Incompleteness and the precautionary principle

I guess I’m still a bit puzzled by the role consequentialism is playing in all this. We need a clear set of principles to govern the use of force in the international order, and we have such a set of principles in traditional just-war doctrine, principles which we could extend or adapt to cover humanitiarian intervention cases.

Given your invocation of the principle of self-defence in fn.1, I guess that your set of principles is pretty much the same as those in this just war doctrine plus.

Now if I were to tell a justificatory story about those same principles, it would proceed differently, by way of a story about individual rights (including the right to self-defence). But I doubt that the content of the principles would differ much — if at all — from yours.

(Maybe it is a strength that we can tell different justificatory stories, since it is important that any set of rules be mutually acceptable, and the chance of universal acceptance is improved if we don’t all have to buy into the same underlying justification — a case of Rawlsian overlapping consensus.)

Thanks for the paper. For this budding economics student, the discussion of more and more general models of uncertainty was enlightning. I think I’ll print it out and use it as a cheat sheet.

Your analysis is VERY dependent on how the decesion maker determines which state of the world is “status quo” and what would be an “innovation”. Where you see preemptive war in Iraq as an uncertain innovation, Bush sees uncertainty in the status quo, i.e. an unknown (and unknowable?) connection between Saddam’s WMD and terrorist willing to use them. This would be analoguous to the “innovation” that you observe in the doubling of green house gas emmisions if we maintain the status quo.

Similarly, there is some fudgy-ness in the statement of the incompletness hypothesis. “Incomplete estimates will generally be over-optimistic.” Your over-optimism may be my over-pessimism.

I’m reminded of Rumsfeld’s famous line about unknown unknowables and some such. What rule of nature implies only bad stuff is more unknowable?

Looking at your last sentence and the accompanying footnote it seems to me that you are departing quite radically from just-war doctrine. Suppose the “true” reason for the invasion was to place Iraq under a leader who would be better for Iraq and for the world in general. Also, suppose the operation had been properly planned and executed. Then you have a war which is justifiable on consequentialist grounds but not a just war in the traditional sense.

“Optimism” in John’s paper is being used in a sense of “cognitive bias toward overestimating the benefits and underestimating the costs of a course of action which is desired”, rather than a general sunny-side-upness.

John: have you seen this paper in Metroeconomica, on evidence theory which looks like a kinda-sorta similar approach?

Basically, it’s an argument about reversibility of courses of action. If we decide to cut carbon emissions and it turns out to be unnecessary, then we can just start burning coal again; we’ll have lost a couple of decades of economic growth in the worst case (which I frankly don’t believe) but no more.

On the other hand, if we make the opposite mistake and London sinks beneath the waves, then it’s pretty much gone and there’s no bringing it back.

Similarly, having not started a war, it’s easy to move to a state of affairs where you have started one, but having started a war, you can’t unstart it.

It is curious that most people who endorse the precautionary principle were and are vehemently opposed to Bush War II.

The PP is an alternative to American-style “cost-benefit analysis.” CBA is a method of decision analysis that converts all uncertainties to probabilities and all consequences to money. The best decision is the one that maximizes expected value (net present value? expected utility?).

The PP says that in situations where the stakes are very high and there is a lot of ambiguity (second-order uncertainty about the magnitudes of the consequences and probabilities) you should give more weight to the worst case scenario and have a bias toward inaction.

The PP is to CBA as Allais and Ellsberg are to Subjective Expected Utility theory (von Neumann & Morgenstern, Savage).

Although Allais and Ellsberg argued that their counterexamples to SEU challenged its normative validity as well as its descriptive accuracy, almost all economists still accept SEU as a normative standard.

The PP rejects the normative status of SEU in a way that is very similar to how Allais and Ellsberg did.

Two concepts from behavioral economics can clarify the PP. PP says that in certain situations, a bias toward inaction (status quo bias) is a good thing. And a bias toward giving “more weight” to negative deviations from the status quo than to positive ones (loss aversion) is a good thing.

