Sunday, November 07, 2010

Judith Curry, positive feedbacks, and AGW bubble

There may be things in which I disagree with Judith Curry but her viewpoint on the positive feedbacks in climate science - and not only climate science - are spot on, I think.
Reversing the direction of positive feedback Part I

Reversing the direction of positive feedback Part II
You may want to read them. Aside from the links, this blog entry is not excessively important. But I just reproduce a comment your humble correspondent has posted on her blog.

Dear Judith, I also think that you focused on actual values – and not on particular people – which is the right thing to do.




While I realize that your writing is still courageous given the atmosphere that persists in certain circles (and the deafening silence of the Academia when it comes to the defense of integrity is something I also know from very different contexts, too), I am among those who find most of your findings so self-evident that it would be hard for me to view you as the discoverer of some novel insights. Sorry for that.

Of course, you are totally right that there have been positive feedback loops that made this panic exponentially grow for years.

You say that some scientists gained, some scientists lost. Well, this sounds balanced except that in the climate science, while 1/2 of the pre-IPCC scientists could have gained and 1/2 of them could have lost, the counting is completely different if you consider the present community to be 100%. In the new community, 90% of the people were either added or “heavily amplified” relatively to the pre-IPCC era, for the money acquired because of the panic, and 90% of the current community has gained.

The losers were more widespread among the general population although, of course, rare scientists who kept their integrity and who realized that no new “radical” insight that could change the climate science had been made, so there was no sudden reason for any panic.

Concerning your historical question how it started, it’s clear that there were feedbacks but if one wants to be more detailed, one would have to talk about particular people and institutions. They would clearly include environmental activists, scientists who honestly believed a threat that was a fringe science a few decades ago, but maybe even Margaret Thatcher who previously wanted to suppress the mining unions (before the IPCC was born).

Many of the initial people did “almost” legitimate things. Again, I have no doubt that some scientists had legitimate concerns and they honestly believed that what they preached followed from the physics. However, there abruptly came a moment for “speculators” – people who made bets that this thing would grow and they could grow with it.

The whole concept of the "consensus science" is a manifestation of this speculation. It wants everyone to think that as the hysteria grows bigger, there will eventually come a moment when everyone agrees that the Earth is threatened by man-made climate disruption. Most people who have promoted this "consensus" implicitly make this assumption about the future trend of acceptance of the climate science.

The people who have driven the monster were in many sectors – media, politics, activism, and of course science – and these parts of the society, with a vested interest to support the alarmism, managed to do so, indeed. They could support each other by having alternative tools. The politicians could bring grants to the scientists, the scientists could have equipped the politicians with an official scientific stamp, the journalists could promote all this propaganda among the readers, and so on.

Why did such bets work at all? Well, for the same reasons why all bubbles – in the financial markets etc. – can grow. When people begin to believe that a thing has a “momentum”, they will make bets that the price of it will keep on increasing, which is why it will indeed be increasing as the demand grows.

This comment is true both for memes as well as houses, dot-com stocks, or anything else. If the “momentum players” begin to dominate, positive feedbacks prevail and one gets closer to an unstable behavior where things easily get out of control.

However, in healthy stable markets – including markets of ideas – most people are not momentum players. They have an independent idea what the absolute price of a house – or an idea – should be, instead of an idea about its time derivative. If they have such an idea, they inevitably drive the price towards an equilibrium where the supply and demand match – given the existing knowledge – and their acts stabilize the market. If they think that too much interest is dedicated to the alarm, they will leave it because it is overpriced, and vice versa.

So the momentum players are the true sources of instability in the markets – and markets of ideas. They’re clearly pure noise-that-turned-systematic and they don’t contribute anything independent to the correct evaluation of prices (in the markets) or the convergence to the truth (in science). What goes wrong that such people begin to dominate?

Well, under stable circumstances, it doesn’t pay to be a momentum player. If you know that things are stable or meritocratic, the rationally unjustified increase of price – or an unreasonably elevated interest in an idea – will be more probably than not followed by a decrease, so it’s unwise to make a bet when the price is peaking or after it just went up.

However, sometimes players just know that the behavior is unstable and goes out of control, so they benefit – until the bubble bursts. Of course, bubbles will ultimately burst but the momentum players may delay this moment almost arbitrarily. People in the markets and/or science will be more eager to participate in new irrational bubbles if they learn that they can gain but they cannot lose much. That’s why I find it important to punish the people who were responsible for this thing’s having gotten out of control as long as we can show that they were not honest (which is hard).

