Problems of Political Moralism

‘Political realism’, as Bernard Williams has preferred to call it, is indeed far from a unified political theory. It consists of a rich and pluralist group of political theorists from diverging quarters. Still, they share enough features to represent a relatively clear dissenting position to the liberal idealist theory represented by, for example, Rawls and Dworkin. They also share enough assumptions to represent an alternative approach to political legitimacy.Footnote ⁵ In brief, political realists unite in two general points of criticisms. The first involves problems of action-guidance and feasibility. In line with non-ideal theorists in the debate on justice, they doubt whether the utopianism of political moralism has any practical import and whether it is of any relevance for political theory. The second criticism concerns conceptual and normative questions pertaining to political legitimacy. These points are certainly deeply connected for political realists, since a defensible theory of political legitimacy (in their view) would have to be feasible and take seriously the empirical conditions under which politics is pursued. However, this article focuses on the second point and hence on what political realists have to offer in terms of a normative theory of political legitimacy (we will say more about this restriction below). It will only bring up criticism related to the first point in cases in which it is relevant to do so.

The core of the second critique is the charge that political moralism cannot offer a coherent and plausible theory of political legitimacy because it does not regard the political as an autonomous domain, but rather as an arena for the application of moral principles. Since it is anchored to an ethics first premise, which prioritizes morality over politics, it is argued that political moralism ‘reduces political normativity to morality’Footnote ⁶ and represents a ‘desire to evade, displace, or escape from politics’.Footnote ⁷ According to realists, this ethics first premise has several unfortunate implications for how political legitimacy is conceptualized.

First, political legitimacy is seen as a pre-political concept in the sense that it assumes that significant consensus must be in place ‘before politics as an activity can, properly speaking, begin’.Footnote ⁸ In other words, pre-political abstract moral principles are allegedly applied to the practice of politics.Footnote ⁹ In Williams’ view, political moralists assume the moral to be prior and external to the political, either expressed in terms of an ‘enactment model’ – according to which political institutions and actions are intended to express moral principles, ideals and values (the prime example being utilitarianism); or a ‘structural model’, according to which such principles offer external constraints on what politics can rightfully do (the prime example being Rawls’ theory of justice).Footnote ¹⁰ In both models, political theory starts ‘from outside politics, whereby the demands of morality give content to the principles of cooperation or legitimate political action’.Footnote ¹¹

Realists argue that political legitimacy on a political moralist account is flawed because it does not take into consideration that consensus cannot be achieved through a philosophical argument that builds moral premises into a stylized choice situation.Footnote ¹² As an ‘artifact’ of politics, consensus cannot exist before the political and thus be a precondition of politics. Rather, it is only by employing political power that general agreement of any sort is possible and achievable. Rejecting the ethics first premise, political realists argue that the normative sources of political legitimacy must be sought solely within politics rather than from an external moral standpoint.Footnote ¹³ Above all, since politics is inherently an exercise of collective decision making, realists cast doubt on the assumption that individual moral norms carry over to collective political action.Footnote ¹⁴ Not only do public decisions differ from private ones in numerous respects; politics is most of the time a dirty business in which ‘we desire principled politicians but expect – even oblige – them to commit unprincipled acts’.Footnote ¹⁵

A second implication of the ethics first premise, realists argue, is that it leads to depoliticization, that is, to a view of the political domain as deprived of any real politics and unable to offer any ‘remotely plausible account of political agency’.Footnote ¹⁶ In Raymond Geuss’ words, ‘ethics is dead politics’.Footnote ¹⁷ By regarding political legitimacy as a pre-political concept, liberal idealists attempt to depoliticize and sanitize real-world politics from the outside by eliminating ‘political sleights of hand from collective decision-making’.Footnote ¹⁸ But this cannot be right, realists complain. Since politics concerns public reasons, there cannot be any general way in which to construct political agency from individual reasons for action.Footnote ¹⁹

A Political Realist Account of Political Legitimacy

Against the backdrop of these criticisms, realists recommend shifting the agenda of political theory away from debating minor revisions of Rawls’ difference principle towards focusing on the contexts and processes through which a citizenry addresses shared problems.Footnote ²⁰ To begin with, Williams stresses, the fact that realists aim to elaborate political legitimacy from normative sources in the political domain does not imply that politics is amoral or immoral. It simply means that the appropriate standards of evaluation arise from within the political rather than from an external moral standpoint.Footnote ²¹ While no direct appeal is made to ‘moral norms of a pre-political nature’, a ‘code of political ethics’ might still exist.Footnote ²²

