Internalist and Integrationist Theories of (Global) Justice

Abstract

The global justice debate has increasingly moved toward the analysis of concrete issues in global politics, such as trade, migration, or climate change. This raises a methodological question: should the demands of justice in these domains be theorized independently or in conjunction with one another? Integrationists have championed the latter approach, arguing that it is better suited to guide our practical judgments. In contrast, internalists maintain that each domain is governed by its own set of principles. This paper defends the plausibility of the internalist approach against integrationist challenges. By examining different interpretations of internalism, it first seeks to provide a clearer overview of the methodological dispute. It then analyzes various arguments for integrationism, showing that their implications are more limited than their proponents believe. Finally, it focuses on the question of practical guidance, highlighting the value of idealized domain-specific theorizing in guiding transitions toward just arrangements.

INTRODUCTION

In recent decades, the global justice debate has moved from the abstract to the concrete. Where once much of the literature focused on the scope of theories developed originally in the context of the state, a growing number of theorists have turned their attention to various applied sub-fields of global politics. Relatively detached yet empirically sophisticated debates have emerged around the global arrangements that govern international trade, finance, climate change, migration, or human rights. This might be said to reflect a turn from “vertical” to “horizontal” approaches in theory-building: rather than deductively applying a general account of justice to a given area, the analysis starts by exploring the moral and empirical questions within it and drawing out the implications for justice—both domestic and global—from the bottom up (Beitz 2019).

With this shift, an important methodological question is brought to the fore: how, if at all, do normative theories within a given area or domain relate to wider considerations of justice? On one possible view, these theories are best developed in relative isolation, reflecting the particularities of the domain to which they apply. On the competing view, they should be developed in conjunction with one another so that the demands of justice in trade, for example, are linked to other normative concerns surrounding the environment, migration, or human rights. While this does not rule out focusing on individual domains to uncover the specific moral issues they raise, it presupposes a general background theory that places domain-specific considerations within a more comprehensive account of what justice, in a larger sense, demands. Following Simon Caney, this view is typically referred to as “integrationism” (Caney 2012, 259). Developed initially in the context of climate justice, it has also been invoked in domains such as international trade, natural resources, or migration (Armstrong 2017; Walton 2020; Yong 2016). As a method of theory-building, its attractions are clear: by taking a broader view of the moral terrain, it promises to avoid myopic principles that neglect the ways in which injustices in different domains interact. Bracketing environmental questions in a theory of international trade, for example, might yield flawed principles whose implementation would not bring about just outcomes. According to its proponents, then, the integrationist method is better suited to guide our practical judgments.

Integrationism can be contrasted with a pluralistic view of (global) justice, according to which separate domains are independently governed by “internal” principles.¹ My primary aim in this essay is to defend the plausibility of this view against integrationist challenges. Specifically, I argue that integrationists have not shown that distributive principles for a given domain must be derived from a higher-order theory, that all domains of justice must be theorized together, or that interlinkages between domains make it impossible to define the boundaries of any given domain. By distinguishing more clearly among different internalist approaches, I also seek to provide a fuller picture of their respective methodological implications. Thus, a first contribution of the essay is to lay out more systematically different claims on both sides of the debate and to assess their strength. Building on this, I advance an argument in favor of internalism that turns the integrationist claim about practical guidance on its head. By drawing on the distinction between idealistic and realistic theorizing, I argue that seemingly implausible practical judgments may follow from an idealized form of domain-specific theorizing that can help provide a fuller picture of how to guide transitions toward just arrangements. At least in one respect, this approach provides guidance where integrationism risks being indeterminate: since there may be many compossible domain-specific principles that bring about the same desired higher-order distributive ideals or values, we require a way to choose among them.

The argument unfolds as follows: in the next section, I introduce the idea of internalism, distinguishing among social interpretivist, teleological interpretivist, and constructivist versions of the approach. I then discuss the arguments put forward in support of integrationism, focusing on the substitutability of goods, concerns about practical guidance, and the difficulty of drawing boundaries between domains. I show that these arguments do not provide direct support for integrationism, that they underestimate the capacity of internalism to appeal to broader moral considerations from within, and that they overstate the difficulty of demarcating domains. The following section returns to the question of practical guidance. Even when internalism does not provide full practical guidance, I outline the value of more idealized forms of theorizing on account of their role in specifying the demands of justice and guiding transitions toward their implementation. The final section concludes.

