Combining Philosophical and Democratic Capability Lists

2.1 The Question of the List

The question of the list concerns how societies should choose what to distribute to their members. Capabilitarians frame the question in terms of substantive freedoms, called ‘capabilities’, to be and do those things that people have reason to value, called ‘functionings’ (Byskov 2019; Khader and Kosko 2019; Robeyns 2003, p. 63, 2005a, 2016, p. 405f., 2017, pp. 169–174). A freedom to satisfy nutritional needs, for instance, is substantive when jointly sufficient conditions to facilitate the act of eating enough are met (Sen 1980, p. 218, 1993, p. 40). People have this capability if they meet conditions such as having healthy digestive systems and access to food. People have the related functioning once they eat. In answering the question of the list, the selected capabilities and functionings ought to capture what is in people’s interests and reflect their priorities (Alkire 2006, pp. 133–136).

Despite capabilitarians agreeing that valuable ends should be distributed, the question of the list remains divisive (Robeyns 2005b, p. 105ff., 2017, p. 169; Terlazzo 2019, pp. 283–288). The question of the list is often considered to impose a dichotomous choice between Sen’s (1999, p. 110, 2004a, 2004b, p. 333, 2009, p. 43f.) and Nussbaum’s views (1992, 2003, 2006, p. 70, 2009, p. 334). The divisions between Sen’s and Nussbaum’s views concern the influence of contingent versus universal considerations when selecting capabilities (Austin 2018, p. 64).

In analysing the divisions between Sen and Nussbaum, Claassen (2011) calls Sen’s view the democratic position and Nussbaum’s view the philosophical position. The democratic position is a procedural selection method. It focuses more on the process of forming lists than on which items are chosen (Byskov 2018, pp. 110–115). More precisely, Sen (2004a, 2004b, p. 333) argues that people should select capabilities collectively. In comparison, Nussbaum’s philosophical position is a foundational selection method (Byskov 2018, pp. 105–110). The philosophical position has ground level philosophical principles play the primary role in capability selections (Nussbaum 1987, 1990, 2003, p. 41f., 2011, p. 33f.). The ground level philosophical principles make up strict boundaries for contextual specifications of ends in, for instance, different times and places (Nussbaum 1992, p. 214f.; Qizilbash 2007, p. 176).

Claassen’s classifications highlight the democratic and philosophical positions’ different functions. The democratic position’s function is to respond to people’s varying assessments of capabilities and make selections that respect people’s autonomy. The philosophical position’s function is to provide a list of incontrovertibly valuable capabilities. Thus, the divisions concern the trade-offs between political legitimacy and philosophical accuracy. Recently, calls for reconciliation have been made to move beyond this purported dichotomy and its trade-offs.

2.2 The Reconciliatory Research Agenda

Different selection methods are suitable for different purposes (Byskov 2018, pp. 100–103). Procedural methods can be useful for gauging a population’s priorities. Foundational methods can be useful for formulating theories. Similarly, ad hoc methods that, for example, depend on research interests or what empirical information is available can be useful to ascertain who is poor. For instance, the Human Development Index and Multidimensional Poverty Indices involve such empirically informed selections for that purpose (OPHI and UNDP 2020, p. 4f.; UNDP 2010, p. 7f., 16, UNDP 2018, p. 9). However, in policy-making, we do not only need to settle descriptive matters about, for example, what people prioritise, or what well-being is, or what we can measure. These descriptive matters often need to be tied to normative views about what to do about them. The reconciliatory research agenda gives equal priority to democracy and philosophy to combine the strengths of both viewpoints.

In formulating his reconciliatory research agenda, Byskov’s (2017, p. 2) ambition is to steer the debate on capability selections toward compromise rather than competition. Compromising, in this context, is about combining the philosophical and democratic positions in some mutually reinforcing way (Byskov 2017, p. 4). In public policy-making, Byskov (2017, pp. 5–10, 2018, pp. 115–121) argues, there are dependencies between the philosophical and democratic positions that can make an imbalanced focus on either of the two illegitimate. Compromises are thus both possible and sometimes warranted, on Byskov’s view.

