Models Do Not Prove Anything About the Real World, Ever
That, I think, is the take away truth from the whole debt debate. Nick Rowe rather arrogantly kept insisting if people just studied the OLG models they would "get it." But Steve Landsburg, Noahpinion, Ken B., Rob, and me, amongst others, did study these models. (I suspect Krugman has as well.) I know I have spent many hours looking over Bob's models, because I would much rather admit that I was wrong than go on being wrong! The problem is, we just don't interpret the meaning of the model in the same way that Rowe does. Of the people I named above, I think every one of then totally understands the flow of apples through Bob's models. What we disagree about is what this represents. The "Rowers" think it shows the burden of government debt being passed forward through generations, while we "Landlubbers" believe the model shows the effects of a series of inter-generational debt transfers each occurring at a single point in time, something that Lerner and Krugman acknowledge can occur.
The great Lewis Carroll nailed this truth about models in his brilliant "What the Tortoise Said to Achilles." The tortoise keeps accepting Achilles's models, while disagreeing about what they prove. The point here is not that the tortoise is right: Carroll was a great logician, and certainly was not using the tortoise as his mouthpiece. The point, rather, is that models do not interpret themselves: I can fully grasp your model (ahem, Professor Rowe) while disagreeing with you about what the model demonstrates.
And the point of this post is not to claim that the "Landlubbers" have been right all along, but to comment on the nature of models. A logically consistent model is a lovely thing. But by itself, it proves nothin' about nothin' except about things internal to the model. No model can contain its own interpretation, since things the model contains are internal components of the model, and thus cannot be interpretations of it: as components of the model, they themselves will need their own particular interpretations within the framework of interpreting the model as a whole.
The great Lewis Carroll nailed this truth about models in his brilliant "What the Tortoise Said to Achilles." The tortoise keeps accepting Achilles's models, while disagreeing about what they prove. The point here is not that the tortoise is right: Carroll was a great logician, and certainly was not using the tortoise as his mouthpiece. The point, rather, is that models do not interpret themselves: I can fully grasp your model (ahem, Professor Rowe) while disagreeing with you about what the model demonstrates.
And the point of this post is not to claim that the "Landlubbers" have been right all along, but to comment on the nature of models. A logically consistent model is a lovely thing. But by itself, it proves nothin' about nothin' except about things internal to the model. No model can contain its own interpretation, since things the model contains are internal components of the model, and thus cannot be interpretations of it: as components of the model, they themselves will need their own particular interpretations within the framework of interpreting the model as a whole.
Gene: I built my latest model just for you!
ReplyDeletehttp://worthwhile.typepad.com/worthwhile_canadian_initi/2012/10/another-debt-burden-counterexample.html
BTW, when I finally figured out Noah's model, this is what it really says: "Debt does not create a burden on future generations, because as long as there are still some people left alive who benefited from the deficit, we can still tax those people to pay off the debt. So it must be our failure to do that, rather than the debt, that causes the burden on future generations".
"Gene: I built my latest model just for you!"
DeleteWell, I feel very special.
But what I especially feel is my comment above is just right. I totally follow your model. I agree total utility, given your assumptions, is lower than it would be without...
Well, you say, "Without the debt."
I say, "Without the intergenerational transfers of goods, to which the debt is a tangential matter."
I.e., we could get the exact same pattern of utility lowering with pure transfers and no debt, as I believe Landsburg pointed out months and months ago.
I.e., we could get the exact same pattern of utility lowering with pure transfers and no debt, as I believe Landsburg pointed out months and months ago.
DeleteYou can't have it both ways, Gene. At my blog in the comments, you were astounded that I didn't see the full importance of the *internally held* nature of the debt, to Krugman and Lerner's argument.
But as you well know, Landsburg thinks it is *irrelevant* if foreigners or Americans hold the bonds. So they are clearly making different arguments.
Also, Krugman would throw up if he read your post right here. The point of models is to clarify thinking and make sure you're not saying something internally inconsistent. If you were someone disagreeing with Krugman's conclusions, and said, "I don't care about what happens in a model, they never change minds, we just all interpret things differently, maaaan," Krugman would think you were worse than a creationist.
Last point: If Nick's models are so obvious and insignificant, isn't it odd that I am running around saying I made mistakes in my textbook? Isn't that a weird thing for me to be doing? And why did Dean Baker throw in the towel and concede that Nick had raised an interesting theoretical possibility, but that Dean still thought it wasn't empirically that important?
