Whom Do You Trust?

To teach police cadets how to search a dark building without getting shot, the Arizona Law Enforcement Academy puts people armed with paintball guns inside a building. They hide behind furniture or in closets, waiting for the trainees to flush them out. The instructors then send in pairs of recruits suited up in pads that protect them from direct hits.

The goal is to find and disarm each suspect before he can "kill" or "wound" an officer. This doesn't always work out as planned.

"During one of the drills, I was on one knee peeking into a dark room while my partner pointed a flashlight into the room from over my head," one cadet reported on his blog. "He then pointed the light directly down on my head. He thought he turned it off first. The suspect in the room immediately saw me and shot me in the hand. Even though we were in full pads, my fingers were unprotected. It is a painful place to get hit. I'm now sporting a colorful welt and blister on my index finger as a reminder of my partner's great tactics."

Five months later, the same cadet, now a Phoenix police officer, stopped a car with an invalid license plate. Two men were inside the vehicle. "One of my squad mates saw me pull over the car and pulled in behind me to provide backup," he wrote. "I approached the driver's window and asked him for his license. He said he had no identification. The passenger was hiding one of his hands under his leg. I told him to let me see his hands. My backup officer walked up to the passenger side, and we both knew there was something wrong."

The initial search turned up a gallon-sized freezer bag of marijuana in the back seat and a bag of cocaine in the passenger's pocket. With both suspects in handcuffs, the officer returned to the car to search it further. "I walked back to the passenger side of the car and felt my heart start racing when I saw a 9mm Ruger handgun sitting between the front seats partially under a piece of paper," he wrote.

"The hammer of the gun was cocked back, and it was loaded. It would have taken two seconds for the passenger to pick it up if he had wanted to." A second loaded gun was hidden between the driver's seat and the center console. "I had two guys committing serious felonies with loaded guns within reach," wrote the rookie. "If my backup partner hadn't arrived a few seconds after I did, I wonder if the outcome may have been different. Two [of them] against one [officer] may have encouraged them to take a chance with running or even shooting."

Every partner needs to be able to depend on his counterpart. Every partner takes a risk that the other person might fail, intentionally or innocently, leaving him with the light shining on him in the middle of a dark room. Every partner needs the dependable backup that is the difference between success and failure.

Statements about trust

This came through forcefully in Gallup's analysis of people's responses about their best and worst partnerships. Three statements about trust form the heart of the working relationship:

• We trust each other.
• We can count on each other to do what the other says he or she will do.
• He or she tells others how good I am, and I tell others how good he or she is.

In a good collaboration, 58% of partners strongly agree that they trust each other, and another 29% score the statement a 4 on a 1-to-5 scale. In a poor partnership, less than 3% strongly agree they trust each other, while 50% strongly disagree.

Trust is the linchpin of a partnership. With trust, both people can concentrate on their separate responsibilities, confident the other person will come through. One brain-imaging study discovered that once trust is created, a person's brain will process his counterpart's cooperative move before it even happens.

Without trust, it's better to work alone. Both people doubt whether the other will fulfill his end of the bargain. Both must verify the other's actions. Both must make contingency plans in case their counterparts fail. The frustration and inefficiency of not being able to count on someone is more hassle than the burden of handling the full load alone. No trust, no partnership.

You face a dilemma every time you interact with someone new. If he can be trusted, and if he learns to trust you, the two of you can be more successful working together than going it alone. But if one of you is not trustworthy, it's better that you never even try to collaborate.

This psychological fork in the road is one of the most important decisions a person makes. It's one of the crucial reasons humans can distinguish so many faces and keep track of so many reputations. For our ancestors living under more dangerous conditions, making the wrong choice could be fatal. For us, success on a project, reputations, promotions, future opportunities, happiness, and careers all hang in the balance.

Collaboration occurs only when both people trust each other and prove themselves trustworthy. This is a risky undertaking. Trust involves exposing yourself to the chance that the other person will fail to keep her end of the bargain.

The American TV game show Friend or Foe? created this dilemma by squaring off two contestants whose trustworthiness determined how they would divide money they won together earlier in the show. Each had two buttons the other person could not see -- one labeled "friend," the other "foe." If both players pushed "friend," they split their winnings evenly. If both pressed "foe," they both went home empty-handed. But if one pressed "friend" and one pressed "foe," the defector got all the money, and his partner received nothing.

One episode brought back two contestants who were betrayed in earlier shows, but who had accumulated $7,500 working together. "Mike, playing the game with you has been fun, and I think you are a trustworthy person," said fellow player Rob, putting his hand over his heart. "I'm giving you my word of honor that I will press 'friend.'"

