A. Freedom-centered: Functionings and Capability

Functionings and capability denote the intrinsically valuable aspects of life. As such, they constitute the evaluative space where the assessment of social states ought to take place. Functionings highlight people’s actual functions, what they do and are. Thus, they are the beings and doings that people value and have reason to value (^{Sen, 1999}). They are achieved states or duly considered valuable types of life (^{Sen, 1993}). That being so, a person’s achievements can be seen as the vector of their functionings (Sen, 1992). Consequently, they are constitutive of a person’s being (^{Sen, 1990}) and, what is more, life can be regarded as the combination of several interrelated beings and doings (^{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 emphasis on the plural signals the CA’s move from opulence and single indicators to the recognition of the incommensurability of the intrinsically valuable things in life. Therefore, the approach advances multidimensionality. The same applies to the opposite of functionings, i.e., deprivations or disadvantages.

Capability, in turn, encompasses all potential functionings from which a person can choose (^{Sen, 1999}). Its name comes from their denoting people’s capability to function (^{Sen, 1992}). In this sense, it captures the notion of freedom, entailed by the possibility people have to choose from different lifestyles regarded as valuable after reflection (^{Sen, 1993}). Hence, both a person’s doings and beings and the freedom to choose them constitute a valuable life (^{Crocker & Robeyns, 2010})^{. Sen (1999)} suggests that development ought to be assessed in the space of capabilities and has redefined development as freedom.

The above notwithstanding, both functionings and capability are important for the approach. The focus on capability means that besides outcomes, processes also matter (Sen, 1997). Given the assessment of a social state, it makes a difference if it is the product of imposition or of choice. Moreover, the actual exercise of freedom is important and, therefore, attention to actual outcomes and functionings is necessary. That is, even when the possibility of choice is available, it is indispensable to know how it is exercised. To address this, Sen (1997) introduces the differentiation between ‘culmination outcomes’ and ‘comprehensive outcomes’. While the former focus only on the result, the latter also encompass how the result is reached. The CA regards the improvement in people’s lives as an expansion of their freedom. Thus, development is about expanding people’s choices in all dimensions of life (^{Haq, 1995}).

The stress on a person’s various possible valued, reflected-upon types of life underscores the CA’s focus on human plurality. There is a myriad of legitimate lives that a person can lead and, in fact, the opportunity to choose is relevant to well-being (^{Sen, 1999}). This has been referred to as inter-end variation (^{Sen, 1992}). Being a parent or not, observing a religion or not, pursuing a career or not, practicing a sport or playing an instrument or not are a few examples of legitimate considered ends. As in the case of functionings, there are multiple dimensions that can be so valued. And, just as the presence of freedoms is relevant, so is their absence, referred to as capability deprivations or unfreedoms.

In sum, while functionings are achievements, capability is the freedom to achieve. “Freedom can be distinguished both from the means that sustain it and from the achievements that it sustains” (^{Sen, 1992, p. 86, emphasis in the original}). At the same time, ^{Sen (1999)} points out that different capabilities can be related to each other. Expanding some freedoms may lead to the expansion of others. Thus, freedom plays a constitutive and instrumental role in development (^{Sen, 1999}).

More recently, the capability framework has been enriched with additional insights. ^{Wolff and de-Shalit (2013)} have explored an interdependence among functionings and among disadvantages. There are important achievements that can lead to the improvement of others. These are referred to as ‘fertile functionings’. Having disposable income is a telling example as it can lead to other achievements. Less evident examples are also possible. Humor, for example, can be a functioning that improves a person’s social abilities from school to the office environment, augmenting their likelihood of success in each, and leading to a better quality of life, in societies that value humor (^{Wolff & de-Shalit, 2013}). By the same token, there are deprivations that can aggravate existing ones. These are known as ‘corrosive disadvantages’. The lack of disposable income can be such a deprivation because its absence limits the consumption of relevant goods and services for a person’s quality of life. Another, less conspicuous, illustration is parents’ education since less educated parents tend to have a poorer vocabulary and talk less to their kids, thereby reducing the latter’s school performance, possibilities for higher education, and job opportunities (^{Wolff & de-Shalit, 2013}).

