������������������������������������ ∗ ∗ (1)
������������������������������������������������ ������������������������������������������������
������������
where j and t represent transportation modes (road, rail, domestic navigation, domestic aviation, and
pipeline transport) and year, respectively. CO2 are transport-related carbon emissions; GDP represents
gross domestic output in million dollars based on 2015 PPP rates; and POP refers to the population size.
Equation (1) can be rewritten as:
������������������������2������������ = � ������������������������������������������������ ∗ ������������������������������������������������������������������������ ∗ ������������������������������������������������ (2)
������������
where CO2 represents the environmental-impact (I) variable in the IPAT identity. The first term on the
right-hand side of equation (2) measures the population effect (P). The second term is GDP per capita, a
measure of economic growth and a proxy for affluence (A). The third term is the emissions intensity per
unit of GDP, which is a proxy for technology (T).
In addition to the first decomposition equation, we specify another identity to further examine the factors
affecting the changes in carbon intensity measured by CO2 emissions per unit of GDP in inland
transportation for which more disaggregated data on travel volumes are available. The identity is given
by equations (3) below:
7
������������������������2������������ ������������������������2������������������������ ������������������������������������������������������������������������������������������������ ������������������������������������������������������������ ������������������������������������
=� ∗ ∗ (3)
������������������������������������������������ ������������������������������������������������������������������������������������������������ ������������������������������������������������������������ ������������������������������������ ������������������������������������������������
������������
where ENECONS represents transport energy consumption. Total KM represents total transport turnover,
i.e., total kilometers traveled by both passenger and freight. The first and second terms on the right-hand
side of equation (3) represent the carbon intensity of transport energy consumption and transport energy
efficiency, respectively. The third term represents transportation demand normalized by GDP. By further
breaking down carbon intensity (measured by emissions per unit of GDP) into carbon intensity of fuel and
fuel efficiency, equation (3) complements equation (1) to help understand the importance of various
factors in determining the trajectory of carbon emissions from transport.
Various index-decomposition approaches have been used in the literature to isolate the impact of one
variable from the other in determining the changes in CO2 emissions. In this paper, we adopt the
Logarithmic Mean Divisia Index (LMDI) approach because it provides a perfect decomposition, i.e., the
changes in the aggregate indicator are fully explained by predefined factors, and the decomposition
results do not leave an unexplained residual term. The LMDI is consistent in aggregation, meaning that
the results obtained from subgroups can be aggregated to a higher aggregation level in a consistent
manner (Ang & Liu, 2001).
Applying the multiplicative LMDI formula to equation (2), the change in the transport CO2 emissions from
the base year zero to year t is given as:
������������������������2������������ ������������������������������������������������ ������������������������������������������������������������������������ ������������������������������������������������
= ������������������������������������ � � ������������������������ ������������������������ � × ������������������������������������ �� ������������������������ ������������������������ � × ������������������������������������ �� ������������������������ ������������������������ � (4)
������������������������20 ������������������������������������0 ������������������������������������������������������������0 ������������������������������������0
������������ ������������ ������������
�������������������������2������������������������ – ������������������������2������������0 �/������������������������� ������������������������2������������������������ – ������������������������ ������������������������2������������0 �
where ������������������������ = ������������������������������������ ������������������������2������������������������ ≠ ������������������������2������������������������
(������������������������2������������ – ������������������������20 )/(������������������������ ������������������������2������������ – ������������������������ ������������������������20 )
������������������������ = ������������������������2������������������������ ������������������������������������ ������������������������2������������������������ = ������������������������2������������������������
The terms on the right-hand side of equation (4) quantify the relative contribution of each term on the
changes in transport emissions. The first term on the right-hand side of equation (4) represents the
population effect. The second term represents the economic-growth effect. The third term represents the
emissions-intensity effect. A similar formula is applied for equation (3).
8
3.3 Panel regression analysis
Lastly, we estimate a dynamic model to understand the effects of various economic factors on the level
of transport-related CO2 emissions. We assume current level of per capita emissions depends on the level
of per capita emissions in the previous year, that is, transport activities respond to changes in economic
factors with some lags. The model is specified as follows:
������������������������2������������������������������������������������ = ������������0 + ������������1 ������������������������2������������������������������������,������������−1 + ������������2 ������������������������������������������������������������������������������������ + ������������3 ������������������������������������������������������������2������������������������ + ������������4 ������������������������������������������������������������������������������������������������ + ������������5 ������������������������������������������������������������������������������������������������������������������������ +
������������6 ������������������������������������������������������������������������������������ + + ������������7 ������������������������������������������������������������ + ������������������������������������������������ + ������������������������ + ������������������������ + ������������������������������������ (5)
Where ������������������������2������������������������������������������������ is the natural logarithm of per capita CO2 emissions in country i in year t. GDPPCit is per
capita GDP. GDPPC2 is the square term of GDPPC. Diesel and Gasoline are diesel and gasoline prices,
respectively. Urban is the percentage of the total population living in urban areas. BRT represents the
number of Bus Rapid Transit (BRT) systems a country has. ������������������������������������ is a vector of variables, including the value-
added share of agriculture, manufacturing, and service sectors. ������������������������ is an unobserved country fixed effect,
which includes country-specific characteristics that are fixed over time, such as culture, climate zone, and
government regulation. We also include year fixed effects, ������������������������ , to control for common cyclical components
such as a common technology shock. ������������������������������������ is an idiosyncratic error term. ������������0 − ������������7 ������������������������������������ ������������ are parameters to
be estimated. We opted for a relatively modest set of explanatory variables because due to limited data
availability the inclusion of some further explanatory variables would have substantially cut the sample
size.
4. Data
Data used in this study come from several sources. The CO2 emissions and energy consumption data are
from International Energy Agency (IEA) Fuel Combustion Statistics database and World Energy Statistics
and Balances database, respectively. The databases report CO2 emissions and final energy consumption
due to fuel combustion of transportation activities at the country level. The transportation sector is
further divided into road, rail, domestic navigation, domestic aviation, and pipeline transport. The study
period is from 1990 to 2018. The full sample contains 145 countries for which data are available. In 2018,
the full sample of countries contributed to 84 percent of global transport carbon emissions. International
aviation and marine bunkers account for the remaining 16 percent. The list of countries included in the
study is presented in Table A1 in the appendix. We obtain PPP-based GDP (based on 2015 PPP rates) and
population data from IEA.
9
For the decomposition analysis of emissions intensity, we obtain annual country-level total passenger and
freight transport traveled by road and rail from the World Road Statistics and the International Transport
Forum databases. Total passenger transport, measured in million passenger-kilometers, refers to the total
movement of passengers using inland transportation. It represents the transport of one passenger for one
kilometer. Total freight transport, measured in million tons-kilometer, refers to the total movement of
goods using inland transport. It represents the transport of one ton over one kilometer. The total
kilometer variable is available for 65 countries during the period from 2000 to 2018. These countries are
listed in Table A2 in the appendix.
The level of urbanization, and the value added from agriculture, manufacturing, and service sectors are
all obtained from the World Bank World Development Indicators database. Data on country-level BRT
systems are obtained from global BRT database. Diesel and gasoline prices are compiled from fuel price
documentation from the Deutsche Gesellschaft für Internationale Zusammenarbeit GmbH (GIZ). The
summary statistics of variables are presented in Table 3.
5. Results
In this section, we describe the results of the decoupling analyses, the general pattern of transport-related
CO2 emissions, and the results of the regression analyses. The descriptive statistics reveal that CO2
emissions of the transport sector account for 24 percent of global CO2 emissions, making the transport
sector the second-largest contributor to global CO2 emissions after the energy sector, which accounts for
42 percent of such emissions. Within the transport sector, road transportation is the predominant driver,
accounting for 74 percent of all CO2 emissions; international sea transport accounts for 9 percent, and
international aviation accounts for 7 percent (Figure 1). The composition of global transport CO2
emissions by mode has been fairly stable over time; the shares in 1990 were about the same as those in
2018 (Figure A1). There is interesting variation between income groups and regions; the CO2 emissions
from road transportation are slightly below the global average in high-income countries, and above
average in lower-middle and low-income countries (Figure A3).
