In Heroes and Cowards: The Social Face of War, Dora Costa and Matthew E. Kahn explore the operation of social norms and social networks among American Civil War veterans. That's the big picture. More narrowly, the authors' primary objectives are to employ modern statistical techniques to “show who deserted his comrades and who was loyal, who survived POW camps and who did not, and how wartime friends influence the veterans' later lives” (p. 28). The authors achieve these objectives. I am confident the Cliometricians (a group inclined to cheer the mapping of elemental human traits, such as courage and cowardice, into the coefficients of a probit model) will be pleased.
After two introductory chapters—one reviewing the character-testing situations faced by soldiers in wartime and the other explaining the importance of the Civil War as a test case—the authors get to the heart of their investigation. Chapter 3 foreshadows the main findings, focusing as it does on the manner in which the War Department assembled fighting units. Some units were manned by troops from socially homogeneous groups, including those from the same town and even the same family; while other units were relatively diverse in their composition. This was important.
Chapters four through seven contain the volume's key empirical findings. The chapters contain analyses of, in order: the performance of black and white soldiers (i.e., who fought and who ran away); the survival of soldiers in POW camps; white soldiers' postwar experiences; and African American soldiers' postwar experiences. A key finding concerns the cost to the Union war effort of socioeconomic and demographic diversity in its units. Eliminating diversity from fighting regiments would have reduced desertion (the authors' measure of cowardice) by more than 75 percent in white units and by more than 80 percent in units manned by former slaves. In short, the social norm of courage (i.e., not deserting) is more likely to be enforced in more homogenous groups.
Qualitatively, POW survival mirrors the battlefield experience. Not being part of a well-defined group could be deadly. Using survival probabilities as an indicator of POW “success,” the authors find that white soldiers who surrendered with ten or more colleagues had a 20 percent greater probability of surviving four months in a Confederate POW camp. In the notorious Andersonville Prison, survival rates of men captured alone or in small groups fell below 50 percent after eight months in captivity. A conclusion is that the social norm of virtue (e.g., aiding a fellow POW's survival) is more likely to be enforced among a group of colleagues who entered the camp together than among strangers.
With respect to the postwar experiences of deserters relative to non-deserters, the authors report that “[d]eserters were more likely to move, and when they did move, they moved farther away” than non-deserters (p. 175). Experiencing, or expecting, the enforcement of social norms, deserters left town or never returned after the war.
One situation in which unit diversity had a positive impact was the postwar experiences of African American soldiers. Newly freed slaves who found themselves in units manned by a high percentage of free blacks were more likely to learn to write, and blacks who served in diverse units were more likely to migrate (in response to better economic opportunities, one assumes) after the war.
The authors have chosen to omit most of the details of how they obtained their results. This choice probably reflects the fact that those details can be found in a series of related articles, by the authors, published over the past few years in a number of journals. It might also reflect a quest to broaden the volume's readership. That strategy is a double-edged sword. Cliometricians might just go straight to the articles for their statistical fix; while the statistically uninitiated might scratch their heads and ask: “Now, how do the authors know that, exactly?”
One lesson from this volume is the importance of institutions in the enforcement of social norms. The War Department's recruiting strategies, which varied over time and space, determined a unit's composition. Assuming one accepts the authors' definition of courage as NOT deserting, then one can conclude that the most courageous, and thus most effective, fighting units were the most cohesive, and the most cohesive were those composed of soldiers drawn from the same family, neighborhood, town, and/or socioeconomic group. On the battlefield, diversity appears not to have been an asset, though it did have a positive impact on the postwar experiences of African American veterans.
Of course, there is another view of valor under fire, one the authors themselves consider. It might simply be the case that no matter how one manipulates the variables, on the battlefield, “[c]ourage is not for everyone” (p. 74).
