18
Sep
2015
Economists Are So Clever, They Can Make Good Climate News Bad
My latest at IER. My conclusion:
To repeat, these researchers at Harvard and other elite institutions are very smart, and they haven’t made a mathematical mistake in their models. But the public should pause and ask if these sophisticated maneuvers match the more populist rhetoric they’ve heard on the issue. When even good news—in the form of a lowered estimate on the likely range of human influence on the climate—is construed as cause for worry, don’t people start to get suspicious that this isn’t really a neutral scientific debate?
“Guys, guys! Good news! I found a technological way to reverse the effects of global warming that only costs $50 billion! It sounds like a lot, but it’s much cheaper than having to cut emissions, and best of all, doesn’t require *any visible effort* from the average person!”
Environmentalists: “Oh, .”
Oops, that last line was supposed to have an “explitive” tag. Like, they’re swearing out of disappointment and contempt.
Silas I get the main gist of your joke, but are you referring to something specific from the article?
Just the general mentality of, “Whatever happens, it’s bad news!”
Humans heavily influencing climate -> bad news
Humans moderately influencing climate -> bad news
Climate change mitigation is expensive -> bad news
CC mitigation is cheap -> bad news
OK… Like I said, I knew that was the general theme of your comment, but I wasn’t sure if you were also referring to something more specific.
Silas, that already happened with ocean fertilization.
They dusted some iron into the ocean and discovered that although the world as a whole has plenty of spare iron, it happens to be in short supply on the surface of the ocean and becomes the limit to growth. The experiment demonstrated a huge success, so of course it was immediately banned.
Yes, they have made mistakes in their models. Almost all of them have overestimated warming, even in the low range. They have not been able to determine cloud feedbacks, haven’t taken enough into account concerning the oceans and understate natural variation and cosmic rays. It’s simply not true to say they have working models. They don’t! They just fit curves backwards in time.
Of course they have made mistakes.
Yes, I know the difference between weather and climate; but their inability to predict weather or create models that can do so accurately lends me to question the predictions of something much longer term and more complex like climate. Nothing is perfect, so there has to be flaws in the models. No model is 100% accurate. Von Mises in Human Action cautioned us about buying into predictions, and he was right.
Mistakes were made. The models have refused to comply with the climate patterns.
I don’t know what the weather will be like tomorrow, but I am pretty certain that July will be warmer than December where I am. If you doubt my ability to make such predictions, I am willing to have a small wager….
People generally tend to be risk averse. The degree of risk preference varies like all other preferences.
However, I think that their argument only makes sense if the reduction in the lower bound involved a symmetrical distribution and that thus any change in in the lower bound means an increase in the implied “probability” of future temperatures at both tails.
In reality the lowering of the lower bound back to what it had been in all previous reports save the last one and in fact making the range the same as it has been since the Charney report in the 1970’s, has only happened because the actual statistical distributions, which are skewed right, have been shifting to the left in recent years, and in particular the “fat (right) tails” have been reduced in many recent studies. So even though the IPCC’s stated range is back to what it’s always been, the actual variance of distributions has decreased. There’s no way to spin that into bad news no matter how risk averse people are.
Although personally I think they’re misinterpreting the statistical distributions of sensitivity estimates. They’re Bayesian “probabilities” not frequency probabilities. They don’t reflect the actual probability of future climate states, they reflect how plausible scientists think future climate states are.*
*Yes I am simplifying hear by treating sensitivity distributions as the same thing as future state distributions, but one roughly maps to the other, plus, that’s implicit in their argument as well.
Unfortunately a lot of times it is aversion to only perceived risk, not actual risk.
Well, there IS an actual risk. It may be infinitesimally small, but it’s there.
“People generally tend to be risk averse.” Why are lotteries so popular, when the expected return is less than zero and the whole enterprise is inherantly risky?
Bob, I wanted to let you know about a concerning issue. After I wrote the 11th comment on this page, at the top, the number of comments was properly listed as ’11 comments.’ However, when I went back to your main page, it still said ’10 comments.’ I tried refreshing the page several times, but it still said ’10 comments.’ Finally, I wrote a 12th comment, and both the number of comments on this page and when viewed from the main page were in congruence as ’12 comments.’ I have no idea why that temporarily happened, but I felt compelled to let you know in case this problem occurs again and affects absolutely no one in any significant way.
Bob, when can we stop reading your IER columns for substance and just read them for the surrealistic humor in the quotes you find?
This video is like Bob Ross ASMR level anarchy:
http://m.imgur.com/b84EOrA
“Suppose the AR5 had instead raised the lower bound from 2°C to 2.5°C. Would Ivy League economists have produced a paper showing that this actually reduced the need for a carbon tax?”
I don’t get this comment, as a raising of the bottom limit would represent an increase of the possible temperature and a decrease in the certainty. How could this be interpreted as reducing the need for carbon tax?
However, I surmise that it would be possible to conclude that an increase in the upper bound could be good news under certain assumptions.
The bit in the paper about mean-preserving functions I do not quite follw. “An MPS results in the variance of ?2being greater than that of ?1, with the means of the two distributions being the same”. Figure 1 shows “The offsetting rightward distribution of mass, so as to preserve the mean, is in the center of the distribution. There is no increase in mass in the right-hand tail.” Is there a good reason to preserve the mean?
Bob,
I still don’t get your summary. To use your lottery example, under any reasonable assumptions, I don’t see how increasing the upper bound in the lottery would harm me.
No one is so risk averse that they would prefer a lottery that might pay $900 or $1,100 to one that might pay $900, $1,100 or $1,200, are they? (Unless you thought that the increase in the top prize told you something about the distribution).