What is reciprocity, anyway?

– by Diego Guevara Beltrán, Jessica Ayers, Lee Cronk, & Athena Aktipis

Thirty-four experts on cooperation walk into a bar and begin to discuss the ‘true’ meaning of the term ‘reciprocity.’ A biologist offers one definition. She says, “we observe reciprocity when the gift of some food, or another favor, increases the probability for the donor to receive the same thing, or something else, from the receiver”.

Perhaps because that approximates one of Robert Trivers’ definitions of reciprocal altruism, all experts nod in agreement with the biologist. However, not being quite content with this definition, an economist retorts “[that the] formulation is rather sloppy.” His colleague adds that “reciprocity should be conceptualized as reciprocal strategies in repeated games, normative obligations, and individual preferences to respond to kind acts with kind acts and to unkind acts with unkind acts.”

“Yes, yes, but how do we operationalize reciprocity?” asks one psychologist. He says, “The best statistical definition is the contemporaneous correlation of two individuals’ responses within time-series regression analyses and as partner effects among these variables.” Intrigued but bewildered, another psychologist adds “reciprocity is reciprocity, that’s what I was taught! Because he gave to him and he gave to me, then I gave to her, albeit indirectly. But I couldn’t keep track of who gave what to whom, what I owed to him, or who I should groom. Maybe we should just avoid this definitional animosity and just call all transfers something-something reciprocity!”

A review of the literature yields a proliferation of reciprocity-related terms – thirty-four at last count. That might lead you to believe that experts indeed agreed with the bewildered psychologist. Perhaps experts want to avoid definitional animosity and decided to call all transfers “something-something” reciprocity. With several conflicting, overlapping, and even a few redundant terms, it was clear there is no consensus among scholars regarding what truly constitutes reciprocity. In fact, these terms continue to proliferate, as evidenced by the preceding paper in the same issue of Evolution and Human Behavior (i.e., “dynamic indirect reciprocity”). This lack of consensus and the continued proliferation of terms motivated us to review the term reciprocity and survey cooperation experts about it.

In our study, we asked 85 cooperation experts to rate the extent to which they believe thirty scholarly definitions of the term reciprocity were truly reciprocity. Some of these definitions (e.g., indirect reciprocity: return is expected from someone other than recipient of benefit) are common in the literature, while others (e.g., homeomorphic reciprocity: exchange of things that are the same) not so much.

We first wanted to know whether experts’ responses would help us identify underlying dimensions of reciprocity (i.e., characteristics that cluster some, but not other, terms together). Applying factor analyses, we find that the scholarly definitions of the term reciprocity can be represented by four broad dimensions: (1) Balanced Transfers (i.e., transfers that are of equal or equivalent value), (2) Reputation-based Transfers (i.e., transfers where individuals give to others who have given in the past and receive from others if they have given in the past, (3) Debt-based Transfers (i.e., transfers where individuals keep track of, and expect repayment for, what they give to others), and (4) Unconditional Transfers (i.e., transfers that do not revolve around concepts of debt or account keeping). This four-factor framework captures greater variation in types, or dimensions, of transfers than previous frameworks (e.g., direct, indirect, and generalized reciprocity).

We encourage experts to consider adopting these terms. In doing so, we can provide specific qualifiers to a transfer while at the same time reducing the possible number of dimensions that describe a transfer based on whether, or the extent to which, they are balanced, debt-based, reputation-based, or unconditional. For example, we saw that several experts used the same term (e.g., indirect reciprocity) to describe multiple different reciprocity terms. Hence, our framework would allow experts to avoid using overlapping terms to describe different dimensions of transfers, ultimately improving communication within and between disciplines.

We then assessed the level of consensus among experts regarding what truly constitutes reciprocity. There was no consensus regarding what should be considered reciprocity. However, over 90% of experts agreed that unconditional transfers (e.g., generalized reciprocity I: non-conditional sharing and giving of assistance) should not be considered reciprocity. So, although experts do not yet agree on what truly constitutes reciprocity, these findings suggest that the scholarly community can move beyond discussions of whether unconditional transfers should or should not be considered reciprocity. They should not be.

Instead, the scholarly community might consider focusing on outstanding disagreements, such as those identified in our study. For example, some experts believe that transfers are reciprocity only if agents have the intention to reward a giver or incentivize a receiver to give back, while others do not think intentionality is a defining feature of reciprocity. Similarly, some experts believe that reciprocity must involve a cost, while others believe that reciprocity does not involve cost or transfers with negative utility.

Focusing on these issues might help experts reach a consensus regarding what truly constitutes reciprocity. However, we might also supersede these disagreements by adopting the language of Balanced, Debt-based, Reputation-based, and Unconditional transfers. Does it matter for the study of evolution and human behavior whether unconditional giving is or is not reciprocity? We think not. What matters is that we reach a deeper understanding of the ultimate causes and proximate mechanisms underlying the diversity of transfers that have shaped the evolution of human behavior. While our framework does not include information about the ultimate causes of behavior, it provides information about the proximate mechanisms underlying an individual’s decision to engage in a transfer. Overall, we hope our study allows the scientific community to gauge the current level of consensus, or lack thereof, regarding the term reciprocity, and ease communication across academic fields.

Read the original article: Guevara Beltrán, D., Ayers, J.D., Munoz, A., Cronk, L., & Aktipis, A. (2023). What is reciprocity? A review and expert-based classification of cooperative transfers. Evolution & Human Behavior, 44(4), 384-393.