Saturday, April 30, 2011

Miscellaneous Links

1. More on public health and junk science.  Taubes seems a lot more sure about some of his conclusions than seems warranted, but still quite though-provoking.

2. Spring offensive begins in Afghanistan tomorrow.  My prediction: Taliban agrees to disarm within 30 days.

3. Another point in favor of the xenophobia story for Libyan intervention: it explains why no one cares about Syria.

4. Israeli government rules out talks with unified Palestinian government.  Dang.  Just when we were on the verge of a final agreement.

5. The myth of 1 billion people hungry people in the world.  Great subtitle: "What if the experts are wrong?"  How could that ever happen?!?

6. Obama orders Gitmo prisoners be transferred to next president.

Directed Readings with PotUS (updated)

Drezner proposes an interesting thought experiment.  Suppose it was in your power to select a handful of books for an incoming president of the United States to read.  Kindred provides an eminently reasonable list here.

I'll give an actual answer, but first I need to do my cranky thing.

Rat Choice Apologetics II

How Not to Evaluate Model Assumptions

In principle, the debate about rational choice and the debate about the use of formal models are separate debates.*  But as we saw with Walt's famous criticism, in practice, the debates are inextricably linked (original article here, also reprinted with all the responses as a book).  When people criticize the privileged position of game theory within the discipline,** they frequently argue that game theory has little to contribute to knowledge either because people are not rational or because rationality is defined tautologically.

There are other critiques made as well that are best viewed as critiques of formal modeling as distinct from critiques of rational choice.  These are important as well, and I intend to discuss them later in the series.  But for now, let's stick to these points.  I attempted to address the claim that people are not rational in the first post in this series.  Now I'd like to springboard off the second point into a more general discussion of how we should or should not evaluate the assumptions of a theoretical model.

Thursday, April 28, 2011

Rat Choice Apologetics I

Basic Myth Busting

The modal political scientist appears to believe the following two statements to be true:

1. Game theoretic models require the assumption of rationality.

2. People are not rational.

Some people conclude that therefore, it must be true that:

3. Game theoretic models are devoid of all value.

Even if the first two statements were true, this conclusion does not logically follow.  Perhaps the conclusion is nonetheless true.  I'll have more to say about the value of game theoretic models in the future.  But the truth value of premises 1 and 2 is of no use whatsoever, on their own, in determining the truth value of the statement 3.

At any rate, the point of today's post is that statements 1 and 2 cannot simultaneously be true. 

Monday, April 25, 2011

Rubinstein on the Exaggerated Usefulness of Game Theory

Podcast here.

Though being a committed advocate of game theoretical modeling myself, I've long been bothered by some of the grandiose and ultimately inappropriate claims that have been made about what game theory can do.  So it's nice to hear such a respected scholar, himself a game theorist, expressing the same concerns.  My views on what game theory can and cannot do, what obligations scholars have to society, and many other matters are almost exactly the same as Rubinstein's.  (And no, I didn't know he held these views before now.)

As you might expect, I have more to say on the topic, but I'll hold off for now.  It looks like there would be some interest in the series of posts on rational choice theory that I proposed the other day.  One of the later posts in that series will articulate my views on the use of theoretical models in general, including the pros and cons of different classes of formal models.

Thursday, April 21, 2011

UNC Proposal to Tackle Grade Inflation

Story here (h/t Kindred).

There are a few things I love about this proposal.

If Democracies Have No Stomach For War, Is That An Advantage?

A number of scholars argue that, relative to other regime types, democracies have no real stomach for war.  The assumption is that mass publics the world over have no stomach for war, and so those regimes that hold their leaders accountable to the public will similarly have no stomach for war.  There are some important debates remaining about whether the public generally cannot stand the costs of war, or is specifically only unwilling to bear costs when victory is unlikely, but there is fairly widespread agreement that wars have greater implications for the ability of democratic leaders to retain power than is true of other leaders.

Hold aside the question of whether this is in fact the case (though it's far from clear that it is).  I'm struck by the fact that we've managed to convince ourselves that, were this true, it would be a virtue.

Wednesday, April 20, 2011

Libyan Mission Creep Update

Yesterday, we learned the Brits and the French were sending advisers (see this post at QP).  Today, we add Italy to the list.  (Much to Munger's amusement.)

