03
Aug
2013
David Friedman versus Me
(I’m never sure whether to refer to myself in the 3rd person on posts like this. I’ve got to do what is best for Bob Murphy, even if it sounds weird.)
Here’s the debate we had at Porcfest in June. The only major miscommunication is that Friedman just wanted to talk about methodology, whereas I, the moderator, and the Porcfest event brochure all thought we were talking about the ethical foundations of anarcho-capitalism as well.
Also, if you’re too busy to watch this, I draw your attention to 11:20 – 12:20, where I correct the woeful deficiency in attention that the economics profession has given to Pat Murphy.
I enjoyed this, thought it was pretty informative. What is your response to Friedman’s assertion that you’re attacking a straw man when you argue against using collected data / statistics as evidence to derive economic laws?
So did you actually disagree on anything?
From what I could tell the methodological argument was:
“Austrians split economics into theory and history, and theory is first developed and then applied to history”
and
“Chicago school economists don’t split economics in this way, but they first develop theory and then apply it to history (or “the world”).”
Austrians do not split economics into theory and history . . . for Austrians, economics is theory. It can neither be proven nor disproved by empirical data. This is not a cop out. It is simply the recognition that empirical data in the social sciences are not fit to render any meaningful economic truth, since this empirical data cannot be produced in a controlled experiment. However, you can use the theory to interpret empirical data, thereby rendering economic history. Note that this use of the theory to interpret the empirical data does not prove the theory . . . the only way to disprove the the theory is to show that its internal logic is inconsistent.
Chicago School economists, on the other hand, develop theory and then apply it to history, but then allow history (or the empirical data) to modify their theory. They do not pay heed to the fact that the empirical data is unfit for this purpose.
I really enjoyed that, glad to see it finally. However the first questioner reminded me why I never want to go to something like Porcfest, watching people get genuinely upset about philosophy (the guy seemed to be blaming you for everything wrong with social science methodology). I really like reading yourself and David Friedman, and seeing you together was (very slightly) like that scene in Heat when Al Pacino and De Nero’s characters finally come face to face.
This was really good.
I have tried over and over again with Prof. Sumner and as best I can tell he is looking at curves or data and that is all.
The first guy in the Q&A was an asshole. Nobody came there to hear his opinion. Didn’t even ask a question.
Yep. The first guy in the MMT debate was an asshole too.
RPM crushed him in the closing by pointing out that empiricism has been ruling the mainstream for quite a while.
^^ First guy in the MMT Q&A I mean 🙂
Very nice debate. Two points that surprised me:
1. The way David F. described how insane the current fractional reserve system is since it obviously economizes a good of which the production cost is zero.
2. The idea that MMORPGs (Massively Multiplayer Online Role Playing Game) could possibly be useful for some empirical observations, which I thought about also already some time ago.
http://www.lewrockwell.com/2006/07/alexander-villacampa/the-economics-of-world-of-warcraft/
http://www.zerohedge.com/news/2013-05-21/diablo-3-case-virtual-hyperinflation
Yes I know those articles. It really is an “irony” that some empirical work on that field was already done by followers of … wait for it … Mises..
Correction. Didn’t know the first one.. Thanks!
That’s pretty good. You gotta love how people were abandoning virtual gold in favor of fiat currency in that game. Pretty ironic.
Isn’t D. Friedman’s proposed tests in online games just an example of ceteris paribus arguing?
If he expects there to be actual people playing on every server on which he does tests, the Austrian response is to point out that you can’t control for individual preferences.
But if he’s just changing one thing at a time, then he’s conceding the Austrians’ methodology.
I thought you let him get away with a lot. A couple of them are
1. The insistence that rationality is an assumption rather than a logical corollary of the action axiom
2. Predictive power as the measure of the worth of an economic theory
Bala,
On (1), I don’t remember when, but at some point I clarified that Mises and Friedman were using “rationality” differently.
On (2), that was basically the whole debate, so I don’t know what you mean that I let him get away with it.
