Black Swans

Paul Krugman writes that you can't think correctly about climate change unless you take adequate consideration of relatively unlikely scenarios for disaster. Psychologically, if there's only a one or two percent change of something happening, we tend to put it in the "not going to happen" file. But sound policymaking would consider a one percent chance of a scenario in which billions die to be something worth worrying about.

I've been thinking about this lately as I re-read Watchmen inspired by the trailer for the forthcoming film adaptation. It's a reminder that in the 1980s, and especially before the Reykjavik Summit there was enormous anxiety about nuclear war. In particular, the criticism that Ronald Reagan's policies were likely to lead to a nuclear war was, though clearly not embraced by a majority, a fairly widespread and mainstream opinion. This comes out in Watchmen and The Dark Knight Returns and "99 Red Balloons" among other places:

Obviously, in retrospect this was wrong. Consequently, from a certain point of view the Nuclear Freeze Movement people and critics of Reagan's "evil empire" rhetoric look foolish. But when you think in terms of probabilities this isn't necessarily right. Say that under Jimmy Carter's policies there would have been a one percent chance of a nuclear war in the 1981-84 period whereas under Reagan's more aggressive policies there was a three percent chance of such a war. That still gives a very good chance that Reaganism will work out -- 97 percent is good odds -- but still probably means that the Reagan option is a bad idea. But then of course things will probably turn out okay, making the skeptics look foolish, and perhaps unduly biasing future policymaking toward aggressive options.

Comments

But sound policymaking would consider a one percent chance of a scenario in which billions die to be something worth worrying about.

This is exactly Dick Cheney's One Percent Doctrine.

It's odd to see Matthew endorsing Dick Cheney so thoroughly.

Posted by Al | July 31, 2008 11:10 AM

I just started re-reading the Watchmen too!

Posted by Seth Pearce | July 31, 2008 11:13 AM

FWIW, I amazingly thought MY's post was odd for the same reason that Al did.

Posted by DAS | July 31, 2008 11:16 AM

The eighties were a weird time to become an adult. The day-to-day expectation, before Thatcher dragged Reagan kicking and screaming to Gorbachev, was that we were all gonna get nuked any day, and they probably wouldn't bother to tell us.

My dad worked at the VA and I grew up around VFW types (boy, those guys can drink), and their foreign policy outlook was basically that Carter was a pussy for not nuking Iran, but that Reagan was a thrashing retard who was making a huge mess.

Government employees. Go figure.
.

Posted by Grand Moff Texan | July 31, 2008 11:17 AM

But then of course things will probably turn out okay, making the skeptics look foolish, and perhaps unduly biasing future policymaking toward aggressive options.

Of course, the key difference is that MY is referring to the cost of "aggressive options" rather than of being what used to be called conservative. Cheney's 1% doctrine is that if there is a 1% chance of gobs of people dying from inaction, we should act ... oddly for a so-called conservative, Cheney doesn't seem to consider the cost of action.

Posted by DAS | July 31, 2008 11:18 AM

You're an idiot Al, and so is Dick Cheney. Cheney's belief was that we should invade Iraq, thereby destabilizing the entire ME in a way that raised, not lowered, the likelihood of future catastrophe.

But then of course things will probably turn out okay, making the skeptics look foolish, and perhaps unduly biasing future policymaking toward aggressive options.

This is basically why the human race is doomed. No way we can handle lots of choices where we have to trade off the pleasures of pounding our chests and satisfying our egotism against low likelihoods of total catastrophe.

Posted by MQ | July 31, 2008 11:19 AM

This is exactly Dick Cheney's One Percent Doctrine.

No, it's not. Dick Cheney's doctrine is, if there is a 1% chance of something happening, we should act as if it will definitely happen.

Matt's statement is, if there is a 1% chance something will happen, we should act as if it could happen.

It's the difference between buying homeowners insurance in case your house burns down, and buying a back-up house.

Posted by joe from Lowell | July 31, 2008 11:19 AM

This was exactly what Bush and his staff expected to happen with their Mid East Policy.

Posted by TDE | July 31, 2008 11:20 AM

Except Cheney's threats never rose to the level of extinguishing life as we know it.

This sort of thing is very familiar to me--I'm a property insurance underwriter. A 1% chance of something happening is a big deal! You don't mentally file those threats away, you charge for them so, over the long term and over your entire book of business, when the big one comes you have enough in the bank to stay afloat. Maybe you don't make any money that year, but over the course of the decade you're profitable.

The market's pretty soft right now. That means there's a lot of competition out there. The people ignoring climate change remind me a lot of the competition who are giving away big limits for rock bottom prices on stuff in CA and FL. Will they probably be ok? Sure, there haven't been any biggies for a couple years--but if anything happens, they're gonna be out of business. Hopefully the human race stays "solvent"

Posted by Jack | July 31, 2008 11:21 AM

This is exactly Dick Cheney's One Percent Doctrine.

No, it's the inverse of the 1% doctrine.

Cheney: if a threat has a 1% chance of happening, we need to move against it.

Yglesias: if there's 1% chance of a policy failing, those are good odds.

Matt didn't say that Reagan pursued escalation because there was a 1% - 3% threat of Soviet attack. Matt said that Reagan's policy had a certain percentage chance of failure.

Truly, you are dumb enough to be a conservative.
.

Posted by Grand Moff Texan | July 31, 2008 11:21 AM

1)Reagan and the Pentagon knew that the Soviet Union was on the verge of serious economic collapse -- and that the historic fix for such a problem has been to start a war. Hence, the need to deter the Kremlin from following that option.

2) The nuclear freeze movement was ignorant of this concern within the US intel/military community.

Posted by Don Williams | July 31, 2008 11:22 AM

I'm skittish here. What we know for sure about Reagan is that there is a 0% experimental probability of nuclear war during his presidency. It feels a bit like grasping to indict his policy on the grounds that it could have failed, even though it didn't.

Posted by Tim Dees | July 31, 2008 11:25 AM

Obviously, in retrospect this was wrong.Consequently, from a certain point of view the Nuclear Freeze Movement people and critics of Reagan's "evil empire" rhetoric look foolish.

I don't think you even need get involved in the type hypothesizing proposed here to see that Reagan's policy were foolish. He lead a military build up, while the Russians were looking for ways to wind down the cold war. The standing conservative line about Reagan winning the cold war is primarily based on fiction. The entire conception of the cold war is based on the faulty premise that every positive development was a response to our actions. It would only be accurate to say his policies were so disastrous as have prevented the soviets collapse.

Posted by Christopher Colaninno | July 31, 2008 11:27 AM

Watchmen sucks. So did Reagan's nuclear policy.

Posted by Freddie | July 31, 2008 11:29 AM

and that the historic fix for such a problem has been to start a war. Hence, the need to deter the Kremlin from following that option.

Analogy fail.

Eisenhower had already handled that.

If Eisenhower was intent on containing Soviet military and political power without imposing an enormous economic burden on the United States and the rest of the free world, he did so by posturing possible nuclear escalation. This strategy may have contributed to the ultimate western victory in the Cold War.

Furthermore, "historic fix" doesn't apply when you're dealing with apocalyptic weapons, because there was zero historical precedent for that.
.

Posted by Grand Moff Texan | July 31, 2008 11:29 AM

Say that under Jimmy Carter's policies there would have been a one percent chance of a nuclear war in the 1981-84 period whereas under Reagan's more aggressive policies there was a three percent chance of such a war. That still gives a very good chance that Reaganism will work out -- 97 percent is good odds -- but still probably means that the Reagan option is a bad idea. But then of course things will probably turn out okay, making the skeptics look foolish, and perhaps unduly biasing future policymaking toward aggressive options.

Of course, you have to apply this to the flip side of the calculus as well, right? That is, presumably Reagan didn't adopt "more aggressive policies" just for the sake of being aggressive, or machismo, or what have you. It was because he thought that this would deter Soviet expansion or an attack on, say, Western Europe. So, if under Carter's policies there was a 3% chance of Soviet attack, but under Reagan's there was a 1%...well,you get the idea.

Now whether in fact Reagan's policies actually changed the probabilities this way is questionable, but then again, that same skepticism applies to Matt's hypothetical increase in the probability of nuclear war.

