Preferences and the original position

Norm writes:

Maybe top executive pay and other very lavish remuneration can't be justified as what fair-minded and reasonable people would agree to in an Original Position.

This invites the retort: so what? Our preferences change with circumstances. What I wanted as a 25-year-old is different from what I want now. It would be silly to live my life according to what I wanted then. So why should we organize our collective lives according to what we’d want in a different situation?
Let’s say Norm is right and in the Rawlsian original position we’d all agree that CEOs should only get 2-3 times average earnings.
But when we shift from the original position to the real world, our preferences change. The guy who in the OP thought a 2-3x multiple of average earnings was a good wage might decide that it’s not so handsome after all now that he‘s CEO; the more we have, the more we want. And non-CEOs might want to pay him more, to bribe him not to mismanage the firms.
Why should our preferences in the OP dominate these preferences?
For that matter, would our preferences in the OP actually be for a multiple of 2-3x? Mightn’t we instead anticipate the combination of adaptive preferences and efficiency wage motives we would have in the real world, and agree upon a greater inequality?
What I’m asking here is not the old question: why should we agree to a hypothetical social contract? Instead, it’s about preferences: why should preferences in one situation have any relevance to a different situation?
The standard answer is that our preferences in the OP are rational and disinterested. But, in my example, they are also rational in the real world - and, for all but CEOs, disinterested too.
One could also reply that OP preferences are “clean” in the sense that they exclude the preferences that result from adaptation to our actually-existing institutions.
But this invites the question: how can we tell what such cleansed preferences look like? Does a context-free preference even make sense?
Another answer here is that in the OP, we wouldn’t content ourselves with merely asking what sort of salary to pay CEOs. We would also ask: should we really consent to the existence of institutions which generate such inequalities? 
But how could we answer this question? The most natural thing to do is to import into the OP our knowledge of capitalist history - that hierarchical firms, tempered by market forces, have generated wealth but also crises and inequality - and ask whether we‘d want this.
This, however, entails two questions. One is: in doing this, are we guilty of sampling bias? Maybe there‘s an alternative possible world of a viable market socialist/cooperative world that we haven‘t experienced because of bad luck or historical accident - one of the accidents being that it just hasn‘t come about yet.
If we under-sample this alternative, we might - in the OP - be more generous towards hierarchical capitalism than we would otherwise be. How can we tell whether we’re making this mistake or not?
The other question is: if we import our experience of hierarchical capitalism into the OP, mightn’t we also import at least some trace of the preferences (eg instrumentally rational self-interest) such capitalism has generated? If so, the OP isn’t truly O.
We are, then, left with the question: what does OP-type thinking do that other forms of political thinking not do?

Comments

[Charlie]

Original position arguments aren't intended to show anything about our preferences: they're intended only to show which arrangements will count as just. At the same time (if I've remembered it right) facts about human psychology generally _are_ known behind the veil, at least they're known in Rawl's version of it. So, if, given those facts about human psychology, only an inequality of wage differential x will serve to maximise the benefit to the least well off through motivating the talented and productive, then wage differential x will be just.

[Charlie]

And reading Norm's post, it _seems_ as though he's making a claim about human psychology: that even a modest wage differential will in fact be sufficiently motivating. Again, I don't see that the original position argument is relevant; it's not the job of that sort of argument to settle the issue of what in fact motivates people.

What might be relevant here is Cohen's argument in 'Rescuing Justice and Inequality', which says - very approximately - that it's wrong for the rich to argue that they are in fact motivated by higher wages and that the poor should therefore accept such inequality, even if what the rich say about themselves in this regard is true. The point is that it's conceivable that the rich _make_ this fact true of themselves.

[Charlie]

(Should be 'Rescuing Justice and Equality' - sorry.)

[chris]

Ta, Charlie. My argument does lead to Cohen's, I think.
Assume the rich say - truthfully - that they need vast sums to motivate them to supply their labour, and that this fact is known in the OP. Actors in the OP then have a choice: to either accept such inequalities on the grounds that the rich's labour actually makes the worst off better off, or to rearrange institutions to take the rich redundant.
It's a good question which to choose. But I don't see why we need an OP to address it.

[Rachit]

Two interesting reads.
@Charlie:
As I understand it (my preface to any comment on Rawls), you are correct in saying that the OP is not about preferences but is about the shape of a just framework within which we exercise our preferences. However, the argument doesn't stop there: Rawls's outlook says that that just framework is more important than preferences. So it might be wrong to say why do preferences of the OP matter but it can still be asked why this just framework matters more than all preferences. In Rawls, the OP's decision takes priority over any amount of agreement in real society.
@Chris
The OP is of course just a device to exclude certain types of reasoning from deciding this just framework - an exclusion of special bargaining. So according to Rawls being rational in the real world isn't a good reason on which to base a just framework because your position will dictate your preference.
However you raise two crucial questions: are we psychologically capable of imagining ourselves without our background? And even if so, would that context-free agreement (borrowed from Kant) even matter?
My own view is no and no. I think that in racist 19th century Britain, a similar thought experiment would yield a racist OP, and if social practices were ruining even a 'perfectly just' institutional system, that wouldn't be a good or just society. This doesn't make Rawls irrelevant but perhaps we should use the good parts of his reasoning (impartiality) and add it to other things we value.

[A K Haart]

"Does a context-free preference even make sense?" Surely not.

"what does OP-type thinking do that other forms of political thinking not do?"

Very little other than provide a starting point and therefore a point of reference. I see it as a framing device, much like an agenda item.

[Keith]

It is a form of special pleading to get all to agree that in theory we should want equality. And actually compels all actual persons to do so in practice. I can always say equality is good in theory but I don't give a damn in practice as can the electorate. There is always an excuse why equality should apply only to other people. So what use is the theory when I will in fact know that my rich aunt is probably going to leave me the family fortune. Hence I will always have an incentive to reject inheritance tax?

[tim]

Perhaps we're missing here the notion of freedom, or as Rawls puts it the idea of persons as free and equal. The OP is supposed to model the conception of persons as free and equal, and then the principles that "come out" of it (however they're supposed to) are supposed to be fundamental for the basic structure of society, in the sense that any society that doesn't live up to them can justifiably be changed. So the OP is a way of thinking about how a society should be if it is to be conducive to human freedom (at least, Rawls' Kantian-liberal understanding of human freedom).