Global CO2 emissions went from increasing 3% per year on average from the 2001-2010 period to 1% per year from 2011-2019, raising hopes that a peak in global emissions may be near. These declines were primarily driven by the increasing deployment of lower CO2 energy sources and accelerating declines in the energy intensity of the global economy. Both of these trends are likely to accelerate in the future as clean energy technologies become more cost-effective and countries become wealthier. At the same time, current trends continuing will fall far short of the rapid emissions reductions needed for the world to meet Paris Agreement goals.
During the period from 2001-2010 global emissions were rapidly increasing at around 3% per year, raising fears that emissions could double or even triple by the end of the century. Global CO2 emissions from fossil fuels increased by 31% — from 25 to 33 gigatons (Gt) — over the course of the decade, while emissions in China increased by a stunning 150% – from 3.4 to 8.5 GtCO2. Rapid economic growth was driving a record expansion of coal use, particularly in China.
The period from 2011 to 2019 saw the situation change substantially. Global emissions continued to grow, but at only one-third the rate — 1% per year — compared to the prior decade. It now seems likely that the world is entering a long plateau in global emissions under current policies — a far cry from the world of rapidly increasing emissions that appeared all too plausible a decade ago.
Here we seek to explore the changing drivers of CO2 emissions and to understand what factors drove this large reduction in the rate of emissions growth. This is important, as it may help inform us better understand how the drivers of emissions growth — and ultimately declines — will change in the future. At the same time, we should also recognize that reduced emissions growth is only the first step toward the peaking and absolute reduction in global emissions needed to ultimately stop the world from warming.
CO2 emissions — and their growth over time — can be thought of as the result of four different factors: global population, GDP per person, energy intensity (per unit of economic activity), and CO2 intensity of energy. This is known as the Kaya identity and has often been used to assess the drivers of emissions over time, though it is only one of many possible ways to explore changes in emissions.
Figure 1, below, shows the drivers of global emissions growth for each year between 2000 and 2019 across the four different parts of the Kaya identity. A growing population and economy tend to drive emissions upward while falling energy use per GDP tends to push emissions downward. CO2 use per unit of energy was either flat or increasing from 2000 until 2012 — reflecting an acceleration of global coal use during that period. In recent years, however, the emissions intensity of energy has fallen as the world began to shift towards more clean energy sources.
To examine changes in decadal drivers of emissions, we first need to decide which specific periods we want to examine. Here we choose to compare two different periods: 2001-2010 and 2011-2019. The former period was selected to include both 2009 and 2010, as 2009 saw large declines in global (and country-level) GDP due to the global financial crisis, while 2010 saw a large recovery in global economic growth. Simply looking at 2000-2009 and 2010-2019 periods would artificially depress emissions in the former due to the great recession and inflate the latter due to the recovery. The year 2020 is not included in the more recent period both because country-level data is not yet available for all required variables, and because COVID-19-related disruptions in global economic activity make it unrepresentative of broader decadal-scale trends.
Figure 2, below, shows the drivers of CO2 emissions growth across both periods. The top panel shows the different drivers of emissions growth for the two periods, with total emissions growth shown by the bars on the right side of the graph. The bottom panel shows the difference in the contribution to CO2 emissions growth between the two periods for each factor.
During the 2001-2010 period, population growth drove global emissions upward by around 1.3% per year, and increasing GDP per capita increased emissions by 2.3% per year. Falling energy intensity reduced emissions by 1% per year while increasing emissions intensity of energy increased it by 0.3% per year. Added together, these factors result in the observed emissions increase of 2.8% per year.
During the 2011-2019 period, global population growth was similar to the prior decade, at 1.2% per year. GDP growth was also quite similar, at 2.2% per year. There were much larger changes in both energy intensity and emissions intensity; energy intensity fell by 1.7% per year, while emissions intensity shifted from increasing to declining by 0.5% per year. Combined, these resulted in an increase in CO2 emissions of only 1.1% per year during this period.