Posted by Glenn | July 31, 2008 11:30 AM

Say that under Jimmy Carter's policies there would have been a one percent chance of a nuclear war in the 1981-84 period whereas under Reagan's more aggressive policies there was a three percent chance of such a war. That still gives a very good chance that Reaganism will work out -- 97 percent is good odds -- but still probably means that the Reagan option is a bad idea. But then of course things will probably turn out okay, making the skeptics look foolish, and perhaps unduly biasing future policymaking toward aggressive options.

Of course, you have to apply this to the flip side of the calculus as well, right? That is, presumably Reagan didn't adopt "more aggressive policies" just for the sake of being aggressive, or machismo, or what have you. It was because he thought that this would deter Soviet expansion or an attack on, say, Western Europe. So, if under Carter's policies there was a 3% chance of Soviet attack, but under Reagan's there was a 1%...well,you get the idea.

Now whether in fact Reagan's policies actually changed the probabilities this way is questionable, but then again, that same skepticism applies to Matt's hypothetical increase in the probability of nuclear war.

Posted by Glenn | July 31, 2008 11:31 AM

I think you could make an equally plausible argument saying that another 4 years of Jimmy Carter could've lead to more weak responses to Soviet aggression, more Soviet miscalculation, and a higher chance of nuclear war (while still very low-say from 1% to 2%).
Regardless, it's pretty silly to criticize Reagan for saying his policies COULD have led to nuclear war. They didn't, and in fact the Cold War ended soon after his presidency. I don't see a plausible scenario as to how it could've been wound up even faster with 4 more years of Carter.

Posted by jamie | July 31, 2008 11:33 AM

I don't think you even need get involved in the type hypothesizing proposed here to see that Reagan's policy were foolish. He lead a military build up, while the Russians were looking for ways to wind down the cold war. The standing conservative line about Reagan winning the cold war is primarily based on fiction.

A fiction that helped launder tax dollars into private profits under the pretext of a superior and threatening Soviet Union, a fiction propped up (when necessary) by Team B.

It wasn't foolish, Freddie. It was profitable and politically useful. Anything that lets you embezzle billions while portraying your opponents as weak is a pretty good business model.
.

Posted by Grand Moff Texan | July 31, 2008 11:33 AM

No, Al, you evil insect. Matthew is saying that the one percent chance should be considered in policy making - Cheney is saying that (for only certain types of potential harm) we should act as if the 1% chance was certain to occur. That's not the same thing.,

Don't you have cancer yet Al? I've certainly prayed very hard for that to happen. Why don't you just put a bullet in your "brain." It might help, and certainly couldn't hurt.

Posted by voice of reason | July 31, 2008 11:36 AM

Sorry, that was addressed to Christopher, not Freddie.

I haven't read Watchmen, so I can't evaluate Freddie's take on it.
.

Posted by Grand Moff Texan | July 31, 2008 11:37 AM

I think that this post suggests that a more in depth discussion of Watchmen's politics would be extremely interesting. I consider them to be one of the most fascinating aspects of the book, as well as one that the upcoming adpatation is likely to get wrong.

Posted by Michael T Sweeney | July 31, 2008 11:38 AM

I think you could make an equally plausible argument saying that another 4 years of Jimmy Carter could've lead to more weak responses to Soviet aggression, more Soviet miscalculation, and a higher chance of nuclear war (while still very low-say from 1% to 2%).
Regardless, it's pretty silly to criticize Reagan for saying his policies COULD have led to nuclear war. They didn't, and in fact the Cold War ended soon after his presidency. I don't see a plausible scenario as to how it could've been wound up even faster with 4 more years of Carter.

Posted by jamie | July 31, 2008 11:39 AM

Ya sure, and if you drive drunk you probably won't kill anyone either.

Posted by Aatos | July 31, 2008 11:40 AM

Re Texan's comment "Eisenhower had already handled that" and "when you're dealing with apocalyptic weapons"

-------------
The US was in a much stronger position relative to the USSR in Eisenhower's administration than in the early years of Reagan's administration.

Nuclear weapons are not "apocalyptic" except in an all-out exchange in which both sides throw everything they have.

The USA and USSR would have survived a nuclear war if less than , say, 100 ICBMs were launched -- although significant percentages of the population in both countries would have died. But Russia endured the loss of millions in World War II.

Posted by Don Williams | July 31, 2008 11:40 AM

Forget not the 7 Seconds cover of 99 Red Balloons.

I saw them at the Reseda Country Club with the Adolescents and some straight edge band. I remember more Valley toughs than nuclear anxiety.

On the other hand we weren't allowed to watch the Day After when it was on TV (although slasher flicks were never prohibited - go figure).

Posted by Linus | July 31, 2008 11:42 AM

"No, it's not. Dick Cheney's doctrine is, if there is a 1% chance of something happening, we should act as if it will definitely happen."

Cheney's doctrine is that if there's a 1% chance of something happening, then we should use that as an excuse for doing whatever the hell we want to. Meanwhile there may be other diastrous possibilities with a 2% or 3% probability (e.g. loose nukes in the USSR getting in the wrong hands; global climate change) which don't match our preconceived agenda, just ignore them completely.

Meanwhile, bomb your enemies, torture the innocent, and just to be consistent, shoot your friends in the face.

Posted by Richard Cownie | July 31, 2008 11:42 AM

This is exactly Dick Cheney's One Percent Doctrine.

People are right to suggest there's a big difference between factoring in unlikely scenarios and Cheney's suggestion that unlikely threat must be moved against.

However, the more important issue is that Dick Cheney's one percent doctine involved Dick Cheney as a decision maker.

Dick Cheney isn't interested in honest evaluation of threats. Dick Cheney isn't interested in accurately presenting evidence on threats to the public. Dick Cheney isn't interested in applying his doctrine in every situation. He's interested in coercing the intelligence community into changing their results, he interested in leaking stories to newspapers and then using those leaks as evidence, he's interested in all manner lie and exaggeration to see that his policies go through.

Nor is the media or the Democrats in Congress remotely prepared to combat his tactics.

To accept the Dick 1% doctrine is to Dick Cheney's opinions and integrity as an judge of threats to our nation.

Even in the absence of Cheney as an actor we would still expected to believe that it would be a good idea to launch wars if you could talk Congress into believing their is 1% chance an occurrence. There is a long history of Congress approving wars based on highly spurious evidence already. Further lowering the bar would be an awful idea.

Posted by Christopher Colaninno | July 31, 2008 11:46 AM

The USA and USSR would have survived a nuclear war if less than , say, 100 ICBMs were launched -- although significant percentages of the population in both countries would have died.

Not only do I see no reason why this would have been the most likely scenario, I also don't see how the Soviet Union was likely to consider such a scenario as an obvious economic fix.

And yes, I've read "Red Storm Rising."

My point was that your historical analogy didn't apply to the situation in question. If anything, the Reagan administration was in denial about the difficulties of the Soviet Union. The "Reagan Shift," iirc, was the result of a near-launch, an accident, not the sudden realization that the Soviet Union was somehow weaker and therefore a greater threat.
.

Posted by Grand Moff Texan | July 31, 2008 11:47 AM

"Nuclear weapons are not "apocalyptic" except in an all-out exchange in which both sides throw everything they have."

Wrong, Recent research suggests that even a "small" regional conflict in which two small nuclear powers
(e.g. India and Pakistan) detonated about 100 nukes would very probably put enough dust and smoke into the atmosphere to cause severe global consequences, e.g. years of crop failures. Though personally, in my book just a couple of megatons on big cities, causing 10M deaths, while arguably not as bad as WW1 or WW2, would be quite "apocalyptic".

You might like to read one or two accounts of survivors of Hiroshima, and use any imagination you've got to scale that up from a 20kton blast to a 25x greater 500kton. You could also consider the aftermath of the 19th-century Krakatoa volcano eruption, which put a similarly large quantity of dust high in the atmosphere and caused worldwide climate disruption.

Posted by Richard Cownie | July 31, 2008 11:51 AM

The basic problem is, is what is the actual % rate of something happening. 'Say 3%' is not a good enough estimate. Unfortunately people tend to adjust their estimates of something happening to higher levels depending of possible risk involved. (That is, if somebody is really really afraid of a nuclear war they tend to increase their estimates of it actually occurring versus someone who isn't worried (or likes the action they're getting) and thus tend to adjust the chances downwards. (Which is where you get the excess of black swans comes from in the financial industry.)

