How Can Small Groups Put a Stop to Bad Behavior?
A small organization — an office, academic department, church or team — wants everyone to comply with its ideas of good behavior. But how does it do this without creating feuds and spending resources on fines and monitoring? Punishment by peers is known to make things worse, so what is left?
We propose and test a device we call the Hired Gun. Someone is appointed, elected or hired to carry out a simple rule: Punish the biggest offender, but just to the point that he or she would rather have been the second biggest offender. This amount could be very small — a slap on the wrist — but the incentive creates a race to second place. That is, everyone wants to be the second biggest cheater to avoid the punishment. The only way everyone finishes second is if everyone stays in the starting blocks, if everyone behaves well. We show this simple rule results in easy monitoring and small and rare fines, and quickly produces good behavior.
Which Incentives Produce Good Behavior?
We know professors like teaching but prefer to do research. How does a department make sure effort is put into classes? Some neighbors want all the yards on the street to be clean, and others are lax. How can neighbors enforce civility while remaining civil? These represent classic problems in social science: When people face choices that benefit themselves at a cost to others, how do we structure incentives so people do the right thing?
Some have suggested that monitoring and punishing by peers can be enough: Good teachers can snub bad teachers, clean neighbors can scold messy neighbors. And if that doesn’t work, they can resort to other forms of vigilantism. Economists have examined punishment by peers in laboratory studies and have found that peer punishment increases good behavior but often makes the group as a whole worse off. Vigilante justice is just too costly: People punish too much. Others have noted that, in reality, punishment by peers is fairly rare.
Instead, small groups appoint or hire someone with authority to enforce the social order. Just as in the Wild West when vigilantes were replaced with hired guns, schools have department heads who review teaching evaluations, and apartment buildings have superintendents to track complaints. These delegated authorities have the responsibility to discipline those who fall short of good behavior. But this leaves a critical question unanswered: If we delegate authority to hired guns, how should they govern? Monitoring everyone and punishing all cheaters is difficult and costly. We need an enforcement method that is simple and inexpensive and has small penalties and works.
The Race to Second Place
Imagine a one-minute race where the person who runs the second farthest gets $100. How would this race end? Poetically, it ends at the beginning. Everyone will be standing at the starting line, but nobody will cross. Why? Crossing the line starts you in first place, and no one else will ever want to pass you.
More realistically, think of driving on the highway. You want to go fast but also want avoid a ticket. If there is at least one car going faster than you, the highway patrol will ticket the faster car. You’re safe if you are the second fastest car. If no one wants to get a speeding ticket, then everyone will try to be the second biggest speeder. Everyone slows down until they’re all going the speed limit — they’re all behind the line. Similarly, when a police car is on the road, typically everyone falls in behind it. To enforce the speed limit, the officer needs only fine the first driver who passes the police car.
We use this intuition to design our enforcement device.
The Hired Gun Versus Peer Punishment
We investigate whether a delegated authority — the Hired Gun — who punishes only the largest deviators can effectively enforce socially desirable outcomes. We also compare the effectiveness of this Hired Gun to vigilante or peer punishment. Can delegated authority replace costly and inefficient peer punishment and get superior social outcomes?
We represent a social dilemma with a public-goods game. A person is assigned to a group of four people and given five tokens to divide between a socially beneficial public good (like teaching well, or being a good neighbor) and a private good benefiting only the individual (like being a lazy teacher or messy neighbor). Each token spent on the public good pays a return of $2 to all group members. This means the group as a whole earns $8 (four people times $2 each). In contrast, each token spent in the private good pays a return of $3 to only the person who made the choice; no other group member benefits.
A purely selfish person will choose to spend all five tokens on the private good and none in the public good. If all the group members do that, then each person will earn $15 (five tokens times $3 each). However, a group taking the socially optimal action would instead spend nothing in the private good and instead spend all five tokens on the public good for a payoff of $40 per group member (five tokens from each of the four group members — 20 tokens total — times $2 each).
We add the ability to delegate punishing authority using the Hired Gun. The Hired Gun identifies the group member who gave the lowest number of tokens to the public good (like the fastest car on the highway). The Hired Gun fines this player just enough so that he or she would rather have been the second most selfish person: This player earns what the second most selfish person earns minus one token.
In our first study, people repeated the public goods game 10 times and then repeated it 10 more times with a Hired Gun. To prevent bias based on the reputation of the players involved, players decided how to allocate their tokens on a computer in anonymous, random groups that changed each round. As shown in Figure 1, players spent an average of only 1.3 tokens on the public good when there was no Hired Gun, but they raised their contributions to 4.5 tokens when there was a Hired Gun. This is a 68 percent increase in average earnings, as shown in Figure 2.
Clearly the Hired Gun had the desired effect of starting a race for second place. To understand if this was a superior mechanism to peer punishment, we ran a second study. Here we allowed players to pay a price to fine one another after they played the public goods game. This type of vigilante justice is akin to a tenant in an apartment complex penalizing a messy neighbor by complaining to the person directly.
After 10 periods of only peer punishment, some groups were given the chance to pay a fee to get a Hired Gun. The players paid for the Hired Gun more than 70 percent of the time and, as can be seen in Figures 3 and 4, it was well worth the cost.
It is clear from Figure 3 that the option to have a Hired Gun causes people to give more to the public good. Figure 4 illustrates the huge increase in money earned.
When the Hired Gun option is available, people stop using peer punishment. Just as when a tenant directs complaints to the superintendent instead of complaining directly to a messy neighbor, when Hired Guns are available, people stop using vigilantism to enforce good behavior.
Our research shows that when things aren’t working, people are clever enough to invent methods, like a Hired Gun, that help them achieve better outcomes. Direct punishment by peers is one option, but it only makes things worse. Instead, delegating punishment that creates a race for second place is far more effective and, moreover, pushes out vigilantism.
So, let the races begin!