It sounds like a rare piece of good news about climate change: emissions of carbon dioxide, the principal cause of global warming, grew at a slower rate after 2000 in the United States, and have actually dropped since 2007. In Europe the story sounds even better, as overall emissions dropped from 1990 to 2008, often roughly matching, or in some cases exceeding, the reductions promised under the Kyoto Protocol.
Yet the apparent progress on emission reductions in rich countries has occurred at a time of widespread outsourcing of manufacturing to China and other developing countries. In the process, we have effectively outsourced our carbon emissions as well. If consumers are responsible for the emissions from making the consumer goods they buy, then we have not solved the problem. We have just made it harder to see - and much harder to measure.
Here's the problem: if a Chinese steel mill sells steel to Toyota in Japan, which uses it to make cars sold to Americans, which country is responsible for the steel mill's emissions? America seems like the logical answer; the Chinese emissions happened in order to make something that was bought and used in the United States. Those emissions, however, belong to China in all standard statistics, and in most discussion of climate targets and responsibilities for emission reduction.
The standard approach, counting emissions based on where they are produced, is much easier to implement. National statistics naturally count factories, vehicles, and other emission sources that occur within a country's borders. The more logical approach, attributing emissions from production to the country where the final product is consumed, requires detailed modeling of international trade flows.
Researchers have attempted to allocate emissions on a consumption basis in recent years. A leading figure in this new field of study is Glen Peters, at Norway's Center for International Climate and Environmental Research-Oslo (CICERO). Peters and several coauthors have estimated that emissions from the production of exports - often referred to as emissions "embodied" in exports - have risen from 20 percent of global carbon dioxide emissions in 1990 to 26 percent in 2008. When they recalculated emissions on a consumption basis, emissions in America, Europe, and other developed countries continued to rise throughout this period, and the rich world missed its Kyoto targets by a wide margin.
Studies of the emissions embodied in trade have been done for many individual countries, finding that almost all high-income countries are net importers of embodied carbon. Australia, and to a lesser extent Canada, are the exceptions, with substantial exports of energy-intensive metals, fuels, and other materials. On the other hand, many developing countries are net exporters of embodied carbon. According to the Peters group, China accounted for half the world's carbon embodied in exports in 2008, but this is in part because of its sheer size. On a per capita basis, South Africa, Kazakhstan, Australia, and Russia all had at least double China's export emissions.
Emissions in lower-income regions are often for the benefit of higher-income consumers elsewhere, within countries as well as globally. Looking at provincial trade within China, a recent study found that more than half of all carbon dioxide emissions are for goods consumed outside the province where they are produced. Up to 80 percent of emissions for goods consumed in the higher-income coastal provinces occur in the less developed interior provinces.
All such estimates are based on elaborate modeling of trade and emissions, and there is unfortunately no consensus on exactly how to do the calculations. A survey of 50 different studies found vast differences in estimates of the emissions embodied in trade. Credible estimates for the export-related share of China's total emissions in 2005, for instance, range from 18 to 45 percent.
Uncertainty about the numbers makes it hard to use consumption-based emissions accounting for policy purposes. But regardless of the exact amounts, the implication is clear: the United States, Europe, and Japan are responsible for much more of global emissions than what occurs within their own borders. The goal of climate policy must be not only to clean up at home, but also to help create low-carbon energy systems in the countries where, increasingly, the world's manufacturing is located.
Staring at these numbers leads to one more surprise. Although China is by far the world leader in exports of embodied carbon, it does not specialize in exports of carbon-intensive products. In a study of carbon in China's trade, I found that China's leading exports are manufactured goods of relatively low energy intensity. China's comparative advantage in trade is based on labor costs and manufacturing experience, not on energy or raw materials. The country's huge export sector involves a lot of carbon emissions due to its reliance on coal-based electricity, as well as inefficiencies in some industries.
This puzzling result is genuinely good news for the prospects for climate negotiations. Although it will take time and money to convert China to low-carbon energy, doing so will not threaten the country's pattern of economic growth. Recent Chinese government policy statements, endorsing rapid moves to reduce coal use and carbon emissions, are compatible with the ongoing export-driven growth model.
Indeed, it may be more difficult to win support for climate mitigation from other countries with high carbon exports. South Africa, Russia, and the oil-exporting countries currently base their economies on the most carbon-intensive products, and will require more fundamental restructuring to exist in a low-carbon world. China, like the developed countries, will find it comparatively easier to make the transition to sustainable energy use.