Moreover, instead of focusing on consensus, political realists start from a notion of disagreement and commonly see ineliminable conflict as the essence of politics. From this standpoint, politics is foremost and unavoidably about trying to achieve some sort of social order. According to Williams, legitimacy is a matter of responding to the ‘first’ political question, which is concerned with securing order and preventing chaos. Peace, stability and security must be ensured first, since they serve as a pre-condition for any other political objectives, such as justice or fairness.Footnote ²³ While this does not require political realists to presume that people must necessarily see peace and security as supreme and overriding goods, they are fundamental goods for politics since they are an essential condition for achieving almost any other goods.Footnote ²⁴ In other words, a basic level of social order must be maintained that is at least sufficient to enable persons to live minimally worthwhile lives and secure the conditions of co-operation.Footnote ²⁵

That said, following Williams’ axiom ‘might is not right’, not all social orders that achieve this can be considered legitimate. For example, ‘a peaceful situation preserved entirely through suppression or tyranny, or at the extreme by the exclusion or even elimination of those who do not accept it’, is not legitimate.Footnote ²⁶ To distinguish legitimate from non-legitimate power, realists commonly anchor political legitimacy in two conditions, which we here call ‘the agreement condition’ and ‘the autonomy condition’. Concerning the agreement condition, Horton claims that for an arrangement to be legitimate it has to be ‘broadly “acceptable” or “agreeable” to those who are party to it’.Footnote ²⁷ As long as such acceptance ‘is not the product of clear, willful, systematic and comprehensive deception by those with political power, the reasons why it is accepted can be many and various’.Footnote ²⁸ Similarly, Newey argues that political legitimacy requires an agreement that is ‘thought of as free’. However, he accepts that political reasons, such as reasons for agreement, are often artifacts of the political process.Footnote ²⁹ Bellamy also argues that legitimacy requires ‘some degree of willing consent on the part of the ruled’.Footnote ³⁰

The autonomy condition is to be understood in a weak sense, defining the subjects who are supposed to reach this agreement (or perceive it as such) and express some form of ‘self-determination’ or ‘self-authorization’ on behalf of those who are involved in, or are part of, the social order in question. Being part of the social order is defined in terms of subjectedness, such that the parties ‘subject to’ the social order or political authority are those who are supposed to agree to itFootnote ³¹ or be offered justification in terms of sufficient reasons for accepting it.Footnote ³² In Horton's view, for example, a political order meets the conditions of legitimacy if it is ‘broadly acceptable to those subject to it’.Footnote ³³

In addition to these shared features, Williams adds further components that seem to make his realist account more universalist. While he stresses that any response to the first question of politics is a temporary solution, the paradigmatic rejection of slavery has universal applicability since the mere circumstance of some subjects being de facto ruled by others is ‘no legitimation of their being radically disadvantaged’.Footnote ³⁴ In Williams’ view, it is deductively true from the axiom ‘might is not right’ that some justifications of power are simply a violation of universal (human) rights, the latter of which can be derived from the ‘universal materials of politics’.Footnote ³⁵

The Normative Wiggle Room of Political Realism

As mentioned above, our analysis is directed at the normative aspect of realism, and leaves the extensively debated questions of feasibility and action-guidance aside. We do not deny that feasibility is a central aspect of realist accounts of political legitimacy. Indeed it is. However, although feasibility sets constraints on the normative demands of a society, this article analyzes the normative claims made by realists within these limits, whatever they are.

As we have seen, instead of relying on prior moral and idealized assumptions, realist legitimacy responds to the historical context of a political arrangement or polity.Footnote ³⁶ At the same time, it relies on normative conditions, mainly on what we call the agreement condition and the autonomy condition, which are meant to ensure that not all forms of agreements count as legitimate. In order for a political order to be legitimate, the agreement (Horton) or willing consent (Bellamy) must be perceived as free (Newey), and thus cannot rely on means that are too tyrannical (Horton), be coerced (Williams) or be a result of total deception (Horton). The question is, in light of the realist emphasis on the actual circumstances of politics, how are these normative restrictions established? Basically, either the two conditions of legitimacy are established by actual political processes, or they are not. If they are indeed established by actual processes, such that a political order is legitimate insofar as it is in accordance with the process in question, we have started on a regress of justification. Because what if this process is due to a too tyrannical order? If all we do is adhere to a new actual (meta) process or arrangement, we are merely pushing the justificatory question one more step, not responding to it. Consequently, realists are well aware that the only way of out this regress is to acknowledge that not all sources of legitimacy are due to the actual political context in which they are formulated. Legitimacy in this sense transcends actual politics, as it were.Footnote ³⁷

