Wednesday, February 22, 2006

Casualty-sensitivity of the demand for war

In the January/February edition of Foreign Affairs, John Mueller and Christopher Gelpi squabble over the determinants of US public support for war. Gelpi believes that wars are supported by the public when the perception of military success is high. Mueller believes, to the contrary, that the main factor impacting on support for wars is casualties - support declines as casualties mounty. The exchange follows a piece by Mueller in an earlier issue attacking Gelpi’s general thesis.

Mueller notes that public support for the Iraq war has followed the same course as wars in Korea and Vietnam. Initially there was public enthusiasm but then, an erosion of support as casualties mounted. He implies that President Bush will be unable to reverse this deterioration short-term to stave off an ongoing anti-war ‘Iraq syndrome’ that might inhibit US foreign policy for decades.

Gelpi responds that while public opinion is concerned with casualties that there is no implication of policy-making paralysis due to the Iraq war experience. He claims the public become more sensitive to casualties during different phases of wars and that a key influence is the probability that a mission will succeed. Thus the public is ‘defeat-phobic’ not ‘casualty-phobic’. His data from the Korean, Vietnam and Iraq wars supports this claims. For example, after the Tet offensive in Vietnam, the public lost hope and casualties became important. He also argues that acceptance of casualties, since the Vietnam War, has declined because technology has reduced, in the public’s mind, levels of 'necessary' casualties. In other words Americans are more fussed about casualties these days.

Mueller responds by pointing out that support does drop during the early stages of a war as ‘weak’ war supporters drop away. But given this, the pattern is still that support declines as casualties mount. He disagrees with Gelpi’s analysis of Tet arguing, instead, that most US support for the war had faded in the two years before the US indignity due to high casualties. He also points out that that the defeat-phobia attitudes analyzed by Gelpi applies (using Gelpi’s data) to only that 20% of the population that is sensitive to ‘setbacks’.

Mueller does not believe that high casualties have permanent or ongoing effects on the public's support for war. The effects are temporary. For example, if the stakes are high enough Americans will accept many deaths. His claim is that in Iraq the stakes were not high enough to justify the high number of deaths.

From an econometric viewpoint one would need here to be able to distinguish the effect on public support for war of casualties and reduced winning prospects. These variables will be collinear. One needs to think of situations where casualties were low but prospects of winning were poor. Here Mueller’s argument would make sense if public support for war was high. If support for war was low then Gelpi’s analysis would work better. My guess is that Gelpi is right - with strong enough motivation the American public will support wars even with a recent history of high casualties. For example, I think that despite the difficulties in Iraq, the US and/or Israel are likely to launch a strike on Iran's nuclear facilities. The stakes are high enough in this case to ignore casualty history.

I am interested in the effects of one conflict on the propensity and ability to engage in another. I notice for example that the Bilmes & Stiglitz study of the costs of the war in Iraq (famously claimed to cost up to $2 trillion, see here) mentions, but does not include, the costs that the war imposes in terms of reduced capabilities to engage in other wars both now and in the near future. B & S are talking about resource constraints while Muller is talking about constraints imposed by public opinion.


P.A. Coplay said...

One argument in favor of the public being less inclined to accept casualties in developed economies is that specific children may be more valued now (investment in a small number of "high quality" children) than perhaps in the past. To make myself clear, it is not that people necessarily love their children more, but the probability of survival of children was lower As late as the second world war, if young men did not die in wars, the death (and injury) rate in childhood diseases was much higher than now, workplace mortality was higher. As well as their being less wars, there has been extensive regulation to attempt to reduce deaths in the workplace, and to increase the quality of life of those children whose quality would have been much poorer sixty years ago.

Hence the greater concern about casualties than before.

hc said...

That is a perceptive observation and sounds right. People have fewer children, invest more in their 'quality' (in Becker's sense) and so value them more.

Maybe a positive thing too, As kids become more valued, less war.

Stiglitz and others hacve written of the extraordinary inability to secure recruits into the US military. We have the same problems here. Yet salaries and conditions have never been better.