The reconciliatory research agenda offers a sympathetic critique of Claassen’s (2011) combination of the philosophical and democratic positions. On Claassen’s view, the philosophical and democratic positions are interdependent, but not entirely equal. Claassen responds to Crocker (2008, p. 55), who, despite rejecting the purported dichotomy between Sen and Nussbaum, sometimes favours the democratic position (2008, p. 139). In his response, Claassen (2011, p. 502) argues that favouring the democratic position can impoverish public debates. Additionally, Claassen (2011, p. 505) argues that the aim of selecting capabilities is not to arrive at mutually acceptable agreements. Instead, a refined version of the philosophical position should be embedded in the democratic position. One of the reasons he proposes for this is that philosophical theories offer clarifications and systematisations that help democratic list-making procedures (Claassen 2011, p. 502).

Claassen’s (2011, p. 504f.) envisioned philosopher–investigator embraces conflicts between selections from philosophers and the public but does not (aim to) resolve them. The philosopher–investigator provides a list based on his or her own individual reflections and data gathered, for example, from public debates and social scientific research. Thereby, its view of the good of people is informed by other people’s conceptions of the good life. By giving recommendations to the population on these two bases, the philosophical and democratic positions become interdependent. Nevertheless, despite seeking out other people’s opinions, people in general and the philosopher–investigator can end up disagreeing about what a good life consists in. When substantial disagreements obtain about conceptions of the good life, the philosopher–investigator is entitled to recommend a capability list that he or she considers best, on Claassen’s view. Thus, the philosopher–investigator is not committed to recommending selections that it and the population mutually accept. Claassen’s reason is that beliefs that a capability is morally called for do not necessarily coincide with them being morally called for (2011, p. 505). As an advisor, the philosopher–investigator may respectfully disagree with other people’s conceptions of the good life and give divergent recommendations.

Byskov agrees with Claassen that the philosophical and democratic positions are interdependent. However, Byskov (2017, p. 4) argues that Claassen’s embedding of philosophical theory in the democratic position is less legitimate than Claassen suggests. Byskov (2017, pp. 11–13) concludes that reconciliation is not about regulating democracy through philosophy, but about reaching selections that gain both philosophical and democratic support.

In developing an alternative view of the interdependence of philosophy and democracy that avoids exaggerating philosophy’s importance, Byskov (2018, p. 117) identifies some selection methods as synthesising methods. They give philosophical considerations and democratic deliberations an equal but not overriding priority (Byskov 2018, p. 118). Their purpose is to produce philosophical and democratic lists as input for a dialectic process that somehow integrates them (Byskov 2018, p. 122).

The extent to which Byskov’s alternative view can be made mutually reinforcing for well-being policy-making purposes remains underexplored, however. Hence, I will provide two necessary criteria for the reconciliatory research agenda to meet with regard to urgent situations involving well-being.

3.1 The Equilibrium Criterion

Equality, in terms of priority, between philosophy and democracy in the reconciliatory research agenda is methodologically important. If equality between philosophy and democracy is not maintained by a selection method, then it will be a comparative method. In comparative methods a list of capabilities is produced as a standard of comparison (Byskov 2018, p. 117). Other capabilities are compared against that standard and given a lower priority if they conflict. A comparative method thus treats one of the philosophical and democratic positions as lexically prioritised. Such lexical prioritisation fails to shift the focus from competing to compromising, which is the aim of the reconciliatory research agenda. In an appropriate reconciliation of the philosophical and democratic positions, the philosophical and democratic grounds for selecting capabilities ought to be balanced in a way that comparative methods by necessity will fail to do.

The commitment to an equal priority of philosophy and democracy in the reconciliatory research agenda respects universal assessments of goodness and badness in accordance with philosophical considerations while maintaining a serious commitment to people’s contingent opinions. In contrast with comparative methods, synthesising methods afford equal priority to philosophical considerations and democratic deliberations, and thus offer a promising starting point.

However, equality of priority in synthesising methods can be realised in different ways. This is because synthesising methods are compatible with different interpretations of what priority entails. As a result, the equalisandum, or the thing to be equalised by synthesising methods, is not fixed. The task is thus to identify an interpretation of priority that is morally salient when significant amounts of well-being are at stake in time-sensitive conditions.