There is a sense, of course, in which all normative principles that have ever been formulated are due to actual processes, namely in the sense that they all, as a matter of physical necessity, have been formed through an historical process. Thus there is no other way than within actual politics (in the broadest sense of the term) to formulate, question and argue for what the justificatory sources of legitimacy are. But this must be kept separate from the claim, rejected by realists, that actual processes decide the conditions of legitimacy. As history has taught us, actual processes come in all flavors. They have resulted not only in the values that realists endorse but also, for example, that ‘might is right’, the idea that some people are worth more than others, that birth constitutes a legitimate ground for ruling and so on. Paying close attention to context, in particular to the limits set by how politics works, realists consider the (historically formulated) available norms. Moreover, similar to any other normative account of political legitimacy, they pick out some and reject others when formulating and justifying their notion of legitimacy. It is only in this way that realists become equipped to say that some actual rule making, or some actual authority, is simply not legitimate (for example, by being too tyrannical). In this sense, the realist notion of legitimacy transcends actual politics.

One possible objection to this line of reasoning would be that political realists are supposed to be interpreted as putting forward a relativist position such that legitimacy stems from the convictions of the public and nothing else. However even if such relativist claims are made, such as when Bellamy says that ‘[u]ltimately, the source of public reasons is the public themselves’,Footnote ³⁸ this interpretation is not overall plausible in view of the commitment made by the same theorist (and by political realists generally) to the restrictions on how such beliefs may be caused in order to constitute legitimacy. The prohibition of agreement by coercion, deception and tyranny is de facto an endorsement of sources of normativity that transcend actual politics. From this we infer that political realism generally is to be interpreted as a normative project that, as Philp says, ‘provides the basis for a contextualist but non-relativist account of “what is to be done”’.Footnote ³⁹

Another objection might be that the normative restrictions on legitimate agreement on the realist account are supposed to be understood as being grounded merely in the empirical conditions of the political community as such – that is, in terms of conditions that must be fulfilled in order for the community to persist. On some realist conceptions, such as Williams’, it is clear from the start that legitimacy is more than this: ‘the power of coercion offered simply as the power of coercion cannot justify its own use’.Footnote ⁴⁰ In a similar vein, Galston argues that the realist emphasis on context does not imply ‘that the concept of universal principles makes no sense; basic human rights exist and apply to all’.Footnote ⁴¹ On other conceptions, however, there seems to be a certain ambivalence concerning the status of the normativity. Horton, for example, claims that ‘political theory can only tell it as it is, mapping rather than seeking to adjudicate rival perspectives’.Footnote ⁴² But since his theory relies on a notion of agreement that shares with other realists similar normative restrictions on what is acceptable for legitimacy – such that it has to be broadly ‘acceptable’ or ‘agreeable’ to those who are party to itFootnote ⁴³ – he cannot avoid the transcendent normativity that he seems reluctant to embrace. Surely, there are other possible arrangements if we were merely interested in the persistence conditions of a political community.

We infer that the realist conception of legitimacy is best interpreted as resting on a source of normativity that is dependent on, but still transcends, the political processes. While realists stress the contextual factors of real politics, such that ‘reasons for acting politically… are largely artefacts of the political culture’,Footnote ⁴⁴ they still endorse that not all reasons are so constituted.Footnote ⁴⁵ Thus they adhere to values or principles that have priority to and transcend the actual political process, in order to distinguish between (1) fulfilled empirical conditions for, say, peace and security to live worthwhile lives (not legitimate) and (2) empirical conditions for peace and security to live worthwhile lives that are fulfilled in an accepted way (legitimate).