Internalism

A domain, as I shall understand it, is the site or sphere within which specific considerations of justice apply. At least since Plato’s Republic, one central domain of justice has been the state.² In contemporary political theory, this concern is epitomized by John Rawls (1999a, 6), who famously took the subject of justice to be the “basic structure of society”—that is, the major social and political institutions that distribute the benefits and burdens of a relatively closed social system. But just as Rawls became the undisputed reference point in debates about justice, attention began to shift beyond the traditional boundaries. The early global justice debate was dominated by the disagreement between global egalitarians who argued for a single global domain of justice and statists who held fast to the focus on bounded societies. If this suggested a form of monism about the domain of justice on both sides of the debate,³ a “third wave” of global justice theorists has attempted to provide a more pluralistic and nuanced picture of the distinct justice-generating phenomena in the global realm (Cohen and Sabel 2006; De Bres 2012; James 2012; Miller 2010; Risse 2012; Valentini 2011a). These phenomena include formal and informal institutions, regime complexes, or social practices that transcend the borders of any single country. Apart from the state, then, other potential domains of justice may be found in the global arrangements that govern policy areas such as trade, migration, human rights, and so on.

But what normative principles apply within each domain? On one possible view, these principles can be derived from a higher-order theory whose formulation and validity are independent from the domain to which it is applied. Utilitarianism, global luck egalitarianism, and some forms of libertarianism are examples of such a view because they treat facts about domains as only instrumentally relevant for the implementation of more general values or distributive ideals. The alternative view, on which I want to focus here, assumes that domains themselves play a constitutive role in justifying the relevant principles. While this approach may also appeal to general values and considerations of political morality (welfare, equality, liberty etc.), the resulting principles are “internal” because they are an expression of the specific social relationships within the domain.

Internalism: normative principles are justified not just for a given domain, but also in virtue of its existence and nature.⁴

Rawls (1999a, 25) expresses this idea memorably when he states that “the right regulative principle for a thing depends on the nature of that thing.” The nature of a domain, in turn, is shaped by the social relationships that comprise it—in this sense, internalist theories are relational theories (on the latter, see Sangiovanni 2007, 5; 2016). It follows that the formulation of theories of justice must proceed by analyzing the particularities of the domain in question. But this general approach allows for more than one interpretation. Consider the following possibilities:

A first type of internalism is social interpretivism. This method is most prominently associated with Michael Walzer (1983), who argues that there are multiple domains even within the state. Goods such as medical care or education have “social meanings,” which must be interpreted to arrive at the principles internal to each distributive sphere. This entails a “situated” but critical reconstruction of shared understandings within the political community. While competition for educational opportunities might be found to follow a meritocratic principle, say, the best interpretation of the social meaning of healthcare might require distribution according to need. It is worth emphasizing that Walzer develops his theory in the context of a bounded political community, a society organized as a state. Hence, although different spheres of human activity are domains of justice governed by internal principles, they are simultaneously framed by a broader conception of citizenship or “complex equality” (Walzer 1983, 19). Nevertheless, there is no principled reason why the method itself could not be applied to domains of justice beyond the state.

A second type of internalism is teleological interpretivism. By this, I do not mean a method that assumes a consequentialist morality, but rather one that derives principles of justice from the “telos” or purpose of the domain in question. In the global justice literature, this approach is exemplified by the “practice-dependent turn” associated with theorists such as Aaron James (2005; 2012) and Andrea Sangiovanni (2008). Rather than appealing to Walzerian social meanings, these authors take their cue from Ronald Dworkin’s (1986) legal theory. Their method proceeds in three broad steps. The first is to identify and describe the object of interpretation. This step aims to delineate a relevant social practice in uncontroversial terms by referring to common understandings among its participants. In a second step, an interpretation of the normative “purpose and aim” of the practice is proposed. This interpretation must provide the best possible fit with the descriptive account of the practice established in the first step. The final step is to outline how the practice can be reformed to best align with its purpose and reconstruct the reasons participants have to affirm it. In James’s (2012, 28) version of the approach, this final step is explicitly contractualist, asking what is “reasonably acceptable to each party affected, as framed by the underlying conception of the practice.”

A third type of internalism also emphasizes justifiability to others, yet without appealing to interpretations of social or teleological meanings. It draws on general moral reasoning that is not constrained by these interpretive foci but is nevertheless applied to the particular social relationships that emerge within the domain in question. I will refer to it simply as contractualism. This method operates in two steps. First, it identifies structured forms of human interaction that affect people’s interests in ways that require justification—typically non-voluntary and consequential social practices. While the state remains the most important domain in this regard, it can be argued that certain practices and institutions beyond the state also meet these criteria. In a second step, the approach asks how the practice can be made justifiable to its participants in view of their individual interests (see, e.g., Valentini 2011b). The objective is to formulate regulative principles that, in T.M. Scanlon’s (2000) well-known expression, nobody can reasonably reject. The approach is internalist in the relevant sense if the reasons for rejection are limited to those that arise in virtue of the existence of the practice in question (cf. Beitz 2014, 234).

