Evolutionary ethics, contractualism and fairness

 

 

What is the connection between contractualism and fairness (distributive justice)?  When you and I agree to collaborate, we agree that the subsequent regulation of our collaborative behaviour – by and on behalf of “us” – is legitimate.  So, if you protest to me that I have not fulfilled my role properly, I accept it as legitimate because it comes from the “us” that I willingly agreed to join.  

The contract or agreement we make is separately enforced, and our ideal collaborative behaviour is obliged, by threats of damage to our reputations, if we are not ideal collaborative partners.  

Being an ideal collaborative partner includes being fair when it comes to sharing the rewards of our collaboration.  Fair distribution will usually be explicitly negotiated before collaboration starts, as part of an explicit agreement.  The proposal is, you and I will only do that which we already endorse, including the ideals of fairness.  Hence, you or I will never agree to something that is too unfair, according to the ideals of fairness.  Contracts confer legitimacy on the regulation of our collaborative behaviour, but not necessarily on the content of the behaviour itself.   

The agreement to collaborate shapes and constrains the collaboration, and legitimises the intrapersonal, interpersonal, and collective regulation of our collaborative behaviour.  

 

As a family of moral theories, contractualism holds that morality is primarily about acting according to what would be agreed by rational agents.  Common to the main contractualist approaches ... is the idea that acting morally is about acting in mutually advantageous ways ..., and that the morality of an action depends on whether relevantly affected parties could reasonably agree to it, or not reasonably reject it ... .

Arthur Le Pargneux, Nick Chater, and Hossam Zeitoun – “Contractualist tendencies and reasoning in moral judgment and decision making” (2024)

 

Rational agents

There are at least two kinds of rationality – instrumental rationality and cooperative rationality (Tomasello, 2016).  

Instrumental rationality recognises that “I need to thrive” and also allows “f*** anyone who gets in my way, if I can get away with it”; i.e., it’s not moral in any way.  Essentially it is acting for the good of “me”, which is instrumentally good, and morally neither good nor bad in itself.  

Cooperative rationality recognises that “we need to thrive”, and is essentially acting for the good of “us”, and this is what morality is primarily about, if it is a regulation of the ways in which “we” collaborate to achieve “our” joint goals of mutual benefit.  Morality is cooperatively rational; it is cooperatively rational to be altruistic and fair and uphold moral norms.  

 

Mutual benefit

The goals-methods model of morality (Perry, 2025), and theories of morality-as-cooperation in general (e.g., Tomasello, 2016; Curry, 2016) recognise that morality is “about” achieving mutual benefit as a joint goal.  This is, of course, consistent with making a contract.  

 

What is moral?

I act morally when I do the right thing.  Sometimes I do the right thing from free choice, because I want to uphold moral principles for their own sake; and sometimes I do the right thing because I have to: I am obliged to for instrumental reasons.  For example, considerations of my reputation may oblige me to behave morally.  

In the goals-methods model of morality, the morality of an action depends on how well it complies with a moral norm or multiple moral norms.  Simultaneously, it might conflict with other moral norms and be wrong according to them.  

We may note that morality contains more principles and domains than just fairness and altruism.  It also includes things like patriarchy, parenting, pair-bonding, family fitness, and the incest taboo.  Women do not agree to patriarchy: it is forced on them, so the definition of morality as consisting of what someone would reasonably agree to fails on that point.  However, patriarchy aims to achieve a joint goal of mutual benefit (reproduction), albeit on men’s terms, so it fits the definition of a moral (regulatory) domain of action as consisting of a joint goal and methods of achieving it (Perry, 2025).  

Even though, as we have demonstrated, morality does not always consist of “what I would reasonably agree to”, in most cases it does.  Fairness would seem to be one of those.  It is reasonable, after all, to expect to be treated on some kind of equal basis with my partners, to receive no more and no less than my due as an equal.  

