He summarizes Rothbard's position:
Rothbard went one step further by "reconstructing" welfare economics along Austrian lines. His main conclusions are simple and austere: every market transaction benefits all participants, while every act of government intervention benefits some people at the expense of others. Rothbard goes on to make a seemingly stronger claim: "If we allow ourselves to use the term 'society' to depict the pattern of all individual exchanges, then we may say that the free market 'maximizes' social utility, since everyone gains in utility." This claim might be re-phrased to say simply that each voluntary exchange benefits all participants, and the free market permits the implementation of all desired voluntary exchanges.
Sounds like govt meddling is a bad thing then, no? Not necessarily, explains Caplan:
Strictly speaking, however, Rothbard could only claim the welfare effects of government intervention upon "social utility" are indeterminate; i.e., since the victim loses and the intervener gains, it is impossible to say anything about social utility without making a verboten interpersonal welfare comparison.
In other words, it's true one person loses when the govt steals his property. But maybe the beneficiary of the theft feels so great from his ill gotten gain that his happiness outweighs [in some cosmic scale of things] the sadness of the victim. Since we cannot see the cosmic scale, we don't know if govt meddling in any given situation is good or bad.
Sounds crazy, no? But wait, it gets worse. Caplan continues:
This is an important point, because it shows that Rothbard's welfare economics provides a much weaker defense of the free market than usually assumed.
I don't know what is "usually assumed", since Caplan provides no links where to find these usual assumptions.
At any rate, Caplan explains why Rothbard's defense of the free market is weaker than usually assumed:
In particular, Rothbard's own theory strips him of the ability to call any act of government "inefficient." By denying the ability to endorse state action in the name of efficiency, Rothbard also implicitly denies the ability to reject state action in the name of efficiency.
Oh, so that's it. The govt using violence to steal and make people suffer cannot be rejected, because maybe they are being "efficient", meaning maybe they really love the results of stealing more than the victims hate it.
Just to make sure we understand his take on Rothbard, Caplan spells it out again:
This is no logical flaw in Rothbard's theory..., but it's political implications are rather different than commonly assumed: Rothbard's welfare criterion justifies agnosticism about - not denial of - the benefits of statism.
In other words, if the govt, or for that matter the Mafia, kills and rapes and steals, we dare not say they have done a bad thing. Because who knows, they might have gotten more pleasure from killing and raping and stealing than the victims suffered from those acts. That's the political implication of Rothbard's theory. We dare not judge; we must remain silent agnostics.
"Hey, don't look at me," Caplan may protest. "That's not my theory, that's Rothbard's. I'm only pointing out what he really said, as opposed to what is commonly assumed. If you don't like the political consequences of his theory, email your gripes to him, not me."
It's clobbering time. I see a couple of flaws in his reasoning.
1. Rothbard is comparing two systems. System A is always voluntary, System B always involves violence and coercion. System A always benefits both parties, System B always harms at least one party. System A always has a guaranteed net benefit [what Caplan calls "efficiency"], System B never has a guaranteed net benefit, and has a possible net loss.
Which system is clearly the superior one? Is it not clear that System B deserves our rejection compared to A, not our agnosticism? Of course it is. What's Caplan going on about?
2. There is a more subtle error here as well. Caplan is confusing the unknown with the impossible. As an example, let me ask which is a deeper shade of green, Justice or Mercy? The answer is not "We don't know because they are both invisible". The answer is, "No scale of greenness can be applied to these concepts. They are incommensurable with respect to greenness."
As an example for those who know a bit of math, when a mathematician is asked, which is the larger number, 1 or zero, he says 1.
Asked which is larger, 1 or some unknown number X, he answers "Unknown."
Which is larger, 1 or i [=the square root of minus one], he answers "They cannot be compared. The very concept of 'larger than' does not apply to these two numbers with respect to each other."
If Rothbard was saying that Mr Smith's benefit cannot be measured against Mr Jones' loss, just as 1 and i cannot be measured against each other, then we are left with a System A that causes benefit to Smith and Jones [=two things in its favor] and System B that causes Smith harm and Jones gain [= one thing against the system and one thing in favor of it]. Clearly, System A is the superior one. Nothing agnostic about it.
[Added May 21, 2011: I found a line in Human Action by Mises which expresses exactly this idea. I think it fair to assume that Rothbard agrees. Here's the quote:
Some economists believe that it is the task of economics to establish how
in the whole of society the greatest possible satisfaction of all people or of
the greatest number could be attained. They do not realize that there is no
method which would allow us to measure the state of satisfaction attained
by various individuals. They misconstrue the character of judgments which
are based on the comparison between various people’s happiness. While
expressing arbitrary value judgments, they believe themselves to be estab-
lishing facts. One may call it just to rob the rich in order to make presents
to the poor. However, to call something fair or unfair is always a subjective
value judgment and as such purely personal and not liable to any verification
or falsification. Economics is not intent upon pronouncing value judgments.
It aims at a cognition of the consequences of certain modes of acting.
There ya go. Any comparison of two people's happiness is an "arbitrary value judgement".
One could argue still that Mises means merely that the value is unknown. But once he says there is no method of measuring even the happiness of one individual, Occam's Razor assures us that there is incomensurability between two people].
Caplan then lays down another whopper, but that's for our next blog.