Mankiw’s Wager Posted by: Jon Henke
on Monday, May 29, 2006
Bear with me, becaise I'm going somewhere with this. Greg Mankiw discusses self-interest...
So what motivates me? Self-interest or respect for the truth? The answer, I believe, is both.
Suppose I filled my blog and books with nakedly self-interested statements: "The world would be a happier and more prosperous place if everyone sent Greg Mankiw $20." First of all, no one would believe me. Second, people would stop reading my blog and buying my books. Making self-interested statements like this is not in my self-interest.
If "send me $20" does not appear to be in Greg Mankiw's self-interest, perhaps that is only because he's framed it poorly. Frame it like this instead:
"If you send $20 to Greg Mankiw, there's a small chance that you'll receive infinite, eternal happiness VS a large chance that you'll only be out a small, finite sum of cash. Let us weigh the gain and the loss in sending a 20 to Mankiw. Let us estimate these two chances. If you gain, you gain all; if you lose, you lose 20 bucks.
Send, then, without hesitation, that Jackson to Greg Mankiw!"
We'll call it "Mankiw's Wager". If history and Adherents.com are reliable guides, Professor Mankiw ought to be a billionaire in short order.
If "Mankiw's Wager" seems like a bad cost/benefit calculation to you, then consider the same argument made with regards to national security threats — terrorism, illegal immigration, nuclear proliferation, global warming, etc. They frequently come in the form of...
Consider the existential threat of [X]. Our country, our freedom, our very lives are at stake! If we Do Something, we might save something of infinite value; whereas, if we Do Nothing, we might lose everything! Therefore, let us estimate these two chances. If we gain, we gain all; if you lose, we lose the finite cost of action.
Send, then, without hesitation, those bombers to Iran! We have everything to gain, and nothing to lose but our opportunity cost!
The threats from actual terrorism and nuclear proliferation are obvious, whereas Mankiw's Wager is, of course, a threat asserted without evidence. On the other hand, it's far from clear that a less restrictionist border policy or a quasi-nuclear Iran would constitute an actual threat. (as opposed to a theoretical one)
I don't wish us to dismiss the genuine cost/benefit calculations that can be made with regards to security threats, but I would like us to view problems in terms more nuanced than "it's a threat! Do something!". That way lies overextension. As Walter Lippman warned during the Vietnam war, there's a natural gap between our strategic goals (which may be infinite) and our resources. (which are necessarily finite) When we reach that state of geopolitical insolvency, the crash can be painful.
I write that in order to urge caution in our approach to problems and modesty in our solutions. The worst case scenario may be a danger, but it is one of many. Before we decide, for example, that military action is necessary against Iran — and I grant that, as Joschka Fischer writes in the Post today, Iran's "combination of hegemonic aspirations, questioning of the regional status quo and a nuclear program is extremely dangerous" — there is every reason to negotiate. It may be that our eventual choice will be a military strike (and the enormous blowback it would bring) or a nuclear-armed Iranian threat. For now, though, the choice is far less dire and the case for bargaining with Iran, far stronger.
the same reasoning applies to many things...such as global warming, water pollution, drunk driving, or any other real, potential, or invented problem...what is the cost for that extra measure of safety - usually freedom, money, or lots of both.
On the other hand, it’s far from clear that ... a quasi-nuclear Iran would constitute an actual threat. (as opposed to a theoretical one)
I don’t know what a "quasi-nuclear" Iran or a merely "theoretical" threat means, but I am confident that an Iran possessing nuclear weapons would pose an actual threat regionally and, as a result, if not directly then proximately to the vital interests of the U.S.
I don’t conclude from this that the U.S. should not engage in discussions with Iran, let alone that military intervention is warranted at this point. However, as between this and your last post (and echoing the earlier comment), it seems to me that the imminent threat of a nuclear Iran is far stronger than the imminent danger of global warming.
I don’t know what a "quasi-nuclear" Iran or a merely "theoretical" threat means,
Quasi-nuclear Iran = Iran that has nuclear energy, but not nuclear weapons. It is, of course, also far from clear that Iran would stop at nuclear enery and most observers seem to think they’re not terribly interested in nuclear enrichment for energy alone. Having no basis for contradiction, I’ll accept that consensus. Still, stopping Iran from pursuing any enrichment at all, and stopping Iran from pursuing nuclear weapons are categorically different things.
