Friday, 21 August 2015

"At the end of the day, everyone was a Ranger"

Two women, First Lt. Shaye Haver and Capt. Kristen Griest, graduated the US Army’s Ranger School at Fort Benning yesterday. That doesn’t mean they’ll be joining the 75th Ranger Regiment. The two are institutionally separate, and combat arms remain closed to female soldiers. But they’ve still passed a course showing their ability to lead a small unit in brutal conditions. It is a tremendous accomplishment.

Their progress through Ranger School is a pretty interesting window into a persistent question in military sociology: the relationship between social cohesion, task cohesion, and unit performance. The concern that women would undermine fraternal bonds among male soldiers, or social cohesion, is often cited as a reason to bar women from combat, along with their supposed inability to meet the physical rigors of the job. Skeptics raised similar objections to integrating openly gay and lesbian soldiers, and for that matter to black soldiers, in earlier eras.

But I think a very good way of thinking through this issue is through costly signals. In a recent article about desertion published in International Studies Quarterly, I argue that when soldiers send costly signals of their commitment to their unit-mates (such as volunteering for service rather than being conscripted), they build trust and desert less often.

Passing through Ranger School is a hugely costly signal – far more than just volunteering. It is brutal. And in forcing these costly signals from its candidates, Ranger School itself seemed to undermine social differences among them. By showing they were capable of enduring and succeeding in the same challenges as their male peers, 1st Lt. Haver and Capt. Griest showed what they could do and overcame stereotypes about women’s physical capabilities. Here are some of their unit-mates quoted in the Army Times:

    "I was ignorant and assumed that because they were women, it was going to be harder for them," said 2nd Lt. Zachary Hagner, who was Griest's Ranger buddy for much of the course.

    During the last day of the Mountain Phase, Hagner said he had been carrying the squad automatic weapon for three days. Exhausted, he asked his teammates if they would help him out and take over the load for a while.

    "Everyone said 'no,'" Hagner said. "But [Griest] took it from me. She, just as broken and tired, took it from me, almost with excitement. I thought she was crazy for that, but I guess she was just motivated."

    Haver was the same way, said 2nd Lt. Michael Janowski, an infantry officer and Haver's Ranger buddy.

    Janowski said he also struggled with extra weight during the Mountain Phase, and Haver "was the only one who volunteered" to help, he said.

    "I probably wouldn't be sitting here right now if not for Shaye," he said.

    Second Lt. Erickson Krogh, an infantry officer, said his skepticism was "smashed pretty fast."

    Griest and Haver almost looked embarrassed as their classmates praised them for their efforts during Ranger School. "I got to know a few of the other females [in the course], they're absolute physical studs," he said. "When I completely believed they would run into physical walls, that was never the issue."


    After a while, the students said, gender didn't matter.

    "When you're out in the field and you're starving and you're tired and you're cold and you're miserable, you really don't think about gender," said 2nd Lt. Anthony Rombold, an armor officer.

    Staff Sgt. Michael Calderon, a sniper section leader from Fort Carson, Colorado, agreed.

    "You're way too tired and way too hungry to honestly care," he said. "At the end of the day, everyone was a Ranger. It didn't matter, as long as the team pulled through and accomplished the mission."

Look at the process of learning. The other candidates learned what Griest and Haver were capable of, and gender stopped mattering in the teams: “everyone was a Ranger.”

Ranger School is a classic costly signal in that it separates, brutally, those who have a certain level of physical fitness and mental and emotional toughness from those who do not. (Were the standards lowered for the female candidates? It seems not: most of their female peers washed out. Maj. Jim Hathaway, a senior Ranger School officer, pushed back strongly on any suggestion that the standards had changed. Indeed, recognizing the concerns that would emerge from an accusation of favorable treatment, senior officers apparently stayed away from patrols involving female candidates, to avoid any notion that top brass was trying to secure their success.)

With costly enough signals of your commitment and your capabilities, soldiers can establish task cohesion – trust in their unit-mates’ efforts towards a common aim – regardless of initial social differences. This suggests a contingent relationship: at high levels of demonstrated commitment, social homogeneity doesn’t matter for intra-unit trust and performance.

