Not Fit for Purpose: Why Climate Classifications of Developed and Developing Countries Have to Change

Say something often enough and it becomes accepted as fact. The idea that developed countries are responsible for climate change has become a common refrain in the climate debate, asserted and repeated by campaign groups and media outlets alike, for example here, here, and here. No wonder that most respondents to an earlier poll thought as much. And no wonder that climate negotiations are beset by rows over climate finance, and who is responsible, with the latest talks in Bonn largely ending in failure.

But it isn’t true.

In a new CGD working paper, published today, we find that the 23 developed countries that are obliged to provide climate finance (under the 1992 UN Framework Convention on Climate Change) have, in fact, produced under 40 percent of all greenhouse gas emissions since 1850.

Our paper presents a comparative analysis of historical emissions from numerous datasets using different measures of emissions, different classifications of developed and developing countries, and covering different time periods. To our knowledge this has not been done before.

We also find that nine major emerging economies—currently classified as “developing” under the UN system—have produced at least 30 percent. While over 100 low-income, lower-middle-income, least-developed countries and small island developing states collectively account for just 15 percent of all emissions. Adjusting for colonial, consumption and per capita emissions makes surprisingly little difference. It is nonsensical therefore not to distinguish major emerging economies from poorer, low-emitting countries—but this is currently what happens under the UN climate system.

Here, we lay out our findings and argue that old distinctions between developed and developing countries are no longer fit for purpose.

What do the data on historical emissions actually say?

Different data sources can vary by over 30 percent in their estimates of global emissions, but estimates of developed and developing country shares of those emissions are much more consistent. We find that developed countries (using the “Annex I” definition of 42 countries agreed by the United Nations Framework Convention on Climate Change (UNFCCC) in 1992 that underpin the climate negotiations) are responsible for about half (49-51 percent) of all GHG emissions going right back to 1850. Figures for the subset of 23 Annex II developed countries—those that have a formal obligation to provide climate finance under the UNFCCC—are just 36-39 percent. Alternative estimates for contributions to rises in global mean surface temperature (GMST), which take account of the short-lived nature of some GHGs such as methane and are arguably a better reflection of contributions to global warming,are a little lower at 46 and 34 percent respectively.

Shares are higher if analysis is restricted to CO₂, and higher again if emissions from land use, land-use change and forestry (LULUCF, around which there is greater uncertainty) are excluded, with Annex I and II shares of CO₂ excluding LULUCF being 62 and 49 percent respectively. But even if we believed this narrower measure was a better reflection of responsibility (and we don’t), these figures are still hardly consistent with a claim that climate change is all the fault of developed countries.

Moreover, developed country shares are typically around 10 percentage points lower if analysis is limited to the period since 1990, when the first Intergovernmental Panel on Climate Change (IPCC) report was published and for which more datasets are available. Specifically, Annex I and II shares of all GHGs including LULUCF are 38-39 percent and 29-32 percent respectively, with shares of contributions to rises in GMST of 33 and 25 percent respectively. These shares are lower because around 40-45 percent of all GHG emissions have been produced since 1990, with emissions falling in most developed countries while rising rapidly in many emerging economies, reflecting the extraordinary economic progress that many have made.

The paper also presents details for other groups of countries, defined for example by income, region, or institutional affiliation, and for significant individual countries. A key point to emerge is that the Annex categories mask a lot of variation within each group, particularly amongst developing countries. While developing (non-Annex I) countries account for about half of all GHG emissions since 1850, much of this is from the nine developing country members of the G20—Argentina, Brazil, China, India, Indonesia, Mexico, Saudi Arabia, South Africa, and South Korea—who collectively account for 30 percent. Low-income countries (LICs) account for just 2-3 percent, while the 100+ countries that are either LICs, lower middle income countries (LMICs), least developed countries (LDCs) or small island developing states (SIDS) collectively account for just 14-16 percent (or around 10% when excluding India).

The interactive chart below summarizes these figures for all four measures of emissions and both Annex categories of developed countries, and allows users to add custom groups of countries, vary the start and end dates, and apply adjustments based on colonial rule.

What about colonial, consumption, and per capita emissions?

We find that taking account of colonial emissions (using Carbon Brief data on the patterns of colonial occupation) makes a significant difference to some individual countries, but not that much difference to the overall picture. Even if all emissions from colonies are attributed to the respective colonial power during the period of occupation (arguably an equally unreasonable position), Annex I and II shares of all GHG emissions since 1850 rise by about 5 percentage points to 55-56 percent and 41-44 percent respectively, and by at most 1 percent for the measure focusing on CO₂ excluding LULUCF (to 63 and 49 percent respectively).

A focus on consumption rather than territorial emissions (which some have argued give a fairer picture of responsibility, though this misses the wider economic benefits derived from their production and is not how the UNFCCC measures emissions) is constrained by data which are generally limited to the period since 1990 and only cover CO₂. However, analysis suggests that doing so would only add a few percentage points to developed country shares, and possibly not much more even going back to 1850.

