One More For the Road
I’m actually in a hotel room right now, and I need to stop hammering Krugman on the bond vigilante stuff. But I see Gene Callahan misunderstood the point of one of my recent posts, and so perhaps I need to step back and explain why this whole thing is so frustrating.
Krugman is saying that we need to show him a historical example of a country in exactly the same circumstances as the US today that suffered from a bond vigilante attack, before he’ll take the deficit scolds seriously. So this rules out all sorts of obvious examples of countries suffering from a bond vigilante attack, and arguably it rules out all examples. But Krugman’s rules for eligibility restrict things to a handful of countries during liquidity crises. Does someone want to come up with an estimate of the percentage of “country-years” for this eligibility? In other words, Krugman basically needs us to show him an example of the US, UK, Japan, and … ? suffering from a bond vigilante attack, at a time when interest rates were 0%, from 1971 onward, or during the interwar period when the major powers were still off gold. If we can’t come up with an example from this incredibly restricted sample of world history, we lose.
Beyond throwing out 99% (?) of the data, though, there’s something more insidious going on here: Krugman is telling us that a bond vigilante attack can’t harm us so long as interest rates are at 0%. Sure, once we leave the liquidity trap, then we might face hyperinflation–Krugman himself tells us this. But so long as we’re still in the liquidity trap, no problem. It’s not even theoretically possible that a bond strike can hurt.
Does everyone see why this is a really really bad argument? Remember that for Krugman, a liquidity trap is defined as a situation with high unemployment and 0% interest rates.
If I look at the price of petrol and decide I’ll just have to walk to work instead… does that make me an oil price vigilante? Should I weak a black mask like Zorro while I’m supposedly “attacking” by not buying something I can’t afford?
If I look at the price of prime steak in the supermarket and then decide to buy a tin of beans instead… does that make me a meat market vigilante? So the butcher is staggering around his shop with knife in one hand and meat axe in the other hand looking dazed.
REPORTER: What happened here?
BUTCHER: The meat market vigilantes attacked.
REPORTER: What did they do? Did you fend them off?
BUTCHER: Worse! They didn’t even show up!
You know how your mum told you always to turn the lights out when you leave the room? What are you, some kind of electricity market vigilante?!?
Absolutely right Tel! I also wonder how people can make up so colorful expressions for people who either just do not buy or at “worst” who bet with other people who are also not directly related if the price of some asset goes up or down.
Those evil vigilantes ATTACK. We need a BIG BAZOOKA to defend us!!
Or in other words: “Our best customers don’t like our products anymore? How dare they stopping to buy our stuff voluntarily. We need to force them to buy our great products” The sad thing is, this is exactly what they are doing. It is called financial repression.
This is so pathetic, really…
Not prostrating under the aura of the state is evil. No wonder statists like Krugman call the mere act of not buying what the state is selling “vigilantism”.
“We need to force them to buy our great products”
Actually it’s more like “If they won’t pay, then we need to devalue their money and gather more real resources before prices rise.”
Bond vigilantes …
James bond vigilantes.
They leave an economy shaken, not stirred.
Paper isn’t forever.
From Bernanke With Love.
Stockfall
The Man with the Fiat Gun
“No, Mr. Bond; I expect you to buy.”
Ken, you mean the man with the printed gun…
http://www.extremetech.com/extreme/133514-the-worlds-first-3d-printed-gun
Dr. No Interest Rate
Tel, this is a very nice way to cut through the snide hyperbole of Krugmans writing style and show in plain English what he is really saying. Well done.
I’ve lost track of how many FreeAdvice ground rules you broke here Dyspeptic.
That’s probably because he made more than the number of little piggy fingers you have.
And we’re back!
I’m still smarting from the last “woosh”. My comment above is part of the recovery process.
Glad to see we’re past denial.
🙂
Denial of what?
Cute.
80 years ago Mr. Hoover did his own anti-fiscal-terrorist thing, in a very gentle manner by today’s standards.
Today terror prevention teams would fish through our underwear looking for unspent $100 bills
“Seeing money-in-circulation increase by $800 million in 1931, Hoover engineered a coordinated hue-and-cry against “traitorous hoarding.” “Hoarding,” of course, meant that individuals were choosing to redeem their own property, to ask banks to transform their deposits into the cash which the banks had promised to have on hand for redemption.”
It is characteristic of depressions that, because of the inherently
fraudulent nature of the commercial banking system, any real
attempt by the public to redeem its own property from the banks
must cause panic among banks and government alike. And so, on
February 3, Hoover organized an anti-hoarding drive, headed by a
Citizens’ Reconstruction Organization (CRO) under Colonel
Frank Knox of Chicago. The hoarder is unpatriotic, ran the hue
and cry; he restricts and destroys credit (i.e., he is exposing the
unsound nature of the credit which was granted against his interests
and in destruction of his property). A group of top-level Anti-
Hoarding patriots met on February 6 to organize the drive: present
were General Dawes, Eugene Meyer, Secretaries Lamont and
Mills, A.F. Whitney, Alvanley Johnston, and industrialist Magnus
Alexander. The CRO urged hoarders to invest in short-term Treasury
securities, i.e., in unproductive rather than productive investments.
