Landsburg is certainly right that Romm can't provide any precise answer to the economist's question of how much we "should" collectively spend on the environment, in order to achieve a kind of optimal outcome. Unfortunately, I don't think Landsburg can provide an answer either. It seems to me that his "six hard questions" are in fact so difficult as to be impossible to answer meaningfully.

* The first two empirical questions about what would happen in the event of climate change seem to require too many assumptions or conclusions about both physical science and human behavior. How much will the chances of a big hurricane in New York go up? How much damage will it do? How much will be spent on precautionary meaures? How much will growth be reduced? How many people will move to Chicago instead?

* The third question, about discounting the welfare of future generations, seems very difficult to assign a number to. I have no idea whether I would discount the welfare of tomorrow's babies by zero, 10%, 50%, or what. The best case is that economists have devised some clever way to figure this out and the preferred number turns out to be similar for everybody. Even if so, good luck convincing policymakers.

* Empirical questions four and five, about whether future generations are likely to be rich or dead or what, are if anything even harder to devise predictions for than one and two.

* Question six, about risk aversion, might actually be reasonably answerable; I don't really know. But one out of six is pretty useless for Landsburg's purpose.

So my conclusion is that a thorough economic analysis of the issue, of the sort that Landsburg is arguing for, is necessarily going to be so hand-wavy that it's practically useless. It's not going to help people agree on an amount to spend because they'll (rightly!) never be able to agree on the proper assumptions to use.

I would prefer not to pretend that we could get the One True Expenditure if only we had the right numbers to plug into the standard economic equations, because I'm pretty sure it's not possible to narrow the range on all those numbers sufficiently. Much as I like economic analysis, you can't really do it without good data.

Instead, I suspect that Romm's or Gore's assertions that the risk is "relatively high" and the price to reduce it is "relatively low", based on converging evidence from many sources, has about the same level of precision as anything Landsburg's numerical approach can realistically achieve.