Finally, freedom is conceived broadly within the approach. Sen (2009; ¹⁹⁸⁸) points to the distinction between opportunity freedoms and process freedoms. While the former can be associated with the notion of positive freedoms, as in ‘being free to’, the latter can be linked to negative freedoms as in ‘being free from’, in the libertarian tradition. Both are relevant to development (^{Sen, 1999}). Sen (2009) stresses that the capability perspective regards substantive, effective, or opportunity freedom widely, of course focusing on choice (opportunity) but also including process. Nonetheless, he also acknowledges that the CA accounts for the former much better than it does for the latter (^{Sen, 2004}).

B. Conversion factors: personal and contextual

What is more, the CA recognizes that the translation from resources to capability or functionings is not direct but mediated. ^{Sen (1999)} identifies the following conversion factors: i) personal heterogeneities; ii) environmental diversities;iii) variations in social climate; iv) differences in relational perspectives; and, v) distribution within the family. These can be summarized as characteristics of the individual as well as features of the context, whether environmental or social, that can affect that conversion (see Figure 1). While personal characteristics can be exemplified by age, gender, ethnicity and civil status, social ones may be illustrated by religious practices, societal norms and family customs, and environmental ones by rurality, region (coast/mountain/jungle), soil fertility, availability of basic public services, and probability of droughts. These are also referred to as internal and external conversion factors, respectively.

If the attention to multidimensionality and diversity entailed by the CA’s recognition that there are various legitimate types of life is a first move towards human plurality, the addition of conversion factors, consolidates this position. The latter has been referred to as inter-individual variation (^{Sen, 1992}). This addition is significant given that different people in different contexts may require different quantities and qualities of resources to reach certain levels of achievement. For instance, a pregnant woman, someone with a disability or terminal illness, or someone whose labor is physically taxing require more resources (for nourishment, medicine, etc.) than people lacking these personal characteristics. The challenge is compounded by the addition of different contexts, as resource requirements can vary for those individuals depending on whether they live in rural or urban areas and whether they have immediate access to public services (health).

Inter-end and inter-individual variations are closely related. A person’s actual freedom to lead their reflected-upon, valued types of life depends on two factors, namely their goals, and their power to convert resources into the meeting of those goals (^{Sen, 1992}).

As illustrated by the above, the focus on Sen’s discussion has been on conversion factors as constraints for the achievement of relevant outcomes. This has been expanded in recent contributions to include their conception as enablers as well (^{Hvinden & Halvorsen, 2017}). Conversion factors, therefore, serve the purpose of establishing the necessary nexus between agents and structures in order to explain the dynamics between them. What is more, conversion factors also interrelate, particularly individual features with social ones, which can help explain unexpected outcomes (^{Hvinden & Halvorsen, 2017}).

Source: ^{Garcés-Velástegui (forthcoming)}

Figure 1 Illustration of the CA’s movement from means to ends incorporating latest contributions 

C. Agency-oriented: Well-being, and Agency

Concerning motivation, the CA introduces two categories: well-being and agency. Well-being refers solely to personal welfare (^{Sen, 1993}). It captures the inducement related to self-regarding goals and individual wellness (^{Sen, 1992}). Being healthy, literate, or able to elect and be elected, or earning the same salary for the same job regardless of gender, ethnicity or another characteristic are some examples.

Agency is a broader concept and denotes the totality of a person’s motivations and goals. It conveys the idea of “what a person can do in line with his or her conception of the good” (^{Sen, 1985, p. 206}). That is, it encompasses both self-regarding and other-regarding objectives (^{Crocker & Robeyns, 2010}). Some illustrations are: participating in philanthropy and volunteering, being able to donate blood/organs, or to demonstrate against climate change.

