This paper argues that incompletely theorized agreements (ITA) offer a robust framework for organizing a pluralistic society of a liberal constitutional democracy. I demonstrate this by showing that incompletely theorized agreements offer a minimalist and plausible framework to accommodate for the need of legitimacy, coexistence, and feasibility within pluralistic societies. My argument goes as follows. In pluralistic societies, there exists value pluralism: viz. that there are multiple value systems that govern legal, political, and social thinking that lead to disagreement (Mason 2015). For some disagreements, it is especially important to reach some form of agreement, as it would create a better state of affairs. Incompletely theorized agreements offer pluralistic societies a framework for reaching an agreement that is feasible in practice, legitimate by being non-arbitrary and non-dominating, and a pathway to coexistence. In addition, incompletely theorized agreements can overcome potential objections. Consequently, incompletely theorized agreements offer a robust framework or organizing a pluralistic society.
I do not focus on incompletely theorized agreements in constitutional law as Cass Sunstein, since this would require a more comprehensive treatment of the subject. I limit my argument to establishing it as a framework for pluralistic societies by reference to legitimacy.
The problem of pluralistic societies
A pluralistic society is one in which there are multiple conceptions of value, truth, and reason. Political pluralism identifies this fact and asks what limitations can a government place on the freedom to act and argue according to these values (Mason 2015). Here, I am concerned primarily with how to reach agreement under the condition of pluralism, which is a matter of fact in many societies. The consequence of value pluralism is that there is widespread disagreement between groups and individuals on the reasons for a particular outcome or practical arrangement.
For some such issues on which there is disagreement, reaching an agreement is highly desirable, such as reaching an agreement that resolves a civil war or provides healthcare to a disenfranchised portion of society. The point is that a better state of affairs hinges on agreement, while unresolved disagreement may result in the state of affairs being considerably worse off (e.g., continued civil war or poor people dying from preventable or treatable medical causes). But what ultimately decides what counts as a better state of affairs? It is the through incompletely theorized agreements that a better state of affairs is determined since it involves generating agreement on the desired end state that is acceptable to all or at least a majority. This is a weak understanding of a better state of affairs, but I give my reasons with reference to legitimacy and non-domination.
By way of caveat, I do not demonstrate here that societies are in fact pluralistic, because I have already limited the scope of my argument to only address pluralistic societies. In addition, I do not make a positive or negative claim about the equal validity thesis of political pluralism, although I do address this as a possible objection.
The solution of incompletely theorized agreements
Incompletely theorized agreements refer to the idea put forward by Cass Sunstein and others that there can be agreement on some outcome or practical arrangement, although there is disagreement on the underlying reasons or justification thereof (Sunstein 2000: 117). Since there is a wide range of disagreement on the reasons for an outcome, there can be no majority support (not to mention unanimity) on the justification of a particular outcome. For instance, a society composed of three equal parts of Kantian ethicists, rule utilitarians, and virtue ethicists attempt to organize a way to agree on laws. Let’s call this society S1. Each argues on the basis of a different reason for why murder should be illegal. The Kantian ethicists state that it should be illegal because it violates the Categorical Imperative and does not treat individuals as ends in themselves (R1). The rule utilitarians argue it should be illegal because it violates the rule that leads to the greatest amount of good (R2). Lastly, the virtue ethicists argue that is should be illegal because murdering contradicts core virtues of a good individual (R3).
Society S1
Reason 1 (R1)
Reason 2 (R2)
Reason 3 (R3)
Murder illegal
Kantian Ethicists
Yes
No
No
Yes
Rule Utilitarians
No
Yes
No
Yes
Virtue Ethicists
No
No
Yes
Yes
Majority of S1
No
No
No
Yes
In this situation, making murder illegal would clearly result in a better state of affairs, whereby there is a good reason to reach agreement on it. If they tried to reach a consensus or simple majority on the justification for this law, it would never be passed. However, if they agree on the outcome, by resorting to incompletely theorized agreements, there is a framework to pass the law. Therefore, it offers a solution to resolving disputes in pluralistic societies.
