A non-ideal global basic structure

Focusing on the basic structure as the subject of justice has tended to lead theorists to make a choice: either there is no global basic structure and therefore obligations of justice remain domestic only (the statist position) or there is sufficient institutional basis at the global level to warrant affirming a basic structure global in scope, meaning that duties of justice must also be global (the cosmopolitan position). Recent literature, however, has pointed out that this might be a false choice between denying and asserting the existence of a global basic structure. There are two main claims that I make in this paper. First, I claim that on a Rawlsian understanding of the basic structure, justice does not require one before its demands arise, but rather that under certain conditions, justice can require that a basic structure be established as an essential part of fulfilling its demands. This has the benefit of not restricting the scope of justice to the domestic sphere. Thus, the second claim is about determining, from a practice-dependent, non-ideal starting point, what those “certain conditions” are. Specifically, I argue that when currently existing global institutions begin impacting on the freedom of individuals to interact against a fair backdrop and pervasively impact on life chances, then the demands of justice will arise and we will need to establish a global basic structure. This paper, then, also has implications for the ideal/non-ideal theory debate, because I argue that the best way to globalize the basic structure is to begin from a non-ideal starting point.


Introduction
In this paper, I take as given that justice is a virtue of institutions, and that the "basic structure" is the subject or site of justice. 1 The basic structure, in Rawlsian terms, is defined as "the way in which major social institutions fit together into one system, and how they assign fundamental rights and duties and shape the division of advantages that arise through social cooperation" (Rawls 2005: 258).My purpose is to explore and elucidate part of the global justice debate that concerns the idea of a global basic structure.Focusing on the basic structure as the subject of justice has tended to lead theorists to make a choice: either there is no global basic structure, and obligations of justice therefore remain domestic only (the statist position) or there is sufficient institutional basis at the global level to warrant affirming a basic structure global in scope, meaning that duties of justice must be global (the cosmopolitan position).Recent literature, however, has pointed out that this might be a false choice between denying and asserting the existence of a global basic structure.The dichotomy arises largely from the assumption that a basic structure must exist before the demands of justice can arise. 2The problem of cosmopolitanism versus statism might be circumvented if we can prove that justice requires the establishment of a basic structure to fulfil its demands.
Reinforcing and further complicating the cosmopolitan/statist divide is another concept introduced by John Rawls, that of ideal and non-ideal theory.Though Rawls discussed ideal and non-ideal theory in terms of "compliance", the terms have evolved into a section of (meta-) political theory of their own. 3Ideal theory has come to mean political theory that abstracts away empirical facts or idealizes certain aspects of society, and non-ideal theory is about guiding action within a society and about reforming practices in a way that is applicable here and now. 4 The ideal/nonideal theory division is important in the global justice debate because focusing primarily on ideal theory tends to lead theorists to assert obligations of global justice, while non-ideal theorists tend to limit themselves to concerns of domestic justice.This paper has implications for the debate, because I argue that the best way to globalize the basic structure is to begin from a non-ideal starting point. 5 Before setting out my thesis, I want to comment on the significance of attempts to globalize the basic structure, which thus far have largely been in response to empirically observed injustices at the global level.Many see the global order as imposing injustices on developing countries and their citizens, and use a global basic structure as a call for justice.Yet most of these attempts have started with the (empirically suspicious) claim that there exists an unjust global basic structure to which we must apply principles of distributive justice. 6Two aspects of the cosmopolitans' quest for a global basic structure have been problematic.First, by assuming that a basic structure is a precondition for the demands of justice to arise, it limits their ability to assert that justice is global in scope.Second, focusing primarily on distributive justice rather than remembering the lexical priority of Rawls's two principles of justice has tended to obscure more fundamental problems with the debate.In this paper, I do not work under the assumption that there is an existing global basic structure, but rather focus on the more philosophically interesting questions of, if, and under what circumstances one might arise.Additionally, I make no normative comment on whether or not we should adopt a global basic structure, only how to understand it and what conditions might give rise to the need for one.
There are two main claims in this paper.First, I claim that on a Rawlsian understanding of the basic structure, justice does not require a basic structure before its demands arise, but rather that under certain conditions, justice can require that one be established as an essential part of fulfilling its demands.This has the benefit of not restricting the scope of justice to the domestic sphere.Thus, the second claim is about determining what those "certain conditions" are.I propose that, beginning from a non-ideal starting point, we look at practices that give rise to the need for global justice, thus requiring the establishment of a global basic structure to fulfil its demands.Specifically, I propose that when currently existing global institutions begin impacting the freedom of individuals to interact against a fair backdrop and pervasively impact their life chances, then the demands of justice will arise and we will need to establish a global basic structure.This conclusion has the added implication of suggesting that the starting point for theorizing about a global basic structure need not be ideal theory.
