Say the word "evolution" and many people hear the word "selfish". From Tennyson’s "nature red in tooth and claw" to Dawkins’ "selfish gene", Darwin’s theory seems to explain the dark side of nature better than the bright side—even to the point of suggesting that there isn’t a bright side.

Yet, selflessness can evolve just as selfishness can. To see how, I decided to study selfishness and selflessness in my hometown of Binghamton, a small city in upstate New York. I began by enlisting the help of the public school district to give a survey to the students in grades 6-12. The survey measured their degree of prosociality, defined as any attitude or behavior oriented toward the welfare of others, with items such as "I think it is important to help other people" and "I am helping to make my community a better place". The survey also asked the students about how much social support they received from their family, neighborhood, school, religion, and extracurricular activities. The correlation between the prosociality of the individual and the prosociality of their social environment was an astonishing 0.7. At least according to their own report, kids in the city of Binghamton who gave were also getting—the basic condition for selflessness to succeed as a Darwinian strategy.

By attaching the students’ residential locations to the survey data, we were able to create geographical maps of prosociality for the city of Binghamton. If students who scored high and low were evenly distributed throughout the city, the maps would appear like a flat plain. Instead, there is a rugged topography of hills representing neighborhoods where most of the kids are highly prosocial and valleys where most of them express little interest in others or their community. The social clustering that we demonstrated with our correlation coefficient is based in part on an impressive spatial clustering.

We have validated our survey results with a variety of different methods, such as door-to-door surveys of adults, experimental economics games played for money with the same students who took the surveys, and stamped addressed envelopes dropped on the sidewalks of the various neighborhoods to see who is kind enough to mail them. We’ve even measured naturalistic expressions of prosociality, such as Halloween and Christmas decorations. The hills and valleys of prosociality are real; the same social forces that select for selfishness and selflessness throughout nature are playing themselves out on the streets of Binghamton.

Yet people are not like Mendel’s peas. Our behaviors are influenced in part by our genes, but in many respects we are like chameleons, capable of matching ourselves to our immediate environments. Imagine that you’re a selfless person who finds yourself surrounded by selfish people. You have exactly four choices: 1) you can try to leave; 2) you can try to change their behavior; 3) you can turn off your own selflessness to protect yourself; 4) you can continue behaving selflessly and suffer the consequences. A smart chameleon will avoid option 4. In this fashion, the kids who express little interest in helping others or their community weren’t necessarily born that way—they could just be responding to the environment around them.

Then there is the question of how long they have been responding to their environment. Take two identical twins, separate them at birth, and have one raised in a nurturing environment and another raised in an abusive environment. When they are teenagers, place them both in a nurturing environment. Will they respond in the same way, chameleon-like, or will a lifetime of abuse prevent one of them from responding to a nurturing environment, even when it has become available?

Slowly, we are beginning to disentangle the short-term and long-term environmental effects on the expression of prosociality in the city of Binghamton. In one set of experiments, my former student Dan O’Brien—now at Harvard employing a similar approach for the city of Boston-- showed photographs of the various neighborhoods to college students who knew little about Binghamton. Their assessment of neighborhood prosociality based on the photographs corresponded to the assessment of people who actually lived in the neighborhoods.

Then Dan gave the college students an opportunity to play an experimental economics game for real money with actual residents of the neighborhood. The less prosocial the neighborhood as assessed from a photograph, the less likely the college students were to choose the cooperative option in the experimental economics game. In other words, just like chameleons changing their color, the college students were calibrating their degree of cooperation to the perceived quality of their social environment. Dan calls this "community perception" and compares it to the way that people quickly assess other people.

Another opportunity to measure how people respond to changes in their environment presented itself when we gave the survey a second time, three years after giving it the first time. Several hundred students took the survey both times. Of these, some had moved their residential location within the city of Binghamton. Some had moved "uphill" to a more prosocial neighborhood, and others had moved "downhill" to a less prosocial neighborhood. We were able to measure whether the student’s own prosociality changed in response to the environmental change. It did. Parents everywhere will be happy to learn that the prosociality of teenagers is not set in stone.

Just as so many people equate "evolution" with "selfish", they also equate "problems" with "poverty", but the relationship between poverty and prosociality is not so simple. People with lots of money don’t need to cooperate with their neighbors to get what they need. They can pay with a credit card. People without money have a greater need to cooperate. They’re the ones who organize car washes, bake sales, and so on. Of course, poor people in neighborhoods where no one cooperates have the worst of both worlds. Our research shows that the relationship between poverty and prosociality is complex, with the most prosocial kids tending to come from neighborhoods with modest incomes but high social support.