This, and the similar experience of DADT repeal, adds a new dimension to what we know about task cohesion and social cohesion. Other research has also shown contingent relationships between task and social cohesion, but with different results. My article shows that social homogeneity mattered more in reducing desertion in volunteer units than in conscript units. Similarly, Peter Bearman found that among North Carolina troops in the American Civil War, more socially homogenous units had lower desertion rates at the start of the war (when things were going well for the South), but higher desertion rates by the end of the war than heterogeneous units (when everything was falling apart). My work, and Bearman’s, suggest that social homogeneity helps military units more among the relatively committed and enthusiastic than among the uncommitted and those who don’t want to be there. If nobody wants to fight (either in a losing Confederacy or among conscript units in Spain), it doesn’t matter if you trust each other—in fact, trusting each other can help you desert together. These results stand against unambiguous, unqualified claims that it’s social cohesion that keeps a unit fighting – a long tradition in military sociology. Social cohesion can’t rescue an uncommitted unit.

But aren’t my findings, and Bearman’s, also in tension with the experiences of Griest and Haver? For them, training among the most committed, the toughest mentally and physically, washed away social differences – they didn’t matter anymore. Based on their experience, shouldn’t we expect social homogeneity to matter less, not more, in more committed units?

But I think costly signalling theory helps us figure out the apparent paradox. As I note in the conclusion to my article, the upper end of commitment in my study and in Bearman’s is still not extremely highly committed – or at least they don’t show it through costly signals. Volunteering for service is a long, long, long way from Ranger School – in an all-volunteer army, only a very small number of soldiers even try to get through Fort Benning, and only a fraction succeed. Bearman and I just don’t cover the full range of possible levels of commitment and costly signals that it’s possible for a soldier to display.

Putting all of this together, it looks as though social homogeneity promotes unit cohesion only at middling levels of task cohesion. It can help reinforce a sense of trust that’s potentially there, but needs a boost. It’s useful for startup armed groups of volunteers to build on social networks. But social homogeneity won’t rescue a group of conscripts who don’t want to be there, and simply won’t matter among the toughest of the tough. Capt. Kristen Griest and First Lt. Shaye Haver, and their Ranger unit-mates, have shown that it isn’t an excuse.

Monday, 29 June 2015

"It is no longer a reproach to be known as a deserter": Desertion in the Civil War

I’m dusting off the old blog because I got a bunch of new followers Friday who seem like they're interested in desertion during the American Civil War. (That's what happens when you get retweeted by someone like Ta-Nehisi Coates.) David Brooks had talked about the dilemma of Robert E. Lee: valiant, but a slave-owner, fighting a treasonous war in defense of slavery. What does America do with him, revered as he is? Coates made this great point about deciding whom to venerate:
So I responded:
It’s easier to ignore that the South fought, manifestly, for slavery, when you focus on how brave its troops were. But to do so, it helps to ignore desertion rates.

(I neglected to point out that it isn’t just the Lost Cause. Emphasizing bravery and downplaying desertion was one way that all of white America came to grips with the Civil War while failing to do very much about responding to its origins. Sentimentality about bravery helps you think past what the war was about. By 1876, forgetting about what the war was about was something North as much as South was happy enough to do. I’m overgeneralizing; but it’s noteworthy that the first book-length study of desertion in the Civil War that I'm aware of took over sixty years to appear.)

In any case I thought I’d write a little bit about desertion in the American Civil War. I’m not an expert; my work is on the Spanish Civil War. But it’s hard to ignore the Late Unpleasantness. There is more extensive research on desertion in this war than in probably any other civil conflict. So, keeping in mind that others know way more than me, here are some basics and further reading.
To begin with it’s hard to get a fix on just how many deserters there were. It’s hard to pin down the difference between surrender, capture, going AWOL temporarily with the intention of returning, and desertion. There were probably general incentives to under-report desertion, telling commanders what they wanted to hear. It also means that the accuracy problem was probably worse for the South than the North, because as it’s losing, there’s more surrender, more capture, and—probably—more desertion. 

Estimates based on official records are comparable for the two sides: 9-10% among white Union troops (see pp. 92-93 here), less than 8% among the US Colored Troops (p. 112 here), and around 10% for the Confederacy (p. ix here gives the figure of 103,400 deserters). On both sides, that’s about half of the total number of dead and wounded – so it was a major source of troop losses.
But keep those caveats in mind. An army historian writing in 1920 estimated desertion in the CSA at 45%, but didn’t document the estimate (p. 4 here). 