The chance of a nuclear war occurring was (and is) actually incomputable, since it depends upon the decisions of a small number of unpredictable human actors.

On the other hand, the global warming thing is very much different, and unfortunately as far as I can tell, estimating a 1% chance of 20 centigrade increase in temperature (with the current doubling of CO2 from nominal 'natural' levels) is pretty much a laffer. A 1% chance of a 2 centigrade increase sounds about right, although maybe a little on the low side.

max
['Or another words, whee the fuck did that come from?']

Posted by max | July 31, 2008 11:52 AM

1) What seemed strange to me during the Reagan administration -- and seems strange still -- is the Mainstream News Media's SILENCE on the discrepancy between US military spending and the spending of our allies.

2) Both Germany and Japan were mere miles from the "evil empire" and yet spent relatively nothing on the military in the 1980s.

Japan, the world'second largest economy at the time, spent less than 1 percent of GDP on defense. Germany, the world's third largest economy, spent less than 3 percent.

Meanwhile, Reagan was piling up a huge debt by spending 7.5 percent of US GDP. How was this in the national interest? Why did this major and obvious FACT never come up in the years of talks show debates over Reagan's military acts.

3) It makes sense only if you assume that our plutocratic elites --who own and totally control the news -- were looking ahead and planning for the global empire they would build after the Communist Soviet Union fell and reduced the risk to foreign direct investment.

Posted by Don Williams | July 31, 2008 11:57 AM

"Regardless, it's pretty silly to criticize Reagan for saying his policies COULD have led to nuclear war. They didn't, and in fact the Cold War ended soon after his presidency"

So if the day you retire you cash in your 401K, fly to Vegas, and bet the whole lot on red, is that wise or foolish ? Or can you not decide until you see where the ball stops ?

Reagan had bad policies and good luck. Not least the good luck to be soon enough after Watergate that everyone was reluctant to impeach him for Iran/Contra.

Posted by Richard Cownie | July 31, 2008 11:58 AM

Let's not forget that one of the reasons that summit came about at all was that the Reagan administration became more serious about talking to the Soviets after the war scare of 1983. Reagan's policies (and the Kremlin's reaction) did increase the chances of nuclear war rather more than the public was aware at the time.

https://www.cia.gov/library/center-for-the-study-of-intelligence/csi-publications/books-and-monographs/a-cold-war-conundrum/source.htm

Posted by Jeff | July 31, 2008 12:01 PM

"... That still gives a very good chance that Reaganism will work out -- 97 percent is good odds -- but still probably means that the Reagan option is a bad idea. ..."

Not necessarily if it reduced the chances of nuclear war in the future.

Posted by James B. Shearer | July 31, 2008 12:02 PM

-By an overwhelming 96 percent to 3 percent, Americans assert that "picking a fight with the Soviet Union is too dangerous in a nuclear world. . . ."

http://www.foreignaffairs.org/19840901faessay8394/daniel-yankelovich-john-doble/nuclear-weapons-and-the-u-s-s-r-the-public-mood.html

The debate would be whether or not Reagan's policies constituted "picking a fight." The Nuclear Freeze movement was about decreasing the relative likelihood of nuclear war and did not stake its arguments on a guarantee that Reagan's actions would lead to nuclear war, but rather that his actions increased the likelihood more than different actions would.

I don't know how you can say that in retrospect Nuclear Freeze was wrong. The point of view from which the Nuclear Freeze movement was/is foolish isn't particularly dependant on facts on the ground. If Reagan's policies had resulted in nuclear strikes/war, the people who consider Nuclear Freeze foolish wouldn't blame Reagan, they'd blame the Freeze movement for making America look weak.

Posted by Sophia | July 31, 2008 12:04 PM

"2) Both Germany and Japan were mere miles from the "evil empire" and yet spent relatively nothing on the military in the 1980s.

Japan, the world'second largest economy at the time, spent less than 1 percent of GDP on defense. Germany, the world's third largest economy, spent less than 3 percent.

Meanwhile, Reagan was piling up a huge debt by spending 7.5 percent of US GDP. How was this in the national interest? Why did this major and obvious FACT never come up in the years of talks show debates over Reagan's military acts."

I thought the fact that the U.S. didn't want other countries spending a lot on their own militaries, so our leadership in military matters would be unassailable, and that other countries didn't want to spend a lot on their own militaries, relying on the U.S. to carry the burden so they didn't have to, was widely understood.

Mike

Posted by MBunge | July 31, 2008 12:11 PM

Re Cownie's comment "Wrong, Recent research suggests that even a "small" regional conflict in which two small nuclear powers
(e.g. India and Pakistan) detonated about 100 nukes would very probably put enough dust and smoke into the atmosphere to cause severe global consequences, e.g. years of crop failures."
----------
It depends upon how the war is waged. Usually, attacks on cities, military bases,etc are done as AIR Bursts -- because such bursts distribute greater blast over a greater distance than do ground bursts. (The shock wave traveling downward from 6 o'clock at the air burst bounces off the ground and reinforces the part of the shock wave traveling more diagonally (2 to 4 o'clock) from the detonation. See Glasstone's "Effects of Nuclear Weapons".

Air bursts generate much less dust and fallout etc than do ground bursts.

Ground bursts tend to be used against buried military command posts or missile silos.

Posted by Don Williams | July 31, 2008 12:11 PM

There is a fundamental problem with dealing with uncertainty, as opposed to risk. As I'm sure many people have read elsewhere, risk is when you know the probabilities of various possible outcomes, and uncertainty is when you don't even know the probabilities.

The problem is this: generally speaking, climate models have a great deal of uncertainty - we don't know the relevant probabilities. It is natural to think, as Krugman points out (referencing Weitzman) that policy should take into account the fat tails. Fat tails are when there are outrageously bad outcomes that are possible but unlikely.

But when you mix this decision procedure with high uncertainty, you get nutty results (unless you take another dubious step, which I'll mention in a second). For example, it is highly unlikely that Earth will be invaded by giant flesh-eating hamsters from outer space. We don't know the relevant probabilities, because SETI has been chronically underfunded. So as long as there is a non-zero probability that we'll be attacked by the Hamster Space Army, we should take measures to bolster our Hamster defense.

The dubious step you can take to avoid this result is to disregard the uncertainty, and assign probability zero to things like the Hamster Space Army invasion. This doesn't sound so dubious in this particular case, but *the whole point of uncertainty is that you lack empirical grounds to assign probabilities.* So in any case that there is indeed uncertainty, to assign a situation probability zero because it sounds wacky is making an empirically unjustifiable, a priori judgment. And deciding which cases it's ok to do that relies on your notion of what sounds wacky, which is hardly scientific.

For the record, I'm NOT a global warming skeptic. I'm just a humble philosopher arguing against a "in the face of uncertainty let's assign non-zero probability to some but not all worst-case scenarios" decision procedure.

Posted by Matthew Non-Yglesias | July 31, 2008 12:13 PM

Re Mike's comment "I thought the fact that the U.S. didn't want other countries spending a lot on their own militaries, so our leadership in military matters would be unassailable, and that other countries didn't want to spend a lot on their own militaries, relying on the U.S. to carry the burden so they didn't have to, was widely understood. "
-----------
What is the point of having "leadership in military matters" is you can't extort huge sums from your clients ..er .. allies?

Have you seen Japan expressing it's gratitude by giving back the hundreds of billions of Reagan IOUS (US Treasury Bonds ) it holds and collects interest on?

Posted by Don Williams | July 31, 2008 12:17 PM

Correction: "What is the point ... is you can't extort" should read
"What is the point ... IF you can't extort"

Posted by Don Williams | July 31, 2008 12:21 PM

"I don't know how you can say that in retrospect Nuclear Freeze was wrong."

Because, in retrospect, things worked out just fine by NOT doing what the Nuclear Freeze folks wanted. Given that what actually happened, I think it's hard to come up with a credible alternative history where adopting the Nuclear Freeze concept would have produced a better result.

First, you have to assume that the Soviet Union would have abided by any Freeze agreement and not used a Freeze by the West as an opportunity to improve it's military position.