The first section of this paper is dedicated to producing a comprehensive understanding of a Rawlsian basic structure, a task that has been neglected in much of the global justice literature.The tendency has been essentially to pluck Rawls's conception of the basic structure out of its domestic context and claim that it applies globally, yet this is done without sufficient attention to the justification Rawls gives for making the basic structure the subject of justice and its role in a theory of justice at large.This section explores the question of how to understand the basic structure as the subject or site of justice with a close analytical look at Rawlsian literature.Section II then looks at how the basic structure as the subject of justice does not necessarily limit its scope, and justifies the claim that a basic structure does not have to pre-exist justice.In Section III, I introduce Ronzoni's idea of a practicedependent, non-ideal approach to justice and suggest it as a plausible understanding of Rawls's approach to formulating justice as fairness.I go on in Section IV to present Ronzoni's suggestion of how a practice-dependent approach can be used to devise a global basic structure, and although I disagree with her interpretation of Rawls, I argue that the overall approach is useful.In Section V, I put forward my own suggestion as to under what circumstances a global basic structure will arise.I ultimately conclude that the demands of global justice arise due to problems with currently existing global institutions that limit the freedom of individuals to interact against a fair backdrop and pervasively impact their life chances.

I. Understanding Rawls's Basic Structure
Rawls begins A Theory of Justice by asserting that "Justice is the first virtue of social institutions" (Rawls 1999: 3).That is, he begins by flatly rejecting an individualistic justice, and asserts that justice as fairness is a strictly political and namely institutional concept.It is important to note that justice as fairness applies only to institutions, which then subsequently regulate and establish just interactions between individuals (and groups), but individuals themselves are not the subject of justice as fairness.
"… [T]he primary subject of justice is the basic structure of society, or more exactly, the way in which the major social institutions distribute fundamental rights and duties and determine the division of advantages from social cooperation" (Rawls 1999: 6).
The basic structure comprises institutions and the way in which they fit together to regulate interactions within a society. 7This comprises the backdrop for social interactions and is the primary venue for justice.
The role of the basic structure is to secure background justice.Because Rawls values the freedom of individuals (and groups), and does not wish the rules of justice to apply directly to them, they must have a fair backdrop against which to act.In other words, the rules of justice apply to the basic structure which ensures "… the fairness of the system of social cooperation" (Abizadeh 2007: 326).In effect, the basic structure is pragmatically the best way to ensure background justice, but Rawls's justification is more nuanced.Rawls gives two distinct, but mutually supporting, justifications for making the basic structure the subject of justice.I discuss each in turn.Following Rawls and Scheffler, I label the first justification the "division of labour" justification (Rawls 2005: 268;Scheffler 2006: 9).Though some theorists have argued it is a purely pragmatic justification-the basic structure is the best way to secure background justice-I argue that this is insufficient, and the purpose of the division of labour justification is actually twofold (Abizadeh 2007: 327-9).Making the basic structure the subject of justice also speaks to Rawls's liberal commitments, and allows individuals to remain as free as possible in their own life choices.The second justification for the basic structure as the subject of justice, which I term the "pervasive impact" justification, speaks to Rawls's egalitarian commitments, noting the pervasive impact that the institutions of the basic structure have on citizens' life chances. 8 The division of labour argument is, in large part, a response to Nozick's assertion that "liberty upsets patterns," the idea being that the background conditions which make interactions fair in the first place cannot be preserved over time due to simple, everyday interactions. 9This division of labour separates the principles of justice that apply to the basic structure from the general rules that apply to individual actions.(These "rules" are not rules of justice, but rather simply rules governing social conduct.)Having the basic structure as the subject of justice prevents us from having to devise complicated rules that apply to complex and vast human interactions, and is therefore the best means to bring about justice, but it also preserves the freedom of individual actions.The basic structure is the subject of justice and provides background justice precisely so justice does not become a property of individual actions.
Rawls's second justification for making the basic structure the subject of justice has to do with the effects of the basic structure on the lives of individuals (Rawls 1999: 7).The basic structure will have a profound impact on individuals' life chances.
"Now everyone recognizes that the institutional form of society affects its members and determines in large part the kind of persons they want to be as well as the kind of persons they are.The social structure also limits people's ambitions and hopes in different ways; for they will with reason view themselves in part according to their position in it and take account of the means and opportunities they can realistically expect" (Rawls 2005: 269).
