There are difficulties, though. One is that “avoided deforestation” is hard to define and quantify. Another, raised by officials in Europe who have chosen not to include REDD in the European carbon-trading scheme, is that the carbon market would be flooded with deforestation credits that will push down the price. Companies would then buy cheap credits and continue doing business as usual rather than cutting their own emissions. Further tricky issues abound: who should have the right to sell credits? How should the money be split between central governments, local governments and indigenous people? And should the money be paid in perpetuity?

REDD schemes will require careful monitoring to ensure that forests really are left intact and that carbon credits for an area are not claimed more than once. Murky goings-on in Papua New Guinea, one of the leading advocates of REDD, highlight such worries.

Even so, it is worth trying, simply because avoiding deforestation is so effective in slowing carbon emissions. So REDD deserves a place in the world climate treaty to be negotiated in Copenhagen in December, to replace the Kyoto treaty when it expires in 2012. As with other forms of carbon credit, today’s voluntary and experimental REDD schemes will need to be replaced by more rigorously accredited and monitored schemes. But they have a chance of working only if the countries in which they operate define forest land rights clearly. Brazil’s flawed attempt to do this is a step forward.

© The Economist Newspaper Limited 2009