Moreover, ^{Sen (1999, p. 19)} regards agency “[…] in its older – and ‘grander’ – sense as someone who acts and brings about change, and whose achievements can be judged in terms of her own values and objectives, whether or not we assess them in terms of some external criteria as well.” In this sense, agency is valuable in and of itself. Associating it to the capability of political participation, it has been argued that the value of agency is threefold: i) intrinsically, instrumentally, and constructively (^{Alkire, 2009}). Intrinsically, by denoting people’s ability to control their environment to pursue their considered goals, agency enables people to be in control of their lives. Instrumentally, agency can be valued insofar as it enables people to enjoy capabilities and functionings. As in the case of capabilities and functionings, free agency “[…] contributes to the strengthening of free agencies of other kinds” (^{Sen, 1999, p. 4}). Constructively, agency entails judgment, not only choice and, as such, it permits the assessment of preferences and values (^{Crocker & Robeyns, 2010}). This is particularly important for the inevitable selection of relevant capabilities and their ranking.

What is more, the evaluative space and motivations are deeply interrelated. This means that capabilities and functionings can be expressed in both well-being and agency, beyond well-being (see Figure 1). Regarding well-being, an individual can achieve functionings or types of life that are exclusively and solely related to their wellness. At the same time, the vector of those potential functionings, their capability can be related solely to them as well. Therefore, both well-being functionings and capabilities are possible. Apropos of agency beyond well-being, i.e., other-regarding goals, an individual can achieve doings and beings that do not advance their own welfare. ^{Sen (1992, p. 56)} has posited that: “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.” Further, the vector of all potential such functionings, would depict their capability. Thus, there can be agency functionings and capability as well.

The differentiation between agency and well-being is momentous. In fact, although they are likely to move in similar directions, this distinction is useful to account for the tension that may ensue between goals related to each. The pursuit of other-regarding objectives, say an organ donation, may lead to an increase in agency achievement but also a decrease in well-being achievement. Additionally, the distinction can highlight the interdependence between the categories (^{Sen, 1992}). First, well-being is a motivation that a person can have as an agent. Second, achieving other-regarding goals can certainly contribute to an agent’s well-being and, likewise, failure to achieve them can prove detrimental to it.

Furthermore, it has recently been pointed out that the CA’s very notion of agency challenges the dominant approach to human conduct, rational choice theory, turning rationality into reasoning. (^{Garcés, 2020b}). ^{Sen (1977)} has stated that the axiomatic rational model renders people ‘rational fools’. This is related to the second part of the approach’s concept of agency, dealing with the agent’s ability to judge their values and objectives, a relatively much less explored aspect of the notion. This points to the relevance that rationality has for the capability perspective. In this sense, ^{Sen (2002, p. 4}) has opposed the convention and proposed an idea of rationality that is much closer to human experience and the CA:

The broad reach [of reason] entails the rejection of some widely used but narrowly formulaic views of rationality: for example, that rationality must require following a set of a priori “conditions of internal consistency of choice” or “axioms of expected utility maximization,” or that rationality demands the relentless maximization of “self-interest” to the exclusion of other reasons for choice.

In his later work, Sen (2009, p. 180, emphasis in the original) specifies: “[...] rationality is primarily a matter of basing —explicitly or by implication— our choices on reasoning that we can reflectively sustain, and it demands that our choices, as well as our actions and objectives, values and priorities, can survive our own seriously undertaken critical scrutiny.” Against the convention, this means that for the CA de gustibus est disputandum; that is, preferences and values are not assumed but must be subject to study. Consequently, in contrast to the dominant rational agent, the CA’s agent has been described as a reasoning agent (^{Garcés, 2020b}).

A. Decision-making: individuals and context matter

BE recognizes that humans in actuality fail to conduct themselves in accordance with the axioms of full rationality assumed by mainstream economic theory, i.e., neoclassical economics (^{Dawnay & Shah, 2005}). This recognition is based on sound, mostly experimental, research. These divergences from the standard model are referred to as misbehaviors instead of irrationalities, in light of the negative connotations of the latter (^{Thaler, 2015}). They ensue because people’s decision-making takes place via a process consisting in two systems, each of which can fail. One is automatic and the other is reflective (‘rational’) (^{Thaler & Sunstein, 2009}).