Second, you have to assume that an effective Nuclear Freeze would either have had no influence on the dynamic that led to Gorby coming to power and ushering in the end of the Soviet Union, or that a Freeze would have produced an alternative dynamic that would have resulted in the same or a better result. Because having no Soviet Union at all is a hell of a lot better, atomic war wise, than having a Nuclear Freeze and a Soviet Union that persists.

Mike

Posted by MBunge | July 31, 2008 12:25 PM

The 1% risk sounds vaguely Cheney-esque.

The error of reasoning, though, is better described as a tendency to treat policy-making as far more discrete than it is. Policy-making is not like buying a lottery ticket and waiting for a result.

It is more like choosing a path. Sure, a particular path might lead over a sheer cliff. The critical question, though, is whether anyone will notice the cliff prior to the plunge and change policy accordingly.

Bad policy, as demonstrated by Bush in Iraq, requires treating the cliff as a diving platform.

Really bad policy, in practice, is not a matter of taking risks; it is a matter of not modeling processes correctly, and ignoring the feedback. It is a matter of steering toward catastrophe, and then compounding disaster, by failing to acknowledge it.

Posted by Bruce Wilder | July 31, 2008 12:27 PM

"Air bursts generate much less dust and fallout etc than do ground bursts."

Less radioactive fallout, for sure. Less dust: not necessarily, if the target is something like a city
which contains a lot of flammable material.
The heat from the air burst is going to start fires everywhere, exacerbated by the shockwave breaking things up and spreading flammable material all over the place. So then you get a firestorm, with associated high winds and lots of smoke being sent high in the atmosphere. And burning millions of tons of fuel releases more energy than the original blast.

Unglamorous, I know. You could do the same thing with lots and lots of 1940s low-tech incendiary bombs, as at Dresden. But really terrible.
Quite precisely "apocalyptic".

Posted by Richard Cownie | July 31, 2008 12:29 PM

Under what possible scenarios would global warming kill billions of people? The most pessimistic assessments I've seen involve gradual flooding over the course of decades, along with more severe bad weather. It's hard to imagine those trends killing on he order of a billion people

Posted by Tim Lee | July 31, 2008 12:33 PM

"Under what possible scenarios would global warming kill billions of people?"

Rapid changes in rainfall and/or temperature cause severe crop failure for several seasons. Then people starve.

And here's a link and some quotes about the
consequences of "limited" nuclear war:

http://www.wunderground.com/education/nuke.asp

The black smoke, they found, absorbed far more solar radiation than the brighter sulfuric acid particles emitted by volcanic eruptions. This allowed the smoke to heat the surrounding air to much higher temperatures, resulting in stronger upward motion of the smoke particles higher into the stratosphere. Once the smoke reached the stratosphere, where there is no rain to rain out the soot particles, it stayed at significantly high levels for over a decade. The black soot blocked sunlight, resulting in global cooling of over 1.2° C (2.2° F) for two years, and 0.5° C (0.9° F) for more than a decade.

This magnitude of this cooling would bring about the coldest temperatures observed on the globe in over 1000 years (Figure 1). The growing season would shorten by 10-30 days over much of the globe, resulting in widespread crop failures. The effects would be similar to what happened after the greatest volcanic eruption in historic times, the 1815 Tambora eruption in Indonesia. This cooling from this eruption triggered the infamous Year Without a Summer in 1816 in the Northern Hemisphere, when killing frosts disrupted agriculture every month of the summer in New England, creating terrible hardship. Exceptionally cold and wet weather in Europe triggered widespread harvest failures, resulting in famine and economic collapse. However, the cooling effect of this eruption only lasted about a year. Cooling from a limited nuclear exchange would create two to three consecutive "Years Without a Summer", and over a decade of significantly reduced crop yields. The authors anticipated that the smoke in the stratosphere would partially destroy Earth's protective stratospheric ozone layer as well, but did not model how large of an impact this would have. Clearly, even a limited nuclear exchange could trigger severe global climate change capable of causing economic chaos and widespread starvation.

Posted by Richard Cownie | July 31, 2008 12:39 PM

The nuclear freeze movement was wrong for the simple reason that there was a zero chance that the Soviets would ever have paid attention to them - and unilateral disarmament in the face of evil is a very bad idea.

So at this point in time, what do you think the odds are that China will pay attention to Al Gore? For that matter, the Europeans are ignoring Kyoto and building lots of coal plants - even the supposed fans know it's bunk.

The goal, clear to everyone who isn't Matt, is to slow the US economy down and make enviros feel good.

Posted by James Robertson | July 31, 2008 12:44 PM

Krugman's argument is correct (that the economic analysis of climate change may be dominated by low probability but catastrophic outcomes), but the fundamental problem is that we don't have any idea just how low the probability of catastrophic outcomes really is. We don't know if it's 1% or 0.001%. Nordhaus goes into this issue in his analysis.

Posted by Mixner | July 31, 2008 12:50 PM

You don't have to look back to the '80s for an example of this calculus at work. Look at the current situation with avian influenza - the H5N1 virus. The threat of a high fatality rate novel influenza virus that passes easily among humans is above 0.

Different people would reckon the risk differently but virology experts generally put it in the 30 to 50% range when asked the likelihood of H5N1 going to stage 6 in the next three or five years. Coupled with the impact on our lives, on all our lives, this would pencil out to be a high priority for any administration.

The policies that should be in place - have costs certainly but they also have ancillary benefits. Things like allowing individuals to stockpile Tamiflu by making it available over the counter have no federal costs. Clear communication of people's adult responsibility for personal preparation could make a huge difference in the event of an event.

There is actually a long list of things a President could do now to improve the situation in two years if there then was a novel, deadly virus easily passing among people. What is the threshold for action? If it is less than 10% I think we've been there for some time.

Posted by Billp | July 31, 2008 12:55 PM

Re Cownie's comment "So then you get a firestorm, with associated high winds and lots of smoke being sent high in the atmosphere.... You could do the same thing with lots and lots of 1940s low-tech incendiary bombs, as at Dresden.
---------

Curtis Lemay firebombed the shit out of roughly 68 Japanese cities --which were exceptionally flammable because of Japanese construction. The firestorms matched the destruction of Nagasaki and Hiroshima. The British firebombed the shit out of several German cities as well.

But I don't recall any "global cooling" effects/ crop failures seen from WWII.

I think that to get the "nuclear winter" cooling effects mentioned you have to get a 100 or so nuclear GROUND bursts that throw the smoke/dust high enough into the atmosphere that it doesn't rapidly rain out. I don't think Ordinary fires and or even firestorms have the necessary energy to throw smoke that high.

Posted by Don Williams | July 31, 2008 12:58 PM

My God - I certainly hope Heads in the Sand doesn't consist of policy insights gleamed from the Watchmen, the Dark Knight, and 99 Luftballoonen - we really are in trouble.

Not to mention the lessons learned from all of the scads of popular culture items from the 80s that never once addressed Reagan's policies and the threat of nuclear war. Unless you're saying that the catfight between Alexis and Krystal was a metaphor...

Posted by Ethel-to-Tilly | July 31, 2008 1:02 PM

"I don't think Ordinary fires and or even firestorms have the necessary energy to throw smoke that high"

Whatever. You can use google and read the peer-reviewed research on this topic and find out for yourself whether your concern has any validity. It seems weak to me: you can easily exceed the energy of a nuclear explosion by burning a lot of stuff. All it takes is wood and air and time.

There seems to be a pretty clear trend that over the last 25 years as scientists have had access to better computers and better climate models, the predictions about possible effects of disruptions - nuclear or non-nuclear - have got more and more alarming. The climate and ecosystems are fragile.

Posted by Richard Cownie | July 31, 2008 1:12 PM

But sound policymaking would consider a one percent chance of a scenario in which billions die to be something worth worrying about.


This is exactly Dick Cheney's One Percent Doctrine.

It's odd to see Matthew endorsing Dick Cheney so thoroughly.

I wonder why Cheney doesn't? There's a 95% chance that global climate will stick within the IPCC range. There's a 2.5% chance, temps won't rise as much. And there's a 2.5% chance for horrifying disaster.

So, Al, why don't you go do what Mr. Goodbar Sez?