Figures 2a and 2b present the distribution of global transport CO2 emissions in terms of GDP and
population by income group and major emitting countries in 2018. The area of each rectangle represents
the contributions of a country/income group to total global transport emissions. The rectangle's width
represents the share of global GDP (population), and the height represents transport emissions per unit
of GDP (population). Figure 2a suggests that the countries with the most carbon-intensive transport
10
sectors are the Russian Federation, Brazil, Mexico, and India. Excluding China, the transport sectors in
middle-income countries are more carbon intensive than those of high-income countries. In terms of the
fuel mix, Russia has a high fraction of “other” fuels, composed of mainly natural gas, compared to other
countries and aggregates. The high share of biofuels in Brazil’s transportation-sector fuels also stands out.
Figure 2b provides additional insights on the global carbon footprint of the transport sector. In 2018, the
global average per capita emissions from the transport sector were 1.09 ton of CO2. Per capita emissions
in the United States were five times the world average. Per capita emissions of other high-income
countries were three times the world average, and those from the EU countries were double the world
average. Although middle-income countries at that time had relatively carbon-intensive transport sectors,
their per capita emissions were significantly lower than those of high-income countries.
In absolute levels, the findings show that high-income countries originated as many transport CO2
emissions as all low- and middle-income countries combined (Figure 3). However, this is rapidly changing
due to high emissions growth in middle-income countries. Lower-middle-income countries have more
than tripled their transport CO2 emissions, and upper-middle-income countries have more than doubled
their transport CO2 emissions since 1990. In 1990, high-income countries still generated about twice the
level of transport-related CO2 emissions as those created by low- and middle-income countries combined.
By 2018, levels generated by high-income countries, and by the combination of low- and middle-income
countries were about even. In per capita terms, the gap between high- and middle-income countries is
still wide. A person in a high-income country is generating around three times as much in the way of
transport CO2 emissions as a person in a middle-income country. But this gap is also rapidly decreasing.
Per capita transport CO2 emissions in middle-income countries have doubled since 1990, while they
remained relatively stable in high-income countries. Since 1990, transport-related CO2 emissions per unit
of GDP have declined for all income groups, but obviously not enough to make up for the large increases
in GDP that we have observed over the same period. We show these trends by world regions in the
appendix (Figure A2).
The United States is by far the leading contributor to global transport CO2 emissions; it generates 21
percent of such emissions, while accounting for less than 5 percent of the world’s population. The United
States is followed by China and the EU-28 countries (still including the United Kingdom), which both
contribute around 11 percent to global transport CO2 emissions (Figure 4). The EU-28 countries contribute
slightly more than their share of the world population and China contributes less than its world population
share. However, transport CO2 emissions in China are rapidly increasing, experiencing a tenfold increase
11
since 1990 (Figure 5). Transport CO2 emissions have also increased by around 20 percent in the United
States. Among the lower-middle-income countries, India is growing rapidly, and has increased its
transport CO2 emissions by a factor of five since 1990; it now contributes around 4 percent of global
transport CO2 emissions, still much less than its population share. There is not a single low-income country
that accounts for more than 1 percent of the transport CO2 emissions compared to the United States.
Looking at the fuel mix, we see that motor gasoline and gas/diesel oil account for the vast majority of the
transport fuel mix, and today both account for around 35 percent of fuels used. Since 1990, the share of
motor gasoline has steadily declined, and the share of gas/diesel oil has steadily increased. The share of
the remaining fuel types has remained relatively stable (Figure 6). Consequently, motor gasoline and
gas/diesel oil also account for the majority of transport CO2 emissions.
Only a dozen countries achieved absolute decoupling over the study period. The remaining countries
experienced either relative decoupling or no decoupling. This implies that transport emissions in most
countries, including the top-emitting countries, are still increasing and strongly linked to economic growth.
Hence, in order to achieve decarbonization in the transport sector, there is an urgent need to break this
link. This is especially the case in middle- and low-income countries. While about 72 percent of high-
income countries are in the absolute or relative decoupling states, only 29 percent of middle- and lower-
income countries have achieved some level of decoupling.
Moving onto the decomposition analysis, the first decomposition results show that since 1990, transport
related CO2 emissions have increased by around 80 percent, despite sizeable improvements in efficiency.
The results also reveal that these efficiency improvements did not fully compensate for the increase in
emissions due to population growth. The net increase in transport CO2 emissions is almost in sync with
the increase that is due to GDP growth (Figure 7). If these patterns continue, achieving the goals of the
Paris Agreement with improvements in efficiency alone seems unrealistic. According to the UN population
projections, the world population will continue to grow rapidly and reach about 10 billion people by 2050.
Then population growth will gradually slow down, and the world population is projected to peak at around
11 billion people in 2100. Notwithstanding the uncertainty surrounding these projections, the slowdown
in population growth will likely come too late to allow further efficiency improvements to outweigh the
increases in emissions due to economic growth.
If we look at the results of the index decomposition by income group, we see that in high-income countries
the improvements in efficiency fully compensate for the effect of population growth, and even partially
12
outweigh the effect of economic growth. The efficiency improvements in upper-middle income countries
are almost as large as in high-income countries, but there they are only just enough to outweigh the
effects of population growth such that total transport-generated CO2 emissions continue to rise with
growth of GDP. Efficiency improvements are much less in lower-middle-income countries, such that in
these countries increases in transport-related CO2 emissions are driven by both sizeable population
growth and GDP growth (Figure 8).
The first decomposition results for top-emitting countries reveal several interesting points (Figure 9). All
countries shown, except Japan and Russia, reveal increases in transport emissions since 1990. In line with
the descriptive analysis, the decomposition results suggest that China experienced the most significant
increase in transport-related CO2 emissions since 1990. This increase was largely driven by economic
growth, which is not surprising, given that the Chinese economy is among the fastest-growing economies
in the world, with an average annual GDP growth rate of 9.5 percent between 1990 and 2018.
Transportation emissions intensity in China declined slightly, but population growth remained relatively
stable. In contrast to China, in India, the country with the second-largest increases in transportation
emissions, the increases have been driven by both population growth and economic growth.
Among the largest emitters, Japan is the only country that achieved absolute decoupling in the
transportation sector over the last three decades (i.e., it experienced economic growth while transport
CO2 emissions fell absolutely). Efficiency improvements and stable population growth in Japan were
sufficient to decrease transport-related emissions. On the other hand, the United States and the EU
countries achieved relative decoupling of transport emissions, meaning that GDP grew faster than
emissions.
To better understand the drivers of the changes in transportation emissions intensity, we further
decompose the emissions intensity of transportation (i.e., transport-generated CO2 emissions divided by
GDP) by carbon intensity of fuel, energy consumption of kilometers traveled, and total kilometers traveled
(Figure 10). Due to limited data and time coverage for total kilometers traveled, the base year for the
second decomposition is 2000. Nevertheless, we observe a similar decreasing trend of transport emissions
intensity as in the first decomposition. Transport emissions intensity in India and Mexico have remained
at the same level where they were in 2000. In India this happened because kilometers traveled per unit
of GDP increased, and energy consumption per kilometer decreased, balancing each other out. In Mexico
all indicators are at levels from 2000, with some movements in between. China also increased kilometers
13
traveled per unit of GDP, but its increase was more modest than that of India, and the reduction in energy
consumption per kilometer was sufficient to outweigh this increase.
The significant decrease in transport carbon intensity that contributed to the absolute decoupling of CO2
emissions in Japan was driven by improvements in the energy efficiency of the transport sector and a
decrease in transport demand. The carbon intensity of transport fuel consumption was relatively stable
in both countries. Results for Ukraine and Kazakhstan are similar to those of Japan and Russia. The United
States results reveal that the decrease in transport demand was the main contributing factor to the
decline in transport carbon intensity. Conversely, the main contributing factor for the decline of transport
carbon intensity in EU countries was the improvements in transport energy efficiency. This implies that
energy consumption per kilometer traveled in EU countries has declined since 2000.
We now turn to a regression analysis of the determinants of per capita transport CO2 emissions measured
by transport-generated CO2 emissions per unit of GDP. The first three columns of Table 2 report results
from static panel regression via OLS, the random-effects and fixed-effects methods.. The last column
report results identified in equation (5) using the generalized methods of moments (GMM) technique
outlined in Arellano and Bond (1991). Because governments may respond to increasing pollutions and
growing traffic congestion by raising energy prices, energy price could be endogenous to carbon emissions.
To address the potential endogeneity concern, we use observations on energy price lagged two to four
periods as instruments in the GMM model. The coefficient associated with the lagged CO2 emissions is
statistically significant, suggesting that the current level of emissions is indeed strongly correlated with
the level of emissions in the previous period. The dynamic panel regression reported in column (4) of
Table 2 is therefore our preferred specification.