Tuesday, April 19, 2011

Building Qaddafi a Golden Bridge

Steve Walt suggests that by offering to arrange an easy exit for Qaddafi, the US risks sending a message to other embattled dictators not to go quietly, lest they end up like Mubarak (who is now in prison).  Presumably, we are to conclude that the US would be making a mistake if it ends up shuffling Qaddafi out of office this way.

Friday, April 15, 2011

Democracy and Civil War

As I've discussed in the past, it is not clear to me that we have either theoretical reason to expect nor unambiguous empirical evidence indicating that democracies are more likely to provide public goods.*  (See this post or this paper).  Basically, we have good reason to believe democracies satisfy a larger % of the population, and governing for 50% plus 1 is much better than governing for a handful of cronies, but even so, we ought to recognize that even in democracies, there are winners and losers. 


My co-author, Nick Nicoletti, and I have toyed with the idea of discussing the implications of this difference with respect to internal conflict.  We'd have to cut some stuff to do that, and so far we feel it's beyond the scope of the paper.  And I'm still not sure how I feel about that...but this week, I did some empirical analysis (not even on simulated data, if you can believe that!) for a separate project on civil wars (see this post or the preliminary version of the paper), and I found some interesting results with respect to democracy and civil war.  I don't know if these results will make it into either paper, because they probably only make sense if you consider both arguments at once...but I'll have to think about it, and discuss the options with the co-authors of each paper.

Tuesday, April 12, 2011

At Least We've Ruled Some Stuff Out

A common critique of the "fetish" for methods in political science is that the use of quantitative and formal approaches hasn't brought us anything of real value in terms of understanding the world.  Recently, an op-ed written by a Princeton grad student made the rounds on the internet making this point, but it's hardly a new criticism.

I'm somewhat sympathetic.  I agree that many advocates of these approaches claim that their use will someday enable us to accomplish all sorts of wonderful feats, including predicting conflicts before they happen (more on that soon).  And so far, we haven't done a great job of providing compelling answers to important questions (and by compelling, I simply mean beyond reasonable criticism), as I discussed here and here.

But I think it's problematic to infer from this that more technical approaches have not born fruit.  Specifically, I think one valuable role the use of these approaches has served is to produce something close to consensus in favor of rejecting arguments that non-specialists often find quite appealing.  That's not as glorious and sexy as what we were promised, perhaps, but it's something.  So look below the fold for a brief discussion of some examples of arguments related to international conflict that I think we've more or less ruled out.

Monday, April 11, 2011

Implications of US Reversal on Yemen

The US is severing its ties with Yemen's Saleh, who had up until recently been considered a critical ally.

John Quiggin argues that this further illustrates that the US is not a hegemon, and goes on to conclude that the US would be irrational to maintain ties with Middle Eastern dictators now that it has proven that it will not stand by these regimes when/if they face domestic challenges.  I agree with a lot of what Quiggin says, but I have a few issues with his conclusions.

Of course, you should first read Kindred's response, if you haven't already.  But below the fold, I offer a few thoughts of my own.

Saturday, April 9, 2011

Miscellaneous Links

1. Military cracks down on protests in Egypt, in which protesters demanded the resignation of the man who replaced Mubarak.  I, for one, am still overcome with joy now that democracy has come to Egypt.  How about you?

2a. US may not leave Iraq this year after all, and 2b. Muktada al'Sadr threatens to relaunch violent campaign if US does not.  Good thing the US didn't elect the guy who was going to keep Guantanamo open, keep troops in Iraq indefinitely, escalate the war in Afghanistan, and maybe even get involved in a third war. 

3. Jim Fearon on Libya and mission creep.  Self-recommending.

4. US-Pakistan intelligence operations frozen since January.  But it's not like any of the experts consider cooperation from Pakistan to be vital to success in the Afghanistan war, so that's cool.

5. XKCD on statistical significance.  Sure would be nice if journalists got this joke.

6. Speaking of not understanding statistical significance, academics only seem to be so much better.  You should see if you do any better on the challenge.  It will only take you two or three minutes. (H/T Matt Zimmerman)