If you're looking for reasons to not be a Rawlsian, please read my colleague Tom Palmer's terrific new paper “No Exit: Framing the Problem of Justice” [pdf] for a profoundly illuminating discussion of why Rawls's theory justice makes sense only within his illiberal and fantastically unrealistic zero-mobility assumption.
There are a lot of good libertarian criticisms of Rawls, but I think Tom's may be the most damning I have ever read, probably because there is nothing really especially libertarian about it, unless challenging deeply question-begging assumptions about the nation-state as the inevitable and inescapable framework of justice is essentially libertarian. If I were to make the argument in one sentence it would be this: Rawls provides accounts of liberalism and justice that are essentially nationalist/anti-cosmopolitan, but since nationalism and anti-cosmopolitanism are pretty obviously illiberal and unjust he fails at a very fundamental level.
Here's a taste:
The sleight-of-hand involved [in establishing “justice as fairness”] is remarkably similar to that typically involved in specification of the choice situation governing the provision of public goods. Once a decision has been made to produce a good on a non-exclusive basis, it is then asserted that the good cannot be produced through voluntary, uncoerced cooperation. Or the demonstration begins by assuming the existence of a good from which consumers cannot be excluded (or can only be excluded at some cost), when the problem is to produce such goods in the first place. Assuming that the good exists is hardly a solution to the problem of how to produce it. Similarly, by excluding exit or choice among options as an option, the problem of distribution of rights and obligations is converted into a pure bargaining game, which sets the stage for Rawls’s voluminous writings and the many jots and tittles added by his followers. By eliminating such options, Rawls does not solve a problem of justice or fair division; he creates it.
Tom's paper is limited to speaking to rights of emigration or exit from a polity, but it points to a rather more general lesson. That it is impossible to discuss questions about the (in)justice of limiting the right to travel across state borders, or to discuss questions of the justice of the distribution of national citizenships within a pure Rawlsian framework goes to show how useless is that framework for thinking about questions of equality and distribution in an increasingly globally interdependent world. It also goes to show just how captive is Rawlsian political theory to what I think are a set mistaken, distinctively 19th and 20th-century background assumptions about the nation-state as more or less co-extensive with society.
Added: Tom's essay may be found in Ordered Anarchy: Jasay and His Surroundings, ed. by Hartmut Kliemt and Hardy Bouillon (London: Ashgate, 2008)