PHD THESIS SUMMARY: Amartya Sen’s Earlier Conception of Economic Agents through the Origins and Development of his Capability Approach (1970–1993)

Amartya Sen’s (1977b, 1982, 1985b, 1987) conception of economic agents assigned a significant role to the analysis of rational economic motivations and their consequences on individual decision-making. In particular, Sen’s conception of economic agents differed from those of his contemporaries—especially those committed to the mainstream idea of homo economicus. This is because he argued that individuals can still be rational even when their choices result in a decrease of individual welfare. This idea was in line with Sen’s broader criticism of mainstream economics and its commitment to homo economicus. However, Sen did not devote any work to outline and defend his own conception of economic agents. Rather, he provided elements of this conception across his oeuvre. Thus, we may rebuild Sen’s conception of economic agents but only through an extensive reading of his writings. Among the existing literature on Sen’s conception of economic agents, I argue that the most interesting work is that of John B. Davis (2003, 2007, 2009, 2012). As mentioned in the epigraph, Sen (2012) admits that Davis is correct to claim that his theory of justice (Sen 2009) largely depends on his understanding of economic agents, although this point is often implicit in Sen’s own writings. Some years later, Giovanola (2013) emphasized a different element compared to Davis’ works, namely that Sen’s criticism of homo economicus 1 leads to a re-thinking of the anthropological and ethical foundations of economics (Giovanola 2009), including his understanding of economic agents. Intrigued by these readings of Sen, I began by asking whether Sen does have a distinct view of economic agents. I then rebuilt the origins and development of Sen’s understanding of economic agents. On the basis of


PHD THESIS SUMMARY: Amartya Sen's Earlier Conception of Economic Agents through the Origins and Development of his Capability
Approach  VALENTINA ERASMO

PhD in Ethics and Economics and History of Economic Thought, June 2020 Università degli Studi "G. d'Annunzio" Chieti-Pescara
Amartya Sen's (1977bSen's ( , 1982Sen's ( , 1985bSen's ( , 1987 conception of economic agents assigned a significant role to the analysis of rational economic motivations and their consequences on individual decision-making. In particular, Sen's conception of economic agents differed from those of his contemporaries-especially those committed to the mainstream idea of homo economicus. This is because he argued that individuals can still be rational even when their choices result in a decrease of individual welfare. This idea was in line with Sen's broader criticism of mainstream economics and its commitment to homo economicus. However, Sen did not devote any work to outline and defend his own conception of economic agents. Rather, he provided elements of this conception across his oeuvre. Thus, we may rebuild Sen's conception of economic agents but only through an extensive reading of his writings. Among the existing literature on Sen's conception of economic agents, I argue that the most interesting work is that of John B. Davis (2003Davis ( , 2007Davis ( , 2009Davis ( , 2012. As mentioned in the epigraph, Sen (2012) admits that Davis is correct to claim that his theory of justice (Sen 2009) largely depends on his understanding of economic agents, although this point is often implicit in Sen's own writings. Some years later, Giovanola (2013) emphasized a different element compared to Davis' works, namely that Sen's criticism of homo economicus 1 leads to a re-thinking of the anthropological and ethical foundations of economics (Giovanola 2009), including his understanding of economic agents.
Intrigued by these readings of Sen, I began by asking whether Sen does have a distinct view of economic agents. I then rebuilt the origins and development of Sen's understanding of economic agents. On the basis of this historical approach, I argue that Sen began to develop his earlier conception of economic agents between the 1970s and the 1990s, alongside the origins and development of his capability approach.
