I just finished reading Shellenberger & Nordhaus' latest, and while I realize I am a bit late to the party, I think they say some fascinating things -- perhaps not for the reasons they intended.
S&N manage to succinctly distill an awful lot of the ideas that are core not only to policy debates on carbon, but to policy discussions of any major change to the economy. Understanding these biases is critical to understanding why S&N write what they write, but also why they are so deeply wrong.
I recently joked to an economist friend that economics is the only discipline where you can get fired if you admit you are unable to predict the future. The comment has too much truth in it to be particularly funny, though. We not only have heaps of economic models out there (on the basis of which we make massive policy decisions), but we also have heaps of evidence that the predictive efficacy of the models is awful. From Laffer curves to future costs of SO2 tradable credits, the world is awash in economic predictions that were proven grossly wrong -- yet we keep making decisions based on the same set of core assumptions.
The irony in this is that the way to get a Nobel Prize in Economics seems to be to prove that classical economic theory is flawed (see Tversky, Solow, etc.). Yet we keep using it, on the basis, so far as I can tell, that we might as well until something better comes along.
I should make it clear that I'm not hostile to economics. It does a great job at providing a narrative to make sense of historical events -- just a lousy job predicting future events. In Consilience, E.O. Wilson differentiated between the sciences and the pseudosciences by noting that only the former are verifiably predictive. By this definition, economics is a pseudoscience. Arguably, so is much of medicine, and this is fine so long as the discipline is honest enough to acknowledge its limitations. My beef with economics is that it keeps on dogmatically asserting the accuracy of its crystal ball in spite of all available evidence to the contrary.
So why does this matter to Shellenberger & Nordhaus? Four reasons:
(1) They implicitly presume that all futures are known and testable.
This is implicit throughout their work, when they assert what prices will have to be to lower carbon and what technologies will participate. But since when are all futures known? Indeed, much of the most interesting work going on in economics right now focuses on the psychological and sociological drivers behind individual action. (Softer than econometric modeling, to be sure, but vastly more accurate -- witness Tversky and the behavioral economics crowd.) Consider this gem from Jeffrey Luke:
Individuals have a natural tendency to choose from an impoverished option bag. Cognitive research in problem solving shows that individuals usually generate only about 30 percent of the total number of potential options on simple problems, and that, on average, individuals miss about 70 percent to 80 percent of the potential high-quality alternatives.
This is both insightful and obvious. Next time you go out to dinner, try to imagine what choices you'll have on the menu and what you're going to pick. Compare that to what your choices actually are and what you actually eat and you'll get to Luke's 70% failure rate. Now expand this to the much larger question of what technologies and approaches society will select in response to a carbon price signal, and it becomes obvious how deeply flawed any prediction based on that impoverished option bag is likely to be.
(2) They claim expertise they lack.
They write that "technically, there simply do not exist the low cost, low carbon technologies that could be quickly brought to scale to replace carbon intensive energy sources." How the heck do they know this? How does anyone know this? Again, we've got that impoverished option bag. Here's a great example: I am willing to bet that we could find a virtually identical quote in the late '60s / early '70s from some economist who concluded that Rachel Carson and her crazy hippie followers were delusional. (Perhaps someone can find one?) And here's a great statistic: If the U.S. still had the same energy intensity (Btus of primary fuel per dollar of GDP) that we had in 1970, we would presently use three times as much primary energy as we actually use today. (Thanks to the folks at ACEEE for this great calculation.) This makes energy efficiency by far the most significant source of new "load" over the past three decades. I challenge you to name any economist this side of Amory Lovins who predicted anything close to that potential. Outside of Skip Laitner at ACEEE, who is seriously talking about our "strategic energy efficiency reserve" as a resource moving forward? No one -- and yet we are making policy decisions as if these options don't exist.
(3) They provide an opening for my latest favorite joke.
It seems to be implicit in much of their writing that carbon capture and storage is going to have to play a major role in a carbon-constrained world, and since CCS is really expensive, that means carbon control will be really expensive. This is as logically bizarre as it is ubiquitous, and yet (outside of Gristmill, at least) we keep having this same tedious conversation. Sing with me folks: "just because something is expensive doesn't make it good!"
Okay, so here's my joke. I was recently at an energy conference in Houston, having that exact same conversation with a guy who had a small consulting business to help companies hedge energy price volatility. He was also helping them hedge carbon, which he figured was eventually going to have to be priced at $100/ton, since below that point, CCS didn't pencil out. Try as I might to convince him that his logic was flawed, he kept clinging to the number. So finally I cut him off and said, "Look: if you paid me a million bucks, I might have sex with a goat. But that wouldn't make it a good use of a million bucks." Needless to say, that killed the conversation, but we all laughed. My own blue humor notwithstanding, it is exactly the same with the CCS nonsense. Get over it.
