Martin O’Neill and Thad Williamson had an article in the Boston Review outlining some of the radical components of Rawls’ philosophy. The short thesis is: “to treat Rawls simply as a defender of Democratic Party liberalism and the welfare state—as he is widely regarded—is to misread him.” The authors then go on to explain Rawls views on political freedom, property-owning democracy, and so on. You should read the whole thing.
It is a great read because it takes on the conventional wisdom on Rawls. I first read Rawls some years ago, and I did so on my own, never in a class with a professor teaching the book. Later when I would engage with those who had learned Rawls in a class, I was shocked to find that he was regarded as merely the proponent of welfare state capitalism (or Democratic Party liberalism as described above). This is apparently how he is taught.
Reading him in my own bubble without outside guidance, I got something quite different from Rawls, something much closer to what this article outlines. Rawls spends a significant chunk of his magnum opus describing a socialist system that is compatible with his theory of justice. How you read that and take away that he is a Democratic Party partisan, I will never know. He is not necessarily anti-capitalist, as he provides for a possible society where capitalism is present that still coheres with his theory. But, as the authors point out in the above article, this capitalist incarnation of his theory requires widely dispersed wealth ownership, not merely food stamps and Social Security.
In my view, Rawls’ most radical move is also his most understated. Sure he goes into all sorts of institutional regimes, talks about political liberty, and all the big topics folks care about. But really the crucial move for Rawls is his departure from the liberal tradition of treating property rights (or property systems) as something which are internally justified. That is, the tradition that preceded him largely took the issue of property to be something to be interrogated a priori by itself. Rawls instead declares property systems to be justified externally by their relationship to an overall system that satisfies the demands of distributive justice.
So instead of abstractly reasoning about the mechanics of property rights, Rawls first figures out what is distributively just, and then he uses that determination as the standard to measure whether a given property scheme is just. That is why Rawls can have multiple property systems that he deems distributively just (property-owning democracy and liberal socialism): if they both achieve the distributive goal, then they both are just even if different from one another in form. The brilliance of this move is that it finally liberates discussions of justice from the completely intractable and rather ridiculous discussions of property. Under Rawls and the tradition he spawns, the property system component of justice is a purely secondary matter, something that can be constructed in multiple ways, and something which is only really justified derivatively.