On March 6, Hoover delivered a public address on the evils
of hoarding: “the battle front today is against the hoarding of currency.”
Hoarding has lowered prices and incomes, and restricted
credit; it strangles our daily life. “No one will deny that if the vast
sums of money hoarded in the country today could be brought into
active circulation there would be a great lift to the whole of our
economic progress.” Hoover then commended Colonel Knox for
his “great battle against . . . the American people, and called on
everyone to serve in protection of the American home.”
_The Great Depression_, pp 306-307, Rothbard
Because: bond vigilantes would make the interest rate higher, thus exiting the trap
Don’t worry Murphy, even if we do find an example that fits Krugman’s (rather arbitrarily restrictive) criteria, then he will just say that these countries are examples where the governments imposed unnecessary austerity (e.g. rising interest rates, rising unemployment, etc), and so are not examples of bond vigilantes responding to deficits.
Keynesianism is not, contrary to protestations otherwise, a falsificationist doctrine.
“So this rules out all sorts of obvious examples of countries suffering from a bond vigilante attack, and arguably it rules out all examples.”
It does rule out all examples. History does not repeat in the same exact ways, and this is precisely why economics cannot be empirical. Krugman’s protestations are an implicit acknowledgement of this fact, but you will never hear him actually recognize what it is that he is relying on.
Because I enjoy the occasional quibbling:
Even if economic history did repeat, economics would still not be an empirical science. This is because the more fundamental reason is that human actors cannot even coherently regard themselves as past causally determined according to constancy in relations. For even if we suspended logic and for the sake of argument believed that our actions as based on constancy, and we tried to discover such constants through research, then that very approach of assuming we are learning would invariably carry with it a presuppostion that we as subjects are changing over time (i.e. we are learning). Since we would be compelled to prseume we are learning, and (crucially) our knowledge influences our actions, it follows that knowledge and actions cannot coherently be regarded as based on constancy even if we believed they did and attempted to exploit that belief.
In short, even if we tried to “falsify” the pesky Austrian argument of non-constancy, we would be compelled to consider ourselves as non-constant in that very attempt.
MF, I believe this is wrong. Science makes predictions but predictions are not guaranteed to come true. If you get 100 similar results it doesn’t mean the 101st will turn out the same, but you still have reason to believe that it will. (If the 101st goes wrong, I don’t see this is different from negative results in other sciences.)
This doesn’t mean there are necessarily economic laws, although perhaps there are, but I think you can still express predictions in probabilities even if there aren’t.
You’re right that knowledge may change outcomes, but I imagine it could change it to confirm more to the expected result rather than less so, but I believe these changes can themselves be modeled. (Maybe it’s a cat and mouse game, but unless people behave completely unpredictably I don’t understand how economics can’t have something to say.)
I think you missed the thrust of MF’s argument, there. His point was that constancy can be assumed with deterministic objects (like baseballs) in physics and the other natural sciences, but human beings can learn – and as learners, they will not act in a deterministic manner with a constant response to situations. Even if we try to demonstrate constancy, we necessarily must assume non-constancy in the process. It is a cogent argument along the lines of similar reasoning I have seen in Hoppe’s work.
I actually agree with it – but my reply would be that this is the reason that history does not repeat the same situations over again. In short, my argument was basically “because Y, therefore Z”, and – in my estimation – MF replied with “however, Y is a result of X, and this is an important thing to note, as it makes not-Y (and therefore not-Z) impossible”.
Matt, it’s an interesting topic, but even if humans can’t be *assumed* to act in a deterministic manner it doesn’t mean they won’t.
But my point was that you don’t even need deterministic laws, because you still have probabilities.
To take it in another direction though. If you want to argue that Keynesian economics is empirically wrong because people could act differently you still would need to say how you expect them to act. In other words, I’m not sure you can escape prediction. (If you don’t make a prediction, but people act differently anyway, obviously the model is wrong, but unless people *prove* to behave unpredictably, which is itself empirical, economics can still have things to say.
(Of course they could behave unpredictably, then in the future start behaving predictably, but you get the idea.)
It’s like you guys would say scientist can’t say anything about the future of human evolution because humans can influence their own evolution. (I guess I’m presuming we’re all assuming some sort of free will and/or consciousness or at least putting the issue aside.)
“even if humans can’t be *assumed* to act in a deterministic manner it doesn’t mean they won’t.”
To do so would contradict the ability to learn.
“But my point was that you don’t even need deterministic laws, because you still have probabilities.”
Probability – and this seems to be a difficult thing for certain kinds of people to grasp – is merely another form of determinism based on partially unknown information. It cannot apply to the decisions of human beings anymore than can static, deterministic laws.
“If you want to argue that Keynesian economics is empirically wrong”
But I am not arguing that, and never would. My argument was that Keynesian economics is *logically* wrong, and that economics cannot be empirically decided at all. Any prediction can only be a forecast based on understanding, and can always be wrong – WITHOUT proving or falsifying anything.