(4) They assume that carbon control is incompatible with economic growth.
As I've said many times, this is totally, grossly flawed, for the simple reason that it assumes that energy markets are economically optimal. They aren't even close ... ergo, we can get massive carbon savings concomitant with drastic reductions in energy cost. The electric sector is half as efficient as it was in 1910, thanks to our goofy regulatory model, so we could slash its emissions and costs in half (and this is the single biggest source of both in the economy) just by deploying 100-year-old technology. Efficient, my butt. Here's another source. The Arizona Climate Change Advisory Group is serious about climate change, and they've got a much less impoverished option bag. They looked at all the technologies out there to control carbon and found that meeting their commitments would an average cost of negative $12.74 a ton. Oh, and it will create 285,000 jobs in the state as new businesses rise to deploy all those low carbon, revenue-generating technologies. The example is far from atypical. As Ken Colburn of Symbiotic Strategies has pointed out, the states that are actually doing something about climate change right now are consistently finding that carbon reduction (done right) has a net negative cost, and the total NPV worldwide in present dollars is in the neighborhood of $1 trillion. This completely changes the game. If China doesn't want to participate, fine -- we'll just take our manufacturing jobs back as our energy costs fall. The economic benefits serve only to double the imperative to act -- but first we have to get beyond those impoverished option bags telling us that the costs of action are too high. To be perfectly blunt: Shut up.
One last point. I do accept S&N's core existential question. When they write, "will green groups transform themselves into institutions motivated by a vision of prosperity and possibility? Or will the remain grounded in the the politics of pollution and limits?" I agree it is a good question. The environmental movement has for too long had a knee-jerk reaction to economic growth, to its great detriment. But I'm not sure it matters. Dow and BP have both made public statements indicating that setting internal goals to reduce carbon created hundreds of millions of dollars of shareholder value. Our company is doing the same. If the green groups get on board and give their imprimatur to those efforts, so much the better -- but the economic incentive is going to keep that train rolling forward regardless.
Comments
View as Threaded
Jason D Scorse Posted 5:41 am
10 Oct 2007
That being said, a few questions back at you:
Are there other disciplines that you think are better at predicting the future than economics? And even on a macro scale?
Do you think economics has any predictive power that we should rely on?
If economics is just a descriptive science with little predictive power, than what would you suggest we use to make informed decisions about current environmental problems? Psychologists? Philosophers? Obviously, this doesn't have to be an either or but what other disciplines in the social sciences do you think should have a more prominent rule in actually crafting the policy details?
Finally, who would you prefer to manage the Central Bank, Ben Bernanke, or someone without a PhD and serious expertise in economics? If your answer is Bernanke, is it because you think he has a crystal ball?
Thanks
I teach environmental economics and blog at http://www.voicesofreason.info.
Permalink
Aklemm Posted 6:41 am
10 Oct 2007
The whole point is that no one can predict the future. The best near term subsitute to a crystal ball is crafting public policy with near and long term feedback loops that constantly inform and allow policies to be adjusted to achieve desired end states.
Permalink
Sean Casten Posted 7:02 am
10 Oct 2007
Answers to your questions.
No, but that's a pretty low bar. The world is rife with examples of how bad economically-literate people are at predicting market changes, from random stock pickers that beat mutual fund managers to forward price forecasts on fuel prices that are extrapolated off linear trends and are wildly inaccurate. So yeah, it's the best tool available, but it is really dangerous to make policy predictions with an assumption that it's got much accuracy to it.
Qualitatively, yes. Quantiatively, no. Do interest rates prime the economy? Absolutely. Are supply and demand related through price? Absolutely. But saying that a 0.5% rate cut will grow GDP by a known percent, or that a $1 increase in gasoline price will lower demand by a known percent is almost pure speculation. This is why I said that economics is very good at explaining what happened - but not so good at explaining what will happen. (For example, an economist would argue - quite rightly - that the last decade has proven that gasoline demand curves are highly inelastic. This is true, and interesting - but largely inconsistent with the vast majority of economic forecasts prior to 2000 which assured us in one way or the other than $70/bbl oil would cripple the US economy.) This is good so far as it goes, but is a shaky foundation upon which to build a discipline, unless the discipline is willing to admit and seek to address it's limitations. After all, given perfect knowledge of the future (as all backwards predictions have), I can make astrology look like a scientifically valid discipline. I'm not suggesting that economics isn't a lot more rigorous than astrology, but just going back to my first point that this is a pretty low bar.
