Half-Truth
Jonathan Chait has written a column, titled Half-Baked, in the current issue of The New Republic that attacks the global warming positions that I have presented in National Review.
Global warming is a complex topic that depends upon inherently speculative predictions about the world’s economy and environment more than a century into the future, so there is plenty of room for principled disagreement. As a general rule, however, people like Mr. Chait who speak about it in tones of certainty mostly reveal that they are uniformed on the subject. While I recognize the constraints on length that prevented him from producing a doctoral dissertation, he never actually engages with the central arguments that I’ve made, at any level of abstraction.
A simplified version of my core argument can be presented in one paragraph:
Carbon dioxide is greenhouse gas, and if you put more of it in the atmosphere, then all else equal the Earth will get warmer. The key unknown, because of the complexities of climate feedbacks, is how much warmer. The UN IPCC forecasts that Earth will get something less than 3C hotter within about a century, and further estimates that 4C of warming would cause the world to lose 1% – 5% of GDP; therefore the expected costs of global warming are on the order of 3% of GDP sometime well into the 22nd century. This is a huge amount of money, but not exactly consistent with Miami becoming an underwater theme park. Given that global consumption is projected to grow from about $6,600 per person per year today to about $40,000 per person per year over the next century, it’s pretty hard to justify massive sacrifice of wealth today for the purpose of preventing our descendants a hundred years from now being only 5.7 times, instead of 6 times, as rich as we are. The real risk is that science has radically understated the greenhouse effect. A carbon tax designed for the expected case can safely be avoided for decades, while a carbon tax high enough to ameliorate a low-odds disaster scenario would be insanely expensive. Relatively low-cost investments in specific technologies that would be useful if such a disaster scenario arose, on the other hand, are a smart insurance policy.
Here is how Mr. Chait presents my views:
Republicans are no longer denying the scientific basis for global warming. That’s good news for those of us who have grown accustomed to the continued existence of things like polar sea ice, various forms of life, and Miami. The bad news is that Republicans, having seen the light, have fallen back on the possibly even more annoying stance of simply refusing to do anything about the problem.
The new conservative stance was presaged last summer in a National Review cover story, “Game Plan: What Conservatives Should Do About Global Warming.” The article, by Jim Manzi, began with a frank acknowledgement that global warming could no longer be denied. After this promising start, the argument swiftly degenerated. Manzi proposed “development of tactical technologies, such as carbon sequestration and cleaner-burning [oil-fueled] engines.” Of course, everybody is for those things. The problem is that dirty energy sources like oil and coal are far cheaper—and will remain cheaper for a long time—unless the government somehow increases the cost of carbon emissions. This, however, is where Manzi begs to differ. “Conservatives,” he writes, “should propose policies that are appropriately optimistic, science-based, and low-cost.”
How, exactly, conservatives can persuade people to give up cheap energy sources without imposing a cost on them Manzi does not say. (Perhaps this is where being “optimistic” comes in.) ….
Rather than address the uncertainty that is at the heart of the problem, Mr. Chait falls back on scare stories and cat calls.
Mr. Chait then goes on to describe the political strategy that I outlined in NR:
Instead, he proceeds to a gleeful discussion of how conservatives could win votes by opposing carbon taxes—most likely by luring “old-line industrial-union members” ticked off by the soulful entreaties of Al Gore and other members of the “Hollywood and political smart set.” The essay concludes on the triumphal note, “Global warming can be the first wedge issue of the 21st century.”
I believe that clever conservative politicians could make this a winning political issue, because upon careful inspection it becomes apparent that the proposed programs to radically reduce carbon emissions through taxes or regulation are very likely to cost much more than the benefits they provide. Mr. Chait finds it impolite to mention this in mixed company because it would disadvantage his preferred political candidates. Sorry about that.
The second part of Mr. Chait’s article is devoted to criticizing a specific oil industry tax advantage about which I have no informed opinion, and about which I do not believe I have ever expressed an opinion.
(cross-posted at The Corner)
Is it just me, or do we run into epistemological problems when talking about climate change. Isn’t predicting the weather not unlike charting the movement of the stock market – there is no reason to suppose that there might be a sudden drop or change in temparatures?
— Mark · Mar 26, 02:50 PM · #
Isn’t defining the effects of a century’s worth of global warming in terms of global GDP sort of glossing over some things? If global warming caused a huge uptick in the amount of species going extinct, which some ecologists seem to think might happen, or, say, caused the death of masses of the world’s poorest citizens, without denting global GDP, wouldn’t it still be a bit glib to look at it all and say “awww, only 5.7 times richer”?
I mean, when we talk about Brave New World, The Matrix, Soylent Green, we don’t dwell on the percentage difference in global GDP is between their societies and ours. We tend to focus on the intangibles.
— Dave · Mar 26, 04:58 PM · #
I’m not sure if you’ve covered this elsewhere, but would you object to a revenue-neutral carbon tax? I believe that British Columbia has recently enacted a similar proposal.
— Will · Mar 27, 05:41 PM · #
“Therefore the expected costs of global warming are on the order of 3% of GDP sometime well into the 22nd century.”