Thus, the basic structure is the subject of justice because it, more than any other aspect of society, has the most profound impact on the lives of the citizens in the society, not just from their starting points, but also by articulating their wants and aspirations (Scheffler 2006: 5).It is important to note here something that is often overlooked in literature on the basic structure: each of these justifications for making the basic structure the subject of justice corresponds roughly to Rawls's two principles of justice. 10The first, the division of labour argument, ensures the maximum amount of liberty for individuals.The basic structure, rather than citizens themselves, is the subject of justice to allow for maximum freedom against a fair backdrop that ensures a similar amount of liberty for everyone (the first principle).The second, pervasive impact justification has to do with the way the basic structure impacts the lives of individuals, and ultimately the way in which it distributes social goods (the second principle).Recall, also, that the principles of justice are lexically ordered, so that only when the first principle is met, can the requirements of the second principle be fulfilled.The consequence of this is that theorists who focus first and foremost on Rawlsian distributive justice run into problems, because they first forget about the lexical priority of the principles of justice, and second forget that the two justifications are mutually supportive.The basic structure is not the subject of justice because it is the best way to secure justice and individual freedom simultaneously OR because it pervasively impacts the lives of citizens: the two must be taken together.
There is one more aspect of the basic structure argument important for our purpose that is often overlooked in the literature on basic structure.Along with the division of labour argument, Rawls introduces an ideal basic structure.
"… [A]lthough society may reasonably rely on a large element of pure procedural justice in determining distributive shares, a conception of justice must incorporate an ideal form for the basic structure in light of which the accumulated results of ongoing social processes are to be limited and adjusted [emphasis added]" (Rawls 2005: 281).
This ideal basic structure is determined by the two principles of justice: we apply them to the basic structure to give it its ideal form and then use this as our guide to make alterations to preserve background justice.Adjustments to the basic structure will be necessary because background justice will eventually erode.In making the relevant adjustments to the basic structure, we need an ideal by which to judge if our adjustments are in line with a just basic structure. 11What we therefore end up with is two versions of the basic structure: an ideal, perfectly just picture, and the actual one. 12 The implications of the ideal basic structure will become important in Section IV below.Now, however, I want to move on to substantiating the claim that the basic structure does not have to exist before the demands of justice can arise and instead, under certain circumstances, justice requires the implementation of a global basic structure to fulfil its requirements.

II. The Scope of Justice and how Justice "Requires" a Basic Structure
Up to this point, we have only discussed the basic structure as the site of justice, but have said nothing of its scope, which is the source of much of the disagreement between cosmopolitans and statists. 13  "The site of justice is not the same as its scope: the site of justice refers to the kinds of objects (individuals' actions, individuals' character, rules, or institutions, and so on) appropriately governed by principles of justice, that is, to which the principles rightly apply, whereas the scope refers to the range of persons who have claims upon and responsibilities to each other arising from considerations of justice" (Abizadeh 2007: 323).
Statists, naturally, claim that the scope of justice is domestic, while cosmopolitans want to claim that it is global.Here is the typical statist argument for restricting the scope of justice to the domestic sphere: 1) The primary subject of justice is society's basic structure 2) A basic structure global in scope does not exist ∴ 3) The scope of justice is not global (Abizadeh 2007: 322).
The first premise is a comment on the site of justice, whereas the second is a comment on its scope.Thus, for the third premise to be valid, the site and scope of justice must coincide; that is, "… its validity requires showing that the boundaries of the basic structure qua site of justice somehow limit the scope of justice to the range of persons whose lives are regulated by the existing basic structure" (Abizadeh 2007: 323).In other words, to limit the scope of justice to those affected by the basic structure depends on the assumption that there is one already in place.This argument relies on the assumption that the basic structure is a prerequisite, or "existence condition" for justice (Abizadeh 2007: 320).
Yet, there are two other ways in which justice can be said to require a basic structure.The basic structure could be a constitutive part of justice, meaning justice includes (by constitutively demanding) a basic structure.Alternatively, justice could require a basic structure as an instrumental means by which to realize its demands (Abizadeh 2007: 320).On the latter two interpretations of the basic structure as the subject of justice, if demands of justice arise sans a basic structure, one must be put in place in order to realize justice.The benefit of affirming one of the second options is that we are not limited by a pre-existing basic structure, meaning the scope is not necessarily limited to the domestic sphere.When we assert that the basic structure is a constitutive or instrumental condition of justice, we then simply have to look to see where (to whom) justice applies and then create a corresponding basic structure to fulfil its demands.
Discovering whether, on any given definition of the basic structure, it is a precondition instead of an instrumental or constitutive part of justice, depends upon the justification given for making the basic structure the site of justice.In other words, the reasons we give for making the basic structure the site of justice will determine whether or not it must exist before the demands of justice can arise.