Whereas system 1 is engaged automatically, is in charge of non-reasoned action, works effortlessly, and involuntarily, system 2 operates only when required, is responsible for reasoned action, requires effort, and is deliberately engaged (^{Kahneman, 2011}). S1 works sufficiently well most of the time, providing satisfactory outcomes. This suggests that, at best better outcomes are possible and at worst, outcomes can be counterproductive. Despite its capacity, S2 is not infallible either and can also produce sub-optimal outcomes. The use of S1 and S2 is a matter of degree, however, not an all-or-nothing situation. The power and effectiveness of S1 and S2 can change. Training them can improve their performance and other factors like age (childhood and elderliness) can diminish it. Cognitive skills and discipline are subject to betterment and can influence both. Indeed, the effort demanded by perfecting talents and abilities decreases as they are practiced (^{Kahneman, 2011}), depending on personal features as well. The above notwithstanding, S2 can also fail and often enough, it does.

S1 and S2 are both necessary and complementary. While S1 is intuitive and impulsive, requiring minimum effort, S2 is conscious, reflective, and deliberate, demanding more energy. Beyond their differences, one commonality worth stressing is that both are prone to fail.

Failures in S1 and S2 frequently, and to varying extents, depend on the context in which the decision-making is performed, or its choice architecture. This is the setup or circumstances in which choice takes place. For example, different arrangements of the options available can induce different choices, which may or may not be optimal or even aligned with the chooser’s (longterm) preferences (^{Thaler & Sunstein 2009}; ^{Sunstein 2020}). This means that for BE, decisions are embedded within a situation. Thus, actual behavior, and the systematic deviations from the rational model, are explained by personal biases and heuristics as well as the features of the context. That is, the environment in which choices are made also matters.

Significantly for policy-making, because choices are necessarily made in a context, deciding on the choice architecture is inevitable (^{Thaler & Sunstein 2009}). Whether aware of it or not, ‘architects’ facilitate, or not, the attainment of people’s goals and the enhancement of their well-being.

B. Decision failing: individuals and context matter

Deviations from the rational convention have been referred to as ‘reasoning failures’⁴. The literature uses this term to cover failures in both S1 and S2, although reasoning is mainly performed by the latter. Although such an approach may be justified in light of the aforementioned fact that both systems are a matter of degree and subject to change, it seems more straightforward to talk about rationality failures to encompass both, so as to avoid confusion when failures in different systems can be clearly identified. The causes of these failures can be categorized into four types of limitations associated to: i) technical abilities; ii) imagination or experience; iii) objectivity; and iv) willpower (^{Le Grand 2008}). Limited technical ability indicates the incapacity to assimilate and interpret information. Limited imagination and experience refer to the incapacity to foresee states of the world not lived or hypothetical, and one’s possible reactions to them (^{Le Grand, 2008}). Limited objectivity refers to the incapacity to be impartial towards one’s view. Limited willpower indicates weakness of will. Information and experience, however, do not guarantee optimal outcomes in the long run.

Many of these limitations, biases and heuristics emphasize the fact that preferences and choices are not made in a vacuum but necessarily within a specific context. Human action takes place embedded in a situation. This context, or the circumstances and conditions surrounding choice, is referred to as ‘choice architecture’, and inescapably influences decision-making, whether deliberately set up or not (^{Thaler & Sunstein, 2009}; ^{Thaler, 2015}). Because of this, and the fact that this architecture can be modified, much of BE’s contributions have sought to contribute to policy-making and address those failures (^{Sunstein, 2020}; ^{Vlaev & Dolan, 2009}).

