Dan Simmons alerted me to this WSJ article by Harvard's Robert Stavins. Now I respect Stavins, because he blasted the absurd economic analysis that the California Air Resources Board put out on its statewide cap-and-trade plan. So even though he is a proponent of carbon legislation, I think he's intellectually honest.

However, in this recent WSJ piece, he's got two different arguments that seem wrong on the face of them. First let's look at the iffy one:

Critics argue that we can afford to wait because the world of tomorrow will be wealthier and better able to absorb the costs. But acting sooner, such as by adopting the emission caps proposed in the U.S. House legislation, will lower the ultimate costs of achieving the target, because there will be more time allowed for gradual transition—which is what keeps costs down.

OK on the face of it, this is a bit weird. You keep costs down by giving firms more time to comply. But, I think we all get what Stavins is saying. He's got in mind a scenario where firms don't have to do anything for the next twenty years, and they don't anticipate they will ever have caps imposed in the future, and then Glenn Beck's coastal house gets flooded and everyone realizes Al Gore was right. So then the government imposes draconian cutbacks in emissions, and everyone slaps his head and realizes the government should have made the hard choices sooner. OK fair enough, but like I said, you need to assume the part about firms being surprised by the caps, or else it doesn't work.

But now this next one I'm really not sure I follow:

As for how much [the Waxman-Markey cap-and-trade] will cost, the best economic analyses—including studies from the U.S. Congressional Budget Office and the U.S. Energy Information Administration—say such a policy in the U.S. would cost considerably less than 1% of gross domestic product per year in the long term, or up to $175 per household in 2020. (That's the cost of one postage stamp per household per day.)

If you read it quickly, it sounds like the cost of Waxman-Markey will be a postage stamp a day. But wait a minute, he said (no more than) 1% of GDP per year. Surely the economy (especially "in the long run") will crank out more than 100 postage stamps per household per day. (In 2007 median household income was over $50,000. Note that $50,000 > $17,500.) So Stavins is giving two different answers here, saying (no more than) 1% of GDP in the long run, OR in the short run, the cost will be up to a postage stamp per day. I don't think it's a coincidence that he didn't actually spell out a concrete number, so that the reader could realize just what "1% of GDP" meant.

Also, there's the little problem that the CBO report that came out last week said Waxman-Markey would cost the economy anywhere from 1.1% to 3.4% of GDP, by the year 2050. (See Table 1 on page 13 of this pdf.)So I'm not sure how that range translates to "considerably less than 1% of gross domestic product per year in the long run."