Agreement confers psychological legitimacy on what we agree to do, but it does not necessarily confer morality; i.e., something can be legitimate without being moral.  Say there are two employees A and B doing the same work at the same company for the same hours, but A earns more than B because A negotiated a higher wage when they started the job.  The wages of A and B are presumably legitimate in their minds, because each separately agreed to their wage, but still it seems a little unfair if you know what they both earn.  

 

What would be agreed by rational agents

 

A recent proposal is that such joint reasoning involves a process of “virtual bargaining” ... .  Each person figures out what would happen if both parties were able to reach an agreement through discussion and negotiation.  They then follow their part in the hypothetical virtual bargain.  Suppose that B is pushing a drinks trolley. If A and B were to discuss explicitly, they might conclude that A should move aside.  If the conclusion of this hypothetical bargaining process is “obvious”, then A may immediately do this — and the interaction is managed successfully.  Crucially, the hypothetical agreement to follow this plan has normative force.  If A violates the “obvious” agreement and barges ahead, B may complain, likely backed up by bystanders.  Notice that the bargain is created “in the moment” to best meet the needs of the immediate circumstances, and its normative force comes from its status as the natural tacit agreement ... between rational or reasonable people.  Following the virtual bargain is, in a sense, the “appropriate” or “right” course of action; violating it is reprehensible.

Arthur Le Pargneux, Nick Chater, and Hossam Zeitoun – “Contractualist tendencies and reasoning in moral judgment and decision making” (2024)

 

This may well be true.  Yet, “what is considered reasonable” is not random or arbitrary, but instead, follows well-worn rules of behaviour: moral norms.  It is a moral norm to help those in need; to make way for someone pushing a bulky drinks trolley in a narrow space.  It carries normative force because it is a method of achieving a biological goal – safeguarding the well being of the most at-risk person in the scenario.  It requires a real agreement – you go first, thank you, you’re welcome – to get the drinks trolley past A.  

It is likely that people will only agree to what they already endorse, and pull out these endorsed rules to use at the appropriate times.  

 

Contractualism, moral legitimacy, and deontology

A contractualist theory hold that the moral legitimacy of a regulatory action comes from having made an agreement to collaborate (Tomasello, 2016).  This is probably true.  Yet, the fact that moral norms are legitimate in themselves if you endorse them points to another, deeper source of legitimacy of moral action.  

According to the goals-methods model of morality, a moral norm is correct and legitimate according to itself, as it is a proven and successful method of achieving mutual well being, survival, and/or reproduction.  For individuals, it is normative to achieve these things, for biologically evolved reasons.  This suggests an overall biological framework for morality.  After all, evolutionary ethics has been called a branch of biology.  

 

Perfect Compassion

Fairness is an aspect of a larger category of ideal behaviour that I call Perfect Compassion.  The formula is:  

 

each person affected by my action is to receive the maximum benefit and minimum harm available to them.  

 

It is called “perfect” because each person achieves the maximum benefit available.  

This is inspired by the self-maximising pressure to achieve personal benefit, and by real life.  Each individual experiences an evolved biological pressure (normativity) to achieve personal benefit.  

Ideally, each person is satisfied to accept what they are given, either because no more is available, or because they have received everything they wanted or needed, or because they feel the distribution is fair.  

Resources may be distributed:

 

Two-step evolution of fairness  

According to Perry (2025), Tomasello, Melis, Tennie, Wyman, and Herrmann (2012), and Tomasello (2016), human morality developed in three major stages:

Fairness had to evolve in two stages:

  1. sharing communally, with free riders discouraged; (small nomadic groups)
  2. restriction of sharing to collaborative partners, with free riders excluded. (large settled groups)

In other words, the first distinctively moral thing that humans learned to do was to share: to give away resources voluntarily.  Great ape sociality is based on dominance relations.  A dominant will take what they want from a subordinate, and a subordinate has to give up what they have to a dominant.  Resources may be given up grudgingly in response to begging and harassment.  Great apes do not otherwise share, or very little, as they live in the deep forest where food is plentiful.  

Without sharing, there can be no cooperation, as partners require a share of the rewards if they are to be motivated to take part.  

