Free Capitalist Network - Community Archive
Mises Community Archive
An online community for fans of Austrian economics and libertarianism, featuring forums, user blogs, and more.

Using empirical support only when it suits us?

rated by 0 users
Not Answered This post has 0 verified answers | 19 Replies | 4 Followers

Not Ranked
Male
28 Posts
Points 1,010
slugmises posted on Wed, Jan 5 2011 10:31 PM

I read this comment on Krugman's blog that I found relevant to a question I've been thinking:

"As a recovering Austrian economist, I think that the key flaw in Austrian economics is its insistence that economics is an a priori rather than an empirical science. As a result, when empirical evidence supports their positions, they use it. When empirical evidence falsifies their positions, they declare that it's irrelevant because economics is a priori."

This got me thinking back to an inconsistency that I first noticed in Jim Cox's "The Concise Guide to Economics". One such example is his discussion on the Phillips Curve: http://www.conciseguidetoeconomics.com/book/phillipsCurve/

His claim: "Once workers realize that inflation is not undermining the value of money as rapidly as they had anticipated, they lower their wage expectations thereby shortening the duration of the job search and reducing the unemployment rate."

...is then supported by brief statistics showing a relationship between year, inflation, and unemployment. Is this not empirical support? As a newcomer to Austrian Economics, everywhere I have read informs me that there are too many variables for empiricism to be a valid way to understand the economy.

  • | Post Points: 110

All Replies

Top 500 Contributor
104 Posts
Points 1,585

It's not a flaw in AE, per se. It's more just a flaw in humans altogether.

Reason being: It's human nature to use empirical evidence when it helps you - and ignore it when it hurts. Keynesians do it just the same. Everyone knows all too well that Krugman (et al) will rattle off some empirical evidence as proof he is right - yet completely disregard it when it goes against his belief.

 

  • | Post Points: 35
Top 75 Contributor
Male
1,434 Posts
Points 29,210

It's human nature to use empirical evidence when it helps you - and ignore it when it hurts.

I agree. Leftists do the same thing (i.e. you use stimulus statistics portraying it to be ineffective, they say "it would have been worse without it").

"Every normal man must be tempted at times to spit on his hands, hoist the black flag, and begin to slit throats." - H.L. Mencken.

  • | Post Points: 5
Top 500 Contributor
178 Posts
Points 2,260
Empirical evidence can help illustrate a point. The key point is that empiricism leads to faulty theories.
  • | Post Points: 20
Not Ranked
Male
28 Posts
Points 1,010

"Reason being: It's human nature to use empirical evidence when it helps you - and ignore it when it hurts."

I find this an insufficient excuse/reason because the difference between Austrian claims and most others is that Austrians make it clear that their arguments are a priori, as a posteriori is said to be not practical for economic science. If this is one of the most distinguishing features of the school, then deviating to a posteriori when the opportunity arises is self-contradicting.

  • | Post Points: 5
Top 150 Contributor
Male
550 Posts
Points 8,575

Empirical evidence can help illustrate a point. The key point is that empiricism leads to faulty theories.

That's how I see it. Claiming that economic theory is not developed through empirical studies does not mean one is barred from citing empirical data. If theory is sound, we should expect there to be some data corroborating it. If theory is sound and yet a historical event appears to contradict it (Cox says, ceterus paribus, unemployment should fall under the specified conditions, yet the unemployment rate rises during a period in which the conditions are met), then the problem is with the application of the theory (there are other factors which caused unemployment to rise).

I'm going to assume that the commentator meant he was a casual reader of Austrian economists, as he failed to grasp Misesian methodology, parroting the frequent mischaracterization. Again, the issue is with the development of economic theory, not the citing of data to persuade an audience of the theory's merit and applicability in interpreting real world phenomena.

"People kill each other for prophetic certainties, hardly for falsifiable hypotheses." - Peter Berger
  • | Post Points: 20
Top 25 Contributor
Male
3,055 Posts
Points 41,895

1.  There is no such thing as empirical evidence.  There is only one type of evidence: evidence.

2.  No evidence has falsified AE.

  • | Post Points: 20
Not Ranked
Male
28 Posts
Points 1,010

That's how I see it. Claiming that economic theory is not developed through empirical studies does not mean one is barred from citing empirical data. If theory is sound, we should expect there to be some data corroborating it. If theory is sound and yet a historical event appears to contradict it (Cox says, ceterus paribus, unemployment should fall under the specified conditions, yet the unemployment rate rises during a period in which the conditions are met), then the problem is with the application of the theory (there are other factors which caused unemployment to rise).

The question is then whether or not the opposing argument would be considered valid if conflicting data is presented, considering the original statistics were likely biased. Would an Austrian counter this conflicting data and continue to argue in terms of empirical support, or would s/he back out claiming there are too many variables for conclusive support? If the latter is the case then using any empirical support in the first place seems contradicting.

@ Caley

1. You're right, I meant empirical support.

2. The issue is contradiction.

  • | Post Points: 20
Top 25 Contributor
Male
3,055 Posts
Points 41,895

In my understanding it is not a contradiction.  Allow me to offer an analogy to make it obvious.  Suppose that a scientist performs an experiment that he interprets as a black swan to the laws of physics.  Should he:

A. publish a book declaring the laws of physics wrong?

B. revise his interpretation of the experiment?

C. file it in the records, unsure what conclude at the present time?

The only stupid option is A.  It is one experiment.  Now suppose the alternate case where the scientist interprets the evidence as expected according the laws of physics.  Is it stupid publishing a book declaring the laws of physics correct?  It is ridiculous as the laws of physics are already believed based on many prior evidences.  Nothing practically changes.  This particular evidence is just one more drop in the bucket.

