A capabilitarian behavioral economics: what behavioral economics can learn from the capability approach

Abstract

Behavioral economics (BE) accounts for how people actually behave has proven to be very influential in multiple domains. Its insights have dislodged settled assumptions and put forward an improved understanding of human beings and their agency. Although it challenges the conventional approach, advanced by rational choice theory (RCT), because it also formally builds on it, BE shares some of RCT’s advantages as well as its shortcomings. One the one hand, it shares an important focus on freedom. On the other, it is limited to self-interest as the main motivation for human action. The latter is a limitation since human experience is much richer than that. To complement BE’s limitation, this paper suggests Amartya Sen’s Capability Approach (CA). This framework places humans and their quality of life at the locus of attention for the assessment of social states. To do so, it challenges RCT in a more fundamental way and it provides a richer account of human agency, which includes both self-regarding as well as other-regarding motivations. This combination enriches BE’s realism as well as practicality.

1 Introduction

The traditional model of decision-making assumes people decide rationally and portrays human beings as virtually infallible in their pursuit of their preferences. Rational choice theory (RCT), which favors elegance and parsimony and relies on a limited set of assumptions (Sen 1977), establishes an axiomatic framework to understand individual behavior. Although the implications of full rationality lack consensus (Wittek et al. 2013), it has been argued that it includes at least three unrealistic features namely unboundedness in rationality, willpower, and selfishness (Mullainathan and Thaler 2001). However, this does not imply that all individuals consistently exhibit all of them; rather, it suggests that most people tend to act rationally most of the time, at best, or that they act as if they were rational, at worst (Corr and Plagnol 2019). This framework assumes that deviations from rationality or instances of unconventional behavior are rare enough not to significantly affect the explanatory power of the theory.

This so-called homo economicus model, which portrays a rational decision-maker, is not purely a descriptive explanation of human behavior; it carries normative implications as well (Elster 1986). It doesn't merely describe how individuals make decisions and act; it also prescribes how they should do both. In this sense, it goes beyond elucidating human behavior to encourage individuals to conform to a predetermined model. This can lead to a self-fulfilling prophecy where people, influenced by the prevailing ideas of rational choice, strive to behave rationally and align their actions with the theory's axioms, ultimately reinforcing the theory's explanatory power.

In reality, there is a substantial gap between actual human behavior and the model of complete rationality proposed by the homo economicus. In practice, people frequently fall short of achieving their goals, even when making conscious efforts to do so (Thaler and Sunstein 2008, 2021; Corr and Plagnol 2019; Kahneman 2011). These deviations from conventional prescriptions have been extensively examined by various disciplines, and their collective insights have contributed to the development of behavioral economics (BE). In brief, BE accounts for how people actually behave (Sunstein 2020). That is, it seeks provide a descriptive framework of decision-making.

As such, BE has proven to be influential among scholars and practitioners. Although it challenges the conventional approach, because it also uses it as a benchmark of sorts, BE shares some of RCT’s advantages as well as its shortcomings. Regarding the latter, much like RCT, BE assumes self-interest is the main motivation for human action, often at the expense of others. This is a limitation because human experience demonstrates that although selfish motivations are important, they are neither the most let alone the only important motivations guiding action (see e.g. Sen 1999; Le Grand 2006). Addressing this limitation is necessary for a framework seeking to account for how people actually behave, and that is what this paper seeks to do.

To that end, it seems warranted to complement BE’s insights at the theoretical level, in its account of human beings and their agency. This paper puts forward Amartya Sen’s Capability Approach (CA) to meet that tall order. The CA is a framework that defies RTC at a fundamental level to provide a richer account of human beings (Author 2020a, b, c, 2022a, b, c, 2023). Most relevantly for this paper, the CA encompasses the totality of motivations guiding people, which means that it accounts for both self-regarding as well as other-regarding goals (Sen 1999; Crocker and Robeyns 2010). To be sure, Sen’s CA does not pay sufficient attention either to other-regarding motivations, in fact it lacks a term to capture them in isolation, or non-human others, which is why this paper addresses both.¹

There is promise in combining the CA and BE. It has been argued that BE can enrich the CA significantly. That case is made in this very forum. Focusing on BE’s insights on rationality and decision-making, it shows that such combination makes two additions to the CA: (1) non-reasoning as an internal conversion factor; and, (2) choice architecture as an external one (Author 2023).

Since both BE and the CA are quintessentially policy-oriented, the implications of such combination can be policy relevant. Since BE’s insights have been powerfully used in approaches to government such as Libertarian Paternalism and the much influential nudge literature and policy agenda, such enrichment can contribute to both decision- and policy-making alike. This is particularly evident in the case of the growing and promising literature using BE’s insights to tackle environmental issues (see e.g. see, e.g. Beckenbach and Kahlenborn 2014; Carlsson et al. 2019; Sunstein and Reisch 2014), which lacks a sound foundation.

Indeed, this is significant because theories that offer an unrealistic account of human behavior can be counterproductive or even detrimental to human wellbeing and the welfare of non-human beings.² Given the impact of assumptions about human agency on policy-making, it is crucial to have an accurate understanding of human behavior, as emphasized by Thaler (2015, p. 25):

We don’t have to stop inventing abstract models that describe the behavior of imaginary Econs. We do, however, have to stop assuming that those models are accurate descriptions of behavior, and stop basing policy decisions on such flawed analyses.

To explore this possibility and elaborate that argument this paper has five sections, besides the introduction. A presentation, however brief, of BE is elaborated in the first section and it exposes the shortcoming identified in the literature. The second section develops the CA and its account of human beings with a focus on their motivations. The specification of other-regarding motivations, the relevance of non-human beings within the CA, and a brief explanation of altruistic behavior in terms of emotions is elaborated in the third. The plausibility of enriching BE with the CA is explored and advanced in the fourth section. The final section concludes.