In conclusion, if you want to say, "I disagree with Nick and Bob, and here's why..." OK fine. But don't act like everybody *else* who disagrees with us, is saying the same thing you are. No, there are at least 4 different camps on this issue, by this point.
"t my blog in the comments, you were astounded that I didn't see the full importance of the *internally held* nature of the debt, to Krugman and Lerner's argument"
DeleteNo, I wasn't.
"Also, Krugman would throw up if he read your post right here. The point of models is to clarify thinking and make sure you're not saying something internally inconsistent. If you were someone disagreeing with Krugman's conclusions, and said, "I don't care about what happens in a model, they never change minds, we just all interpret things differently, maaaan,""
But I said nothing like that. I would agree with the view you ascribe to Krugman about what models are for. I've spent most of my adult life building models!
"Last point: If Nick's models are so obvious and insignificant..."
I never said anything remotely like "These models are obvious and insignificant"!
"isn't it odd that I am running around saying I made mistakes in my textbook?"
You misinterpreted the meaning of the model!
Bob, since you so wildly misunderstand what I've been saying, it is absolutely understandable why you think the objections to Rowe's conclusions are so divergent.
"I.e., we could get the exact same pattern of utility lowering with pure transfers and no debt, as I believe Landsburg pointed out months and months ago."
DeleteYep. This really is a key point. Bob argues that debt imposes a burden all by itself above and beyond nay burden transfers impose. He has to argue that to show PK is wrong. Yet that is the one thing he *cannot* show.
Gene, I chime in on your side here http://consultingbyrpm.com/blog/2012/10/my-final-word-this-generation-the-debt-really-is-fundamental.html#comment-47815
ReplyDeleteRe your last paragraph:
ReplyDeleteNot sure I understand what you mean. Do the components of the model have to be interpreted in the same way that signs in general (including combinations of letters (i.e. words)) have to be interpreted* for them to become meaningful and represent something? If so, then your point is true, but relatively trivially so.
If not, then in what way would for example the components of a simple model of the solar system or of an atom have to be interpreted other than in the same way as the words that are typically used for the entities and relations in the solar system or the atom have to be interpreted?
*the interpretation of course comes through use, not through a link with another representation. h/t Wittgenstein
You should just read it in the context of "Can a model contain its own interpretation."
DeleteWittgenstein would be with me here, I suspect; see his work on following a rule.
Okay, that doesn't really answer the question or clarify things. I mean, I will just rephrase my question as: So, is the question 'Can a model contain its own interpretation?' basically the same as "Can a word contain its own interpretation***"?
DeleteIf so, then your point is trivially true.
Or are you saying anything over and above this in the case of models and their interpretations?
if so, what is it?
***the Wittgensteinian answer to this latter question-within-a-question is of course: no, it cannot. Use is what breathes life and meaning into a word
Ok, I am just making the trivially true point, but i am making it because some people seem to miss it.
DeleteNot sure that that's true, at least not in the sense that this discussion would demonstrate that or anything.
DeleteThe thing is, that last paragraph in your post was simply different from and unnecessary in relation to the paragraphs before it (which did make a non-trivial and more specific point).
I'm not saying Murphy or Rowe would think that! But we do have one lurker at this site who might.
DeleteBob said:
ReplyDelete"isn't it odd that I am running around saying I made mistakes in my textbook?"
To which Gene replied, "You misinterpreted the meaning of the model!"
Right! I interpreted the model (of no crowding out, etc.) to mean that government debt couldn't burden future generations on net. Then Nick Rowe's model showed that was wrong. Woo hoo!
(And yes, I was being sloppy in my statements above. But Krugman would indeed vomit in his mouth if he read your statement here, in the context of you disagreeing with him. E.g. if he showed more deficit spending would restore full employment in his model, and then you said, "Ah, we interpret his model in different ways. What *I* see is that restoration of Aggregate Demand is key. In principle, Warren Buffet and Microsoft could invest another $30 billion and get the same effect. So this isn't 'government stimulus' in action at all." ... then Krugman would indeed engage in projectile vomiting in your general direction.)
Gene, I think you will like my latest comment as it bolsters you quite directly.
ReplyDeletehttp://consultingbyrpm.com/blog/2012/10/my-final-word-this-generation-the-debt-really-is-fundamental.html#comment-47904
Wait a minute Gene, for some reason it seems you believe George Selgin and Larry White make a great case for Fractional Reserve Banking when they cling thoroughly to Keynesian mechanics, and specifically the velocity of money model. Are you contradicting yourself here? I mean you're he one with a PhD
ReplyDelete"Wait a minute Gene, for some reason it seems you believe George Selgin and Larry White make a great case for Fractional Reserve Banking when they cling thoroughly to Keynesian mechanics..."