Indeed, this is in accordance with what is commonly seen as a truism about normativity in the philosophical literature, namely, that it is always in principle possible for a person, group or community to be wrong about what should be done.Footnote ⁴⁶ Regardless of how empirically informed we are, and how contextual a criterion of legitimacy may be, a realist account that equates normative criteria of legitimacy with people's beliefs about what is legitimate shoulders a heavy justificatory burden.Footnote ⁴⁷

We next turn to the second question about the source of normativity, viz. how these realist transcendental (grounding) values should be characterized in order to analyze how the relationship between the moral and political domains is construed by political realists.

Flawed Conceptions of the Ethics First Premise

Does the fact that political realists’ values transcend actual politics entail that realists must also subscribe to the ethics first premise? Thomas Hurka, in his review of Geuss’ Philosophy and Real Politics, believes so.Footnote ⁴⁸ In what bears structural similarities to our argument above, he reasons that when justifying every situation-dependent value, there is another value that justifies this value; we eventually reach a value that is not contextual but holds universally, which means to ‘revert to the ethics-first view’.Footnote ⁴⁹

These structural similarities notwithstanding, Hurka puts forward a different, much more controversial argument. Even if we accept this argument, his conclusion about an ethics first view does not follow. Our claim in the previous subsection was that the realist conception of legitimacy fundamentally rests on values that transcend the actual processes of the political arena. But we have not yet said anything about the form that these values must take. Hurka makes the stronger claim that they have to be in the form of universal principles, and concludes from this that realists must accept an ethics first view. But this is a question-begging argument, since it starts out by assuming that realists (or at least Geuss) deny that there are any generally true principles and goes on to state that underneath every contextual principle there has to be a general one. In the present context, there is an even more crucial problem with Hurka's argument: it simply assumes that any universal value or principle is moral. As we saw in the previous subsection, political realists typically do not deny grounding principles; on the contrary, they are committed to the agreement and autonomy conditions of legitimacy. So they would deny particularism for political legitimacy and instead ‘use a principle that doesn’′t need special justification in this society because it holds universally’.Footnote ⁵⁰ However, they could still insist that these principles are political and not moral – but the latter is precisely what the principle-denying theoretician would resist.Footnote ⁵¹

As will be shown in the next subsection, political realists do in fact rely on the ethics first premise and thus prioritize morality over politics. However, we will argue that they have misconstrued what this priority consists of. Before turning to our argumentation, we should first say a few words about the notions of the political and moral domains on the realist conception. To our knowledge, political realists have given little in the way of an unambiguous characterization of the two domains. Indeed, they sometimes even argue that we should not even try, since ‘disputes over the extension of “politics” are themselves irreducibly political’.Footnote ⁵² Still, in practice they seem to rely on a typical understanding of the two domains. According to this view, the moral (or ethical)Footnote ⁵³ domain is primarily about how we, as individuals, ought to live our lives. Moral values and principles are those that justify our actions, beliefs and intentions, distinguishing what is right from wrong, or good from bad. Such moral values and principles may be structural, such as that of equal consideration for everyone,Footnote ⁵⁴ or they may be more substantial, such as W. D. Ross’ duty of non-maleficence.Footnote ⁵⁵ The political domain, on the other hand, concerns how we ought to live our lives on the societal level. According to realists, politics is primarily about collective action and collective decision making pertaining to governmental bodies and institutions, and consists of social relations involving power and authority.Footnote ⁵⁶ Typical political values and principles are collective self-determination, political equality and the principle of inclusion. Naturally, all characterizations of the moral and political domains are bound to be controversial, but since our arguments are directed at political realism, we will accept this realist reading of the two domains unless explicitly stated otherwise (see below for further discussion).

Let us now turn to the ethics first premise. Political realists claim that this premise entails that morality is ‘prior’ or ‘external’ to politics. Matt Sleat provides a typical example of how the ethics first premise is conceptualized:

Moreover, realists often perceive such ‘prior moral stipulations’ as fixed and static, in comparison to the dynamic nature of politics. Newey, for example, argues that ‘[l]iberal political design takes morality as fixed, and politics as negotiable’.Footnote ⁵⁸ In contrast, legitimacy on the realist account ‘must be generated from within politics itself rather than reflect the demands of an external moral requirement’.Footnote ⁵⁹

To begin with, a dichotomy of morality as fixed and pre-given and politics as dynamic and negotiable does not hold, whether we refer to the practices or the content in question. Interpreted as practices of morality/politics, the several thousand-year-old debate in moral philosophy bears witness to the fact that morality is just as ‘negotiable’ as politics. Indeed, influential moral philosophers have long taken the moral domain to be one of ‘essentially contestable’ claims.Footnote ⁶⁰ However, if we interpret them as justificatory principles or claims themselves, they (as we have demonstrated above) pace the contextual factors of politics, are also essentially fixed on the realist conception of the political. In other words, in both the moral and the political domains, we justify our actions or processes with reference to principles and claims that we treat as fixed, even if the content of these fixed points is always under debate.