This list is neither exhaustive nor mutually exclusive, but I single out these three internalist methods because their influence can be traced in contributions to the recent global justice literature. Although there are important differences among them, they share the assumption that there are domains of justice governed by internal principles that are not derived from a higher-order theory of justice. Before examining integrationist challenges to this view, let me make three clarificatory remarks. First, internalists need not deny that there are considerations of justice that are not grounded in specific relationships, such as those within communities, social practices, or institutions. For example, few reject general humanitarian obligations or the existence of pre-institutional duties of allocative justice (such as those over natural resources in a global commons). The view I am interested in defending holds that in addition to whatever non-relational considerations exist, there are independent demands of justice that arise in virtue of the relationships within certain domains (for an attempt to systematize both types of considerations in the global realm, see Risse 2012). Internalists deny that these various considerations all derive from a single, higher-order theory of justice.

Second, internalists oppose the integrationist impulse to assess the justice of domains through the lens of a comprehensive theory of global justice. But they may well resort to holistic assessments regarding the distribution of goods and bads within domains.⁵ For example, Rawls’s theory judges the basic structure of society based on an all-things-considered judgment regarding the distribution of various “primary goods” within it. Dworkin’s theory, similarly, attempts to assess whether the combined holdings of different resources among people subject to sovereign rule conform to a master principle of equal concern. However, while these theorists resort to holistic assessments of primary goods or resources within the domain of a political community or society, they resist the idea of integrated assessments across domains.⁶ As such, they are internalist, not integrationist theories.

Finally, I take internalism to be a view about the nature of the demands of justice. Isolationism, as I shall understand the term here, is the view that a given distributive realm should be theorized in isolation. This view can be advanced purely for pragmatic reasons without drawing on internalist reasoning.⁷ It could be argued, for example, that excluding non-climate related considerations makes for more tractable principles for the distribution of climate responsibilities, which may in turn help with their practical implementation. Integrationists have rejected such claims and argued that taking a broader view is preferable even on pragmatic grounds (Caney 2012, 278–80; Walton 2020, 62). These arguments turn largely on empirical assumptions about political feasibility, and I shall not say much about them here. My interest is in internalism and integrationism on the level of philosophical principles and their methodological implications for theory-building.

INTEGRATIONIST PRESSURES

I now want to contrast the internalist approach with integrationism, the view that what justice requires within a given domain must be theorized in conjunction with considerations of justice in other domains. There are at least two ways of understanding this idea.

Strong integrationism: normative principles for any given domain are derived from higher-order, global principles.

Modest integrationism: normative principles for any given domain must be formulated with an eye to other domains so that practical conflicts are avoided.⁸

In this section, I examine various ways in which these two integrationist approaches can put pressure on internalism. I start with Simon Caney’s influential argument for integrationism in climate justice. I then focus on what I take to be the most powerful integrationist challenge to internalism, its alleged inability to provide plausible practical guidance. I finally examine an important conceptual challenge to the viability of internalist theorizing.

Substitutability

The term “integrationism” was first introduced by Caney in an article that challenges the popular view that greenhouse gas emission rights should be distributed on an equal per-capita basis (Caney 2012; see also Bell 2008). This view focuses exclusively on emissions and treats their distribution separately from other issues of global justice, such as poverty, trade, or development. But this, according to Caney (2012, 271), is a mistake: “If distributive justice is concerned with the fair share of a ‘total package’ of goods, then we have no reason to endorse a principle that applies solely to one particular item, such as greenhouse gas emissions.” To make this point, Caney examines and rejects several possible reasons for treating emissions in isolation. Central to his argument is the idea of substitutability. The right to emit greenhouse gases is valuable because it advances certain ends (e.g., heating or transport), but these ends can also be secured through alternative means. Furthermore, a shortfall in one type of good can often be compensated with a different type of good; for example, two people might be equally well off if one has fewer emission rights but more money. To be sure, not all goods are like this. For example, voting rights have a certain symbolic and expressive status, making them non-substitutable (Caney 2012, 273–4). But Caney finds no reasons of this kind in the case of emission rights. If emission rights are substitutable, then it follows that one should reject an equal per capita principle, or indeed any principle that targets this good in isolation. Instead, “one must then start with an account of what persons are entitled to as a matter of justice, and work back from that to deduce what share of emissions they are entitled to” (Caney 2012, 292).

Caney mounts a persuasive case against thinking of emission rights as a good to which specific principles apply. While his discussion is focused only on emission rights, making no claim about other issues, he concludes his discussion by formulating a multi-step method of theory-building that could be applied beyond the issue at hand (Caney 2012, 291–9). This method starts with a higher-order, global theory of justice that includes distributive, intergenerational, historical, and environmental concerns. This theory is then applied to the distribution of a given good, then subjected to substitutability tests, and finally a political implementation process. But although Caney intends this method to be neutral among substantive theories of justice, in fact it seems difficult to reconcile with the possibility that the distribution of goods within a given domain is governed by internal principles rather than principles derived from a higher-order theory of justice.