Not every interpretation of priority will be morally salient for capability selections in urgent situations. For instance, synthesising methods are compatible with interpreting priority in terms of the number of selections that conflicting parties are afforded. Such a synthesising method could be motivated by the view that there is no privileged position from which to gauge the value of selections being made. If the various selections cannot be gauged in terms of how valuable they are, giving each interested party the same number of things that they want appears equitable and justified. We may call this interpretation of priority the equal-numbers view. Reconciliations in urgent situations require a different interpretation of priority than the equal-numbers view offers, however. To see why, consider a situation where one of the philosophical or democratic positions determines what shelter-related capabilities people get in the aftermath of a natural disaster while the other position is only considered with regard to what hobbies people can take up to pass the time while the shelters are built. If both positions consider the same capability to be more important than the other in terms of well-being, then one of the philosophical or democratic positions is substantially disfavoured on the equal-numbers view.

Reconciliations in time-sensitive situations that involve significant losses of well-being ought to take into account how much interested parties value different outcomes. Consequently, a suitably restricted interpretation of priority is needed. As a viable complement to the comparatively underspecified accounts of priority in the synthesising methods, consider the following criterion:

Equilibrium: A reconciliatory selection method treats philosophical considerations and democratic deliberations as equally weighted when selecting capabilities.

The equilibrium criterion, in contrast with the equal-numbers view, captures salient aspects by gauging equality of priority in terms of how much influence each position has in a selection process. It expresses a normative commitment to treating democratic deliberations and philosophical considerations in terms of how strongly capabilities are valued by the philosophical and democratic positions. The equilibrium criterion equalises the degree of input each interested party has in deciding what a valuable outcome is, without requiring that they base their evaluations on a shared conception of the good. Thereby, the equilibrium criterion captures what and how much is at stake for disagreeing parties, whether their respective evaluations are based on ground level principles or on people’s opinions. Since the equilibrium criterion is about whose input matters, rather than what input matters, it offers a fair interpretation of priority in urgent situations. Thus, the equilibrium criterion is suitable when both political legitimacy and philosophical accuracy are at stake simultaneously. The equilibrium criterion will consequently be adopted to design a suitably restricted synthesising method.

3.2 The Problem of Overgeneration

While the equilibrium criterion captures a morally salient aspect of equal priority, treating philosophy and democracy as equally prioritised can cause implementation issues. This is because equalities of priority can lead to indeterminacy (Byskov 2018, p. 120). The first step of a synthesising method is to create a comprehensive list that includes the items from philosophical and democratic proposals. Synthesising methods thus partly make public policy about covering potential blind spots in the philosophical and democratic positions. However, the comprehensive list will not necessarily be consistent. Two examples follow.

First, one list might include capabilities to have bodily autonomy, including decisions over tests and treatments for infectious diseases. In contrast, the other list might prioritise capabilities for bodily health, even at the expense of drawbacks to people’s independence from interference. Both capabilities may be valuable for well-being related purposes, but implementing them simultaneously in a well-being policy is unfeasible.

Second, consider a disaster such as a flood or wildfire that causes the destruction of a significant portion of people’s homes. As a result, people urgently need shelter to survive. One list detailing which capabilities shelters should facilitate may be grounded in what would maximally benefit individuals. However, another list might, for example, hold that it is more important to secure a near-minimum level of shelter for all, rather than ideal shelter for some. Consequently, a conflict between prioritisations arises that ought to be handled so as to avoid significant losses of well-being from both perspectives.

Further examples of urgent situations can be produced, and the thresholds for time-sensitivity and losses of well-being can vary in different situations. Setting these thresholds is important for classifying situations as urgent or non-urgent. For the current purpose, however, it suffices to note that the there are paradigmatic urgent situations that ought to be handled.

While urgent situations may be numerous, philosophical lists and democratic lists will not necessarily disagree with each other over what to do about them. It is not unlikely that the lists will partially overlap regarding some capabilities for meeting basic needs such as eating and drinking. But, as the examples above illustrate, consensus is not guaranteed.

When disagreements obtain, capabilities are overgenerated in the sense that they are selected for distribution but cannot all be distributed since they are mutually exclusive. If the overgenerated capabilities are afforded equal importance, we need some form of tie-breaker.

To see why overgeneration is a problem, consider an alleged paradox of democratic theory (Walzer 1981, p. 385f.; Wollheim 1962). First, imagine that a citizen deems some policy A to be better than another policy B, where A and B are mutually exclusive. Every other citizen deems B to be better than A, however. The first citizen, being a devoted democrat, is drawn to conclude that both A and B should be implemented. This, however, is inconsistent.