2 Behavioral economics: its scope and limitation regarding motivation

BE challenges the conventional 'rational' approach in economics and aims to provide a more accurate depiction of human behavior.³ “The standard economic model of human behavior includes (at least) three unrealistic traits: unbounded rationality, unbounded willpower, and unbounded selfishness” (Mullainathan and Thaler 2001, p. 1095). To provide a more realistic depiction, BE integrates insights from the social sciences, particularly psychology, into economic thinking. In essence, BE “conducts economic analysis while acknowledging how humans genuinely behave” (Sunstein 2020, p. 2). Unlike mainstream economics, which is based on RCT and relies heavily on a few foundational assumptions and axioms for building theories and predictions, BE takes a different approach, while using RCT and its yardstick (Corr and Plagnol 2019, p. 4). It recognizes that human decision-making deviates from the model of full rationality because it is limited internally as well as externally, and explores the implications of these limitations. As this section elaborates, BE has focused on mostly on rationality and willpower, leaving the issue of selfishness or motivation virtually unaccounted for.⁴

2.1 Bounded rationality

Herbert Simon, Nobel Laureate, played a crucial role in shedding light on departures from rationality through his work on ‘bounded rationality’. This proposal essentially posits that individuals strive to make rational decisions, but they can only achieve a limited degree of rationality (Simon 1955). In simple terms, people do not act randomly but rather deliberately attempt to make choices that align with their interests and preferences but, despite their best efforts and intentions, they often fall short of maximizing their utility, but managing only to improve it to some extent, at best.

'Bounded rationality' suggests that while there is room for decision-making, there are also inherent constraints on its actual execution (Simon 1955). Two primary constraints affect what would otherwise be rational behavior: the limitations of the human mind and the characteristics of the environment (Ghisellini and Chang 2018). Simon (1990, p. 7) articulated this by stating, "Human rational behavior is influenced by a scissors whose two blades are the structure of task environments and the computational capabilities of the actor." In doing so, he proposed a dualistic approach that recognized both the strengths and limitations of decision-making (Bendor et al. 2006). Consequently, 'bounded rationality' does not imply optimization under constraints but rather making adequate choices based on personal attributes (including preferences) and contextual factors, which continuously interact and evolve (Simon 1956; Ghisellini and Chang 2018).

Despite these inherent bounds, individuals cannot avoid making decisions, so they employ two suboptimal strategies. Both strategies involve subjecting their options to reason and striving to make rational choices while operating within unavoidable constraints. First, because reality is too complex to be fully grasped, humans simplify it and make choices within this simplified construct, approximating what would be optimal within this framework. Alternatively, they adopt a 'satisficing' approach.

2.2 Satisficing: satisfying and sufficing

Simon (1956) made another significant contribution with his theory of ‘satisficing’. Simply put, he argued that people employ a second suboptimal strategy in their decision-making. When confronted with a challenging situation, individuals preselect an outcome that would be satisfactory for them, rather than aiming for the optimal one, and then search among the available options for the one that most closely aligns with or exceeds that chosen outcome. This strategy can be seen as a natural extension of bounded rationality. Essentially, “[a] decision maker who adopts a satisficing strategy implicitly accepts cognitive and analytical limits and focuses attention on the most feasible alternatives” (Fry and Raadschelders 2013, as cited in Roberts and Wernstedt 2019, p. 10).

Therefore, in practice, individuals do not primarily pursue optimal results in their decision-making; instead, they settle for outcomes that are "good enough," a concept encapsulated by the term 'satisficing,' which combines "satisfying" and "sufficing" (Gigerenzer and Goldstein 1996). Importantly, satisficing inherently considers criteria beyond mere cost–benefit analysis, allowing for the inclusion of values, norms, and institutions, among other factors. This underscores the importance of factoring in individuals' perceptions and preferences in the decision-making process (Roberts and Wernstedt 2019).

With this framework, Simon established a middle ground between approaches based on utility theory's assumptions of broad individual freedom in decision-making and insights from behavioral social sciences, which suggest that choices are constrained by factors like culture, history, power relations, and cognitive limitations (Roberts and Wernstedt 2019).

His “satisficing” coinage encapsulates a strategy of searching through available alternatives until an acceptable alternative is found, a middle ground between maximization at one extreme and choice unguided by any decision process at the other (Roberts and Wernstedt 2019, p. 10).

Furthermore, this contribution delineates a middle ground between utility maximization on one side and randomness on the other. Therefore, when people make choices that lead to secondary outcomes, it does not necessarily mean they are detrimental to achieving their goals. Goal attainment is not an all-or-nothing condition; there can be varying degrees of success. If individuals can partially meet their aspirations, and this is sufficiently satisfactory (satisficing) without causing long-term harm, then their decision-making should not be deemed flawed. Only when biased or heuristic-based choices result in damaging or harmful outcomes regarding an individual's objectives should the decision-making process be considered malfunctioning or defective, warranting closer scrutiny.

These insights suggest that human decision-making often performs adequately, as assessed by the individual, though not necessarily all the time. This means that deviations from the model of full rationality can be sensible when considering the personal characteristics of the decision-maker and the circumstances in which choices are made. However, the dominant approach typically labels only optimal behavior as 'rational,' branding everything else as the opposite or 'irrational.' From its early stages, behavioral economics has indicated that some suboptimal choices can be reasonable, if not 'rational,' provided we consider internal and external constraints on choice.

