Sunday, July 16, 2006

A problem with libertarianism.

I'm going to finish up today with a problem for classic libertarian political theory. This theory is founded on the idea that each individual has the (moral) right to dispose of his or her legitimately acquired property in any way he or she sees fit. There is also a suggestion under the surface that these individuals, since they have no obligations to others, would be morally permitted to look out only for their own interests.

That's the background. Here's the problem. Let's say we have two people, Algernon and Barbary. Algernon has $1B (billion) in assets that he could use to buy goods, while Barbary has $1K (thousand) in assets that he could use to buy goods. Both want the same bundle of goods, which we will represent by X. A third actor in this little problem, Clancy, has X, and knows that Algernon has $1B and Barbary has $1K. It would, it seems, be perfectly permissible for Clancy to charge Algernon 1,000,000 times what he charges Barbary, because Algernon has 1,000,000 times more assets at his dispoal than Barbary, yet wants the same goods. To put the reason another way, since Algernon and Barbary both want X, but Algernon has far more that he could use to pay for it, Algernon should be willing to pay far more than Barbary.

To compound the problem, suppose Algernon balks at paying more, and Barbary buys X. Algernon may then attempt to buy X from Barbary, but why shouldn't Barbary sell X to Algernon for whatever he (Barbary) paid for it plus 1,000,000 the original asking price? That is, why shouldn't Barbary reason exactly the same way as Algernon, and also attempt to recoup what he's lost in acquiring X?

On the face of it, the natural (libertarian-styled) response is to say that's it not fair to "punish" Algernon for accumulating more assets than Barbary, but this notion of fairness cannot have any traction. Clancy (and Barbary, in the extended version of the example) both have good reason to try to acquire as many assets as they can; and, furthermore, have no good reason to treat Algernon more favourably than Barbary (this is particularly true for Barbary himself).

It could be argued, of course, that Clancy (or Barbary) might not want to ask for 1,000,000 times more if Algernon is willing to give 100 or even 10 times more -- by a simple cost-benefit analysis, Clancy (or Barbary) will still win, even if Algernon hoards his assets to a greater degree than Barbary does. But this misses the point that, given Algernon has so much more assets than Barbary, it follows that Algernon's assets are worth less to him than Barbary's. If the price of X is $10, then Clancy might charge Algernon $10M for it, while Barbary $10 for it; and, if Barbary buys it and is then asked to sell it to Algernon, he would charge $10,000,010 for it. Note, though, that paying $10M for X would leave Algernon with $990M at his disposal, while paying $10 for X would leave Barbary with $990 at his disposal. That is, although Algernon would pay more than Barbary under either situation, he would also retain more than Barbary, and hence would have more assets left at his disposal than Barbary. (And, under the other version, Barbary would end up with $10,001,000, while Algernon would have $989,999,990.)

So (at long last), the problem gains its strongest form: although Algernon very likely would object to Clancy's differential pricing (and to Barbary's), he should not. Even though he is being gouged, he ends up better off than Barbary under the originally proposed offer. He might try to haggle out of simple greed, but has no rational justification for doing so (because what he might gain is not worth that much to him).

This result is hopelessly counter-intuitive, and hence stands as a significant argument against classical libertarianism.


Cidney said...

In such a situation, wouldn't it be extremely unwise (and unlikely) for the potential buyers to reveal their assets to C?

ADHR said...

Unwise, yes; unlikely, certainly. But it's not impossible, and that's where it gets the libertarian into trouble. If it were impossible, then the problem I'm sketching is no problem at all. Since it could happen, though, then the libertarian has a nasty dilemma: either (1) accept that the theory leads to highly counter-intuitive consequences (in at least some cases) or (2) give up the theory. Consider how a Rawlsian egalitarian might approach the same scenario: even though C knows that A is much better off than B, he has no reason to differentially price against A unless (a) this somehow betters the worse off (unlikely). Furthermore, he has reason not to differentially price against A because (b) this would violate fair equality of opportunity. So, Rawlsian egalitarians have a much more pleasing result: C should not discriminate against A because, prima facie, this deprives A of his fair equality of opportunity (to X); but, if by doing so C helps the worst off, then C should do so. The libertarian analysis is not so complex, and hence it gets into trouble.