1913 Fed Reserve Act, 1947 end of Bretton Woods, 1971 Nixon Gold shock, nothing will ever repeat after these events. Not to mention the income withholding tax and direct election of senators which led to Obama Presidency. The world has forever been changed, precise historical events will never repeat, Krugman knows this…
Fully digital fractional reserve banking with less than zero reserves.
Digital Global Fiat.
This is new found territory. When Mises said inflationary policy was the most radical institution known to man, he was not kidding. His words were not an empty exercise. Mises could see the future with near clarity. I am not into false idols, but Mises was rare.
Krugman is a master at tricking us into chasing our tales rather than facing the encroaching statism head on.
Krugman is a full blown, die hard, statist and will do or say anything to protect his godless communism…
Matt, my phrasing was bad.
I should have said that people can behave “as if” by deterministic laws. (You said deterministic “manner” which I interpreted somewhat loosely.)
And you’re right you’re saying it’s logically not empirically wrong, but one thing I was trying to say was that saying humans will behave unpredictably is itself a prediction so it seems hard to escape prediction.
But Matt, I don’t know what you mean about probabilities being deterministic. (You mentioned a baseball earlier, what about baseball players?)
And I don’t know why you say that forecasting is “only” a forecast.
I think maybe you think I said something stronger than I did. I was responding to MF saying economics can’t be an empirical science, but you seem to the empirical science is the same as “hard” science.
But even the hard sciences can, at least in principle, be wrong. But even putting that aside I’ve never claimed economics will be decided once and for all.
So I don’t understand how you’re saying an empirical question (whether something can be predicted) can be logically wrong.
If a probability model fails the probabilities are updated. If you want to say they will fail every time because you believe logically they must, this is still an empirical question: whether they will fail or not.
It’s not enough to say, in theory, people can do something else. It’s a question of whether they will. And of course something may change in the future, but then the models will change, too. (Models can change the outcome themselves and this could perhaps be a problem.)
“one thing I was trying to say was that saying humans will behave unpredictably is itself a prediction so it seems hard to escape prediction”
It would be more accurate to say that any prediction can only be an educated guess. It’s not that humans will “behave unpredictably” – it is that there is no logical basis for any prediction of their action.
“I don’t know what you mean about probabilities being deterministic”
The nature of probability is such that it is a hard deterministic law about frequency of certain chains of events.
“You mentioned a baseball earlier, what about baseball players?”
Baseball players do not behave probabilistically – they have statistics recorded of the frequency of similar events in the past. But a player having a batting average of .250 does NOT mean that he will – in the future – hit 25% of all at bats. It tells you absolutely nothing about his future endeavors as a player, unless constancy holds (and in reality, we can see that it does not).
“I don’t know why you say that forecasting is “only” a forecast.”
It is an educated guess at the future based on understanding of the unchangeable aspects of the nature of the world, how they apply to the current state, and certain assumptions about human character that may or may not hold. It is not (and cannot be) deterministic, and thus can easily be wrong.
“So I don’t understand how you’re saying an empirical question (whether something can be predicted) can be logically wrong.”
Because that isn’t an empirical question, but an epistemological one. Logically, there is no way for prediction to be anything more than educated guesswork.
“If a probability model fails the probabilities are updated. If you want to say they will fail every time because you believe logically they must, this is still an empirical question”
This is not the case. For one, as history does not repeat, if you have a probability model that says in a unique situation, 50% will result in A, and 50% in B, you aren’t making any sense. But more importantly, the prediction cannot be more than a one-shot guess, based in but never proving theory.
I suggest reading Mises and Rothbard – they do a good job of explaining the problems with empirical study of economics.
anon:
people can behave “as if” by deterministic laws.
What does that mean? “People can behave ‘as if’ by deterministic laws”? Is that “as if” in the same family as “People can behave ‘as if’ they are being told what to do by an invisible talking unicorn”? If not, why not?
humans will behave unpredictably is itself a prediction so it seems hard to escape prediction.
One is a logically necessary proposition. The other is an empirical proposition subject to falsification. It is a rather misleading usage of the word “prediction” in the field of logical deduction.
But sure, if I argue that all place right triangles have inner angles totaling 180 degrees, then you could call that a “prediction” of place right triangles, but it’s really a logically necessary proposition inherent in right triangles themselves.
But Matt, I don’t know what you mean about probabilities being deterministic. (You mentioned a baseball earlier, what about baseball players?)
Probabilities are deterministic if you assert that A if followed by B, C, D, E, etc in particular probabilities represented by a percent or what have you. If you say B follows A 25.573% of the time, that is a deterministic proposition. All the supposed outcomes, B, C, D, E, etc total up to 100%.
But even the hard sciences can, at least in principle, be wrong.
The concept of constancy in relations is never addressed. The only leeway falsificationist methodology gives is whether or not the theory fits the data.
But even putting that aside I’ve never claimed economics will be decided once and for all.