Controlling for country and year fixed effects, the parameter estimates suggest that per capita GDP,
energy price, urbanization, and structural transformation are strongly correlated with transport related
emissions. The squared term of GDP per capita is negative and statistically significant, but if we calculate
the maximum of the inverted U, the point where the slope changes from positive to negative, we find that
it is out of sample, and that all observations are on the increasing part of the parabola.
The estimates of economic structure measured by the relative share of the value added of agriculture,
manufacturing, and services show that as countries move away from agriculture, transport-related
emissions decreases. Another interesting finding that emerged from the results is that CO2 emissions first
increase and then decrease with the increase of the percentage of people living in urban area. The turning
14
point occurs at 58 percent of urbanization. Finally, higher diesel prices is associated with lower per capita
carbon emissions. The coefficients associated with gasoline prices are of the expected sign but are not
statistically significant. Because gasoline prices are highly correlated with diesel prices, including the diesel
price in the regression could absorb most of the explanatory power.
In the above analysis, we do not control for the existence of transportation policies aimed at reducing
transport-related emissions. As a robustness check, we collect data from IEA transportation policy
database and control for the existence of various transportation policies in the estimation. We use the
principal component analysis method to create one comprehensive transport decarbonization policy
variable that summarizes several transportation policies: fuel efficiency standards, tailpipe emissions
regulations, driving restrictions, promotion of electric vehicles, promotion of biofuels, R&D and vehicle
emission standards. Including transportation policy variable in the regression reduces the sample size by
half due to limited data availability for many developing countries. The results are reported in Table 3.
The existence of transportation policies has a strong impact, sharply decreasing per capita emissions. The
coefficients of the other variables have the expected sign, but the statistical significance reduces due to
the smaller sample.
6. Conclusions
Emissions from the transportation sector – which includes travel of people and goods by road, rail, aircraft,
and marine vessels; and the pipeline transportation of many fuels – represented more than 24 percent of
global CO2 emissions in 2018. Worldwide, emissions from transportation are increasing faster than those
from any other sector. Decarbonization of the transport sector therefore plays a crucial role in achieving
the Paris Agreement target of limiting global temperature rise to 2 degrees Celsius – and preferably to
below 1.5 degree Celsius – compared to pre-industrial levels.
This paper examines the status of the decoupling of transport-generated carbon emissions and economic
growth of 145 countries. In addition, it studies the patterns and drivers of transport-related CO2 emissions
of 145 countries during the period from 1990 to 2018. We observe considerable differences in the
decoupling status of countries according to their income levels. Most high-income countries are in the
relative decoupling states while most of the countries with low incomes are in no decoupling states. This
implies that there is considerable room for breaking the linkage between economic growth and carbon
emissions in low- and middle-income countries.
15
The empirical results in this paper point to several opportunities for developing countries to decouple CO2
emissions from transportation. First, transport’s reliance on fossil fuels needs to shift dramatically.
Renewables and electricity still account for a negligible share of transportation fuel. Thus, there is large,
unexploited potential to reduce CO2 emissions from mobility by leveraging technological advances (in
batteries, for example) that have made “green” vehicles possible and more practical transportation
options than previously has been thought. Nevertheless, the impact of electrification on transport-related
CO2 emissions will be much higher once the power system itself has been decarbonized. Second,
continued energy price reforms to phase out subsidies for fossil fuel could lead to a sizable reduction in
emissions. Third, rapid urbanization in developing countries could provide both challenges and
opportunities for the decarbonization of transportation. On the one hand, the high concentration of
people and activities in cities could lead to a rise in vehicle ownership and traffic congestion. On the other
hand, urbanization allows for the development of complex public transportation systems and other
economies of scale to facilitate emissions reductions. It is important for policy makers to integrate land-
use planning and transportation options to encourage low-emission mobility so as to avoid urbanization
being associated with a phase of extremely high carbon intensity of transportation.
16
References
Amin, A., Altinoz, B., & Dogan, E. (2020). Analyzing the determinants of carbon emissions from
transportation in European countries: the role of renewable energy and urbanization. Clean
Technologies and Environmental Policy, 22(8), 1725–1734. https://doi.org/10.1007/s10098-020-
01910-2
Andreoni, V., & Galmarini, S. (2012). European CO2 emission trends: A decomposition analysis for water
and aviation transport sectors. Energy, 45(1), 595–602.
https://doi.org/10.1016/j.energy.2012.07.039
Ang, B. W. (2004). Decomposition analysis for policymaking in energy: Which is the preferred method?
Energy Policy, 32(9), 1131–1139. https://doi.org/10.1016/S0301-4215(03)00076-4
Ang, B. W., & Liu, F. L. (2001). A new energy decomposition method: Perfect in decomposition and
consistent in aggregation. Energy, 26(6), 537–548. https://doi.org/10.1016/S0360-5442(01)00022-
6
Bai, C., Chen, Y., Yi, X., & Feng, C. (2019). Decoupling and decomposition analysis of transportation
carbon emissions at the provincial level in China: perspective from the 11th and 12th Five-Year
Plan periods. Environmental Science and Pollution Research 2019 26:15, 26(15), 15039–15056.
https://doi.org/10.1007/S11356-019-04774-2
Chen, J., Wang, P., Cui, L., Huang, S., & Song, M. (2018). Decomposition and decoupling analysis of CO2
emissions in OECD. Applied Energy, 231, 937–950.
https://doi.org/10.1016/J.APENERGY.2018.09.179
Commoner, B., Corr, M., & Stamler, P. J. (1971). The causes of pollution. Environment, 13(3), 2–19.
https://doi.org/10.1080/00139157.1971.9930577
de Freitas, L. C., & Kaneko, S. (2011). Decomposing the decoupling of CO2 emissions and economic
growth in Brazil. Ecological Economics, 70(8), 1459–1469.
https://doi.org/10.1016/J.ECOLECON.2011.02.011
Ehrlich, P. R., & Holdren, J. P. (1971). Impact of population growth. Science, 171(3977), 1212–1217.
https://doi.org/10.1126/science.171.3977.1212
Fan, F., & Lei, Y. (2016). Decomposition analysis of energy-related carbon emissions from the
transportation sector in Beijing. Transportation Research Part D: Transport and Environment, 42,
135–145. https://doi.org/10.1016/j.trd.2015.11.001
FY Fan, Y. L. (2016). Decomposition analysis of energy-related carbon emissions from the transportation
sector in Beijing. Transp Res D, 42, 135–145. https://doi.org/10.1016/j.trd.2015.11.001
Georgatzi, V. v., Stamboulis, Y., & Vetsikas, A. (2020). Examining the determinants of CO2 emissions
caused by the transport sector: Empirical evidence from 12 European countries. Economic Analysis
and Policy, 65, 11–20. https://doi.org/10.1016/j.eap.2019.11.003
17
González, R. M., Marrero, G. A., Rodríguez-López, J., & Marrero, Á. S. (2019). Analyzing CO2 emissions
from passenger cars in Europe: A dynamic panel data approach. Energy Policy, 129, 1271–1281.
https://doi.org/10.1016/j.enpol.2019.03.031
Hang, Y., Wang, Q., Zhou, D., & Zhang, L. (2019). Factors influencing the progress in decoupling
economic growth from carbon dioxide emissions in China’s manufacturing industry. Resources,
Conservation and Recycling, 146, 77–88. https://doi.org/10.1016/J.RESCONREC.2019.03.034
Huang, Y., Zhu, H., & Zhang, Z. (2020). The heterogeneous effect of driving factors on carbon emission
intensity in the Chinese transport sector: Evidence from dynamic panel quantile regression. Science
of The Total Environment, 727, 138578. https://doi.org/10.1016/J.SCITOTENV.2020.138578
Huo, T., Ma, Y., Yu, T., Cai, W., Liu, B., & Ren, H. (2021). Decoupling and decomposition analysis of
residential building carbon emissions from residential income: Evidence from the provincial level in
China. Environmental Impact Assessment Review, 86, 106487.
https://doi.org/10.1016/J.EIAR.2020.106487
IEA. (2020). Energy Technology Perspectives 2020. www.iea.org/t&c/
IPCC, Edenhofer, O., Sokona, Y., Minx, J. C., Farahani, E., Kadner, S., Seyboth, K., Adler, A., Baum, I.,
Brunner, S., Kriemann, B., Savolainen Web Manager Steffen Schlömer, J., von Stechow, C., &
Zwickel Senior Scientist, T. (2014). Climate Change 2014 Mitigation of Climate Change Working
Group III Contribution to the Fifth Assessment Report of the Intergovernmental Panel on Climate
Change. Cambridge University Press. www.cambridge.org
Jiang, X.-T., Dong, J.-F., Wang, X.-M., & Li, R.-R. (2016). The Multilevel Index Decomposition of Energy-
Related Carbon Emission and Its Decoupling with Economic Growth in USA. Sustainability 2016,
Vol. 8, Page 857, 8(9), 857. https://doi.org/10.3390/SU8090857
JJ Jiang, B. Y. D. X. J. T. (2017). Provincial-level carbon emission drivers and emission reduction strategies
in China: combining multi-layer LMDI decomposition with hierarchical clustering. J Clean Prod, 169,
178–190. https://doi.org/10.1016/j.jclepro.2017.03.189
Kim, S. (2019). Decomposition Analysis of Greenhouse Gas Emissions in Korea’s Transportation Sector.