Thus, for our purposes, we look at the justifications Rawls gives for making the basic structure the subject of justice, and from there see whether the basic structure is an existence condition or whether it can be said to be instrumental or constitutive. 14 Given that the two principles of justice roughly correspond to the two justifications for basic structure as the subject of justice, I propose that the basic structure is a constitutive part of justice.It is essential for fulfilling the demands of justice.On the first, division of labour justification, the point of the basic structure is twofold: it is the best way to distribute justice, but is also the best way to ensure that individuals are free to act in a way that is compatible with a similar amount of freedom for everyone.This makes the basic structure a constitutive requirement of justice.Ensuring that people have rights and are free in a way compatible with the same rights and freedoms for everyone certainly does not require a pre-existing basic structure, but it does mean that one is essential to fulfil the requirements of justice.
On the second, pervasive impact justification, at first glance it might be an existence condition for justice: "… considerations of … justice arise if and only if there exist major social institutions (i.e.society's basic structure) with a pervasive impact on the life chances of persons" (Abizadeh 2007: 343).Yet, the basic structure comprises only institutions that have a pervasive impact on the life chances of persons if we define it in that way, and our definition of the basic structure is wider than that.Thus, there will naturally be institutions in the basic structure that have a pervasive impact on people's lives, but it does not mean that an entire basic structure must pre-exist justice.It simply means that the institutions that do impact on people's lives must be included when we devise a global basic structure.
Ultimately, I conclude this section with the assertion that on our interpretation of Rawls, there are good reasons to assert that justice does not need a pre-existing basic structure, but rather that one simply needs to be established when the demands of justice arise. 15This has the added benefit of not (immediately) limiting the scope of justice to the domestic sphere.The next question, then, is under what circumstances the demands of justice will arise, and to determine if it will be global in scope.Before we move on to exploring these circumstances, I wish to return to the ideal/non-ideal theory debate, and introduce the "practice-dependent approach" to formulating a theory of justice.This is a strictly non-ideal approach, and, in addition to being a plausible way to interpret Rawls's approach to formulating his theory of justice, it has also been used to propose under which circumstances the need to establish a global basic structure arises, as will be discussed in Section IV.

III. The Practice-Dependent Approach
I now turn to the division between ideal and non-ideal theory in recent literature on global justice. 16I want to look at the so-called "practice-dependent" account of justice, which is a paradigmatic case of non-ideal theory. 17The practice-dependent account maintains that "[t]he content, scope, and justification of a conception of justice depends on the structure and form of the practices that the conception is intended to govern" (Sangiovanni 2008: 138).We have established the claim that justice does not require a basic structure as a precondition for its demands to arise, and have now turned to the question of which circumstances might necessitate the development of a basic structure.Thus, the practice-dependent account seems promising because looking at global practices might tell us under what circumstances a basic structure is required to fulfil the requirements of justice.
Institutionalists, a subgroup of practice-dependent theorists, 18 maintain that "social and political institutions fundamentally alter the relations in which people stand, and hence the first principles of justice that are appropriate for them" (Sangiovanni 2008: 138).Further, institutions "… establish a set of background conditions which alter the way in which participants interact.And these institutionally mediated relationships, in turn, shape reasons we might have for endorsing (or rejecting) a given set of principles" (Sangiovanni 2008: 147).
Sangiovanni's talk of "background conditions", and the way in which institutions form special relationships is in keeping with the idea of the basic structure.Additionally, though the practice-dependent approach in general does not endorse justice as fairness itself, it is a plausible reading of Rawls to assume that justice as fairness is an institutionalist idea, and that Rawls himself utilized a type of practicedependent approach.
When justifying his conception of justice, Rawls pays close attention to "the culture of the social" (Rawls 2005: 14).In other words, society's history, culture, and the practices that derive from the two, play a large role in formulating a theory of justice.Additionally, the institutions of the basic structure put people in a special relationship.This means that the principles of justice must be mutually justifiable to all citizens as participants in this basic structure.The reasons individuals have for endorsing these principles of justice will be determined by their societal history and culture, meaning that, again, the content of, and justification for, justice as fairness is dependent upon practices.Further, Scheffler notes in talking about the reason for making the basic structure primary, "… [G]iven the feasibility constraints that apply to the rules regulating individual conduct, the institutional forms required to insure background justice will have to go beyond and help fix the content of those rules" (Scheffler 2006: 9).Thus, we can see that it is a plausible reading of Rawls to interpret him as a practice-dependent institutionalist (Sangiovanni 2008: 150-4).