The PP says that sometimes, it’s rational to be biased.

When is war a rational policy decision?

JQ: If war is to be an instrument of policy, it should only be used under conditions of overwhelming superiority in all phases (including occupation), for clear and feasible objectives and with a clearly formulated exit strategy (except where the threat is so clear and imminent that standard self-defence arguments can be invoked).

These objectives need to be justifiable in terms of the interests of the people of the world in general, and not of the national interest of one country or the personal interests of its rulers. A nation or group that pursues self-interest through military force is an enemy to all and will ultimately attract retaliation.
—
The United States’ attack on Iraq in March, 2003 fails on at least three counts.

1.Extremely vague plan for occupation and no exit plan.

2.Gross neglect of well-being of occupied people. The US attack on Baghdad damaged critical infrastructure (electricity, water). The US cleanup crew (Cheneyburton) repaired damage to the oil facilities before it repaired damage to the infrastructure of Baghdad.
The failure to attempt to keep track of civilian casualties also might be viewed as “failure to give sufficient weight to the objectives of the occupied people.”

3. If one of the objectives is to deter a non-imminent attack, the evidence of opponent’s intent and capacity to attack should be very strong.

There is no evidence that Iraq’s intention and cabability of attacking the United States were very high.

The logic went something like this:

Saddam Hussein will not allow U.N. weapons inspectors in. Therefore, SH is hiding WMDs.

We know now that SH had another reason for wanting to keep UN inspectors away – he was embezzling money from the UN food-for-oil program. The UN knew that, of course. So did Uncle Sam. But Sam pretended that Saddam’s refusal to allow UN inspectors was more diagnostic of having weapons than it really was. And Americans bought it. (Americans will buy anything.)

The “lackadaisical attitude to weapons exhibited after the invasion” is further evidence for the hypothesis that “they knew the WMD case was bogus, and needed to start the war before it collapsed altogether.”

I think Uncle Scam gets an F, according to Quiggin’s theory of when war is a rational policy decision.

John@AUS: Since war is a negative sum game, rational decision makers do not fight wars.

Lizard@ENG:rational decision makers do not fight wars with other rational decision makers.

Deb@USA: I think Liz’ hypothesis is much more plausible than John’s. But I think they are both false. If the carrying capacity of the earth were greatly decreased (due to global war, global warming, global capitalism, etc.) and the citizens of one nation could not survive without the resources of another, it might be rational for both sides to participate in a war. The US asks Canada to help us out with food and water. Canada says no. So US attacks Canada.

It might be rational, if Americans weigh the well-being of Americans more than the well-being of Canadians.

Would it be irrational for Canada to fight back?

I’m 99% sure it’s possible to conduct a scenario where rational decision makers choose to engage in war.

Where’s Kevin McCabe when you need him?

“One thing that struck me over and over and over again was how many military conflicts started with the parties completely off base as to reality.”

Many biases demonstrated by Kahneman, Tversky, Lichtenstein, Fischhoff, Slovic and other cognitive & decision psychologists apply to military decisions including overconfidence, self-serving bias, outgroup homogeneity and cognitive dissonance. [See Fog of War thread.]

When there’s this much garbage in, cost-benefit analysis gives garbage out. The precautionary principle attempts to counteract overconfidence and self-serving bias.

Skeptics of the PP rightly note that it endorses biased decision making, most notably, the status quo bias. Mistakes of omission are treated less harshly than mistakes of comission. In ambiguous situations where the decision makers are under stress, don’t have all the information, etc. it is rational to prefer to err on the side of caution.

Stephen M:’One thing that struck me over and over and over again was how many military conflicts started with the parties completely off base as to reality.’

That’s another point made by the great Australian historian Geoffrey Blainey in ‘The Causes of War’. Blainey said wars are always fought by two parties who believe they can win, which means at least one of them was badly mistaken. I would qualify Blainey’s point slightly: just occasionally you do seem to get countries fighting despite knowing the odds against them to be almost certainly overwhelming (Greece 1940-1, Norway 1940). But yes: anyone who reads military history comes across this time and again- the utter unrealism of so many policymakers’ view of the world.