Apropos, here’s a video in which Tom Woods mentions that Irving Fisher, a guy who Milton Friedman called “the greatest economist the United States ever produced”, failed to see the crash of 1929 coming, but an Austrian economist, Ludwig von Mises, did:
Neoconservative David Frum Hearts the Fed
http://www.youtube.com/watch?v=1d1rcaX-lzU
I wish Austrians would stop using such examples. It rather reinforces the notion that predictive power is the proper yardstick to judge a theory.
True. Robert Wenzel is constantly reminding of that, too.
I take this back, now that I think of it.
Austrian predictions are of a ceteris paribus nature; This is why we can’t set dates.
We don’t say, “such and such will happen”; We say, “all other things being equal, such and such will happen”.
What we know is that government stimulus is suspending individuals’ pursuits of their own ends, and since that’s what all economic activity reduces to [individuals pursuing their own ends], the artificially stimulated sectors will be liquidated as a matter of course.
For example, digging ditches and filling them again doesn’t create wealth – it will create trading, but it will be trading that, on net, destroys wealth.
“Paying” people, with newly printed money, to do dig and fill won’t change the fact that the work isn’t creating wealth. The workers are benefiting at the expense of later users of the new money.
Austrian theory is why the Austrians were able to predict the housing crash:
New York Fed: Leave the Building!
[WWW]http://mises.org/daily/6028/New-York-Fed-Leave-the-Building
One economist asked me how I knew the housing market was going to crash. I responded that because of Austrian theory, I understood that money created by the Fed enters the economy at specific points and that it was obvious the housing market was one of the those points. I told him that I also knew that this would eventually result in price inflation (as the money spread through the economy) and that at that point the Fed would slow printing and the housing market would collapse, which is just what occurred.
Ron Paul: “This real-estate bubble will burst, as all bubbles do” (part 3)
[WWW]http://www.youtube.com/watch?v=KONpt9a6HrI
The Federal Reserve credit created during the last eight months has not stimulated economic growth in technology or the industrial section; But a lot of it ended up in the expanding real estate bubble …
“Austrian predictions are of a ceteris paribus nature;”
As Bob said in the debate, more than being of a ceteris paribus nature, they predict outcomes in comparison to counterfactuals. Once again, paraphrasing Bob, an Austrian would say “increasing the supply of money would cause prices to be higher than they would be in the absence of the increase in the money supply”.
This “predictive” power, however, is not precise or quantitative. It can be neither simply because you will never have the numbers associated with the counterfactual in hand to compare the actual numbers with. Friedman seems to expect a different kind of predictive power – one that is numerical and precise.
Yes. I agree with the way you put it.
What is the proper yardstick to judge a theory?
I’ll turn the question around. Is empirical evidence the proper yardstick to judge a theory based on deductive reasoning? Is predictive power the proper yardstick to judge a theory that operates in a field where precise prediction is made impossible by the nature of the subject?
If your theory is based entirely on deductive reasoning, then it belongs in the world of pure mathematics. That means starting with axioms, and then expounding the consequence of those axioms.
In such a situation, empirical evidence is not useful because such a theorem says nothing about the physical world. It may be an interesting deduction, but it remains abstract.
At some stage, to get beyond pure mathematics and into applied mathematics, you need to find real-world scenarios, and then you are indeed facing empirical scrutiny.
Incidentally (and I am sure you are aware of this), but Austrian economic theorising goes in the very direction you have described –
“starting with axioms, and then expounding the consequence of those axioms.”
In such situations, as you rightly pointed out,
“In such a situation, empirical evidence is not useful”
though your reason is way off the mark when you say
“because such a theorem says nothing about the physical world”
because you are essentially dismissing deductive reasoning as a method to “know”. The way you seem to put it, direct sense perception is the only correct route to “knowledge”. Isn’t that what is called empiricism? (Correct me if I am wrong on the term. I sometimes get them mixed)
I disagree. Both inductive and deductive reasoning inform us about reality and lead us to “knowledge”. It is just that they operate in different realms, operate on different methods and are judged on different yardsticks. If you want to attack a deductively worked out theory, you need to attack its axioms, not look for empirical evidence that “defeats” the conclusions.