One might note that this could be a potentially problematic interpretation of Rawls due to his assertion that we need an ideal picture of the basic structure (one that does not depend on practice) to guide our adjustments to the actual basic structure.This makes Rawls seem, at least on the face of it, to favour a practiceindependent approach.My response to this objection is to look at the origins of the ideal basic structure.While it is true that the ideal basic structure does not depend on practice in the sense that its justice will not be eroded by individual actions, its very formulation comes from the shared political culture.The ideal basic structure is a picture of the main institutions of society to which the two principles of justice have been applied.These two principles are derived from reflective equilibrium (a practice-dependent process), and the institutions that comprise the basic structure are still the main institutions of the actual society.Thus, although Rawls does want to include an ideal basic structure against which we measure the justice of the actual one, even the ideal version is non-ideal in an important way.Now that we have established that reading Rawls as a practice-dependent institutionalist is a plausible interpretation, I want to move on to a proposition of how to globalize the basic structure, also made from a practice-dependent perspective.As I have noted before, many attempts to globalize the basic structure have used Rawls's idea of a basic structure, but have largely ignored the process and justifications that Rawls used in theorizing about it.Thus, an approach consistent with Rawls's practice-dependent methodology seems promising.

IV. Globalizing a Non-Ideal Basic Structure
Up to this point we have said that justice is a virtue of institutions, and that a lack of a global basic structure seems to be the problem that cosmopolitans have in successfully asserting obligations of global justice.As we have established that the basic structure need not be a precondition for the existence of justice, if we look at the injustice of the global order from a different perspective, the problem might not be the unjust institutions of a basic structure, but rather a lack of institutions needed to regulate the global order (Ronzoni 2009: 229).
The practice-dependent view tends to lead us to assume that the basic structure is an existence condition for the obligations of justice to arise.Because any given conception of justice will depend on the practices that it regulates, we have to refer to a particular practice within that conception (i.e.participation in the basic structure).Further, because obligations of justice arise due to the special relationships in which these practices put people, the content of the obligations of justice depends on the individuals' participation in these practices.Ultimately, this leads most people who adopt the practice-dependent approach to presuppose that a basic structure will give justice its content and justification (Ronzoni 2009: 233).This is also why the practice-dependent approach tends to assume that the scope of justice is domestic, because states will have an existing basic structure.
Yet, Ronzoni proposes that it is also consistent with the practice-dependent approach to "… recommend the establishment of new practices under certain circumstances, namely when this is the only way of preserving the justice of other, already existing ones, and that the case of background justice is one such case" (Ronzoni 2009: 231).This also seems compatible with the Rawlsian division of labour argument.
"… [W]hat matters is that, given a certain account of when agreements are free and fair, that very account cannot be sustained in sufficiently complex social settings unless institutions with the power to maintain background conditions are in place" (Ronzoni 2009: 239).
On one point, however, Ronzoni and I begin to differ.When Ronzoni writes that specific types of interaction can give rise to the need to implement a global basic structure, she does not specify precisely what type of interaction that is.She simply says that it is the type of interaction that erodes background justice."When the socioeconomic interaction between agents is so complex and intense that it is bound to erode just background conditions in the long run," then we need to "… establish the appropriate institutions that will maintain just background conditions" (Ronzoni 2009: 240).She is clearly referencing Rawls's division of labour argument when she talks about interaction between individuals eroding background justice.The problem is that background justice is a property provided by the basic structure.Hence, if we do not have a global basic structure in place, it is hard to see how there could be an erosion of background justice in the first place.Additionally, "interaction that erodes background justice" on the Rawlsian account, is essentially normal social interaction suggesting that, on Ronzoni's account, almost any human interaction at the global level is going to count as eroding background justice.Ultimately, then, we need a slightly fuller picture of the conditions that give rise to the need for a basic structure.
Ronzoni comes to this interpretation of Rawls, I think, by overlooking two things: first, Rawls's justifications for making the basic structure the subject of justice, and, second, the concept of the ideal basic structure.Instead of talking about the division of labour argument, Ronzoni labels it the background justice argument, suggesting that she prioritizes the maintenance of background justice over all else.While the regulatory role of the basic structure is indeed important, it is also meant to ensure the freedom of individuals and distribute the main social goods of society.She continues, "[i]f we are to follow the background justice argument, the principles of justice as fairness are a response to the problem of unjust background conditions, rather than a given standard to check whether background conditions are problematic" (Ronzoni 2009: 243-4).This is true, but unjust background conditions do not equal an erosion of background justice.She seems to conflate injustice with an erosion of background justice.While the erosion of background justice might always count as an injustice, not all injustices are erosions of background justice.Injustice can exist without the erosion of background justice.