The benefits of incompletely theorized agreements
My argument diverges from Sunstein’s account through the addition of the normative component of legitimacy, which I argue gives ITA a certain forcefulness. How do incompletely theorized agreements offer legitimacy to a decision (i.e. agreement)? Here, I understand a decision as having legitimacy if it is procedurally based on majority support or consensus within the context of a liberal constitutional democracy. If a non-majority group imposed its standard of justification on others, it would be dominating and thus violating equal citizenship and pluralism. That is, it would be arbitrarily interfering with the self-determination of equal individuals to impact a decision. Such interference would be arbitrary, since it ascribes asymmetric power to a single group solely on the basis of their value system. Since we are dealing with a pluralistic society, where there are competing value systems and justifications for an outcome, a dominating standard of justification would be arbitrary since it is unaccepted by other groups and individuals. By using the framework of incompletely theorized agreements, the focus is on consensus in outcome rather than justification, which allows the decision to be representative of pluralistic society rather than merely one value system among many others. Thus legitimacy of the agreement is derived from the inherent authority of a democratic agreement.
Is this conception of legitimacy plausible? The rationale for this definition of legitimacy is rooted in the attempt to find a decision in a pluralistic society that is representative of the whole. This definition of legitimacy cannot be arrived at by merely implementing a standard of justification, since such standards on the basis of value systems. Therefore, the answer lies in locating a procedural mechanism that can bring about legitimacy through representation, and hence, non-domination (Pettit 2012: 10). This is precisely what is provided by incompletely theorized agreement.
Incompletely theorized agreements also have the benefit of providing a framework of coexistence and, as Sunstein argues, a demonstration of mutual respect (2000: 124). Coexistence, being a state of relative harmony despite competing value systems, is attained by the shift to consensus outcomes. As such decisions are the joint decision of at least a majority, then most people will accept it and not claim to be arbitrarily interfered with by another group. This offers social stability by not aggravating division or highlighting difference, and generates unity through joint decision.
Moreover, incompletely theorized agreement is feasible in practice for two reasons. First of all, it avoids placing demands for a standard of public reason, which would require individuals and groups to conform to only using certain types of justifications or reasons. By doing so, it would filter out some kinds of justifications, such as those that are religious, divergent, or otherwise do not meet the epistemic or substantive standard of appropriate public reason. Such requirements seem unlikely to be feasible, because individuals holding to different value systems (e.g., worldviews, religious beliefs, ideologies) are unlikely to change their form of reasoning merely for the sake that they are metaphysical, epistemically weak, or otherwise inconsistent.
Secondly, incompletely theorized agreements place a premium on efficiency. Value pluralism will necessarily limit agreement on many issues, which can be avoided by focusing on the consensus in outcomes and practical arrangements. This allows important disagreements to be settled, instead of being endlessly debated without prospect for resolution. That is not to say that disagreements on justification are unimportant. Here, Sunstein’s epistemological observation about agreement in society is pertinent: “People can know that X is true without entirely knowing why X is true” (2000: 119). Just like in the case of the society S1, each group knows that murder should be illegal, but cannot ultimately know why this is the case.
Objections to incompletely theorized agreements
I now turn to four possible objections to incompletely theorized agreements. Firstly, since it does not address differences in justification, it runs the risk of solidifying and even promoting strongly diverging and potentially harmful value systems. This is because incompletely theorized agreements implicitly affirm the equal validity thesis. However, the focus on an outcome on which there is agreement would likely inhibit rather than aggravate divisions between groups, since it is unifying and offers groups mutual respect in their justifications. Secondly, although incompletely theorized agreements do not focus on a standard of justification, it does not mean that discussion and debate on these issues will not take place societally. Moreover, it is not incompatible with attempts to find the best forms of justifications and reasons within other domains of discourse.
A second objection would be that incompletely theorized agreements become infeasible if intense divisions in a pluralistic society make it impossible to even reach agreement on outcomes. In response, if divisions in a pluralistic society are already so stark that there can be no agreement on justification or outcomes, then it is at an ultimate impasse: absolute disagreement without prospect for any resolution. To this, there does not appear to be any obvious solution within any conception of democracy, since there needs to be some form of agreement for it to function.
A third objection might be that the account of legitimacy outlined is circular, in that incompletely theorized agreements is legitimate only because the procedure, being ITA, is necessarily legitimate by definition. The account of incompletely theorized agreements is thus artificially self-validating. However, this is incorrect. My account of legitimacy is derived from the democratic ideal: namely, that a decision should be the result of maximal agreement. It is because such a decision is representative and necessarily non-arbitrary that it is legitimate, not because it itself merely happens to align with incompletely theorized agreements.