I develop this argument in my PhD Thesis in three chapters. In the first chapter, "The Foundational Decade to the Capability Approach (1970)(1971)(1972)(1973)(1974)(1975)(1976)(1977)(1978)(1979)(1980): An Ethical-Economic Analysis of Sen's Social Choice Theory and Welfare Economics", I analyse the intellectual origins of Sen's capability approach. In the second chapter, "The Development of the Capability Approach (1980)(1981)(1982)(1983)(1984)(1985)(1986)(1987)(1988)(1989)(1990)(1991)(1992)(1993): Three Conceptual Extensions Compared to Mainstream Economics", I focus on the development of capability approach, emphasizing that this development extended mainstream economics. In the third and final chapter, "Amartya Sen's Earlier Conception of Economic Agents: A Detailed Outline", I provide a synthesis and extension of those elements that arise in the previous chapters with a detailed outline of Sen's conception of economic agents. I discuss each of these chapters in turn.
The first chapter shows that understanding the intellectual origins of the capability approach-which is found in Sen's work on social choice theory-is useful for the following reason: it brings attention to important questions such as collective choices, inequality, social welfare functions, and rationality, which are crucial for a reconstruction of Sen's earlier conception of economic agents. These intellectual origins of the capability approach are located in Sen's debates with Kenneth Arrow, John Rawls, and John Harsanyi.
In particular, Sen differed from Arrow (1951) with respect to the idea of rationality. Arrow understood rationality as logical consistency. This is characterized by moral and ethical neutrality in economic behaviour, as in the neoclassical tradition. Sen extends this with the claim that liberalism is also important to social welfare evaluation. Specifically, he emphasized the role of 'minimal liberty'. As Gaertner, Pattanaik, and Suzumura (1992) have pointed out, Sen (1970) introduced individual rights in terms of a 'recognized personal sphere' that everybody should have and nobody can interfere with. This sphere is of importance in the context of his earlier conception of economic agents because it views economic agents as independent in their individual decision-making (Chung 2019).
Although Sen appreciated Rawls' focus on justice and fairness, which comprised Rawls' ethical and economic rationality, Sen (1974) rejected both the Rawlsian and utilitarian decision criteria as measures to study and evaluate inequalities. This is because they "run into some fairly straightforward difficulties, since each leaves out completely one of the two parts of the total picture" (Sen 1974, 308). In particular, Rawls leaves out comparisons of levels of welfare, and utilitarians leave out comparisons of welfare gains and losses. This need to elaborate "a more complete theory" (Sen 1974, 308) for studying and evaluating inequalities prepared the ground for Sen's redefinition of economic agents in terms of responsibility.
Finally, the exchange between Sen (1976Sen ( , 1977a and Harsanyi (1977) is characterized by their debate on the representation of social welfare functions. Harsanyi used linear social welfare functions, which are "the arithmetic mean of the different individuals' von Neumann Morgenstern (vNM) utility functions" (Harsanyi 1977, 293). In contrast, Sen promoted non-linear social welfare functions which "refer to representations other than those emanating from the von Neumann Morgenstern system" (Sen 1977a, 299). Sen criticized Harsanyi because his theorems are "about the representability of social preferences and not about utilitarianism, as the latter requires a concept of utility that is not based exclusively on preference" (Weymark 1991, 256). Indeed, when such a notion of utility is used, utility functions might not be vNM functions, although the expected utility axioms were satisfied. Sen (1977a) thus concluded that Harsanyi's theorems imply that social welfare functions are rather "nonlinear […], and hence obviously nonutilitarian" (Weymark 1991, 257). In particular, Sen rejected linear welfare functions because they are not able to represent inequalities. As with Sen's debate with Rawls, his debate with Harsanyi also helped lay the groundwork for Sen's redefinition of economic agents in terms of responsibility.
The second chapter claims that, simultaneously with the development of the capability approach (Sen 1980(Sen , 1985a, Sen makes three conceptual extensions of mainstream economics-these are (i) from welfare to wellbeing; (ii) from agent to agency; (iii) from Paretian to Liberal Efficiency. 2 These conceptual extensions are relevant for Sen's redefinition of economic agents.
The first extension-from welfare to well-being-advances Sen's (1987) criticism of welfarism. In a nutshell, this criticism was based on the idea that the informational basis of social choices is inadequate. This is why Sen argued that economics should return to ethics, emphasizing the role of normativity in economics, for a more complex analysis of the reality that surrounds us. This return to ethics enabled Sen to create those methodological conditions for considering responsibility in individual decision-making.
