1. Skip to navigation
  2. Skip to content
  3. Skip to sidebar
Source link: http://archive.mises.org/6069/mises-and-hayek-refused-to-do-empirical-work/

Mises and Hayek “refused to do empirical work”

December 28, 2006 by

This article by Mark Skousen makes the preposterous claim that Mises and Hayek “refused to do empirical work” whereas Milton Friedman did do empirical work, which accounts for his influence and their marginalization.

Austrians grow tired of saying this but, once again, the point is not that empirical shouldn’t be undertaken. After all, that is the whole point of Mises’s history books on World War I and II, as well as his close examination of banking and money before and after the Great Depression (Causes of the Economic Crisis). But facts, data, and history cannot and should not be used as a “test” for theory, else the essential causal story is missed, precisely as Friedman and others have missed the actual cause of the 1929 crash.

This point was a frustration in particular for the Austrians in the early 1930s. Everyone was screaming to find a reason for the economic downturn, and looking in all the wrong places. The Austrians pointed out that a robust theory of the business cycle was necessary before one can truly understand the events taking place around them. Whereas their colleagues were all trying to find causation in statistics–a method that turns post hoc ergo propter hoc from a fallacy into a principle–the Austrians pointed to the underlying theory of the discoordinated capital structure, as caused by monetary expansion, and then pointed to evidence of prior monetary expansion.

The irony for Skousen and Friedman is that it was the Austrians who actually warned of the coming depression, whereas the pre-Friedmanites had predicted increases in stock prices as far as the eye could see. And, actually, Thornton recently uploaded an entire paper documenting this. The “data” led the monetarists astray; theory led the Austrians to see precisely what was coming.

For Mises on the errors of positivism, see The Ultimate Foundation of Economic Science. Maybe someone can send Skousen a copy?


coyote December 28, 2006 at 9:50 am

Does Stephen Hawking do empirical work? Did Einstein?

Dennis Sperduto December 28, 2006 at 11:36 am

Unfortunately, the positivists (such as Milton Friedman) can not admit that because there are no constants and no underlying, non-varying probability distributions regarding economic “data,” mathematics and statistics are not appropriate tools to utilize in theoretical economics.

Economic data can not be “tested” in the manner in which an experimental physicist tests the data that is the subject of his investigations. The economic world is characterized by uncertainty, which by its nature can not be predicted in the way a physicist can describe the motion of objects by utilizing the equations of mechanics.

Joe Cesarone December 28, 2006 at 1:21 pm

Dennis, what you say is true for the simplified world of Newtonian physics, but at the quantum level uncertainty is a constant companion, i.e. the (Heisenberg) Uncertainty Principal. The path of an individual electron cannot be accurately predicted. This is not even to mention the uncertainties introduced by the Observer Effect. Perhaps the studies of Austrian economics and physics share more similarities than is commonly thought!

Jason December 28, 2006 at 2:20 pm

What is not being said here, is that an experiment is nothing more than a action or a set of actions whose consequences are closely recorded, thus experimentation is guided by the same subjective theory as normal action. Observation is also an action. Since action is guided by certain apriori assumptions, thus is experimentation. Action requires plans, thus do experiements.

Jason December 28, 2006 at 2:20 pm

What is not being said here, is that an experiment is nothing more than a action or a set of actions whose consequences are closely recorded, thus experimentation is guided by the same subjective theory as normal action. Observation is also an action. Since action is guided by certain apriori assumptions, thus is experimentation. Action requires plans, thus do experiements.

Dennis Sperduto December 28, 2006 at 2:53 pm


You are correct; I should have said the “equations of Newtonian mechanics”.

rtr December 28, 2006 at 4:12 pm

From the article:

“Ironically, it was his painstaking, objective analysis in the landmark work, “A Monetary History of the United States, 1867-1960,” that gave him such labels. In that work, he and coauthor Anna J. Schwartz asserted that the Great Depression was not a failure of market capitalism, but of government policy. They showed that the Federal Reserve acted ineptly in allowing the money stock to decline by “more than a third,” converting a garden-variety recession into the worst economic catastrophe of the 20th century.”

HAHAHAHAHAHAHAHAHAHA, I’d just love to see the citation and data showing that “the money stock” declined by more than a third. Do these people not even know what money is? What a colossal error which extends way beyond data collection!