What mainstream economists do is find one "killer" study and go straight to option A.  At that point we're expected to immediately call the demolish firm to raze the Mises Institute and burn the books in every library and start QE1-100 or whathaveyou.

The question is then whether or not the opposing argument would be considered valid if conflicting data is presented, considering the original statistics were likely biased.

There is no such thing as conflicting data.  Only interpretation of data can conflict.

  • | Post Points: 20
Top 50 Contributor
2,162 Posts
Points 36,965
Moderator

slugmises:

"As a recovering Austrian economist, I think that the key flaw in Austrian economics is its insistence that economics is an a priori rather than an empirical science. As a result, when empirical evidence supports their positions, they use it. When empirical evidence falsifies their positions, they declare that it's irrelevant because economics is a priori."

Exactly, I've had that thought for a long time.

Sure, it's fine to use an example in history to illustrate your point, but it's simply demagoguery to act as if you're doing anything other than giving an example. And, in fact, a non-historical example would work just as well for this, but usually they favor the historical ones for two purposes: (a) if their audience believes that it actually happened, the example will be more "vivid" in their minds (where it would take some solid fiction writing ability to pull that off with something made up), and (b) it catches the people who still take empirical evidence seriously into being more likely to get convinced by their argument. The former reason is unremarkable, but the second is pure demagoguery. 

Actually most AEers constantly act as if they take empirical evidence seriously. I mean, it's pretty hard not to. So they give historical examples and raise them up as more than just examples. But then whenever somebody presents a superficially conflicting empirical example, they fall back on one of their main tenets: Empirical examples aren't anything but examples, and they can't prove anything. And so on. So they depend on empirical examples because of their own skepticism about such a radical conclusion, but fall back on that central tenet whenever it's turned around on them. Either that, or it's just a bunch of demagoguery: Use it when it suits us (for converting layman), and deny it when it doesn't suit us (when arguing with more capable people). I mean, I still don't know why so many of the AE theorists care about history. It's fine for somebody like Tom Woods or whatever to care about history: I mean, that's his interest, right? But, for one of the theorists to constantly bring up the historical examples, and not be careful to explain their limitations (that is, their absolute uselessness for their purpose at hand), well, that just strikes me as propaganda. Or something.

(By the way, I'm pretty tentative about these opinions, so don't take them too strongly.)

If I wrote it more than a few weeks ago, I probably hate it by now.

  • | Post Points: 20
Top 500 Contributor
283 Posts
Points 5,580

Caley wrote:

What mainstream economists do is find one "killer" study and go straight to option A.  At that point we're expected to immediately call the demolish firm to raze the Mises Institute and burn the books in every library and start QE1-100 or whathaveyou.

Exactly.  And it's important to note that whenever Austrians do cite a historical example, mainstreamers hardly ever (I want to say never) provide any theoretical interpretation of that example to show why it's not worth considering; it seems they would rather just ignore it and go about their "book burning."  Austrians, on the other hand, can and do couple historical exposition with theory.  To be frank, I'm not quite sure who the economists are that resort to empirical data for tactical reasons.  From my experience, Austrians often resort to statistical data to demonstrate that mainstream intrepretations of that data are simply wrong.  To do this, one needs to apply theory as well.

One instance comes to mind - DiLorenzo has a lecture where he breaks down Standard Oil's pricing history, but his point was that the actual data does not support the common understanding that prices were increasing and production decreasing rather than to use it as any kind of example of the workability of Austrian theory.  Granted, he's responding largely to the laymen, but I've seen many a mainstream economist foolishly cite Standard Oil as an example of monopoly.

  • | Post Points: 20
Top 25 Contributor
Male
3,592 Posts
Points 63,685

It is a total mistake to use empirical evidence, especially the way they do. It assumes an instant relationship between cause and effect. "The tax rate was 50% and we had 5% gdp growth. High tax rates are okay!". I have a rant/solution here.

Banned
  • | Post Points: 5
Not Ranked
Male
28 Posts
Points 1,010

Thank you I. Ryan for explaining well the point I'm trying to get across. To ensure credibility and consistency, I would think these tactics would be highlighted and discouraged here.

@ Caley, I'm not sure I completely understand your analogy but I think I get your gist and appreciate your contribution.

@ Sieben, impressive thread. Hopefully someday I'll be able to understand it. :P

  • | Post Points: 20
Top 50 Contributor
Male
2,687 Posts
Points 48,995

I've noted that a lot of Austrians have tried to use empirical evidence to try and prove a point, rather than simply illustrate it (or apply theory to history), and it's something I've commented on (one and two) before.  However, I don't think it's a widespread problem, or a general contradiction within the Austrian School.

  • | Post Points: 5
Top 25 Contributor
Male
3,592 Posts
Points 63,685

The math on the thread is just junky. The actual idea is pretty simple.

I don't think models like that prove anything at all. Just that you can fit a curve to data. The point is that there is no such thing as a reality that can't be made to fit your worldview. This is bad if you let leftists get away with it. This is good if you show leftists you can pull their crap too (in my case, in a much more sophisticated way). So you force them to retreat back to theory which is where you want to be anyway.

Banned
  • | Post Points: 20
Page 1 of 2 (20 items) 1 2 Next > | RSS