As Goodin and Spiekermann (2018, p. 32) note, it is not paradoxical that an individual accepts conflicting recommendations on different bases. The crucial point, however, is that it would not be morally reasonable to forgo, or practically possible to implement, both options. The comprehensive recommendations from synthesising methods give too little guidance for handling ties in urgent situations. In such urgent situations, the problem of overgeneration ought to be better addressed. Thus, a further criterion is needed.

3.3 The Decisiveness Criterion

The overgeneration problem can have a negative impact on the balance between philosophical accuracy and democratic legitimacy. Incompatibilities in selections can affect many people in significant ways, even if incompatibilities are rare. There can sometimes be a large intersection between the philosophical and democratic lists. There can even be a large set of non-conflicting capabilities exclusively in the philosophical list or exclusively in the democratic list. However, one single conflict can nevertheless affect the capabilities of an entire population. To see how, consider a list that includes the capability to protest. Next, imagine a competing list that instead includes the capability to live in a stable political order. Even those who value democracy may be conflicted about which capability to select. For instance, Przeworski (2019, p. 37) proposes that anti-government demonstrations can be a part of everyday life, but notes that the stability of democratic governance is at risk of failing when protests turn violent. The proposals are jointly incompatible, and indeterminacy fails to secure either proposal’s benefits.

Synthesising methods will not on their own prioritise any alternative from conflicting selections, but they ought to do so in urgent situations. People’s well-being entitlements, as determined by the relevant philosophical and democratic considerations, are at stake. A tie-breaker is needed to prevent impending losses of well-being in urgent situations. Therefore, consider:

Decisiveness: A reconciliatory selection method for use in urgent situations should determine consistent capability selections.

The decisiveness criterion should be met in urgent situations to make the synthesising methods appropriately determinate. Meeting it ensures that people’s well-being entitlements are not systematically left aside when promoting equality between philosophy and democracy.

In non-urgent situations, the decisiveness criterion may be overdemanding. The level of urgency depends on how much well-being is at stake if conflicts remain and on how time-sensitive they are. When relatively insignificant levels of well-being are at stake, conflicts may not need to be resolved. Claassen’s philosopher–investigator securing room for respectful disagreement might suffice. And even if a lot of well-being is at stake, if the situation is not time-sensitive, then Byskov’s pure synthesising methods may suffice. For urgent situations, however, decisiveness is a necessary normative criterion that selection methods should meet.

4.1 Regulated Voting

The selection method I am proposing involves a determinate philosophical list and a determinate democratic list. I restrict the discussion to disagreements between groups. The setup idealises from real-world conditions where philosophers and laypeople rarely agree fully within their respective groups. However, in so doing, it helps develop the pure synthesising methods without depending on a general solution to social choice problems. I assume that neither philosophers nor democrats face issues that aggregating sets of conflicting judgments or preferences can lead to, such as cyclical orderings (see List and Pettit 2002, 2004; Gehrlein 2006, pp. 16–19). For the sake of argument, assume that philosophers agree about which capabilities matter for people’s urgent well-being entitlements, even if they sometimes disagree about peripheral ones. Similarly, assume for the sake of argument that a population facing an urgent situation understands what the affected people would prioritise. This setup mirrors a representative philosopher and a representative democrat compromising on capability selections together.

I will now describe a regulated voting procedure that lets synthesising methods meet the decisiveness criterion while maintaining an equal prioritisation of philosophy and democracy. First, a comprehensive list of capabilities based on philosophical and democratic proposals is compiled. Next, conflicting capabilities in the comprehensive list are separated from the compatible ones. Finally, a philosophical representative and a democratic representative independently weigh the importance of including each capability by voting on the conflicting capability pairs. The aim is to create local imbalances on selections of capabilities while maintaining the equality between philosophy and democracy globally in the selection process. By following this regulated voting procedure, disagreements do not result in conflicts between capabilities, but produce selections.

The regulated voting procedure’s decisiveness relies on three conditions. First, that there is more than one pair of mutually exclusive capabilities to vote on. Second, that each vote is used. Third, that votes are reassigned by both parties when a draw over the comprehensive list of mutually exclusive capabilities obtains. Once an imbalance of weights across options arises by voting, additional capability selections are produced among the overgenerated options.