Sustainability, 11(7), 1986. https://doi.org/10.3390/su11071986
Kwon, T. H. (2005). Decomposition of factors determining the trend of CO2 emissions from car travel in
Great Britain (1970-2000). Ecological Economics, 53(2), 261–275.
https://doi.org/10.1016/j.ecolecon.2004.06.028
Lakshmanan, T. R., & Han, X. (1997). Factors underlying transportation CO2 emissions in the U.S.A.: A
decomposition analysis. Transportation Research Part D: Transport and Environment, 2(1), 1–15.
https://doi.org/10.1016/S1361-9209(96)00011-9
Li, A., Zhang, A., Zhou, Y., & Yao, X. (2017). Decomposition analysis of factors affecting carbon dioxide
emissions across provinces in China. Journal of Cleaner Production, 141, 1428–1444.
https://doi.org/10.1016/J.JCLEPRO.2016.09.206
18
Li, F., Cai, B., Ye, Z., Wang, Z., Zhang, W., Zhou, P., & Chen, J. (2019). Changing patterns and
determinants of transportation carbon emissions in Chinese cities. Energy, 174, 562–575.
https://doi.org/10.1016/j.energy.2019.02.179
Liang, Y., Niu, D., Wang, H., & Li, Y. (2017). Factors Affecting Transportation Sector CO2 Emissions
Growth in China: An LMDI Decomposition Analysis. Sustainability, 9(10), 1730.
https://doi.org/10.3390/su9101730
Lu, I. J., Lin, S. J., & Lewis, C. (2007). Decomposition and decoupling effects of carbon dioxide emission
from highway transportation in Taiwan, Germany, Japan and South Korea. Energy Policy, 35(6),
3226–3235. https://doi.org/10.1016/j.enpol.2006.11.003
Lu, Q., Yang, H., Huang, X., Chuai, X., & Wu, C. (2015). Multi-sectoral decomposition in decoupling
industrial growth from carbon emissions in the developed Jiangsu Province, China. Energy, 82,
414–425. https://doi.org/10.1016/J.ENERGY.2015.01.052
M, W., & C, F. (2019). Decoupling economic growth from carbon dioxide emissions in China’s metal
industrial sectors: A technological and efficiency perspective. The Science of the Total Environment,
691, 1173–1181. https://doi.org/10.1016/J.SCITOTENV.2019.07.190
OECD. (2002). Indicators to Measure Decoupling of Environmental Pressure from Economic Growth.
Papagiannaki, K., & Diakoulaki, D. (2009). Decomposition analysis of CO2 emissions from passenger cars:
The cases of Greece and Denmark. Energy Policy, 37(8), 3259–3267.
https://doi.org/10.1016/j.enpol.2009.04.026
Rasool, Y., Zaidi, S. A. H., & Zafar, M. W. (2019). Determinants of carbon emissions in Pakistan’s
transport sector. Environmental Science and Pollution Research, 26(22), 22907–22921.
https://doi.org/10.1007/s11356-019-05504-4
Saboori, B., Sapri, M., & bin Baba, M. (2014). Economic growth, energy consumption and CO2 emissions
in OECD (Organization for Economic Co-operation and Development)’s transport sector: A fully
modified bi-directional relationship approach. Energy, 66, 150–161.
https://doi.org/10.1016/J.ENERGY.2013.12.048
Shiraki, H., Matsumoto, K., Shigetomi, Y., Ehara, T., Ochi, Y., & Ogawa, Y. (2020). Factors affecting CO2
emissions from private automobiles in Japan: The impact of vehicle occupancy. Applied Energy,
259, 114196. https://doi.org/10.1016/j.apenergy.2019.114196
Shuai, C., Chen, X., Wu, Y., Zhang, Y., & Tan, Y. (2019). A three-step strategy for decoupling economic
growth from carbon emission: Empirical evidence from 133 countries. Science of The Total
Environment, 646, 524–543. https://doi.org/10.1016/J.SCITOTENV.2018.07.045
Tapio, P. (2005). Towards a theory of decoupling: degrees of decoupling in the EU and the case of road
traffic in Finland between 1970 and 2001. Transport Policy, 12(2), 137–151.
https://doi.org/10.1016/J.TRANPOL.2005.01.001
Timilsina, G. R., & Shrestha, A. (2009a). Factors affecting transport sector CO 2 emissions growth in Latin
American and Caribbean countries: An LMDI decomposition analysis. International Journal of
Energy Research, 33(4), 396–414. https://doi.org/10.1002/er.1486
19
Timilsina, G. R., & Shrestha, A. (2009b). Transport sector CO2 emissions growth in Asia: Underlying
factors and policy options. Energy Policy, 37(11), 4523–4539.
https://doi.org/10.1016/j.enpol.2009.06.009
Wang, Q., Zhao, M., Li, R., & Su, M. (2018). Decomposition and decoupling analysis of carbon emissions
from economic growth: A comparative study of China and the United States. Journal of Cleaner
Production, 197, 178–184. https://doi.org/10.1016/J.JCLEPRO.2018.05.285
Wang, W. W., Zhang, M., & Zhou, M. (2011). Using LMDI method to analyze transport sector CO2
emissions in China. Energy, 36(10), 5909–5915. https://doi.org/10.1016/j.energy.2011.08.031
Wu, Y., Tam, V. W. Y., Shuai, C., Shen, L., Zhang, Y., & Liao, S. (2019). Decoupling China’s economic
growth from carbon emissions: Empirical studies from 30 Chinese provinces (2001–2015). Science
of The Total Environment, 656, 576–588. https://doi.org/10.1016/J.SCITOTENV.2018.11.384
Xie, R., Fang, J., & Liu, C. (2017). The effects of transportation infrastructure on urban carbon emissions.
Applied Energy, 196, 199–207. https://doi.org/10.1016/j.apenergy.2017.01.020
Xu, B., & Lin, B. (2015). Factors affecting carbon dioxide (CO2) emissions in China’s transport sector: a
dynamic nonparametric additive regression model. Journal of Cleaner Production, 101, 311–322.
https://doi.org/10.1016/J.JCLEPRO.2015.03.088
Yasmeen, H., Wang, Y., Zameer, H., & Solangi, Y. A. (2020). Decomposing factors affecting CO2 emissions
in Pakistan: insights from LMDI decomposition approach. Environmental Science and Pollution
Research, 27(3), 3113–3123. https://doi.org/10.1007/s11356-019-07187-3
Zhang, J., Fan, Z., Chen, Y., Gao, J., & Liu, W. (2020). Decomposition and decoupling analysis of carbon
dioxide emissions from economic growth in the context of China and the ASEAN countries. Science
of The Total Environment, 714, 136649. https://doi.org/10.1016/J.SCITOTENV.2020.136649
20
Figure 1: Sectoral CO2 emissions and global transport emissions by mode, 2018
(a) CO2 emissions by sector (b) Transport emissions by mode
Pipeline
2%
Others Rail
Domestic
8% Industry 1%
aviation
18%
5%
Road
74% Intl. marine
9%
Energy
42% Transport
25% Intl.
