Biological Markets

Biological Market Theory (BMT) and Reciprocal Altruism (RA)

Biological Market theory (BMT) is not, as some seem to think, an 'extension' of Trivers' reciprocal altruism (RA) paradigm. However, RA and other ‘partner control models’ can describe cooperative relationships reasonably well when the costs of partner choice and partner switching are high. Two of the examples Trivers used in his 1971 classic on RA, coalition formation among male baboons and cleaner fish mutualisms, are explained more accurately by BMT. A review of the literature on food sharing in vampire bats, another example Trivers used, suggests that this might be better described as a market too (Carter & Wilkinson, 2013. Does food sharing in vampire bats demonstrate reciprocity? Communicative & Integrative Biology, 6, e25783) pdf. Gerald Carter advocates the use of the term 'reciprocity' in a broad sense, however, and sees partner control and partner choice models as part of a continuum (The reciprocity controversy). See also Gerald Carter's website & blog: http://socialbat.org/

Olof Leimar and John McNamara give their opinion on the significance of BMT for the theoretical analysis of the evolution of cooperation in their review paper Leimar, O. & McNamara, J. M. (2023). Game theory in biology: 50 years and onwards. Philosophical Transactions of the Royal Society B: Biological Sciences, 378(1876), 20210509

A common fallacy: reciprocity and time frames

In the primate literature ‘reciprocity’ is often confused with Trivers’ ‘reciprocal altruism’ (RA). Reciprocity, however, is an observed phenomenon: two animals exchanging altruistic acts do that more or less equally often. RA is a so-called ‘partner control’ mechanism that can result in reciprocity. Indeed, the lack of reciprocity disproves RA. However, the opposite is not true as reciprocity can result from other mechanisms, such as the ‘partner choice’ mechanisms advocated by biological market theory (BMT). Reciprocity is often found to be more pronounced when measured over longer time frames. This is not surprising; variable things tend to equal out as time passes. A recent paper by Campennì & Schino confirms this (Campennì, M., & Schino, G. (2014). Partner choice promotes cooperation: The two faces of testing with agent-based models. Journal of Theoretical Biology, 344, 49-55)

To prove RA, however, one should look at extremely short time frames, say one to three interactions in a row between the same individuals. Crucial is the contingency in the behaviour shown in those interactions. RA is usually modelled using repeated games such as the 2-player Iterated Prisoner’s Dilemma to make this point obvious. Partner choice can be based on interactions over much longer time frames. If ‘attitudinal partner choice’ (Fruteau et al 2009) plays the role I think it plays, then certain partners are preferred on the basis of their long term reliability or willingness to help. BMT can explain why such partnerships can be lop-sided over both shorter and longer time frames. It is a mistake to think that BMT predicts fast fluctuations only. It predicts adaptations to supply and demand, but it depends on the mechanism involved how fast these adaptations are. It also depends on the commodities involved whether partner values are likely to change over short or long periods. For example: a high-ranking monkey will remain a valuable partner when it comes to giving support in conflicts over long periods of time, but a monkey mother of a young infant may loose value quickly as a partner when new infants are born and her own infant grows older.

written by Ronald Noë  (last update  20 MAR 23)