The second extension-from agent to agency-dealt more directly with a redefinition of the economic agent by developing two key concepts. One was the concept of agency, and the other was a dualistic conception of a person in ethical calculations. On the first, Sen (1982) addressed the relationship between rights and goals, arguing for his alternative to welfarist consequentialism and constraint-based theories. This alternative is a consequentialist approach but is devoid of its welfarist contents. Sen achieves this by placing an intrinsic value on rights. In particular, the "fulfilment and nonrealization of rights are included among the goals, incorporated in the evaluation of states of affairs, and then applied to the choice of actions through consequential links [in what] will be called a goal rights system" (Sen 1982, 15). In this way, Sen elaborated his ethical and moral approach where the economic agents became individuals who have agency. Thus, they are responsible for the consequences of their behaviours, instead of being neutral economic agents in their decision-making, as in mainstream economics. This consideration of consequences in individual economic behaviour justifies why individuals are still rational even when their choices cause a decrease of individual welfare. For instance, responsible economic agents might rationally accept a decrease of their individual welfare in order to reduce inequalities.
On the dualistic conception of a person in ethical calculation, Sen (1987) distinguished agency from well-being (thanks to the extension from welfare to well-being). This was done by including more motivations for action than just self-interested (or selfish) ones. In this way, Sen confirmed the reintroduction of responsibility in economics. So, a person might be understood, on the one hand, as an agent with the ability to form goals, commitments, and values which could be "exercised at the individual level, or in groups, or through democratic participation" (Alkire 2005, 219). On the other hand, we may understand a person in terms of their well-being, which consists in being "concerned with a person's achievement: how 'well' is his or her 'being'?" (Sen 1985a, 5). This is in contrast with the tautological formulation of well-being in mainstream homo economicus (Sen 1987). This is because whatever homo economicus chooses contributes the most to the agent's well-being as it represents the alternative that maximizes utility-solely interpreted in terms of wellbeing and ignoring agency. This espouses a monistic conception of a person in ethical evaluation.
The third and final extension-from Paretian to Liberal Efficiencywas articulated during the 1980s and the beginning of the 1990s. This consisted in Sen (1985cSen ( , 1993 carrying out an ethical and economic analysis of the market and its relationship with individual freedoms. According to Sen (1985c), the market and its mechanisms have a moral status, which he analysed through three different hypotheses, namely: (i) the moral necessity of the market; (ii) the freedom of choice deriving from the market, and (iii) its legitimation ascribable to antecedent rights. In "Markets and Freedoms", Sen (1993) showed how the market might limit or promote freedoms, introducing the distinction between the 'process which leads to freedom' and 'opportunity-freedoms'.
In this way, efficiency is understood in terms of 'efficiency of opportunity-freedoms'. Because of the centrality of freedom to the evaluation of states of affairs, Sen was arguing for the need for more information apart from Paretian efficiency 3 in the analysis of the market and its mechanisms. In particular, through the reintroduction of rights in economics, the centrality of market freedoms was replaced by that of individual freedoms, completing the extension from Paretian to Liberal Efficiency. This element is in line with Sen's redefinitions of welfare as well-being (from welfare to well-being extension) and of responsible economic agents (from agent to agency extension). In turn, this reintroduction of rights in economics shows that Sen's earlier conception of economic agents also leads to an anthropocentric understanding of the market because the centrality of market freedoms was replaced by that of individual freedoms.
Despite this schematization, these extensions are not discrete categories. Indeed, the categories of normativity, responsibility, and rights reintroduced by Sen inform each other and jointly offer a superior theory of well-being compared to that which prevails in mainstream economics. More specifically, the reintroduction of normativity in welfare economics has contributed to Sen's conception of responsible economic agents. The reintroduction of responsibility in economics also finds its origins in the question of individual rights. Consequently, individual rights are the category from which the extension from Paretian efficiency to opportunityfreedoms began. Together, these categories offer a superior theory of well-being because they include more ethical elements in the analysis of individual decision-making than mainstream economics does.