That’s precisely the problem. By keeping economics in a “skeptical, let’s wait and see, and we can’t know anything for certainty until we try it”, presupposes social engineering and control of markets. It leads to such absurdities as “We can’t know for sure for sure that the state imposing a price floor on labor that is above the market clearing rate of labor, will generate unemployment. No, we have to TRY to raise minimum wage by 100,000% before we can know for sure. And if it is followed by massive unemployment, then that won’t definitely show anything either, because maybe we forgot to control for certain factors, or maybe we included too many control factors.
It also leads to “We have to try Socialism first we can know for sure whether or not there will be a price system for the means of production.”
So I don’t understand how you’re saying an empirical question (whether something can be predicted) can be logically wrong.
It’s logically wrong when a constancy based model presupposes non-constant learning of the outcome of utilizing, “testing” said model. That’s the contradiction. You can’t apply a constancy based model to knowledge and learning itself. It can only be applied, if it is going to be applied, to non-acting phenomena like atoms and molecules.
If a probability model fails the probabilities are updated.
That process presupposes constancy in the truth of things over time. You’re saying that if a theory proposed yesterday, is not observed today in the way predicted, then the model “fails.” But why should it be considered a “failure” if the truth of things yesterday were different such that yesterday the model is correct, and today it is not correct?
Only if you make the a priori proposition that the truth of things are unchanging, can you even claim that a theory proposed in the past, is “falsified” or “confirmed” in the present.
If you want to say they will fail every time because you believe logically they must, this is still an empirical question: whether they will fail or not.
What about the question of constancy? That is not an empirical question, because constancy is required in the very process of empirical approaches.
It’s not enough to say, in theory, people can do something else. It’s a question of whether they will. And of course something may change in the future, but then the models will change, too. (Models can change the outcome themselves and this could perhaps be a problem.)
It’s a “problem” that people learn of the very models used to attempt to predict them? I should think that is a benevolent, truly remarkable outcome.
Ok, guys, I’m not sure where to begin here, but it seems to me all you guys are saying is that economics is not a hard science like physics and then acting like I said it was just like physics.
If you guys want to say economics is “only” “educated guesswork” then that seems good enough, I guess.
I’m certainly no expect on statistics or probability but it seems you guys outright reject Bayesian statistics or at least think it is “merely” educated guesswork. But look, if it’s the latter, I’m not looking to get bogged down in endless semantic arguments about whether soft sciences are really science.
You guys want to make it seem like I said that all human action is deterministic. MF, you keep using the word “constancy” as if I said all human action is always “constant.”
But I’m not saying that, it’s seems it’s you guys who want to say humans show no constancy at all. That seems like a very strong claim. You guys act like the fact that an economic model may not be useful 400 years from now then we must throw it out today.
You don’t have to assume “truth is unchanging” only that it hasn’t changed.
You guys seem to allow that humans face some constraints in their actions (at least currently) then it seems you should allow that these constraints open up a space where economics can make simplifying assumptions about those constraints to say something about how they will act. (Not to say they can’t be wrong.)
It seems economics data by itself could overwhelm the ability of the individual to act “unpredictably” because they simply cannot keep track of all the predictions made about them. (If Google successfully predicts your search term do you throw your copy of Mises at the computer screen.)
It’s a “problem” that people learn of the very models used to attempt to predict them? I should think that is a benevolent, truly remarkable outcome.
Note that I was thinking about whether a prediction may “cause” the outcome. Like if a probability for the outcome of the election caused people to vote or not vote either because they thought they would win or thought they would lose. (Perhaps these people cancel each other out.)
I wondering what you guys think of Nate Silver’s election probability now.
anon:
it seems to me all you guys are saying is that economics is not a hard science like physics and then acting like I said it was just like physics.
We’re obviously saying more than that. We’re describing WHY the methodology of economics isn’t like the methodology of physics, and that “why” is the crucial argument.
If you guys want to say economics is “only” “educated guesswork” then that seems good enough, I guess.
I am not arguing that. I said economic propositions are logical categories of action. Your conception of economics is what is guesswork.
I’m certainly no expect on statistics or probability but it seems you guys outright reject Bayesian statistics or at least think it is “merely” educated guesswork. But look, if it’s the latter, I’m not looking to get bogged down in endless semantic arguments about whether soft sciences are really science.
Why are you responding to the argument that your conception of economics is guesswork, with the claim that you don’t want to argue over semantics?
You guys want to make it seem like I said that all human action is deterministic.
You did say that. Although you used the weasel words “as if” we’re deterministic.
MF, you keep using the word “constancy” as if I said all human action is always “constant.”
No, I keep saying constancy “as if” you keep talking about expected results, predictions, models, of which presume constancy in action.
But I’m not saying that, it’s seems it’s you guys who want to say humans show no constancy at all.
What do you mean “humans show no constancy at all”? Remember, I have only said human action is not constant. I didn’t say anything about biology.
Do you have a problem with the argument that human action is not constant?
That seems like a very strong claim. You guys act like the fact that an economic model may not be useful 400 years from now then we must throw it out today.