In terms of behavioral policy, the recent literature has made a distinction between nudges and boosts (^{Grüne-Yanoff & Hertwig 2016}; ^{Hertwig & GrüneYanoff 2017}). Deriving from behavioral insights, both assume that people make decisions employing a limited number of heuristics and that whether they work or not depends on the characteristics of the environment in which the decision takes place (^{Grüne-Yanoff & Hertwig 2016}). Additionally, nudges and boosts are both intended to change human behavior without resorting to conventional interventions; that is, neither using substantial material incentives nor resorting to coercive mandates. The difference, briefly put, lies in how they affect behavior. While nudges instrumentalize people’s cognitive limitations to influence their behavior, boosts expand people’s set of competences to induce theirs (more on this presently). In this sense, the distinction seems to be differentiating behavioral interventions roughly into those targeting S1 failures as ‘nudges’ and those aiming at S2 as ‘boosts’ (^{Garcés-Velástegui, forthcoming}).

C. Behavioral agency: fallibility in multiple motivations

BE’s account of human behavior challenges the framework of full rationality and presents a description of how people actually act. This means that rather than establishing elegant axioms suggesting how people ought to act, and often do, to achieve optimal outcomes, BE acknowledges that human beings are fallible and founder in their pursuit of optimal welfare outcomes. These failures are systematic and, therefore, exploring them contributes to economic explanation and prediction. For BE, human beings are “plural, more and less reflective choosers, and multi-motivated” (^{Garcés-Velástegui, 2022b}).

In brief, S1 makes suggestions to S2 which, under normal circumstances, are endorsed by it without adjustment or revision, turning impulses into voluntary action. Most of the time, this less reflective process dominates human action. Only when the situation demands it, S2 takes over. This complimentary interaction between S1 and S2 is efficient, leading to sufficiently good outcomes, but not necessarily optimal because people fail.

These have been referred to as reasoning failures, which can ensue due to personal limitations, and contextual features, or a combination thereof. Therefore, for BE, human action is embedded in a situation, which means that context and history matter. While perhaps the best illustration of the influence of context on human behavior is the inescapable ‘choice architecture’ (^{Thaler & Sunstein, 2009}; ^{Sunstein, 2020}), history can be exemplified by the habits, talents, and skills, learnt (or not), which is related to inter alia their social norms, and culture. This exposes BE’s attention to human diversity.

The literature has added other-regarding goals as motivations orienting human action within BE. This is another challenge to the model of full rationality and its exclusive focus on the maximization of self-interest. For BE, this remains an important objective but not the only one. Egoistic as well as altruistic aims are increasingly being incorporated by the approach (^{Garcés-Velástegui, forthcoming}).

A. Behavioral mandates for human development

Under the label of mandates, for current purposes, fall all interventions employing substantial material incentives or coercive legal instruments. Although BE has mainly influenced the use of alternative policies leading to the interesting proposition of libertarian paternalism (^{see Sunstein & Thaler, 2003}; ^{Thaler & Sunstein, 2003}), this does not exclude the possibility of using behavioral insights for more traditional policies.

At their core, mandates establish an incentive, rewards, or punishment, in order to produce a particular behavior. Underlying these interventions is usually the conventional model of instrumental rationality, stressing the individual’s sole concern with the maximization of its utility and the calculation involved therein. Since BE uses the convention as a benchmark, its insights can be put at the service of these interventions as well.

Indeed, this is arguably the intuition behind one of the most widespread poverty-alleviating policies: conditional cash transfers (CCTs). In brief, they are monetary transfers provided to households under an income poverty line and paid upon verification of certain conditions, most often school attendance of children within the household and periodical medical check-ups of pregnant women as well as infants. By changing the situation through a substantial material incentive, the policy seeks to produce a behavior in beneficiaries that they would not otherwise engage in, despite it being in their own self-interest. It becomes a mandate for beneficiaries due to the condition with which they are obliged to comply in order to receive the transfer.