ReplyDeleteI imagine you are just a troll, but let's see!
A) Where did I say that?
B) What are "Keynesian mechanics"?
C) In what sense are Selgin and White Keynesians? (Please, please don't say they feel money can be in short supply! That view goes back at LEAST to JS Mill.)
D) "Models do not prove anything" is not equivalent to "models are useless."
A) It seems here in your own blog (http://gene-callahan.blogspot.com/2012/07/stunning-lack-of-historical-curiosity.html), because you met Selgin and love the guy for his interesting historical information, you are praising him as the all-knowing man of how free banking works, even if Rothbard came out with some very factual evidence for his historical claims. Let us remind ourselves that both men can have evidence to support their argument, but perhaps this may not mean either man is more incorrect. I remember many Mises Institute scholars claiming Rothbard was the smartest man he ever knew, because he read (and wrote) so many darn books. This claim is like your blog praising Selgin I linked to above. Here is Mises: "Now, a real reproduction of the past would require a duplication not humanly possible. History is not an intellectual reproduction, but a condensed representation of the past in conceptual terms. The historian does not simply let the events speak for themselves. He arranges them from the aspect of the ideas underlying the formation of the general notions he uses in their presentation. He does not report facts as they happened, but only relevant facts. He does not approach [p. 48] the documents without presuppositions, but equipped with the whole apparatus of his age's scientific knowledge, that is, with all the teachings of contemporary logic, mathematics, praxeology, and natural science."
ReplyDeleteI'm sure it's safe to say Rothbard has done an excellent job in cobbling facts about economic history, even if you favor Selgin for your friendship with him.
B) It's safe to say that Keynesian mechanics are everything that attempts to calculate economic activity and not explain it. So perhaps everything taught at universities all across the world. Because all universities teach one variation or another of economics that includes GDP in it. Is this not Keynesian Mr Callahan?
C) And if after my last statement you don't see how Selgin and White are not Keynesians, then you are missing a lot of your own preachings in this very blog post to be exact. Models indeed never prove anything in the real world. Yet Selgin loves the velocity of money, and claims he uses Keynes' mechanics: http://hnn.us/node/141060
Here's Mises: "The main deficiency of the velocity of circulation concept is that it does not start from the actions of individuals but looks at the problem from the angle of the whole economic system. This concept in itself is a vicious mode of approaching the problem of prices and purchasing power. It is assumed that, other things being equal, prices must change in proportion to the changes occurring in the total supply of money available. This is not true."
and Hazlitt: http://mises.org/daily/2916
GDP is Keynesian, that should be enough to prove that ballyhoo completely incorrect. More importantly, read the title of your own blog post we are currently commenting back and for on here....or are you being ironic and witty?
D) In your statement "models do not prove anything," you are basically making a claim that they can exist insofar as the person using them finds them useful to press his own initiatives. Thus this is the only use found in them. In essence, "models are useless" is indeed tantamount to "models do not prove anything." Models do not prove anything in the real world, they are thus useless to everyone but the biased creator of them.
"It seems here in your own blog (http://gene-callahan.blogspot.com/2012/07/stunning-lack-of-historical-curiosity.html), because you met Selgin and love the guy for his interesting historical information, you are praising him as the all-knowing man of how free banking works..."
DeleteOh God. "All-knowing man"?! No, I think George is a very bright and studious guy. I was criticizing those who had not studied these historical episodes at all and yet saw fit to pontificate upon them.
"I'm sure it's safe to say Rothbard has done an excellent job in cobbling facts about economic history,"
See Yeager's critique of Rothbard's "historical" method.
" even if you favor Selgin for your friendship with him."
I have had a beer with George one night. That hardly makes me intellectually subservient to him!
"It's safe to say that Keynesian mechanics are everything that attempts to calculate economic activity and not explain it."
But Keynes was attempting to explain it, not "calculate" it! There are many problems with GDP calculation, recognized by many, many economists, many of them not Austrian. And many non-Keynesians use GDP.
"GDP is Keynesian, that should be enough to prove that ballyhoo completely incorrect."
Here you are assuming your conclusion.
"In your statement "models do not prove anything," you are basically making a claim that they can exist insofar as the person using them finds them useful to press his own initiatives."