What supports this strong integrationist conclusion? It is ultimately the premise that “justice is concerned with the fair share of a ‘total package’ of goods.” If this is correct, then the domains in which goods arise play no constitutive role in shaping the demands of justice: the total package necessary cuts across domains. This is at odds with the view that justice is at least sometimes concerned with the distribution of goods in the specific context of shared social, economic, or political relationships. On this latter view, for example, an unequal distribution of educational opportunities between Swedes and Finns may not be a matter of justice, even if their distribution within each respective country is. Now, the point I want to make is that arguments about substitutability cannot in themselves support strong integrationism. It may be true that the inferior educational opportunities of the Swedes could be compensated with substitutable goods provided by the Finns, thus satisfying some distributive principle that governs the “total package.” But the fact of substitutability alone does not show that justice requires this. Hence, the case for strong integrationism is simply entailed by the premise that justice concerns the distribution of one “package,” rather than several.

In fairness, Caney (2005) has elsewhere given powerful arguments for a non-relational theory of global egalitarianism that is consistent with this view. For example, he has challenged the idea that state membership can generate entitlements of justice. Because state membership is typically an arbitrary fact, Caney argues, assigning it fundamental moral importance “appears to get matters back to front.” That is, it starts “with the world as it is and the boundaries that exist,” rather than first specifying an independent, global theory of what individuals are entitled to and working back from it to see what social and political arrangements are best suited to realize it (Caney 2008, 507).⁹ It would go beyond the scope of this article to discuss this theory. My point is merely that the argument from substitutability only supports strong integrationism if such a theory is correct. But that, of course, is what internalists deny.

Practical Judgments

A more critical way of advancing the case for integrationism is to scrutinize the implications of internalist theories of justice. If these theories generate implausible practical guidance, this will count as support for the integrationist method. Andrew Walton has made a series of arguments to this effect by focusing on the domain of international trade. His critique is aimed at internalist theories such as that of James (2012, 17), who attempts to assess the justice of the world trade system “independently of other moral or justice concerns which might equally apply in the global economy’s absence.” According to Walton, judgments about trade justice must necessarily draw on these wider concerns to yield the intuitively correct practical judgments. In order to assess the distribution of the gains of trade, the imposition of harms, or claims about exploitation, one must take into account the “overall level of advantage” of trading parties, which will be a product of factors beyond trade (Walton 2020, 54). For example, it seems unjust to divide the gains of trade between a rich country and a poor country equally, given the greater needs of the latter. Similarly, it seems unjust to impose competition harms on farmers in poor countries, even if the same harms may be imposed on farmers in rich countries, given their respective levels of advantage. “If we do not consult these wider concerns,” Walton (2020, 63–4) concludes, “the basis for this judgement is unavailable.”

But it is unclear what exactly these examples show. Consider two possibilities. On the strong interpretation of integrationism, they show the need to derive principles for trade from a higher-order theory of justice that applies some distributive principle across all domains. For example, we might seek to equalize overall levels of advantage among all people in the world, and benefiting poorer countries through our trade arrangements would advance this distributive goal. In contrast, the modest interpretation is that principles for trade must be sensitive to external considerations. Specifically, they need to be formulated so as to satisfy certain minimal moral concerns regarding absolute poverty, basic needs, or subsistence rights. This would be compatible with the view that there are distinct principles that govern the distribution of benefits and burdens within the domain of trade, provided they are circumscribed or adapted to reflect these minimal moral obligations.

All that is required to elicit the intuitively correct practical judgments in the examples above is the modest version of integrationism. What is driving these intuitions is most likely a concern for those who fall beneath some absolute moral minimum. To see this more clearly, consider a scenario in which a rich country and an even richer country engage in trade relations, generating an economic surplus to be distributed. Would we object to an equal distribution of this surplus on the grounds that this would not contribute to the realization of a higher-order distributive principle, such as one that seeks to equalize overall levels of advantage across all domains? The case seems much less intuitively clear. As such, it cannot decisively undermine the view that there are distinct principles of justice that would apply within the domain of trade.

But if we accept that principles of justice for any domain must be sensitive at least to considerations about basic needs, does it not follow that any plausible theory of global justice must be integrationist in the modest sense? In a recent piece, Alex McLaughlin (2022, 610) has indeed suggested that “integrationism is so inclusive it is trivial” because virtually all theories of justice accept some minimal moral constraint. To satisfy such a constraint, “we need to know whether benefits and burdens across domains combine in a way which leaves a particular agent below the minimal threshold. If they do, that agent should not be burdened in the context of global justice under consideration.” If this is correct, there would be no significant distinction between internalism and (modest) integrationism: to ascertain whether some minimal constraint is met, any plausible internalist theory would need to draw on broader assessments that consider distributive concerns beyond the domain in question.

However, it seems to me that this conclusion moves too quickly, for at least two reasons. The first, which I will flesh out in the following section, is that there are valuable forms of internalist theorizing that do not resort to these broader assessments. The second reason is that speaking of external “constraints” or “broader concerns” masks an ambiguity about the source of these considerations.

