The straightjacket of formal methods
(1 voti, media: 4 su 5)
Loading ...March 14th, 2007 by Admin
1,777 Views - 
Between the Austrian School and contemporary orthodox economic thought (often called, albeit improperly, "Neoclassical") there are noticeable differences, both from the theoretical and the methodological points of view (e.g., “economics and meta-economics”). One may be tempted to dismiss methodological disputes as pointless metaphysical quibbling, but, in my opinion, these divergent philosophical viewpoints can explain many differences in economic theory.
One of the topics on which methodological differences are evident is the role given to formalization in economic theorizing. Austrians claim that formal methods are not adequate in economic theorizing, while Neoclassical theorists deem axiomatization a necessary step to obtain a proper scientific theory.
To be sure, one shall not equate formalization and quantification: a formalized theory is a theory based on axioms, and whose theorems are derived from these axioms by the use of formal logic. Hence, it has nothing to do with the debite on the measurability of economic entities, such as individual preferences.
Where are creativity, entrepreneurship, subjective information and valuations, intelligence, disequilibrium, production time, capital structure? Is it scientifically meaningful to discard relevant economic topics only because they can’t be forced to fit methodological (i.e. metaphysical) principles?
Economists had to choose between two alternatives.
They could:
- Avoid formalization, and keep on studying relevant economic problems, such as the structure of production and subjective valuations, waiting for the eventual development of adequate formal methods to include these topics, whose relevance in real life is beyond doubt;
- Avoid realistic economic theory, by blindly following the formalization diktat, eliminating whatever economic problems resisted formal treatment: that is, every problem concerning the individual, his knowledge, his intelligence.
They chose the second path, and one of the results is that the understanding of the economic theories prior to Keynes and Friedman (e.g., prior to the overwhelming success of macroeconomics and methodological positivism) has been severely hampered (that’s the only reason I can imagine why an easy book such as Hayek’s "Prices and Production" is considered cumbersome and obscure by most economists, such as DeLong and Blanchard). Besides, the theories of economic calculation, subjective information, capital structure and long-term disequilibrium are now forgotten and esoteric themes.
Anyway, there are two topics on which the Austrian aversion against mathematics is unwarranted and overstated:
-
Ceteris paribus (i.e., if formal methods were adequate) a formal treatment is preferable to an informal (i.e., verbal) one: the formal method is based on calculus (in a wide sense of the word), hypotheses are always explicit, logical passages can be mechanically tested, powerful abstract mental tools can be employed;
-
Entrepreneurs and historians do use and do have a need for quantitative methods, such as economic indicators, statistical tools, numerical models… even Mises made this clear in his collection of essays on the theory of the trade cycle ("On the Manipulation of Money and Credit"); and his theoretical investigations on the theory of economic calculation prove the economic relevance of these techniques, even from the Austrian point of view: no algebra, no accounting; no accounting, no markets; no markets, no economics.
L.Baggiani Says
I do thank you for trying to re-equilibrate Austrian and non-Austrian positions on the use of math-logic, and the righteous distinction between “formalization” and “quantification”.
I’d like to stress that “caeteris paribus” clause’s rationale stems from the very reason of stylising reality for formalization porpouses: analysing a single “branch” of reality in a Marshallian-fashion, i.e. admittedly avoiding external feed-backs, which is surely unrealistic but is just one in a million step to (qualitatively) analysing the whole economy.
It is not a matter of limitedness of the tool you use, it is a matter of consciousness of its limits.
E se ‘r mi’ ‘ngrese fa onco, sempre meglio ‘he parlà ‘n dialetto. Grandesuperpietro!
Mar 14th, 2007 at 4:41 pm
libertyfirst Says
The coeteris paribus clause surely performs a fundamental task in economic theorizing: it would be impossible to discern the intricated web of interrelations among individual actors without it.
When the clause is employed, the result is surely “approximated”, but if relevant factors have been neglected, it is relatively straightforward to reduce the level of aggregation and/or increase the number of causal factors (as far as there is some underlying understanding of these factors at a disaggregated level,
).