We’re not there yet because it’s not really clear who is in control right now. But I am telling you: if and when the sensible people who don’t like bandwagons and uncontrollable strengthening of ideologies take over, it is necessary to wisely, moderately, yet resolutely punish those who drove science – and policies – in wrong directions at uncontrollable accelerations. Such a punishment, however cruel it may sound to the nice people on the sensible sight, is necessary for these (and perhaps worse) things to be avoided in the future.

While I agree with your observation that the positive feedbacks may also accelerate the system in the opposite direction, I don't think that this has actually occurred during the last year. Since the ClimateGate, we've just seen an unusual slowdown of the AGW panic - the lack of the usual exponential growth of the panic - which looked like a positive feedback loop. But it was actually just its absence.

In reality, I don't really know how the opposite mechanisms could exponentially grow. Will we try to increase our CO2 emissions exponentially just for the fun of it? I don't know who would advocate such an attitude. The maximum thing I can imagine is that the AGW panic will almost completely disappear - it will return to the fringe activist groups where such things used to belong. But if that's true, the final point is a stable equilibrium so there are not excessively positive feedbacks in the convergence towards this more sensible world.

2 comments:

  1. The stock market/climate issue analogy is pretty good, I think.

    However, I think you neglected a key element: the role of the professional journalist/stock broker.

    I have been informally observing press hysteria (i.e. positive feedback, bubble) for a few years. Many people assume that it is the duty of the professional journalist to check facts and opposing views. But like a stock broker makes money on trades regardless of whether his clients make or lose money on them, the professional journalist makes money on the popularity of his articles regardless of whether his readers gain or lose realisticity from them.

    A typical example that I am familiar with is the Toyota recalls. Toyota had some minor but statistically significant problems with accelerator pedals that warranted action. Although the problems were on the same order of magnitude as the usual problems that almost all other manufacturers routinely have, the press was able to exaggerate the problem to make it seem much bigger.

    Then, when it finally became obvious, the same publishers jumped to the other extreme and made a big story out of claiming there not really being any problem at all, not realizing, or hoping that their readers would not realize, that the press itself was instrumental in propagating the exaggerated story. Thus the press profits from the bursting of its own bubble.

    I am of the opinion that there is a systematic flaw in the press system: Once an article gets into one of the press agencies, it is often reproduced and distributed thousand fold without any further examination. Journalists seem to be mainly occupied with rewriting articles from the press agencies so that they fit into the allotted space, rather than examining the accuracy of the content.

    Why is there the distinction “investigative reporter?” Apparently they are in the minority. By the time the rare “investigative reporters” get around to debunking the parrotive reporter’s tales, the AP & Co. has already fixated them into the minds of millions.

    In the “source – press – public” system, the sources are the inputs, the press is a signal accumulator and modifier, and the public is a sometimes unpredictable feedback. This is oversimplified, but the general idea. In an ideal world, the source and the public, which are supposed to be uncontrolled, would rely on the journalists adhering to professional standards to control themselves and maintain their own output signal quality.

    ReplyDelete
  2. The stock market/climate issue analogy is pretty good, I think.

    However, I think you neglected a key element: the role of the professional journalist/stock broker.

    I have been informally observing press hysteria (i.e. positive feedback, bubble) for a few years. Many people assume that it is the duty of the professional journalist to check facts and opposing views. But like a stock broker makes money on trades regardless of whether his clients make or lose money on them, the professional journalist makes money on the popularity of his articles regardless of whether his readers gain or lose realisticity from them.

    A typical example that I am familiar with is the Toyota recalls. Toyota had some minor but statistically significant problems with accelerator pedals that warranted action. Although the problems were on the same order of magnitude as the usual problems that almost all other manufacturers routinely have, the press was able to exaggerate the problem to make it seem much bigger.

    Then, when it finally became obvious, the same publishers jumped to the other extreme and made a big story out of claiming there not really being any problem at all, not realizing, or hoping that their readers would not realize, that the press itself was instrumental in propagating the exaggerated story. Thus the press profits from the bursting of its own bubble.

    I am of the opinion that there is a systematic flaw in the press system: Once an article gets into one of the press agencies, it is often reproduced and distributed thousand fold without any further examination. Journalists seem to be mainly occupied with rewriting articles from the press agencies so that they fit into the allotted space, rather than examining the accuracy of the content.

    Why is there the distinction “investigative reporter?” Apparently they are in the minority. By the time the rare “investigative reporters” get around to debunking the parrotive reporter’s tales, the AP & Co. has already fixated them into the minds of millions.

    In the “source – press – public” system, the sources are the inputs, the press is a signal accumulator and modifier, and the public is a sometimes unpredictable feedback. This is oversimplified, but the general idea. In an ideal world, the source and the public, which are supposed to be uncontrolled, would rely on the journalists adhering to professional standards to control themselves and maintain their own output signal quality.

    ReplyDelete