[Dora Costa and Matthew Kahn, guest-blogging, January 14, 2009 at 12:20am] Trackbacks
Social Networks during War Time:

In our new Princeton University Press book, Heroes and Cowards: The Social Face of War, we examine the war experience for Union Army soldiers. We weave a single narrative from the life histories of 41,000 Union Army soldiers, diaries and letters, and government documents. Our core questions are not those typically asked in a military history. When are men willing to sacrifice for the common good? What are the benefits to men of friendship? How do communities deal with betrayal? And what are the costs and benefits of being in a diverse community?

In this post and our next post, we would like to provide an overview of some of our key findings. For readers who enjoy applied econometrics, we encourage you to go Dora Costa's UCLA website where you can download our key academic papers.


A soldier who sought to survive the U.S Civil War should have deserted and roughly 200,000 Union Army soldiers did (about a tenth of the army). Out of the roughly 80,000 men who were caught, 147 were executed. Those who stayed faced a death rate of 14 percent, with half of the deaths from wounds and half from disease. In contrast, during World War II, Stalin's armies had special detachments that formed a second line to shoot at any soldiers in the first line who fled and the families of all deserters were also arrested. Out of the roughly 35,000 German soldiers tried for desertion by the Third Reich, about 22,750 were executed. Democracies cannot inflict such punishments. Lincoln recognized that "you can't order men shot by the dozens or twenties. People won't stand it."

Given these facts, is it surprising that only a tenth of men actually deserted? We argue that desertion is a great measure of "community participation". In a group of 100 men, if one man deserts the army he raises his probability of survival but puts his fellow men at risk to be crushed by the enemy. Unlike in the modern corporation with bonus pay and pay for performance, the diaries these men left makes clear that they were fighting for each other.

We use our unique longitudinal data to document several facts about the determinants of desertion. Here we focus on our most salient findings. We encourage you to read our book to learn all! In what follows, please keep in mind that these findings are based on multivariate statistical analysis so we are holding all other factors constant and varying one explanatory variable at a time. When the Union Army was winning battles, desertion rates declined. Just like in professional sports, everyone loves a winner. Desertion probabilities were higher in more diverse war companies. Turning this statement around, men who fought in more homogenous war companies (based on occupation, age, and place of enlistment) featured lowered desertion probabilities. A generalization of this finding is that people are better citizens in social settings when their community looks like them. In our next post, we will relate this finding to ongoing social science research on the costs of living in a diverse community.

Surviving POW Camps

Many of us have enjoyed watching Hogan's Heroes on television. While Bob Crane outwitted Clink and Sgt. Schultz in his WW2 Nazi POW camp, U.S Civil War soldiers sent to Andersonville had a lot less fun. An estimated 211,411 Union soldiers were captured during the Civil War. Seven percent of all U.S. Civil War soldiers were ever imprisoned compared to 0.8 percent for World War II.

Civil War POWs suffered from poor and meager rations, from contaminated water, from grounds covered with human excrement and with other filth, from a want of shoes, clothing, and blankets (having often been stripped of these by needy Confederate soldiers), from a lack of shelter in the open stockades that constituted camps such as Andersonville and Millen, from the risk of being robbed and murdered by fellow prisoners, and from trigger-happy guards. Our data show that at Andersonville, the most notorious of the POW camps, roughly 40 percent of all men of who passed through the camp died, and half of the deaths occurred within three months of entry. The chief causes of death were scurvy, diarrhea, and dysentery.

How did men survive such horrific conditions? The accounts of survivors provide some clues, as do the accounts of survivors of Nazi concentration camps, the Soviet gulag, and Japanese and Vietnamese POW internment camps. But some accounts conclude that death is random; others emphasize psychological defense mechanisms; others emphasize the importance of leadership; and still others emphasize the role of friends. We can use our longitudinal data and a data set on almost the entire population of Andersonville put together by the Park Service to examine the effects of age, social status, rank, camp population, and the presence of own officers on survival.

The single most important determinant of camp survival was how crowded the camp was. Another important determinant of camp survival was age. Those of higher rank fared better, as did those with useful skills. Men with officers from their own companies were more likely to survive than those without or with fewer officers.

Holding these factors constant, social networks within the camps increased a soldier's survival probability. We can establish this because we know each POW's war company and home town. We document that men who were in the camp with "more friends" had higher probabilities of survival then men with similar demographics who were in the camp with fewer friends. Ties between kin and ties between comrades of the same ethnicity were stronger than ties between other men from the same company.

Why did friends increase the probability of surviving POW camps? Did friends provide extra food or clothing, tend to the sick, protect against the predation of other prisoners? Or did simply having a friend have a positive effect on men's immune and endocrine systems? Monkeys randomly assigned to stable or unstable social conditions and inoculated with the simian immunodeficiency virus face shorter lives if they live in unstable social conditions conditions. We cannot run such tests on humans. But, in our final post, we will discuss some intriguing evidence for how social networks can cushion psychological shocks. We would like to conduct similar survival research based on records from the Nazi holocaust camps. We have not been able to identify credible network measures (analogous to our War Company identifiers) to be able to establish who knew who within these camps.