The interpretation of morality as ‘prior to’ politics is equally hard to get off the ground. If what realists refer to here is epistemological or conceptual priority, it seems to be too strong a claim. We seem to be able to gain knowledge of – hence understand – what the political domain is without relying on prior knowledge of the moral domain. Similarly, our political concepts do not seem to presuppose moral ones; indeed, we seem perfectly able to understand the political domain through our understanding of collective action, the processes of actual politics and so on – aspects that are all internal to the political domain.

Another interpretation of ‘prior to’ that fits with realist claims about ‘external’ to or ‘outside’ of politics is a temporal, presumably causal, relationship. But the idea that moral concerns always come first, and somehow by themselves cause political concerns, seems equally problematic. A question of whether a decision is politically legitimate would, from this perspective, start with a moral concern. But this interpretation seems just as implausible as the conceptual claim. Although there are plenty of cases in which moral concerns cause political concerns, such as when infringements on people's ownership rights cause them to call for political action (and legislation), in many other cases political concerns are the initiating force, such as when people encounter collective action problems. Moreover, even when moral considerations cause political action, why could it not be in the very context of politics that these considerations become salient? Although this is of course controversial, the above example of infringements of the right to private property may be a morally relevant fact only due to its political context, rather than because it has any independent or prior force. A moral consideration may, it seems to us, be ‘generated from within politics itself’ and still reflect the demands of a moral requirement. There is no reason (apart from a rhetorical one) to claim that this requirement is somehow ‘external’.

To conclude, the realist claim that the ethics first view would entail that the moral domain is prior or external to the political domain – either conceptually, epistemically, temporally or casually – builds on a misconstrual of the relationship between the moral and political domains. However, realists fail to address at least one notion of priority. While explanatory priority may refer to the sort of epistemological or conceptual priority that we have already dismissed, it may also be interpreted as a justificatory notion. We call this notion justificatory priority, and will argue below that morality does indeed have this priority to politics on any plausible notion of political legitimacy.

The Justificatory Priority of Morality

In order to show that political realists seem to – and must – subscribe to the ethics first premise understood as a justificatory priority, let us return to the agreement and autonomy conditions of political legitimacy.

Evidently, political realists do not wish to reduce legitimacy to any social order that is stable and secure. The normative work of the agreement and autonomy conditions is meant to distinguish legitimate (acceptable) stability and security from non-legitimate (unacceptable) stability and security. However, it is not self-evident (or an inherent feature of political decision procedures and processes) that those who are subject to political authority, in the sense of being governed by it, would have to agree on it for acceptability.Footnote ⁶¹ From the point of view of acceptable stability and security, there are many alternative versions of normative conditions. For example, why not instead let citizens who contribute most to stability have twice as much to say in the authorization (acceptance) of political power, and prohibit those who contribute nothing or very little from having a say as long as they benefit from this social order?

Based on our analysis, the only plausible explanation for why realists stick to the agreement condition as a prerequisite of acceptability is that they ultimately ground political legitimacy in an egalitarian premise of some sort, for example ‘equal acceptability’. But such a premise is constituted by the moral values of freedom and equality, regardless of whether these values (and the way they are integrated into the account of political legitimacy) are part of a political morality – that is, a morality that emerges in the political domain – or are in some sense ‘before’ or ‘external’ to it.Footnote ⁶²

The same applies to the autonomy condition, which is expressed in terms of ‘equal subjectedness’. Here realists rely on a version of what is sometimes referred to as the ‘all subjected principle’ in democratic theory.Footnote ⁶³ The basic structure of this principle is egalitarian, such that those who are subjected to an authority's laws should have an equal say of some sort. Indeed, even Newey's minimalist version, which merely requires an agreement that is ‘thought of as free’,Footnote ⁶⁴ presumes that every individual subject to the authority should perceive the agreement as free, rather than just a limited group of subjects within the citizenry. Even if the source of normativity in ‘free agreement follows from the conditions on collective action’,Footnote ⁶⁵ it does not follow that legitimacy requires political power to ‘be authorized by those on whose behalf the collective is taken to act’,Footnote ⁶⁶ unless some further normative premises are added.Footnote ⁶⁷