Great question, and I'm honestly not sure. But I go back to the policy point. One can make the broad observation that we choose from an impoverished option bag (a psycological point) and that virtually all policies that were anticipated to be economically disastrous (think SOx trading, the first generation of CAFE, etc.) have actually had way lower costs than we forecast, in large part because of that impoverished option bag. And if one makes that point, one quickly would conclude that the future is much less knowable than economics would have us believe. And yet we don't. We get the S&Ns of the world assuring us that carbon will have to get to $100/ton because their bag doesn't include stuff that works for less... and policy makers act on that information. So maybe my answer is not that there is a better tool, but that the tool needs to be presented with a lot more humility.
I think a PHD is overrated, but I get your question. : ) Again, there are important economic tools out there, and we clearly need people making major economics decisions in this country to be familiar with those tools. But as anyone who's ever built an econometric model knows, (a) you can never build enough complexity into a model to represent all the things that will actually affect your forecast and (b) complex systems behave in unpredictable ways. And so we make massive simplifying assumptions. Capital is efficiently allocated, all market participants are rational, people knowingly maximize their utility function... We don't build these assumptions in because they're true, but rather because if we don't, then we can't apply any mathematical models. And so what I really want out of our Fed chief is not a fancy degree, but the gut level experience to look at a model and say "your math is right, but your answer is crazy." And the ability to do that ultimately comes from experience - and it's not very mathematical when it's done really well.
One last point, to illustrate my larger beef. At any given time, the Congress has numerous policies to consider, all of which have a "score" based on their net fiscal cost. Decisions are made between competing provisions based on that score. And the score is based only on the net impact on tax revenues. So let's say we want to pass a tax break for some clean technology. You and I may be quite confident that this industry is poised for explosive growth, would create a ton of jobs and lower the cost of power. None of that matters for the scoring - just the tax impact. The economists at the Congressional Budget Office put very precise, very accurate forecasts together on the costs, but put none into the benefits. And so we make policy with one hand clapping. And clapping wrong, since even the option bag they have is impoverished.
Permalink
mkayser Posted 9:43 am
10 Oct 2007
There are mathematical ways to account for uncertainty. We could put some "give" in our cap-and-trade schedule. I expect this often doesn't happen because it's too complicated for the policymaker.
I'm appreciative of your helpful counterbalancing arguments in favor of experience-driven intuition over fancy math. But there are computers right now that basically run hedge funds, and some of them make a ton of money. That's just straight up math. Often using ridiculously unrealistic models of stock pricing, I might add.
The two schools of thought should continue to have a healthy tension. Humility on both sides is warranted. Your post is well taken.
Permalink
mkayser Posted 9:49 am
10 Oct 2007
Policymakers really like clear answers and can't handle complexity. I don't deal with policymakers, I'm just stereotyping.
Cap-and-trade policies etc. really are based on "first-best-guess" as opposed to a "smeared guess" that gives rise to many possible paths depending on what happens. I don't work on policy so I don't know these details, I'm just assuming.
Permalink
Jason D Scorse Posted 12:53 pm
10 Oct 2007
Also, I agree that economists' point estimates should be taken with a grain of salt. We can't know virtually any future economic activity down to the dollar. But we can predict reasonable ranges and orders of magnitude, and I would argue that this is crucial and very useful. Economists might not be able to tell you exactly how much gas consumption will go down if we had a $1 a gallon gas tax, but they can present a reasonable range that is probably pretty accurate. I'll take that. And I wouldn't trust anybody else's guess that's for sure.
My point about Bernanke- who I think is doing an amazing job- is that just because he doesn't have a crystal ball doesn't mean that he doesn't have a very good sense of trying balance these very complicated forces that have such huge implications for us all. Maybe a PhD in economics is overrated, but not expertise in economics. We have people like Bernanke at the helm because this stuff is serious serious business and we want people with a profound understanding of economic forces to set things like interest rates. If Bernanke's salary was $100 million it would be a bargain.
Anyway, as to economists not being humble or admitting their imperfections, I don't know. I see a lot of hubris and arrogance in just about every profession, especially throughout the humanities and the rest of the social sciences. And a lot of it is spewed by people who don't have a clue about economics. But anyway, that's a secondary issue.
Thanks again.
I teach environmental economics and blog at http://www.voicesofreason.info.
Permalink