Posted by Jeffrey Davis | July 31, 2008 1:24 PM

Cownie:
I won't continue to beat on this topic -- here's some closing info:

From http://www.princeton.edu/~globsec/publications/effects/effects2.pdf , Fig 2.16, page 34:

You can see that a 10 Kt nuke fireball only carries a cloud up to 19,000 feet, a 100 Kt to 40,000 feet, a 1 Megaton nuke fireball can carry a cloud up to 65,000 feet and a 10 megaton to about 100,000 feet.

The stratosphere (area of stable air with little precipitation rainout effect ) doesn't begins until around 25,000 to 35,000 feet (depends on latitude). Most modern day nukes are 250 Kt or less because of MIRV's greater efficiency.

Smoke from non-nuclear fires doesn't travel as high because it disperses/cools in the lower atmosphere and is not lifted upward by the buoyency of a superheated nuclear fireball.

Posted by Don Williams | July 31, 2008 1:25 PM

But I don't recall any "global cooling" effects/ crop failures seen from WWII.

http://www.solarpowerfor.us/images/Instrumental_Temperature_Record.png

Posted by Jeffrey Davis | July 31, 2008 1:31 PM

The goal, clear to everyone who isn't Matt, is to slow the US economy down and make enviros feel good.

Flee! All is discovered!
.

Posted by Grand Moff Texan | July 31, 2008 1:36 PM

Psychologically, if there's only a one or two percent change of something happening, we tend to put it in the "not going to happen" file.

Please provide any evidence for the use of the phrase "tend to" - rather than, say, the word "sometimes".

Humans are miserably bad at assessing the relative risks of unlikely events. There are countless examples of humans putting WAAAAAY too much emphasis on low percentage (but high impact) events. Responses to 9/11 is a great example: yes, let's ban water bottles, but not take very basic steps on baggage matching. Let's wage a war against a country that had nothing to do with 9/11 - but which might acquire WMDs someday and give them to terrorists - while we ignore port security.

So too: people - in general - feel great about driving but scared that planes will crash, even though driving is far, far more dangerous. And then there's the Freakonomics point that a pool in your backyard is something like 100 times more likely to result in a child's death than a gun in the house; but the hysteria is around guns, not pools.

While the overall point is good - that we need to properly account for low prob, high impact events - the premise of your point is that we need to pay more attention to those events. In fact, much evidence suggests the opposite, that we overvalue those events.

Posted by Anon | July 31, 2008 1:37 PM

"From http://www.princeton.edu/~globsec/publications/effects/effects2.pdf , Fig 2.16, page 34:"

Published 1977. That's 31 years ago. There's a lot of progress in climate modelling since then.

"Smoke from non-nuclear fires doesn't travel as high because it disperses/cools in the lower"

Firestorms are not just ordinary fires. They generate their own wind system. And the dispersion/cooling of the updraft air is going to be less effective on a really large (wide and fast) updraft. Anyhow, I'll trust the guys who have been researched this in the last couple of years, with modern CFD codes and modern computers. And they say it's bad.

On the question of global climate effects from the firebombing of Japan, there are some data which show a postwar drop in average sea temperature. But alternative explanations have been proposed. I think it's possible that there *was* an effect on global climate, but it's probably hard to disentangle from all the other political and economic chaos that was happening in 1945-1948.
It doesn't make me feel comfortable about giving the atmosphere a similar shock in future ...

Posted by Richard Cownie | July 31, 2008 1:41 PM

"Say that under Jimmy Carter's policies there would have been a one percent chance of a nuclear war in the 1981-84 period whereas under Reagan's more aggressive policies there was a three percent chance of such a war. That still gives a very good chance that Reaganism will work out -- 97 percent is good odds -- but still probably means that the Reagan option is a bad idea."

The problem is that it isn't at all obvious that we should assume Reagan's policies increased the chance of war. We know that the USSR intitially saw Kennedy as weak, which is why they were willing to try out the Cuban Missiles--putting us as close to nuclear war as at any other time.

Carter was seen as weak. Fair or not, that was the impression. I'm not convinced that it is correct to assume that Reagan's policies were more likely to cause war than Carter's.

Posted by Sebastian | July 31, 2008 2:00 PM

Jeffrey Davis,

And there's a 2.5% chance for horrifying disaster.

No there isn't. You're treating probability estimates that at this point represent little more than educated guesses as if they were empirical facts.

Posted by Mixner | July 31, 2008 2:01 PM

Here's the paper which really goes into the weeds on the regional nuclear war scenario: 15kton weapons with airburst detonation over population centers, then it's the firestorm and convection that does the climate damage:

http://climate.envsci.rutgers.edu/pdf/acp-7-1973-2007.pdf

"However, the fires generate by the explosion take several hours to fully develop. Hence the injection height of the smoke is controlled by the energy release from the burning fuel not from the nuclear explosion. The energy released in a fire initiated by a nuclear explosion is much greater than that released by the explosion itself. ... the energy release from the Hiroshima fires would have been about 3x10^16J. This is more than 1000 times greater than the energy release from the atomic explosion"

Posted by Richard Cownie | July 31, 2008 2:03 PM

Under what possible scenarios would global warming kill billions of people? The most pessimistic assessments I've seen involve gradual flooding over the course of decades, along with more severe bad weather. It's hard to imagine those trends killing on he order of a billion people

Posted by Tim Lee | July 31, 2008 2:12 PM

"On the question of global climate effects from the firebombing of Japan, there are some data which show a postwar drop in average sea temperature."

This was recently dicovered to be a measurment problem--it went from buckets over the side before the war, to engine intake during, to both after.

Posted by Sebastian | July 31, 2008 2:13 PM

"You're treating probability estimates that at this point represent little more than educated guesses as if they were empirical facts."

How could a probability estimate for climate change ever be anything more than an "educated guess", in these terms ? Obviously we can't be *certain* what the climate is going to do until it's done it: so in that sense the "probability" of catastrophe is either 0% or 100%. But we need to have some kind of framework for thinking about these issues, and we don't have 100 other planets to play with to do experiments. This planet is the only one we'cve got, and the climate scientists are doing their very best with all their knowledge and tools to predict what might happen. Their "educated guess" is not just the best information we have; it's the best information we can possibly ever imagine having.

Ignore it at your peril.

Posted by Richard Cownie | July 31, 2008 2:19 PM

Richard Cownie,

How could a probability estimate for climate change ever be anything more than an "educated guess", in these terms ?

In what "terms?" The accuracy of a probability estimate depends on, among other things, the amount and quality of the data on which it is based. The data relating to the magnitude and effects of future climate change is still very limited, so the probability estimates for various outcomes are still very poor. This is especially true for estimates relating to extreme outcomes.

Their "educated guess" is not just the best information we have; it's the best information we can possibly ever imagine having.

Nonsense. If there's one thing climate scientists stress over and over again it's the need for more and better data. No serious climate scientist would argue that current information is "the best information we can possibly ever imagine having."

Posted by Mixner | July 31, 2008 2:36 PM

Re Cownie's comment "Published 1977. That's 31 years ago. There's a lot of progress in climate modelling since then."
------------
A hilarious remark --given that the "recent study" you then cite 3 posts further downward has extensive references to my 31 year old source (Glasstone) whenever your study needs actual empirical data to bulwark its airy mathematical handwaving.

I cited Glasstone because we don't detonate ground burst 10 megaton nukes anymore and then actually measure how high the cloud goes.

Ah, for the Good Old Days! The days when nuclear physicists were men and the women were glad of it. Way before the enviros starting wondering why all that radioactive iodine was showing up in childrens' thyroids.

Posted by Don Williams | July 31, 2008 2:44 PM

"The data relating to the magnitude and effects of future climate change is still very limited,"

And it always will be, unless someone invents a time machine. Look, "data" is what you measure from the real world: it's in the past. Then you apply a whole lot of theory and models to extrapolate from that data to make predictions about the future. And that's what you're calling "an educated guess". What could ever possibly happen to make that anything more than an "educated guess" ? When we're dealing with predictions, there are "educated guesses", and there are "uneducated guesses": I'll take the former every time.

Posted by Richard Cownie | July 31, 2008 2:48 PM

"I cited Glasstone because we don't detonate ground burst 10 megaton nukes anymore and then actually measure how high the cloud goes."