The above analysis finds that, on a closer look at the realists’ justification of political legitimacy, they all rely on some moral value(s). Hence political legitimacy is, in fact, irreducibly moral and, consequently, the moral domain does have a certain priority over the political domain in terms of a justificatory priority. Justificatory priority of morality over politics means that among the values that justify (a reasonable notion of) political legitimacy, moral values are at the core. Note that we are not saying that behind any value justifying political legitimacy there must be a moral value. Our claim is rather that, while non-moral values may play a justificatory role among the grounding values of political legitimacy, there are also moral values, which are tied either to the procedural or substantive features of political legitimacy. And this is true, we believe, not only of the accounts of current realists but of all (defensible) normative notions of political legitimacy. Even if we grant that it is a rather bold move to go from the accounts of political realists to all accounts of legitimacy, we think it is motivated. When realist accounts of legitimacy, given their aim of taking ‘political reality’ into account, also turn out to include moral values, there is simply not enough wiggle room left for a plausible account that does not include moral values of any sort. Without such values, it will always lack the sought-after normative force.

Hyperautonomy and More on the Relation between the Moral and the Political

An evaluation of the claim that the political domain would lack autonomy according to the ethics first view depends not only on what we mean by ‘the autonomy of the political’, but also on how we understand the relationship between the political and the moral. Let us start with hyperautonomy. According to this demanding notion, the normative source of political legitimacy lies exclusively within the political domain.Footnote ⁶⁸ Since we interpret the ethics first premise to entail that political legitimacy is justificatory dependent on moral values, and the ethics first premise is necessary to all reasonable accounts of political justification, it directly follows that hyperautonomy is unattainable.Footnote ⁶⁹

Note, however, that ‘unattainable’ wrongly signals that hyperautonomy would be the ideal that, unfortunately, cannot be reached. But as we have argued in the previous section, this is the very idea that realists should eschew. In order to fully appreciate this point, let us return to the distinction between the moral and the political domains, which we briefly explicated above. We concluded that the way realists use the terms, the moral domain typically concerns individual actions, whereas the political domain typically concerns collective actions and procedures pertaining to society's basic institutions. Our main point has been that this difference in focus does not entail that the normative, justificatory grounding of legitimacy can be fully separated: moral values are also necessary in the political domain. In other words, we should refrain from conceiving of the moral and the political as entirely separate spheres, which realist expressions like ‘prior to’ and ‘external to’ might delude us into believing.

An important reason why we hold on to the broad characterization of the two domains made by realists is precisely because our argument does not rely on any specific view of morality (which is why we use ‘morality’ and ‘ethics’ interchangeably). For example, we do not have to defend an individualist morality, as do liberal idealists. Also a more collectivistic approach to morality, according to which the ethico-cultural community is a prerequisite for possessing a moral language (such as communitarianism) uses the ethics first premise as a basis for construing and grounding political legitimacy.Footnote ⁷⁰ Further, our argument does not imply that the moral values that ground political legitimacy must be tied to a particular conception of justice or democracy. However, even in what Jeremy Waldron calls ‘the circumstances of politics’, which are permeated by radical disagreement over matters of justice, moral values play an essential role in justifying political legitimacy.Footnote ⁷¹

In addition, our argument does not presume any particular theory of moral justification. We have framed our normative discussion in terms of ‘moral values’ and ‘moral principles’ since, as we noted above, political realists typically do not deny grounding principles. However our argument is also compatible with doubts concerning the applicability of principles or the level of generality of principles. The fundamental claim is that the basic justification of political legitimacy includes moral considerations, whether or not it refers to general principles or values. Hence, our argument is compatible with claims that justification in the political domain need not come in the form of general principles that always hold. The basic justification of political legitimacy may be irreducibly contextual, so that we may not find a principle that always captures what is politically legitimate, no matter the situation.Footnote ⁷² We take no stand as to whether or not this is possible – our argument is merely that, even in this case, moral considerations will figure into the justification (which is why our necessity thesis is not also a ‘sufficiency thesis’). Hence our argument is also valid for those who claim that justification – when it comes in the form of principles – comes in the form of contextual principles that are inherently bound to the institutional structures to which they are supposed to apply, rather than in the form of abstract, general ones.Footnote ⁷³ As we have stressed throughout, while theorists disagree on how big a role contextual factors – such as historical traditions, narratives of a polity, held values, etc. – play in the justification of political legitimacy, our claim merely states that as long as the justification is supposed to have normative force (as in the current debate), moral values are necessary.

