In a recent article I explored how different income groups and world regions compared in terms of their share of the global population and versus carbon dioxide (CO2) emissions. The key take-home message was that there were large inequalities in global CO2 emissions: the world’s richest countries accounted for half of the global population but emitted 86 percent of annual CO2 emissions.
From this, two key questions from readers emerged:
- How does this comparison look at the national level [covered in a related blog post here]; and
- How does this look when we correct for emissions embedded in trade, so that we are comparing the emissions caused by a country’s consumption rather than production?
Here I aim to answer the second question.
The initial comparison of emissions by income group and region was based on ‘territorial’ emissions (those emitted within a country’s borders) — these are termed ‘production-based’ and are the metrics by which emissions are commonly reported. However, these emissions do not account for traded goods (for which CO2 was emitted for their production). If a country is a large importer of goods its production-based emissions would underestimate the emissions required to support its standard of living. Conversely, if a country is a large goods exporter, it includes emissions within its accounts which are ultimately exported for use or consumption elsewhere.
‘Consumption-based’ emissions correct for this by adjusting for trade. Consumption-based emissions are therefore: (production-based emissions – embedded CO2 in exported goods + embedded CO2 in imported goods). The Global Carbon Project (GCP) publishes estimates of these adjustments in their carbon budget.1 You can find much more information and data on emissions in trade in our full entry here.
How do consumption-based emissions change the emission shares by income group and region? In the table I compare each group’s share of the world population, production- and consumption-based CO2 emissions.
On a production basis we had previously found that the richest (high and upper-middle income) countries in the world accounted for half of the population but 86 percent of emissions.2 On a consumption basis we find the same result, but resulting from the fact that upper-middle income countries primarily export emissions to high income countries. High income countries’ collective emissions increase from 39 to 46 percent when adjusted for trade (with only 16 percent of the population); upper-middle income countries’ emissions decrease by the same amount (7 percentage points) from 48 to 41 percent. Overall, this balances out in the top half of the world population: upper-middle income countries are net exporters whilst high income net importers.
In the bottom half, it appears that very little changes for the collective of lower-middle and low income countries: their production and consumption emissions shares are effectively the same.
By region we see that traded emissions tend to flow from Asia to North America and Europe (Asia’s share reduces when adjusted for trade whilst North America and Europe’s share increases).
Note here that consumption-based emissions are not available for all countries. Collectively, countries without consumption-based estimates due to poor data availability account for approximately 3 percent of global emissions. Many of the missing countries are at low and lower-middle incomes. With the addition of these countries, we would expect small percentage point shifts across the distribution. The challenges in accounting for carbon embedded in global trade3 mean these estimates are not perfect; nonetheless they should provide a good approximation of the global transfers across the world.
On a consumption basis, high-income countries (Europe and North America in particular) account for an even larger share of global emissions (46 percent — nearly three times their population share of 16 percent).
| Income or regional group | Share of population (%) | Share of production-based CO₂ emissions (%) | Share of consumption-based CO₂ emissions (%) |
|---|---|---|---|
| High income | 16% | 39% | 46% |
| Upper-middle income | 35% | 48% | 41% |
| Lower-middle income | 40% | 13% | 13% |
| Low income | 9% | 0.4% | 0.4% |
| North America | 5% | 17% | 19% |
| Europe | 10% | 16% | 18% |
| Latin America & the Caribbean | 9% | 6% | 6% |
| Asia | 60% | 56% | 52% |
| Africa | 16% | 4% | 3% |
| Oceania | 0.5% | 1.3% | 1.3% |
The above analysis looks at income groups and regions on an aggregate level. What about at the country level? In the visualization I show individual countries’ share of consumption-based emissions (on the y-axis) versus its share of the global population (x-axis). Countries which lie above the grey line have consumption-based emissions higher than their ‘share’ of the global population based on population. Those countries below have a lower share than equitable distribution based on population.
In a separate post I discuss the distribution of emissions at the country level in more detail (and on a production basis). The key take-homes from this are similar to that of the production-based comparison: high-income countries consume more than their ‘share’; upper-middle income countries are mixed; and lower-middle and low-income tend to consume less than their ‘share’. On a consumption basis high-income countries’ emissions typically increase, meaning they move further from the grey line.
The biggest stand-out result is that some of the countries we assume to be very larger emissions exporters do not shift as much as we’d expect. China here is the notable example. Although China is a net exporter (around 13 percent of its emissions), even when adjusted for consumption it emits more than its population ‘share’. This challenges the common misconception that most of China’s increase in emissions is attributed to the production of goods sold elsewhere. Much of its growth has resulted from very fast development domestically [which have led to rapid improvements in living standards].
