Awkward Utopia

This might appeal to UES political science junkies like myself. Titled “War and Political Instability,” my Friday morning session included three paper presentations, by Ed Glaeser of Harvard, Carlos Seiglie of Rutgers, and Daria Sevastianova of the University of Southern Indiana, respectively. I’ll overview each.

—

Mr. Glaeser’s paper poses the questions: if wars are so incredibly destructive, why are they also so common? He admits his thesis takes the view that there is an “agency” problem: wars help leaders while hurting the country as a whole. Consider a couple of models:

1. The Rational “Wag the Dog” Model: Here, war increases political stability and creates an incumbent advantage. Because the incumbent’s ability to wage war is a known variable (and the challenger’s military prowess an unknown or “random draw”), regime change while a country is fighting an outsider can be disastrous, and people know this. There are four “periods” here: (1) the incumbent decides to go to war, (2) the challenger decides to support or oppose the decision to go to war, (3) regime change either does or does not occur, and (4) conflict is resolved. He also begins with several assumptions: (1) that leaders prefer to govern during times of peace, (2) that voters have perfect information, and (3) that a challenger may not be able to reverse the costs of a war already in progress.

The incumbent is indifferent between war and peace at u* and the value of u* decreases with higher gains from being in power and the number of votes on the margin, and decreases with a higher cost of war and the leader’s non-war skills. The problem here is that while politicians may benefit, voters lose from wars. And even when they stick with the incumbent, they are aware of the costs.

2. The Munich Syndrome Model: This model adds a fifth period during which another country might decide to strike first. The results of this addendum are that wars become more attractive when the probability of a future war is higher, when another country is growing quickly, or the war is likely to do less damage. The problem here is that under this model, wars will only have popular support if going to war is what Mr. Glaeser describes as “quite sensible,” yet no war can be optimal for both sides. So how can we have two sides both deciding to go to war?

Mr. Glaeser then spoke about “hatred,” which he defines as a person’s willingness to pay to impose costs on others. For clarity he noted that racism and rate hatred are not technically the same thing (eg. hatred against women is actually very rare; gender discrimination is far more common). Now he allowed for the possibility that politicians can create false beliefs that lead to maybe-unjustified hatred. A great example is global anti-Americanism. Dislike of America, he said, doesn’t line up with the places we’ve “done bad things.” For example, Argentina is far more unfavorable to America than is Vietnam. Empirically speaking [and I think the club's Ron Paulians will find this especially intriguing], more military involvement on our part actually correlates with people liking us more!

Obviously, citizens hear a signal (”you should hate group x”) and then downgrade it based on how often they think such signals are false. Given these factors, the paper derives a couple of conclusions: (1) challengers who face a low cost of spreading misinformation and a low probability of success are more trigger happy, and (2) the difference between the value placed on leading a country during a war versus a time of peace is extremely important.

The major historical example here is Napoleon III. The Franco-Prussian war is often seen as a blunder; Mr. Glaeser suggests otherwise. Napoleon was in a position to lose power anyway if he did not act, and such circumstances will endear a leader to high-risk gambles. The situations during America’s wars in 1812, 1898, and in Vietnam are similar, and involve the profusion of much misinformation.

—

The second paper, Globalization and Arms, notes the inverse relationship between trade and conflict and seeks to explain it. Mr. Seiglie mentioned two possible explanations: (1) that trade leads to the proliferation of arms, which is a deterrent; and (2) that trade leads to increased wealth, which is a deterrent. He then delved into the former.

The amount a country allocates to defense (as compared to the amount its competition spends on defense) determines the country’s probability of retaining its resources r in a dispute. Through backwards induction, Mr. Seiglie was able to calculate a country’s optimal defense expenditures. Even more interesting was his realization that the amount a country allocates to defense does not depend on the country’s own endowments, but only on the other side’s. Within this model, a country could legitimately spend all of its resources on military might. The conclusions? First, that a greater degree of openness in trade does indeed lead to more money spent on weapons. And second, that the more a country’s enemy spends on arms, the more the country itself will spend on arms regardless.
—

Ms. Sevastianova’s paper (a joint project with Sol Polachek of SUNY) addressed the question of whether conflict disrupts growth. More specifically, they were looking at civil versus interstate war and its effects on developing versus developed nations. Their conclusions were ultimately that war is nearly always accompanied by a decrease in the growth rate.

1. The short-term effects of war were more dramatic than the long-term effects.
2. Both civil and interstate war cause a decrease in growth, though civil wars were more likely in poor countries and interstate wars were more likely in rich countries.
3. Fatalities sustained was a more powerful measure than was the length of the conflict.
4. Civil wars negatively affect all countries.
5. International wars especially affect developing (eg. African) countries.
6. Democratic countries were more severely affected by international wars and non-democratic countries by civil wars.

Seems pretty straightforward, right? But the paper makes a real contribution because it verifies that measures of the severity of the conflict, and not just measures of time (which is usually utilized as a sort of proxy for severity) are needed. For instance, while wars always negatively affect growth rates (see numbers 2 and 4 above), the same can not be said about other forms of instability (eg. assassinations, coups, revolutions). Much research until this point has overlooked severity as an indicator. This paper employs a fixed-effects framework incorporating variables like life expectancy, access to a body of water, and openness, and in so doing better accounts for the nature of the conflict, (thereby resolving, at least in part, this problem - see number 3 above).