In Michael Sandel's Justice, he gives a wonderfully clear account of Kant's and Rawls' social philosophy. In placing Kant's views of the social order Sandel reviews the whole tradition of social contract theories. Sandel then shows how Rawls ingeniously solves the problem of the historical implausibility of an actual social contract by turning the whole metaphor into a thought experiment. We do not have to believe that there ever was an actual social contract, Rawls argues - we can imagine ourselves in an "original position" as generic human beings without particular qualities to see what kind of social contract we would make in such circumstances.
I was surprised at how much space Sandel gave to the argument about whether there was an historical social contract, and if so whether a contract made by our social predecessors could really bind us. This book grows out of his long and rich experience teaching about justice to Harvard undergraduates. Do those smart kids really think that Hobbes, or Rousseau, or Mill thought that society was born in an actual gathering in the woods?
I assume from the fact that Sandel takes the time to explain how the idea of a social contract works without entailing an historical contract-making that this is an important issue to his students. My best guess is that what they are concerned about is not the historicity of the event. Rather, they believe that if they or their predecessors did not consent to society, then they are not bound by it. No agreement, no contract.