In their thorough and insightful article, Maestripieri et al. summarise evidence comparing the dominant economic, social, and evolutionary explanations for the social and employment biases favouring attractive individuals. They justifiably conclude that these biases are better explained by an evolutionary theory (relating to access to high-quality mating partners) than they are by theories put forward by economists and social psychologists.

The authors' argument implicitly invokes Tinbergen's (Reference Tinbergen1963) four levels of explanation (“Tinbergen's four questions”). Tinbergen argued that complete accounts of behaviour comprise four levels of explanation: the (1) causal mechanism and (2) lifetime development (ontogeny) of the behaviour (both proximate explanations), and the (3) adaptive function and (4) phylogenetic origin of the behaviour (both ultimate level explanations). Explicitly applying a Tinbergian perspective to the authors' arguments reveals that the authors' evolutionary theory is the preferred option of those theories considered because it is the only one providing an ultimate, in this case, functional, explanation. The authors' evolutionary theory both considered the adaptive function and made predictions about the causal mechanisms of the behaviour. The other theories are strictly proximate explanations, describing only the causal mechanism of the behaviour. This is why Maestripieri et al. describe the social and economic theories as descriptive – proximate theories frequently are, as they describe how behaviours develop and manifest in an immediate sense. But when seeking to understand why behavioural mechanisms develop and manifest the way they do, only an ultimate-level explanation will do.

The authors' theory focuses on an ultimate (functional) explanation for favouring attractive individuals. The authors have, however, overlooked the potential utility of considering phylogenetic and comparative evidence – the other half of the ultimate-level explanation. Although non-human species do not interview applicants for jobs, choice of social partners for cooperative enterprises is an area where the authors' “mating opportunity” theory could be tested against comparative evidence. One starting point may be the acceptance or rejection of new individuals into groups in social species. For example, female chimpanzees disperse into new groups at times of high reproductive value (during oestrus and at late adolescence); they risk attack from resident females during the migration process; and resident males will defend immigrant females who are in oestrus (possibly as a way to elicit mating) (Hemelrijk et al. Reference Hemelrijk, van Laere and van Hooff1992; but see Hemelrijk et al. Reference Hemelrijk, Meier and Martin1999), but attack immigrant females who are not in oestrus (Nishida Reference Nishida, Heltne and Marquardt1989). Such comparisons may reveal important similarities and differences between humans and chimpanzees in prosocial treatment of high mate-quality individuals, thus providing clues to both the evolutionary precursors of the human attractiveness bias and the more recent selection pressures that may have shaped it.

A formal application of Tinbergen's framework also reveals that some of the evidence presented as support for the authors' evolutionary theory is not necessarily relevant to it. For example, the evidence reviewed of brain areas activated by attractive opposite-sex faces is no more consistent with an evolutionary explanation than it is with any of the other explanations (even if the proponents of other theories are less likely to look for such evidence). It is important to understand which brain areas are involved in perceiving facial attractiveness, and informative to know that attractive opposite-sex faces (for heterosexual observers) activate neural reward circuitry. However, all of this is evidence only of the proximate, causal mechanisms involved in making attractiveness judgements, and all of the other theories reviewed by the authors are proximate, causal theories, which could as easily incorporate this evidence as could the evolutionary theory for which they argue.

We agree with the authors that evolutionary explanations are crucial for any comprehensive explanation of the attractiveness bias. The evidence that mating motivations play an important role in these biases is strong and well articulated by the authors. Some of the evidence put forward, however, is actually difficult to reconcile with mating motivations being the sole ultimate explanation for prosocial attractiveness biases. For example, the mating motivations theory is not obviously consistent with biases favouring attractive children and same-sex individuals (because they are not potential mates). Such biases suggest that attractive individuals might also be favoured for nonmating functions, perhaps because facial attractiveness serves as a reliable cue to a range of desirable traits, and forming coalitions with, or doing favours for, such individuals confers other kinds of advantages. The authors argue against some of these possibilities, but there is a positive correlation between intelligence and attractiveness (Kanazawa Reference Kanazawa2011), for example, and if attractiveness is a cue to health (see Stephen & Tan Reference Stephen, Tan, Sheppard and Haque2015 for a review) and developmental stability (Perrett et al. Reference Perrett, Burt, Penton-Voak, Lee, Rowland and Edwards1999), as the mating motivation theory suggests, then it is likely to also correlate with other traits that are useful in social partners. If modern hiring decisions had analogues in the social dynamics of pre-industrial or pre-agricultural human societies, then favouring attractive individuals in these situations might have been advantageous. One way of testing this possibility would be to examine how widespread preferences for attractive individuals are in modern human groups across a broad spectrum of cultures and levels of industrialisation.

West-Eberhard (Reference West-Eberhard2014) provides a cogent summary of the complexity of behaviours expected to appear under social selection pressures (where social selection encompasses sexual selection, but includes inter-individual competition over any kind of resource, not just potential mates). In this vein, it would be worth examining whether there are systematic patterns, beyond the opposite-sex biases towards (and, in some cases, same-sex biases against) attractive individuals. Perhaps the effects are stronger in jobs requiring extensive teamwork, or for positions where the target individual's competency is especially important, or even for positions where the target individual's attractiveness may benefit the employer directly through the attractiveness bias the target will elicit in others (for example, it may be beneficial to hire attractive salespeople, mating motivations of the hiring team aside).