One way to rescue this proposal might be to suggest that the erosion of background justice would happen domestically.That is, if we start with the assumption that states have basic structures, and therefore have cases of background justice that can be eroded, we might be able to claim that global interaction could erode domestic background justice, thus giving rise to the need for a global basic structure.A contract between citizens of different states would have to lead to domestic interaction of the sort that erodes background justice.The problem with this thought (in addition to seeming not to be in keeping with what Ronzoni intended) is that if justice is eroded at the domestic level, it will be corrected over time.As Ronzoni recognizes, we have to continually adjust our institutions with reference to the ideal (domestic) basic structure, giving us no impetus to correct justice at the global level.Essentially, the problem is that we have no reason to assert that justice is global in scope, and consequently no incentive to establish a global basic structure.
Despite my disagreement with Ronzoni, her practice-dependent approach gives us an incredibly useful and relatively unexplored starting point.Additionally, as I have mentioned above, one of the problems with much of the global basic structure debate is its primary focus on distributive justice.By instead focusing on the establishment of a global basic structure to satisfy the demands of justice, we push the focus on distributive justice to the background, and look at it only once other demands of justice (the establishment of a basic structure) have been fulfilled.Using Ronzoni's practice-dependent suggestion as our starting point, if the trigger for a global basic structure is not the erosion of background justice, as she suggests, our job is now to figure out what it actually is.

V. Injustice and a Non-Ideal Global Basic Structure
The task now at hand is to establish an alternative theory for when the demands of justice will arise and thus trigger the need to establish a basic structure.Ronzoni proposes that it is the erosion of global background justice, signifying that the basic structure will automatically be global in scope.Yet, as we have rejected that thesis, but are also theorizing about a global basic structure, we must also ensure that our conditions will tell us when the demands will be global in scope, thus giving rise to the need for a global basic structure.(Note that on the practice-dependent account, the question of these circumstances arising will be an empirical one.) Ronzoni does not give us a definition for the content of justice, making it harder to figure out when the demands of justice will arise.If we claim that the content of justice is the same as domestic justice as fairness, we come to the same problem that Rawls faced in The Law of Peoples.The principles of justice were agreed upon and mutually justified by citizens of liberal democratic state qua citizens of that society.The world at large is not liberal, and we do not want to impose liberal values on non-liberal societies.Thus, it does not seem immediately plausible to assert that the content of global justice will be the same as justice as fairness.
We started, in the first place, with the Rawlsian justifications for making the basic structure the subject of justice, so it makes sense to continue in a Rawlsian vein.These justifications are already value-laden in some way, but not necessarily laden with liberal values.The basic structure is the subject of justice because it allows the individuals to be most free, and we value not restricting their freedom by directly applying the rules of justice to them.We can make this claim without any reference to liberal freedoms or rights, and simply claim the importance of free agency in a conception of justice.Second, the other justification is that the institutions of the basic structure will pervasively impact the lives of the individuals it affects.This in itself has no liberal implications.So essentially, all we are looking for in fleshing out the content of justice is "… an account of when agreements between individuals or other noninstitutional actors are free, and when the social circumstances under which they are reached are fair" (Ronzoni 2009: 238).This sounds like Rawlsian justice without the liberal overtones.
Thus, now that we have a picture of the content of justice, we can move on to exploring what circumstances will give rise to its demands.This may perhaps sound unexceptional, but the thing that is going to give rise to demands of justice is injustice.After all, looking once again at the parallels between the domestic and global cases, on the practice-dependent account justice as fairness arises because society is unjust.We want society to be structured in a way that is fair.So empirically, we look at the circumstances that give rise to these injustices, and apply the rules of justice to correct them.
The parallel to domestic justice complicates things, because, as noted above, whether the basic structure requires justice before its demands arise or not, the fact is that a basic structure did pre-exist the demands of justice.Therefore, we can turn the question around, and look not at what circumstances give rise to global justice, but rather ask when the basic structure will be required to fulfil justice's demands.This brings us back to the justifications for making the basic structure the subject of justice in the first place: 1) it allows individuals to remain free in their interactions; and 2) it will pervasively impact on their life chances.
Empirically, institutions already exist at the global level.Some of these will pervasively impact on people's life chances, but that is not a sufficient condition for calling global institutions a "basic structure".We must establish them as part of a global basic structure when interactions on the global level are complex enough, and interdependence is full enough to give rise to injustices that can be remedied by the basic structure.The idea is, as Ronzoni stated, that these institutions already directly affect individuals and their interactions.At the point when these institutions begin impacting on the freedom of individuals to interact against a fair backdrop and pervasively impact life chances, then the demands of justice will arise and we will need to establish a global basic structure.This is essentially because, as global interdependence and global interaction increase, we get an increase in the possibility of unjust interactions.