Geoffrey Blainey: excellent historian, very good prose style, said nasty things about immigrants before he died but still- every intelligent person should read ‘The Causes of War’.

I have to say, though, that I’d like to go through his archives to discover what his data set was when he asserted that increased trade between nations did not decrease the likelihood of war between them. I suspect that he was thinking of the onset of the First World War, which was indeed marked by an increase in international trade, despite the tariff barriers of all the Great Powers bar the UK- still, I’m a bit dubious about that one.

d*d : I guess that was a joke, because o.w. it seems you agree that the two cases are analogous.

I rather think it does, or at least something similar to it does. There does appear to me to be an important sense in which the flooding of London (if it happens) is an irreversible event whereas the decision to restrict carbon emissions is one that can be reversed. Or to put it crudely (and I’ve always been a big advocate of the options approach that John mentions in the paper), time is on your side when you postpone a war (as long as it is genuinely a war of choice), but not when you postpone doing something about global warming.

DD–if it’s true that, as some have argued (included Kip Viscusi, for example) that regulatory expenditures can lead to an increase in mortality, then the secondary effects of the restriction on carbon emissions (a regulatory expenditure) include things that can’t be undone. Death is, still, irreversible. (And uncertainty about whether this effect actually holds would, I gather, argue in favor of doing nothing, wouldn’t it?)

John Q–A lawyerly comment and then a real one: When someone hasn’t complied with a requirement, lawyers argue that they have “substantially complied.” In truth, in early 2003 the US and its allies had a number of options. The option that provided the most certainty was the invasion option. One can think of a variety of options for dealing with the threat of global warming, including a simple reduction in the amount of energy used to massive investments in nuclear power, and so on. But each of those options must be run through the same decision process. A costly and bad outcome–invasion–is still preferred to the uncertainty of the consequences in the scenario you described.

How is the flooding of London irreversible? Said flooding will not occur overnight; the technology to reclaim land from the sea is pretty well understood (e.g. Netherlands), so you put a dam across the Thames. I think you’re assuming your conclusion.

‘Two possibilities’ because there was a third but I’ve developed Doubts about it.

Giles:’Thus over the period your utility might be higher if you are over optimistic about the war which you then lose, than you are if your pessimistic during the whole war, but then get a good outcome and win.’
This is a joke, right? In the context of a *war* (including, you know, deaths, maimings, destruction of property, invasion of territory and other minor disutilities of that nature) I might be *better off* if my country loses a war but we all spent the preceding few years misguidedly *feeling happy* about winning it, than if we won the war but everyone had been a little bit pessimistic?

If I spot you London sinking into the ocean (a decidely unlikely event under even the more ridiculous global warming circumstances) you get to analyze the war from the unlikely circumstance of Saddam giving a nuclear weapon to terrorists who successfully explode it in New York City.

But then we also analyze the war from the somewhat less unlikely (ex ante) circumstance of the war angering Pakistani radicals to the point to which they assassinate Musharraf, take over the govt, give nuclear weapons to terrorists etc.

I think a point to be made here is that if we insist on concentrating on the worst possible outcome it will always be very very bad. (Do we know that there’s not a crazy god who will destroy the world if we do/do not follow a course of action. Yes we do, but never mind.) The kind of uncertainties at issue in global warming at least are such that we can say that irreversible consequences are more likely if we fail to combat global warming. (Even if we stipulate that flooding London is unlikely.) And I think it was pretty easy to tell ex ante that bad irreversible consequences were more likely if we went to war than if we didn’t.

Of course there’s a problem about how we define irreversibility, because no action is truly irreversible. I think that getting the US military stuck in Iraq for several years is somewhat irreversible. Certainly the deaths of people killed in the war are irreversible, but there are people alive today who would not be if we had not fought the war, and those deaths would also have been irreversible. So we need to look at a higher level.