aviation
7%
Domestic
Agriculture navigation
/forestry 2%
Residential
6% 1%
21
Figure 2: CO2 footprint of the transport sector, 2018
22
Figure 3: Trends in transport carbon emissions by income group (1990 – 2018)
Figure 4: Transport CO2 emissions by country
LICs World
1% aviation
Other HICs Other MICs bunkers
9% 19% 7%
World
marine
bunkers
9%
EU-28 Mexico
11% 2%
Japan
3%
Russia
United 3%
China India
States
11% 4%
21%
23
Figure 5: Top transport CO2 emitters by income group
Figure 6: Global transportation fuel mix (1990 – 2018)
24
Figure 7: Trend in global transport emissions (1990 – 2018)
Figure 8: Trend in global transport emissions by income group (1990 – 2018)
25
Figure 9: Trend in global transport emissions by country (1990 – 2018)
Figure 10: Decomposition of global transport emissions per unit of GDP (1990 – 2018)
26
Table 1: Status of the level of decoupling of the link between transport emissions and GDP
90- 91- 92- 93- 94- 95- 96- 97- 98- 99- 00- 01- 02- 03- 04- 05- 06- 07- 08- 09- 10- 11- 12- 13- 14- 15- 16- 17-
country 91 92 93 94 95 96 97 98 99 00 01 02 03 04 05 06 07 08 09 10 11 12 13 14 15 16 17 18 1990-2018
High-income countries
Australia AD RD RD RD ND RD RD RD RD ND AD RD RD ND RD RD RD ND AD ND RD RD RD RD RD RD RD ND RD
Austria ND RD ND RD RD ND AD ND AD ND ND ND ND RD ND AD RD AD ND ND AD AD ND AD ND ND RD RD ND
Bahrain RD ND RD ND ND RD ND ND RD ND ND ND ND ND ND RD ND ND AD RD AD ND RD RD ND RD RD ND ND
Belgium RD ND ND RD RD ND RD ND RD RD ND AD ND ND AD RD RD RD ND AD AD AD AD RD ND AD AD RD RD
Brunei
Darussalam ND ND ND ND ND ND ND ND AD AD ND ND ND ND AD ND ND ND ND ND ND ND ND ND ND ND AD ND ND
Canada ND ND RD ND ND ND RD RD RD AD AD ND RD RD RD AD ND AD ND RD RD RD ND AD AD RD RD ND RD
Chile RD RD ND ND RD ND RD ND ND RD AD ND AD RD ND AD ND ND ND RD RD RD ND AD ND ND ND RD RD
Chinese Taipei RD ND ND RD RD RD RD ND RD RD ND RD AD RD RD AD AD AD ND RD RD AD AD RD ND ND AD AD RD
Croatia ND ND ND ND RD ND ND ND ND RD RD ND ND RD RD ND ND AD ND ND ND ND ND ND ND ND ND AD ND
Curacao ND ND ND ND ND RD RD ND RD RD ND ND RD ND ND RD RD RD ND ND ND ND AD ND ND ND ND ND ND
Cyprus AD ND ND RD RD ND ND RD RD RD RD AD ND ND RD AD ND RD ND ND AD ND ND ND ND RD RD RD RD
Czech Republic ND ND ND ND AD ND ND ND ND RD ND ND ND RD ND RD RD AD ND AD AD ND ND ND RD ND RD RD ND
Denmark ND AD ND RD RD RD RD RD RD AD RD ND ND ND RD RD ND ND ND AD AD AD AD RD ND RD RD RD RD
Estonia ND ND ND ND RD ND RD ND ND AD ND RD AD RD RD RD RD ND ND ND AD RD AD RD ND RD RD RD RD
Finland ND ND ND ND AD AD RD RD RD AD RD RD RD RD RD RD RD AD ND ND AD ND ND ND AD ND AD ND AD
France ND ND ND RD RD AD RD ND RD RD RD AD AD RD AD RD AD AD ND RD ND AD AD RD RD RD RD AD RD
Germany RD RD ND AD RD RD RD RD ND AD AD ND ND AD AD AD AD RD ND RD RD AD ND RD ND RD RD AD AD
Greece ND ND ND RD RD ND RD ND RD AD RD RD RD RD ND RD RD ND ND ND ND ND ND RD ND ND AD RD RD
Hong Kong
(China) ND ND ND ND AD AD RD ND ND AD AD AD AD AD ND AD AD AD ND AD AD AD RD RD ND ND ND AD RD
Hungary ND ND ND AD ND AD ND ND ND AD ND ND ND RD ND ND ND AD ND AD AD ND AD ND ND RD ND ND ND
Iceland ND ND AD ND ND ND AD RD RD RD RD ND RD RD ND ND RD AD ND ND AD AD RD RD RD ND ND RD RD
Ireland ND ND RD RD RD ND RD ND ND RD RD RD RD RD ND ND RD ND ND AD AD AD RD RD AD ND AD RD RD
Israel RD ND ND ND ND ND ND AD ND RD AD ND ND AD AD RD RD ND AD RD AD ND AD RD ND RD RD RD RD
Italy ND ND ND AD RD ND RD ND ND RD RD ND ND ND AD RD RD ND ND AD AD ND ND ND AD AD AD ND RD
Japan ND ND ND ND ND RD RD ND ND AD ND AD AD AD AD AD AD ND ND RD ND AD AD AD AD AD AD AD AD
Note: AD: Absolute decoupling (color-coded: dark green); RD: Relative decoupling (color-coded: light green); ND: No decoupling (color-coded: red). The columns are one-year-
interval time periods from 1990 to 2018.
1
90- 91- 92- 93- 94- 95- 96- 97- 98- 99- 00- 01- 02- 03- 04- 05- 06- 07- 08- 09- 10- 11- 12- 13- 14- 15- 16- 17-
country 91 92 93 94 95 96 97 98 99 00 01 02 03 04 05 06 07 08 09 10 11 12 13 14 15 16 17 18 1990-2018
High-income countries
Korea ND ND ND RD ND RD RD ND RD RD RD RD RD RD AD AD RD AD ND RD AD ND ND RD ND ND RD AD RD
Kuwait ND ND RD RD ND ND ND RD ND RD ND ND RD RD RD ND RD ND ND ND AD RD ND ND ND RD ND ND ND
Latvia ND ND ND AD ND AD RD AD AD ND ND RD RD RD RD RD ND ND ND ND AD AD RD ND ND RD ND RD RD
Lithuania ND ND ND AD ND ND ND RD ND AD ND RD RD ND RD RD ND AD ND ND AD RD AD ND ND ND ND ND RD
Luxembourg ND ND RD RD AD ND ND RD ND ND ND ND ND ND ND AD AD ND ND ND ND ND AD AD AD AD ND ND RD
Malta ND RD ND ND ND AD ND AD AD AD AD AD ND ND ND ND RD ND ND ND AD AD RD RD RD AD RD ND RD
Mauritius RD ND ND RD AD AD AD RD ND RD RD ND RD RD ND RD AD RD ND ND RD RD RD RD RD ND RD RD RD
Netherlands RD ND ND AD RD ND RD RD RD RD RD ND ND RD RD RD AD RD ND ND RD ND ND AD RD RD RD RD RD
New Zealand ND ND RD ND ND RD RD ND RD ND RD RD RD RD RD RD RD ND AD AD RD AD RD RD RD RD ND RD RD
Norway RD RD ND AD ND RD RD ND RD AD RD RD ND RD RD ND ND AD ND ND AD AD ND RD RD AD AD ND RD
Oman ND RD ND RD RD ND ND ND ND RD ND ND ND ND ND ND ND ND RD ND ND ND ND ND ND AD AD AD ND
Panama RD RD ND ND ND ND ND ND ND RD AD ND ND RD ND RD AD ND ND RD RD RD AD ND ND ND RD AD RD
Poland ND RD AD RD RD ND RD ND ND AD AD AD ND ND ND ND ND ND RD ND RD AD AD RD ND ND ND RD ND
Portugal ND ND ND ND RD ND RD ND ND ND RD ND ND ND AD AD RD RD ND AD ND ND ND RD RD ND RD RD ND
Qatar ND RD ND ND ND ND RD RD AD RD ND ND ND RD ND RD RD ND RD RD RD AD ND ND ND ND ND AD RD
Romania ND ND AD ND AD ND ND ND ND RD ND RD ND RD AD RD RD ND ND ND RD ND AD RD RD ND RD RD RD