Lastly, incompletely theorized agreements would suggest that a wrong or even harmful agreement would be legitimate, as it does not stipulate a criterion of justification. This is true and unavoidable if societies themselves do not regulate the justifications they use as public reason. However, it is unlikely that a strict standard of public reason could do much better, because individuals can still hold nonpublic reasons privately and derive public one from these.
Conclusion
I have argued that incompletely theorized agreements offer a minimalist and plausible framework to accommodate the need for legitimacy, coexistence, and feasibility within pluralistic societies. Pluralistic societies face the problem of disagreement due to having multiple value systems. For some disagreements, it is especially important to reach some form of agreement, as it would create a better state of affairs. Incompletely theorized agreements have the virtue of conforming to a democracy-based conception of legitimacy, avoid being dominating, enable coexistence, and appear feasible in practice. However, whether this argument is empirically valid remains to be seen, but, nonetheless, it may very well be since incompletely theorized agreements have a modicum of feasibility.
In this paper, I argue that global inequality matters only do to instrumental reasons, which are limited to severe suffering. My argument is divided into five sections. In section I, highlight the moral significance of inequality and infer a normative obligation from it. In section II, I establish the conditions on which the moral significance of global inequality is contingent. In section III, I argue against the intrinsic value of equality by using Derek Parfit’s Levelling Down Objection. In section IV, I make the case for an account of inequality that is narrowly instrumental by modifying Harry Frankfurt’s sufficiency view. In section V, I show that morally relevant inequalities are also relevant in the global context, except when an obligation to address them is overridden by feasibility, knowability, or counter-productivity constraints. Throughout this paper, I address possible objections to each point and demonstrate why this account remains nonetheless persuasive. Lastly, I offer a brief summary of my argument as a conclusion.
I. What does it mean for inequality to matter?
Broadly speaking, equality refers to a balance between individuals in opportunities, outcome, and welfare. However, such inequalities matter only insofar as they are morally relevant. Although appearing circular, the purpose of this claim is to limit the scope of the term ‘matter’. Inequalities may matter in other ways that are morally irrelevant, but I define it strictly to mean morally relevant for sake of clarity. I take this to be an uncontroversial move.
If they are morally relevant, this means that there exists a normative obligation to correct them. If a state of affairs has moral relevance, there is something bad about it that requires something of us in the sense of obligation. Therefore, morally relevant inequalities demand that we ought to correct them.
II. Conditions for the moral significance of global inequality
Global inequality is morally relevant iff conditions A, B, and C hold.
A Inequality is morally relevant;
B This relevance extends into the global context; and
C This obligation is not overridden by a more crucial factor.
I first address condition A: is inequality morally relevant? In answering this question, there are two routes of argumentation. As Charles Beitz points out, global inequality can be bad because it is either intrinsically bad whereby we have a direct moral reason or instrumentally bad whereby we have a derivative moral reason (2001: 97). This same dichotomy can be looked at in the non-global context.
III. Why equality is not intrinsically valuable
I argue that inequality is not intrinsically bad because there is good reason to think it is not valuable as a thing in itself. Consider the Levelling Down Objection introduced by Derek Parfit in his critique of a wide version of the Telic View (1997: 211). If inequality is intrinsically bad, then it would better if individuals were equal and less well-off rather than better off in absolute terms. For instance, take the hypothetical scenario of two societies, society 1 (S1) and society 2 (S2), in which the welfare of three equally portioned groups is quantified.
S1 Group 1 Group 2 Group 3
Welfare 100 50 25
S2 Group 1 Group 2 Group 3
Welfare 25 25 25
According to the intrinsic account of equality, there is something inherently bad about the inequality in S1, which makes S2 the morally desirable scenario. However, this conclusion is strongly counterintuitive, since individuals are much better off in absolute terms in S1. The question remains, what possible value is offered by S2 over S1 to make a better state of affairs?
One could alternatively consider a different scenario in which one group is super rich and another is merely very rich. It is intuitive to think that such inequality is not inherently bad or necessarily even problematic, suggesting that the intrinsic value account of equality faces an absurdity problem. Therefore, as there is a strong counterintuitive objection, I consider this sufficient reason to think that equality is not intrinsically valuable, but instrumentally valuable.