In the following, I illustrate the regulated voting procedure through a numerical example. The two parties that vote are Philosophers and Democrats. The number of conflicting capability pairs is set to an arbitrary but manageable number, n, greater than one: say, four. The conflicting capability pair members are designated by A and A*, B and B*, C and C*, as well as D and D*. Philosophers propose A, B, C, and D while Democrats propose A*, B*, C*, and D*. A and A* are mutually exclusive but compatible with any configuration of B, B*, C, C*, D, D* (and so on for the rest). Philosophers and Democrats are then assigned an equal number of votes, v, equal to or larger than n, say 100 in this case.

Since Philosophers and Democrats indicate their priorities by voting, assigning an equal number of votes to both parties ensures that the equilibrium criterion will be met. In short, the disagreeing parties will have an equal say in the decision-making process. The reason that the number of votes should be at least equal to the number of conflicting capability pairs is that the philosophical and democratic proposals are about valuable outcomes. It should thus be possible to place some importance on each proposal. By allocating more votes than conflicting capability pairs, each option can be given some weight even when some option is considered to be more important. Following this reasoning, we might at first reach a result such as this:

Since a draw over the complete set of conflicting capabilities obtains, the voting process starts again. An outcome like the following might be produced, where wins are indicated by bold text:

Because cardinal information is used to reflect the prioritisations of the outcomes, comparisons between them can be facilitated (see Sen 1984, p. 4). In this numerical example, the gains for Philosophers and Democrats amount to a value of 55 votes, respectively. The numbers representing the valuable outcomes can vary between Philosophers and Democrats, but the outcome will be a positive result, r, from their respective viewpoints of 0 < r ≤ v. Thereby, indeterminacy stemming from ties is mitigated while equilibrium is maintained.

Beyond tallying the positive result r, respectively for Philosophers and Democrats, we should consider the external effects of making further capability selections. These further selections produce social good beyond the voting procedure’s numerical result. The effects of those selections will be a benefit for the members of society who receive them. The regulated voting procedure thereby provides a weak Pareto improvement over pure synthesising methods, since everyone has more value by using it than is produced by pure synthesising methods (see Sen 1979, p. 539, 1984, p. 24).^[1] Each party in the selection process can achieve more than ties allow for. Consequently, in urgent situations, impending losses of well-being will be mitigated.

Despite the regulated voting procedure’s benefits, its three conditions do not guarantee that decisiveness is met in two steps. There is a possibility for votes to serially mirror each other across the options, leading to further restarts of the voting procedure. Instead of moving from STEP 1 to STEP 2, we can imagine scenarios moving from STEP 1 to something that requires another restart, like STEP 2*:

However, as the number of conflicting capability pairs increases, this mirroring problem becomes less likely in practice. Similarly, as the number of votes increase, the likelihood of mirroring decreases. This second point is particularly important, as it does not rely on Philosophers and Democrats disagreeing broadly. It illustrates a possibility of increasing the number of votes to avoid mirroring while also allowing for fine-tuned settings of weights.

The regulated voting procedure goes some way toward achieving a consistent integration of philosophical considerations and democratic deliberations. However, more work remains to be done. The regulated voting procedure is silent on what to do when only one pair of capabilities present mutually exclusive options. Moreover, it assumes that philosophers and democrats are willing to reassign votes in case they vote with exactly the same number of votes on all conflicting capabilities. The regulated voting procedure, moreover, does not guarantee that there will not only be one further capability that each party wins, while some remainder is excluded. Thus, whilst this addition improves on how pure synthesising methods handle ties, future work ought to develop this voting procedure to secure even higher levels of decisiveness. To that end, the voting procedure could in principle be run with a modified third condition that disallows all draws to increase the number of wins by clearing all conflicts. However, doing so may be costly timewise and the weights set may not accurately represent philosophical and democratic opinions. Efficiency in terms of time and representation trades off against the degree of decisiveness achieved. The moderately regulated voting procedure strikes a reasonable balance. It will lead to better outcomes while maintaining the equal status of disagreeing parties.