This decline in CO2 growth from 2.8% to 1.1% between 2001-2010 and 2011-2019 is primarily due to two factors: less energy use per unit of GDP as economic growth is increasingly driven by lower material-intensity sectors like information technology and services, and lower emissions per unit of energy as cheap clean energy provides an increasing share of global energy use. Both factors are roughly equally responsible for declines in the rate of emissions growth, with energy intensity accounting for 0.7% and emissions intensity 0.8% of the total 1.8% reduction in the rate of emissions growth between the two periods.
Different Drivers in Different Groups
A global Kaya decomposition provides a necessarily incomplete picture. For example, while it attributes a sizable portion of emissions growth to population increases, it neglects the fact that much of the growth of the global population is occurring in countries where per-capita emissions are quite low.
We can get a more complete picture of changing drivers of CO2 emissions by examining changes in major emitting countries in regions. To do this, we divide the world into the US (representing 15% of global emissions in 2019), the EU (8%), China (28%), India (7%), Japan (3%), Russia (5%), and the rest of the world (35%), and perform a separate Kaya decomposition on each.
Figure 3, below, shows the results of our analysis for each region across the 2001-2010 and 2011-2019 periods. Each region has a bar for each period, while the black dots represent the average rate of CO2 growth (or decline) during that period. Note that there may be some slight differences between the sum of the individual factors and the overall rate of CO2 growth due to interaction effects not included in the figure.
The US saw modestly faster emissions reductions in the 2011-2019 period (-0.8% per year) compared to the prior decade (-0.5% per year). Population growth slowed in the 2011-2019 period compared to the prior decade, while economic growth accelerated. Reductions in energy intensity stayed the same between the two periods, while emissions intensity of energy fell as the US replaced its coal generation with renewables and natural gas.
In the EU emissions reductions accelerated in the 2011-2019 period (-1.8% per year) compared to the prior decade (-0.4% per year). Population growth slightly fell and economic growth slightly increased in the 2011-2019 period. Energy intensity fell at a notably faster rate, while emissions intensity also fell faster in the recent period.
China saw particularly dramatic changes between the two periods. Their emissions grew at a rate of 9.8% per year during 2001-2010 but only grew at 2.1% per year during 2011-2019. This was due to a combination of lower economic growth, significant reductions in energy intensity, and reduced emissions intensity as they switched from carbonizing to decarbonizing their energy. Population growth in China slowed slightly in the recent period, but this had a relatively minor impact on emissions.
India’s emissions growth remained relatively unchanged between the two periods, increasing by 5.6% per year during 2001-2010 and 5.1% per year during 2011-2019. Economic growth was similar between the two periods, though population growth modestly declined in the recent period. Energy intensity declined faster in the recent period, but their emissions intensity actually increased more rapidly.
Japanese emissions fell notably faster in the 2011-2019 period (-1% per year) compared to the 2001-2010 period (-0.4% per year). This was largely due to a reduction in energy intensity and was despite a modest increase in the rate of economic growth. Emissions intensity went from being unchanged during the 2001-2010 period to actually increasing at around 0.4% per year during the 2011-2019 period. This was largely due to large increases in emissions intensity following the Fukushima-Daiichi nuclear reactor accident and the subsequent shutdown of much of Japan’s nuclear generation in 2011 and 2012, though it has been somewhat balanced by more rapid reductions in emissions intensity as some reactors have come back online and other clean energy sources are deployed.
In Russia, emissions grew by 1% per year in the 2001-2010 period, which fell to 0.5% per year during the 2011-2019 period. This was primarily due to slower economic growth, though it was counterbalanced by slower reductions in energy intensity. Russia also saw slightly higher population growth and modestly lower emissions intensity in the recent period.
The rest of the world saw very similar rates of population growth and reductions in energy intensity between the two periods. However, notably lower population growth and modestly lower emissions intensity of energy resulting in the rate of emissions growth declining from 2.7% in 2001-2010 to 1.6% in 2011-2019. That said, this region represents an aggregation of many different countries, each of which may have quite different drivers of emissions growth (or declines) across all of these different factors.
Figure 4, below, shows the drivers of the differences in growth between the 2001-2010 and 2011-2019 periods.
All regions experienced a slower growth — or faster decline — in emissions during the latter period compared to the former. China saw the largest absolute declines, with the growth rate of emissions falling by around 7.8% between the periods. This was driven by lower economic growth (3.1%), lower energy intensity of GDP (2.6%), and lower emissions intensity of energy (1.7%).
