Reciprocity is an old issue in economics. Given its focus on human interactions, economics may even be defined as the study of relations of reciprocity. Such a thesis however would be true only if we understood reciprocity in the old Latin and Greek original meaning. Reciprocity comes from reciprocus or reciprocitate, which means ‘returning the same way, alternating’: reci-procus, where reci is from recus (from re- ‘back’ + -cus, adjective formation), and procus (from pro- ‘forward’ + -cus, adjective formation). The verb reciprocate means ‘to return, requite’. Etymological dictionaries translate reciprocity as ‘retrogression, alternation, ebb’, or ‘move back and forth’. Aristotle, for example, uses the expression antipeponthos in order to express both commercial relations and civil ones, because an idea of proportionality and mutuality exists in all relations within the polis.

The idea of reciprocity in contemporary economics is rather different, being much more limited than reci-pro-city. In economic theory, reciprocity is normally associated with other-regarding preferences, which have most commonly been represented by the assumption of altruism – that is, a positive concern of one person for another person's welfare or utility. A few marginalist economists (from Edgeworth to Pantaleoni) at the end of the XIX century had in fact hypothesized, without departing from the homo oeconomicus paradigm, that agents might undertake actions not motivated by self-interest. Such behaviour, though, was not considered as especially significant in accounting for economic interaction, which was conveniently limited to the simpler but realistic assumption that agents, when operating in the market, do not take into account other agents’ well-being – or, at least, not in a noteworthy degree.

In Serge-Christoph Kolm's Reciprocity. An Economics of Social Relations, there is no reference to the historical development of the concept of reciprocity. This absence was surprising to me, the issue of reciprocity being a classical locus in the western tradition. Kolm's book takes instead the issue of reciprocity as if it were something totally new. Reciprocity is nevertheless an important book full of insights, inspiring analyses, and original discussions of key issues in economics and social science more generally. It is also a big book that aims at offering a general theory of reciprocity, in particular an economics of reciprocity intended as the paradigm of all social relations.

The book is organized in four parts: (I) Facts and forms (of reciprocity), (II) Motives, (III) Values and reasons, (IV) The economics of reciprocity. The core of the analysis is a ternary theory of reciprocity: (1) Balance (or comparative) reciprocity, (2) Liking reciprocity, (3) Continuation reciprocity. Balance reciprocity ‘results from the urge or desire to reciprocate the gift or favour with a return gift or favour that has a certain relation of equality with the former’ (p. 105). In liking reciprocity, instead, ‘liking someone is a standard, direct reason for giving or to favoring her’ (p. 117). Continuation reciprocity is reacting ‘to an act, an attitude or an expressed judgment in order to induce its repetition’ (p. 134).

The book in all its four parts discusses and analyses, in a sort of ‘progressive approximations’ methodology (that inevitably generates many repetitions), the various forms of reciprocity, mentions also other ways of looking at reciprocity (for instance direct, indirect), and ends with some formal models of reciprocity built using the technical tools of neoclassical economics: individual preferences and utility functions.

In this note I will discuss critically two issues that, in my opinion, are at the core of both Kolm's theory of reciprocity and all his studies on the subject that began in the early 1980s. (Kolm is in fact a true pioneer of the contemporary theory of reciprocity.)

The first point I would like to address is a critical discussion of the model ‘The effect of liking on satisfactions’ (pp. 275 ff.), where Kolm presents the key elements of his theory of liking reciprocity. In this model the utility of the individual i (Ui) depends, among other things, on the utility of j, the person i likes or loves (Uj), on the liking of j towards i (lj), and j's other general conditions (Sj). So the utility function of i is: Ui = f(Uj, lj, Sj). Kolm compares two situations: one case (li) in which i loves j not too much, with the other case (l'i) when i loves relatively more her partner (then l'i > li). The two cases are represented in a graph with two lines with positive slopes (the slope of l'i is steeper than that of li). Because of their different slopes, the lines intersect: on the left l'i is above li, on the right below (see Figure 1).

Figure 1.

The structure of liking reciprocity.

Kolm's discussion of his model goes as follows: ‘When individual i's liking li increases, these effects make individual i more or less satisfied . . . If the other person is happy, otherwise in a good situation, and likes you, liking her more tends to make you appreciate these facts more and a priori makes you happier for this reason. However, if the other is unhappy, otherwise in a poor situation, or if she likes you too little, liking her makes you resent more these insufficiencies and lowers your satisfaction’ (p. 278).

This model can be interpreted in two different ways. The first is in line with the Aristotelian idea of the fragility of happiness, a tradition of thought rediscovered in particular by Martha Nussbaum (Reference Nussbaum1986). In fact, the model of ‘liking reciprocity’ can be interpreted as an argument in favour of the ‘tragicity’ of good life, because the possibility of a greater happiness coming from reciprocal love in good times is ‘paid’ by the greater pain in the moments of suffering of friends and partners. Good life, says classical Greek culture, along with the Bible, is a twisted experience of pain and joy, which are two sides of the same coin.