We have so far assumed that minimal moral constraints pertaining to basic needs or human rights are external considerations that influence or modify, in a modest integrationist way, the internal principles of justice within a given domain. But it is not clear that internalists need to accept this premise. Recall the scenario in which a rich and a poor country divide the gains of trade equally, despite the more urgent needs of the latter. It is not clear why internalist theories could not find internal resources to condemn such a division as unjust. Social interpretivists could point to social meanings within the domain of global trade; conceivably, principles of trade justice could reflect shared norms of solidarity with the disadvantaged, much like the norms of solidarity that exist within liberal-democratic states. Teleological interpretivists could argue that principles expressing a concern for the disadvantaged follow from the best interpretation of the point or purpose of the practice of trade. This concern might be said to be found in the “canons of interpretation” of the practice, including its history or its public rules and implicit meanings.¹⁰ Contractualists could argue that such a concern is an expression of the regulative principles that can be justified to all members of the social arrangements that govern global trade. For example, the poorest members within a system of trade might have a valid complaint against a distributive principle that does not optimally improve their position.

Why should we think of these as internal considerations? After all, the relative disadvantage of poorer trading parties may well be explained by circumstances beyond trade—that is, by domain-independent facts. Yet the internalist arguments sketched here, if they can be fleshed out successfully, would show that the existence of trade creates internal reasons to act on these facts of poverty within the domain, regardless of their sources outside it. Compare: a theory of justice for the basic structure of society may well require duties to redress or compensate for pre-institutional disadvantages, such as congenital health deficits, which are not caused by the basic structure (Daniels 2008). This duty derives from the relationships created by the basic structure, even if the source of the disadvantage itself lies elsewhere. Hence the facts and circumstances that shape internal principles of justice are not limited to those that would not arise but for the existence of the domain.

In the same way, a theory of global trade justice may be able to draw on internal resources to reach what appear like (modest) integrationist conclusions. To be sure, there are universal duties of humanity or justice that would exist even in the absence of any world trade system. There are also external considerations of justice pertaining to other domains. And it is certainly possible to use the trade system instrumentally for the pursuit of genuinely external goals, ranging from climate duties to geopolitical objectives (Miller 2010, 71). My point here is merely that integrationists are too hasty if they assume that internalist theories must necessarily draw on these external considerations to reach intuitively plausible practical judgments.

Demarcating Domains

But the integrationist challenge may run deeper: perhaps it is not simply that principles of justice for a given domain must draw on external considerations to reach plausible practical judgments. On a more conceptual level, it may be thought impossible to clearly separate the phenomena governed by internal principles, given the interlinkages between domains. Take, once again, the example of international trade: trade is linked to many other phenomena, such as global monetary policy, foreign investment, or climate change. Do principles for the domain of trade apply only to trade in goods and services, or also to international monetary cooperation, capital markets, or emissions trading schemes? Where we draw the borders of a given domain may naturally influence the content of the normative principles that guide it. Hence, if we cannot provide a principled answer to questions of scope, the theory cannot get off the ground (Walton 2020, 68–9; for a statement of this challenge in the context of Walzer’s theory, see Gutmann 1995).

A first reply to the demarcation problem is that it does not arise with equal force in all cases: whether the boundaries of a domain are blurry will depend on the domain in question. Take, for instance, the state. It is true that states are not fully isolated and discrete social forms; war, migration, trade, and diplomacy have always connected states and their citizens to people and institutions beyond their borders. There are also famously vexing questions about who counts as a member for the purposes of exerting and being a subject of state power. Yet despite these complications, the idea of the state as a domain of justice is widely thought to be coherent and meaningful. As Rawls (1999a, 7) puts it, “the significance of this special case is obvious and needs no explanation.”

Granted, some domains are harder to demarcate. This will be the case especially when we cannot point to clearly defined institutional structures that serve as the target for principles of justice and simultaneously define their scope. Think again of international trade: although the World Trade Organization (WTO) provides a forum for multilateral trade agreements, there are many phenomena that could plausibly fall under the domain of trade but are not governed by the WTO, including global capital markets or foreign direct investment. How can we define boundaries in these cases?

Different internalist methods will provide different answers. The social interpretivist approach attempts to derive principles of justice from shared social meanings. This suggests that the problem of justice arises only within bounded political communities, where shared social meanings can exist. To the extent that political communities with shared social meanings do not exist in the global realm, this approach suggests a narrow set of domains of global justice. Teleological interpretivism identifies a social practice, interprets its purpose, and derives normative principles from said interpretation. Identifying the boundaries of a domain in the first step does not require shared understandings among participants of the practice, but it does require some level of consensus about what the practice is and who its participants are—this needs to be established in “relatively uncontroversial, sociological terms” (James 2012, 28). Cases where that is not possible might simply not be domains of justice on this view. The contractualist method suggests a more permissive approach. What is needed here is a plausible account of a social realm structured by a system of rules that impose non-voluntary constraints and shape individual interests in meaningful ways. Even if we acknowledge a high degree of interdependence between several social realms, we can start from a provisional descriptive account of the realm in question and draw and redraw the boundaries of the domain by evaluating the implications of different iterations.¹¹ There may be more than one possibility. For some purposes, the question “how should the WTO be structured to treat its members fairly?” is perfectly valid and important. Separate internalist theories can be developed for adjacent domains, such as the global financial markets (Wollner 2014) or the foreign investment regime (Kniess 2018). For other purposes, perhaps, we may be interested in the wider global economy, or as James (2012, 19) puts it, the “international practice of market reliance.”