By "theoretical threat", I mean the uncertainy about whether Iran would progress from energy enrichment to weapons-grade enrichment. That a nuclear-armed Iran would be a danger to us — either by using nuclear weapons or threatening their use as geopolitical leverage — is fairly clear. Whether Iran would progress from energy to arms is less so.
Seperately, I think a case can be made that a nuclear armed Iran would probably find their weapons more useful as geopolitical leverage than as mushroom clouds. That, again, is a different calculation.
it seems to me that the imminent threat of a nuclear Iran is far stronger than the imminent danger of global warming.
Of course. The threat from global warming is about 20-100 years off. Nobody even speculates (in these terms) about what Iran will be 20-100 years hence.
For that matter, the threat that I’ll be attacked by a gypsy is more "imminent" than the threat of global warming, but I don’t plan to take any immediate defenses against gypsies.
Jon I think you have the cart before the horse with Iran. Iran with nuclear power will FOLLOW Iran with Nuclear Weapons. I believe the reactors are the precursors to nuclear weapons, just as they were in the US.
Ah, but "stronger" applied to "threat," not "imminent." I can’t speak to your gypsy situation, but the threat of global warming, however real it may be, does not pose the sort of catastrophic chain of events that a nuclear Iran poses unless one believe both the tipping point and worst case scenerios about global warming, and I am convinced of neither.
In any case, given that a future contingency is always in principle possible of not happening, one also has to go "of course" about the possibility that Iran would not proceed to build or to use nuclear weapons, but little about the situation suggests to me that it would not be folly to expect the best.
I understand the point you are making but you undermine your point by lumping a lot of things together which are not the same, nor have anything like the same level of threat/reality. From where I sit, in southern Texas, I can tell you that unrestricted immigration is not an inconsequential problem. I know you don’t agree with me, but, well, you are wrong. And that is just one example. If we are doing strict cost benefit analysis then you have to consider each step to deal with each perceived problem discretely. They are all different.
Iran with nuclear power will FOLLOW Iran with Nuclear Weapons.
Maybe so. But it’s a VERY large step from nuclear power to nuclear weapons. It’s not a foregone conclusion that they’ll take that step. A reasonable assumption, perhaps, but not a foregone conclusion.
the threat of global warming, however real it may be, does not pose the sort of catastrophic chain of events that a nuclear Iran poses unless one believe both the tipping point and worst case scenerios about global warming, and I am convinced of neither.
I suppose it depends on what kind of scenario you posit. I could come up with scenarios in which either one would be worse, but most such scenarios would depend on some pretty unlikely stuff.
I do agree that we should work to prevent Iran from getting nuclear weapons, and the best way to do that — at this point — is negotiation. Skipping straight to military intervention is the next-to-worst of all possible paths. (and I know you’re not advocating that)
Don’t you pay taxes that support a police force and prisons that will hopefully deter the random Gypsy attack?
No. They exist to detain criminals, not to prevent crimes.
Yes Jon it’s about assumptions.... I ASSUME Iran has the nuclear reactor to develop nuclear WEAPONS. You assume that they have the reactor for power. It depends which you assume is the by-product, the power or the weapons. I am assuming that it’s the power, you the weapons.
And no, actually it ISN’T a big step from power to weapons. Stagg Field went critical in ’42, Hanford in ’44/45? First weapons in July 1945. You seem to assume that first one must have a reactor, producing power, to move to weapons. And the US Manhattan Project shows that to be untrue. You build a reactor to "cook" plutonium, or power the Uranium enrichment process, which uses a lot of energy.
So it is NOT sequential... It’s based on what you want the nuclear infrastructure FOR. I assume the worst, you mayhap do not. But there is no reason that peaceful power will emerge first.
Enriching to the low levels required for nuclear energy requires a different set-up that enriching to weapons-grade levels. For example, if Iran only had light-water reactors, they couldn’t go on to produce weapons-grade uranium. The question is how certain we can be that they don’t upgrade their facilities, or carry on concurrent weapons-grade enrichment. That may or may not be possible to do, depending on their level of cooperation. But it’s important to remember that nuclear weapons are not merely the result of "more of the same" activity that producers nuclear energy. They require somewhat different infrastructures.