Why do you keep saying “seems” and “as if” all the time? Why aren’t you engaging the actual arguments being presented? Why do you keep making assertions as to what we “seem” to be saying, and what “as ifs” our arguments somehow imply.
You use a very evasive and non-engaged manner of discussing these issues.
You don’t have to assume “truth is unchanging” only that it hasn’t changed.
If you utilize constancy in your theories, then you are in fact assuming that truth is unchanging. The present use of the constancy model assumes present unchanging truths.
You guys seem to allow that humans face some constraints in their actions (at least currently) then it seems you should allow that these constraints open up a space where economics can make simplifying assumptions about those constraints to say something about how they will act.
The certainties are the logical categories (form). The uncertainties are the actual decisions made (content).
It seems economics data by itself could overwhelm the ability of the individual to act “unpredictably” because they simply cannot keep track of all the predictions made about them.
Nice try. However, if you refer to data, then you refer to known data. If the data is known, then it does not “overwhelm” the actor. The actor makes use of the information.
(If Google successfully predicts your search term do you throw your copy of Mises at the computer screen.)
No, becase Google doesn’t predict. It recalls websites based on your input, according to a constancy based algorithm.
Note that I was thinking about whether a prediction may “cause” the outcome.
That comment is a good example of why I keep saying “human action is not constant” all the time.
Like if a probability for the outcome of the election caused people to vote or not vote either because they thought they would win or thought they would lose.
Probability is based on actual choices. It is not independent of choices, after which choices are influenced by the probability. It’s going in circles.
You did say that. Although you used the weasel words “as if” we’re deterministic.
Actually, I just reread Matt’s post that I was originally responding to and he does say that we could try to demonstrate constancy if we assume non constancy “in process”, so my response was probably unwarranted.
But that was my point. Not that all human behavior is necessary constant but that some human behavior may be constant enough in practice that it can be used to make useful predictions.
You however appear to rule that out and for some reason keep trying to imply I originally was saying something much stronger.
Why are you responding to the argument that your conception of economics is guesswork, with the claim that you don’t want to argue over semantics?
Because from what I can tell you’re only criticism of what I said is that economics is either not a hard as science as physics or possibly even pure mathematics or logic. Since I never claimed it was, I don’t know what your point is except to argue endlessly over definitions or “natural rights”.
Do you have a problem with the argument that human action is not constant?
Only if it means it can’t be constant. Not if just mean it shouldn’t be assumed to be constant.
Why do you keep making assertions as to what we “seem” to be saying, and what “as ifs” our arguments somehow imply.
I’m not a mind reader MF. I’m also not omnipotent. It’s strange that you express (what “sounds like”) hostility towards someone expressing doubts as to their understanding of another’s position. If I’m misrepresenting you, please say so. I don’t think that’s being evasive, this isn’t a scientific journal, we’re having what is, at least on my part, a somewhat off the cuff discussion.
Nice try. However, if you refer to data, then you refer to known data. If the data is known, then it does not “overwhelm” the actor. The actor makes use of the information.
I don’t know why you assume the actor must “make use of” the information or why this would necessarily change their behavior. But I think you’re also overlooking predictions about large groups of people where any individuals behavior won’t necessarily change the group’s behavior in any significant way at least for the purpose of the prediction.
No, becase Google doesn’t predict. It recalls websites based on your input, according to a constancy based algorithm.
This “seems like” another semantic argument where you’re changing “predict” to “recall”. You seem to want to say probabilities are only valid when they’re deterministic, even though deterministic probabilities are not necessarily very interesting. (A fair coin toss is 50/50, so what?)
But doesn’t Google update their algorithms? How is this different from updating an economic model with new results or data?
Probability is based on actual choices. It is not independent of choices, after which choices are influenced by the probability. It’s going in circles.
MF, I think you need to look up and read about Bayesian probability and prior probabilities. I never said probabilities are “independent of choices”. But there may well be a “circularity” to the human sciences, that’s why I said it could be a “problem”.
I meant I’m not all knowing above rather than all powerful, but you get the idea.
And just to be clear when I object to the idea that “human action can’t be constant” I’m not saying human action might be constant, only that some human action might be constant enough that we can (for whatever reason) make useful predictions based on it.
I’m not saying even if we can make some useful predictions, that even in those predictions we should assume “constancy” (that are prediction can’t be wrong) only that the predictions are possible and possibly useful.
“But my point was that you don’t even need deterministic laws, because you still have probabilities.”
Look into the issue of case probability:
http://mises.org/humanaction/chap6sec4.asp
That comment is directed at anon.
anon:
Matt, it’s an interesting topic, but even if humans can’t be *assumed* to act in a deterministic manner it doesn’t mean they won’t.
If you understand that we are necessarily not able to even coherently regard our actions as past causally determined; that you are presupposing you are capable of learning; that we must, by virtue of how we think, regard our actions as not past causally determined, but “free”…then, even if our actions really were past causally determined and even if our actions could be predicted by a super-human entity with a mind that is able to transcend action, and transcend logical constraints of non-contradiction, identity, and excluded middle, i.e. God, then we as actors will never be able to know such things. Insisting that we are past causally determined would become a religious doctrine.