To be sure, none of this is to suggest that every area of global politics can be theorized in an internalist fashion. Take the case of climate justice. It is not clear that all the normative considerations raised by climate change are best described as deriving from the existence and nature of a political community, social practice, or institution. If that is the case, there may simply not be a clearly demarcated domain characterized by the types of social relationships that give rise to internal demands.¹² Of course, that does not mean that climate change does not raise considerations of justice. Some of these may be non-relational (such as, say, the allocation of initial atmospheric shares or emission rights); others are based on historic patterns of social interaction or shared institutions (such as, say, duties of social justice to compensate for past harms to outsiders, or requirements of fairness and inclusion in climate governance). This suggests that there is not a single domain of climate justice but rather a complex patchwork of normative considerations emanating from various sources. But this complexity is not itself a reason to adopt the integrationist approach; on the internalist view, it simply reflects the nature of justice.

PRACTICAL GUIDANCE, REVISITED

I suggested in the preceding section that internalist approaches can avoid implausible practical judgments by drawing on internal resources to address seemingly external considerations. But someone might object that this possibility does not disarm the worry entirely, for the strategy is not guaranteed to succeed in every case. There may simply not be internal resources to address issues that, on reflection, should be captured by principles of justice for a given domain. To put the worry differently, the internalist method appears to only provide contingent reasons to care about pressing issues that may go beyond the domain in question, such as global poverty, human rights, or the environment. These issues inform justice within the domain only if they form part of social meanings, teleological interpretations, or provide grounds for the rejection of principles that govern the relationship among members of the domain. If they do not, the problem of practical guidance rears its head again.

The worry, as I have been describing it, is that internalist theories produce implausible practical judgments. A different way of putting the same point is to say that the practical guidance they provide is incomplete: it only tells us what justice demands within the domain, but not all things considered. In this section, I argue that there is nevertheless an important place for internalist theories that focus on a given domain without drawing on the wider perspective championed by integrationists. And perhaps surprisingly, this form of theorizing has distinct advantages in helping to guide our practical judgments.

To elucidate this point, it is important to reflect on the aims of political philosophy more generally. One important aim is to provide normative orientation in the face of political disagreement as it arises at a specific time and place. By clarifying what justice demands of us here and now, political philosophy can provide practical guidance for the reform of existing social and political arrangements. But theories of justice can also abstract away from our present circumstances and formulate a more utopian ideal as a benchmark for the assessment of existing arrangements—a distant goal to strive toward. These, of course, are two familiar aims that are often discussed in the context of Rawls’s (1999a, 216) famous distinction between ideal and non-ideal theory. But I am here drawing the distinction in somewhat broader terms. For one thing, Rawls’s focus on compliance with the demands of justice is but one of the ways in which theories can be realistic or idealistic (Valentini 2012). It would also be a mistake to think of the distinction in strictly dichotomous terms. What Rawls calls ideal and non-ideal theory are two possible points on a continuum of possible presuppositions about the world to which our theories apply (cf. Carens 2013, 300–6). There are degrees of realism, just as there are more idealistic conceptions of justice beyond Rawlsian ideal theory. The level of abstraction we choose must depend, at least to some degree, on the kinds of questions we seek answers to.

Now, much like theories of justice can be more or less realistic, they can be more or less integrated (see also Christensen 2017, 6). One type of internalist theorizing is realistic because it focuses on a specific issue “from the bottom up,” attempting to capture its complexities and nuances the way they are manifested in a concrete social setting. A practical version of this approach can also be observed in some “single-issue” political movements that focus exclusively on a specific injustice to arrive at actionable and tractable demands. Another approach is internalist, but modestly integrated in light of realistic assumptions about the interactions between separate domains. Suppose we are interested in what justice demands in the realm of international trade. For concrete practical guidance, we might do well to consider how the rules that govern the World Trade Organization (WTO) impact, and in turn are impacted by, a host of issues ranging from climate change to migration (assuming, of course, that these are external issues). Injustices in other domains may well influence the ways in which we ought to reform the WTO in our present circumstances.¹³

But it would be a mistake to assume that internalist theorizing is necessarily realistic. Much like we can theorize the best way to reform the WTO, for example, we can also ask a more general question about the normative principles that would regulate an idealized structured trade system among nations under favorable conditions for its implementation, including compliance with principles of justice in other domains. Granted, the extent to which such idealization is deemed worthwhile or even possible will depend on the method used to derive domain-specific principles. Social interpretivism does not lend itself to this task readily: because it draws on existing practices, institutions, and social meanings, its interest lies mostly with the social world as we find it. Teleological interpretivists, it seems to me, can abstract away from existing conditions somewhat more, since the “point” of a practice can be rather removed from its actual instantiation.¹⁴ Finally, the contractualist approach can be developed equally under realistic or idealistic assumptions. Drawing on the Rawlsian method, for instance, we can imagine two different hypothetical original positions to formulate principles of trade justice: one in which contract parties are aware of present and past injustices in other domains, and another in which these empirical facts are not to enter their deliberations.