To be sure, this policy has been conventionally explained in terms of opportunity costs. Under normal circumstances (without the intervention), poor households cannot afford to invest in education or health since the little they have barely suffices to survive. The merits of this perspective notwithstanding, it does not exclude the presence of reasoning failures. It seems sensible to argue that not all beneficiary households are equal. Despite all falling under a poverty line, there are those that are closer to it and those that are farther, those that have more and those that have less vulnerable people (children, elderly, etc.). For those families that can afford to invest in education and health and still do not unless CCTs are implemented, an argument can be sensibly made on the basis of behavioral economics.

In this respect, although such interventions may be regarded as freedom restricting, in both the short and long term, people’s choices can be expanded, as the CA advances. Not only are education and health valuable things in and of themselves, these outcomes can lead to other opportunities for the exercise of substantive freedom and achievements as well.

Another relevant policy debated lately is the possibility of vaccination mandates to fight the spread of COVID-19. Governments are considering this policy to safeguard people’s well-being in light of some groups’ reluctance to get vaccinated on autonomy-related and other grounds. Behavioral economics can explain many of these positions as being based on reasoning failures. Based on such explanations, via a mandate, people’s health, welfare, and even life itself, can be prioritized at the expense of some freedom because it is compelling people to do something that is in their self-interest despite their refusal to deem it so.

B. Nudges for human development

Nudges are the first, and perhaps most famous, type of interventions inspired by behavioral economics that are intended to change people’s behavior without changing material incentives substantially or resorting to coercive legal instruments. They have been defined thusly:

[A] ny aspect of the choice architecture that alters people’s behavior in a predictable way without forbidding any options or significantly changing their economic incentives. To count as a mere nudge, the intervention must be easy and cheap to avoid (^{Thaler & Sunstein 2009, p. 6})

These interventions assume that people employ a limited set of heuristics in their decision-making and seek to steer individuals towards a given conduct through adjustments in the context in which decisions are made so as to trigger a specific heuristic strategy. In other words, “the nudge approach instrumentalizes these cognitive limitations to influence behaviour” (^{GrüneYanoff, Machionni & Feufel, 2018, p. 4}). Heuristics, thus, are regarded as sufficiently stable so that changes in the environment consistently activate the same heuristic leading to the same expected behavioral changes. In fact, the initial relevant literature treated them as similar to optical illusions (^{Tversky & Kahneman 1986}; ^{Kahneman & Tversky 1996}) that cannot be corrected despite one’s awareness of them. That is, heuristics are not easily changed and training to improve them can be difficult.

There are numerous successful illustrations of the use of nudges to increase people’s well-being. From the now famous ‘Save More Tomorrow’ program designed to increase the contributions people make for their pensions (^{Thaler & Sunstein 2009}), to the warnings of various types (texts and images) on cigarette boxes, the goal is to harness a particular heuristic with each intervention on the environment to produce a predicted behavior.

Although such experiences have been conceived in developed nations, there is evidence from developing contexts as well (^{see e.g., Alpizar et al., 2020}; ^{Nelson, Partelow & Schlüter 2019}; ^{Sudarshan 2017}). Some examples are Colombia’s pop-up messages to remind people to pay their social protection contributions online (Alm et al., 2019) and Ecuador’s ‘nutritional traffic light’ which alerts potential consumers about the levels of fat, sugar, and salt color-coding them so that red is high, yellow is medium, and green is low.

Consequently, since nudge interventions affect choice, they can lead to the achievement of reflected-upon valued doings and beings. This means, that complementary policies are advised to allow people to expand their own choices.

C. Boosts for human development

Boosts are the second, and perhaps lesser known, type of interventions inspired by behavioral economics seeking to produce a change in behavior employing neither a substantial change in material incentives nor coercive legal mandates. As its name suggests, “boosts have been characterized by their ‘goal of expanding (boosting) the decision maker’s set of competences and thus helping them to reach their objectives’” (^{Grüne-Yanoff & Hertwig 2016, p. 156}).