4.2 Consensus and Non-ideal Alternatives

The regulated voting procedure offers a way to maintain equilibrium while increasing the decisiveness of synthesising methods. It locates the place of compromise at the level of all options, rather than each conflict. To avoid the problems of indeterminacy found in pure synthesising methods, we should not aim at universal consensus between philosophers and democrats. To see why, I introduce the consensus view. The consensus view selects the overlap between capability and functioning selections made by philosophers and democrats.

A notable issue with the consensus view is that it significantly limits the scope of capabilities relevant for urgent well-being entitlements. Using the consensus view, only those capabilities that the philosophical list and the democratic list both agree on are considered relevant for well-being. However, equalising the priority between philosophy and democracy in this way gives too much power to some (variable) party in conflicting capability selections.

The consensus view introduces a problem similar to the dictator in social choice theory. The dictator is a fixed individual who gets what he or she wants regardless of the opinions of others (see List 2018, p. 466f.; Sen 1977, p. 1543). In practice, what we may call the naysayer is a non-fixed party designated as whoever is the most restrictive about including a capability in a conflict. Given the consensus view’s way of equalising the weight given to philosophical considerations and democratic deliberations, the naysayer gets what he or she wants in each instance. The drawback is that a philosophical or democratic list that included no items at all would be the most effective decider, but gets no political work done.

It may be objected that we can seek consensus on different levels of generality, and that people may, at some level, promote the same things. For example, an overlapping consensus obtains when reasonable people from different groups with different particular beliefs and norms agree on a set of more general beliefs and norms (see Claassen 2018, p. 19; Claassen and Düwell 2013, p. 501; Nussbaum 2006, pp. 388–392). If the objection holds, the naysayer will be avoided.

The objection fails, however. The naysayer is problematic for reconciliators because it breaks the equilibrium criterion. The equilibrium criterion could, according to the objection, be met by formulating competing suggestions at higher levels of generality. However, urgent situations also demand that the decisiveness criterion is met. Urgent situations typically demand resolutions of disagreements on comparatively particular levels beyond general claims such as ones about promoting well-being or involving everyone equally. Seeking a consensus on higher levels of generality does not guarantee that we are left with sufficiently specific selections for handling urgent situations well enough. For disagreements that are relevant for handling urgent situations, something besides an overlapping consensus can be necessary. Even if further discussion can eventually lead to agreements that handle urgent situations, the process may be too slow. Therefore, the regulated voting procedure is a welcome addition in this context.

It may further be objected to my counterargument against seeking an overlapping consensus that there are non-idealised ways of securing consensus on suitably particular levels. It may, briefly put, be argued that pure synthesising methods identify what disagreeing parties value and that these things can enter into various non-ideal political bargaining processes.

However, this objection in favour of non-ideal political bargaining processes only goes so far. Firstly, such processes can also be time-consuming beyond thresholds of time-sensitivity in urgent situations. Secondly, in situations where the processes reach a determinate suggestion quickly enough, it can be because those with the most to lose tolerate concessions imposed by those in a better position. Compromises should meet the morally salient equilibrium criterion even when urgency calls for meeting decisiveness. It would therefore be useful to have a viable selection method, such as the regulated voting procedure, that meets both criteria by design.

Nevertheless, it should be granted that non-ideal procedures are likely to remain in use. They may, moreover, have other merits beyond those relevant to meeting the equilibrium and decisiveness criteria. For instance, they may reflect deep-rooted and stable political practices. However, even if the regulated voting procedure idealises by abstracting away from some of these non-ideal procedures, it can be used to evaluate them. For instance, failures to meet the equilibrium and decisiveness criteria in some decision-procedure can be identified by using the voting procedure. Once failures have been identified, claims for compensation can be made.

In summary, instances of overlapping consensus can fail to meet the evaluative criteria for reconciliatory selection methods. They are consequently not always reliable in urgent situations. Non-ideal practices offer alternative ways of integrating philosophical and democratic considerations. However, given that these non-ideal practices can take many forms, it is useful to have an ideal to compare them against. The regulated voting procedure that I have provided offers one such ideal that shows what a better outcome than the status quo would be. The regulated voting procedure is not only relevant for decision making. It also has an evaluative use for decisions that have been made in other ways. It therefore has two areas of application in our continued efforts to democratically achieve valuable outcomes.