Thread by @ProfPaulPoast

You've heard the quote, "a single death is a tragedy; a million deaths is a statistic."

What if international relations scholars took that quote seriously? What if we brought tragedy back into our conception of war? What would we learn?

[THREAD]

I am referring to how studies of war, especially interstate war, use "fatality thresholds" to decide if a conflict is a war. For example:

- 1,000 battlefield deaths (COW)

- 500 deaths (IWD by @mchorowitz)

- 250 deaths (MARS by @jaylyall_red5)

- 25 deaths/year (@UCDP)

There are issues with such thresholds, as @tanishafazal pointed out in her @Journal_IS 2014 piece

mitpressjournals.org/doi/abs/10.116…

A fundamental problem is that such thresholds can lead to perverse/dehumanizing discussions such as "well, ONLY 900 soldiers were killed in this conflict, so it doesn't count".

Good example is the Falkland war (🇬🇧v🇦🇷), where "just" over 900 died.

nytimes.com/2018/03/28/wor…

Moreover, these thresholds can lead to "fudging" of the numbers.

For example, guess which "# of battle deaths" appears the most in the COW dataset? Yep, exactly 1,000.

But what if ANY conflict that resulted in ANY deaths was counted as a war? That would mean, for example, that the latest "standoff" between China and India at the LOAC is actually a war

bbc.com/news/world-asi…

If we were to include such conflicts in our conception of "war" how would it change what we know/claim about war?

For instance, would the "Long Peace" claim still hold: that there have been no wars b/w great powers since 1945?

jstor.org/stable/pdf/253…

One way to gain a sense for how our understanding of war would change is to look at the Militarized Interstate Disputes (MID) data collected by the Correlates of War Project

correlatesofwar.org/data-sets/MIDs

MIDs include wars, but also levels of military conflict below war, including those that do not result in fatalities (such as threats to use force).

Not all MIDs (indeed, most) result in fatalities

https://twitter.com/jljzen/status/1047296954814750725

But some do, as is captured by the "fatality" variable in the dataset

As this tab shows, the VAST majority of MIDs result in no deaths.

But the next highest # is fatality=1 (b/w 1 and 25 deaths), with 213 conflicts. Notice that fatality=6 (which is for MIDs at the COW "interstate war" threshold) only has 87 conflicts.

So lowering the threshold will DRAMATICALLY increase the number of wars we consider.

How will that alter our inferences about war?

Consider this chart. The maroon bars are the number of "Traditional >=1000 deaths wars" that started each year from 1816 to 2010.

The dark blue bars are the number of "Fatal MIDs" that started each from 1816 to 2010.

VERY different inference regarding the trend in war.

Moreover, some of these additional "wars" would indeed involve major powers on opposing sides. A good example is the 1970 “Black September” Crisis, a fatal “militarized dispute” b/w USA-Israel-Jordan on one side, and Soviets-Iraq-Syria on the other (it's MID # 1039).

Most of the deaths in this "war" were incurred by Jordan and Syria (notice below that it seems the "dispute" *might* have actually reached "war" threshold)

source: en.wikipedia.org/wiki/Black_Sep…

Of course, whether someone dies in conflict shouldn't be the ONLY factor we consider in classifying an event a war. We might need to consider:

- duration (more than a day)
- level of organization of forces
- intentions (was it directed toward a political aim?)

But that's exactly the point: THOSE are the factors that should drive our consideration of whether a "conflict" is a "war", not whether a "sufficient" number of soldiers died.

In sum, treating seriously conflicts where any number of deaths occur could change our understanding of war and the debates we have about the causes and trends in war.

[END]

Addendum/Correction: Project MARS uses a 500 threshold, similar to IWD (though it uses a broader conception of state actors, which is useful). Forgot to correct that tweet before posting. cc @jaylyall_red5.