Blanche Dubois may have famously depended on the kindness of strangers in the closing scene of Tennessee Williams'A Streetcar Named Desire, but in a world fraught with treachery and deception, most of us prefer to depend instead on the kindness of friends and family to help us through hard times, and reciprocate when called upon to return the favor. But are there limits to what we'll do to sustain the bonds of friendship?
In their new book Heroes and Cowards, economists Dora Costa and Matthew Kahn use the Civil War as their laboratory to study what men will do in the name of friendship. They find that men serving in companies with tight social connections—like shared birthplace and occupation—were more likely to stand and fight than those in less tight-knit companies, where desertion rates were up to four times higher. The bonds of friendship also mattered for Union soldiers who ended up in Confederate POW camps: Soldiers imprisoned with others of similar backgrounds were much more likely to survive to see the war's end.
When economists look at friendship and social networks, what they see is people trading favors—you scratch my back and I'll scratch yours. A friendship's value is determined by the benefits of favors you receive weighed against the cost of the favors you'll need to do in return. A friendship built on cold economic foundations can be sustained only as long as the gains of the long-term trading of favors exceed the benefit of taking one last back scratch before putting an end to the relationship (though news travels fast, so retaliation from others in your social circle may help to keep you from taking advantage of others).
Of course, for all but the most calculating of individuals, there's more than naked expediency to friendship—we help others because we care about their well-being rather than what it gets us in return. At an extreme, this concern for others can lead us to do things like giving up kidneys or falling on hand grenades to save our friends—acts that can't be good for our long-term health. So friendship can help us in our daily lives but can also prove very costly because of the extreme sacrifices that emotional bonds may inspire.
Much of the evidence on the nature of favor-giving among friends examines situations that aren't exactly life-or-death. Costa and Kahn look at the larger stakes decision of whether Union soldiers chose to risk death by remaining to fight or desert and save their own skins. The authors reason that social bonds are stronger among soldiers from similar backgrounds. New England-born soldiers, for example, will feel greater kinship with other New Englanders, the Irish with other Irish, blacksmiths with other blacksmiths. In companies where men had shared backgrounds, fewer soldiers would be expected to abandon their comrades, both because of the greater kinship among men with a sense of social connection and because their shared network would make it easier to punish and censure cheaters back home.