If we accept this general picture, one immediate upshot is that initially implausible practical judgments are not in themselves reasons to reject domain-specific principles at the level of ideal theory. To illustrate the point, consider theories of justice in migration. The first two thirds of Joseph Carens’ The Ethics of Immigration discuss the subject under realistic assumptions regarding the political constraints and injustices of the world we live in. The final part defends a more idealistic and radical proposal for open borders. This proposal is still developed through what Carens (2013, 230) calls “political theory from the ground up”: instead of deriving the case for open borders from a general theory of freedom or equality, he reconstructs it from existing democratic commitments and practices. Now, Carens readily acknowledges that applying his proposal in practice would create considerable issues. Not only would it presumably produce a backlash in liberal-democratic societies, it would also fail to solve (and according to some critics even exacerbate) the problem of global inequality, given that many of the world’s poor are unable to migrate to wealthy countries. And yet Carens (2013, 234) insists that global freedom of movement is an important moral goal at the level of principle, “quite apart from its effects on the overall level of inequality.”

Consider now a quite different ideal theory of justice in migration. Like Carens, Kymlicka (2001, 253) starts “from ‘actually existing’ liberal democracies to see which practices have developed over time within them regarding boundaries and membership.” He finds that there is value in borders that demarcate discrete national political communities, as certain forms of national identity-building can promote liberal values such as democracy, equality of opportunity, and solidarity. Kymlicka (2001, 270) himself recognizes, however, that immigration restrictions to promote liberal nationalism may be indefensible in a world in which some people are condemned to live in poverty if they cannot migrate to a wealthier country. Both Kymicka’s and Carens’ theories, then, could be criticized on account of implausible practical implications that can be avoided by introducing broader considerations pertaining to human rights, poverty, or historic injustice. But this would miss the deeper purpose of these accounts, namely what they tell us about the nature of justice in migration on a more abstract plane.

There are reasons to think that internal principles are in fact necessary for a full account of how we ought to go about creating more just arrangements. These mirror the reasons Rawls advances in favor of ideal theory. “By showing how the social world may realize the features of a realistic utopia,” says Rawls (1999b, 128), “political philosophy provides a long-term goal of political endeavor, and in working toward it gives meaning to what we can do today.” Identifying the long-term goal helps us understand what is required under non-ideal circumstances, “for until the ideal is identified, (…) non-ideal theory lacks an objective, an aim, by reference to which its queries can be answered’ (Rawls 1999b, 90; see also Simmons 2010). In short, ideal internalist principles can help guide the transition toward just global arrangements. There are, after all, many alternative paths through which we could reduce present injustices. By formulating standards to aspire to, internalist theories shed light on the desirability and potential opportunity costs of these different paths. Hence, even if an ideal theory of justice in migration, say, is so detached from present circumstances that it cannot provide immediate guidance as to what justice requires here and now, it can be used as a long-term yardstick to evaluate alternative courses of action. Among other things, this can help guide normative judgements when empirical circumstances change (say, if severe poverty were eliminated, so that it played no major role in migration policy).¹⁵

Now, integrationists too can claim that their method is applicable to ideal circumstances (Walton 2020, 66). Assuming perfect conditions, we imagine how different domains (trade, migration, and so on) could be governed to bring about some higher-order distributive ideal. Yet such an account would face a problem of indeterminacy. To illustrate, suppose our ideal theory of global justice requires outcome X, where X might refer to an egalitarian, prioritarian, or sufficentarian distribution of the overall package of benefits and burdens among all people in the world. On this view, all parts of the global order must be regulated so that they together bring about a distributive outcome that satisfies this principle. Could such a theory help us assess whether a given way to regulate the domain of international trade or migration is just or unjust? Because the integrationist theory is concerned with overall outcomes, it would not be enough to show that a certain WTO policy or immigration law does not optimally advance X. After all, the policy could be part of a set of global arrangements that, in their combination, ensured that X was brought about. Indeed, to reject the policy or law in question, it would need to be shown that it makes impossible the overall outcome of X. Given that X could be achieved through means other than international trade or migration, it is not clear how this could be shown conclusively. Hence the integrationist approach cannot tell us whether trade or migration regimes are just or unjust on their own merits.¹⁶

I should clarify that this is not a claim about the complexity of integrationist theorizing. Some have argued that identifying and drawing all relevant connections between diverse domains of justice presents significant informational challenges (Meyer and Roser 2006, 239; Zellentin 2015). Whatever we make of this epistemic claim, my argument is about the indeterminacy of principles that target the overall distribution of benefits and burdens at the global level. To assess the justice of a policy in domain A and its effect on some global distributive outcome, one would need to hold all other domains (B, C, D, etc.) fixed. But the arrangements that govern B, C, D, and so forth could also be governed differently. Because different combinations of domain-specific regulative principles can bring about the same desired overall pattern of benefits and burdens, these principles must be made determinate somehow.