What about the moral case for acting instead of sitting idly by doing nothing, even though acting may cause some unintended consequences?

If a woman is getting beaten up across the street, and you decide to intervene, there is a chance that you will make the situation even worse and cause more harm to everybody. Should you walk on by to avoid that possibility? Or do you act?

JQ: On irrationality, Lizardbreath is right. At least one party must be irrational.

DF: This is still up for grabs. I’d wager $100 that it’s pretty easy to come up with a game theoretic toy problem a la Becker and Posner where war is rational.

Two rational, ethical agents probably would never engage in war. But rational agents, without the assumption of ethical probably would.

JQ: Now apply symmetry.

DF: Huh?

JQ: The occurrence of a war implies that somebody is acting irrationally and if you’re the one starting the war, it’s probably you.

DF: Yikes. This is an oversimplification.

you get to analyze the war from the unlikely circumstance of Saddam giving a nuclear weapon to terrorists who successfully explode it in New York City.

The fact that this completely untrue scenario seemed to form a fairly substantial part of the calculation about going to war looks to me to be a pretty good example of the cognitive bias part of the incompleteness principle.

That is quite a dodge considering how often we have to contend with how often the highly unlikely scenarios get high profile play in the global warming arena.

The London underwater probability is certainly in the marginal/vanishing probability zone. The chance that Saddam if left alone would have eventually built a nuclear device is I would say excellent, and that he would pass it off to a terrorist organization is certainly non-zero. You raised the London underwater possibility as part of the discussion.

You either know very little about global warming or are aware that the probability of your scenario is vanishingly small. I’ll assume that you weren’t making a point about completely ignorant bathering in reasoned calculations so you knew it was vanishingly small.

So……other than the moral of taking swipes at the pro-war crowd, did you have a point about making choices based on incredibly low probability but really bad scenarios? Because at the moment it appears that you employ the precautionary principle for advancing leftist aims and some other principle when critiquing everybody else.

The chance that Saddam if left alone would have eventually built a nuclear device is I would say excellent

This is precisely why I didn’t recommend leaving him alone, and said so, in writing, repeatedly. I’m actually now quite disappointed because I thought you were a regular reader, Sebastian.

Global warming is not a “leftist” issue and I don’t really understand how anyone might think it could be. Political policies can’t affect the level of the ocean; this is a sociological fact which was proved by experiment by King Canute. I’m referring to London being underwater because I have a map in front of me which I cut out of the Standard, showing all the postcodes where it is now very difficult to get flood insurance because on latest modelling the existing Thames barrier won’t be able to save them in a few decades time. It doesn’t include my own house, but there are several billions worth of real estate in the cross-hatched area, which houses are presumably fucked whether they vote Labour or Conservative.

To recap, my policy projections were:

Problem: An underplanned and poorly resourced war might go really badly wrong.
DD solution: Don’t start a war until you’ve planned it properly.
Alternative solution: Start the war anyway because we might be lucky.

Problem: Saddam might blow us all up with a nuke.
DD solution: Put inspectors in place to make sure he doesn’t.
Alternative solution 1: Don’t bother because we might be lucky.

Problem: Global warming might flood us all.
DD solution: Do something about it.
Alternative solution: Keep on burning vast amounts of hydrocarbons because we might be lucky.

Problem: Reducing hydrocarbon consumption might plunge the world into recession.
DD response: Start off with a sensible small reduction like the Kyoto treaty.
Alternative response: Ban all hydrocarbons today, because we might be lucky.

You see that the common thread here is that the “DD response” usually involves taking precautions today against an identified future danger, while the “alternative response” tends to rely on assuming that unknown outcomes will be favourable.

Problem: A problem.
DD Solution: Solve the problem!
Alternative solution: Do nothing.

Yes, well if it were only that easy, and if it were only what we were talking about in the first place.

Kyoto is a mild step? What were the mid-range estimates? $700 billion or so? For delaying global warming 1-3 years? The entire cost of Social Security in a year could go a long way in making things better for people in non-speculative ways.