No, I am not. Recall, in the context of the post in question, that I was disputing *Murphy's* use of a model. Have you posted this view of models at Bob's blog?
"Models do not prove anything in the real world, they are thus useless to everyone but the biased creator of them."
Yeah, so Kepler's model of the solar system was useless to everyone but the biased Kepler, hey?
You cite Garrison. He writes: "In the General Theory, impenetrable uncertainty about the future clouded the decision processes of investors and wealth holders"
DeleteSo much for your notion that Keynesianism relies on certainty about the economic future!
Great point Gene, but more importantly one can see that in this method of thinking is where Keynes set about to create false mechanics by which to stir the market in its appropriate direction, to stimulate the economy through government intervention in Keynesian terms. So, for example, he could not understand why people would invest resources in risky adventures that helped keep the economy growing at full employment. He therefore substituted "animal spirits" for the profit motive. These spirits allow entrepreneurs to proceed with a naïve confidence and to set aside concerns over losses. Also, the failure to invest was also a psychological problem that he dubbed the "liquidity trap." This trap occurs when investors seek liquidity in cash and when monetary policy — in terms of cutting interest rates — no longer produces an increase in investment.
DeleteThe problem with Keynes is that he thought that if entrepreneurs lose their collective nerve, the government should socialize investment, prop up demand and employment, and provide assurances to drive the economy back to full employment. Keynes' greatest failures were to attempt to define the uncertainty of the market, to create unrealistic mathematical models by which to snub the uncertainty and steer towards certainty. This is Garrison's point.
Thus I hope you see how Selgin and White are both falling into this trap of attempting to steer an uncertain world towards certainty. I have seen Horwitz use the beloved IS-LM model, something many mainstream economists use. This stuff is so far fetched it makes me think people like the eclectic economists Sanchez writes of simply admire those Keynesian mechanics because they teach them at the university and their pay check relies on them. Selgin wrote a book on praxeology and still uses these statist models, that's why Jeffrey Hummel questioned him in that link I posted above. It's quite ironic and makes them less credible. Rothbard and even Robert Murphy are approaching economics in the most appropriate way, with ordinal value scales, free from this attempt to use unrealistic models and mathematical jargon. Mises had all the reason to state that Rothbard laid Keynesianism to rest, MES is a genius book. Uncertainty cannot be calculated nor modeled, the market can only be described. This is the only way economics can stay value-free.
"So, for example, he could not understand why people would invest resources in risky adventures that helped keep the economy growing at full employment."
Deleteadringuti, you have no idea what you are talking about.
"Thus I hope you see how Selgin and White are both falling into this trap of attempting to steer an uncertain world towards certainty."
Repeat the comment above.
adringuti, I can't deal with this volume of ignorance. Further comments by you won't be posted.
Finally a disclaimer: I would hope that after noticing the complete statist approach Selgin and White use in attempting to prove that FRB is viable and appropriate in a free society (not fraud, and does not lead to ratcheting of state power), you come to terms with the fact that they are simply running in circles like Milton Friedman did with their arguments against other economists. But of course Friedman used those unrealistic models of GDP, velocity of money, phillips curve and so forth, and looked so much a fool in his approach, that the Austrians found conspicuous his Keynesian love all along:
ReplyDeletehttp://www.auburn.edu/~garriro/fm2friedman.htm
Let us take time to thoroughly appreciate the Misesian descriptive approach to economics, and remember that value scales are only ordinal. Here is perhaps where other economists like Selgin and White fumble in their arguments with other statists. They fall into the trap of using those statist mechanics, and so run in circles like poor Mr Milton Friedman. Perhaps Robert Murphy and Rothbard can help you through this time of difficulty.
As far as I know, White and Selgin are anarchists. And, as far as I can see, you are a nutjob.
DeleteEconomic models prove nothing Gene, let's be fair if you want to break down semantics in your argument.
ReplyDeleteFurthermore, is someone still an Austrian if they use models and mechanics that deviate from the descriptive methods of the Austrians? Seems contradictory to me. Mises makes it quite clear that it makes no sense to be normative in economics, thus at that point they might just be posers or something. Like Selgin...
http://bastiat.mises.org/2012/09/distinctly-austrian-vs-eclectic-economics/
Yeager's critique is a good one? I can find tons of Rothbardians that could challenge that biased view very easily. Yeager stands minuscule in his efforts as compared to Rothbard.
"Furthermore, is someone still an Austrian if they use models and mechanics that deviate from the descriptive methods of the Austrians?"
DeleteWho cares who is "still an Austrian"? Only cultists.