So, what kept soldiers fighting? A righteous cause, the barrel of a gun, or the "band of brothers" of a unit? I find a useful way of thinking through the phenomenon is in the norms of military units. Do the soldiers around you want to fight, or not? Let’s say you wanted to fight, in principle. If you couldn’t trust your unitmates to have your back in combat, would you? If it was "no longer a reproach to be known as a deserter," as one South Carolina officer wrote in August 1863, desertion became more thinkable (p. 29 here). Or, on the flipside, let’s say you really didn’t want to be fighting. Social pressure might make you join up in the first place, or you could be among the 10% of Union troops or 20% of Confederates who were conscripts. But once in your company, if everyone around you is gung ho, you might keep fighting because you don’t have much choice.

So where did norms come from? I think a good framework is the sense of accomplishing something together, underpinned by a sense of community. If soldiers believed that their unit-mates were committed to the fight, they could trust that their unit-mates would have their backs and be more willing to endure. So as the war was ending, desertion became an epidemic in the CSA as fighting seemed futile, hardships at home ramped up, and many soldiers’ homes were now behind Union lines.

Trust could be easier to establish with social ties—coming from the same hometown or similar occupations. Among (white) Union volunteers, for example, desertion rates were considerably lower in companies where the troops came from the same state, or had the same job, or were similar in age.
But social homogeneity, helping you see your unit-mates as buddies, didn’t always help. Sometimes it hurt, quite seriously. Take North Carolina, which really was “first in flight,” as Katherine Giuffre once joked. At 13%, it had the highest desertion rate of any Southern state. Peter Bearman found that Tar Heels in homogenous units deserted less often at the start of the war, but more often by the end. Basically, in 1861, groups of hometown buddies fought together. By 1865, they often deserted together. Social ties need to be linked to commitment to a task or a project. 

So the “band of brothers” idea doesn’t mean much for desertion and battlefield endurance unless there’s a common sense of doing something together. This “something” need not be ideological, or part of the master narrative of a conflict. Civil wars connect lots of little political aims up with big ones. It need not be the same thing for every in a unit (though organized factional competition definitely drives up desertion rates, as my research in Spain shows). It doesn’t need to be very sophisticated—simply a sense of, do my unit-mates care whether the unit succeeds or not?

(Side note: lots of proponents of racial segregation and, later, of Don’t Ask, Don’t Tell in the US military have argued that social heterogeneity undermines unit cohesion. Does my argument mean I agree with them? No, actually. I think that if soldiers’ signals of commitment are strong enough – not only voluntarism but also going through rigorous training and constant opportunities to show their unit-mates what they are made of – then social homogeneity doesn’t matter. The aftermath of DADT repeal makes this pretty clear. In the American Civil War, raising armies rapidly and in desperation in a divided country, really clear signals of loyalties could not really be clearly sent, and social homogeneity could make more of a difference.)

There were other dimensions of desertion in the American Civil War. One I’m particularly interested in is hill country. Hill country soldiers were lots more likely to desert. Katherine Giuffre’s “First in Flight” doesn’t just have a fantastic title, it also shows this pretty clearly in the case of North Carolina troops. Giuffre argues that this was because they were less connected with the lowland slave economy, so they had less skin in the game.  That is plausible, but I think another reason might be just that it was easier to hide in the hills. I make that argument in the case of Spain here.

There’s lots more work to be done, and with so many proximate causes to leave—a desperate letter from home; hating one’s officer; horrible medical care; Lincoln just issued the emancipation proclamation, alienating white soldiers who do not want to fight for black people—the literature is immensely rich. Here are a few things to read. Ella Lonn’s Desertion During the Civil War was the first book I know of on the subject (1928), and it’s still packed with insights. James McPherson’s book For Cause and Comrades is terrific on ordinary soldiers’ combat motivations. Mark Weitz’s More Damning than Slaughter is a great history of desertion in the Confederacy. Though I think their main finding needs to be contextualized a bit, Dora Costa and Matthew Kahn’s Heroes & Cowards is truly magnificent, based on a study of over 40,000 soldiers in 300 Union volunteer companies. There is so much in there: how deserters were ostracized when they went home, how unit buddies banded together in the horrors of POW camps. Peter Bearman’s “Desertion as Localism” is a classic piece of insightful, rigorous, and fascinating sociology.