Were a third of the outstanding paper dollars burned? Did M3 decline from a base of 100.0 to 66.6? Hell no! Just going by recollection guess work, since the data has been wiped from the ferderal reserve website, M3 might have dropped 0.3, like say from 38.4 to 38.1, from 1928 to 1929. Perhaps he meant the “gold stock”? But it still goes to show people have no clue even what a simple thing as money is. Hell, even the Austrians are just less confused, which is why they overestimate the impact of malinvestment and underestimate the impact of cessation of trade when it comes to the Depression.

Even statistics does not establish the validity of statistics. Statistics is logically deduced from mathematical theorems. How does one expect to statistically derive or statistically prove philosophy or economic methodology?

If the term “money” was forever stricken from use and instead replaced by the actual things which are traded such as units of paper, grams of gold, credit grade promise to repay in the future, etc. much improved economic thought would soon follow. Instead of “do the math” it should be “do the description”. Which is more accurate?

1.) Person A traded $100 to Person B for a book.

2.) Person A traded a green and white piece of paper with the # 100 written on it to Person B for a book.

That’s the only way anyone is going to understand why goods are services are exchanged and how they move through different hands through time. Otherwise, people like Friedman get dumbfounded by money and people like Mises get dumbfounded by such concepts like “hyperinflation”.

Anon December 28, 2006 at 7:00 pm

“But facts, data, and history cannot and should not be used as a “test” for theory, else the essential causal story is missed”

I’m sorry but I don’t understand. How do you know the theory has any validity then? You could say the easter bunny created the depression but since the data should not be used to test the theory you can never ‘disprove’ it. You could make predictions with the theory but the predictions can only bear fruit by ‘testing’ the theory against that final data (i.e. either it predicted correctly or it did not). If it is purely a tool for post-facto explanation without any recourse to actual data then why is the easter bunny theory not equally valid as an explanation.

“Does Stephen Hawking do empirical work? Did Einstein?”
Their theories were validated against empirical evidence.

Antony Mueller December 28, 2006 at 7:08 pm

Let us imagine a talk between Mises, Hayek, Friedman, and some modern econometrician. And then give out the task to discuss an issue about the “real, empirial world” (current or past). From whom could one expect to gain insight and understanding? In my view, no question: Almost nothing (other than utter confusion) from our contemporary econometrician, some stories and some funny but unrelated bits of information from Friedman), and then maybe some thoughtful insights from Hayek, but, from Mises, so I am sure, you would get precise facts, an explanation of their relationship, and many, many empirial knowledge about human actions in the past (history), and thereby the Mises approach would give us an understanding of our present situation. History, of course, cannot be measured. Any science has had its ridiculous period, now it is modern economics in its positivist or econometric stage that is going through this phase.
antony mueller: http://www.continentaleconomics.com

David J. Heinrich December 28, 2006 at 8:33 pm


As Austrian methodology explains, it is perfectly valid to check empirical data to see if a particular economic theorem applies. E.g., if we stipulate that massive credit-expansion caused a boom-bust, we can check the empirical data to see if there actually was a credit-expansion, and if the theory would even apply in that given situation. It is possible that an alternative explanation could be valid; e.g., if the King stole all of the gold, or the world entered into a devastating world-war, or an asteroid wiped out a hemisphere.

Given that the specified conditions necessary for the theory to apply exist, we cannot further use “empirical evidence” to determine whether or not an economic theory is correct.

quasibill December 29, 2006 at 7:30 am

It’s not uncertainty, per se, that causes economics (and other social sciences) to be non-empirical in nature. The truth is that the ability to isolate one variable is either a) impossible or b) unethical, immoral, etc, in the social sciences.

As David Heinrich notes, the best one can do is say that there is historical evidence that tends to support a given theory, but you can almost never “prove” the theory by eliminating all null hypotheses, like one would in the “harder” sciences.

All of which just sidesteps the underlying issue – the scientific method is nothing more than a widely accepted method of applying reason to answer a question – it is shorthand. But it is only valid for those questions that reason tells us it can be applied fruitfully to. For those problems where reason (or ethics!) tells us the scientific method can’t be properly implemented, one falls back to pure reason to address the question.

Jason December 29, 2006 at 11:53 am

“the scientific method is nothing more than a widely accepted method of applying reason to answer a question”
Reason does not exist in a vacuum, it is formed and limited by human action, thus is science. This fundamental fact is ignored by modern science and so they come up with theory that cannot and will not be tested. The so called “end of science”. You may put together a hypotheses that exists seperate from reality, but it still must be steeled against what action can accomplish.

