Fifteen years ago... no, thirty years ago... no, forty years ago we should have impose a carbon tax. Pollution externalities and the national security externalities of relying on and paying people in an unstable and undemocratic part of the world to be the key energy link on our economy both strongly militated for conservation and alternative energy--and a carbon tax was the best way to do that.
Twenty years ago with the recognition of human carbon-emission caused greenhouse gas global warming, a third decisive reason was added to the case for a carbon tax.
Yet here we are now, without one. And many of those who believe in a carbon tax argue that our absence of one now is not that big a deal--that ideally we would start with a low carbon tax now and gradually ramp it up over the next generation or two:
Paul Krugman tries to understand the issues:
http://www.princeton.edu/~pkrugman/toyclimate.pdf Nordhaus and other [ramp-up-gradually] modelers, making their best possible estimates, come to the conclusion that while emissions must eventually be brought way down and carbon concentration stabilized, it’s not worth doing this until K has risen a long way above current levels. So it’s a mega-St. Augustine: "O Lord, make us carbon-neutral, but not yet."... [W]... [have] a long way to go up the saddle path, and hence a low carbon price is appropriate now. This is the “climate policy ramp”.... [I]n Figure 1... [because] there doesn’t seem to be much disagreement about the economic costs of carbon abatement... it’s about dλ/dt=0. Now, it’s obvious if we’ve gotten this far that there are two ways to argue that this locus should be set higher, and hence imply a lower long-run level of atmospheric carbon: you can either increase the numerator in equation (3) or reduce the denominator.
What Nicholas Stern did was reduce the denominator, arguing that we should use a much lower discount rate than the private sector appears to. I’m still wrapping my head around what I believe about that. But what about the numerator? This... [is] the sensitivity of temperatures to carbon concentration... [times] the sensitivity of [social] welfare to temperature.... Lately, climate models have begun suggesting a lot more sensitivity to concentration, with a number of groups doubling their predicted temperature rise.... Marty Weitzman has managed to scare me, by pointing out that there’s a pretty plausible case that a rise of 5 degrees C – which is no longer an outlandish prediction – would be utterly catastrophic. You don’t have to be sure about this; just a significant probability is enough...
I think Krugman goes astray when he characterizes Stern. Nick Stern assumes something different than a "low discount rate." He assumes:
- No pure time preference--that the welfare of somebody born in 2100 does not count less than the welfare of somebody born in 1960 simply by virtue of birth date.
- The costs of global warming scale with the economy: double the size of the economy and you double the costs of global warming.
Stern's first assumption seems to me to be trivially and obviously correct as a proposition in utilitarian or, indeed, any other view on moral philosophy.
Stern's second assumption is debateable: if you thought that the costs of global warming rose not with the size of the economy but with the square root of the size of the economy, then all of a sudden the r in equation (3) would be much, much bigger.