(Washington Post illustration; iStockphoto)

Twenty-five years ago, civil war erupted in the coastal African nation of Sierra Leone. Civilians were massacred, mutilated and gang-raped; children were kidnapped and injected with drugs to make them fight. By the end of the conflict, which lasted more than a decade, more than half the population had fled their homes, and some 70,000 people had died, according to U.N. estimates.

A few years later, the Sierra Leone government conducted door-to-door interviews to gauge the rebuilding and reconciliation process. Surveyors asked people about their experiences with the war, about their current incomes, their participation in community meetings and their political engagement.

The war had shattered many lives. About 44 percent of people reported someone in their household had been killed. Yet a deeply surprising change had come over the survivors. Compared to their neighbors, those who had been exposed to more violence were also more trusting of others. They were more active in their communities and more likely to vote.

How could that be?

“We all expected these terrible adverse consequences for the community, for people’s trust in one another.” said Edward Miguel, an economist at the University of California, Berkeley, who traveled to Sierra Leone in the early 2000s to study the post-war recovery. “Instead, we were finding positive effects of war.”

At first, even Miguel was skeptical of his results. But in the past decade, 19 studies have come out documenting the same thing, not just in Sierra Leone, but also in Uganda, Burundi, Georgia, Israel and Nepal. The evidence has been adding up.

In a forthcoming paper for the Journal of Economic Perspectives, Miguel and his colleagues — a team including Harvard evolutionary biologist Joseph Henrich and University of Chicago economist Chris Blattman — went back and reanalyzed all the available data from these studies. Their conclusion is that the experience of wartime violence somehow changes people for the better, making them more cooperative and more trusting. And they have some theories explaining why.

None of the studies they looked at were true experiments, of course. Scientists cannot assign certain people to be pillaged. Instead, they relied on the fact that war itself often strikes randomly. In Sierra Leone, for instance, raiders would come through a town like a tornado, ravaging some homes but leaving others untouched. Rebel kidnappings in northern Uganda operated with similar arbitrariness. Nepal’s civil war cut an unpredictable swath of destruction through the nation — there was not much rhyme or reason to which villages got caught in the violence.

It’s important to recognize how horrific some of these conflicts were. Rebels in Uganda raped women and cut off their lips, ears and breasts. In Sierra Leone, armed parties routinely went on missions to hack off villagers’ hands and legs. One survivor who lost both his hands told aid workers:

Blood was spurting out of my arms. I was weak and kept falling as I tried to get up. They started laughing at me and I shouted, “just kill me, kill me, look at how you've left me." They spit on me and started pounding me and then several of them took a hammer, held me down and started knocking out my teeth.

In light of these atrocities, the resilience of the survivors seems all the more remarkable. People who suffered more violence during war not only reported more civic and political engagement on surveys conducted years later. They also behaved in more cooperative and altruistic ways when researchers tested them with laboratory games.

For instance — some studies paired up anonymous strangers, giving one person a sum of money that could be shared or hoarded. This is called "the dictator game" because the person with the money has all the power. In Sierra Leone, researchers found that people who had been exposed to more wartime violence were more generous — they shared more money in the dictator game than neighbors who hadn’t seen much violence.

Another kind of experiment tested people’s willingness to cooperate. In Nepal, villagers were put into groups where they could vote to cooperate or vote to be selfish. For each vote in favor of cooperation, the researchers gave all participants four rupees. If people voted for themselves, they received 20 rupees on top of whatever the group payout was.

For instance, if four out of 10 people voted to cooperate, then everyone received 16 rupees each. However, the selfish voters also received a 20 rupee bonus, for a total payout of 36 rupees. In this kind of game, it makes sense for individuals to vote selfishly no matter what everyone else does. People only vote the other way if they care more about the community than about their own well-being.