Moreover, because 'satisficing' incorporates preferences and perceptions into the decision-making process, it sheds light on the inner workings that the dominant approach obscures. How exactly do individuals select a satisficing outcome that suits them in advance, especially when considering both personal and contextual factors? The answer lies in individual biases and heuristics, which play a central role in decision-making. The research program that exposed these shortcuts has been gradually developing. Perhaps its most telling illustration is the work carried out under the umbrella of prospect theory (see e.g. Kahneman and Tversky 1979), which provided an elegant framework for parsimoniously explaining many of the deviations from the model of full rationality (Thaler 2015).

2.3 Decision-making and systems

Despite Simon’s seminal insights, later contributions to BE failed to adequately incorporate his findings, particularly in terms of how the wider context (e.g. social and cultural factors) affect decision-making. Broadly, two main streams have emerged: the heuristics-and-biases framework (see e.g., Kahneman et al. 1982; Gilovich et al. 2002) and the fast-and-frugal heuristics approach (see e.g. Gigerenzer and Goldstein 1996; Gigerenzer 2001). The fast-and-frugal heuristics approach primarily focuses on how people make reasonably good choices by employing simple decision-making shortcuts (heuristics). On the other hand, the heuristics-and-biases framework emphasizes people's shortcomings and failures in judgment (Bendor et al. 2010). While both streams contribute to our understanding of human decision-making, the former highlights the effectiveness of heuristics, and the latter delves into the biases and errors individuals often exhibit. For the purposes of this discussion, the focus will be on the heuristics-and-biases framework, which has had a more significant impact in both academia and practice (the widely influential nudge program is an offspring of this framework).

In reality, human behavior deviates from the idealized depiction of homo economicus, a term often used to describe individuals in the normative representation of human behavior proposed by the dominant approach: RCT, which is a product of neoclassical economics and, as such, emphasizes objectivity, observables, and mathematical modeling (Corr and Plagnol 2019). This perspective portrays human behavior as driven by the pursuit of optimal outcomes and provides guidance on how people should achieve them. While this behavior is not assumed to be flawless, it is believed to be consistent enough to become a regularity when aggregated.

According to the dominant framework, deviations from this standard model are viewed as exceptions and are considered irrational. However, from a behavioral economics perspective, these deviations are systematic (Elahi 2014). Rather than labeling such behavior as 'irrational' or 'irrationalities,' which carries negative connotations,⁵ this conduct has been referred to as 'misbehaving' (Thaler 2015). Misbehaving occurs because humans have two distinct systems of thinking: an automatic system and a reflective one (Thaler and Sunstein 2009, 2021), both of which can lead to errors in judgment and decision-making. Nobel laureate Daniel Kahneman (2011) has extensively elaborated on these two systems:

• 1. The Automatic System or System 1 (S1): This system operates quickly and intuitively, making decisions based on heuristics, gut feelings, and automatic responses. It is prone to biases and can lead to errors.

• 2. The Reflective System or System 2 (S2): This system operates more deliberately and analytically. It involves careful reasoning, logic, and conscious deliberation. While it is less prone to biases, it can be resource-intensive and may lead to cognitive overload.

S1 operates as the default system and handles simple activities that require minimal awareness and cognitive effort. It excels at tasks involving quick, intuitive responses, associations, and managing information about individual items. However, S1 struggles with more complex tasks that involve handling multiple distinct items or employing statistical analyses. In essence, S1 deals with operations that demand lower levels of consciousness and effort. The simpler the task, the less conscious effort it requires, making it well-suited for S1.

As the automatic system, S1 generates impressions, feelings, intuitions, and intentions, which it communicates to S2. These signals are based on past experiences and often provide valid predictions in the short term, allowing for efficient problem-solving, especially with repetition (Kahneman 2011).

In contrast, S2 is responsible for reasoning and activities that demand increasing levels of awareness and cognitive energy. As tasks become more complex, S2's involvement in decision-making intensifies. When issues reach a sufficient level of complexity, S2 can take full control of decision-making. The level of cognitive effort required by S2 depends on the amount of mental energy available. More energy enables longer S2 activity or the handling of more complex tasks. As mental energy is depleted, S2 relinquishes control, and S1 resumes command.⁶

Unlike S1, which operates automatically, S2 is activated on demand. When S1 is active, S2 is on standby, conserving energy until needed. S2 functions only when deliberately triggered and as much as the task demands.

Nonetheless, S2 is not infallible. Despite its strengths compared to S1, S2 is also susceptible to errors and can lead to suboptimal outcomes. This becomes evident in situations where individuals struggle to process large amounts of seemingly straightforward information or fail to accurately estimate the probabilities of events and predict uncertain situations. In other words, while S2 offers advantages over S1 in certain contexts, it does not guarantee flawless decision-making⁷ (Kahneman 2011).

Importantly, while S1 often performs well, it can lead to detrimental results. S1 relies on shortcuts like biases and heuristics, which make decision-making efficient but not necessarily optimal and can result in systematic errors under certain circumstances (Author 2022a). The latter has been referred to as "choice architecture," emphasizing the importance of the decision-making context or setting (Thaler and Sunstein 2009; Sunstein 2020), albeit in a very limited way, unlike the account provided by Simon’s work.

2.4 Motivation within BE

Whether S1 or S2, BE’s literature has assumed that decision-making is oriented towards the decision-maker. That is, following RCT’s tradition, BE has assumed that individuals are driven (exclusively) by self-regarding motivations. In other words, by and large, BE maintains RCT’s assumption that people seek their self-interest. This most conspicuous in the nudge literature, which focuses on the chooser’s freedom and welfare (Thaler and Sunstein 2021, 2008; Thaler 2015).