"Rob, we've both been here before, and it would be a shame to go away with nothing," Mike reassured his counterpart.

They reached down and made their choices. Rob voted "friend." Despite his reassurances, Mike chose "foe," taking all the money for himself.

"I'm sorry," he said as the show's host berated him. "I couldn't be sure. I couldn't be sure."

"Things happen," said a clearly angry Rob. "Maybe you'll have a conscience sometime in the future. . . . All the money in the world can't buy you a clean conscience."

Betrayal -- often mutual betrayal -- was quite common on Friend or Foe? "Thousands of dollars were left on the table," wrote University of Chicago researcher John A. List. "In nearly 25% of the 117 games, both players chose not to cooperate, resulting in a net loss of nearly $100,000."

Half the time, one of the players burned a cooperative partner, despite friendly overtures. Occasionally, however, mutual trust made both people richer. "I think we worked really well together. I have a really good feeling about you," a contestant named Stacia told her counterpart, Jennifer. "We are both good people; I can just tell. It's better for both of us to go home with half the money than one or the other to be broke."

"Stacia, the other two teams voted 'foe' and walked away with nothing," replied Jennifer. "Let's go home feeling good about ourselves and with some money."

They both voted "friend" and split the money. "See," said the show's host, "nice people play Friend or Foe? too. They're not all evil little back-stabbers." The same lesson is true in real life.

Doing the right things

In a working relationship, being trustworthy is not a matter of hitting the right button. It is showing up for an important meeting on time, doing more than your share of the work, quickly returning e-mails and phone calls, giving all your creativity to a project, jumping in rather than having to be asked, not being a burden to your partner, fighting for the success of the project, working hard on physical jobs and smart on mental ones, and hundreds of other acts large and small.

In the early stages of a collaboration, both partners take a wait-and-see attitude, still in the clutches of the dilemma, not sure the other person is going to do his part. But if one cooperative move is matched by another, solid reputations form. The fear of being taken advantage of fades. Trust removes doubt. Trust eliminates the dilemma.

More than 350 years ago, the philosopher Thomas Hobbes concluded that the incentives to be selfish when dealing with someone else are too strong to be resisted. He would have told you to hit the "foe" button because the other person was almost certainly going to do the same.

The incentives in life naturally pull people toward "war of every man against every man," he wrote. As each person pursues his or her selfish interests, human existence is doomed to be "solitary, poor, nasty, brutish, and short."

Nobel Prize winner John Nash, portrayed in the movie A Beautiful Mind, looked for the patterns of behavior into which games of trust would settle if each player did what was best for himself. The "Nash Equilibrium," as it's called, of the TV game show is for both people to choose "foe" every time.

John von Neumann, one of the pioneers of this type of strategic thinking, even recommended the United States launch a nuclear strike against the Soviet Union before the Soviets struck first. Agreeing with von Neumann, his contemporary Bertrand Russell argued a first strike was "as simple and as unescapable as a mathematical demonstration."

For many decades, educated theorists dismissed people who are trusting as just too stupid to really understand the game. "Evidently," concluded one book on the subject, "the run-of-the-mill players are not strategically sophisticated enough to have figured out that strategy DD [mutual defection] is the only rationally defensible strategy."

And, of course, every experienced collaborator has several stories of serious breaches, when trust was repaid with treachery, when perhaps Hobbes and Nash and von Neumann were right. At one retail firm, the highly effective partnership between two regional vice presidents, one responsible for the west half of the United States and the other responsible for the east, ended when one of them resigned and joined a competitor. "I invested a lot in him, and this is how he repays me?" said the one left behind. "To hell with him."

Tit for Tat

In the midst of this cynicism, political scientist and professor Robert Axelrod organized a curious competition. In 1979, Axelrod decided to use computers to seek the best strategy for when to extend and when to withhold one's trust. He invited experts to submit programs that were, in essence, a set of rules stating when a player would cooperate and when it would defect in a series of interactions similar to the culminating event in the TV game show.

Fourteen people agreed to participate in the tournament. They came from the disciplines of psychology, economics, political science, mathematics, and sociology. "Most of the entrants were recruited from those who had published articles on game theory," wrote the professor. No one could claim they were "not strategically sophisticated enough" to understand the power of being distrustful. They submitted a diverse group of strategies. Just like the variety of potential partners you encounter in any setting, the programs included jerks, saints, and many permutations in between.