You can know a lot about pure mathematics with deductive reasoning, you just can’t know anything about the physical world if you refuse to take measurements.
Even when you do take measurements, they will be flawed of course, but that’s a lot better than nothing.
Your present reply fails to address my point that your previous reply dismisses deductive reasoning as a method to “know”, as a route to “knowledge”.
Tel, if the premises tell us something about the real world, don’t the deductions from those premises necessarily tell us about the real world as well?
If our premise that persons act, using means to achieve ends, is correct, and does indeed tell us about the real world, don’t deductions from this premise tell us about the real world as well?
Also, it is not necessarily correct to say that flawed measurements are better than no measurements at all. Flawed measurements can be so flawed as to lead to incorrect conclusions. Of course, a consistent empiricist would say we’d have to measure that to figure out which is true 😉
I think this will help:
In Defense of “Extreme Apriorism” by Murray N. Rothbard
[WWW]http://mises.org/daily/5195/
Whether we consider the action axiom “a priori” or “empirical” depends on our ultimate philosophical position. Professor Mises, in the neo-Kantian tradition, considers this axiom a law of thought and therefore a categorical truth a priori to all experience. My own epistemological position rests on Aristotle and St. Thomas rather than Kant, and hence I would interpret the proposition differently. I would consider the axiom a law of reality rather than a law of thought, and hence “empirical” rather than “a priori.” But it should be obvious that this type of “empiricism” is so out of step with modern empiricism that I may just as well continue to call it a priori for present purposes. For (1) it is a law of reality that is not conceivably falsifiable, and yet is empirically meaningful and true; (2) it rests on universal inner experience, and not simply on external experience, that is, its evidence is reflective rather than physical;[7] and (3) it is clearly a priori to complex historical events.[8]
“you just can’t know anything about the physical world if you refuse to take measurements.”
This is decidedly false. In fact, it refutes itself. In order to take measurements and know things about the physical world based on those measurements, you must know quite a few things about the physical world prior to this. Constancy of physical material, continuity of events, universality of physical law, reliability of observation, etc.
All that knowledge must be a priori to empirical measurement, and can only be deduced by deductive reasoning.
He repeated 10 times that economic hypotheses have to be predictive, and you never corrected him to say, no prediction in economics is nonsense. You implied that, but never really spelled it out.
Precisely my point. Thanks.
You did touch briefly on the two notions of rationality.
Still, I was jumping in my chair hoping the discussion would take a detour and go deeper into that distinction. I didn’t feel David Friedman understood the Austrian position there.
Maybe next time and thanks for the debate 🙂
David Friedman said:
Bob addressed this, but maybe I can bring something a little more.
As Bob said, this is not the Austrian’s view. Our view is that people will take the actions they BELIEVE will allow them to achieve their objections, and that they will use means (time, physical exertion, tools, etc.; [If it didn’t require means, then there would be no need to pursue or act on their desires]) to do so.
This significance of this is that, whether he’s right or wrong about what his actions and means will accomplish, he is expressing a valuation of them. He is valuing his ultimate objection more than what he believes the use of the means he employs will cost him. That’s what he believes.
Whether or not he is being “rational”, he has expressed willingness to pay a certain price for all the means he believes is required to fulfill his ends. And this is where prices come from.
When government force is used to suspend one’s pursuit of his ends, it will make the means which have been forcibly withheld from him to be worth less to him in the pursuit of his ends.
At some sufficient level of coercion, it becomes cost prohibitive to pursue his ends using his intended means, and he will either seek an alternative means (with alternative tools, goods, etc.) to pursue his ends or suspend his pursuit.
Government is artificially changing people’s valuations of certain goods. This is why when the government tries to help the poor with a Minimum Wage, all other things being equal [a concept which I think David Friedman missed] this will make lower skilled people less attractive to hire, and it actually ends up hurting the very people it was intended to help. The whole point of the employer hiring anyone in the first place is to make money – that’s the only reason the position exists.