Saudi Arabia RD ND ND ND AD ND ND RD ND RD ND ND RD RD RD ND ND ND ND RD ND ND RD ND ND AD ND AD ND
Singapore ND AD AD RD ND AD AD ND RD RD ND AD RD RD RD RD RD ND ND RD ND AD AD RD ND AD AD AD RD
Slovak Republic ND ND AD ND ND AD ND RD ND AD ND ND AD RD ND AD RD ND ND ND AD AD RD AD AD ND ND AD RD
Slovenia ND ND ND ND ND ND RD AD AD AD ND RD RD ND ND RD ND ND ND ND ND ND ND RD AD ND RD AD ND
Spain ND ND ND ND RD ND AD ND ND RD ND RD ND ND RD RD RD AD ND AD ND ND ND RD RD ND ND RD RD
Sweden ND ND ND RD RD AD RD RD RD RD AD RD RD RD AD RD RD ND ND RD AD ND AD AD RD AD AD AD AD
Switzerland ND ND ND ND AD RD ND RD AD ND AD RD ND RD RD RD RD ND ND AD AD RD AD AD AD AD AD RD RD
Trinidad and
Tobago ND AD ND ND ND RD ND RD RD RD RD RD AD ND RD ND ND ND ND ND ND AD AD ND RD ND ND ND RD
United Arab
Emirates ND RD ND AD RD RD AD AD ND RD ND ND ND RD ND RD RD ND ND ND RD RD ND RD AD ND ND AD ND
United Kingdom ND ND RD RD AD ND RD AD RD AD AD RD RD RD RD RD RD ND ND AD AD AD AD RD ND ND RD AD RD
United States ND RD RD RD RD RD RD RD RD RD RD RD RD RD RD AD RD ND ND ND AD AD ND AD ND RD RD RD RD
Uruguay ND RD ND ND ND ND ND ND ND ND ND ND AD ND RD ND ND RD ND RD RD RD ND AD ND ND ND AD ND
2
90- 91- 92- 93- 94- 95- 96- 97- 98- 99- 00- 01- 02- 03- 04- 05- 06- 07- 08- 09- 10- 11- 12- 13- 14- 15- 16- 17-
Country 91 92 93 94 95 96 97 98 99 00 01 02 03 04 05 06 07 08 09 10 11 12 13 14 15 16 17 18 1990-2018
Upper-middle-income countries
Albania ND ND ND ND AD AD ND ND ND RD RD ND ND ND RD AD ND ND RD AD ND AD ND RD AD AD RD AD ND
Argentina RD RD RD AD ND ND AD ND ND ND ND ND RD RD RD RD RD ND ND RD RD ND ND ND ND ND RD ND RD
Armenia ND ND ND AD AD AD ND RD ND ND AD AD RD RD RD AD ND ND ND ND RD RD AD ND AD AD RD ND AD
Azerbaijan ND ND ND ND ND AD ND AD AD ND ND ND RD ND RD RD AD ND AD ND ND ND ND ND AD ND ND ND RD
Belarus ND ND ND ND ND ND RD RD AD AD AD ND RD RD ND ND RD ND AD RD ND ND ND AD ND ND ND ND RD
Bolivarian Republic of
Venezuela RD AD ND ND RD ND AD ND ND RD ND ND ND RD ND RD RD RD ND ND ND ND AD ND ND ND ND ND ND
Bosnia and Herzegovina ND ND ND ND RD RD RD ND AD ND RD AD RD ND AD ND RD ND ND AD RD ND AD ND ND ND ND AD RD
Botswana ND ND AD AD ND RD RD ND RD ND ND AD ND ND RD RD RD ND ND RD RD RD RD RD ND ND RD RD ND
Brazil ND ND RD RD ND ND ND ND AD RD ND RD AD ND RD RD RD RD ND ND ND ND ND ND ND ND RD AD ND
Bulgaria ND ND ND AD ND AD ND ND ND AD RD RD ND ND ND RD AD RD ND AD ND ND AD ND ND RD RD RD RD
Colombia ND ND RD AD ND ND ND AD ND AD ND AD RD ND AD RD RD RD AD RD ND ND AD RD ND ND AD RD RD
Costa Rica ND ND RD ND AD RD RD ND RD AD ND ND RD ND AD RD RD RD ND RD RD RD RD RD ND ND RD RD ND
Cuba ND ND ND AD ND ND ND AD AD AD RD AD AD RD AD AD RD AD AD AD AD AD ND ND ND AD ND ND AD
Dominican Republic AD RD ND ND ND ND ND ND ND RD RD ND ND ND RD RD RD AD ND RD ND ND AD AD ND ND AD ND RD
Ecuador RD RD RD ND AD ND ND AD ND AD ND RD ND RD RD RD AD AD ND ND ND RD ND ND ND ND ND ND ND
Equatorial Guinea ND ND AD AD ND ND RD AD ND ND AD ND ND RD RD ND RD RD RD ND RD RD ND ND ND ND ND ND RD
Gabon ND ND ND AD ND RD ND ND ND ND ND ND AD ND AD ND ND ND ND ND ND RD ND AD AD ND AD ND AD
Georgia ND ND ND ND AD ND AD AD AD AD ND RD RD AD ND RD ND AD ND RD RD RD ND ND ND ND AD AD ND
Guatemala ND RD ND ND ND AD ND ND ND ND ND ND AD AD ND RD RD AD ND AD AD RD ND ND ND ND ND ND ND
Indonesia ND RD ND ND ND ND ND ND AD ND ND RD RD RD AD AD RD ND ND ND ND ND RD RD AD ND ND ND ND
Iraq ND RD RD RD AD AD AD RD RD ND ND ND ND RD ND AD AD AD RD ND ND RD AD AD AD RD ND ND RD
Islamic Republic of Iran RD ND ND ND AD ND ND ND ND ND ND ND RD ND ND ND AD ND ND AD RD ND ND ND ND AD RD ND ND
3
90- 91- 92- 93- 94- 95- 96- 97- 98- 99- 00- 01- 02- 03- 04- 05- 06- 07- 08- 09- 10- 11- 12- 13- 14- 15- 16- 17-
country 91 92 93 94 95 96 97 98 99 00 01 02 03 04 05 06 07 08 09 10 11 12 13 14 15 16 17 18 1990-2018
Upper-middle-income countries
Jamaica AD ND RD RD ND ND ND ND ND RD AD ND AD ND ND AD AD ND ND ND AD ND ND ND AD ND ND ND ND
Jordan AD RD RD RD ND RD ND RD RD ND RD RD RD ND ND AD RD AD ND RD RD ND ND AD ND ND ND AD RD
Kazakhstan ND ND ND ND ND AD AD ND AD ND AD ND AD RD RD RD ND ND AD RD RD ND AD ND ND ND AD ND RD
Kosovo ND RD ND RD ND RD ND RD ND ND AD RD RD AD ND ND ND ND ND ND
Lebanon RD ND ND AD ND RD AD ND ND AD AD RD ND RD RD AD AD ND ND AD RD ND AD ND ND ND ND AD RD
Libya RD ND ND ND ND ND ND ND ND RD ND ND RD RD RD RD AD ND ND ND ND RD ND ND ND ND AD AD ND
Malaysia RD RD RD ND RD ND ND ND ND RD ND RD ND ND AD AD ND ND ND RD RD AD ND ND AD RD AD RD RD
Mexico ND RD RD ND ND RD RD RD ND RD ND ND ND ND ND ND ND ND ND RD RD RD AD RD AD ND AD ND RD
Montenegro ND AD ND RD ND ND AD ND AD AD ND ND ND ND ND
Namibia ND ND ND ND ND ND RD ND RD AD ND AD ND RD ND AD RD ND ND RD RD AD ND RD ND RD ND ND ND
Paraguay AD ND ND ND ND RD ND ND ND ND ND ND ND AD AD ND ND RD ND ND ND ND RD ND ND ND ND ND ND
People's Republic of
China RD RD RD AD RD ND AD AD ND ND RD RD ND ND RD RD RD RD RD RD RD ND ND RD ND RD RD RD RD
Peru AD ND RD ND ND ND RD ND ND AD AD AD ND ND AD RD RD ND ND ND RD RD ND AD ND ND RD RD RD
Republic of North
Macedonia ND ND ND ND ND ND AD AD ND AD ND ND RD RD RD RD ND RD ND ND ND ND ND ND ND ND ND AD ND
Russian Federation ND ND ND ND ND ND AD ND RD AD RD RD RD RD AD RD RD ND ND ND RD AD RD ND ND AD ND ND AD
Serbia ND ND ND AD RD ND ND AD ND AD ND ND ND ND RD ND AD ND ND RD AD ND ND ND AD RD ND RD RD
South Africa ND ND RD ND ND AD ND AD RD AD RD RD ND ND RD RD ND AD ND ND ND RD ND AD ND AD ND RD RD
Suriname ND RD ND ND RD ND ND RD ND RD ND AD ND AD ND ND ND AD RD ND
Thailand RD RD ND ND ND ND ND ND RD AD RD RD RD ND AD AD ND AD ND RD ND RD AD AD ND ND ND RD RD
