A reader writes:
I think the real point is that this is the debate we should have been having years ago. The right wing attack on science, especially of global warming, has meant that in fact the real necessary arguments about cost, level of mitigation, etc are not being played out. I happen to disagree with Manzi, and can site some pretty compelling arguments, but the point is that's an argument we should have. An honest one not about the existence of a fact, but the proper way to manage it going forward. Fundamentally, there is no cost to a polluter for polluting, so there are no incentives to prevent pollution. Mitigation by counting on declining supply of energy to drive pricing still ignores this basic fact.
Carbon taxes and cap-and-trade models attempt to put a value to society of this pollution and charge that back to the polluter (we can also talk about the disproportionate effect on the poor of such models.) Discounting models such as Manzi use are acceptable for middle of the road projections of results and costs, but since both are actually only known as ranges I would suggest Black-Scholes is a better approach. However, I argue that discounted cash flow models are not sufficient, because there are finite probabilities in global warming of near infinite cost (human extermination or mass death) where standard discounting models break down (divide infinity by zero). See this this for a bit dated but still useful article.