I’m not sure what you mean by “expected” costs. That might be the “most likely” costs, but it’s not the “expected” costs in the sense of “expected values” from statistics. To find that you need to consider the tails of the distribution as well. The actual mean costs of global warming are probably heavily affected by the risk of really bad outcomes if the warming is worse than expected. It doesn’t make any sense to ignore those outcomes in assessing the risk posed by global warming.
(Another example: if your investment strategy is to sell leveraged out-of-the-money puts on the S&P 500, your “most likely” outcome will be a healthy gain, but your “expected outcome” would be small or negative, because the downside is very large relative to the upside.)
— ed · Mar 27, 07:15 PM · #
I’m not defending the Chait piece, but I think this line represents an error. You’re assuming a discontinuity in our available responses to global warming that I don’t believe exists. You talk about a “massive sacrifice of wealth”, but if we really think the sacrifice of wealth is so massive, we could just have a smaller tax. It doesn’t follow from the fact that the world will be richer in 2100 that we should do nothing whatsoever to guard against future dangers, but that’s essentially what you’re claiming.
Whether we face the expected or the low odds catastrophic scenario, modest measures to decelerate the rise of atmospheric CO2 will modestly improve our lives (or chance of survival if things became REALLY catastrophic). You’ve got good arguments for why those measures should be modest until more data is gathered, but I haven’t seen any argument for why there should be no measures at all, unless there is some threshold I’m unaware of below which carbon-reduction does absolutely no good.
— Consumatopia · Mar 28, 01:31 AM · #
Mark:
I’ve written a fairly detailed article on uncertainty in climate modeling here: http://article.nationalreview.com/?q=ZmViY2Y3YzY1YmVkYTg4NjczODhkYWU1Mjg1YzhjMTI=
Dave:
Money is not everything, however, I don’t think the idea of losing 3% of wealth in a world that is 5 times as rich as ours comports with “a global emergency” or whatever. Alternatively, what China and India, for example, would have to give up today is some economic growth that would otherwise translate into imporved health care, jobs and so forth that I think most people would see as very basic.
Will:
I’ve written a very long post on that topic: http://theamericanscene.com/2007/12/05/coase-club
Ed:
I actually mean precisely expected value, as in the odds-adjusted loss using a PDF of climate sensitivites. In turns out, by the way, that this is very close to the projected economic loss from the central tendency – what you’ve called “most likely” – for climate sensitivity (about 3C per doubling of atmospheric CO2), because the PDF and the cost / unit warming are both very roughly symmetric around the mean. See, for example: http://nordhaus.econ.yale.edu/dice_mss_091107_public.pdf
Consumatopia:
I’ve gone into this point in some detail in numbered items 2 and 3 of the follwoing post: http://theamericanscene.com/2007/12/05/coase-club
— Jim Manzi · Mar 28, 04:35 AM · #
Overlapping obviously does not mean coincident. Coal and oil’s min, max and mean were way higher than the corresponding values for wind and solar. Again, you’ve got a good argument for a smaller tax, but not one for no tax.
Yeah, the price is going to be set politically, but given the thorny politics of the situation that you so wonderfully described, that politically set price is almost certainly going to be less than the likely actual externalities. In any event, the taxes would almost certainly start very small then rise (or disappear) as new data emerges.
You’re using the low cost of the tax as an argument against the tax. No.
I think markets work better than government at choosing research priorities (and, boy, if you think that tax policy is politicized, you haven’t seen anything yet), and because there are reasons not related to climate change why we should consume less oil (foreign war externalities), I would definitely prefer the former. I mean, why are you such a fan of government direction in this instance?
Eh, I suspect a smaller or more gradually rising tax would have proportionally smaller political baggage and pork. And there’s definitely no reason to think it would have more pork than directed research, which could be entirely pork. So it’s also cheap to not wait. We might as well go with whatever action has the highest expected return (a small tax), because both paths are cheap. If more data indicates that AGW is bogus, we’ll abolish the tax. It’s a risk, but it’s not a very dangerous one. The annoyance of hearing you “I told you so” us would probably exceed the actual pain of the tax itself.
On the other hand, if data changes in the other direction and AGW turns out to be a bigger problem than we think, setting up a small tax now would make it easier to restrict carbon further in the future.
There’s a strange loop here in that you personally agitating against the carbon tax because it’s too politically complicated to pass only makes it more politically complicated to pass.
— Consumatopia · Mar 28, 02:02 PM · #
Consumatopia:
Thanks for the detailed and thoughtful comments.
I guess I consider it a judgment call about whether a distribution of possible costs that ranges by a factor of 100,000 (no typo) for coal, as an example, contains sufficient information to be the basis for setting global energy policy. Yes, the means are different for external cost estimates by fuel type, but the overlaps are so signficiant that it’s a real leap of faith to believe that when we use a tax to push some energy production from, say, oil to nuclear, we’ve really reduced net external costs.