Yeah, but Glasstone is completely irrelevant. The question was whether a firestorm could put smoke into the stratosphere. And you cited a paper that only talks about direct effects of nuclear weapons, *not* about what firestorms can do. Which is completely f*cking irrelevant.

Read the damn paper. Forest fires can put smoke into the stratosphere. Recent modelling (and computational fluid dynamics is way better than it was in 1977) indicates that firestorms can do it too.

Heck, the original "nuclear winter" theory was developed in 1982 (and confirmed by later modelling). And you're citing something that's 5 years before even that.

Posted by Richard Cownie | July 31, 2008 2:58 PM

Richard Cownie,

And it always will be,

No, it will likely get much better over time, as scientists collect more data about how the climate changes in response to changes in GHG concentrations, and how other aspects of the environment change in response to changes in climate. This will allow them to develop better theories and models to better predict future changes. With respect to probability estimates regarding catastrophic outcomes, the data is at this point so limited and the models so crude and uncertain, that any actual number is little more than an educated guess. That's why statements like your "there's a 2.5% chance for horrifying disaster" are such nonsense.

Posted by Mixner | July 31, 2008 3:06 PM

"No, it will likely get much better over time, as scientists collect more data about how the climate changes in response to changes in GHG concentrations, and how other aspects of the environment change in response to changes in climate"

Well, sure, if we keep on pumping CO2 into the atmosphere then we'll get more and more knowledge about what happens when there's a lot of CO2. And eventually we'll be 100% certain about what happens when there's a lot of CO2.

But it might be really really bad. In fact the current models suggest that it probably will be quite bad, and that it's worth changing our behavior *right now*, because modest changes right now will be cheaper than drastic changes later.

So really it's too late for "wait and see". The climate scientists aren't saying "wait and see" - they're strongly in favor of curbing C02 emissions. They're also, obviously, strongly in favor of spending more money on climate science.

Posted by Richard Cownie | July 31, 2008 3:17 PM

Re "Read the damm paper. Forest fires can put smoke into the stratosphere. Recent modelling (and computational fluid dynamics is way better than it was in 1977) indicates that firestorms can do it too."
-----------
Actually, your cited paper indicates that most of the smoke would be limited to the troposphere. We have lots of forest fires --and I haven't noticed any "forest fire winter" this summer.

The issue is how much soot would be injected all the way into the stratosphere and whether it would have any noticeble effect on the climate. Your cited paper doesn't really answer that question -- at best, it skirts dangerously close to edge of saying something and then says more studies are needed.

Posted by Don Williams | July 31, 2008 3:20 PM

99 Luftballons (99 Red Balloons) really is better in the original German:

http://www.youtube.com/watch?v=9whehyybLqU

Posted by Dilan Esper | July 31, 2008 3:26 PM

Possibly, you mistakenly cited Toon when you meant to cite Robock's studies?

Posted by Don Williams | July 31, 2008 3:29 PM

"Actually, your cited paper indicates that most of the smoke would be limited to the troposphere. We have lots of forest fires --and I haven't noticed any "forest fire winter" this summer."

Well, you didn't read the paper very carefully. To get stuff that high, the fire needs to have a large radius and a high burn rate. Most forest fires involve a narrow travelling front, and don't really meet the conditions; a few do (providing direct experimental evidence to refute your guess that fires can't do that). Firestorms over cities - especially with multiple simultaneous ignition points, as from an airburst - offer much more favorable conditions. The estimate that the Hiroshima firestorm released 1000 times as much energy as the original bomb should surely indicate that your focus on the location and size of the detonation is missing the point completely (BTW, at Hiroshima the
firestorm didn't develop until 2.75 *hours* after the detonation, another indication that the firestorm convection effects are not very dependent on the details of the detonation event).

Further, in the tropics, once the smoke plume gets high enough it can take advantage of solar heating to rise further. That is a difference from the earlier models of USA-USSR high-latitude nuclear war.

Posted by Richard Cownie | July 31, 2008 3:36 PM

I guess the bottom line I take from this discussion is that if Big Oil does cause global warming to occur, it can easily fix the problem by provoking Al Qaeda into setting off 50 small nukes-- to adjust the planetary thermostat with a small nuclear winter effect.

Posted by Don Williams | July 31, 2008 3:37 PM

Richard Cownie,

Well, sure, if we keep on pumping CO2 into the atmosphere then we'll get more and more knowledge about what happens when there's a lot of CO2.

We'll get more and more knowledge about what happens when there's a lot of CO2 the more we study it, period. We'll get more and more knowledge from further study even if we stopped emitting any more CO2 tomorrow.

But it might be really really bad.

Yes, it might. And there might be a nuclear war next year. There might be a huge asteroid impact. There might be a global disease pandemic. Merely pointing out that a catastrophic event or outcome is possible doesn't tell us anything useful about how to respond to that possibility.

In fact the current models suggest that it probably will be quite bad,

No they don't. Or, at least, if any clear sense is to be given to this assertion, you'll have to explain more clearly what you mean by "quite bad."

and that it's worth changing our behavior *right now*, because modest changes right now will be cheaper than drastic changes later.

The changes proposed by people like Stern and Gore are not "modest."

Posted by Mixner | July 31, 2008 3:39 PM

Re Cownie's comment "The estimate that the Hiroshima firestorm released 1000 times as much energy as the original bomb should surely indicate that your focus on the location and size of the detonation is missing the point completely"
------------
No, it's not. A 1 Megaton nudet injects debris into the stratosphere because an intensely hot fireball carries debris rapidly upward -- similar to a bubble rising in water.

Conventional fires , by contrasting, slowly emit much cooler smoke into the troposphere where it is rapidly and constantly dissipated -- including by rainfall. Kinda like a squid leaking ink into water.

Posted by Don Williams | July 31, 2008 3:48 PM

"We'll get more and more knowledge about what happens when there's a lot of CO2 the more we study it, period. We'll get more and more knowledge from further study even if we stopped emitting any more CO2 tomorrow."

Yes, scientists can continue to evolve and refine their models. The question for you, Mixner, is whether there's ever any criteria by which you might be prepared to admit that a scientific study with a prediction you don't like is actually worth acting on, rather than dismissing as a mere "educated guess".

"Yes, it might. And there might be a nuclear war next year. There might be a huge asteroid impact. There might be a global disease pandemic. Merely pointing out that a catastrophic event or outcome is possible doesn't tell us anything useful about how to respond to that possibility."

Indeed. Which is why scientists try hard to produce quantitative estimates: they don't just say "we might be hit by an asteroid", they say, "within the next 50 years there's an X% probability of an impact by an asteroid of diameter greater than Ykm". And with that estimate - an "educated guess" -you make a rational choice about what action to take and how much to invest in preparing for the possible disaster.

In the particular case of global climate change, we're at the point where detailed policies and targets are being discussed by governments. There are plenty of things we can do, and plenty of estimates of the likely effects. But you prefer to dismiss the whole thing as an "educated guess" and bury your head in the sand. Feeble.

"No they don't. Or, at least, if any clear sense is to be given to this assertion, you'll have to explain more clearly what you mean by "quite bad.""

This link should do.

http://www.marklynas.org/2007/4/23/six-steps-to-hell-summary-of-six-degrees-as-published-in-the-guardian

"Scientists estimate that we have at best 10 years to bring down global carbon emissions if we are to stabilise world temperatures within two degrees of their present levels. The impacts of two degrees warming are bad enough, but far worse is in store if emissions continue to rise. Most importantly, 3C may be the “tipping point” where global warming could run out of control, leaving us powerless to intervene as planetary temperatures soar. The centre of this predicted disaster is the Amazon, where the tropical rainforest, which today extends over millions of square kilometres, would burn down in a firestorm of epic proportions. Computer model projections show worsening droughts making Amazonian trees, which have no evolved resistance to fire, much more susceptible to burning. Once this drying trend passes a critical threshold, any spark could light the firestorm which destroys almost the entire rainforest ecosystem. Once the trees have gone, desert will appear and the carbon released by the forests’ burning will be joined by still more from the world’s soils. This could boost global temperatures by a further 1.5ºC – tippping us straight into the four-degree world.