Formalization is different: it builds the entire edifice of economic theory in such a way that what cannot be formalized is eliminated.
For example, let’s consider:
1. a production theory based on input/output vectors: how can time be included in the analysis, if the formal theory has a timeless productive sector by construction?
2. an economic theory based on equilibrium: how can errors and sunk costs be included in such a framework?
One could imagine a formalization process that does not define away these economic topics: but this formalization, actually, does not exist.
Actually, formalization is a language that does not enable the speaker to conceive several difficult topics: economic theorizing made in this way is as absurd as a treatise written in the language of a four-years old child.
The outcome of the formalization diktat, as reported in the article, is not that the result is approximated, but that relevant economic themes become literally inconceivable: this is not the same as being imprecise. It means being blind to several topics because the underlying theory cannot include them.
If Brad DeLong thinks that Hayek’s “Prices and Production” is cumbersome, and I think it’s an elementary book (as it was intended by Hayek), the reason cannot be (only, at least) that I’m a genius.
Mar 14th, 2007 at 5:26 pm
L.Baggiani Says
I do not agree: first of all in case a theory would miss relevant factors it is simply wrong whether formalized or not.
Coeteris paribus means “let us suppose everything else stays the same”, I repeat: no feed-back. It is helpful in a formalized framework, but it works in any context also.
As to formalization, which it seems to me as what you were discussing about, I can tell you that you should consider what you’re working for: in case you consider the effect of prices onto consumption you can stylize the facts and analyse them (conscious you are not explaining everything, but just a part of reality, that one you are concerned of) In case you consider the effect of culture onto consumption you hardly can “formalize” that. It is the pourpose your analysis which tells you whether any formalization is allowed. It’s you who must discern what instrument to use with regard to what you want to discover.
In case you are trying find a theory to explain everything at once, like Marx did, that’s another business.
But it is anyway unfair to state that every economist eliminates from his mind what he cannot formalize. Someone is slave to diktat, someone is not. Someone is slave to considering anyone else stupid.
Mar 14th, 2007 at 7:57 pm
libertyfirst Says
I agree that economists do consider other factors besides the ones which are formalized in their theories. What I state is that the theory they teach and they consider as the summa of knowledge in economic theory is just a caricature of economic reality.
In order to improve realism, they shall dismiss many of their own theoretical assumptions: long term equilibrium, continuity and infinite divisility… they will probably also have to take into account factor complementarity, production time and whatever else they shall add to their views in order not to fall prey to the formalization diktat.
Good economists surely do it: they are not stupid (sometimes I really have doubt on it: when I hear Blanchard saying “Increase demand! Whatever you do is good: anything which affects demand is equivalent”, I do consider this bad economics, but it is fortunately not the rule).
The problem is that noone has ever been taught, in the university, to take these factors into account (otherwise, they would have learnt Austrian Economics).
At the end, the way Neoclassical and Austrian economists reason on real cases is more similar than their respective theories.
But when it comes to decide who’s the economist, who shall teach, who shall win the Nobel prize, what students shall learn… methodological prejudices keep on being alive…
Is it more meaningful to teach the market process in an Austrian manner, or teach formal models that can be obtained only by sacrifice of relevant economic aspects of reality?
Austrian economics is the general theory of the market process; under particular irrealistic assumptions, formal models can describe several aspects of this process. Unfortunately, because of methodological prejudices, the general theory is dismissed and neglected, and the special theory is considered the only real economic point of view…
Isnt’ it funny?
Mar 15th, 2007 at 10:00 am
libertyfirst Says
Many economists, I add, claim that unrealistic assumptions are such as the physical hypothesis of frictionless motion, which is not realistic in the atmosphere.
The similarity is apparent: behind mechanics there is a formal theory, which is valid both in frictionless motion and in a viscous fluid; behind formal economics there is not a general economic theory, but a particular set of special assumptions. The general economic theory has resisted formalization, so far.
