Rothbard on statistics

You should probably send this question to someone who really knows Rothbard's writings as well as a lot of statistics...so email Robert Murphy or Walter Block. 

[email protected]

[email protected]

 

I just looked at the PDF quickly, and I found it strange. I'm only a first year economic student and I wouldn't make such an elemntary mistake, which seems odd since Rothbard was very proficient in maths.

the fellow that wrote the pdf about what Rothbard said, then goes on to comment about it

 

It may still
remain true, however, that in contexts where random
collectives do not exist (that is, contexts characterized
by lack of independent repetitions), as will often be the
case in economics, objective probabilities cannot be
used. Given that Rothbard embraced an objective,
frequency interpretation of numerical probability, his
rejection of statistics is a defensible and logically
consistent corollary. Moreover the rejection of the use
of objective probabilities in economics is in agreement
with the conclusions of some of the most recent research
about these matters, and with general arguments for
interpreting probabilities in economics as
epistemological rather than objective.8

How does epistemological contrast with objective?

im sure if i had a PhD i could explain,

but i dont... so.....

nirgrahamUK:

may still
remain true, however, that in contexts where random
collectives do not exist (that is, contexts characterized
by lack of independent repetitions), as will often be the
case in economics, objective probabilities cannot be
used. Given that Rothbard embraced an objective,
frequency interpretation of numerical probability, his
rejection of statistics is a defensible and logically
consistent corollary. Moreover the rejection of the use
of objective probabilities in economics is in agreement
with the conclusions of some of the most recent research
about these matters, and with general arguments for
interpreting probabilities in economics as
epistemological rather than objectiv

Yet none of this is Rothbard's argument.  It's one thing to give a different reason for Rothbard's conclusion, but something else entirely to explain Rothbard's reasoning.

is it not analgous to other differences between austrians and neo-classicals.?

yes, you have models and given the assumptions you make, the models are good models. but contrariwise your assumptions are nonsense so i dont find your models useful. etc. etc.

 

im really speculating as i know next to nothing about statistics aside from basic stuff about averages and whatnots.

Jon Irenicus:

ow does epistemological contrast with objective?

I believe the difference is that objective probability results from uncertainty in the case, whereas epistemological uncertainty results from incomplete knowledge.  We say that the outcome of a pool shot is not likely to put a ball in the hole, but it might, which seems to express epistemological probability, since (many believe anyway) if we knew all the forces, the outcome could be calculated in advance.  An example of the distinction seems to be in the various interpretations of quantum mechanics - whether the Heisenberg uncertainty principle refers to uncertainty in the thing (the electron doesn't have a certain position and velocity), which would be objective, or to uncertainty of knowledge (to measure the position, we use a wave of such a high energy that it changes the velocity) which would be epistemological.  Thus, when we give the probability cloud, the probability of it being in some portion of the cloud is, in the first view, objective, and in the second, epistemological.

Ah I see, so it's basically a term of art then.

Kling on probabitily.

Argh, yet another article to look into for my paper on methodology! Luckily it's brief enough.

it was very informative article, i like how he categorised the three approaches to probability. (i wonder are there more?)

he said this:

Frank Knight was famous for describing a notion of uncertainty that could not be reduced to an objective probability distribution.

does that mean it could not be reduced to something that can be translated into the normal distribution that was mentioned earlier? is this then the rothbardian position?

Knight's uncertainty is something like the probability that Hans Hermann Hoppe will oppose a proposed bailout bill. There's no calculation or formula for figuring it out, and it's just silly to say that Hoppe is 90% certain to oppose the bill. What does the 90% stand for? It isn't his past record of positions on similar bills (frequentist.) It isn't logically derivable from some set of axioms (axiomatic.) What is it? The best we can say is that he's very likely to oppose it, but we cannot give a number. Some might want to give a number of 100%, but that's not true either. People do change their positions, and an Austrian could decide tomorrow to support socialism. Giving a probability of 100% would imply that was impossible. Hoppe relates Knightian uncertainty directly to human action. Although it's possible, I don't think this is what Rothbard had in mind. My reason for thinking that is the he was talking in the context of a statistics class, and so he should have in mind the kinds of things that people claim statistical inference works with. An example would be GDP figures over long time frames, measured annually. Such numbers are not at all normally distributed. As far as I can tell, Rothbard is claiming that, for their calculations, statisticians simply assume that GDP is distributed normally, find the average, find sigma, and assign probabilities to the future from there. But everyone knows it's not normal, and no one does this. Instead, statisticians make use of the CVT, and do this kind of procedure: 1. Find the distribution annually. Notice that it isn't normal, or close to normal. 2. Break the interval into 2-year periods, and assign each 2-year period the average of the GDPs. Notice that it still isn't close to normal. 3. Break the interval into 3-year periods, averaging in each interval. 4. Continue the process. The CVT says that the distribution will approach normality.

under what conditions would it be incorrect to apply the CLT when wanting  to run some statistical analysis? do economic matters of the kind that concerned rothbard share these conditions or no?

its just that this on the wiki page for clt caught my eye>>

let X₁, X₂, X₃, ... X_n be a sequence of n independent and identically distributed (i.i.d) random variables each having finite values of expectation µ and variance σ² > 0.  etc.

 

i am a layman so forgive me if what im saying seems particularly stupid.