Re Kyoto, I’ve always wondered if the real benefits don’t include the fact that it will put in place, and test, a worldwide carbon-emissions caps system (and consequent measurement, enforcement and cap-trading mechanisms). Current economic theory does say that caps and trading schemes are the best way to reduce harmful emissions, but we don’t know if such a scheme can work on a massive, international basis across many different countries and legal systems, as opposed to within one legal system, like the EU or the US. Secondly, even if it can work, the chances are that there will have to be a lot of ‘learning by doing’ before it works anywhere near optimally. I suspect Kyoto, now that it is off the ground, may function mainly as a feasibility test. If it works, with amendments, then it can form the basis for a larger successor scheme; if it doesn’t work then we have to either design an entirely different international emission control system, or start the bargaining necessary to amend international law and strengthen international co-operation such that an international capping scheme is feasible.

This isn’t a rhetorical point: I’m genuinely interested as to whether this might prove to be the main result, intended or unintended, of Kyoto.

John,

All Posner said was “A rational decision to go to war should be based on a comparison of the costs and benefits (in the largest sense of these terms) to the nation.” He did not say that a nation should give equal weight to the well-being of the citizens of all nations.

Your claim that rational nations will never choose to go to war with each other needs to be defended, not merely stated repeatedly.

Dear John,

I’d been trying to avoid having to read d-squared’s comments (my time spent blogging is already > optimal). And I doubt (p is less than .05) that d^2 has proved that “two irrational parties are necessary for war” any better than you have (i.e., not at all).

But you’re the main man here, JQ. So I’ll check out d^2 and get back to you. If you could direct me to the particular post of the prolific d^2 that addresses my concern, i’d be most grateful.

Yours,

Deb

I’d like to see you apply this so-called principle to global warming. Wait, that would require that “wait and see” alternative.

Deb, it seems that JQ and DD are trying to invoke the Nash bargaining solution. I can’t really argue with that as it is in fact “standard stuff”, but it would have been easier if they’d just referred to it rather than the ongoing handwaving and poorly worded defenses. And no, the Nash bargaining solution doesn’t establish that rational actors don’t go to war. It only establishes that they don’t go to war if bargaining is an option. Also, JQ started out by criticizing Posner for not incorporating that the other side might be acting rationally. Now he claims that Posner’s “analysis makes no sense on his own assumptions”. But all we’ve seen from JQ and DD is that war doesn’t happen if both sides act rationally, which is something Posner doesn’t assume.

Thomas Schelling ‘Strategy of Conflict’

A great book, and at one point he describes what seem to me situations where rational people can commit to war.

For instance, it is sometimes rational to bind yourself to your statements/threats, so that the other party will be forced to concede because, yes, you will destroy the entire planet if you don’t get the last piece of Thanksgiving pie. Problems can arise when two parties bind themselves in incompatible ways (for instance, due to communication channel errors/delays).

John,

I am really not sure why this is so difficult. You made a very specific claim:

“According to game theory, two rational agents will never choose war.”

I disagreed. I provided a counterexample – a plausible scenario in which Nation A (e.g., USA) would threaten to wage war against Nation B (e.g., Canada) unless Nation B did x, y and z and Nation B said “Bring it on” and Nation A said “You got it.”

Instead of responding to my counterexample to your conjecture, you said:

“I think we are arguing at cross-purposes.”

Sometimes you wave your hands in the direction of “what everyone knows is true about game theory.” Other times you wave your hands in the direction of “Posner’s definition of rationality.” When that fails, you resort to deflection and suggest our disagreement is illusory.

You seem to think that Posner and game theorists in general believe that people are omniscient. You seem to think that game theorists think that the payoffs in the cooperation/defect matrix can be specified with certainty.

In a prisoner’s dilemma, this might be true. The warden says “If you say X and he says Y, you go to prison for 4 years and he goes to prison for 8; if you say Y and he says Y, you both go to prison for 2 years, etc.” In the real world, the consequences/payoffs are not known with certainty.

