I argue that the authors accept too casually the neo-classical framework of thought that was incapable of predicting choices in 2-person and other experimental games in the 1980s and 1990ss. The ex post hoc hypothesis that social preference can describe homo socialis reduces inevitably to a rescue of neo-classical economics in which Max-U (own payoff, other payoff) substitutes mechanically for Max-U (own payoff) in our personal groupings. This static procedure unnecessarily and inappropriately robs human conduct of its sociality as a process relationship. The model I articulate was masterfully developed by Adam Smith, which back-predicts the results of these earlier small group experiments, and argues the central importance of context—a finding of experimentalists in their attempt to come to terms with the predictive failures of Max-U (own payoff).
Vernon L. Smith (2015), "Adam Smith: Homo Socialis, Yes; Social Preferences, No; Reciprocity Was to Be Explained", Review of Behavioral Economics: Vol. 2: No. 1-2, pp 183-193. http://dx.doi.org/10.1561/105.00000028