Carbon markets set up under climate-change agreements are supposed to reduce emissions of carbon dioxide. Credits are issued that correspond in some way to the desired carbon emissions (details vary). Companies that produce a lot of greenhouse emissions can then purchase credits from companies that produce fewer. Supposedly, this will fund new clean companies and projects that lead to a decrease in carbon emissions.

Here’s the problem. Some of those new companies and projects would have been undertaken anyway, without the credits. In that case, the credits won’t actually lead to less emission. An article in today’s Wall Street Journal describes one such case.

For a carbon-credit system to work, it seems that you’ve got to have a competent regulatory body that can give credits only to companies and projects that need the credits to succeed. But this requires a detailed understanding of industries and economies all over the world, as well as new technologies. According to the Journal article, it’s taken years for a UN regulatory body to figure out that it was issuing credits to projects that didn’t need them. Overall, this sounds like an inefficient system.

A carbon dioxide tax, which assigns a cost directly to the thing that’s supposed to be regulated, would work better. Make carbon dioxide emissions expensive, and then let the market work out the best way to deal with those costs.