Without the self-domestication of humans, the removal of dominance relations, and the substitution of male-male egalitarianism, there could be no sharing.  Self-domestication could have occurred in the human ancestor, the pithicene, Ardipithecus ramidus, 4-6 million years ago, as environmental conditions were growing harsh, sharing became necessary to prevent starvation, and females sexually selected for males who would share and not compete (Perry, 2025).  Self-domestication simultaneously gave birth to sharing, egalitarianism, and monogamous pair-bonding (from polygynous pair-bonding).  

The second stage of the history of fairness, restricted sharing, had to have occurred within large impersonal culturally mixed settlements, rather than in small nomadic groups.  

 

Self-other equivalence and equal treatment

As egalitarians, humans wish to be recognised equally as human beings, albeit allowing for differences in power or status.  Natural selection works on relative differences between individuals as well as on individual behaviour, and so, it is psychologically necessary for individuals to receive benefits equally with each other in some respect.  

Collaborative partners have a sense of self-other equivalence with each other (Tomasello, 2016), as they are equal and equivalent in their collaboration in several ways.  As a result of this, and of mutual value as collaborators, partners see each other as equal, deserving and respected, and this motivates other-directed fairness: I wish to be fair to my partners because they deserve it, and I respect them as equal collaborators.  

 

Nicolas Baumard, Jean-Baptiste André, and Dan Sperber’s (2013) account of the evolution of fairness by partner choice

Baumard, André, and Sperber (2013) propose that fairness evolved as a result of the proximate trade-off between “maximising my benefit” and “maximising my reputation”; i.e., between “my benefits” and “how likely I am to be chosen as a partner in the future”.  This motivates me to ensure that my partners are satisfied with what they have received.  What satisfies them is to be rewarded on some kind of equal basis, without my cheating or dominating them.  

There are two drawbacks to this model:

  1. it is only a partial story of the evolution of fairness; and
  2. as a result, the proposed model of fairness, and by extension, morality, looks terribly cold, calculating, transactional, and cynical.  It misses the warmth, humanity and compassion of human morality entirely, and gives no basis for those things.  

However, it is true (i.e., plausible) as far as it goes.  We locate it within stage 2 of the history of fairness, in large impersonal culturally mixed settlements, in open marketplaces of potential collaborative partners.  In a free marketplace of partners, my reputation is my livelihood, and other-directed fairness is compelled by this source of obligation.  

In stage 1, in small nomadic groups, sharing is not restricted to collaborative partners only, so performance of my role has less effect how well I eat, and reputation is not important in the same way.  Benefits are unconditional.  A good reputation is one for sharing and contributing.  

A likely scenario is that when sharing was first restricted to collaborative partners only, it was found that partners would only accept equal and not unequal treatment.  Proportionality represents equal treatment as there is an equal unit of output per unit of input.  

 

Jean-Baptiste André, Stéphane Debove, Léo Fitouchi, and Nicolas Baumard’s (2022) account of the evolution of morality by partner choice

 

We use an evolutionary approach to explain the existence and design features of human moral cognition.  Because humans are under selection to appear as good cooperative investments, they face a trade-off between maximizing the immediate gains of each social interaction and maximizing its long-term reputational benefits.  In a simplified game, we show that this trade-off leads individuals to behave according to the generalized Nash bargaining solution at evolutionary equilibrium.  From this result, we derive the psychological proposition that moral cognition is a calculator of this bargaining solution.

André, Debove, Fitouchi, and Baumard (2022) – “Moral cognition as a Nash product maximizer – An evolutionary contractualist account of morality”

 

This account proposes that moral cognition itself, and not just the psychology of distributive justice, evolved as a result of the trade-off between “maximising my gains” and “maximising my reputation”.  We recognise, again, that this is true as far as it goes, but it lacks some important elements of real human morality.  It accounts for moral obligation – acting morally “because I have to”, but not moral volition – “because I want to”, or compassion – “because I care”.  All of these are consistent with each other.  