Freedom of thought would remain a necessary illusion for actors even if our actions really were past causally determined from the perspective of a God. But since there is no apodictic proof that can be apprehended by us, it follows that there is no scientific reason to assume that were are in fact past causally determined.
But my point was that you don’t even need deterministic laws, because you still have probabilities.
As noted above, probabilities does not get around the constancy in relations. It just restates it. Instead of “B follows A 100% of the time”, it becomes instead “C, D, E, etc follow A 100% of the time, where C follows 20% of the time, D follows A 30% of the time, E follows A 5% of the time, etc”.
To take it in another direction though. If you want to argue that Keynesian economics is empirically wrong because people could act differently you still would need to say how you expect them to act.
This is incorrect, and quite frankly a rather silly demand. If I reject the notion that predictions can be made on the basis of constancy in relations, then can’t you see how absurd it would be to demand that the person who rejects it must himself come up with a constancy based prediction mechanism of his own?
Or did you mean a non-constancy based prediction? How would that work? How can I predict your future knowledge and actions, if I am not to utilize constancy? I would be unable to do it, because I would have to ground all my predictions on my existing knowledge, and any relation between my existing knowledge, and my future actions, would invariably be presumed as constant.
Now, this does not mean that I am unable to know or say anything about your future actions. If there wasn’t something to be known a priori about my future actions, then acting itself would be impossible. Action would be a chaotic mess. The world would appear as completely unintelligible. What I can know about my future actions are the logical categories that constrain all possible actions. Action itself is delimited, finite, and definite. These categories of action, such as the law of marginal utility, profit, loss, cost, gain, etc, are the conceptual building blocks of action itself. No action can be regarded as violating these concepts, because they are a part of action itself.
Mises argued that the best elaborated science of action in general, is economics. Economists are the ones who learn and teach the logical constraints of action. They don’t predict the content of action based on constancy, but rather they discover the form of action.
For example, I cannot know in advance what your future knowledge and hence actions will be regarding the economic good table salt. There is no information today that I can use to know what the demand for table salt will be next year. I cannot know this because I cannot know your future path of knowledge, in the present. What I do know however are the logical constraints that are associated with table salt as an economic good. I know that the law of marginal utility will be true for table salt next year, 10 years from now, a trillion years from now, for eternity. I know this because the law of marginal utility is grounded in action itself. As long as action exists, so will the law.
In other words, I’m not sure you can escape prediction. (If you don’t make a prediction, but people act differently anyway, obviously the model is wrong, but unless people *prove* to behave unpredictably, which is itself empirical, economics can still have things to say.
Actually, that people learn unpredictably is not an empirically derived understanding. It is in fact an a priori derived understanding, grounded in self-reflection, not observing people’s brain waves on a scanner.
And nobody is saying that without predictions according to constancy, that “economists have nothing to say.” On the contrary, economists do have something to say. They have something to say about the form of actions, not necessarily the content of actions.
What is true is that positivist minded economists, those who try to mimic the method made successful in physics and chemistry, those economists have nothing to say about economics. What they do have something to say about is economic history. They research what they and others did in the past, which is a function of what they and others knew in the past, and what their actions in the past. They are not discovering any “laws” of economics through looking in the rear view mirror. They are not discovering a constancy in human action that can enable them to predict future actions based on that constancy. As a stylistic argument, just ask yourself why economists are not the wealthiest people in the world, if they really could discover constants of action. Why are so many of them wasting time in windowless offices, working for peanuts, when they could be utilizing these allegedly discoverable constants of action to make predictions in the market and make themselves billionaires?
Entrepreneurship is an art, not a science. Nobody can learn to be like Bill Gates through textbook study, the way physicists and chemists can learn to be like their role models through textbook study.
(Of course they could behave unpredictably, then in the future start behaving predictably, but you get the idea.)
It’s more fundamental. You cannot even coherently regard knowledge and action as predictable according to constancy. The very concept of learning precludes it.
It’s like you guys would say scientist can’t say anything about the future of human evolution because humans can influence their own evolution. (I guess I’m presuming we’re all assuming some sort of free will and/or consciousness or at least putting the issue aside.)
No, evolutionary biologists can’t say anything about human KNOWLEDGE and ACTIONS by utilizing constants. You are falsely generalizing from an argument that is limited to knowledge and actions. If I argue X is not constant, it makes no sense to respond with “It’s like you guys believe Y is not constant.”
MF, I don’t want to have a philosophical discussion about free will.
I was responding to Matt saying humans cannot act in a deterministic manner.
I was trying to say even if human actions are not determined it doesn’t mean human won’t act as if they were determined.
As noted above, probabilities does not get around the constancy in relations. It just restates it. Instead of “B follows A 100% of the time”, it becomes instead “C, D, E, etc follow A 100% of the time, where C follows 20% of the time, D follows A 30% of the time, E follows A 5% of the time, etc”.