Three degrees alone would see increasing areas of the planet being rendered essentially uninhabitable by drought and heat. In southern Africa, a huge expanse centred on Botswana could see a remobilisation of old sand dunes, much as is projected to happen earlier in the US west. This would wipe out agriculture and drive tens of millions of climate refugees out of the area. The same situation could also occur in Australia, where most of the continent will now fall outside the belts of regular rainfall.

With extreme weather continuing to bite – hurricanes may increase in power by half a category above today’s top-level Category Five – world food supplies will be critically endangered. This could mean hundreds of millions – or even billions – of refugees moving out from areas of famine and drought in the sub-tropics towards the mid-latitudes. In Pakistan, for example, food supplies will crash as the waters of the Indus decline to a trickle because of the melting of the Karakoram glaciers that form the river’s source. Conflicts may erupt with neighbouring India over water use from dams on Indus tributaries that cross the border.

In northern Europe and the UK, summer drought will alternate with extreme winter flooding as torrential rainstorms sweep in from the Atlantic – perhaps bringing storm surge flooding to vulnerable low-lying coastlines as sea levels continue to rise. Those areas still able to grow crops and feed themselves, however, may become some of the most valuable real estate on the planet, besieged by millions of climate refugees from the south."

Posted by Richard Cownie | July 31, 2008 3:53 PM

"No, it's not. A 1 Megaton nudet injects debris into the stratosphere because an intensely hot fireball carries debris rapidly upward -- similar to a bubble rising in water."

Yeah, but who cares ? The regional-war scenario assumes small (15kton) airbust weapons. It doesn't matter what 1Mton weapons do. The mechanism is different. Volcanic eruptions can also put debris into the stratosphere, but without any such superhot fireball. There's more than one way to do it.

"Conventional fires , by contrasting, slowly emit much cooler smoke into the troposphere where it is rapidly and constantly dissipated -- including by rainfall. Kinda like a squid leaking ink into water."

Yeah, that's your theory. Except a) direct experimental evidence from forest fires shows that they *can* put smoke into the stratosphere under certain conditions, disproving your theory
and b) for the paper I cited they ran models and showed that large firestorms in the tropics can do it. In particular, a big firestorm violates
your "slowly" condition: it burns a very large quantity of fuel at a very high power output
(thousands of megawatts).

The fact that a firestorm can release 1000x more energy than the original detonation should indicate that it's a very unusual phenomenon which defies your simple intuition about "conventional fires".

Posted by Richard Cownie | July 31, 2008 4:04 PM

Re "for the paper I cited they ran models and showed that large firestorms in the tropics can do it"
-----------
I'm confused. In the post to me , you indicate that tropical firestorms could inject soot into the stratosphere --causing global cooling.

In the post immediately preceding that -- to Mixner -- you indicate that a tropical firestorm in the Amazon would cause massive global warming. Which is it?

Or would we have global cooling initially --from the soot -- and then global warming later -- from the carbon dioxide? Or would the two effects cancel out?

Posted by Don Williams | July 31, 2008 4:12 PM

"In the post immediately preceding that -- to Mixner -- you indicate that a tropical firestorm in the Amazon would cause massive global warming. Which is it?"

Well, I was just quoting that piece, and I think he's being loose in talking about an Amazonian "firestorm". I would have to dig up the underlying research to understand in more detail what mechanism is predicted for the destruction of the rainforest. But my uneducated guess would be that drier hotter conditions in that region would cause some species to die off, and a greater prevalence of fires leading to deforestation. Probably not putting soot in the stratosphere, but causing lots of global warming by a) releasing a lot of CO2 from burnt/decayed biomass, and b) reducing the CO2-consuming live biomass of the rainforest.

Anyhow, my impression, particularly from accounts of the aftermath of the 19th-century Tambora and Krakatoa eruptions, is that the climate effects of atmospheric dust tend to last only a few years (though that's enough to cause massive disruption of food supplies and thus great hardship), whereas changes in CO2 level persist for decades or even centuries. On that basis I would expect a few cool years followed by persistently warmer conditions. But doubtless there's active research on this.

Posted by Richard Cownie | July 31, 2008 4:24 PM

Hmm... What goes up must comes down. I guess you could make a rough estimate based on the particle size and weight, and the air conditions, you could predict the terminal velocity for a particle (i.e. the speed at which dragforce balances gravity). And then that tells you how long the particle will take to fall back down.

At first the dust/soot/smoke will be concentrated, and solar heating can make the air around it hotter, and thus create a convection current to keep it up
(or putting it another way, the higher temperature causes the hot air + dust to be less dense than
cold air without dust). But once the dust gets throughly dissipated and spread around, any such heating will be uniform and thus will not cause
convection.

Posted by Richard Cownie | July 31, 2008 5:23 PM

Richard Cownie,

Which is why scientists try hard to produce quantitative estimates: they don't just say "we might be hit by an asteroid", they say, "within the next 50 years there's an X% probability of an impact by an asteroid of diameter greater than Ykm".

Right. Which is why your statement that global warming might be catastrophic is so worthless. Merely listing possibilities doesn't tell us anything useful about the proper policy response.

And with that estimate - an "educated guess" -you make a rational choice about what action to take and how much to invest in preparing for the possible disaster.

You keep pretending that all probability estimates are educated guesses. They're not. As I said, the reliability of some probability estimates is much better than others because our knowledge and understanding of some phenomena is much better than others. With respect to catastrophic outcomes from climate change, our knowledge and understanding of the relevant phenomena is so poor that probability estimates are very, very unreliable. They amount to an educated guess. That is not true of probability estimates regarding outcomes for which our knowledge is much better.

This link should do.

Then your claim is simply false, because there is no scientific consensus that the outcomes you list are "probable." Again, you're simply listing possibilities. You don't even try to assign any probability estimate to each of them, let alone try quantify the confidence level of those estimates.

Posted by Mixner | July 31, 2008 5:26 PM

I vaguely remember hearing from someone the notion that significantly improving the general public's fundamental understanding and appreciation of probability would have enormous effects on how they view most of our most hotly debated political problems. I wish I could remember who said it ... maybe it was Carl Sagan, but I'm not sure. Anyway, this is a good example. And thanks for the Nena video!

Posted by Andrew Pulrang | July 31, 2008 5:49 PM

I vaguely remember hearing from someone the notion that significantly improving the general public's fundamental understanding and appreciation of probability would have enormous effects on how they view most of our most hotly debated political problems. I wish I could remember who said it ... maybe it was Carl Sagan, but I'm not sure. Anyway, this is a good example. And thanks for the Nena video!

Posted by Andrew Pulrang | July 31, 2008 5:53 PM

"Then your claim is simply false, because there is no scientific consensus that the outcomes you list are "probable." Again, you're simply listing possibilities."'

Well, yeah. Which is why you'll need to go do some actual work of googling and reading the peer-reviewed research to get a more meaningful and quantitative estimate. Obviously what I linked to was a summary and review of that research, not the hard-core stuff. But since you yourself never cite anything quantitative or talk about numbers, I doubt you'll go any deeper.

I encourage everyone to study the issue for themselves.

Posted by richard cownie | July 31, 2008 7:03 PM

Richard Cownie,

Which is why you'll need to go do some actual work of googling and reading the peer-reviewed research to get a more meaningful and quantitative estimate.

Sorry, it's not my job to look for evidence for your claims. That's your job. Unless you can produce evidence to support your claims here, we need not take them seriously.

Posted by Mixner | July 31, 2008 7:22 PM

"You keep pretending that all probability estimates are educated guesses. They're not."

My point really is that all probability estimates that happen to have inconvenient consequences can be attacked as mere "educated guesses". Just as you do.

"As I said, the reliability of some probability estimates is much better than others because our knowledge and understanding of some phenomena is much better than others.""

On the whole, the phenomena of which we have good "knowledge and understanding" are those which can either be replicated in the laboratory, or occur repeatedly and frequently in nature. Global climate change doesn't meet those criteria and never will. So you're not going to get very precise probability estimates.

We're in a bus heading towards a cliff: I want us to start putting on the brakes; you think we should calibrate the speedometer to be sure whether we're going at 50mph or 60mph, and try to figure out just how high the cliff is, before taking making any attempt to slow down.

Posted by Richard Cownie | July 31, 2008 7:29 PM

Matt makes another "Duh!" post.