Moral regulation has three sources:

The conscience is the internal motive to follow moral principles, values or norms.  The question is, why do we value moral principles or norms?  We propose that it is because “I” value the results of upholding them: i.e., mutual benefit.  Mutual benefit has several advantages for “me”: 1) I benefit in the process; 2) peace reigns in my social world; 3) my partners, upon whom I may depend, benefit too.  

The Stakeholder Principle (Roberts, 2005; Tomasello, 2016) describes and accounts for the human instincts to care for others, both in kin (a special case) and non-kin who depend on each other.  The theory is that “I care about you because I depend on you”.  Since humans are so tightly and widely interdependent, it makes sense that human compassion is also strong, and wide in scope.  

Roger Crisp (2006) proposes that only instrumental considerations of personal welfare can compel us to act morally: are the only source of moral obligation.  Threats to my reputation, protests from my partners, and a troubled conscience, are all examples of instrumental disutility.  Moral reasons for doing moral things are voluntary: “because I want to” and “because I care”.  

Another source of regulatory pressure is instrumental success: the pressure to successfully achieve our benefit that we are seeking through our cooperative behaviour, whether that is justice (as a social norm, a goal in itself), material reward, or whatever.  

 

Le Pargneux and Cushman’s (2025) study of moral judgment and bargaining power

Arthur le Pargneux and Fiery Cushman’s (2025) study “Moral Judgment Is Sensitive to Bargaining Power” seeks to find out whether moral judgment of participants by third parties is sensitive to the bargaining power of the participants, in terms of a) their stake in a successful outcome; b) outside options.  That is, to see whether the same action will be judged more or less harshly depending on the bargaining power of the participants.  This is taken as a test of contractualist theories of morality: if the result of the vignette is as if negotiated by rational agents, then it should be considered morally correct, consistent with contractualism.  

Studies 1-4 mainly looked at vignettes where partners had differing stakes in the outcome.  

 

Participants were presented with seven vignettes (an additional vignette was also used to check attention) constructed according to the following structure: Two characters can perform a mutually beneficial but unpleasant action, one of them has higher bargaining power, the other has worse bargaining power.  The bargaining power asymmetry is mainly manipulated via the value attributed to the outcome at stake by each party (well, deal, paintings, canoe, boss, stadium), and, for one vignette, via the parties’ available alternatives (cab).  Participants were asked: How morally appropriate would it be for X to ask Y to do Z?  

Le Pargneux and Cushman (2025)

 

Study 5 looked at another set of vignettes where partners had differing levels of outside options.  

In most or all of these vignettes, “my bargaining power” is inversely proportional to “what I have to lose”.

“Bargaining power” is less relevant to rationality than “what am I about to lose right now?”.  

Instrumental rationality is defined as the achievement of personal benefit and welfare.  In the moment, “who will save my camera?” is more urgently relevant to my interests and welfare than “how can I get out of doing this unpleasant action?”  Can I rely on the person who only has their lunch box to lose, to do it for me?  Should I?  It’s my camera, and I’m the one who needs to keep it.  Importantly, no real negotiation of this point is required – and not only because of time constraints.  Need is non-negotiable; it either exists or it doesn’t, and it’s outside the scope of negotiation.  What is being tested, therefore, is straightforward, urgent need, rather than negotiation.  Negotiation will always back up this result, in this context.  

If neither side could afford to lose their stake in a successful outcome, then perhaps the asymmetry of the situation would be removed.  

 

Arthur Le Pargneux, Xavier Roberts-Gaal, and Fiery Cushman’s (2025) study of fairness, negotiation and bargaining power

 

Humans constantly negotiate about how to best allocate money, food, effort, rewards, and punishments.  When dividing resources between two agents that differ from each other, some differences (like one’s initials) are morally irrelevant.  Others, such as desert, need, effort, ability, and luck, can be.  Intuitively, we sometimes favor the disadvantaged, donating more to the needy.  Other times we favor the deserving, rewarding performance, effort, or ability.What is the fair way to divide a sum of money between two people in asymmetric positions?  Drawing on contractualist models of moral cognition, we suggest that a key driver of our moral intuitions in such contexts is the logic of bargaining.  When the situation is one of negotiation and asymmetries between the recipients translate into bargaining power differences, we favor those in better bargaining positions.  When the logic of  bargaining does not apply, our intuitions are reversed, instead reflecting egalitarian or redistributive concerns.  

