Blog Name

The Challenge of Climate Change (in One Chart)

Arvind Subramanian (PIIE) and Aaditya Mattoo (World Bank)



Is the world making progress on climate change? Recently, the OECD struck a hopeful note, reporting that emissions were growing more slowly than GDP in both the high-income and developing countries, including China. This decoupling of emissions and growth, if true, would be good news indeed, since it would suggest that the world can cut emissions without hurting the economic growth needed to lift hundreds of millions of people out of poverty.

Alas, what matters is global emissions. And while it's true that global emissions grew slower than GDP between 1960 and 2000, since then decoupling has stalled and global emissions have grown as fast—or even faster—than global GDP.  The graph shows that the emissions/GDP ratio declined in high income and developing countries until the late 1990s but since then the global emissions/GDP ratio has been climbing. How can this be?

Global emissions as ratio of global GDP (kilograms per dollar of GDP, measured in US dollars at 2000 prices)

Global Emissions as ratio of Global GDP
Source: World Bank's World Development Indicators, and IEA's Emissions data (2012) . Data for 1960–2009 are from the World Bank and data for 2010 are extrapolated using the International Energy Agency's most recent data for emissions.

The answer is that fast-growing countries such as China and India account for an increasing share of global GDP, and these fast-growing countries have higher emissions/GDP ratios than the slower growing high income counties.  So even though these countries' emissions/GDP ratio is falling, as they account for a larger share of the global economy, the global emissions/GDP ratio is rising.

Focusing on global emissions is the right approach, since that's all that matters for the atmosphere. It's also the right approach for a proper understanding what drives emissions. Dieter Helm, in his terrific new book, Carbon Crunch: How We're Getting Climate Change Wrong and How to Fix It, argues that emissions measures should focus on consumption rather than production.  If a steel plant in the United Kingdom closes, the UK production emissions will decline.  But if UK steel consumption is unchanged, and steel previously provided by the shuttered plant is instead imported from China, UK consumption emissions will increase if the Chinese steel is produced in a more carbon-intensive manner than the UK production it replaced.  Result: Although the United Kingdom gets to brag about falling emissions, global emissions are higher.  

Similarly focusing on global GDP is important to recognize the role played by energy demand in driving global emissions. As global GDP rises, demand for energy will rise in high-income countries less than proportionately. But there will be much greater increases in energy use from developing countries. These increases from very low current levels are not only unavoidable but arguably a moral imperative, since increased energy use will be crucial to lifting vast numbers of people out of extreme poverty.

How should this play out? One can ask what needs to happen by 2050 for the world to have a fighting chance of limiting temperature increases to below 2 degrees centigrade. Put more precisely, how much does the global emissions/GDP ratio need to fall to avert runaway climate change?

Let's assume global GDP-per-capita growth of about 3 percent, which is less than seen in recent years and assumes some under-performance by low-income countries; and annual population growth of about 0.8 percent. These assumptions, combined with the goal of limiting emissions in 2050 to about 15 gigatons of CO2 emissions, which climate scientists say is near the upper end of what might be safe, imply that the emissions/GDP ratio must fall from the current level of 0.8 to about 0.08—that is, a ten-fold decline.

In the graph above, the barely visible bar at 2050 is the amount of emissions that the atmosphere just might be able to tolerate without rapidly overheating. For all practical purposes, we need to be in a zero-emissions world by 2050 to avert catastrophe. Note that the drop between now and 2050 will have to be dramatically steeper than anything we have witnessed in history (bending the health cost curve in the United States seems, in comparison, a piece of cake). The basic assumptions can be altered but they will not change the basic fact of how much the emissions-GDP curve needs to be bent.1

If, on the other hand, the emission-GDP curve heads anywhere close to the current trajectory, we are looking at emission levels of about 90-100 gigatons (2050 GDP is about $185 trillion, at 2000 prices), orders of magnitude larger than the target for 2050 of about 15 gigatons.

Discouraged? Clearly we need an entirely different approach to climate change. That is why in Greenprint: A New Approach to Cooperation on Climate Change, our new book, we argue that nothing short of radical technological change can help reconcile the need for energy to lift hundreds of millions of people out of extreme poverty while still avoiding frying the planet.  If you will be in Washington on Friday, March 15, I hope you will join us for presentation and discussion of the book's key recommendations.

Also posted on the Views from the Center blog at the Center for Global Development.


1. Expressing the emission-GDP relationship in terms of GDP measured at purchasing power parity does not alter the conclusions. There will be some natural decline in the demand for energy and in the carbon-intensity of world GDP as the composition of world output shifts from manufacturing to services.

More From