Nevertheless, this seems to be at odds with actual human experience. Self-interest captures only one type of motivation and, at times, it may not even be the most relevant one (see e.g. Sen 1999, 2009; Le Grand 2006). Therefore, an account of how people actually behave, as BE seeks to provide, requires the inclusion of goals oriented towards the interests of others. Indeed, there is growing literature, mostly policy-oriented, seeking to use BE to address interests beyond the individual. Perhaps the most telling example is the case of environmental protection and climate change (see e.g. Beckenbach and Kahlenborn 2014; Carlsson et al. 2019; Sunstein and Reisch 2014). Since the relevant concerns can be with future human generations, or the regard of nature as intrinsically valuable, such efforts clearly move beyond self-regarding goals and require an account of motivations beyond mere self-interest (Author 2023).

Hence, BE primarily treats human beings as self-interested, plural, impulsive actors, as well as reflective choosers. Individuals are seen as mainly self-interested, although other-regarding motivations are increasingly acknowledged (Author 2022a). They are considered plural due to the various factors that characterize them and their environments. This is particularly so if a move is made beyond the immediate setting of decision-making is made or choice architecture, and a richer notion of context is adopted by returning to Simon’s seminal insights in the model (see Fig. 1). Moreover, they exhibit both impulsive and reflective decision-making tendencies, acting on impulse as well as through reasoned choices.

Fig. 1

Source: based on Thaler and Sunstein (2008, 2021); Kahneman (2011) and Simon (1955). Elaboration: author

Behavioral economics’ decision-making model.

Although this would suggest the admission of a diversity of motivations, BE’s theoretical literature is yet to include an explanation of non-self-interested preferences and choice. Nonetheless, this seems to be at odds with the growing practical literature striving to tackle interests beyond the individual’s own wellbeing, such as climate change. These efforts offer increasingly compelling results and clearly move beyond self-regarding goals. Therefore, an account of motivations beyond mere self-interest is warranted.

3 The capability approach: accounting for the totality of motivations

Elaborated at the intersection of economics with philosophy, the Capability Approach (CA) offers a descriptive and normative framework for understanding ‘development’, suggesting where assessments of social states should take place.⁸ The CA places humans and their quality of life at the center of attention, considering them the ultimate goal of development. Consequently, development should be assessed based on people's quality of life (Sen 1999).

By prioritizing people and their quality of life, the CA also provides an account of human beings and their agency. This account is reflected in the CA's key concepts, including functionings and capabilities as the evaluative space (the framework’s normative part), conversion factors as mediating aspects that influence the transformation of resources into capabilities and functionings, and wellbeing and agency as motivating factors. For the purposes of this argument the focus is on the descriptive part of the approach.

3.1 Capability and functionings: the evaluation space

How can we recognize development through this perspective? If humans are the focal point of development, the CA must outline the dimensions of their quality of life that should be prioritized for improvement. The CA introduces two key aspects: functionings and capability. Functionings represent the actual 'doings' and 'beings' that individuals value and have reason to value (Sen 1999). These achievements are referred to as functionings because they encompass the activities and states that enable people to function effectively. Functionings are integral to a person's identity and wellbeing, reflecting their chosen ways of living (Sen 1990), and they constitute the essence of human existence (Sen 1993). Hence, living can be regarded as intertwined doings and beings (Sen 1992).

The valued functionings may vary from elementary ones, such as being adequately nourished and being free from avoidable disease, to very complex activities or personal states, such as being able to take part in the life of the community and having self respect (Sen 1999, p. 75).

The CA exhibits two significant features through the concept of functionings: plurality and multidimensionality. Plurality arises because functionings are personal and reflect individuals' values and perceptions of the good life. As they are integral to a person's identity, they hold significance in discussions regarding conversion factors, an issue addressed below. This stands in stark contrast to economistic approaches, which predominantly rely on monetary indicators, focusing on resources or bundles of goods rather than on the individual (Sen 1988). Additionally, functionings are multidimensional, encompassing a wide range of valued activities and states that individuals may consider important. This recognition acknowledges the incommensurability of human experiences, rejecting the assumption prevalent in conventional economistic approaches that a single indicator can encompass all aspects of human welfare (Sen 1999).

Capability, on the other hand, pertains to the range of possible functionings that are valuable upon reflection and accessible to an individual. It represents a person's capacity to engage in activities and possess states that they deem valuable after careful consideration (Sen 1999). Capability comprises all the potential functionings available to a person, as long as they hold value upon reflection. Therefore, capability signifies the freedom individuals possess to lead various types of lives that they consider valuable after careful deliberation⁹ (Sen 1993). A valuable life encompasses both achievements and the freedom to choose and achieve them (Crocker and Robeyns 2010).

3.2 Conversion factors: mediating elements

The translation from resources to capability and functionings is not straightforward or automatic. This process is influenced by what the literature terms as conversion factors. Conversion factors encompass both personal attributes of the individual and contextual elements that can impact the translation of inputs into outcomes (refer to Fig. 2). These factors are categorized as internal and external conversion factors. Sen (1999) identifies the following conversion factors: (1) Personal Heterogeneities; (2) Environmental Diversities; (3) Variations in Social Climate; (4) Differences in Relational Perspectives; (5) Distribution within the Family.¹⁰

Fig. 2

The CA’s original model moving from means to ends. Elaboration: author

The introduction of conversion factors recognizes the diversity among individuals and their contexts, as well as the intricate relationship between them.¹¹ Internal or personal conversion factors include characteristics like gender, class status, age, ethnicity, and more. External or social and environmental conversion factors encompass factors such as a location's climate and topography, a family's religious beliefs, societal patriarchal norms, and so on. This underscores the idea that different individuals in different contexts require varying types and amounts of resources to achieve similar levels of achievements and freedom.

