19
Nov
2013
The Misesian Approach to Economic Laws
For my current course on Basics of Economics: Action and Exchange (where I go through my textbook Lessons for the Young Economist), I “pre-recorded”* one of the lectures because I was going to be in Chile that week and wasn’t sure my internet connection would allow for a live broadcast, which is the norm. Since we argue about these methodological issues all the time–it was the crux of my Porcfest debate with David Friedman–I thought I would share it with you folks.
And hey, if you like what you see, consider signing up for Part II of the course, where we cover the economics of a free-market economy, starting in January.
[*As George Carlin asked, does that mean I recorded it before I recorded it?]
Okay Bob is a serious question. Have you read QED by Feynman? For those who have not, he talks about particles wearing a wrist watch. (I think I have the right book I might be confusing books.)
Oops. To continue. This image of a clock or wristwatch is a pretty good analog of the phase of a particle. It explains the rules that physicists apply to combine phases and make predictions. But no one no one has ever said that the phase of a particle is directly observable. Quantum vectors and phases are simply not directly observable. But no one argues from that physics is not ultimately tested by the accuracy of its predictions. Even if we grant your premise that human beings act or that humans respond to incentives or any other such premise about acting man why should that exempt the subject from the need to be tested against empirical fact just as is physics or chemistry or any other science which might rely on fundamental motions which are not directly observable?
It is because of the nature of the propositions that the science comes up with. It does not tell you what reality will look like in absolute terms. It tells you what reality would be as a consequence of a particular approach (you may call it policy) as against what it would be were that approach (or policy) were not to be followed.
Take for instance the proposition that increase in money supply would lead to prices being higher. Here, the word higher stands for
In simple terms, it compares prices as they exist post implementation of the said policy with prices as they would have been had the policy not been implemented. The latter data is something you are in no position to gather as it does not occur. That is a counter-factual universe that does not exist.
The only data you would have gathered is that which is seen – prices before and after the implementation of the policy. Comparing these two, however, is to completely fail to test the economic proposition.
If you are still obsessed with “testing” economic propositions, you need to figure out a way by which you can come up with what the magnitudes of the economic phenomena (say prices) would be sans implementation of the said policy. These need to be empirical data, not numbers churned out by a model that attempts to mirror reality. In such a case, it is the model that is under scrutiny all the time, not the economic proposition.
The biggest problem all economic models face (as against models built in the natural sciences) is the absence of causal constants in human behaviour. A stone falls back to the ground in the same manner if thrown in the same manner under the same circumstances any number of times. Any other stone would do the same at the same time and at different times. A human being, in contrast, could react differently at different points in time. Different human beings could react differently at the same time under the same circumstances. So, attempts to model human behaviour, predict outcomes and test them against empirical observations must necessarily bite the dust. To insist that in spite of all this, empirical testing of economic propositions is a necessity for economics to be a science, do try to understand that you are not being scientific but scientistic.
If you still insist on being scientistic, be my guest.
Well if they really want to know, they should immediately scrap all Keynesian policy and replace tem with sound money.
They then can report their scientific findings to the world.
(1) The idea that “people respond to incentives” is not in any way based on empirical reality is so much nonsense.
It is nothing but a synthetic a posteriori proposition, and our prior experience of people.
(2) You still cannot distinguish between pure geometry and applied geometry, or pure mathematics or applied mathematics. A geometrical theorem in pure geometry has necessary truth and is not refuted by reality, but only because it remains analytic a priori and tells you nothing necessary true of the real world.
As soon as geometrical theorems are asserted as true of reality they get an empirical hearing. This is why we know the universe is ultimately explained by non-Euclidean geometry.
If Murphy thinks praxeology is like pure geometry, then Murphy is implying that praxeology is an analytic a priori system that says nothing true of reality.
(3) tell us whether you think that you can sit in arm chair and use deduction from the action axiom to determine how all prices are set in the real world and with apodictic truth?
An example I like to give is addition mod p. Its only experience that leads us to believe that addition as in peano arithmetic “works”. But maybe we live in a world where the real addition is mod p for a large p we haven’t observed yet.
Dear Lord Keynes,
“The idea that “people respond to incentives” is not in any way based on empirical reality is so much nonsense.”
Empirical reality also includes inner reflection. This is why the action axiom is unique; it comes from analyzing one’s own inner thoughts. Exact laws derived analytically are apodictic precisely because of this unique certainty the action axiom holds.
“If Murphy thinks praxeology is like pure geometry, then Murphy is implying that praxeology is an analytic a priori system that says nothing true of reality.”
If you say analytic propositions convey no information, you live in a bizarre universe. You use the principles of logic to make analytic propositions in your writing very often. Why do you do this if not to convey information? Why talk at all? According to you, we should just be constantly pointing at things.
“to determine how all prices are set in the real world and with apodictic truth?”
Economics is not about the prediction of specific prices, this is finance. Economics would be formulating a theory about how prices are formed.
