How to assess the costs of global warming?

Jim Manzi has a fairly scathing take on this global warming piece from the Washington Post:

What’s so funny is that Eilperin never seems to be willing do the work to pick up the trail of breadcrumbs that all her interviewees leave behind them. She writes that “Most scientists warn that a temperature rise of 2 degrees Celsius (3.6 degrees Fahrenheit) could have serious consequences.” Really – how serious? Well, according to the UN IPCC a 4C increase – twice this amount – would reduce global economic output by 1% – 5%. Oh yeah, that’s in the world of the 22nd century which is expected to have per capita consumption of something like $40,000 per year versus our current consumption of about $6,600 per year. So we are condemning future generations to be only 5.7 times richer than us, rather than 6 times richer. She quotes a scientist’s “tremendous” finding that under a business-as-usual scenario Earth will warm by 7.2 degrees Fahrenheit by 2100, without mentioning that this is 4C, or well within the forecast range of the current business-as-usual projections for warming by 2100 of the most recent UN IPCC report. Also note that this is the amount of warming that is projected to cost a much richer world about 3% of its consumption.

Most reporting on global warming is, to say the least, craptacular. The science journalists don't understand the economics, and so they unquestioningly accept projections from advocacy groups who want drastic action on emissions; most economics journalists don't understand the science, so they are equally hostage to bad information from a different set of advocates. Indeed, this is also true of the economists and scientists. Doing a cost-benefit analysis is very hard to do when you don't have a commensurate measure of costs and benefits.

So I understand where Manzi is coming from, but I think his position is way too strong. It's not adequate to simply project GDP forward, and compare the costs of global warming to the cost of abatement. There are serious distributional justice issues here: it is particularly reprehensible for rich countries to inflict this kind of negative externality on poor ones in order to consume more. Bjorn Lomborg's argument that we should use the savings to do poverty abatement programs has some merit, but there are multiple problems with this approach. For one, we aren't very good at poverty abatement in the third world; for another, there's currently no political movement saying "let's ignore global warming and spend $50 billion on the third world". Moreover, carbon has a very long shelf life--ca. 100 years--and there's no way to make a credible committment to have our descendants transfer money to future poor people in exchange for letting us emit carbon now.

The philosophic and economic problems of intergenerational discounting and international distributive justice are gigantic, and so far no one has a good handle on them. For all that I disagreed strongly with his method, Sir Nicholas Stern had the right idea: you need to figure out a way to fairly weigh the effects on various (cringe) stakeholders, not merely calculate the monetary costs of various potential policies.