Tell you to ask Kant, or at least, so I am briefly about to argue. Last week, I gave a paper in which I argued that right libertarians - and implicitly with them a substantial number of post-Rawlsian political philosophers - make a fatal mistake about how to go about the justification of political institutions. Basically, the thought is that they, at least if Robert Nozick is representative, assume that our intuitions about single-shot, often two person cases unproblematically give us direct instruction about what we ought to do when designing political institutions which are defintionally not single-shot - what it means to be an institution is to create a structure within which many interactions between many people take place, over a lengthy period of time. The Wilt Chamberlain example says nothing about what happens the day after Wilt Chamberlain gets his $250,000, for example. Since there are all kinds of effects which are related directly to an institution being extended both over time and over space, this misses any normative significance which might attach to those effects, and therefore gets what institutions ought to look like wrong.
Anyway, whether or not this claim is correct, I attributed it - not prominently enough, I think, actually - to Rawls in the paper; indeed, I regard what I took thirty-odd pages to say as an extrapolation of remarks Rawls made in about two pages in Political Liberalism which have been unfortunately largely ignored. Obviously, so far as I'm concerned, that's pretty much enough to show it's correct, but just in case anyone still wasn't convinced, I want to invoke another name in moral philosophy to demonstrate that this isn't just me and Jack being totally crazy, that of Immanuel Kant. Kant notoriously says that a maxim is only one which it is morally permissible to act on if it can be shown to be a Categorical Imperative. Maxims that are Categorical Imperatives all, at least so far as Kant claims, fulfil three criteria: the criteria of the universal law, the criteria of treating humanity always as an end in itself, and the criteria of the kingdom of ends. Now, ignoring the idea that only maxims which treat humanity as an end in itself rather than solely as a means for the time being, notice something about the other two criteria. They both invoke institutions.
The idea in the criteria of the universal law is that something is only a permissible maxim on which to act if it could be willed as a maxim by everyone when they are relevantly similarly situated. Kant obviously from his discussion in the Groundwork takes this to mean it being an institution; his examples of maxims failing that test are related to either it being impossible to act on the maxim when it exists as an institution - as in the case of telling false promises, where we when all tell false promises, there is no institution of promising for us to be parasitic on - or it being impossible to will the institution which would result from the universal willing of the maxim - as in the case of a refusal to give assistance, where we come to the realisation that if no-one ever helps us, we will not be able to reliably achieve our ends. Likewise, in the formula of the kingdom of ends, the thought involves testing a maxim against the possibility of whether or not it could exist as a viable and acceptable institution, one in which all participated, and all were treated as lawmakers. In that case, though, Kant, with Rawls, thinks that single-shot, two-person cases are not going to tell us anything helpful about what we ought to do. Now, I think that Kant and Rawls likely actually think this for different reasons, but isn't it interesting that Nozick, who invokes Kant in favour of his view, so singularly fails to notice this relatively central aspect of Kant's own doctrine? Equally, isn't it interesting that a whole bunch of deontologically-inclined egalitarians are similarly uninterested in the claims of what must a central figure in that tradition?
PS: Wasn't Spooks good?