Strong Reciprocity, Human Cooperation and the Enforcement of Social Norms

In Fehr, E., Fischbacher, U., Gächter, S., 2002a. Strong reciprocity, human cooperation and the enforcement of social norms, the author challenge the self-interest assumption, and indicates that people are more willing to voluntary cooperate, with the punishment to non-cooperators. This behavior is called “strong reciprocity”.

According the Ginits (2002), a person is a strong reciprocator if she is willing to sacrifice resources to be kind to those who are being king(strong positive reciprocity). On the other hand, negative reciprocity is when she is willing to sacrifice resources to punish those who are being unkind.

The author introduces the enforcement of ‘Nonbinding’ Agreements, and a real life example would be the exchange between a taxi driver and his passenger in a big city, this is a sequentially played one-shot PD because the probability of repeated interactions is extremely low, and the taxi driver has to choose whether to cooperate or not. The author mentions some well-controlled laboratory experiments by Fehr. In these experiments, subjects could earn real money according to their decisions. Exchange partners were located in different rooms. In the experiment, a subject in the role of an employer can make a job offer (or contract offer for one unit of a good) to the group of subjects in the role of worker (or seller). Each worker can potentially accept the offer. The number of workers is more than the number of employers. A job offer consists of a binding wage offer (or price offer) w, and a nonbinding ‘ desired effort level’ (or ‘desired quality level’)

Once the worker accepts an offer, he or she has to determine the actual effort level. In the experiment the choice of an effort level is represented by the choice of a number. The higher the chosen number the higher the effort the higher the monetary effort costs borne by the worker. The desired and the actually effort level have to be in the interval from 0.1-1 and the wage offer has to be in the interval. The higher the effect level is the larger is the material payoff for the employer but the higher are also cost of providing e for the worker. (shown in the paper). The payoff for the employer is 100e-w, and the payoff for the workers is w-cost of e.

The author notices that a rational and selfish employer (who believes that there are only selfish workers) will offer a wage above 1, the reason is that the employer knows that the workers will incur no effort costs and will accept a wage offer at 1.The observation is that the vast majority of contact offers imply a much larger profit than 1 for worker.( on average 35) Based on the figure 1, subjects have a propensity for strong positive reciprocity, and the strong positive reciprocity is not diminished if the monetary stake size is rather high.

The result above confirms that strong reciprocity substantially contributes to the enforcement of cooperative agreement in bilateral sequential exchanges, it provides the incentives for the potential cheaters to behave cooperatively or limit at least their degree of non-cooperation. Then the author gives two further foundings: even purely selfish employers have an incentive to make a cooperative first move, such as a generous job offer, if they except sufficiently many workers to behave in a strongly reciprocal manner. Similarly, even purely selfish workers have an incentive to provide a high effort in case of a reward/punishment opportunity if they expect sufficiently many employers to be strong reciprocators. To be honest, this experiment wasn’t explained well, the author did not give us the information on the background of the participants and the number of participants and how many rounds are there in this experiment.

However, when I look at the original experiment (Fehr, Ernst, Georg Kirchsteiger, and Arno Riedl. "Does fairness prevent market clearing? An experimental investigation.", there are two stage sof the experiment, and the  original purpose was meant for studying the fairness on market prices. Still there is not enough or clear information about the background of the participants,( but the communication of contract was throughout phone.) However, I know that there are four supervisors engaged. The aim of this experiment was “to detect whether agents exhibit intrinsic fairness preferences, and whether they base their actions on the others act fairly.”  What surprises me most is that when people focus on their actions instead of focusing on other’s act, it makes the outcomes better.

...