I don’t get this. You guys understand probabilities can be changed.
Or did you mean a non-constancy based prediction?
MF, you’ve posted a very long post where you’ve used the word “constancy” what seemed like dozens of times. You seem to think I’ve said somewhere that humans can’t learn or act differently or that economics is an all knowing God or future conditions can’t change in fundamental ways.
You seem to have some strange ideas about what economics is. (I don’t know what your point about Bill Gates alleged entrepreneurship is.) You seem to have set up some arbitrary distinctions, but I don’t have time to respond in full right now, and so will try to return to this later.
(I’m wondering why you think economists should be so much richer.)
“I was trying to say even if human actions are not determined it doesn’t mean human won’t act as if they were determined.”
This is a spurious distinction. If human beings are not determined in their actions, then any “law” formulated regarding such action will be the meaningless tautology of “if people choose X, they will have chosen X”.
“I don’t get this. You guys understand probabilities can be changed.”
Then they are meaningless as anything more than historical statistics. If I claim you will choose chocolate ice cream 50% of the time, and it turns out that of the next 10 times you have ice cream, you never choose chocolate, what good is my probabilistic system? Either I have some grain of truth – that people *must* act in accordance with certain probabilities – or I am merely guessing off some piss-poor assumptions.
“You seem to have some strange ideas about what economics is.”
I think that it is you that have a very strange idea of economics. Particularly in the realm of epistemology, and the distinction between derived logical analysis and empirical statistic nonsense.
MF, I don’t want to have a philosophical discussion about free will.
No, you’re rather just deny it and reject it and assume that human knowledge and action are deterministic. Remember, you are already in a debate about determinism versus free will the moment you criticized the Austrian approach on the basis of determinism.
I was responding to Matt saying humans cannot act in a deterministic manner.
I was trying to say even if human actions are not determined it doesn’t mean human won’t act as if they were determined.
As asked above, what does that mean “as if” human are determined? How can you distinguish between what humans are actually doing because of something true, versus what they are actually doing because of something “as if” it were true?
It makes little sense to accept the logic that people learn, and are hence not constant, and then right after say “Yeah well, that doesn’t mean people won’t behave “as if” their knowledge and actions are based on constancy.
“As noted above, probabilities does not get around the constancy in relations. It just restates it. Instead of “B follows A 100% of the time”, it becomes instead “C, D, E, etc follow A 100% of the time, where C follows 20% of the time, D follows A 30% of the time, E follows A 5% of the time, etc”.
I don’t get this. You guys understand probabilities can be changed.
You do understand that changing probabilities does not remove their constancy nature. If you hold a set of probabilities as valid for today only, then you are holding them as constant for today.
MF, you’ve posted a very long post where you’ve used the word “constancy” what seemed like dozens of times. You seem to think I’ve said somewhere that humans can’t learn or act differently or that economics is an all knowing God or future conditions can’t change in fundamental ways.
You seem to have some strange ideas about what economics is. (I don’t know what your point about Bill Gates alleged entrepreneurship is.) You seem to have set up some arbitrary distinctions, but I don’t have time to respond in full right now, and so will try to return to this later.
(I’m wondering why you think economists should be so much richer.)
This is too hasty and empty to warrant a response.
It makes little sense to accept the logic that people learn, and are hence not constant, and then right after say “Yeah well, that doesn’t mean people won’t behave “as if” their knowledge and actions are based on constancy.
See you’re assuming that learning alters behavior when what someone learns may not alter their behavior significantly. (Or at least the rate of change of behavior (in at least some cases) may be slower than our ability to make predictions about that behavior.)
anon:
See you’re assuming that learning alters behavior when what someone learns may not alter their behavior significantly.
So you agree it alters behavior. I never said the altering is “significant” instead of “insignificant”, whatever that would mean.
(Or at least the rate of change of behavior (in at least some cases) may be slower than our ability to make predictions about that behavior.)
The rate of change is equal to the rate of learning.
So you agree it alters behavior. I never said the altering is “significant” instead of “insignificant”, whatever that would mean.
Significant towards the prediction. I said that because I was anticipating, based on what you said about “human action”, you say something like what you said next:
The rate of change is equal to the rate of learning.
MF, the rate of change of the outcome of the prediction, not the change in what the person “knows”. Perhaps you can define “learning” as a change in behavior (whether predicted or not), but that’s not what I’m saying.
Matt:
I actually agree with it – but my reply would be that this is the reason that history does not repeat the same situations over again. In short, my argument was basically “because Y, therefore Z”, and – in my estimation – MF replied with “however, Y is a result of X, and this is an important thing to note, as it makes not-Y (and therefore not-Z) impossible”.
Precisely.
anon:
I noticed Matt Tanous answered your arguments below, after I typed out my response. So this response is not to “overrule” Tanous in any sense, but merely to reiterate and restate.
MF, I believe this is wrong. Science makes predictions but predictions are not guaranteed to come true. If you get 100 similar results it doesn’t mean the 101st will turn out the same, but you still have reason to believe that it will. (If the 101st goes wrong, I don’t see this is different from negative results in other sciences.)