Caney is not unaware of the problem. If what matters is only the overall distribution, we must choose among different ways to bring this distribution about. Hence the final step of his integrationist method is a political process in which participants decide how precisely people’s entitlements are to be secured (Caney 2012, 298). However, Caney says little about how such a political process would work, including what kinds of institutional arrangements would be necessary to confer it the legitimacy to make the demands of global justice actionable. This is no minor detail. Even if one does not take the existence of legitimate global political agency to be a requirement for duties of global justice to arise (Meckled-Garcia 2008), it appears to be necessary (at least on the strong integrationist view) to render these duties determinate and to produce action-guiding principles. Furthermore, it is not clear that, once a democratic process is introduced, it will always deliver the best or most just way to achieve the correct distributive principles, by integrationist lights.

Now, an integrationist might reply that it is no embarrassment to their view that a given theory of justice is realizable in multiple ways, requiring some way to choose among them—that is simply what the integrationist view is. But if we care about practical guidance, this lays bare one respect in which integrationism fares worse than internalism. Even in the absence of a global political process that may or may not materialize, internalist theories can help specify long-term ideals for the regulation of diverse domains of global justice. Take, for instance, an ideal theory of migration that defends a right to global freedom of movement, analogous to the domestic freedom of movement that exists in liberal societies. Defending such a theory need not entail advocating open borders in the non-ideal circumstances we face presently. In this specific sense, it is not immediately action-guiding. But by formulating an ideal principle, the theory can help shed light on the moral costs and trade-offs involved in pursuing policies that move us away from this principle. Among other things, it can provide at least pro-tanto reasons to follow alternative courses of action if these are feasible and effective in addressing the moral considerations that tell against implementing the ideal principle. This way, it can help make the demands of justice actionable by combining the idealistic and realistic dimensions of theory-building.

Let me end, however, on a conciliatory note. While the integrationist cannot say what is to be done within a domain before the full picture is provided, the internalist can tell us how to render a domain internally just, but not whether we ought to do so, all things considered. The reason is that advancing justice within the domain may involve external repercussions that undermine the pursuit of justice in other domains. Suppose, for example, that realizing domestic justice in a given country would require a protectionist trade policy aimed at lifting the economic opportunities of the least advantaged members of society, but that this policy is at odds with what trade justice requires globally. The theory would not deliver decisive practical guidance because the demands of justice in two separate domains pull in different directions.

Now, the possibility of conflicts among considerations of justice cannot vindicate strong integrationism, any less than conflicts among values can demonstrate the truth of value monism (for instance, the attempt to reduce the values of beauty, friendship or liberty to their contributions to utility). But what it does show is that we need somehow to navigate the trade-offs between competing demands involved in concrete circumstances. One possibility is to propose general priority rules among various principles that govern different domains—but even those who have attempted this, such as Risse (2012, 330), concede that “the lack of conclusive arguments to support my ordering reflects the nature of a pluralistic theory.” The alternative is to eschew formal rules and rely on contextual judgments informed not only by moral reflection but also considerations of feasibility and political strategy. But this, presumably, is also what integrationists must pin their hopes on in the absence of a global democratic process. We should expect theories of justice to provide practical guidance, but perhaps not all the way.

CONCLUSION

Contemporary global political philosophy continues to move toward ever more applied normative analyses of specific sub-fields of global politics. This trend, which reflects a growing willingness to engage with applied issues and the empirical social sciences, has led to fruitful developments in the theory of global justice. According to integrationists, however, theorizing at the level of domains like trade or migration should not operate in isolation from broader considerations of global justice. At its most radical, this view suggests that normative principles in these domains must stem from a higher-level theory of justice, for only a higher vantage point allows us to see the connections and common moral considerations that cut across domains. Internalist theories, on this account, risk failing to see the wood for the trees. Contrary to this view, I have attempted to clarify and defend the idea that different domains of justice are governed by internal principles specific to them. First, we need not embrace the strong integrationist method. Even arguments for the modest version overlook the possibilities of internalist theorizing. Furthermore, the question of demarcating domains does not constitute an insurmountable challenge to internalism. Finally, we can distinguish several versions and uses of the internalist approach. Despite appearances, “idealistic” perspectives that exclude external considerations can be valuable for practical guidance, for they shed light on the nature of the demands that diverse domains of justice make on us.