IV. Instrumental account of inequality: a view from sufficiency
To argue that inequality is bad for instrumental reasons is to say that it is bad because it has negative consequences. So what are these instrumental reasons? I adopt a modified variant of Harry Frankfurt’s Sufficiency View as having the strongest reasons for thinking particular inequalities are bad. He writes that “[e]quality may often have […] a very considerable indirect or derivative moral importance” (Frankfurt 2000: 89). Specifically, its moral relevance is derived from the inequality of those in severe conditions of suffering, such as chronic hunger, illness, pain, want, and other forms of extreme deprivation (ibid. 94). The existence of such conditions when there exist the means to correct them is an injustice, since these often result in prolonged suffering and even death. It is these conditions of severe suffering (i.e. insufficiency) that make inequality morally relevant, which in turn generates a normative obligation.
Frankfurt’s description of derivative reasons is nonetheless vague, since it alludes to rather than establishes clearly what counts as insufficiency: “treating people arbitrarily entails a failure to respect what they truly are” (Frankfurt 2000: 102). A better approach is empirical rather than theoretical. Suffering globally appears to have four dimensions of deprivation: shelter, nutrition, health, and basic education. The causes of these may be structural, circumstantial, gender-specific, or income related. It is these four dimensions of suffering that are most pressing because they are the most fundamental aspects of a basic existence. If an individual is deprived of any combination of these, they are at risk of prolonged suffering, subject to a minimal existence, and possibly risk early mortality. Therefore, sufficiency is having enough access and resources to guarantee a basic existence.
A critic might object to this ‘minimalist’ nature of this account of morally relevant inequality. For instance, Simon Caney argues that the ideal of equality of opportunity should be applied globally so that individuals have the same opportunity to achieve an outcome, regardless of nationality, ethnicity, or religion (2001: 114). However, for reasons articulated by David Miller, this principle does not offer a way to address the inequality of opportunities or resources in a way that is culturally neutral, which he argues is “indispensable in a global principle of justice” (2008: chapter 3). It is this objection that my account of sufficiency can avoid, since the four dimensions mentioned are universally true of human existence in the modern day. That is, they are existentially significant dimensions of human life everywhere. The degree to which these dimensions are morally relevant depends on how these manifest themselves comparatively inside different cultures. For example, the minimal basic education requirement in the U.K. would be very different to that of Papua New Guinea, but nonetheless there needs to be some education provided. If we adopted a more maximalist account of morally relevant inequalities, the argument would become weaker since the demand is less urgent, more complex to address, and their forcefulness less acute.
Another objection might be that this sufficiency view trivializes inequality, since it does not give it much instrumental value. I disagree with this assessment, because the sufficiency view only makes sense from a comparative standpoint of inequality. If all were equally suffering, as in scenario S2 in section III, then there would no longer be a morally relevant circumstance that could be addressed.
V. Why inequalities matter in the global context
So does the moral relevance of such suffering extend into the global context as per condition B? Arguably, if there exists an obligation to correct an inequality that is morally relevant, then there are only three factors that could override an obligation (condition C): minimal feasibility, knowability, and counter-productivity. The extent to which it is feasible is morally relevant only insofar as the requirement is not and cannot even be made achievable. For instance, suppose a human space colony was in danger of being completely destroyed if it is not evacuated soon. If we didn’t have a spacecraft to reach them, nor the ability to procure one in time, our obligation would be overridden by feasibility constraints.
Nor could we demand a moral obligation if we did not have knowledge of the fact, since ignorance would prevent the danger from being realizable. However, this would no longer apply if it was a matter of strategic ignorance, in that one willfully avoids knowing about a morally relevant circumstance in order to avoid having an obligation to address it.
Lastly, in some circumstances, attempting to address severe suffering may generate more suffering and thus be counter-productive. For example, the wealthy nation X might have an obligation to aid the suffering in poor nation Y, but if this triggers a conflict that makes the suffering greater in the short and long term, then it overrides the obligation as well.
Beyond these three factors, no other factor in the global context seems relevant to override the obligation to address the particular inequalities according to the Sufficiency View.
VI. Conclusion
In summary, my argument is that global inequality is morally relevant if and only if inequality is morally relevant, this relevance extends into the global context, and the obligation is not overridden by a limiting factor. A sufficiency view of inequality is best equipped to fulfill these conditions, since it focuses on the types of suffering that are the most urgent, the most severe, and which generate a moral obligation that is acute. In addition, the focus on universally relevant characteristics of suffering avoids the problem of cultural bias faced by more expansive accounts of morally relevant global inequalities. Therefore, this account offers the strongest possible argument for establishing an obligation to address global inequality.