In Nepal, there was already a strong sense of unity within villages following its civil war. A study found that about two thirds of participants voted to cooperate overall. But the communal feeling was more intense in places that were hit harder by war violence, where 76 percent of people voted to cooperate — compared with 60 percent of people in places that were less directly damaged by the war.

When the team of researchers reviewed all these studies on war and cooperation, they noticed two interesting patterns. First, the positive effect of war didn’t seem to fade. Some of the studies had been conducted over a decade after the conflict in question had ended, yet they still found that the scars of war had made people more generous, more politically active and more likely to be leaders in their communities.

Second, some studies suggested there were limits to the increased altruism. Certain lab games found that exposure to war violence made people nicer toward members in their own village but not necessarily toward people outside their own community.

This is a topic that needs to explored further, Miguel said, but it lends support to a provocative idea about the origins of cooperation and generosity.

People don’t behave in perfectly rational, self-serving ways. We all feel, for instance, some kind of moral or social obligation to be generous toward strangers — as shown by psychological experiments conducted among many different cultures. This instinct for cooperation almost certainly contributed to the survival of our ancestors. We are much more powerful in groups than we are alone.

But there’s a slight problem with that idea. Suppose you lived in a community where everyone was genetically predisposed to be generous, but you had a mutation that made you selfish. You would prosper by living off of everyone else’s hard work, and eventually, your selfish offspring would dominate the community.

In other words, it seems that natural selection would extinguish any altruistic impulses in the long run — unless people were constantly under threat and could survive only if they banded together.

In recent years, evolutionary psychologists have argued that war may have played an important role in making us more cooperative. We know that humanity has a bloody past; constant conflict between different tribes would have extinguished any groups where people couldn’t work together or sacrifice themselves for the common good. At the same time, it doesn’t make sense to be indiscriminately kind toward others. If war made us nicer to our neighbors, it did not make us any more trusting of outsiders.

“[I]t seems likely that, for many groups and for substantial periods of human prehistory, lethal group conflict may have been frequent enough to support the proliferation of quite costly forms of altruism,” economist Sam Bowles wrote a few years ago in Science. “This might help explain why altruism often does not extend across group boundaries, and how this kind of ‘parochial altruism’ may have evolved in humans and perhaps even other animals.”

How does this explain the patterns of generosity among victims of war violence? The researchers speculate that our evolutionary history may have predisposed us to behave in certain ways following stressful or violent encounters with outsiders. In times of peace, we might jostle selfishly among ourselves, but in times of heightened conflict, we might instinctively become cooperative — and this trait may have helped our ancestors triumph over their rivals.

In countries that have suffered civil war, there's perhaps a note of caution here. If the consequence of war violence is to harden local solidarities, it may increase the chances that a nation will fracture again. But there's plenty of evidence that people can change their minds about who they consider an outsider. Rwanda's remarkable recovery from ethnic genocide owes itself in part to the government's campaign to emphasize a single national identity, which includes a controversial ban on any speech that would highlight divisions between the Hutus and the Tutsis.

The researchers emphasize that we have much to learn about how nations rebuild after a war. What is clear, though, is that trauma can strengthen, not weaken civic life. This may be a universal reaction to stress. Related studies show that survivors of violent crime or natural disasters also become more engaged in their communities.

We should not be so pessimistic, then, about "war-torn" regions of the world; we should not be so dismissive; nor should we be so surprised when countries beset by violence make miraculous recoveries.

"The data strongly reject the common view that communities and people exposed to war violence will inevitably be deprived of social capital, collective action, and trust," the researchers write.

These ideas should not seem so foreign. Folk wisdom has long held that conflict knits people together — centuries ago, for instance, the wars between European powers solidified national identities. The researchers also point out postwar societies have been more likely to pass laws strengthening the social safety net and gender equality. Humans might have an instinct for conflict and competition, but it seems we also have an instinct for rebuilding.

“While the human costs of war are horrific,” the authors write, “there may also be at least some reason for optimism once the violence ends.”