The much larger point is that the “optimal” tax ramps up over time from now to 2250. The tax is “relatively” (relative, that is, to the later tax rate and to the rate that would be required to immediately push carbon consumption way down) low through about 2050, then goes way up. Relative to most other human projects the tax is a huge amount of money right away – trillions of dollars. If our climate models are correct, then (in my view) we would never want to implement such a tax. Only if the climate is far more sensitive to carbon than current science estimates would we want it. So in the relevant situation, the amount of emissions abatement that such a tax would accomplish wouldn’t be very valuable by, say, 2050. What we would want is to have done the long-lead time engineering / basic research that can’t be easily time-compressed that would provide emergency technological options.
— Jim Manzi · Mar 29, 01:08 AM · #
I don’t see how you can call the most probable outcome a leap of faith. The 25-75% bands are a lot narrower than the full range. While the government policy might not set exactly the right externality, it’s a lot more likely to be correct than the value of zero that the market would set. There’s nothing magical about the marketplace that automatically makes its decisions safe. In fact, in this case, quite the opposite—if you want to avoid leaps of faith, you should avoid using the form of energy with a 100,000 factor of possible external costs.
I still don’t see why we couldn’t have a small tax relative to most other human projects in the short term then adjust it as more data accumulates. I guess by scare-quoted “optimal” you mean optimal with no time discounting or something, and I can understand why you wouldn’t want to implement that, but rather than being a hard choice between no tax and the “optimal” tax, it seems like we have a continuous selection of tax rates in between those.
There’s also a continuous range of climate outcomes, is there not? Just because carbon sensitivity is such that 2100 will be much worse than we think that doesn’t mean it will be so bad that none of our prevention measures will do any good at all—or that the magic tech fix might be so great that no prevention is necessary. It would awesome if some magic technology just fixes all of our problems without creating new ones, but in practice that never actually seems to happen. At least, prevention would give us a few more years to find the magic technology.
Technology isn’t the only thing with a long-lead time. The decisions of businesses in what kind of factories making what kind of products, or the decisions of consumers in where they should live, where they should work, and how they get between the two, are difficult to redo once done. It seems like it would be wiser to let the marketplace decide which adaptations should be made and which technologies should be researched rather than have the government decide by fiat.
— Consumatopia · Mar 29, 03:48 PM · #
I am with Consumatopia. Jim appeares to conceed that the optimal tax is not zero. And I believe that is before taking into account the distortionary effects of other taxes.
Jim talks about the “cost” of a carbon tax in the “trillions”, but I believe he is tallking about the size of the tax itself (discounted over decades I assume ..). But this is not the cost. The cost is the relative deadweight loss of a carbon tax versus whatever the alternate means of funding the government. The first order cost of government is set by our spending decisions not our tax decisions (assuming the tax system is not truly awful).
Given that, I assume Jim believes that this new type of tax will lead to increased government spending. The last decade shows anecdotally the opposite and I believe the research bears that out as well. Cutting taxes does not mean cutting spending and raising taxes does not mean increasing spending.
best,
Tom
— Tom G. · Mar 30, 12:06 PM · #
Tom G:
I think I’ve addressed these points in the post that I highlighted for some previous commenters: http://theamericanscene.com/2007/12/05/coase-club
Consumatopia:
I don’t accept that the mid-point of these studies is the most probable outcome (or said more precisely, that even if one were to accept this is the mid-point of some PDF as the true “best estimate” that the confidence interval is so incredibly broad that it provides almost no information).
You say: “I still don’t see why we couldn’t have a small tax relative to most other human projects in the short term then adjust it as more data accumulates.”
There are several problems with this. The first is that we already have a tax on carbon fuels – for gas, for example, it averages about 46 cents per gallon in the US and several dollars per gallon in Western Europe. Most carbon tax proposals call for something like 10 cents per gallon today ramping up to 40 – 50 cents er gallon by 2050. So we are already charging a “carbon tax” on gas greater than what AGW costs supposedly justify. Should we just assume this away and raise the tax more? Do you really think that raising the tax on carbon by an average of say 25 cents per gallon over the next 40 years will more than marginally reduce CO2 emissions? If you accept my premise (which obviously you may not) the only situation in which emissions reductions can be cost-justified is one in which climate is far more sensitive to GHGs than current models estimate, and in that case marginal reductions are, in effect, useless. (Yes, I recognize that these functions are continuous and in speaking of “marginal” “useless” and so forth, I’m simplyfying).
I only put quotes around “optimal” to indicate that the exact tax ramp that is best is a subject of debate.
— Jim Manzi · Mar 31, 12:45 AM · #
confidence interval is so incredibly broad that it provides almost no information
Which is definitely not the case unless you force the confidence interval to include the outliers at the very top and bottom, which is absurd.
_So we are already charging a “carbon tax” on gas greater than what AGW costs supposedly justify. _
Because there are additional externalities involved with petroleum consumption beyond those of AGW.
Do you really think that raising the tax on carbon by an average of say 25 cents per gallon over the next 40 years will more than marginally reduce CO2 emissions?
Two points here. The ramping up of rates sends a signal to private research that renewable energy will be lucrative in the future. But, again, if the tax isn’t large enough to help anything, it’s also not large enough to hurt anything.
And I think your simplification is carrying more weight than you realize.
— Consumatopia · Apr 1, 02:16 PM · #