When David Friedman talks about there conceivably being a scenario where a Minimum Wage doesn’t result in there being more unemployment among the poor, he’s not saying something that the Austrians would disagree with.
Rather, the Austrians say that when a Minimum Wage is imposed, ALL OTHER THINGS BEING EQUAL [key], it WILL hurt the poor. You don’t need to test this. There is a restriction on how competitive poor people are being permitted to be as far as wages.
If there was a Minimum Wage and the government also taxed rich people and gave it to the poor, would the poor be worse off? At the moment, no – Austrians wouldn’t disagree with this. But this is not a scenario in which the ONLY change is the introduction of a Minimum Wage.
(In the longer term, what will happen in that last scenario – all other things being equal, of course – is that the taxes will disincentivize production, and goods would cost more than they otherwise would, thereby hurting the poor.)
There’s two quite separate issues you have brought up here.
First being people’s subjective valuation. That is to say, the Austrian position is that even in a world of absolute rationality and perfect information, still people would make different choices, simply because they have different preferences, different objectives.
The simple-minded economic model is merely to presume all people have the same objective, i.e. to make as much money as possible. That’s obviously wrong because money is a means to an end, not an end unto itself (Say’s Law) never the less this model remains strong.
The second issue is imperfect information. No one can really guarantee to be able to predict the future, but all action requires some expectation of what the future will being. Thus, if we are to predict human behaviour, we must also meta-predict what type of predictions those humans are making about their own behaviour and that of others. Being human ourselves this leads to endless recursion. To the best of my knowledge, this second issue has not been solved… you can go some distance down the rabbit hole and then draw a line and say “good enough”.
That’s not useful in a real-world scenario because all actions affect all other actions, in a non-linear and highly interconnected world. Thus, you do have to test these things (as best as you can) and worse, you have to regularly go back and test again.
Joseph Salerno gave a good break down of this:
Calculation and Socialism | Joseph T. Salerno
[WWW]http://www.youtube.com/watch?v=alqUqdbfxhk#t=15m54s
This is precisely why tests couldn’t tell you anything about the policy effects. Is what is observed in the tests the result of the policy, or in spite of the policy?
But since it can be known with certainty that all economic activity reduces to individuals pursuing their preferences [by way of the Action Axiom], we can deduce that, whatever tests show, there would be more wealth absent a Minimum Wage.
Really? How do you deduce that the concept of “more wealth” in a global sense even exists, when no method is even defined for summing subjective preferences between individuals?
Even if you did come up with a definition, there’s no reason to presume that everyone else agrees.
Wealth is subjective to the individual, so there’s no need to think in aggregate terms; There’s no reason to attempt to sum subjective preferences.
And there is, as you say, no method to do so.
What I’m saying is that since the Minimum Wage suspends individuals’ pursuit of preferences, they are forced to choose a lower ranked preference. This is a situation in which the individual is not acquiring as much wealth as he would in the absence of the Minimum Wage.
Therefore, all other things being equal, absent the Minimum Wage, there would be more wealth as determined by the individual.
The worker has an added burden placed upon him in his search for a job; He could likely find a job right away if he were able to offer a lower wage.
The employer isn’t hiring employees as fast as he otherwise would be able to, because unskilled labor is being priced out of the economy.
Therefore, the production costs are higher than they would be, absent the Minimum Wage.
And this, in turn, makes it that much more difficult for other businesses to compete; This disincentivizes production, such that prices will be higher than they would be, absent the Minimum Wage.
You don’t have to test for these things. All other things being equal, this follows logically.
On rationality, I did note that you mentioned that there is difference between the ways Mises and Friedman used the term. However, my point was that you did not take him on over his repeated statement that rationality is an assumption, which I thought a Misesian should dispute strongly.
On predictive power, I didn’t see you saying (in this debate) that predictive power is not the proper measure of the worth of (an) economic theory. Friedman somehow made it look like the purpose of economic theory is to come up with conjectures that need to be tested against the data while you did not question this position. Shouldn’t you have emphasised the point that the purpose of economic theory is to explain rather than to predict? He seemed to be claiming the intellectual high ground on that point. He even said that you refused to take up the challenge.