This is called the problem (some might call “fallacy”) of induction. Only if you assume that B following A in the past implies that we have good reason to believe B follows A in the present as a matter of determined causality, i.e. the very concept of constancy under discussion, does your claim “you still have reason to believe” even make sense.
Yet my argument that there is not constancy would nevertheless remain unchallenged, because you would still be presuming to be a learning subject that allegedly learns that B following A in the past and that this knowledge allows you to conclude that B would follow A in the present.
This doesn’t mean there are necessarily economic laws, although perhaps there are, but I think you can still express predictions in probabilities even if there aren’t.
Predicting that B follows A in 50% of the cases, C follows A in 30% of the cases, and D follows A in 20% of the cases, is still presuming constancy in relations. You can’t even suggest definite probabilities for B, C, and D, unless the concept of constancy remains.
You’re right that knowledge may change outcomes, but I imagine it could change it to confirm more to the expected result rather than less so, but I believe these changes can themselves be modeled. (Maybe it’s a cat and mouse game, but unless people behave completely unpredictably I don’t understand how economics can’t have something to say.)
How can your future path of learning and actions based on that learning be “modelled” in anything other than constancy? Let’s suppose for the sake of argument that you did claim to discover such a model. That very action would, as expected, necessarily entail a learning subject (you) that would a priori know of his own future path of learning before he actually goes out and learns it. But that would imply that you are able to apprehend knowledge in advance of you actually going out and learning that knowledge. That is another contradiction. Knowledge cannot coherently be regarded as anything other than not based on constancy in relations, and hence unpredictable according to such (non-existent) constancy.
Bob, Krugman is making a prediction that there won’t be a sudden spike in interest rates, but if there is it would be stimulative.
He did provide the historical example of France in the 1920s, but even though it is not an exact match, he claimed things turned out pretty well for them. (You’re disputing his interpretation, but this is not the same as not providing an example at all.)
I think his saying France’s situation is different is not that just that if we had a bond strike now it wouldn’t be as bad as them (I’m not even sure he’s saying this), but that we won’t even have a bond strike. (Remember Krugman doesn’t think there will be a bond strike now, he’s saying *if* those other guys are right then it shouldn’t (doesn’t have to be depending on policy response?) the catastrophe they say it will.)
But Krugman does have an example of a country with high debt at the zero lower bound *not* having a bond strike: Japan. So I don’t know what you mean there.
That said Gene’s right that your transition vision post raises an interesting question. But even assuming you’re right about France (that the austerity was necessary rather than gratuitous), there would be the question of whether France was better off anyway. (One year of high unemployment may be preferable to five years after all.) Plus all of this seems to fall roughly in the idea of standard countercyclical policy. (That is we’re depressed now and need stimulus, but later we may overheated and need to slow down.)
I have a naive question.
The Austrian definitions of “economic” and “value” are voluntary and subjective. Why engage people over GDP at all? It is a simple argument to understand that more GDP is not more valuable. Government policy isn’t voluntary, and can therefore never be economic.
etc etc…
It’s useful to criticize non-Austrians on their own terms. If we can concede a number of assumptions to them and still prove how wrong they are, it strengthens our position.
“If we can concede a number of assumptions to them and still prove how wrong they are, it strengthens our position.”
No. It only weakens their position.
Our position must still be built upon many precepts they reject outright – like GDP being unimportant. And if we grant them these for free, what is left to build on except their garbage?
Agreed, but maybe he meant it strengthens our relative position.
Spot on. I hate getting into discussions where people take GDP seriously for the same reason.
“Krugman is saying that we need to show him a historical example of a country in exactly the same circumstances as the US today that suffered from a bond vigilante attack, before he’ll take the deficit scolds seriously”
Does anyone have a link to the place where Krugman asks for this ?
I think Bob is characterizing this, which he quoted PK:
“So let’s step back for a minute, and consider what’s going on here. For years, deficit scolds have held Washington in thrall with warnings of an imminent debt crisis, even though… the historical record shows no examples bearing any resemblance to our current situation in which such a crisis actually did happen.”
Yes, the problem is that “resembling” and “exactly alike” are not synonyms!
When current deficits are advocated, and past “success” stories have to be lauded, then the argument is that the past “resembles” the present.
When deficits are advocated, and past “failure” stories have to be dismissed, then the argument is that the past “is not exactly alike” the present.
You can never win!
That’s precisely the point.
Paul Krugman keeps talking about bond vigilantes; It’s like he’s inadvertently acknowledging that Peter Schiff was right about the coming bond market crash:
Don’t believe the hype — the U.S economy is not recovering, it’s getting sicker.
http://www.youtube.com/watch?v=Vcd5IO_LMBs
Peter Schiff: Ben Bernanke is Public Enemy No. 1
http://www.youtube.com/watch?v=0LC1RIljtp8
Ron Paul predicted a bond market crash, too. I just don’t have a link at the moment.