Basic Autonomy

Thus hyperautonomy is an implausible notion. But looking more closely at the criticism that realists actually put forward, the thinner notion of basic autonomy reveals itself through the explicit charges made against political moralism:

1. 1) that politics is construed as an arena for the realization of moral principles (for example, Galston, Rossi)

2. 2) that political normativity is reduced to morality (for example, Newey)

3. 3) that moral reasons trump all other kinds of reasons when deciding what is politically legitimate (for example, Newey)

Basic autonomy is thus the rejection of these three propositions. However, the ethics first premise is in fact compatible with basic autonomy. To demonstrate this, let us take a look at each of these propositions. The first is arguably the most rhetoric of the three, since we wonder whether any political theorist conceives of the political domain merely in these terms. Certainly Rawls, the main target of the realist critique, does not. More importantly for the present article, however, the first proposition definitely does not follow from the ethics first premise. Even if moral principles and values, such as fundamental egalitarian ones, figure into the basic justification of any normative account of political legitimacy, there are numerous other tasks that are also part of politics. And if the political domain includes institutional aspects and practical concerns that do not follow from the moral justificatory basis – which is a necessary aspect of political justification – the political domain is not merely an arena for realizing moral principles.

The second proposition takes us explicitly into the normative sphere. While normativity comes in many varieties (for example, semantic, epistemic, aesthetical), the relevant sense here is practical – that is, practical normativity primarily concerns how we should act. As explicated above, realist accounts typically distinguish between moral and political values and principles. The second charge claims that political normativity is reduced to morality in the political moralist view. Admittedly, this is a possible theoretical position. Some influential moral theorists indeed interpret the basic normative term ‘ought’ as merely expressing what we should do, all things considered.Footnote ⁷⁴ Doing so would arguably reduce all practical normativity to morality.Footnote ⁷⁵ However, this reductive view does not in any sense follow from the ethics first premise. On the contrary, most contemporary philosophers are skeptical of the possibly of reducing one domain of discourse to another, and among moral theorists it is much more common to acknowledge a variety of different types of non-moral considerations – political, prudential, altruistic, egotistical, instrumental – and to deny rather than defend the claim that all normativity is reduced to morality.Footnote ⁷⁶ Above all, we have found no textual support that political moralists would in fact embrace such a strong and controversial reduction claim.

The third proposition seems to be the only one that comes close to being incompatible with the view of some moral theorists. The idea that moral reasons trump other kinds of reasons does, indeed, have its followers.Footnote ⁷⁷ But as with the previous claims, it is far from obligatory. If anything, many moral theorists reject that moral reasons always trump other reasons.Footnote ⁷⁸ Crucially, however, whether or not moral reasons trump other kinds of reasons lacks relevance for the issue at hand. For even if we assume that moral reasons trump other kinds of reasons – so that if there is a conclusive moral reason to φ, then we should φ regardless of what other reasons (here: political) we have for doing something else – this fact does not entail that moral reasons decide what is politically legitimate. On the contrary, the trumping claim implies that political reasons are distinct from moral reasons (since otherwise political reasons would, by necessity, be aligned with moral ones). In general, we cannot expect two distinct concepts to have the same extension.Footnote ⁷⁹ Thus just as something may be a kind act even though it is not morally right, we should expect that a decision could be politically legitimate even if it is not morally right. Indeed, the very notion of justificatory dependence that we have defended opens this possibility, since we claim here that moral values are necessary (but not sufficient) for political legitimacy.