Well, obviously if you walked away thinking I didn’t challenge him on it, then I didn’t do a good job making my position clear. It’s hard in these debates to dot every “i.” But the whole theme of my remarks was that what economists teach the public comes from logical introspection, not from looking at the data.
Well, that’s sort of what I am trying to say. However, I was not trying to nitpick. It was just that he was so in-your-face over predictive power that to a less-informed listener, you could very well have come across as ducking an important issue that your opponent was bringing up. It was almost like “Austrian’s are so scared to discuss the lack of predictive power in their theory that Murphy is refusing to take up the challenge.”
Bob Murphy said
Some guy in the audience took issue with that.
A person who thinks that the claim is falsifiable is engaging in circular reasoning.
He is presuming the falsifiability of the claim: “To be scientific, a claim needs to be falsifiable”.
The claim is a theory, the truth of which cannot be ascertained by experimentation.
What Is a Scientific Theory?
http://mises.org/daily/5860/What-Is-a-Scientific-Theory
Empirically, you can’t prove that ability to falsify is strictly necessary, but then empiricism doesn’t strictly prove anything, it just gives you a pretty darn good idea, once you have taken enough measurements.
So how many people do you see who walk down the street blindfolded? Not many I guess, given a choice we can empirically observe that most people prefer to look where they are going. That is to say, they take information from the outside world, and react to what comes. Even people who are permanently blind, will choose to use a cane to help them, or a guide dog, or listen carefully to the outside world, or collect whatever information is available to them.
If deductive knowledge is sufficient to do the job, why not simply close your eyes and deduce the correct moment to cross a road? Start with some axioms, put together a theorem and forget all about looking both ways… But we know from experience this is not going to work.
So getting back to the need to be able to falsify… if you put together a statement that is impossible to falsify, what this means is no possible empirical evidence can ever divert your belief in this statement. To all intents and purposes this is equivalent to crossing the road blindfolded.
So yeah, if we could observe people regularly deducing how to do everyday actions while completely blocking out all outside input, then we would have falsified one of the basic principles of empiricism. The statement is thus potentially falsifiable. Empiricism is a self-consistent philosophy.
As I pointed out above, it can be tested, and it doesn’t purport to be universally true. Someone tomorrow may discover a new way of thinking that elevates their capabilities above and beyond what anyone can do right now. Someone tomorrow may discover perpetual motion, they may discover faster than light travel, no scientist can prove otherwise.
What is meant by “falsify” is not simply that “there must be a way of determining the truth of a claim”. That’s not what we mean when we say that certain claims don’t need to be falsifiable.
Rather, what we’re objecting to is that all truth must be ascertained through empirical research.
Successfully navigating the street requires empirical knowledge, but some claims are not about the empirical realm, and so can not be empirically verified.
For example, you can’t test the Scientific Method using the Scientific Method. That’s circular reasoning. You can EXEMPLIFY the Scientific Method by applying it to empirical claims, but exemplifying is not the same as proving.
You test the Scientific Method using logic, while you test theories about the empirical realm using the Scientific Method.
“If deductive knowledge is sufficient to do the job”
Your error is in assuming that when one says that empiricism is not enough that they are also asserting that one cannot gain knowledge by empirical means. That’s not so. There is a place for deductive reasoning (and it DOES apply to the “real” world) and there is a place for empiricism.
“As I pointed out above, it can be tested, and it doesn’t purport to be universally true.”
On the contrary, one cannot test the proposition that knowledge can only be gained through empirical research. And the claim that the scientific method is the only way to know things about the physical world is a claim that purports universality.
