Does peacekeeping work?

A couple weeks back I blogged Bill Easterly’s critique of military intervention in poor nations.

As usual I think Bill is brilliant, but in this case I have a higher opinion of peacekeeping interventions, in small part due to my experience in Liberia, and in large part because of the work of scholars like Page Fortna and her new book, Does Peacekeeping Work?

Now I see that Andrew Gelman, a Columbia professor and blogger, has made Page’s argument picture-perfect. From his excellent blog:

Here’s the graph (which I made from Page’s data when she spoke in our seminar a few years ago):

peacekeeping.png

The red points are countries with peacekeeping, the black points are countries without, and the y-axis represents how long the countries have gone so far without a return of the conflict.

One of the best things about Page’s book is that she tries to investigate the (obvious) selection problem that could be driving the result: namely peacekeeping missions tackling the easiest conflicts. It’s difficult to measure, but her evidence actually points in the opposite direction: peacekeepers pick the tougher cases. If anything, we may be underestimating the effect of peacekeeping.

Page took a shot at a question that eludes easy causal analysis, and shows that with careful case study and thoughtful statistics, you can make an important and convincing argument. In this spirit, my favorite part of Andrew’s post is the last bit:

Easterly also writes, “To a social scientist, the world is a big laboratory.” I would alter this to “observatory.” To me, a laboratory evokes images of test tubes and scientific experiments, whereas for me (and, I think, for most quantitative social scientists), the world is something that we gather data on and learn about rather than directly manipulate. Laboratory and field experiments are increasingly popular, sure, but I still don’t see them as prototypical social science.

4 thoughts on “Does peacekeeping work?

  1. I’d be interested to know what kind of controls she builds in for the scale of each peacekeeping mission. If you put 500 peacekeepers on the Macedonia/Serbia border, sure, that’s probably enough, but 17,000 in the Congo won’t delay conflict for a week.

  2. An excellent graph. There are at least two issues here though. It’s clear to me that peacekeeping keeps conflict away for at least a while. Whether or not it allows space and time enough for reconciliation, and the extent to which it affects domestic incentives for long run political settlement and development are the important things. Experience so far in Sierra Leone suggests yes; in eastern DRC, less so. Context is all.

  3. Easterly writes:
    “As far as ending the Darfur tragedy, the Darfur expert and veteran humanitarian worker Alex de Waal notes that the diplomatic energies of the United States and its allies have been consumed by the “clamor for UN troops,” that such a force would not be adequate to protect civilians anyway, and that the clamor is
    “diverting efforts from achieving a peace agreement that was within grasp…but has now slipped away.”[11]”

    Easterly wants to have it both ways: on the one hand to dismiss attempts at data analysis (“it is just correlation”), on the other hand to accept analyses such as those of the very knowledgeable de Waal, but which basically amount to an opinion. De Waal claims that a “clamor” both scuttled a peace deal between belligerent parties (so talk may be cheap, but “clamor” isn’t) and also “consumed” (as in exhausted) the negotiating abilities of the United States. What on earth would constitute evidence for weighing the merits of such a claim? Cell phone logs of the negotiators showing they spend 18 hours a day talking to “clamorers”, and a counterfactual where the clamorers were not around and a peace deal was signed?

    The “pro-expert, anti-clamorer” bias of de Waal is an opinion, and Easterly helps us precious little in determining whether that opinion is more or less relevant after conducting data analysis. Does Easterly think that cross-country analyses of effects of peace-keeping/intervention that do not measure and account for the endogeneity of “clamor” are deeply flawed? How does he propose to measure clamor? As these questions suggest, my view is that of course these statistical analyses are very problematic. But that doesn’t mean that an “expert’s” opinion is any less problematic.

    And then Easterly gets even more medieval, by throwing in Kuperman’s argument about “my goodness, if we have interventions a lot then rebels will not accept defeat so easily and might even start wars just to provoke interventions.” Gosh the world is a lot better off now that a theorist has “discovered” that insight! Proceed to the next paradox of fallacious equilibrium reasoning: With more interventions government budget deficits will be bigger, so taxpayers will elect parties promising fewer interventions, so more interventions will actually generate fewer interventions!

  4. I am confused. Clearly while peace keepers are present conflict is less likely. But I am not sure that is what is of interest. Rather the question is: given that peacekeepers where present for X years, what is the ex post hazard rate of peace?