Even though I think game theory is kooky, I don’t think it’s as kooky as you do. I think that game theory would say that the consequences associated with waging war are uncertain. The consequences are very different depending on whether you “win” or “lose” and this is up for grabs unless and until you actually wage war.

I provided a blurb about von Neumann showing that like me, he thought that a rational agent might choose to go to war.

I agree with you that in general, nations that choose to attack others are almost always irrational. I think Sam’s nuts, Saddam’s kooky and Osama’s certifiable. It’s an empirical fact that most people who wage war are irrational.

But the British citizens who lived in North American in the mid-1700s who chose to wage war against their government were not irrational.

The fact that most warmongering nations are ruled by lunatics is an empirical fact, not an analytic truth.

JQ: There are some standard game theoretic conditions under which the proposition “No two rational players will choose a Pareto-dominated outcome” is true. For example, these include full Bayesian rationality (not omniscience – the argument copes fine with uncertainty as long as the players are EU maximisers) unbounded reasoning capacity, and free communication between players.

DF: Great. Now we’re getting somewhere. I’m 98.235% sure that we agree that two omniscient rational agents would never fight a war, because they’d just make a contract for the end state and skip the blood and gore. You can fancy it up and call it a Pareto optimal Nash equilibrium, but that would be kind of silly, since it’s an utterly trivial “result in game theory,” given the impossible, implausible assumption. Glad to see you’ve abandoned this line of reasoning.

Now you’ve redefined the claim “At least one irrational agent is necessary for war” to mean “Two rational Bayesian expected utility maximizing decision makers will never engage in war.”

This is a more interesting hypothesis than the one about omniscient agents, but it still seems to me it’s blatantly false. My USA/Canada war example assumed rational Bayesian expected utility maximizers. I think it is easy to construct scenarios where two subjective expected utility maximizers (as you know, this implies they are Bayesians) choose to engage in war.

The hypothesis on the table is:

Will SEU maximizers ever choose to engage in war with each other?

I said yes and I provided an example. JQ says no, and provides more handwaving. He says “the argument copes fine with uncertainty as long as the players are EU maximisers.” This is a tad sketchy – what exactly is the argument for why Bayesians won’t engage in war?

I am 97.59% sure that if you asked mutually acknowledged experts in Bayesianism whether the new version of Quiggin’s conjecture is true (e.g., Shafer, Edwards, von Winterfeldt, Clemen, Hacking (is he still alive?), etc.), 100% would say two Bayesians might choose to duke it out, even if we allow the false assumption of “unbounded reasoning capacity.”

[This post refers to a discussion on JQ’s blogsite. Sorry about the discontinuity and my sympathies regarding JQ’s problems with the lunacies and criminalities of cyberspace.]

JQ wrote:

“If the expected benefits of investment are large enough, they outweigh the gain from waiting.”

According to the model proposed, if the expected benefits are large enough, I would have thought that the aggressor will decide for war at period (n).

I’m trying to think of conditions that might improve the expected rewards of war within realistic time constraints.

Some possibilities:

1. Discovery of new resources.

2. Development of new technologies that cause a revaluation of already known resources. (Mesopotamia, for example, was just a bunch of sand in the eyes of Westerners until the discovery of uses for oil.)

3. The lure of a tied market for domestic industries. (There was a huge literature on this aspect of imperialism spearheaded by Lenin and Hobson.)

4. Domestic political benefits arising from acting on demonisation of the enemy regime or the politico/cultural arrangements promoted and/or imposed by the enemy regime. (The Cold War and its episodic hot spots, such as the Vietnam War, conforms to this pattern.)

But with the exception of the lure of tied markets, it seems to me that these benefits develop too slowly to be encompassed in a coherent decision-making sequence. (This question of the persistence of memory and the framing of a coherent purpose may need to be added to any robust model for decision-making of the type that leads countries to form foreign policies, including starting wars.)