3.3 Wellbeing and agency: the motivations

In the context of human agency, which requires an assessment of action relative to an individual's goals (Alkire 2009; Hempel 1962), it becomes essential to examine people's motivations. The CA introduces two key concepts in this regard: wellbeing and agency. Wellbeing refers to a person's quality of being (Sen 1999). It essentially captures an individual's personal state of "wellness" (Sen 1992) or welfare (Sen 1993). This category is limited to the agent's own states, aligning with the conventional model of rationality's assumption of self-interest as the sole motivation.

On the other hand, agency is a broader concept that encompasses the totality of an individual's goals. It is an evaluation of "what a person can do in line with his or her conception of the good" (Sen 1985, p. 206). Agency, thus, includes self-regarding objectives, as captured in wellbeing, as well as other-regarding aims (Crocker and Robeyns 2010). While self-regarding motivations revolve around personal interests, such as going on holiday, seeking a promotion, or saving oneself after an accident, other-regarding motivations encompass activities like donating blood, money, or organs, engaging in various forms of volunteer work, or participating in philanthropic endeavors. Consequently, the concept of agency challenges conventional approaches that assume self-love and self-interest as the sole motivation, as it acknowledges that an individual's own wellbeing is significant but not the exclusive motivation.

The relationship between these two concepts is crucial. Although distinct, wellbeing and agency are interdependent (Sen 1992). Firstly, wellbeing, or the pursuit of self-interest, can be one of the objectives valued by individuals. Secondly, achieving other-regarding goals can contribute to an individual's wellbeing, while not achieving them can result in unhappiness due to the frustration of failing to attain a valued objective. However, there can be tensions between wellbeing and agency, where seeking other-regarding goals may conflict with self-regarding ones. For instance, mothers who sacrifice their own nourishment to ensure their children have enough to eat are examples of such tensions. In such cases, an individual may enhance their agency by achieving someone else's valued objective at the expense of their own wellbeing.

Exploring the core aspects of the Capability Approach has revealed its emphasis on plurality. The framework explicitly identifies two dimensions of plurality: inter-individual and inter-end variation (Sen 1992). Firstly, there is significant diversity in the ends, functionings, or capabilities that individuals can legitimately pursue. In the exercise of their freedom, people value and have reason to value a wide range of types of life. Secondly, even when individuals have similar ends in mind, they may require different types of resources and inputs to achieve comparable levels of functionings or capabilities in diverse contexts. The attention given to both internal and external conversion factors allows for the recognition of this diversity, which challenges conventional assumptions that people are essentially similar, possessing identical maximal potentials, or adhering to the idea that "all people are created equal."

4 Self-interest and others’ interests: distinguishing motivations

The advantages of the CA’s expansion of motivation notwithstanding, it seems useful to make a further and formal differentiation between self-regarding and other-regarding goals. Although this seems to be well within the framework’s purview, it has not been explicitly stated. The approach has kept the notions of wellbeing and agency as the sole categories capturing motivation, excluding other-regarding goals as a type of motivation warranting explicit attention. Additionally, since other-regarding motivations may be oriented towards non-human others, it this issue warrants some elaboration as well. Finally, a brief discussion about the causes of altruistic behavior is also addressed in terms of emotions.

4.1 Selfishness and selflessness

Self-regarding motivations capture exclusively those that RCT focuses on, i.e. self-interest. They denote the goals that individuals have that improve, advance or favorably affect their own welfare (as judged by themselves). An important, often disregarded, aspect of these motivations is that they can be detrimental to the welfare of others.¹² As such, although they have been covered by the term wellbeing within the CA, they can also be referred to as selfish or egoistic motivations.

Other-regarding motivations go beyond the previous ones, and fall outside RCT’s scope, and capture others’ interests. These are goals that individuals have that ameliorate, further or positively affect the welfare of others (as judged by the latter). As Sen (1993) asserts, pursuing them could even prove to be at odds with self-regarding motivations and be detrimental to the chooser’s welfare.¹³ Therefore, they can be referred to as selfless or altruistic motivations and, for lack of a better term, perhaps they could be labeled ‘other’s being’ within the CA. Agency as the totality of motivations, thus, encompasses both egoism or selfishness, what the CA calls wellbeing, as well as altruism or selflessness, which lacks a specific term in the framework.

This highlights the connection between motivation and the evaluative space. Not only can functionings and capability be expressed in terms of wellbeing and agency (see Crocker and Robeyns 2010), they can usefully be expressed in terms of self-regarding and other-regarding goals (see Table 1). In the case of self-regarding functionings, it is evident that an individual's welfare is directly linked to their achieved doings and beings, as these constitute an essential part of a person's being (Sen 1992). Concerning self-regarding capability, it can be understood as a person's freedom to experience wellbeing since capability encompasses the set of possible valued functionings, and functionings are constitutive of a person's being (Sen 1992). Another way to illustrate this relationship is that self-regarding achievements depend on the capability to function, as capability can directly contribute to wellbeing (Sen 1992).

Table 1 Combination the CA’s evaluative space with motivations

Similarly, the relationship between other’s interests (as opposed to self-interest) and functionings and capability is akin to the nexus mentioned above. In terms of functionings, other-regarding achievements also constitute an individual's being, as all of a person's objectives reflect their identity (Sen 2007). Other-regarding objectives, like self-regarding ones discussed earlier, form an integral part of a person's being. Other-regarding achievement revolves around realizing an outcome in line with one's conception of the good and, as such, it is part of their agency. As Sen (1992, p. 56) has asserted:

A person's agency achievement refers to the realization of goals and values she has reasons to pursue, whether or not they are connected with her own well-being. A person as an agent need not be guided only by her own well-being, and agency achievement refers to the person's success in the pursuit of the totality of her considered goals and objectives.