However, there is also a different and more problematic reading of that model and of Kolm's whole theory of reciprocity, which is consistent with Douglas Bernheim and Oded Stark's well-known paper ‘Altruism within the family reconsidered: Do nice guys finish last?’ (1998). The way altruism is treated in that paper is very similar to Kolm's theory of liking reciprocity. Both theories represent pro-social behaviour (altruism and reciprocity) as a matter of individual preferences. In their paper, Bernheim and Stark demonstrate why ‘nice guys finish last’, where ‘nice’ for them means to have preferences sensitive to the utility of the other/s: ‘The explanation is quite simple. An altruistic type A would be depressed by his partner's low level of felicity. Since the type B cares about her partner, she would in turn be disturbed by the fact that she has made him unhappy’ (Reference Bernheim and Stark1988, p. 1036). Kolm's model of reciprocity is also based on the hypothesis of the interdependence of utility, and for that reason his theory is vulnerable to the same criticism addressed to Bernheim and Stark's model: ‘this hypothesis is contrary to the facts of experience. When we are unhappy, we want sympathy; other people's sympathy makes us feel less unhappy, not more. We want partners who will sympathize with our pleasure and our pains’ (Sugden Reference Sugden, Bruni and Porta2005: 97).

The kind of paradox highlighted by Sugden is the consequence of a key methodological issue, namely the idea that a relationship, such as reciprocity, can be described and analysed in terms of individual preferences and beliefs – an issue that is even more relevant for Kolm's way of approaching reciprocity than for Bernheim and Stark's theory of altruism (because while altruism can be understood as a unilateral action and not as a relationship, reciprocity cannot). From this perspective, Kolm's theory of reciprocity is perfectly consistent with the way contemporary economics deals with reciprocity and pro-social behaviour, the so-called theory of social preferences, which considers phenomena that are essentially relationships as a matter of individual preferences. In the last few years, in fact, several context-specific theories of ‘social preferences’ have been developed in economics, along with the rise of experimental economics and behavioural economics. Matthew Rabin (Reference Rabin1993) for example has proposed that people have preferences for benefiting those who benefit them and for harming those who harm them. According to this theory, which has greatly influenced the whole social preferences theory (Fehr and Gächter Reference Fehr and Gächter1998), agents are not generous or non-generous indistinctively towards everyone, but show a certain degree of conditionality and selectivity in their reciprocating action. It is common to all these theories that the authentic social element of a person's preferences is revealed in her willingness to sacrifice her own interests in order to benefit or harm others. In the so-called trust game or ultimatum game, this nexus between social preferences and sacrifice of individual pay-off is particular evident.

The theory of social preferences therefore restates an old and deep-rooted idea that is at the core of modern economics, namely the opposition between genuine social behaviour and ordinary market interaction: if one wants to be authentically social, she has to renounce, or be ready to renounce, individual payoffs. Interactions are represented as a game in which gains in genuine relational terms correspond to losses in strict economic terms. From this point of view there is no genuine novelty in social preference theory, because it is fully consistent with this old assumption of economic theory (Bruni and Sugden Reference Bruni and Sugden2008).

For this reason, although Kolm's Reciprocity is a genuine attempt to go beyond the neoclassical approach to human behaviour, it is still based on the same two key ideas: (1) that true sociality, i.e. positive reciprocity, is a matter of individual preferences; (2) that the ‘sociality’ of preferences is measured by the willingness to sacrifice individual self-interest, or by a gift, intended as a costly and self-sacrificing action of A in order to benefit B.

After reading Kolm's theory of reciprocity I am more and more convinced that it is very difficult to understand and describe reciprocity, that is a relationship, remaining within the framework of rational choice theory, with its utility functions and individual preferences. Both Kolm and the social preferences theorists analyse reciprocity by complicating the argument of the utility function, introducing the hypothesis of more complex individual preferences. Something new, also in terms of formal and mathematical language, probably is needed. If we take seriously the fact that reciprocity is a relationship, namely something more than a sum of individual separate actions, than we have to challenge one of the most basic hypotheses of neoclassical economics: methodological individualism, namely the idea that individuals and their preferences are the primitive building blocks of every economics analysis. The theory of ‘we rationality’ developed by Michael Bacharach, Robert Sugden and others, is an interesting move in this direction. Social interactions and reciprocity are analysed using collective preferences and ‘team’ utility functions.