Ultimately, this shows that, in contrast to many theorists' ideal starting points, we can get a picture of global justice starting from a non-ideal perspective.It is perfectly coherent to look at empirical facts to see what conditions are going to give rise to considerations of justice.Yet, as Rawls has shown us, non-ideal theory in the form of practice-dependence is non-motivational without an ideal guiding it.Thus, while the establishment of a global basic structure will depend on practices, in order for it to be a vehicle to motivate the changes justice requires, it must have an ideal aspect.
Here, I want to raise two related objections to my position.The first is that theorists have unsuccessfully asserted that interdependence at the global level gives rise to injustices that must be corrected.The response here is simple: those theories assumed a pre-existing basic structure, or that there was already a sufficient amount of interdependence at the global level to warrant calling it a basic structure.That is not what I am claiming.I am asserting that when we reach that point of sufficient interdependence, the demands of global justice will arise, requiring us to establish a global basic structure.
The second objection has to do with my non-ideal starting point.Ultimately, it claims that justice must be an ideal, and that starting with empirical facts obscures the true form of global justice and its demands.My answer to this continues from my answer to the first part.Theories that assert a pre-existing global basic structure tend to start from an ideal position (assuming the counter-factual existence of a global basic structure), and have failed for that reason.In approaching the problem from a practice-dependent account, we look at the circumstances that give rise to the ideal of justice, and then apply them to a non-idealized subject.Thus, I am not denying the importance of ideal theory in the way it motivates us to achieve an ideally just society; I am simply asserting it cannot be our starting point.
Ultimately, when existing global institutions start to give rise to injustices in the form of restricting free agency while also pervasively impacting on individuals' life chances, the demands of justice will arise and require the establishment of a global basic structure.

Conclusion
Let us now assess what we have established.In Section I, I said that the basic structure is the subject of justice and comprises the main social and political institutions of society and the way in which they fit together to regulate social interaction.Additionally, Rawls gives two mutually supportive justifications for making the basic structure the subject of justice: the division of labour argument, which means that the principles of justice do not apply directly to individuals, and the pervasive impact justification, which argues that the basic structure is the subject of justice because of the profound effects it has on people's life chances.I then established in Section II that, given these two justifications, there need not be a basic structure already in place for the demands of justice to arise.Rather, when the demands of justice do arise, we must establish a basic structure to fulfil those requirements.This means that we are not reliant on a pre-existing basic structure that will limit the scope of justice to the domestic sphere.Instead, we can assert that if the demands of justice are global, our basic structure must be global.Section III introduced us to the practice-dependent approach and showed in what ways it was plausible to read Rawls as taking a practice-dependent approach to his theory of justice.We also showed that even the ideal parts of Rawls's basic structure have a root in non-ideal theory.Section IV presented an approach to establishing a global basic structure from a practice-dependent standpoint.I also expounded on my own suggestion for the conditions under which we must establish a global basic structure: when currently existing global institutions begin impacting on the freedom of individuals to interact against a fair backdrop and pervasively impact life chances, then the demands of justice will arise and we will need to establish a global basic structure.Ultimately, I have shown that on a Rawlsian interpretation of the basic structure, one does not have to exist before the demands of justice can arise.Therefore, on a practice-dependent account, we can establish that in certain circumstances, namely in the case of pervasive injustices brought about by global institutions, the demands of justice can arise at the global level requiring the establishment of a global basic structure.This conclusion also has implications for the use of ideal and non-ideal theory in theorizing about global justice, because I have proposed that, contrary to common practice, global justice, like domestic justice, can originate from non-ideal theory.

Notes
1 Though it is outside the scope of this paper to justify the basic structure as the subject of justice, for arguments as to why the basic structure is not the proper site of justice, see Cohen (1997) and subsequent responses in defence of the basic structure as the subject of justice (Scheffler 2006, andRonzoni 2008).
2 Note here that Rawls does not comment on whether the basic structure is an existence condition for justice.The fact is that it was not relevant for Rawls's purposes because he started with an existing society with an existing basic structure.It did not matter if the basic structure had to precede justice, because, empirically, it did.
3 Early in its conceptual development, there was a debate as to which type of theory was correct for political theorizing.This debate remains in the practice-dependent versus practice-independent accounts of justice, but now most theorists seem to have agreed that both are needed in a political theory; the question simply becomes how much of which type at what stage in the theorizing.For a useful picture of the current state of the ideal/non-ideal theory debate see Laura Valentini (2011). 4See Stemplowska (2008), Robeyns (2008), and Swift (2008). 5Essentially, I echo Valentini's (2009) conclusion that we cannot idealize the subject (in this case, the basic structure) of justice.