The relationship between other’s interests and capability can be described in the same terms as that between self-interest and freedom. Other-regarding freedom pertains to the actual opportunities available to individuals to attain valued achievements, reflecting the extent to which they can exercise their agency (Sen 1992).

4.2 Human and non-human others

So far so capabilitarian, the argument has adhered to Sen’s work in placing people and their quality of life at the center of the argument. The implications for altruism, as can be attested, are that such a motivation is directed towards the freedom and achievements of other human beings. That is, people are the sole normative unit of interest. In Sen’s wider work (see e.g. Sen 2009) he has stressed this even when rackling increasingly important issues such as the climate. Indeed, he has made the case for environmental protection in terms of securing the possibility for future generations of people to have lives that the find valuable after reflection available (i.e. capability or freedom) and eventually achieve one (i.e. functionings or achievement). That is, non-human beings seem to have instrumental value only.

There are, however, multiple contributions from different perspectives seeking to expand the CA with the inclusion of non-human beings as entities with a normative standing (see e.g. Nussbaum 2023, 2018; Wichert and Nussbaum 2021; Bendik-Keymer 2021; Holland 2021; Linch and Holland 2017). Needless to say, Sen has not endorsed this expansion of his framework. Perhaps the most evident difficulty is the fact that the CA has been called by Sen (1999) freedom-centered and agency-oriented and it is rather difficult to make those categories apply to non-human beings.

Nonetheless, such endeavor seems warranted given that, in actual experience, people can and do attach intrinsic value to non-human beings. While the jury is still out regarding the normative standing of the latter within Sen’s version of the approach, for the purposes of this paper suffice it to stress that the CA has two sides, one descriptive and one normative. Since this argument employs only the account of human beings that the framework advances, it focuses on the descriptive part, which includes the elaboration on the plurality of motivations guiding human behavior.

4.3 Emotions, motivations, and actions

Why do people act altruistically? Even though a thorough examination of the causes of selfless behavior is beyond the scope of this argument, this issue warrants some attention since it points to a further possibility of enrichment. At some distance from economics, including even BE, the question posed finds a plausible answer in the field of psychology, with some contributions from philosophy. There is a significant body of literature suggesting a persuasive explanation for selfless action based on the role that emotions play in human behavior (see e.g. Gazzaniga 2018; Goldie 2009; Sousa 1990). As Prinz (2009, p. 520) puts it “Egoists credit self-love and altruists credit love of others”. For example, a mother giving her food to her child so they can be better nourished, a person walking their dog even when the weather makes them uncomfortable, or an individual donating money to another homeless one are all instances of altruistic actions and they can be plausibly explained by the role that emotions play in decision-making. Broadly,

Emotions are episodes that have a felt character and are directed at particular objects provided by their cognitive bases. They appear moreover to be intimately related to evaluations of these objects and are subject to assessment by means of a variety of distinct standards, most prominently among them standards of correctness and justification (Deonna and Teroni 2012, p. 11).

Relevant for this discussion, in addition to evaluation, emotions can also relate to motivation. Prinz (2009) clarifies the difference and distinguishes between three types of other-directed emotions. First, emotions can have epistemic and motivational roles jointly, as when one performs a ‘good’ action because they know it is regarded socially as good, in which case they are morally relevant, but they can also have a motivational role absent an epistemic one, as when one does a good deed without doing it because it is good, which are referred to as altruistic or fellow-feelings leading to prosocial behavior. Second, within fellow-feelings, three emotions loom large: empathy, sympathy, and concern. Empathy is a response that denotes one person experiencing the emotion that other person experiences and is a vicarious response, so it is associated with the situation of the counterpart. Sympathy is negative response, is linked to the emotion of the counterpart rather than the situation, and the emotions between the sympathizer and sympathized need not be the same. Concern is a negative emotion induced by the perception that a counterpart is in a bad situation, whether the latter experiences a negative emotion or not. Out of the three, concern is the most likely to prompt action that is altruistic and empathy the least likely to motivate action and if it does, it may have vestiges of self-interest.

This brief discussion tackles other-regarding motivations when those others are human beings, begging the question of non-human beings. There seems to be an interesting debate on this issue. For example, regarding empathy, whereas some belief it to be plausible, stating that empathy is “the capacity to experience the feelings of another person or an animal, cognitively or emotionally (O’Connell in Phillips 2009, p. 47), others are skeptical about this very possibility (see e.g. Kasperbauer 2015). Nonetheless, for current purposes, suffice it to say that given the prevalence of altruistic behavior towards non-human beings, it seems likely that in some form of at least one of the aforementioned responses or emotion provides an explanation for this behavior.

Even though emotions can help elucidate much of human action, they have not received adequate treatment in either the CA or BE. Within the CA, at most, when tackling some of the differences in human action, in this early work, Sen (1977) makes a distinction between sympathy and commitment. While sympathy refers to a state in which someone’s wellbeing is dependent on that of someone else’s (as when a person feels discomfort due to another’s sickness symptoms), commitment indicates choices that expectedly lead to inferior welfare outcomes compared to available options. Indeed, even if these choices led to superior welfare results, they would still fall under the label of commitment as long as that outcome was not the reason for the action but because it was ‘the right thing to do’ (Sen 2002). Therefore, Sen’s (1977, p. 329) challenge is clear when stating that commitment “[…] drives a wedge between personal choice and personal welfare, and much of traditional economic theory relies on the identity of the two”. Since sympathy seems to fall within self-regarding motivations, commitment seems to have become other-regarding motivations in Sen’s later work. What little is said about emotions, however, can be summarized in his assertation that: “even the importance of emotions can be appreciated within the reach of reason”¹⁴ (Sen 2009; p. 39). That is, for Sen, emotions are relevant in guiding human agency but also subjectable to reason.

