From a recent study by the US National Center for Atmospheric Research:

Trade Imbalance Shifts U.S. Carbon Emissions to China, Boosts Global Total

The growth of Chinese imports in the U.S. economy boosted the total emissions of carbon dioxide (a primary greenhouse gas) from the two countries by over 700 million metric tons between 1997 and 2003, according to a study published online in the journal Energy Policy. The analysis, prepared by two scientists at the National Center for Atmospheric Research, suggests that American emissions of carbon dioxide in 2003 would have been 6% higher if the United States had manufactured the products that it imported from China. Meanwhile, China's 2003 emissions would have been 14% lower had it not produced goods for the United States.

"These results show the importance of world trade in accounting for the emissions that drive climate change," says Shui Bin, an environmental policy analyst who authored the Energy Policy paper with geochemist Robert Harriss. Their research was supported by the National Science Foundation, NCAR's primary sponsor.

The world's two biggest emitters of greenhouse gases are the United States (about 25% of the global total) and China (about 15%). The Shui and Harriss study implies that the United States is indirectly responsible for even more carbon dioxide than widely perceived.

So: the US is an even more profligate emitter of CO2 than we thought, and is responsible for more emissions than if it had produced all those consumer goods domestically. (And we should note, by the way, that this is just carbon emissions - the marginal production of "classical" pollutants such as sulfur and nitrogen compounds must be disproportionately greater, insofar as we can assume that the lower efficiency is associated with a lower standard of antipollution technology). But what are the implications for other Western nations?

The BBC has also picked up on this study, and explores the implications for Europe's Kyoto signatories in an article. In between citing the facts of the NCAR study, the correspondent describes the issue of "carbon leakage" - the possibility that CO2 emissions that are attributable to Kyoto-signatory consumer states get offshored to China, a non-signatory, distorting the measure of the former's compliance with treaty targets.

This proposition is not undisputed:

Michael Grubb from Imperial College London believes rates of "carbon leakage" are likely to be small.

"The idea that there are leakage effects has been comprehensively discredited," he told the BBC News website.

The Carbon Trust, a UK government-backed company of which Professor Grubb is policy director, has researched the likely impact of emissions trading on European business.

"What people do talk about is 10-15% leakage - that's the Intergovernmental Panel on Climate Change (IPCC) estimate - but those are model-based estimates, and in reality, you don't even get that," he said

"As a reality check, Europe has had more expensive energy than just about everywhere for decades, yet we still produce virtually all our own steel and our own cement."

Professor Grubb also points up the difficulties involved in measuring and calculating emissions according to end user rather than producer.

Now I know that there are people here who can translate "10-15% leakage" into millions of barrels of oil and generally deconstruct this statement better than I can. I would merely note that Grubb's denial of IPCC estimates is unsubstantiated, and that the remark about steel and cement seems to miss the point of trade with China. That last remark sounds to my jaded lay understanding like the most plausible statement of all.

Nevertheless, [NCAR's Dr] Shui Bin is adamant that the global community should have a try; the way carbon emissions are measured currently is flawed, she believes, and could penalise developing countries unfairly.

"The equity issue should be addressed in the Kyoto Protocol," she said, "but the current Protocol is based on a flawed accounting system.

"A country can intentionally or unintentionally decrease its CO2 emissions by reducing its domestic production but increasing its trade.

"Total CO2 emissions would increase, but the current carbon accounting framework will show a decrease."