Between the Todd Seavey mini-bru-ha-ha, the Cato Unbound discussion, our awesome symposium on libertarianism and land, several recent BHL posts, one from Bryan Caplan and one from Ilya Somin, we keep coming back to a single question: What is social justice? A number of us have provided definitions and refinements. I thought I’d add a few conceptual notes so that no one can say we BHLers weren’t clear about what social justice consists in, along with some replies to Bryan and Jacob.

Social Justice, The Concept 

How does the term “social” modify the term “justice” such that we are left with an important and illuminating concept that is a kind of justice that libertarians should accept? I’m going to give a Rawlsian answer to this question by holding that social justice is justice with regard to the arrangement of a society’s basic structure. Let’s take the second term first. Rawls defines a society’s basic structure as follows:

By the basic structure I mean a society’s main political, social, and economic institutions, and how they fit together into one unified system of social cooperation from one generation to the next (PL, 11).

Rawls’s theory in Political Liberalism is meant to apply to modern constitutional democracy, such that the subject of social justice is the structure of the modern constitutional democratic state and the institutions it governs. For Rawls the basic structure is “the first subject of justice” (257). He states again that,

The basic structure is understood as they way in which the major social institutions fit together into one system, and how they assign fundamental rights and duties and shape the division of advantages that arises through social cooperation. Thus the political constitution, the legally recognized forms of property, and the organization of the economic, and the nature of the family, all belong to the basic structure (258).

So the basic structure is one great big social thing and serves as a subject of evaluation. A basic structure is socially unjust when it is not arranged in accord with principles that can be justified to each reasonable comprehensive doctrine (kind of like my discussion of public reason, but not the same).

Note that Rawls’s primary criticism of libertarianism in his later work is that “Libertarianism Has No Special Role for the Basic Structure” (262-264). I don’t throw this in just for fun. Rawls’s argument is that libertarians fail to acknowledge that societies must have a system of public law which inevitably makes use of coercion and that this use of coercion raises a distinct justificatory problem. Now whether Rawls is right to charge libertarians with this is another matter. The point is just that to embrace social justice is to embrace the basic structure of a given society as a subject of evaluation as just or unjust.

Now to the second term, the idea of an “arrangement.” Arrangements are made in accord with principles, like Rawls’s two principles of justice.

In my Gausian view, we evaluate basic structures piecemeal in accord with evaluations of particular rules and practices. Different rules and practices can be publicly justified in accord with different criteria. The justification for the basic structure, social justice, that is, is a spontaneous moral order that arises from the justifications of its parts. On the Rawlsian view, however, we stay holistic and evaluate the basic structure in one complex, mostly “top-down” process.

But on both views, for a basic structure to be arranged in a particular fashion is to say that it rests on certain principles and shared ideas that are the subject of moral and political evaluation. To put it another way, we see that a basic structure is arranged when it has a rational structure.

Social Justice, Conceptions 

So there’s the concept of social justice. Now let’s turn to conceptions. Consider the conception of social justice articulated by my friend, fellow Arizonan and co-blogger Jason Brennan:

Advocates of social justice believe the moral justification of our institutions depends on how well these institutions serve the interests of the poor and least advantaged.

I don’t think this statement describes the concept of social justice but rather the general range of conceptions of social justice. It is conceptually coherent to have a conception of social justice where the rich are given special preference by a society’s basic structure. It’s a bad conception, but it is still a conception. Instead, J simply (and correctly) observes that nearly all conceptions of social justice hold that the arrangement of a society’s basic structure must be based, among other things, on principles that require institutions to serve the interests of the poor and least advantaged. As J says, conceptions of social justice vary but they usually include these elements.

Now, what could justify this focus on the poor and least advantaged? It depends on how one justifies a particular conception of social justice. On the Rawlsian view, we’re looking for a set of principles that persons have reason to endorse given their diverse comprehensive doctrines (if you want me to be more precise here, ask). As David Friedman has noted, Harsanyi used a similar justificatory device to generate the principle of average utility as a key principle of justice. In my view, Rawls got the first principle right and Harsanyi more or less got the second principle right. But on the Rawlsian view, the liberty principle plus the average utility principle comprise a conception of social justice, namely a conception of the principles that arrange the basic structure in particular configurations. Most of us accept versions of sufficientarian and prioritarian principles of social justice. If people want more specific versions of what sufficiency and priority come to, much has been written on the topic, so I suggest you begin there.

Levy against Social Justice 

On Monday, Jacob Levy made two complaints against conceptions of social justice advanced by friend of the blog John Tomasi and my fellow co-bloggers Matt Zwolinski and Jason Brennan. His first complaint is that talk of social justice suggests a kind of justice fundamentalism, a claim that justice is the overarching metric of evaluating institutions. But, and I don’t think Jacob will disagree with this, nothing about the concept of social justice or the range of conceptions of social justice discussed here succumb to that problem. And we can immunize Jason’s definition above by adding replacing “depends” with “depends, among other things.”

Jacob’s second complaint is that Rawlsian conceptions of social justice, by focusing on “members of society” are objectionably nationalistic. But notice that focusing on the “basic structure” of a society does not restrict us to nationalism. For instance, if a global economy has a basic structure, then it too is the subject of social justice as I have defined it. Again, I don’t think Jacob will disagree.

Caplan against Social Justice

Finally, to Bryan. Bryan runs familiar hard libertarian counterexamples to BHL concerns about the least well-off largely by testing our claims about social justice against micro-level examples, such as an island with ten people, many of whom wish to redistribute the fruit of “Able Abel’s” labor. Many replies suggest themselves, but let me offer one.

A conception of social justice helps to specify what counts as the fruits of one’s labor in a massive system of social cooperation which depends on a vast range of conventional rules. That’s why Bryan’s testing social justice against micro examples is problematic, because it masks indeterminacy in macro-level social norms that we want a standard of social justice to evaluate. Consider the many “incidents” of property rights, such as whether homesteading land means that you own the air miles above your plot. In Bryan’s case, disputes over the “height” of our property don’t matter. But in our complex society, such disputes matter a lot because the efficiency of airline travel depends on how they are resolved. A conception of social justice can help us determinate what system of efficient property rights can be morally justified in this case of indeterminacy.