P.S. Also this example is taken from Blanchard… he shall have “robbed” me the girlfriend…
Mar 15th, 2007 at 10:04 am
libertyfirst Says
Ops.. in english “apparent” is the opposite of the italian “apparente”… I meant “deceiving”, not “apparent”.
Mar 15th, 2007 at 10:06 am
L.Baggiani Says
I see you keep on avoiding my considerations about the pourpose of each analysis and the usefulness of approximations at a first stage of analysis (the same thing rothbard states while defending stereotypising) and the difference betweeen marshall and marx.
But it is ok, I know how to consider neoclassical analysing and what’s the austrian point of view and discern what I consider useful in either cases.
You explanations are always useful.
Mar 15th, 2007 at 11:39 am
libertyfirst Says
It seems that you consider economic theory to be a set of heterogeneous tools, useful to solve specific problems, and not a systematic body of thought.
Is my interpretation correct? If it is, I almost completely agree with you: the attempts to formalize a common ground have failed, and we only have partial tools. It seems that economics is not much different from analog electronics… an eclectic art and not a systematic science.
Some of the tools of economic theory are taught in the universities (e.g. partial equilibrium analysis) others are not (e.g. the structure of production): is there any reason behind this discrimination? That’s an interesting question!
My opinion is that there are tools which are relevant but are not accepted in mainstream economics because of a methodological bias.
Anyway, your eclectic “cherry-picking” approach of using the partial tools which is best suited in the particular case is exactly the opposite of the goal of sistematic formalization: and this proves my point (if I understood you). In the absence of such a formal general theory, I think that your approach is the best…
I add that this surely wasn’t the desired outcome of the attempts toward formalization: the intended goal of a formal theory is to make the bulk of common economic knowledge esplicit… but if economic theory is considered a set of semi-integrated tools… there is no theory
P.S. Ironically, in my attempts to grasp the intricacies of capital theory, I reached a similar conclusion: capital theory does not exist; there are only sparse properties of capital goods, to be put together in order to understand economic problems. This is also a limit of Austrian economics: coeteris paribus, it is always a good thing to be systematic and formal (by the way, Hayek probably lost his interest in economics after realizing this problem, and became a social philosopher).
Mar 15th, 2007 at 12:33 pm
Jacopo Says
Someone is slave to diktat, someone is not. Someone is slave to considering anyone else stupid.
Il realismo di Baggiani è da stimare. Il fatto che consideri l’economia come un insieme di strumenti eterogenei dipende dal fatto che evidentemente lui con l’economia ci lavora, non la studi soltanto come noi.
Ps. per piacere, basta con l’inglese. Oltre ad essere anche un po’ ridicolo è anche controproducente per i lettori che non lo sanno. (a meno che non ci leggano anche dagli states…
)
Mar 15th, 2007 at 1:20 pm
L.Baggiani Says
I think we are mid-path in reciprocal understanding.
In my opinion explaining the whole economy is the righteous goal of any system of thought, but because of the intricate web of relations and feed-backs and hard-to-be-formalized variables (one I like is “fear”) the best we can aim to in this moment is a “modular” system of thought, with a framework for the whole economy and particular and far more precise theories for each branch (for instance: financial markets, capital structure, consumption setting and so on) to enter deeply into the problem and easier to be revised.
A mere sum of detatched tools, in my opinion, may lead to inconsistent results, hence a way to glance to the whole is necessary (whether mainly formalized like MV=PQ or more intuitively like in AE). Each branch can be then formalized or quantitativized or not depending on its features (with no “a priori” diktat). Who will be right? the one who will show the greater power of explanation, simply.
That’s all as to what you seem to be concerned of to me.