How do you know, for instance, that looking both ways will work? How do you know that when you don’t see a car at T1 when looking left, there won’t be one that appears instantly and hits you as you cross? How do you know you don’t need to look up? Down? Behind you? How do you know that this time the physical laws that you accept as valid will continue to hold? You can’t observe that they continue to hold all the time…
I agreed with DF once when he said fractional reserve banking can exist without government and that depositors would rather earn interest on their savings rather than pay someone to hold their money. Consequently, I’m not opposed to FRB because even under an anarcho-capitalist system, a central authority would be required to prevent it. On the other hand, under a stateless society, banks would operate with higher reserves.
Here’s the “Peter Schiff was Right” video that Bob Murphy was talking about:
Peter Schiff Was Right 2006 – 2007 (2nd Edition)
http://www.youtube.com/watch?v=2I0QN-FYkpw
A fractional reserve system, being inflationary, is synonymous with counterfeiting, no?–regardless of whether the contract is fraudulent because of the multiple claims on the same property? Anyway, even if fractional reserves served the purpose in defraying the cost to mining gold, why forget to acknowledge that the production cost is important in developing the proper supply of the particular form of money on which the market has decided? I think it was David’s point that it is bad that the fractional reserve system today has no costs to defray–production of money doesn’t cost anything??? I think the Austrian view would be that it was always bad because in order to defray the cost to mining before, the system had to fundamentally change what kind of money is used because the notes would not be certificates of the real money, anymore. In other words, doesn’t the cost of producing money create the mechanism to have the correct supply of money?; and, if fractional reserves were used to defray mining costs but not printing costs then doesn’t that in and of itself distort the market?–because it is making money out of something that wasn’t decided on by the market. Is this making sense to anyone–if it does, then why didn’t Bob suggest it? Where am I getting this wrong? Tricia.
David Friedman sees the costs of mining gold as an unnecessary burden, rather than a valuable piece of information about consumers’ demand for gold.
As Joseph Salerno explains in a discussion about the Theory of Imputation, prices [or values] CAUSE costs, and not the other way around:
The Birth of the Austrian School | Josep T. Salerno
[WWW]http://www.youtube.com/watch?v=dZRZKX5zAD4#t=43m09s
A fractional reserve system misrepresents the costs of mining gold, resulting in artificial purchasing power being created in opposition to individual valuations.
Wanting to help with a question above, I thought it might be useful to quote a larger portion of an article by Murray Rothbard, but there wasn’t enough space to do so, there:
In Defense of “Extreme Apriorism”
[WWW]http://mises.org/daily/5195/
There’s more good stuff in the rest of the article.
On that point I agree with Rothbard, but there’s a big danger here which is to forget that we have taken a shortcut for “convenience of analysis” and jump to the belief that an absolute truth exists that all rational people must seek to maximize their profits… indeed, many economists go on to insist that any other behaviour is necessarily corrupt and evil and giving a job to your nephew is going to fray the entire fabric of the economic universe.
Right. But keep two things in mind:
1. What the Austrians mean by “rational” is that, believing certain means will attain their highest-ranked preference, people will employ those means.
People associate means with ends; And given that they at lease believe those means will result in the fulfillment of their preferences, it is rational – that is, it follows logically – that they would employ those means.
This is true for everyone. All human action is rational in this sense.
2. All people seek to maximize their PSYCHIC profit.
Meaning, they are always attempting to substitute a less desirable situation with the most desirable situation possible.
This is always the case, for everyone. It’s just that sometimes people consider the hiring of a family member more important than money profits.
The employment of his family member holds a higher rank than money profits on his preference scale.
I do not actually think it is necessary to introduce the fourth postulate, that is to say, that firms always aim at maximizing monetary profit (nor, once having admitted it into consideration for analytical convenience, as Rothbard says, to say that his theories will not be applicable in cases such as the “ne’er-do-well nephew”). The reason I say this is that an approximation of the fourth postulate is, in my view, already derived from the fundamental axiom and the other empirical postulates; that is to say, firms aim at maximizing psychic profit, which is manifested in an approximation of maximizing monetary profit, which in turn follows from the theory of indirect exchange and the preceding fundamental axiom/empirical postulates. Feedback from the ultimate constraint of profit and loss keeps this approximation of a maximization of monetary profit very close to an actual attainment of monetary profit.