Once we see selflessness as a strategy that can win the Darwinian contest on our own streets, we can begin to think about changing our current environments to stack the deck in favor of selflessness. We have a number of social engineering projects taking place in Binghamton. One of the most challenging is a program for at-risk high school students called the Regents Academy. The public school district helps us to give the survey, so helping them design the Regents Academy is the only selfless thing to do.

Improving the academic performance of at-risk students is difficult at any age, but especially for at-risk teenagers, who are thoroughly adapted to living in harsh environments. The Promise Academy, a school associated with Geoffrey Canada’s famed Harlem Children’s Zone, began with a first and sixth grade class, as recounted by Paul Tough in Whatever it Takes. Intensive efforts to improve academic performance, based on the same educational principles, succeeded for the 1st graders but failed for the 6th graders. The Promise Academy has since improved its success with the older students, but only with a full court press that includes an extended day, extended school year, meal and healthcare programs, and so on. Other successful schools for at-risk teenagers, such as the KIPP academies, are similarly intensive.

For the Regents Academy, there would be no extended day or year and no cradle-to-career resources outside of school that the Harlem Children’s Zone provides. Just a dedicated principal and staff of four teachers, working with the meager resources that a typical city public school system can provide. To qualify for the Regents Academy, students had to have flunked at least three of their courses during the previous year. Their lives were tough, even heartbreaking after we got to know them.

Working with Rick Kauffman, one of my current graduate students, we designed a school environment informed by evolutionary theory, not just educational theory narrowly conceived. Using what we knew from evolution, we stacked the deck in favor of selflessness. We created a strong group identity. The program was held at a single location so the students didn’t wander from class to class. We allowed the students to participate in the decision making as much as possible. We were abundant in our praise and mild in our punishment—but we were prepared to ratchet up the punishment if necessary. The principal was a large presence as a caring individual who nevertheless was capable of enforcing the rules. Conflicts were resolved quickly and in a manner perceived as fair by the students. The staff was given considerable authority to organize the program as they saw fit.

An attitude of relaxed playfulness was cultivated, since fear and pressure to succeed are motivating over the short term but toxic over the long term. The teachers began with the basics to make up for the deficits of past years. Good behavior and learning accomplishments were rewarded over the short term. Students competed for good behavior and learning goals in groups, not just as individuals. The curriculum included, but did not obsessively focus upon the state-mandated tests. In fact, at the students’ request, the teachers agreed to make a half-day available every week for activities that the students found most engaging. Research shows that even the most talented students don’t fulfill their gifts unless what they do is rewarding on a day-to-day basis, and this is even more true for at-risk students.

Some of these features might seem familiar, but we brought them together in an unusual package. A group that functions well is a bit like an organism with numerous organs. Remove any single organ, and the organism dies. The Regents Academy had all the necessary organs to function as an effective group—or least that was the plan.

We also employed the gold standard of assessment—the randomized control trial. This means that we identified twice as many students than the program could accommodate, randomly selected half to enter the program, and tracked the other half as they experienced the normal high school routine. If the Regents Academy succeeded—or failed—we would be able to prove it with a high degree of certainty.

What happened? By the first marking period their grades had popped up like a cork, compared to their comparison group. Like chameleons, it seemed that they were quickly adapting to the more prosocial environment provided by the school, even though the rest of their lives hadn’t changed. This result was encouraging but not definitive, because the curriculum and grading criteria of the Regents Academy were not exactly the same as for the regular high school. The definitive test would come at the end of the year when everyone would take the same state-mandated exams. We waited with mounting suspense and were astonished by the results. Not only did the Regents Academy students greatly outperform their comparison group, but they performed on a par with the average Binghamton high school student. Years of academic deficits had apparently been erased in a single year, at least as measured by the state-mandated exams. There was also no difference in the performance of blacks, Hispanics, whites, boys or girls. Everyone benefited to an equal degree.

Time will tell whether the Regents Academy continues to perform this well, along with our other social engineering efforts in the city of Binghamton. But who would have thought that Darwin’s theory of evolution, which so often is equated with selfishness, might actually provide the tools for making the world a more selfless place?