(1) “If you say analytic propositions convey no information, you live in a bizarre universe”
I did not say that “analytic propositions convey no information”. You’re just making things up. Of course, a proper analytic proposition can convey information: they convey definitions, for instance: “all quadrupeds have four legs”. This is analytic a priori and does convey information, but it is tautologous. .
What I said is that an analytic a priori proposition says nothing *necessarily* true of reality. Learn the difference between necessary and contingent truth.
(2) “Economics would be formulating a theory about how prices are formed.
No problems: tell us whether you think that you can sit in armchair and use deduction from the action axiom to determine how all prices are **formed** in the real world and with apodictic truth?
Bob, even the physical knowledge is not the empiricist caricature that our opponents draw. To make an observation of a planet at a certain point in time you need to be sure at the point of observation that it actually moves in time and space, that it is the same planet that you decided to track at the start of the experiment.
That implies that science is not just, and not even primarily ‘guess and measure’. Any science is based on understanding that is not provable or falsifiable though experiment because such experiments require it. But there is no need for Kantian a priori categories, either. We need just to accept the obvious thing that our minds are capable (albeit usually imperfectly) of abstracting the nature of things, and that experiments are only needed for precision and ensuring that we haven’t missed something.
May I suggest that you consciously avoid using the word assumption when referring to the axioms? I do understand that you are trying to relate to people who are used to thinking of these starting points as assumptions but I am afraid this sloppy (pardon my use of this strong word) usage muddles things up in those very people’s minds.
I do understand that this is only a methodological quibble but it sounds important to me.
Are you saying that you do not assume the axioms are true? Or are you saying you don’t need to assume it, they are known to be true?
Yes.
An important point is that the action axiom can only be known empirically anyway:
http://socialdemocracy21stcentury.blogspot.com/2013/07/what-is-epistemological-status-of.html
And even Mises admits that the disutility of labour axiom is empirical too:
“The disutility of labor is not of a categorial and aprioristic character. We can without contradiction think of a world in which labor does not cause uneasiness, and we can depict the state of affairs prevailing in such a world …. Experience teaches that there is disutility of labor. ” (Mises 1949: 65).
Bob,
I guess you can see in Ken B’s response the point I made above.
Then you’re being foolish. I’m a trained mathematician, I understand axioms and axiomatics. I am trying to understand just what you think the difference is.
Whether an axiom is an assumption or otherwise is not an issue that mathematics settles or resolves. It is an epistemological issue. That you use the platform of mathematics to label me foolish adds no credibility to your statement.
You assumed I didn’t know what an axiom is as opposed to an assumption. All in all since I know you’ve read my comments for a long time, that counts as foolish.
But you keep dodging. Are you saying you don’t care if the axioms are true, because you are only doing deductions and don’t care about their truth, or are you saying the axioms are clearly true?
Where did I assume it? I inferred it from your statement.
Are you saying that you would take something as axiomatic even if you knew that it is not clearly true (please do not confuse this with clearly not true)? The response to that would be interesting.
The problem has always been and still remains your epistemology.
Just think about this question. Why would anyone ask me the question you asked me unless he thinks
1. I am a fool who doesn’t know an assumption from an axiom
2. axioms can even be considered as assumptions.
I was being fair to you to take it for granted that you wouldn’t start think thinking I am a fool. Maybe I was being too fair.
In mathematics, we use axioms all the time that may not be obviously true. Is the axiom of replacement obviously true? It states that if a class C has the same cardinality as a set X, then C is also a set.
As Keshav notes it is very common in mathematics to investigate axioms whose truth is unknown. It has been proven that the axiom of choice is independent of the other axioms of ZF set theory. Mathematicians investigate both ZFC and ZF~C as these are called.
Axiomatics is a tool and an approach, not a commitment to what is true.
At the risk of using 15% of the space in a post on Mises to talk about Ian Stewart, who is not a major figure in economics, or of being called patronizing for providing a suggestion, I recommend the book Concepts of Modern Mathematics by Ian Stewart.
I know I know, CONFIDENT TROLL!!
Mathematics uses the axiomatic method extensively. So what Mathematicians have to say about what axioms are is the last word on the issue. Epistemology be damned!!!
However, that does sound familiar. It’s just like saying that what scientists have to say is the last word on the scientific method. It’s sounds like “Science is what scientists do“. Someone out here seems to be practising consistency in context-dropping……
Do I hear someone say Randroid?
Having just completed listening to all 3 parts, I should confess that I just do not tire of listening to you running through these fundamentals over and over again. Thanks a ton for making this available (even if it is for shameless self-promotion).