BE has also failed to incorporate emotions in its framework. Although emotions have been mentioned within the framework, the treatment given to it has been rather partial. For example, Kahneman (2003) asserts that “Utility cannot be divorced from emotion, and emotions are triggered by changes. A theory of choice that completely ignores feelings such as the pain of losses and the regret of mistakes is not only descriptively unrealistic, but also leads to prescriptions that do not maximize the utility of outcomes as they are actually experienced”. In this sense, by and large, emotions have treated at best as a heuristic (see Slovic et al. 2002) and at worst as a type of bias, being associated to the endowment effect (Ghisellini and Chang 2018), hyperbolic discounting, sunk cost fallacy, ambiguity version, regret and affect¹⁵ (see Mauldin 2005). That is, for the most part, emotions have been considered as an independent variable but its meaning has been reduced to a concept related to the idea of ‘mood’ or ‘gut feelings’ to explain inconsistency of action (Kahneman 2003) or preferences and violations of rationality (Ghisellini and Chang 2018).

Despite offering a compelling explanation of selfless behavior, emotions have not received adequate treatment within the CA or BE. Although providing an explanation for either egoistic or altruistic actions falls outside of the purview of this paper, such an undertaking is a meaningful further addition that would complement the CA and BE independently as well as jointly.

5 Capabilitarian behavioral economics

Against this backdrop, it may be intuitive to anticipate how the CA can complement BE’s insights in terms of motivations. By going beyond self-interest and including other-regarding motivations, the CA goes beyond RCT, which is the basis for BE. Therefore, BE can be enriched with the inclusion of a most relevant type of motivation. By so doing, it enhances its theoretical foundations with a sound account of human agency with no apparent cost elsewhere to its framework. Moreover, this capabilitarian BE is better equipped to further BE’s stated project of accounting for how people actually behave. Finally, because of these literatures have tackled the issues of freedom and wellbeing, this combination entails some important implications for this topic as well.

5.1 Plurality in motivations

Table 2 shows how this enrichment can be expressed by combining the CA’s description of motivation with BE’s explanation of decision-making. What may be broadly considered egoistic (i.e. self-regarding) and altruistic (i.e. other-regarding) motivations are combined with the two systems of decision-making (i.e. S1, in charge of non-reasoned choice, and S2, in charge of reasoned choice) mentioned above.

Table 2 Combination of the CA’s motivations with BE’s decision-making systems

As can be attested, the combination proves insightful by showing that decisions made by both systems can pertain both types of motivations, not only self-interested ones as the conventional model of rationality may suggest. Thus, it is exposed that egoistic choices can be reasoned as well as non-reasoned, and so can altruistic ones.

Human experience corroborates this insight. First, regarding BE’s standard model, there is reasoned and non-reasoned egoistic choices. On the one hand, people aim at enhancing their own wellbeing in deliberate ways. Most evidently, this can be seen in the pursuit of immediate gratification, pleasure or utility. Eating the food we like, reading literature we enjoy, traveling on holidays, playing (video) games, spending time with friends and family, or even some activities with a trade-off in terms of (long-term) wellbeing such as smoking or compulsive buying, provide telling illustrations. At the same time, this can also be attested in choices striving for long-term wellbeing (even at the cost of some immediate pleasure). From savings and pension schemes to following diets and workout regimes, individuals willingly seek to improve their own welfare in the long run.

On the other hand, individuals also act according to their own wellbeing in ways that show diminishing degrees of awareness. This are activities that become habitual and little to no mind is payed to them when performed. The impulse of protecting one-self (usually one’s face) in the case of an accident, running out of a building when its emergency siren goes off (even during a drill), or waiting for traffic lights before crossing a street, particularly for pedestrians, are but a few examples. The interest with the short- and the long-term applies to this case as well. To be sure, in all cases, particularly when it comes to long-term wellbeing, this does not mean that people succeed in bettering their situation, because often times we fail, but the motivation to pursue one’s own wellbeing is there.

Second, and this is the insight of enriching BE with the CA, there is also reasoned and non-reasoned altruistic choices. Epitomizing reasoned altruism is people deliberately intent on improving the wellbeing of others, whether due to empathy, sympathy or concern. As in the case above, these choices can be directed towards the short-, medium- or long-term. The enhancement of the immediate welfare of others can be shown by people treating the wounds of others when they do not have to or the personal handouts given to random individuals upon request (e.g. beggars). The improvement of the medium-term utility of others is embodied by the different forms of institutionalized charity and philanthropy. The amelioration of long-term wellbeing can be exemplified by blood and organ donation. Moreover, individuals may also seek to ameliorate the welfare of non-human others. Volunteering in reforestation or river-cleaning initiatives in places far from their residence, caring for injured animals when one does not have to, and participating in demonstrations demanding for environmental protection policies with no direct or immediate impact to those involved illustrate such interests. As can be gathered from the above, this concern with non-human others can also be expressed throughout the time spectrum, from the short- to the long-term.