The winner of the tournament turned out to be the simplest of all the programs. It was submitted by Russian-born psychologist Anatol Rapoport. Called "Tit for Tat," it began by cooperating on the first move, and then it just mimicked what its counterpart did on the previous move. In its simplicity, Tit for Tat was an elegant solution to the trust problem. It had several features of a good human partner that made it most successful in the tournament.

It got things off on the right foot by displaying trust on the beginning move, and unless it was betrayed, it never proved untrustworthy. In these ways, Tit for Tat and the others that performed best in the contest were not what Hobbes or Nash or von Neumann would have predicted.

"Surprisingly, there is a single property which distinguishes the relatively high-scoring entries from the relatively low-scoring entries. This is the property of being nice, which is to say never being the first to defect," wrote Axelrod. When two trusting strategies met each other, they formed an elementary partnership, cooperating through almost the entire game, raising each other's scores along the way.

While it was friendly, Tit for Tat was no fool. As soon as it was betrayed, it retaliated on the next move and would continue refusing to cooperate until the other player ceased the hostilities. Because of this reflex, the strategy was "not very exploitable," wrote Axelrod. Such a fallback is crucial to the survival of otherwise obliging individuals. Without retaliation to keep them in check, just a few egotists or attackers can quickly overrun a benevolent population. Appealing to a sense of fair play works with most people, but a pernicious minority will exploit their colleagues' trust if nothing stands in their way.

Another winning feature of Tit for Tat was that as fast as it went to battle stations, it just as quickly returned to trusting when its counterpart did so. It was forgiving. "Of all the nice rules, the one that scored lowest was also the one that was least forgiving," Axelrod found.

The professor decided to try the experiment again, distributing the results of the first tournament to drum up additional interest. He wanted to know if more players and a wider variety of strategies could improve the results. Axelrod invited the initial set of players to try again. He also took out ads in computer journals. The response was more than he anticipated. Sixty-two entries came in from the United States, Canada, Great Britain, Norway, Switzerland, and New Zealand. They came from professors of computer science, physics, economics, psychology, mathematics, sociology, political science, and evolutionary biology. One came from a 10-year-old kid.

There were more than a million moves in the second tournament. All those who entered a strategy knew Tit for Tat won the first tournament, and why. Yet when all the strategies had a chance to interact with each other, the winner was once again Rapoport's simple rule, which he resubmitted unaltered from the first tournament. "So Tit for Tat, which got along with almost everyone, won the second round of the tournament just as it had won the first round," wrote Axelrod. It is "clearly a very successful strategy," he concluded.

The lesson in Axelrod's tournament lies in how forcefully it contradicts the supposedly savvy strategy of being selfish. Hobbes was wrong. Nash was wrong. Von Neumann, thankfully, was wrong. People are not purely selfish and calculating; they are reciprocating, both positively and negatively. They reflect what they receive. Reciprocity is one of the most powerful forces in human nature.

In many ways, Axelrod's tournament confirmed the wisdom of a passage in the Edda, a 13^th-century collection of Norse epic poems. "A man ought to be a friend to his friend and repay gift with gift," it states. "People should meet smiles with smiles and lies with treachery." The positive side of these deep-rooted emotions is the glue that holds together a partnership.

The most important element in forming and maintaining a variety of strong partnerships is not your craftiness, but your willingness to take the risk of trusting numerous potential partners and your diligence in repaying the trust they place in you. If you're not careful, you could be so "strategically sophisticated" that no one wants to work with you or that you fail to recognize or reciprocate collaborative overtures from people all around you. Just like Tit for Tat, you need to be eager to cooperate; to make early, friendly overtures to your partner; to stubbornly refuse to make the first hostile or neglectful move; and to be quite willing to forgive.

The ultimate twist to the research on trust is what it reveals about your collaborative environment. If you are like most people, you assume you are simply making reasonable reactions to the people with whom you interact. You believe you are working with the hand you were dealt. To the degree you compete, you probably feel as though you are just reacting to the competition around you. You probably attribute your lack of partnerships to a lack of good partners.

But in a Tit for Tat world, where most people return good for good and bad for bad, the world you inhabit is the world you make. Your reputation precedes you, biasing the way new colleagues deal with you. Your first moves, friendly or hostile, tip the balance for future interactions. When you exhibit trust, you will most often find trustworthiness. When you are selfish, you will most often find selfishness. When you compete, others must resort to competition. If you choose to play the game strictly for your own advantage, your attempts at collaboration will indeed be "solitary, poor, nasty, brutish, and short."

In the end, the degree to which you succeed in forming trusting partnerships is less a reflection of how much people trust you than how much you trust them -- less a reflection of their trustworthiness than of your own.