Le Pargneux, Arthur; Xavier Roberts-Gaal; and Fiery Cushman (2025, May 16) – “How the logic of bargaining shapes moral intuitions about resource divisions”

 

This study appears to find that when negotiation is involved, people expect and require the rewards to be proportional to bargaining power (it is “appropriate”), but when there is no negotiation, people expect equal or charitable awards.  This is not surprising, as proportionality, conditionality, and negotiation are associated with each other.  

On the other hand, we would not negotiate with a starving child whether they have earned the right to eat – their benefit is unconditional, in response to need, and proportional to their need, which may be attributed to bad luck and misfortune.  If we do not approve of their reason for need (e.g., laziness), then we will literally feel their pain less (Decety, 2011), and consequently, wish to help them less.  

 

 

 

 

 

 

Necessity plays a role in judgments of fairness.  We judge it fair if people do what they have to do.  If I was expecting 20 apples from our apple picking expedition, but half of the apples were lost on the way home, and I only get 10 apples, I would not be dissatisfied.  

 

 

References

André, J; S Debove; L Fitouchi; and N Baumard (2022) – “Moral cognition as a Nash product maximizer: An evolutionary contractualist account of morality”. https://doi.org/10.31234/osf.io/2hxgu

Baumard, Nicolas; Jean-Baptiste André; and Dan Sperber (2013) – “A Mutualistic Approach to Morality: The Evolution of Fairness by Partner Choice”; Behavioral and Brain Sciences, 36 (1), 59-78. http://dx.doi.org/10.1017/S0140525X11002202

Crisp, Roger (2006) – “Reasons and the Good”; Oxford University Press

Curry, Oliver Scott (2016) – “Morality as Cooperation: A Problem-Centred Approach” in book: “The Evolution of Morality” (pp.27-51); Chapter: 2; Springer International Publishing; https://www.researchgate.net/publication/281585949_Morality_as_Cooperation_A_Problem-Centred_Approach  

Decety, Jean (2011) – “The Neuroevolution of Empathy”; Annals of the New York Academy of Sciences 1231

Le Pargneux, Arthur; Nick Chater; and Hossam Zeitoun (2024) – “Contractualist tendencies and reasoning in moral judgment and decision making”; Cognition 249 105838; https://doi.org/10.1016/j.cognition.2024.105838

Le Pargneux, Arthur; and Fiery Cushman (2025) – “Moral Judgment Is Sensitive to Bargaining Power”; Journal of Experimental Psychology: General; 2025, Vol. 154, No. 1, 263–278; https://doi.org/10.1037/xge0001678

Le Pargneux, Arthur; Xavier Roberts-Gaal; and Fiery Cushman (2025, May 16) – “How the logic of bargaining shapes moral intuitions about resource divisions”; PsyArXiv; https://doi.org/10.31234/osf.io/3uqks_v1​

Perry, Simon (2025) – “Understanding morality and ethics”; https://orangebud.co.uk/web_book_2.html  

Roberts, Gilbert (2005) – “Cooperation through interdependence”: Animal Behaviour, 70, 901–908, https://www.academia.edu/28485879/Cooperation_through_interdependence

Tomasello, Michael (2016) – “A Natural History of Human Morality”; Harvard University Press, Cambridge MA

Tomasello, Michael; Alicia P Melis; Claudio Tennie; Emily Wyman; Esther Herrmann (2012) – “Two Key Steps in the Evolution of Human Cooperation – The Interdependence Hypothesis” – Current Anthropology, vol. 53, no. 6, Dec 2012

 

IN RESPONSE TO

BENEFIT

deservingness

proportional

conditional

negotiated

need

proportional

unconditional

donated