Similarly, people often care for the wellbeing of others in ways that have decreasing levels of consciousness. Illustrations of non-reasoned altruism are: when someone protects their significant other from harm in an accident as if by instinct or when people leave their spare change as a donation/tip after a purchase, by force of habit. The betterment of non-human others’ wellbeing can also be a goal of non-reasoned altruism. The avoidance of plastic products such as straws, grocery bags, etc. when their use does not entail an additional cost or burden for consumers provides an illustration. Certainly, whatever the level of awareness, and whether oriented towards human and non-human others, this motivation can also be manifested form the short- to the long-term. This is reinforced by the insights from the study of emotions since, whether causing selfish or selfless actions, they can be conscious or unconscious. From deeply unconscious motivations many of which are of an emotional nature, to emotions people only recognize once they are pointed out to them, to the existence of emotions not acknowledged by the individual, there are different degrees of awareness (Deonna and Teroni 2012). Also, in all instances, there may be underlying emotions such as concern or responses to emotions such as empathy and sympathy. Again, in all these cases, decision-making is prone to failure and these other-regarding objectives, regardless of the level of awareness, may not be reached as intended.

5.2 Reasoning and motivations

There are also implications for the dynamics between systems 1 and 2. To recall, while S1 is automatic, S2 is deliberate and responsible for self-control, which means that it enables self-critique. Therefore, beyond rationality, BE seems to be describing reasoning. The division of labor between S1 and S2 allows to account for reasoning processes and the provision of reasons or justifications for preferences and choices. Of course, this means that both egoistic as well as altruistic motivations can be subject to criticism and justification, just like their underlying emotions can. Furthermore, this coincides with and can undergird the CA’s focus on freedom and achievements that people not only value but have reason to value (Sen 1999). The description of decision-making via these two systems explains the innerworkings of reasoning, which is what the CA proposes (Author 2020b). Moreover, because self-control is not produced in a vacuum but is influenced, to different extents, by social control and expectations, S2 also captures the latter (Author 2022b). Accordingly, S2 can account for these external influences that give shape to reasons and explanations, giving them legitimacy and validation, and distinguishing them from unwarranted defenses and excuses. This approximates what the wider CA literature refers to as reasonableness (see Sen 2009).

5.3 Adding to BE’s model

Expectedly, this combination has implications for the wider model of decision-making proposed by BE, which includes also the context in which choices are made. Synthesizing the discussion above, Fig. 3 depicts the elements involved in this process as well as the interactions between them.

Fig. 3

Combination of the CA’s proposal with BE’s decision-making insights. Elaboration: author

In synthesis, BE’s model of decision-making enriched by the CA (which may be called a capabilitarian behavioral model) recognizes that humans deviate from the model of full rationality and systematically fall short not only of achieving optimal outcomes but of their sufficiently satisfying ones as well. While many suboptimal outcomes may be acceptable most of the time, there are cases where they can be detrimental to people's wellbeing. These decision-making failures take place because rationality is limited by both cognitive limitations and the environmental context, which includes history and culture (Simon 1955). These aspects influence the systems intervening in decision-making: S1 and S2.

Hence, a capabilitarian BE treats human beings as multi-motivated, plural, impulsive actors, as well as reflective choosers. Individuals are regarded as multi-motivated because they are driven by egoistic or self-regarding goals as well as altruistic or other-regarding motivations, which can be oriented toward human and non-human beings. They are considered plural because of the various factors that constitute them (internal or personal) as well as their environments (contextual or external). Finally, they have both impulsive and reflective decision-making tendencies, since they act on impulse as well as through reasoned choices. What is more, with the addition of emotions undergirding self- and other-regarding motivations, this agent becomes also an affective one, moved by feelings as well, which are more than biases, can be subjected to reasoned scrutiny, and can be conscious or unconscious. All of these features resonate and complement both the CA and BE. This description of human decision-making seems to account for how human beings actually behave and, therefore, it brings BE that much closer to achieving the project it has set out for itself.¹⁶

6 Conclusion

BE seeks to provide an account of how people actually behave. To do so, it has used the standard model of RCT as a yardstick and studies how individuals systematically deviate from it. Selfless motivations and altruistic behavior can be considered an additional meaningful deviation from homo economicus. BE already challenges the assumptions of (1) unbounded rationality, and (2) unbounded willpower. Both are increasingly covered by the biases and heuristics literature, for example. In light of the increasing evidence, showing selfless behavior, it is only logical that the premise of (3) unbounded selfishness should be scrutinized and the deviations from it be accounted for. That is what this paper has sought to do. Additionally, since the current literature seems to disregard Simon’s seminal contributions in terms of the contextual influences over decision-making, a return to those insights has been advanced.

Amartya Sen’s Capability Approach has been proposed to complement BE. The CA offers an account of human beings underscoring the totality of people’s goals by employing two relevant categories: wellbeing and agency. While wellbeing refers to self-regarding motivations, agency refers to the totality of goals, including other-regarding goals as well. Therefore, for the purposes of the argument, a further differentiation has been made, that between egoism, self-interest, and selfishness on the one hand, and altruism, others’ interests, and selflessness on the other hand. Finally, a plausible explanation, particularly for altruistic behavior, has been made with the inclusion of emotions, which together with the issue of ethics offer a promising avenue of future research to enrich the CA and BE, individually or jointly.

Using the CA’s contributions in terms of motivations to fill the void in BE’s approach to decision-making and human agency shows that a plurality of scenarios can be produced. It builds on the two systems proposed by BE and adds the two motivations stressed by the CA. This moves beyond identifying choice as the use of either S1 or S2 oriented to improve mainly, let alone solely, personal self-interest or one’s own wellbeing. The CA’s addition means a recognition that S1 and S2 are also engaged when acting to improve the welfare of others. That is, besides non-reasoned and reasoned egoism, this combination adds non-reasoned and reasoned altruism. These insights add to those made in this forum in terms of how the BE can enrich the CA. Because of the emphasis these bodies of literature have placed on freedom and welfare, exploring the implications of a capabilitarian BE in those terms can be an additional fruitful endeavor.