The problem with Matt's logic is that there's ALWAYS a "one percent chance that billions will die." If it isn't nuclear war or global warming, it's some other potential disaster.

Do we worry about asteroid strikes? Yes, we do. More precisely, somebody who knows about it is paid to worry about it. The rest of us don't care because there's nothing we can do about it and according to the experts the odds are so low it's simply pointless to worry about it when the odds of our personal deaths from various other causes are so much higher.

The disconnect with global climate change is that it's impossible to discern at this point either when it will occur, how bad it will be, and especially what technology options will be utilized to adjust the situation one way or the other. Predictions of major changes in the world are more often wrong than right.

Meanwhile, technology options that can be predicted to make major changes in the situation are dismissed by wannabe pundits who have their own axes to grind or who have no imagination to even conceive what the options are.

Here's the bottom line: whatever happens in global warming is going to happen regardless of anything Matt says - because he isn't an influence on the factors involved.

Thirty years ago the predictions agreed to by many experts were that most of the Earth's population would be dead of starvation by now and the rest would be fighting wars over food - and pollution would have destroyed a large part of the environment.

Didn't happen.

Today I'm supposed to just start jumping up and down because the climate might change in 25 or 50 years and sea levels might rise, and agriculture might suffer, and storms might be worse (or not, the experts aren't sure).

Sorry, not interested. When I start to see effects that might actually negatively impact me, I'll concentrate on doing what I can to minimize those impacts on me personally. Until then, it's somebody else's problem. If they do their job, I'll never be impacted. If they don't do their job - which wouldn't surprise me in the least - then I'll deal with it.

Here's the bottom line: however bad it gets, the human race isn't going anywhere - until we Transhumans decide you are. That's what you chimps really should be worrying about, if you want something to worry about.

Posted by Richard Steven Hack | July 31, 2008 7:58 PM

Sorry, it's not my job to look for evidence for your claims. That's your job. Unless you can produce evidence to support your claims here, we need not take them seriously.

Posted by Mixner

Shut up, you sci fi pompous nimrod. No one needs your 'permission' or evaluation to take a subject or claim seriously, nor does anyone else give the slightest damn what you weigh in on as the last word in how much evidence has been presented to you. And it is your f***ing job to go look for evidence for major claims unless you think this is some sort of high school debate league, where intellectual topics are limited to what convinces you.

Is this some sort of return to parody of medievalist argument, whereby if a complete argument is not sat nicely in front of you, and arranged for your comprehension, it doesn't exist?

On second thought, don't answer that. No, really, just don't.

Posted by El Cid | July 31, 2008 7:58 PM

And, RSH: please try to remember, you're not a Transhuman yet. If you want to call yourself a member of the Future Transhumans of America, fine. But stop trying to talk to the rest of us like we're the dumb meat people hearing you broadcast on one of your many cybernetic frequencies.

Posted by El Cid | July 31, 2008 8:06 PM

Richard Cownie,

My point really is that all probability estimates that happen to have inconvenient consequences can be attacked as mere "educated guesses".

And my point is that, regardless of how they "can be attacked," some of them really are just educated guesses. Probability estimates of catastrophic outcomes from climate change are a prime example of such educated guesses. Scientists simply don't have any clear idea how likely such outcomes are.

We're in a bus heading towards a cliff:

No, we're in a bus heading into unknown territory. There is evidence that the territory will be somewhat hazardous, but we just don't know yet how hazardous it will be. In response to this situation, I want to drive carefully, keep a sharp lookout, and take some sensible precautions to limit the risk of harm. You want everyone to panic.

Posted by Mixner | July 31, 2008 8:14 PM

El Cid,

Shut up, you sci fi pompous nimrod.

Piss off, you insufferable ignorant fool.

Posted by Mixner | July 31, 2008 8:16 PM

Piss off, you insufferable ignorant fool.

Posted by Mixner

No, you, else I'll ruthlessly mock you for misquoting some classic piece of science fiction, thus establishing myself as a Tru Thinker!

Posted by El Cid | July 31, 2008 8:22 PM

El Cid: "And, RSH: please try to remember, you're not a Transhuman yet."

Well, to be precise, I'm a TranshumanIST.

OTOH, since I do think differently from you chimps, I think it's fair to distinguish myself from the rest of you as a Transhuman.

Given that genetic variation between humans exceeds the variation between humans and chimps, according to one source I read, I'd say the odds that I'm sufficiently genetically different from you to call myself Transhuman are pretty good.

One could also argue that the important thing about being Transhuman is thinking like one, the conceptual approach - even if you don't have the million-fold brain speedup yet.

Not to mention that for some of the posters here, such as SLC, Al, Chris Ford, Mixner, I probably could claim a million-fold brain speed advantage.

And, yes, many of you are "dumb meat people". And you are hearing me over the Internet, which is one of my "cybernetic frequencies" (well, my only one, actually, unless you count my cell phone.)

Trust me, it's going to get worse. At some point here, I'll have my own blog and then I will become truly insufferable. Yglesias will be tearing his hair out.

Posted by Richard Steven Hack | July 31, 2008 8:44 PM

"And my point is that, regardless of how they "can be attacked," some of them really are just educated guesses. Probability estimates of catastrophic outcomes from climate change are a prime example of such educated guesses. Scientists simply don't have any clear idea how likely such outcomes are"

Who decides what's "just an educated guess" ? You do. How very convenient. And whatever climate scientists find out in the future, they'll never be able to reproduce the whole climate in their labs, so you'll always be able to use this as an excuse for inaction.

If you could suggest some objective criteria for deciding what's "an educated guess"; and furthermore, such objective criteria as might possibly be met by developments in climate science, then you'd actually be saying something meaningful.

For myself, I'm pretty convinced that making big changes in CO2 concentration could have climate and ecological effects at least as serious as the "normal" droughts and heatwaves we've seen within living memory. And since that includes the dustbowl of the 1930s, that's quite bad enough to worry me: it doesn't take much climate change to convert most of the USA into an unproductive arid desert.

Posted by Richard Cownie | July 31, 2008 9:51 PM

No there isn't. You're treating probability estimates that at this point represent little more than educated guesses as if they were empirical facts.

You can push that little nugget all over the place. It's a reductionist argument. (Speaking of reductionism, you argue like an Objectivist. Am I close?)

More concretely, the IPCC represents the most conservative, broadest-consensus view of Global Warming. It had to pass muster with people like Richard Lindzen. If it's in the IPCC report, it's as "empirical" as we get in this vale of tears.

Now, turnabout is fair play. Where did BushCheneyMixner get "1%" probability for Iraq? Empirical fact, jack? Not on the life of hundreds of thousands. Massaged data, cherry picked facts, and withheld evaluations, it was a slam dunk.

Posted by Jeffrey Davis | August 1, 2008 9:14 AM

Freddie said, "Watchmen sucks."

I now know that I don't have to worry about any thoughts he has being intelligent or worth reading.

Posted by Captain Noble | August 1, 2008 11:00 AM

I think everyone is missing MY's broader point - that we should fake a polar bear attack on New York City in order to rally the people around a serious solution to climate change. That was your point, wasn't it MY?

Posted by Charles | August 1, 2008 12:18 PM

I think everyone is missing MY's broader point - that we should fake a polar bear attack on New York City in order to rally the people around a serious solution to climate change. That was your point, wasn't it MY?

Like MY and others, the recent teaser trailer for Watchmen prompted me to reread the book. It's strange to revisit the book after 15 years and have a totally enhanced view of it -- kind of like reading Catcher in the Rye as a high schooler and then again years later as a parent.

It seems to me that Adrian Veidt was just plain wrong, and that after a period of some years, free of horrific alien destruction, eventually nations would regress back to their combative ways. But then, I'm not the Smartest Man in the World (tm), so perhaps I'm missing something there.

Also: Putting a polar bear on 5th Avenue wouldn't do it. You'd have to submerge half the city in rising sea levels. I'm just sayin.

Posted by Ben | August 1, 2008 12:44 PM

hmm.. thank you very much. usefull information

Posted by Miley-Cyrus-Fan | August 1, 2008 1:08 PM

Super. It took almost a day to find this info. Thanks, great job. :)

Posted by Janet14 | August 3, 2008 3:45 PM