A point I think I am not clear yet: you may be interested in analysing a particular problem, because you are curious or an organisation asked you to; in this case you could find the case of focusing on a narrow range of variables coeteris paribus, i.e. assuming other variables (formalizable or not) don’t vary creating loops or feed-backs. Certes the goodness of your analysis heavily depends on the relevancy of what you are excluding, but e.g. if you want to test that car demand tends (I stress: tends) to decrease against car prices, then considering a whole range of variables like interest rates, tastes, education, carmakers’ capital structure, may simply not be this useful. But in this case your work can be like the smallest of atoms in a wider theory, not necessarily a theory itself, even less than one in a million tools.
Why some “tools” are not taught? In my opinion this has little to do with formalization ability (AE is not an economic alien, and some parts could be formalized in my opinion) a more with AE’s anti-State stance.
For instance: I don’t think formalizing capital’s time structure is impossible, as intertemporal consumption settin is so, so why Hayek business cycle theory has not been taught to me? Simple: it concludes that central banks should not exist and no teacher is willing to fight against a central bank.
I am sorry, I have written too much.
Mar 15th, 2007 at 1:57 pm
libertyfirst Says
E io che volevo invitare qualcuno dagli States… tanto alla fine la prima versione dell’articolo è comparsa sull’Austrian Forum…
P.S. Il fatto che Baggiani consideri l’economia un insieme di strumenti è la mia interpretazione della sua dicotomia Marshall/Marx, non è detto che abbia compreso.
Mar 15th, 2007 at 2:07 pm
libertyfirst Says
P.S. How could non-english speaking visitors be interested in reading the comments of an article they cannot read?
Mar 15th, 2007 at 2:08 pm
L.Baggiani Says
Grandeliberty!
E poi via, sarebbe pure l’ora che un americano leggesse qualcosa di serio!
Mar 15th, 2007 at 2:19 pm
L.Baggiani Says
Jacopo è sempre molto gentile.
Non è che io lavori con l’economia (nel senso di venir per questo pagato, come voi credo); sono un appassionato, semplicemente.
Ho incrociato diverse teorie, e questo mi ha fatto pensare che la “verità” non sia ancora stata rivelata a nessuno, e che tutti (più o meno) abbiano qualcosa che può esser utile.
L’ostracismo e l’iconoclastismo non mi sembrano utilissimi. One measure does not fit all.
Mar 15th, 2007 at 2:29 pm
libertyfirst Says
The idea that AE is not taught because…
Ops… volevo dire: l’idea che l’AE non sia presa in considerazione perchè ciò sarebbe contrario agli interessi della classe politica è interessante e fondata.
Modelli formali della teoria del capitale, che io sappia, però, non esistono. Per capire il motivo dovrei comprare un libro di 1700 pagine che costa 700$ e si chiama “Capital Theory” (Bliss, Cohen, Harcourt)… ma non è che le mie tasche siano molto d’accordo.
Esiste, sembra, un modello limitato di capitale specifico, che fu proposto da un economista per studiare l’edilizia non-residenziale, sarebbe fico avere il PDF (se qualcuno lo trova, me lo spedisca!):
Bischoff, Charles W. 1970. “A Model of Nonresidential Construction in the United States.”
American Economic Review
Di fatto, però, tutta la controversia di Cambridge verteva proprio sulle difficoltà di tener conto della struttura del capitale… e la risposta di Samuelson a Sraffa fu “Hai ragione tu, ma continuiamo sulla nostra strada”…
Mar 15th, 2007 at 3:51 pm
jacopo Says
l’idea che l’AE non sia presa in considerazione perchè ciò sarebbe contrario agli interessi della classe politica è interessante e fondata.
Secondo me è tutt’altro che fondata. Non amo i complottismi. Gli accademici, specie quelli statunitensi, non sono certo finanziati dalle varie amministrazioni (anche perché sennò dovrebbero cambiare idee ogni 4 anni), e non pendono certo dalla loro bocca. Prendete questa frase in senso lato, è chiaro che un minimo di commistione ci deve essere: in fondo l’economia è politica. Ma affermare che esiste una correlazione diretta è fuori luogo. Affermare di essere scomodi mi sembra più che altro nascondersi dietro ad un dito.