The US saw the growth of emissions (or in its case, the negative growth of emissions) decline by 0.3%. This was driven primarily by lower emissions intensity of energy (0.7%) and population growth (0.3%) and was offset by higher economic growth.
The EU saw the second-largest reduction in emissions growth (or, in its case, more rapid emissions decline) between the periods, with emissions falling by an additional 1.4% per year. This was primarily driven by falling energy intensity (0.9%) and lower emissions intensity of energy (0.5%).
India saw a 0.5% reduction in the rate of emissions growth, driven by a combination of lower population growth (0.4%) and lower energy intensity (0.4%), and offset by higher emissions intensity of energy and slightly higher GDP growth.
Japan saw a 0.6% faster decline in emissions, driven primarily by lower energy intensity (1.2%) and population (0.3%) and offset by higher economic growth and emissions intensity (with the latter concentrated in the 2011-2013 period after nuclear shutdowns).
Russia saw a 0.5% reduction in the rate of emissions growth, driven by large declines in economic growth and modest reductions in emissions intensity, but offset by a large reduction in the rate of energy intensity improvements.
The rest of the world saw a 1.1% decline in emissions growth, driven (at an aggregate level) by lower economic growth (0.8%) and lower emissions intensity (0.3%).
While all major emitters and regions examined saw lower emissions growth or faster emissions declines over the past decade, the specific drivers varied between countries. The figures below show the annual drivers of emissions growth and declines in major countries in regions between 2000 and 2019.
US emissions show a lot of year-to-year variability in the underlying drivers. Recessions (e.g. 2008/2009) and recoveries (2010) notably stand out, but even during other periods there is a lot of variability in energy use per GDP; it positively contributes to emissions in 2010, 2013, and 2018, for example. The US also transitioned from increasing the carbon intensity of its energy mix in the early 2000s to strongly decreasing it in the period after 2006 or so. This corresponds to the rise of shale gas production — and its displacement of coal generation — though a rapid expansion of wind and solar generation also plays a large role in more recent years.
The EU also saw falling emissions during much of the 2000-2019 period, with a notable fall in the energy intensity of GDP in more recent years. Emissions intensity has also declined from 2004 onwards.
The growth of Chinese CO2 emissions over the past two decades was primarily driven by rapid economic growth. More recent slowdowns are due to a combination of slower economic growth, faster declines in the energy intensity of GDP, and a shift from increasing carbon intensity of emissions in the 2000-2011 period to declining carbon intensity from 2012 onward. This is primarily due to the rapid deployment of solar, wind, and nuclear generation in China in recent years.
The growth of Indian emissions has declined modestly in recent years — though there is an interesting annual oscillation in emissions growth rate during the 2012-2019 period. Lower population growth has served as a driver of a portion of India’s emissions declines, along with lower economic growth in the past few years and some declines in the emissions intensity of energy in 2017-2019 (following increases in emissions intensity from 2009-2016).
Japanese emissions declines have accelerated in recent years, though they were set back by a large increase in the emissions intensity of energy following the Fukushima Daiichi nuclear disaster. Japan also saw large declines in the energy intensity of GDP during the 2009-2016 period, though it has declined more slowly over the past few years.
The world as a whole has slowed the growth of emissions over the past decade compared to the prior one. This was primarily driven by declines in both the energy intensity of GDP — as economic growth is increasingly driven by the service sector and information technology rather than traditional manufacturing — and falling CO2 intensity of energy due to the replacement of coal by natural gas and renewable energy. However, this global picture obscures the role of very different factors across different regions, including slower economic growth in China and Russia.
Some of these factors — such as falling population, declining energy use per GDP, and emissions per energy use — are likely to accelerate due in the future as clean energy technologies become more cost-effective and countries become wealthier. Economic growth is more of a wildcard, but even here there are likely diminishing returns to growth as countries become wealthier. The confluence of these factors suggests that global emissions will likely plateau or even slightly decline in the coming decade; at the same time, current trends continuing will fall far short of the rapid emissions reductions needed for the world to meet Paris Agreement goals.