It is the consideration of per capita emissions that is perhaps the most interesting. There is no doubt that per capita emissions have historically been much higher in developed countries overall, and while Annex I countries may account for “only” 49–51 percent of cumulative GHG emissions since 1850, that is still significantly larger than their 17 percent share of world population in 2024. But what matters for the atmosphere and the climate is cumulative aggregate emissions (as Professor Friedlingstein, chair in mathematical modelling of climate systems at the University of Exeter and lead author of the Global Carbon Budget, has said). Of course, historical emissions are not the only relevant variable for assessing climate finance contributions (we discuss this in our earlier work on “fair shares” here and here). And per capita emissions are much more relevant for debates about emissions reduction pathways and fair shares of the remaining carbon budget. But as a metric for assessing country-level historical responsibilities, it is deeply flawed.

Moreover, if we were to use this measure, and we elaborate further in the paper on why we do not think this is a good idea, developed country shares of cumulative per capita emissions would typically be 10-20 percentage points lower than for cumulative aggregate emissions over the period 1850-2024, and typically 5-15 percentage points lower over the period 1990-2024, using three different calculation methods which essentially apply different weights to each country’s per capita emissions in each year (unweighted, all weight on current population, or weighted by population in each year).

All three approaches estimate cumulative per capita emissions for each country, and sum these into totals for the Annex I and II groups (and globally) to calculate group shares in exactly the same way as done for aggregate emissions. We would get much higher results for the Annex I and II groups if we calculated cumulative per capita emissions at the group level (e.g. by dividing the sum of all Annex I emissions by the sum of Annex I populations). However, because each total is then effectively a weighted average it doesn’t lend itself to calculating shares—we get some bizarre results, including that the share of some individual Annex II countries (such as the US) is greater than the share of all Annex II countries together. This illustrates the difficulties of using and interpreting data on cumulative per capita emissions, and reinforces the case for focusing on shares of aggregate emissions.

Others have estimated shares of emissions that exceed “fair shares” based on equal per capita rights to a safe global emissions budget, which yield Annex I figures of 90 percent for the period 1850-2015. This raises important ethical questions that need to be considered, but doesn’t address Friedlingstein’s point flagged above, and results are sometimes misrepresented as shares of aggregate emissions that have only added to the confusion.

What does this all mean for climate negotiations?

Rows over responsibility and finance have turned COP climate negotiations toxic, with battle lines drawn between developed and developing countries. This is not surprising. Developing countries are rightly angry at what they see as decades of broken promises, and an international financial system that is biased against them. Pressure from developed countries to force developing countries to share the climate finance burden are understandably interpreted as attempts to shirk their responsibilities. The negotiation of the NCQG and the EU’s introduction of the CBAM (widely perceived as a tool of protectionism rather than an attempt to incentivize the green transition, not helped by the EU’s refusal to provide time-bound exemptions to poorer countries or to allocate a share of the revenues to support the green transition in developing countries) have contributed to a breakdown in trust that developed countries must seek to restore. But recent cuts to GEF replenishments and GCF commitments amidst declining ODA budgets are only exacerbating the problem. This matters because fights over finance are negatively impacting all countries’ willingness and ability to make bolder emissions reduction commitments.

Yet these divisions appear to be fueled by a misperception of what the data on historical emissions actually say, an adherence to the Annex definitions of developed and developing countries that appear increasingly inaccurate and out of place, and ambiguity between these strict Annex definitions and more colloquial understandings of “developed” and “developing”.

As argued previously, none of this analysis is intended to absolve developed countries of their responsibilities. Their actions to date in both reducing emissions and providing climate finance have been woefully inadequate. But we do worry that the battle lines are based on a false premise—that Annex I developed countries are responsible for GHG emissions and climate change—that is leading to highly polarised positions that are poisoning climate negotiations and damaging efforts to prevent climate crisis.

Some have suggested that attempts to reclassify the Annex categories will cause the entire UNFCCC and Paris Agreement to unravel, but the Paris Agreement avoids reference to the Annexes which were never intended to be static. The International Court of Justice, in its landmark advisory opinion on the “Obligations of States in respect of Climate Change”, seems to agree. As indeed does the Baku to Belem roadmap, which suggests that climate finance is the responsibility of all countries. Moreover, many developing countries are already stepping up to the plate. But given what the data say about historical emissions, it’s clear the binary, Annex-based definitions of developed and developing countries are no longer serving the principles of equity and “common but differentiated responsibilities and respective capabilities” that underpin the UNFCCC. It is those countries least responsible for and most vulnerable to climate change that are most disadvantaged by the current state of affairs. A new, more nuanced approach is required that removes these divisions, acknowledges a broader shared responsibility, and takes the poison out of UNFCCC negotiations.