Turkey AD RD ND ND ND RD AD AD ND RD ND RD RD RD RD ND ND AD ND AD RD ND ND ND ND ND RD RD RD
Turkmenistan ND ND ND ND ND ND ND ND ND ND RD ND ND RD RD AD RD RD AD ND ND RD RD RD ND ND ND ND RD
4
90- 91- 92- 93- 94- 95- 96- 97- 98- 99- 00- 01- 02- 03- 04- 05- 06- 07- 08- 09- 10- 11- 12- 13- 14- 15- 16- 17-
country 91 92 93 94 95 96 97 98 99 00 01 02 03 04 05 06 07 08 09 10 11 12 13 14 15 16 17 18 1990-2018
Lower-middle-income countries
Algeria ND AD ND ND AD AD RD RD ND RD RD ND ND RD ND ND ND ND ND ND ND ND ND ND ND AD AD ND ND
Angola ND ND ND ND AD AD RD AD ND AD ND ND ND ND AD ND ND ND ND ND ND AD ND RD RD ND ND ND ND
Bangladesh ND ND RD RD ND ND ND RD AD AD ND RD RD ND ND RD RD ND ND ND ND RD AD ND RD RD ND ND ND
Benin AD AD ND ND ND ND ND ND ND ND ND ND ND RD ND ND ND RD ND ND ND RD ND RD ND ND RD ND ND
Cambodia ND RD RD ND AD AD ND AD RD RD RD RD ND RD ND ND RD RD RD ND ND ND RD ND ND
Cameroon ND ND ND ND AD RD ND RD AD ND RD ND RD RD AD AD ND RD ND ND RD RD RD ND AD RD AD AD RD
Cote d'ivoire AD ND ND AD ND RD ND AD ND ND AD ND ND ND ND ND RD ND RD AD ND ND RD RD ND ND AD ND ND
Egypt ND RD ND ND ND ND ND ND ND RD AD ND ND RD AD ND ND RD ND ND ND ND RD ND RD ND AD AD ND
El Salvador ND ND ND ND ND AD RD ND ND AD RD AD ND ND ND AD ND AD ND AD RD AD RD RD ND ND AD ND ND
Ghana AD ND RD ND ND ND ND ND ND AD AD ND AD ND AD RD ND RD ND RD RD ND RD RD ND AD AD ND ND
Honduras ND ND ND ND RD AD ND ND ND AD ND RD RD RD AD AD ND AD ND RD RD RD AD ND ND RD AD ND ND
India ND RD RD RD ND RD RD RD RD RD AD RD RD ND RD RD ND ND ND RD ND ND RD RD ND RD ND RD RD
Kenya ND ND AD AD ND ND AD RD AD ND AD ND AD ND RD ND RD ND ND ND AD ND ND ND ND ND AD ND ND
Kyrgyzstan ND ND ND ND ND RD RD AD AD AD RD ND ND ND ND ND ND ND ND ND ND ND RD AD AD ND AD AD AD
Lao People's Democratic Republic ND ND RD RD RD RD RD ND ND ND RD AD ND ND ND RD ND RD RD ND
Mongolia ND ND ND AD ND ND AD AD ND ND AD ND AD ND AD ND ND RD ND AD ND ND AD RD RD AD ND RD RD
Morocco ND ND ND RD ND RD ND RD ND ND RD ND RD RD ND RD ND ND ND ND ND RD RD RD ND ND RD ND ND
Myanmar ND AD ND ND ND ND AD ND RD ND AD ND RD RD RD AD RD AD RD RD RD ND ND ND ND RD ND RD RD
Nepal ND ND ND ND ND RD ND ND RD ND AD ND AD ND AD ND RD ND ND ND RD ND ND ND AD ND ND ND ND
Nicaragua ND ND ND ND ND RD ND ND RD RD RD ND AD AD AD RD RD AD ND RD RD RD AD ND ND RD ND ND ND
Nigeria ND ND ND ND ND ND ND AD ND ND ND RD RD RD AD AD AD ND AD ND ND ND ND ND AD ND RD ND ND
Pakistan RD ND ND ND RD ND ND ND ND AD AD RD ND ND AD RD ND AD RD ND ND RD RD ND ND ND ND RD ND
Philippines ND ND ND ND ND ND ND ND AD AD ND RD AD RD AD AD AD RD ND RD AD RD RD RD ND ND RD RD RD
Plurinational State of
Bolivia RD AD ND RD ND RD AD RD RD AD AD RD ND ND RD ND ND ND ND ND ND ND ND ND RD RD ND ND ND
Republic of Moldova ND ND ND ND ND ND AD ND ND AD RD ND ND ND AD RD ND ND ND ND ND ND RD RD ND ND RD RD AD
Republic of the Congo ND AD ND ND ND AD ND ND ND ND ND AD ND ND RD ND ND ND ND ND ND AD ND RD AD ND ND ND ND
Senegal ND ND AD ND RD ND ND ND ND ND AD ND ND ND AD ND ND ND AD ND ND AD ND RD ND ND AD RD ND
Sri Lanka RD ND ND ND ND ND AD RD ND AD ND RD ND RD ND AD ND AD AD ND RD AD ND ND ND ND ND AD RD
5
90- 91- 92- 93- 94- 95- 96- 97- 98- 99- 00- 01- 02- 03- 04- 05- 06- 07- 08- 09- 10- 11- 12- 13- 14- 15- 16- 17-
country 91 92 93 94 95 96 97 98 99 00 01 02 03 04 05 06 07 08 09 10 11 12 13 14 15 16 17 18 1990-2018
Lower-middle-income countries
Tunisia RD RD ND ND ND RD ND RD ND RD RD ND RD RD ND AD RD AD ND ND ND AD AD ND ND ND ND RD ND
Ukraine ND ND ND ND ND ND ND ND ND AD RD RD AD RD AD AD ND AD ND RD AD AD ND ND ND ND ND AD ND
United Republic of Tanzania AD ND ND ND ND ND AD ND AD ND ND ND RD RD ND RD AD ND AD ND ND ND ND AD RD AD RD RD ND
Uzbekistan ND ND ND ND ND ND RD ND RD ND RD AD AD AD RD RD AD RD AD RD AD AD AD AD AD AD ND ND RD
Viet Nam AD ND ND ND AD ND ND RD ND ND ND ND ND ND RD AD ND ND ND ND AD AD AD RD ND ND RD AD ND
Zambia ND ND AD ND AD AD ND ND AD ND RD ND RD RD RD RD AD ND ND RD ND ND RD ND ND AD AD AD RD
Zimbabwe AD ND AD AD ND AD RD AD ND ND AD ND ND ND ND ND ND ND ND RD ND RD ND AD AD AD RD ND ND
Low-income countries
Democratic People's Republic of
Korea ND ND ND ND ND ND ND ND ND ND ND ND ND ND ND AD ND ND ND ND ND AD ND AD ND ND AD AD ND
Democratic Republic of the Congo ND ND ND ND AD ND ND ND ND ND ND ND ND ND ND ND ND ND AD ND ND AD ND ND AD AD ND AD ND
Eritrea ND ND ND ND ND AD AD ND ND ND ND ND AD AD ND AD ND RD ND ND ND RD RD RD ND ND ND ND
Ethiopia ND ND ND ND ND RD ND ND RD ND ND ND ND RD AD RD RD RD RD RD RD ND ND ND ND ND ND ND ND
Haiti AD ND ND ND ND ND RD ND ND ND ND ND ND ND RD RD ND AD RD ND RD AD AD ND ND RD ND RD ND
Mozambique AD ND ND AD ND RD RD RD RD RD AD RD ND RD AD RD ND AD ND ND ND AD ND ND ND RD ND ND RD
Niger ND RD RD ND ND RD RD ND RD ND ND AD ND ND RD RD AD RD ND ND
South Sudan ND AD ND ND ND ND ND ND
Sudan AD AD AD ND AD AD ND AD AD ND ND ND AD ND ND ND RD ND AD ND ND AD ND ND ND ND AD ND RD
Syrian Arab Republic RD AD RD RD AD ND ND RD ND ND AD ND ND ND ND ND ND RD AD AD ND AD AD ND ND ND ND ND AD
Tajikistan ND ND ND ND ND ND ND AD AD AD ND ND AD ND AD ND ND AD AD ND RD ND RD ND AD AD AD RD ND
Togo ND ND ND ND ND ND AD ND ND ND ND ND ND AD AD AD ND ND ND AD AD AD AD AD ND AD ND ND RD
Yemen ND RD AD RD AD AD RD ND ND AD ND AD ND RD ND AD AD ND ND AD ND AD ND ND ND ND ND ND AD
Note: AD: Absolute decoupling (color-coded: dark green); RD: Relative decoupling (color-coded: light green); ND: No decoupling (color-coded: red). The columns are one-year-
interval time periods from 1990 to 2018.
