I started a version of this post a couple weeks ago, but since then the dispute between libertarians about the place of "social justice" in their philosophy has become white-hot, and I might as well jump in. The debate kicked off with the responses to Matt Zwolinski and John Tomasi's lead essay in the April Cato Unbound, especially David Friedman's. Zwolinski and Tomasi of course stood up for themselves at Cato Unbound, but the debate has spread far and wide--farther and wider than I'm able to follow, I'm afraid. Thankfully, the Bleeding Heart Libertarians blog is your online one-stop-shop for the great libertarian war over social justice. On Monday, Jacob Levy wrote a typically insightful post, "Against Social Justice," which comes close to much of what I was going to say, especially this:
The Rawlsian definition of social justice is constructed on the idealized model of a closed society, entered only by birth and exited only by death. The sense in which both Jason and John use “social justice” bears the legacy of that deep moral mistake.
The basic institutions of society must sufficiently benefit all, including the least advantaged and most vulnerable members of society.
Like Bryan Caplan and Will Wilkinson (different as those two are!), I think that we who care about freedom should be deeply outraged by the wrongs done by the system of border controls to keep people in poverty. I think this is a central, defining issue for bleeding-heart libertarianism. And the language of social justice renders it invisible, because the poor people being hurt are not already “members” of the “society” whose institution are being evaluated. It doesn’t just say that the harm done to the non-members is less important than the effect on the poorest members; it denies that the former is a consideration at all. When combined with the “justice is the first virtue of social institutions” mindset, that leaves my bleeding-heart libertarian colleagues in the paradoxical position of hiding from view arguably the greatest-magnitude source of state harm to impoverished human beings.
Yet I agree entirely with Jason Brennan's reply to Jacob:
1. You can think social justice matters without claiming that social justice (or even justice more broadly) is the first virtue of institutions.
2. I don’t take it for granted that “society” is co-extensive with the modern nation-state. Contrary to Rawls or Sam Freeman, I think immigration restrictions violate social justice. Immigration restrictions impose poverty, suffering, pain, and death on some of the most vulnerable people in the world. From my point of view, if you oppose free immigration (or something very close to it), then it had better be because you have an empirical disagreement with me about the expected consequences of free immigration. Otherwise, any claim to be concerned about social justice or the well being of the poor is mere pretense.
Jacob says the language of social justice renders invisible the profoundly immiserating injustice of the status quo system of border controls. I don't think this is quite right. I'd say the assumption of "analytical nationalism" -- taking it for granted that the nation state is the relevant level of socio-political analysis -- not the language of social justice, renders the injustice of immigration policy invisible. Now, it's certainly true that almost all contemporary conceptions of social justice are built atop the error of analytical nationalism. As for Rawls specifically, he's guilty of a nationalist conception of "the basic structure" of institutions to which the principles of social justice are to apply.
I agree with Jason (and Kevin Vallier) that social justice and analytical nationalism can come apart. That is to say, there is nothing unintelligible about a non-nationalist, cosmopolitan conception of basic structure and social justice. The obvious difference between nationalist and cosmopolitan conceptions of social justice is that, according to cosmopolitan conceptions, the basic structure encompasses the institutions that shape international patterns of trade and migration, and principles governing those institutions must be justified to the international motley who act within them. This is messy. Analytical nationalism is neat; it's nice to pretend that the basic structure fits inside the jurisdiction of the nation-state like a hand fits inside a glove. It's easier that way. But this sort of analytical convenience is bought at the price of irrelevance to the messy real world, even if we see ourselves as doing "ideal theory".
In Free Market Fairness, John Tomasi calls the libertarian aversion to social justice "social justiticis." The present controversy shows that this is the perfect term. The aversion is more an allergy than anything. For those who continue to sneeze at every mention of "social justice," Kevin Vallier's latest post might help.