The core phenomenon discussed in the target article is strong reciprocity: a predisposition to reward cooperators (altruistic rewarding) and punish norm violators (costly punishment), at a personal cost, even when the probability that this cost will be repaid is very low (Gintis Reference Gintis2000; Gintis et al. Reference Gintis, Bowles, Boyd and Fehr2003).
Guala calls our attention to the fact that the existence of strong reciprocity appears to be strongly supported by laboratory experiments; yet there is no evidence from outside the laboratory supporting the claim that strong reciprocity (especially costly punishment) sustains cooperation in real life.
We agree with Guala's point that evidence gathered from controlled laboratory experiments is not sufficient to reach the conclusion that strong reciprocity sustains cooperation in real life. However, we further argue that it is not only the absence of real-life data that is problematic for strong reciprocity theorists. There also exist laboratory experiments which present strong counter-evidence against the presence of strong reciprocity and the claims for its evolutionary origins.
The significance of observing a certain behavior pattern (strong reciprocity) in the laboratory for understanding a real-life phenomenon (cooperation) depends crucially on the extent to which the real-life phenomenon is appropriately captured in an experimental setup. Consider a real-life situation in which a group of individuals engages in a cooperative act, say, hunting a stag. In an experimental setting this corresponds to an economic game (e.g., Ultimatum Game, Public Goods game, or Trust game) played out by a group of experimental participants. In real life there needs to be something that motivates individuals to engage in this cooperative act. For the hunters, this could be the idea of satisfying their hunger. In the laboratory one can try to elicit this effect with the use of financial incentives – so far, so good.
However, in order to satisfy their hunger, the hunters have to make a plan, then run after the stag, then use their tools or weapons to kill the stag, then cut it into pieces and distribute them amongst the group. In short, the hunters have to put in some real effort to get this material benefit. This is where real life and standard laboratory experiments diverge: In the laboratory, experimental subjects are generally endowed with an amount of money by the experimenter. They sit on a chair, make a decision, push a button on the keyboard, and leave the experiment with some money that they “earned” without putting in any actual effort. In the hunting scenario this would be like stumbling upon a recently deceased stag and simply having to decide how to share the spoils.
Importantly for the arguments developed in the target article, experiments which incorporate this naturalistic property of expending real effort to earn wealth or status show behavior that clearly diverges from the predictions of strong reciprocity. For instance, in one-shot, anonymously played Dictator games, a huge proportion of the allocators give tiny amounts or even zero to the receivers when the allocators have earned the amount to be distributed (Cherry et al. Reference Cherry, Frykblom and Shrogren2002). However, when the receivers have to put in effort to earn their share, the allocators give more than half of the pie to the receivers (Oxoby & Spraggon Reference Oxoby and Spraggon2008). Similarly, in an Ultimatum Game, the proposers offer less to the responder when the proposers have to earn their status and wealth than when they are simply allocated to their roles and endowed with money by the experimenter (Hoffman et al. Reference Hoffman, McCabe, Shachat and Smith1994).
These examples constitute a problem for the claim that strong reciprocity is an important mechanism for sustaining cooperation outside the laboratory because windfall wealth is much rarer in real life than earned wealth. If one's theory only applies to the minority of real-life situations, then to conclude that costly punishment, a component of strong reciprocity, sustains cooperation in real life appears unwarranted. Put simply, including consideration of such “earned-wealth” experiments strengthens Guala's conclusions about the impact (or lack thereof) of altruistic rewarding and costly punishment on the emergence of cooperation in evolutionary history.
The core phenomenon discussed in the target article is strong reciprocity: a predisposition to reward cooperators (altruistic rewarding) and punish norm violators (costly punishment), at a personal cost, even when the probability that this cost will be repaid is very low (Gintis Reference Gintis2000; Gintis et al. Reference Gintis, Bowles, Boyd and Fehr2003).
Guala calls our attention to the fact that the existence of strong reciprocity appears to be strongly supported by laboratory experiments; yet there is no evidence from outside the laboratory supporting the claim that strong reciprocity (especially costly punishment) sustains cooperation in real life.
We agree with Guala's point that evidence gathered from controlled laboratory experiments is not sufficient to reach the conclusion that strong reciprocity sustains cooperation in real life. However, we further argue that it is not only the absence of real-life data that is problematic for strong reciprocity theorists. There also exist laboratory experiments which present strong counter-evidence against the presence of strong reciprocity and the claims for its evolutionary origins.
The significance of observing a certain behavior pattern (strong reciprocity) in the laboratory for understanding a real-life phenomenon (cooperation) depends crucially on the extent to which the real-life phenomenon is appropriately captured in an experimental setup. Consider a real-life situation in which a group of individuals engages in a cooperative act, say, hunting a stag. In an experimental setting this corresponds to an economic game (e.g., Ultimatum Game, Public Goods game, or Trust game) played out by a group of experimental participants. In real life there needs to be something that motivates individuals to engage in this cooperative act. For the hunters, this could be the idea of satisfying their hunger. In the laboratory one can try to elicit this effect with the use of financial incentives – so far, so good.
However, in order to satisfy their hunger, the hunters have to make a plan, then run after the stag, then use their tools or weapons to kill the stag, then cut it into pieces and distribute them amongst the group. In short, the hunters have to put in some real effort to get this material benefit. This is where real life and standard laboratory experiments diverge: In the laboratory, experimental subjects are generally endowed with an amount of money by the experimenter. They sit on a chair, make a decision, push a button on the keyboard, and leave the experiment with some money that they “earned” without putting in any actual effort. In the hunting scenario this would be like stumbling upon a recently deceased stag and simply having to decide how to share the spoils.
Importantly for the arguments developed in the target article, experiments which incorporate this naturalistic property of expending real effort to earn wealth or status show behavior that clearly diverges from the predictions of strong reciprocity. For instance, in one-shot, anonymously played Dictator games, a huge proportion of the allocators give tiny amounts or even zero to the receivers when the allocators have earned the amount to be distributed (Cherry et al. Reference Cherry, Frykblom and Shrogren2002). However, when the receivers have to put in effort to earn their share, the allocators give more than half of the pie to the receivers (Oxoby & Spraggon Reference Oxoby and Spraggon2008). Similarly, in an Ultimatum Game, the proposers offer less to the responder when the proposers have to earn their status and wealth than when they are simply allocated to their roles and endowed with money by the experimenter (Hoffman et al. Reference Hoffman, McCabe, Shachat and Smith1994).
These examples constitute a problem for the claim that strong reciprocity is an important mechanism for sustaining cooperation outside the laboratory because windfall wealth is much rarer in real life than earned wealth. If one's theory only applies to the minority of real-life situations, then to conclude that costly punishment, a component of strong reciprocity, sustains cooperation in real life appears unwarranted. Put simply, including consideration of such “earned-wealth” experiments strengthens Guala's conclusions about the impact (or lack thereof) of altruistic rewarding and costly punishment on the emergence of cooperation in evolutionary history.