6
Table 2 Correlates of transport-related carbon emissions per capita
(1) (2) (3) (4)
OLS Random Effects Fixed Effects GMM
GDP per capita 5.55e-08** 5.53e-08** 5.53e-08** 4.79e-08*
(1.72e-08) (1.94e-08) (1.94e-08) (1.99e-08)
GDP pc sq -2.01e-15* -1.38e-15 -1.38e-15 -1.38e-15*
(1.00e-15) (7.39e-16) (7.39e-16) (6.86e-16)
Agriculture -0.0597*** -0.0238*** -0.0238*** -0.0143***
(0.00228) (0.00165) (0.00165) (0.00131)
Manufacturing 0.00338 0.00232 0.00232 -0.00321*
(0.00212) (0.00160) (0.00160) (0.00137)
Service 0.00942*** -0.00140 -0.00140 -0.00151
(0.00157) (0.00111) (0.00111) (0.000775)
Urbanization 0.0317*** 0.0765*** 0.0765*** 0.0211***
(0.00377) (0.00457) (0.00457) (0.00385)
Urbanization sq -0.0000710* -0.000574*** -0.000574*** -0.000183***
(0.0000290) (0.0000373) (0.0000373) (0.0000324)
Diesel price 1.087*** -0.102* -0.102* -0.0632*
(0.0856) (0.0404) (0.0404) (0.0254)
Gasoline price -1.218*** -0.0528 -0.0528 -0.00659
(0.0843) (0.0403) (0.0403) (0.0257)
BRT 0.0383 -0.0720*** -0.0720*** -0.00156
(0.0384) (0.0190) (0.0190) (0.0175)
L.co2pc 0.723***
(0.0170)
Obs. 2816 2816 2816 2689
Year fixed effects Yes Yes Yes Yes
Standard errors in parentheses
="* p<0.05 ** p<0.01 *** p<0.001"
1
Table 3 Correlates of transport-related carbon emissions per capita (with transport policy)
(1) (2) (3) (4)
OLS Random Effects Fixed Effects GMM
Log GDP per capita -8.51e-08*** 9.31e-08*** 13.0e-08*** 2.89e-08***
(1.55e-08) (1.41e-08) (1.50e-08) (7.55e-09)
Log GDP pc sq 3.88e-15*** -2.39e-15*** -3.13e-15*** -8.21e-16***
(8.10e-16) (4.60e-16) (4.68e-16) (2.32e-16)
Agriculture -0.103*** -0.0436*** -0.0351*** -0.0128***
(0.00566) (0.00467) (0.00474) (0.00234)
Manufacturing -0.00213 -0.00262 -0.00240 -0.000847
(0.00302) (0.00152) (0.00148) (0.000719)
Service 0.0149*** -0.000240 -0.000670 -0.00358***
(0.00281) (0.00175) (0.00170) (0.000829)
Urbanization 0.0321*** 0.0777*** 0.0574*** 0.00889
(0.00739) (0.00874) (0.00951) (0.00471)
Urbanization sq -0.000151** -0.000487*** -0.000403*** -0.0000593
(0.0000525) (0.0000601) (0.0000637) (0.0000316)
Log diesel price 0.290** -0.00163 -0.0309 -0.0268
(0.101) (0.0465) (0.0452) (0.0220)
Log gasoline price -0.677*** -0.167*** -0.141*** -0.0349
(0.0955) (0.0391) (0.0378) (0.0184)
Transportation policy 0.117*** -0.0812*** -0.0898*** -0.0105**
(0.0214) (0.00769) (0.00746) (0.00393)
BRT -0.128*** 0.00421 0.0126 -0.00119
(0.0352) (0.0148) (0.0142) (0.00692)
L.ln_co2pc 0.820***
(0.0157)
Observations 1010 1010 1010 936
Year fixed effects Yes Yes Yes Yes
Note: Standard errors in parentheses
* p<0.05 ** p<0.01 *** p<0.001
2
Appendix
Table A1: List of countries included in the study
Albania Cyprus Ireland Netherlands Spain
Iran, Islamic
Algeria Czech Republic Rep. New Zealand Sri Lanka
Korea, Dem.
Angola People’s Rep. Israel Nicaragua Sudan
Argentina Congo, Dem. Rep. Italy Niger Suriname
Armenia Denmark Jamaica Nigeria Sweden
Australia Dominican Republic Japan Norway Switzerland
Austria Ecuador Jordan Oman Syrian Arab Republic
Azerbaijan Egypt, Arab Rep. Kazakhstan Pakistan Tajikistan
Bahrain El Salvador Kenya Panama Thailand
Bangladesh Equatorial Guinea Korea Paraguay Togo
Belarus Eritrea Kosovo China Trinidad and Tobago
Belgium Estonia Kuwait Peru Tunisia
Benin Ethiopia Kyrgyzstan Philippines Turkey
Venezuela, RB Finland Lao PDR Bolivia Turkmenistan
Bosnia and
Herzegovina France Latvia Poland Ukraine
Botswana Gabon Lebanon Portugal United Arab Emirates
Brazil Georgia Libya Qatar United Kingdom
Brunei Darussalam Germany Lithuania Moldova Tanzania
Bulgaria Ghana Luxembourg North Macedonia United States
Cambodia Gibraltar Malaysia Congo, Rep. Uruguay
Cameroon Greece Malta Romania Uzbekistan
Russian
Canada Guatemala Mauritius Federation Vietnam
Chile Haiti Mexico Saudi Arabia Yemen, Rep.
Chinese Taipei Honduras Mongolia Senegal Zambia
Hong Kong SAR,
Colombia China Montenegro Serbia Zimbabwe
Costa Rica Hungary Morocco Singapore
Mozambiqu
Côte d'Ivoire Iceland e Slovak Republic
Croatia India Myanmar Slovenia
Cuba Indonesia Namibia South Africa
Curacao Iraq Nepal South Sudan
3
Table A2: List of subset of countries
Argentina Denmark Kazakhstan North Macedonia Turkey
Armenia Ecuador Kenya Norway Ukraine
Australia Estonia Korea, Rep. Pakistan United Kingdom
Austria Finland Kyrgyzstan Peru United States
Azerbaijan France Latvia Poland Vietnam
Belarus Georgia Lithuania Portugal
Belgium Germany Luxembourg Romania
Bosnia and Herzegovina Greece Mexico Russian Federation
Bulgaria Hungary Moldova Serbia
Canada Iceland Mongolia Slovak Republic
Chile India Montenegro Slovenia
China Ireland Morocco Spain
Croatia Israel Myanmar Sweden
Cuba Italy Netherlands Switzerland
Czech Republic Japan New Zealand Tunisia
Table A3: Descriptive Statistics
Variable Obs Mean Std. Dev. Min Max
Log CO2 per capita 3,915 -0.4257 1.311756 -4.96185 2.755506
Log GDP per capita 3,828 9.361465 1.136264 6.154889 11.81192
Squared log GDP pc 3,828 88.92779 20.96595 37.88266 139.5214
Agriculture va 3,554 10.68645 10.43098 0.028407 63.83134
Manufacturing va 3,394 14.53169 6.285805 0 50.03699
Service va 3,454 52.31527 11.73463 10.56928 91.92164
Urbanization 3,915 61.91061 20.78944 8.854 100
Log diesel price 3,120 -0.55517 0.869384 -4.60517 0.854415
Log gasoline price 3,125 -0.3038 0.714075 -3.91202 0.932164
Existence of BRT 3,886 0.147967 0.355113 0 1
Existence of Metro 3,886 0.348945 0.476698 0 1
4
Figure A1: Global transport emissions by mode (1990-2018)
Figure A2: Trends in transport carbon emissions by regions (1990 – 2018)
Figure A3: Transport emissions by mode by income groups and regions, 2018
5
6