In the third chapter, I provide a synthesis and an extension of the elements that arise in the previous two sections, especially those of rationality and agency. The aim here is to locate Sen's earlier conception of economic agents in a broader intellectual history where, on the one hand, he was simultaneously influenced by Greek philosophical tradition and the Enlightenment; while, on the other hand, he was influencing the works on relational goods and Ricoeur's (2005aRicoeur's ( , 2005bRicoeur's ( , 2006 later works. 4 Sen explains how economic rationality enables us to know individual preferences, such as the set of human needs, while ethical rationality integrates and is in conversation with this perspective (Erasmo 2019). This is possible thanks to the extension of the horizon of individual values that direct people's choices, by considering 'rational' motivations that are distinct from selfish or self-referential ones, such as sympathy and commitment (Sen 1977b(Sen , 1985b. These two concepts are set out by him as follows: the former corresponds to the case in which the concern for others directly affects one's own welfare. If the knowledge of torture of others makes you sick, it is a case of sympathy; if it does not make you feel personally worse off, but you think it is wrong and you are ready to do something to stop it, it is a case of commitment (Sen 1977b, 326).
Commitment may be defined "in terms of a person choosing an act that he believes will yield a lower level of personal welfare to him than an alternative that is also available to him" (Sen 1977b, 327). This consideration of others' goals and choices in decision-making leads to a primacy of actions over outcomes (Sen 1997) and is the opposite of what one sees in the idea of homo economicus. Consequently, persons are just not economic agents, but protagonists of actions who are responsible for their actions, as has been previously discussed.
All these elements can be located in a broader intellectual context where Sen was simultaneously influenced by Greek philosophical tradition and the Enlightenment, while also influencing the works on relational goods, and Ricoeur's later work in particular. Sen was especially influenced by Aristotle's concepts of kalokagathia and phronesis, respectively 'beautiful and good' and 'practical wisdom' (Petrochilos 2002). In particular, Aristotle's phronesis has the same value as Sen's ethical rationality in decision-making in that it shows how to attain collective well-being in a way different to homo economicus' ethical individualism. Hence, phronesis is that virtue at the basis of commitment which modifies individual goals and choices by including those of others and adopting fair and virtuous behaviours.
The influence of the Enlightenment comes from Adam Smith and Marquis de Condorcet, who were the pioneers of what Rothschild (2001) calls the 'warm Enlightenment of economic sentiments'. Here, rationality is not indifferent, uniform, and cold. According to Smith and Condorcet, rationality is 'warm', including motivations in decision-making other than those selfish or self-referential ones, as in Sen's earlier conception of economic agents. In particular, Sen's concept of commitment appears to be influenced by Smith's (1759) 'sympathy'. Their respective analyses have in common the same 'plastic' strength of social interactions (Khalil 1990) and are a criterion of moral approval and disapproval (Raphael 1985). According to Walsh (2003), this proximity makes Sen a key figure for a 'renaissance or second phase of classical economics' in the contemporary age.
The emphasis Sen placed on social interactions in economic relationships was influential to authors committed to relational goods, such as Nussbaum (1986). This is seen most clearly in the influence on Paul Ricoeur's The Course of Recognition, where Ricoeur attempts to bridge the gap between self-recognition and mutual recognition through Sen's ideas about agency and social capacities. Social capacities are extremely heterogeneous, and their most fully developed form equates with "rights and capabilities", while their "anthropological ground" is "the power to act, agency" (Ricoeur 2005b, 135). Indeed, Ricoeur is the first who recognized the anthropological value 5 of Sen's capability approach and is the most authoritative voice for supporting my argument about the strong relationship between Sen's capability approach and his conception of economic agents.
In conclusion, this historical analysis has shown that Sen elaborated his earlier conception of economic agents alongside the origins and development of his capability approach. More specifically, the study of intellectual origins of the capability approach has uncovered the relevance of a different understanding of rationality compared to that of mainstream homo economicus since the 1970s. In addition, the analysis of the