E’ che l’uomo secondo me è socialista per definizione…
:D
(’sta qua è grossa..)
Mar 15th, 2007 at 6:02 pm
jacopo Says
A prosito di scambi di paper.. qualcuno di voi ha per caso l’articolo di Samuelson “Proof that Properly Anticipated Prices Fluctuate Randomly ” 1965?
ps. sarebbe bello estendere i dibattiti a più interlocutori.. individuate delle soluzioni a proposito?
Mar 15th, 2007 at 6:05 pm
libertyfirst Says
BTW, ho trovato la prova che Friedman era comunista: guardate che sito pubblica il suo “Meth of Pos. Econ.”.
Mar 15th, 2007 at 6:05 pm
libertyfirst Says
A proposito di cosa?
Mar 15th, 2007 at 6:38 pm
jacopo Says
Sul come introdurre nuovi interlocutori…
Mar 15th, 2007 at 7:03 pm
L.Baggiani Says
il mio Milton al tempo era ancora piccolo…
Mar 15th, 2007 at 10:19 pm
Fox Says
Ma permettere a chi conosce solo la lingua italiana di capire cosa c’è scritto, no?
Mar 16th, 2007 at 2:16 am
L.Baggiani Says
Liberty, ritraduciamo tutto per fox? magari ne facciamo qualcosa da pubblicare su IHC?
Mar 16th, 2007 at 9:01 am
jacopo Says
It’s up to you… se traduci e rielabori io pubblico volentieri..
Mar 16th, 2007 at 9:39 am
L.Baggiani Says
Ok, mi prendo la briga se Liberty mi dà l’OK (o se intanto vuol tradurmi il suo)
Mar 16th, 2007 at 10:05 am
libertyfirst Says
Io ho recuperato l’uso della mano, ma non il diritto di proprietà sul tempo libero (sto progettando un circuito integrato da zero e da solo :O )… comunque traduco al più presto… diciamo per lunedi mattina…
Mar 16th, 2007 at 10:47 am
libertyfirst Says
Quando ho letto “traduciamo tutto per fox”, ho pensato al network televisivo… pensavo di stare per diventare famoso..
Mar 16th, 2007 at 12:48 pm
L.Baggiani Says
cala liberty cala…
Mar 16th, 2007 at 12:55 pm
Fox Says
Grazie Ragazzi! Se ciò non comporta problemi di rilievo ve ne sarei estremamente grato visto che non mi ci metto nemmeno a provare a capirci qualcosa data la mia scarsa conoscenza della lingua inglese.. per correttezza aggiungo che non sono un esperto della materia e che se l’autore non ha voglia di perdere tempo solo per un curioso come me lo capisco.Comunque trovo questo sito molto interessante, nonostante magari qualche divergenza di opinione(sacrosanta), e avrei piacere nel continuare a confrontarmi con voi (ribadisco che raramente interverrò data la mia inesperienza) Grazie ragazzzi..
Mar 16th, 2007 at 7:23 pm
jacopo Says
Welcome aboard…
Mar 16th, 2007 at 7:50 pm
Libertyfirst Says
Troppo tardi: già tradotto. A domani!
Mar 16th, 2007 at 8:19 pm
L.Baggiani Says
Jacopo, ti proporrei ti creare una specie di casella dove persone curiose (tipo fox) possano inserire domande specifiche, chiarimenti, o proporre argomenti per un articolo; oltre che essere d’ispirazione per le sinapsi, mi pare un servizio interessante per attirare gente.
Mar 16th, 2007 at 10:10 pm
CyNyC Says
mi aggiungo alla richiesta di traduzione commenti di Fox (anche io fatico un po’ con l’inglese) e concordo con la proposta di L. Baggiani.
Sono un curioso non-esperto della materia che vi legge più che volentieri per cercare di approfondire le mie (scarse, per ora) conoscenze.
Approfitto per ringraziare autori di articoli e commenti.
Mar 17th, 2007 at 11:43 am