6 This is the route that early Rawlsians like Thomas Pogge (1989) and Charles Beitz (1979) took in an attempt to globalize the scope of justice. 7Institutions that make up the basic structure include: "… the political constitution, the legally recognized forms of property, and the organization of the economy, and the nature of the family" (Rawls 2005: 258). 8Here I want to note my departure from Scheffler's discussion of the basic structure on two related points.First, Scheffler identifies three reasons why the basic structure is "primary" (Scheffler 2006: 4-6).This is not quite the same as justifications for making the basic structure the subject of justice.Scheffler's discussion of making the basic structure primary relates to the reasons why justice applies to the basic structure instead of individuals, but are not reasons for making the basic structure the subject of justice, as such.Secondly, Scheffler identifies two "divisions of labour" at work in Rawls: a "division of moral labour" and "institutional division of labour" (Scheffler 2006: 9).For my purposes, the division he makes need not be adhered to, and a general "division of labour" justification will suffice.It is worth nothing, however, that Scheffler's "institutional division of labour" relates closely to the ideal basic structure, a concept to which he does not make reference. 9"… [T]he conditions necessary for background justice can be undermined, even though nobody acts unfairly or is aware of how the overall result of many separate exchanges affects the opportunities of others.There are no feasible rules that it is practicable to require economic agents to follow in their day-to-day transactions that can prevent these undesirable consequences….To conclude: we start with the basic structure and try to see how this structure itself should make the adjustments necessary to preserve background justice.What we look for, in effect, is an institutional division of labour between the basic structure and the rules applying directly to individuals and associations and to be followed by them in particular transactions" (Rawls 2005: 266-9). 10Here, it should be noted that, in addition to a division of labour between the basic structure and individual actions, Rawls also writes that he imagines a division of labour between parts of the basic structure, with each part corresponding to one of the principles of justice."The first principle of equal liberty is the primary standard for the constitutional convention.Its main requirements are that the fundamental liberties of the person and liberty of conscience and freedom of thought be protected and that the political process as a whole be a just procedure.The constitution thus establishes a secure common status of equal citizenship and realizes political justice.The second principle comes into play at the stage of the legislature.It dictates that social and economic policies be aimed at maximizing the long-term expectations of the least advantaged under conditions of fair equality of opportunity, subject to the equal liberties being maintained.At this point the full range of general economic and social facts is brought to bear.The second part of the basic structure contains the distinctions and hierarchies of political, economic, and social forms which are necessary for efficient and mutually beneficial social cooperation" (Rawls 1999: 174-5).
11 The idea of an ideal basic structure also speaks to the concept of reflective equilibrium and public reason.Instead of establishing justice at the beginning through the bargaining process behind the veil of ignorance, we now come to a more practical picture of society, realizing that justice will be an on-going process that will require constant adjustments.
12 Here we begin to see Rawls moving away from the ideal of compliance as ideal theory, and adding in other idealizations.Yet, the idea of compliance still plays a big part in his justification of the basic structure.The ideal basic structure clearly assumes full compliance.
13 I want to note here that we have been working under Rawls's assumption of a closed society.The basic structure arguments made above rely on Rawls's assumption of the self-sufficiency states.This means we assume that the scope of justice is limited domestically to the body of individuals who form the citizenry of the closed society. 14In two of Abizadeh's interpretations of the basic structure, he finds that they do require a basic structure to be in place before the demands of justice can arise (but he also concludes that their scope should be global instead of domestic).I, however, am not interested in proving that there is an existing global basic structure; rather, I am interested in cases "where the scope of justice isn't limited by an existing basic structure" (325).Yet, on Abizadeh's third interpretation of the basic structure, he argues that its justification-that the basic structure is the best way to bring about justice-means that the basic structure is an instrumental condition of the realization of justice.My interpretation of the justification for making the basic structure the subject of justice is slightly different.I see the two justifications given above as being mutually supportive (Abizadeh sees them as distinct). 16Many theorists now agree that a combination of both ideal and non-ideal theory is the best way to devise a theory of justice.They will endorse non-ideal theory at the level of implementing a theory of justice, but want to use ideal theory to justify their conception of justice. 17Arguably, because there are still not steadfast and universally accepted definitions of ideal and non-ideal theory, suffice it to say that it is a primarily non-ideal account. 18Here, it should be noted that Sangiovanni's institutionalism is not to be confused with the proponents of global justice who label themselves institutionalists (as opposed to interactionalists).The two are related in their commitment to the way in which institutions affect justice, but Sangiovanni's practice-dependent institutionalism better lends itself to a statist position than a cosmopolitan one.