Stefan Karlsson December 29, 2006 at 1:44 pm

Strange that Skousen would claim that Friedman’s empirical works accounts for why he is more influential than Austrians. While he may be unaware of Mises’ Causes of the Economic Crisis, he is definetly aware of Rothbard’s America’s Great Depression as he made numerous references to it in his own article on the depression in the Elgar companion to Austrian economics. Rothbard’s book, like the Friedman book, appeared in the early 1960s, and as it too was mostly empirical , so if it was a question of that, it should have been as influential as the Friedman book.

Brian Macker December 29, 2006 at 3:56 pm

I’ve always regarded the purported non-empircism of Austrian economics to be a weak point of the school. I can not fathom why some Austrians abandoned the high ground on this issue. It’s quite apparent to me that what they are doing is not completely a-priori.

I think it is quite unlikely that anyone no matter how smart would deduce something like the business cycle merely from knowing that people make choices. It is in fact not how the science of economics progressed. It instead progressed as most other sciences with theories and then the refutation of those theories either against empircal data or because of internal inconsistencies.

The fact of the matter is that all Austrian economist’s work went in the other direction. They saw empircal outcomes and then made theories to explain those outcomes based on prior theory.

Yes, economist have to operate differently than other sciences, yes this is not physics, and yes trying to apply the tools of physical modeling to economics is a mistake, but no that does not mean economics is non-empircal. Problem is as I recall Mises made statements too strong in the other direction.

He either was trying to say economics was not empircal or was communicating in such a way that strongly gives that impression. I think this was a big mistake on Mises part and I think it would do well for Austrians to break free of this.

BrianMacker December 29, 2006 at 4:38 pm

Unlike Skousen I think the main reason for the differential in popularity lies in the policy recommendations made by the different schools. That and the fact that most economists are employed by the government. Monetarist theory calls for much more government power over the economy.

Eric December 29, 2006 at 11:48 pm

Of what use is any theory that does not make predictions? And once it does, of what use is it if the predictions cannot be verified.

Any such theory that is not testable may as well be a work of fiction.

Suppose I have a theory that the boom and bust cycle is caused by aliens who use mind control on the human race and that they cannot be detected. I begin with some postulates on aliens and mind control, then I show with geometric logic how it follows that if the aliens manipulate humans in a specified manner then a boom and bust results.

Of what use is this if I can’t find any evidence of the aliens, or I claim that the theory only works when the aliens visit us and since they are undetectable, we’ll never know when they visit.

John December 30, 2006 at 7:44 pm

Just to be clear: Skousen considers himself an Austrian economist. His latest book is a comparison between the two schools title Vienna and Chicago, Friends or Foes. He was stating his opinion as to why Friedman has much greater standing than Mises in the economics community.

Mark Humphrey December 31, 2006 at 11:48 am

Mises and Hayek explained that economics, by its nature, required a different methodology than the physical sciences. The reason lies in the fundamental difference in the ultimate data scrutinized by economics, versus the kind of data studied by hard science.

Hard sciences isolate cause and effect in the realm of unthinking unconscious physical reactions. For example, scientists can predict that combining two molecules of oxygen and one of hydrogen always yields a liquid; or that a billiard ball, banked at a certain angle and speed, will roll into the corner pocket at a precise time. To discover cause and effect, hard science relies on creative imagination to posit an hypothesis, which is then literally tested in an experiment that holds all variables constant, except one.

In contrast, economics is a study of a different kind of data–the realm of human choice. For distinct from unthinking matter, humans possess the capacity of self-directed volitional consciousness. As such, human behavior invariably is characterized by an element of the unpredictable. As Rothbard expressed it, economics is a study of the implications of the empirically observable fact that humans “act”: they employ scarce means to achieve chosen ends.

Economics proves certain broad “predictable” outcomes in human action–outcomes that always remain unpredictable in a more narrow, more specific sense. So, although Austrian theory can reliably predict broad outcome from monetary inflating, i.e. that prices will be higher than they would otherwise be, no one has the ability to predict precisely how high, or when, prices will rise. For the precise outcome is the product of millions of individual choices, each of which is necessarily unknown to the economist.

Ultimately, all knowlege–from physics and chemistry to economics and history–stands on the same epistemological foundation: the integrative power of logic applied to the evidence of the senses. However, the appropriate method of logical investigation depends on the kind of subject under study.

Anthony June 4, 2007 at 1:16 pm

For those interested, here is an article on Hayek and action as an empirical concept:


mempko March 8, 2009 at 4:11 pm

>Does Stephen Hawking do empirical work? Did Einstein?

They are important because their ideas were tested in the real world and proven to be true. Einstein had others to do the empirical work and he was very excited about their results.

Comments on this entry are closed.

Previous post:

Next post: