Carbon Pricing Looks Expensive Only If You Ignore the Benefits

It seems the George C. Marshall Institute has come out with a study which shows that pricing carbon emissions would be costly to the economy. But one problem with the study, as Tony Kreindler points out, is that they forgot that run-amok climate change is costly too. This seems like an important point! If I added up the ruinous costs of auto ownership — thousands of dollars in up front costs, fuel costs, repairs, insurance, etc. — but forgot to mention that you get to drive your car around it would seem baffling that anyone buys one.

The same principle applies to carbon pricing. If there were no climate change, or if climate change wasn’t caused by carbon dioxide, or if rising global temperatures were totally harmless, then cap and trade would look like a terrible idea. It’s when you put the mild short-term decrease in the rate of GDP growth against the alarmingly catastrophic impact of doing nothing that cap and trade starts to look good.


I wouldn’t say you should think about this in a strict GDP cost-benefit framework. In part, this is because many of the harms of climate catastrophe would occur outside the United States and it’s just wrong, morally speaking, to ignore them. In part it’s for technical reasons related to the “broken windows” fallacy — destroying half the buildings in Miami would be a more disastrous loss of wealthy than a look at the growth impact would suggest. And in part it’s because the impact on world GDP doesn’t really properly capture the “cost” of whole nations being rendered uninhabitable. But even if you look at it in pretty narrow economic terms, the long-term costs of doing nothing are quite grave. It’s important to recognize that every year we do nothing, things get much worse. And the time frames on which these things are looked at are often inadequate. Kids born last year should be alive into the 2080s — barring ecological catastrophe, that is. Barack Obama’s grandchildren should be living well into the 22nd century.