The upshot is that political reasons may trump moral reasons in political decision making, and that political outcomes may be legitimately amoral or even immoral. Whether this is so will depend on what is decided upon, what moral and other reasons are involved, and what kind of decision it is. For example, say that we ground political legitimacy in the moral value of equal worth, specifying this value in terms of the equal consideration of everyone's interest. Moral reasons pleading to this value would most likely trump political reasons used against it (for example in support of the political self-determination of a specific religious or ethnic group) if the decision concerned fundamental constitutional matters (for example basic rights of the members of the polity). However, even concerning such matters, not all moral reasons would trump, since other moral values (for example, individual freedom), might be set aside in favor of specifically political values (for example, universal suffrage), if, say, voting was mandatory.

By the same token, decisions to give financial state support to a sexist political party or allow a racist political demonstration might be justified by political reasons (for example, by valuing political competition and opposition or the freedom of assembly). These reasons are all regarded as indispensible sources of normativity in a democratic polity, even though there are strong moral reasons – from the standpoint of the equal concern of everyone's interest – to deny support and to prevent these protesters from shouting disparaging and insulting remarks about certain groups in society.

Beyond Basic Autonomy

On the one hand, realists (wrongly) accuse political moralism of failing to embrace basic autonomy. On the other, it is quite clear that realists hope for something more substantial than basic autonomy in theorizing political legitimacy. So what are realists worried about with regard to the autonomy of the political? Foremost, they seem worried that the ethics first premise leads to depoliticization in the sense that the political domain is deprived of any real political agents, and that political moralism cannot offer any plausible account of political agency. One reason for this is that since politics concerns public reasons, there is no general way to construct political agency from individual reasons for action alone.Footnote ⁸⁰ However, that the political is collectively autonomous in this way does not imply that it is not an individual political practice as well. In their eagerness to see the moral domain as the arena concerned with individual action and the political domain as the arena concerned with collective action, realists neglect that political legitimacy is premised on both individual and collective autonomy and thus involves individual and collective agency.

Acknowledging both these dimensions, together with the conclusion that basic autonomy is a property of any reasonable conception of political legitimacy, opens up considerable room for moving beyond basic autonomy in the theorizing of political legitimacy. Let us conclude by looking at how this could be done. Contrary to what realists may lead us to believe, the problem with basic autonomy is not that it leaves too little space for a substantive account of autonomy; it leaves a fairly large space. The problem is rather that this space has thus far been relatively unexplored and left empty of theoretical content. Indeed, it is here that the realist critique of political moralism hits the target, articulated for example in terms of undertheorized notions of political agency and collective self-determination.

From the point of view of our metatheoretical argument, it is possible to substantiate and justify legitimacy in numerous ways from sources of normativity specifically from within the political, since the only requirement dictated by the ethics first premise on a justificatory reading is that this is not its only source. We hold that it is precisely within the framework set by the ethics first premise that political realists can (and should) perform conceptual and normative work that takes the debate forward. In fact, not only does it lead to a dead end conceptually to go outside of this framework; by working explicitly within it, realists will also avoid setting the bar so low for political legitimacy that most countries in the world today are already above the threshold by a good margin. Needless to say, to solely require of political power that it be ‘not entirely tyrannical’ or ‘not entirely suppressive’ in order to be legitimate means to demand less of legitimacy than, say, even Zimbabweans would expect from Mugabe. We see no reason why realists would want to do that.

One possible explanation of why realists are satisfied with such a low bar is that they try as hard as they can to deny the ethics first premise. But to effectively and convincingly substantiating the notion of autonomy, realists must not only acknowledge the ethics first premise but also further unpack the necessary conditions implied by basic autonomy and thus their particular version of this premise. This would mean to make explicit, rather than rebuff, the moral values embedded in their agreement and autonomy conditions. For example, what notion of equality is involved in each of them and how does a bar as low as ‘not entirely tyrannical’ or ‘not entirely suppressive’ fare with it?

Up until today, realists have been so preoccupied with finding faults with political moralism that they have wrongly assumed that the political domain, in order to be autonomous, must somehow be separated and ‘rescued’ from morality when theorizing political legitimacy – hence the use of terms such as ‘prior to’ or ‘external to’ – instead of exploring overlapping sources of normativity. Since political legitimacy concerns institutions and authority in the political domain, and politics irreducibly concerns collective decision making, both individual and collective aspects of autonomy are central for normative reasons. Therefore, moving beyond basic autonomy would require that we explore in tandem what values (moral and other) underpin both individual reasons for action and individual agency (for example, individual freedom and equal consideration) and public reasons for action (for example, collective decision making and self-determination).