Coutesy of Lord Keynes, from a discussion on his site. http://socialdemocracy21stcentury.blogspot.com/2013/11/a-simple-question-for-austrian.html?showComment=1384970656825#c3100228315950952898
When Mises’s Human Action was first published, one astute review called G. J. Schuller pointed out:
“Acceptance of Mises’ stated axioms does not necessarily imply acceptance of the ‘principles’ or ‘applications to reality’ which he has drawn from them, even though his logic may be impeccable. When a logical chain grows beyond the limits set by stated assumptions, it uses unstated assumptions. The number of unstated assumptions (axioms, postulates, or other) in Human Action is enormous. If Mises denies this,let him try to rewrite his book as a set of numbered axioms, postulates, and syllogistic inferences using, say, Russell’s Principia or, closer home, Von Neumann’s Theory of Games as a model” (Schuller 1951: 188).
Schuller, G. J. 1951. “Mises’ ‘Human Action’: Rejoinder,” American Economic Review 41.1: 185–190, at p. 188.
———
Austrians have had more than 60 years. They have sharp, mathematically able adherents like Robert P. Murphy. Any response to Schuller’s challenge yet?
Have you tried reading MESPM? It comes quite close.
You have a good sense of humour Bala, I’ll give you that.
Look Rothbard made some good arguments. I liked his multiplier example. But nothing I have seen in mes is remotely like formal axiomatics.
That’s why I said it comes close. If you apply your mind to what he has said, you will be able to figure it out for yourself. It’s not formal axiomatics, but comes close.
No it does not. Read the challenge again.
I did read it. That’s why I gave a qualified answer. If that’s not enough, there is nothing more I can do. At the same time, has any detractor managed to demonstrate that the conclusions are not valid because they depend on incorrect premises that were smuggled into the argument? If not, then what’s your point? 60 years is a looooong time for that as well.
p.s – LK’s rants don’t count. I just showed an example of how Rothbard demolishes his disutility of labour rant.
In fact, the challenge is in itself the product of a lazy mind that, frustrated by a deep inability to find the flaws in the arguments put forward, decides to place a big additional responsibility on the shoulders of the propounders of the argument so that failure to respond may be used as the deficiency on account of which the (non-)challenger may justifiably reject the theoretical framework being propounded.
Sounds interesting.
Of course the burden is on the person making an argument to justify it, not on its detractor to find a flaw. Otherwise you can just say “Blah blah blah. What’s the flaw in that?”
I agree, but will someone first find the flaws rather than make demands? If you notice, all Schuller has done is making sweeping statements without providing specific criticism. So merely saying “He has not presented it in this manner” is fairly childish IMO. Smear is not a valid technique of argumentation.
Keshav,
Try reconciling this
with this
and tell me why the latter is not smear if the author accepts the former to be true.
Sorry about the second post, but do focus on the word impeccable.
lol.. because the truth of the vast number of empirical propositions used as premisses can be questioned.
Unless you think that Mises derived everything in Human Action from action axiom with no other premisses by sheer magic, that is.
And your pathetic answer to Ken B above only underscores how Schuller was right.
If you think every theorem of praxeology is really formally derived by valid deduction from the action axiom, then you should be able to set it out formally.
Rothbard did no such thing.
Nor did he “come close”.
How about identifying the vast number of empirical propositions? Please do not drag out the corpse of the disutility of labour that Rothbard buried long ago.
Just to give you an example, our friend LK pointed out in the midst of his crusade against the Austrian approach that Mises refers to disutility of labour as empirical. Rothbard, in contrast, explains that labour implies the loss of leisure, a consumers’ good. He thus disposes of LK’s claim, though LK doesn’t realise it.
Try reading it. It’s quite good. Bob even wrote a study guide to that wonderful book.
Why is leisure a good and labor a bad and not vice versa?
Who said labour is a bad? I just said they are mutually exclusive. I just said there is a trade-off. Labour is a good as well – a producers’ good.
Why isn’t labor a consumer good for the person doing the labor? Why doesn’t leisure have a disutility instead?
Because labour is human effort applied in production to eventually be transformed into consumers’ goods?
Because leisure is the consumption of the means, time?
But here’s the fundamental question: why is the exertion of effort not a consumer good for the person doing the exertion?
That brings us to the purpose of the application which is known to the person alone. If the purpose is the (eventual) production of a consumers’ good, it is necessarily a producers’ good.
Human effort applied for the same of applying it is not a producers’ good because it is the very consumption of energy that is the end. No Austrian ever claimed that human energy expended in physical exercise is a producer’s good.
Will you next ask me why water applied in the manufacture of Coca Cola is not a consumers’ good for the person applying the water?
The question Keshav is asking is, why is it logically impossible for the exertion to be a good for the exerter?
Because if there is no reason why then it’s an empirical result not an a priori one.
First, the point is that even labour is a good for the person who applies it. It is a producers’ good that he offers and applies in the production process. I have been saying this from the beginning and am surprised that both of you are missing it.
Second, labour is by definition human effort expended towards eventual production of a consumers’ good. It is therefore logically always a producers’ good because it is so in the very definition of the concept labour. While plain human effort can be a consumers’ good, labour cannot because that would imply a contradiction – something being simultaneously a producers’ and a consumers’ good.
Just in case you think I am defining labour now and not in my initial response to Keshav, this is what I said then
Did you read this before you reiterated the question?