Much like nudges, boosts also assume that people’s decision-making is based on a limited set of heuristics. However, there are also, at least, two important differences. First, boosts endeavor to produce a change in behavior by enhancing people’s competences; i.e., overcoming people’s cognitive limitations, instead of harnessing or instrumentalizing them. Second, rather than targeting exclusively the environment in which decisions are made, boost policies can also address the individual’s set of heuristics directly.

This being so, the relationship assumed between the environment and the heuristic is distinct for boost interventions. The environment, in this case, does not activate or trigger a heuristic. Instead, it furnishes the individual with informational cues leading them to select a heuristic from a repertoire (^{GrüneYanoff, Machionni & Feufel, 2018}). If cues in the context in which the choice takes place do not match a given heuristic, people resort to their inventory of heuristics and select one that accommodates the situation more readily. Since the individual exerts agency in the selection of heuristic, rather than it simply being triggered, the assumption is that it is possible to improve that selection and learning of other heuristics via training. This is how boosts intervene on the person’s heuristic repertoire instead of on the environment.

The literature has suggested that simple accessible rules can be more effective to make sufficiently accurate assessments than more complicated, technical, or sophisticated training (^{see Gigerenzer et al., 1999}). From stock exchange to food consumption, there is growing evidence of the benefits of such an approach.

For developed and developing nations alike, the main lesson seems to be that boost interventions can be effective when there is the possibility to teach simple rules of thumb or simple ways to train people to undertake complicated procedures. They improve people’s competence to make better choices in the pursuit of their own goals. For example, people’s financial decisions can ameliorate if they are taught simple financial and accounting rules (^{Drexler et al., 2014}), and parents can improve the nutritional habits of their children by modelling healthy behaviors during family meals (^{Dallacker et al., 2019}).

Accordingly, given that, much like nudges, boost policies affect choice, they can contribute to people leading the lives they value and have reason to value. At the same time, however, since boosts can expand the heuristics available for individuals, they seem to be expanding the choices people can make. Put otherwise, such interventions can help people to reach both ends advanced by the CA: functionings and capability.

D. CAand BE-inspired policies: Freedom and Achievement

As the discussion in this section suggests, there is considerable coincidence between behavioral public policies and human development policies, based on the CA. Both place freedom and achievements at the locus of attention. On the one hand, the CA has redefined development as freedom and, since development proposes ends as well as means, it suggests that freedom is the end as well as its principal means (^{Sen 1999}). Functionings, whether as well-being or agency more broadly, can depend on the exercise of freedom and be a consequence of it. Policies inspired in the CA, therefore, can enhance capability, and thereby functionings, or functionings alone (^{Sen 1992}). The CA admits that functionings can be improved directly, but such policies would be inferior to capability-enhancing ones as they overstep people’s freedom (^{Sen 1999}). On the other hand, behavioral public policies are mindful of freedom and highlight the extent to which it may be curtailed by welfare promoting interventions. They do so by focusing on the exercise of choice in the pursuit of well-being (whose definition seems closer to the CA’s notion of agency functioning since it increasingly encompasses other-regarding goals), as defined by people themselves, and also by exposing why people often settle for secondary outcomes. Well-being, thus, can be induced by suggesting effective ways in which people can be helped to reach their goals, preferably without hurting their freedom.

In this sense, BE makes a policy-relevant contribution to human development. It can be argued that whereas the capabilitarian framework proposes what development is, behavioral insights clarify and classify some pathways on how to attain it. Should freedom, or the expansion of meaningful choices, as proposed by human development, be enhanced, then boosts are the safest bet. If there is levelled concern between freedom and achievements in human development, nudges are most likely to reach that balance. When achievement is prioritized in human development, even at the expense of some freedom, then mandates become an option. As the discussion above suggests, however, these alternatives are not necessarily mutually exclusive. There might be issues and circumstances for which they can be complementary and different combinations can prove useful. Ultimately, the policy choice is political. It is up to the people in the public sphere to decide, and this is yet another aspect in which the CA and BE would find agreement. The discussion elaborated here can prove useful in that debate by shedding light on the implications of different behavioral policies for human development.