Now my second remark. At page 103, Kolm presents a graphical scheme (figure 5.2) that summarizes his general theory of reciprocity. The building block at the basis of his three forms of reciprocity is the ‘gift received’. All the branches of the tree depart from the received gift (balance, liking, continuation reciprocity): in that graph the ‘returned gift’ closes the circle of reciprocity. This central and essential role of the gift is a cornerstone of Kolm's book. At the very beginning of his essay, Kolm gives a general definition of reciprocity: ‘Reciprocity is treating other people as other people treat you, voluntarily and not as a result of a binding exchange agreement’ (p. 1). This broad definition of reciprocity raises however two basic questions:

1. (1) Why contrast ‘voluntarily’ and ‘binding agreement exchange’?

2. (2) Why, if Kolm defines reciprocity ‘treating the other people as the other treat you’ does he exclude ‘negative’ forms of reciprocity, such as retaliation or punishment (p. 1)? I don't address question 2 here, but only note that ‘negative’ reciprocity is no less relevant than ‘positive’ one, if one aims, as Kolm does, at a general ‘theory of social relations’.

Question (1) is instead crucial. Contracts and market exchange are surely forms of reciprocity: not in the sense of gift exchange, but in the old and traditional way of understanding reci-pro-city. In a market contract, in fact, two or more people voluntarily choose to bind themselves in order to build a mutual advantageous relationship. Why, then, a contractual relation between entrepreneur and employee, baker and client (in Adam Smith's words), nurse and patient, cannot be understood as a form of reciprocity, when surely the principle ‘treat the other as the other treat you’ applies in these contexts? Of course these are expressions of a different form of reciprocity with respect to a gift exchange relation; nonetheless it is reciprocity too, an essential part of those ‘social relations’ that Kolm intends to analyse in his book. We know from social and economic history that in order to make the reciprocity of gift exchange sustainable in the long run (which Kolm calls continuation reciprocity), gifts have to be combined with contracts.

Why then must the search for mutual benefit in market be read as unilateral egoism, and, conversely, reciprocity must be considered an exclusive characteristic of non-market relations (i.e. in politics, civil society, family)? These are often considered paradigmatic cases of true ‘authentic community’ because they are seen as non-instrumental, intrinsic, and gift-based. ‘Relationships’, ‘community’, ‘friendship’, ‘love’ and so on repeatedly appear in Kolm's pages as objects that might be sought only by intrinsically motivated individuals, and such a theory seems unable to make sense of the idea that the achievement of mutual benefit through exchange in markets might be the content of an authentic relationship, i.e. might be a form of positive reciprocity. The implication of such a theory is that the ordinary economic transactions on which we all depend for our survival are somehow inauthentic.

The opposition between true or genuine reciprocity and market relations cuts through the whole of Kolm's analysis. From this point of view, Kolm's theory of reciprocity is fully consistent with the communitarian approach to market and reciprocity. For both (Kolm and the communitarians) market is a separate domain or ‘sphere’ of social life, depicted as the realm of instrumental, external and extrinsic dealings, the place of non-genuine dimensions of human interactions. Normal economic life crowds out, corrodes morality (in the sense of what is valuable in human life). Hence the need to constrain markets within strict bounds (Walzer Reference Walzer1983). Markets can play a positive role, but within their own sphere, which is radically distinct from the sphere of gift-exchange and from reciprocity (family, kinship, community): again the opposition between market and genuine social relations is stark. Such an attitude toward markets can also be seen as a romantic yearning for a lost Rousseauian age of economic innocence.

Although this opposition between truly moral behaviour or reciprocity and ordinary market relations can find support in some sectors of public opinion (particularly in this time of economic and financial crisis), such a cultural attitude can nevertheless be dangerous, because the economic system and civil society depend on exchange to satisfy mutual needs, on various forms of reciprocity, including a market-based one. This reciprocity has necessarily to be somehow instrumental and not gift-based, because each person has to satisfy others’ needs as a means to satisfying her own. Many essential social activities are scarce, and there is a strong need for coordination in a large free society: it is not possible for each of us to have our wants satisfied while each of us also does what she finds intrinsically motivating only. A gift-exchange society requires also the reciprocity of market interactions. Even a small community or a family cannot function properly on the basis of gifts and intrinsic motivations. Those who contrast moral societies to market interactions forget that market prices are also, in efficient market systems, signals of social scarcity. Markets, in fact, allow demand and supply of socially useful activities to meet in a liberal and non-authoritarian system (as in the patriarchal family or in feudal and illiberal communities).

Neither in the conceptual framework of modern economics nor in Kolm's theory of reciprocity do we find a way of conceiving of a relationship between individuals as both a mutually beneficial exchange, in which neither partner makes a sacrifice for the benefit of the other, and a genuinely social interaction, carrying moral value by virtue of this social content. If reciprocity is captured by the principle ‘treat others as the others treat you’, then gift-exchange, although important, cannot be enough for a good society based on reciprocity.