[On Distinctions and Provisional Convictions in Economics] - A New Kind of Science: The NKS ForumA New Kind of Science: The NKS Forum
Pages:1
On Distinctions and Provisional Convictions in Economics
(Click here to view the original thread with full colors/images)
Posted by: Fiona Maclachlan
The stimulating exchange begun with Gunnar Thomasson's post of October 3rd has prompted the following thoughts. In particular, I was intrigued by the remarks posted by Jason Cawley on October 10th on the question of method in the social sciences and on the need to make the right distinctions.
First, as one who has spent a long time puzzling over questions of methodology in economics, I agree that NKS does have something "essentially new" to say with its distinction between computationally reducible and irreducible systems, and I'm excited about the prospect of witnessing the debate in economics moving forward in the right direction as NKS ideas gain acceptance.
I’d argue, however, that the relevant debate in economics is not between theory and history, or between theory and application. Rather the issue is of the type of theory that is appropriate. Von Mises, for instance, takes the side of the theoretician against the Prussian Historical School, but he was also against those who wished to import methods directly from the physical sciences.
Where I would draw the distinction in economics is between the modern view that economic theory must take the form of an axiomatic-deductive model that generates falsifiable predictions, and the older approach exemplified in the work of most leading economists from Adam Smith through J.M. Keynes.
In the brief history of models of markets in the notes to the NKS book (1015R) this older view is labeled common sense, which is probably as good a term as any.
I recall Stephen Wolfram being quoted in the press saying that he runs his business using “common sense.” Certainly, there’s an overlap between arriving at business decisions and traditional economic analysis-- the determination of the profit maximizing price of Mathematica 5, for example, is a case of both.
The common sense approach does entail an element of rough prediction, so it extends beyond historical description. It can also make use of formal models as a check on the logical consistency of one’s conclusions. The decision to charge existing users of Mathematica less for an upgrade than what is paid by a new user may seem intuitively obvious, but the economist’s toolkit has the formal model of price discrimination to show exactly under what circumstances it is, in fact, the profit maximizing strategy. In the traditional approach formal models are used as tools to leverage and refine one’s economic intuition and judgment.
In the modern approach, on the other hand, the formal model is all there is to theory.
A typical case for the modern view can be found in the introductory chapter of Von Neumann and Morgenstern's Theory of Games and Economic Behavior (3rd ed. 1953 7-8). They argue that economists must start with the “very simplest facts of economic life and try to establish theories which explain them and which really conform to rigorous scientific standards.” They continue:
This preliminary stage is necessarily heuristic, i.e. the phase of transition from unmathematical plausibility considerations to the formal procedure of mathematics. The theory finally obtained must be mathematically rigorous and conceptually general. … Beyond this lies the field of real success: genuine prediction by theory. It is well known that all mathematized sciences have gone through these successive phases.
Against this view consider Keynes’ remarks on methodology contained in his correspondence with Roy Harrod regarding Harrod’s paper “Scope and Method in Economics”:
It seems to me that economics is a branch of logic, a way of thinking; and that you do not repel sufficiently firmly attempts … to turn it into a pseudo-natural-science. … I also want to emphasize strongly the point about economics being a moral science. … It is as though the fall of the apple to the ground depended on the apple’s motives, on whether it is worth while falling to the ground, and whether the ground wanted the apple to fall, and on mistaken calculations on the part of the apple as to how far it was from the centre of the earth. (Keynes, Collected Writings, 14:296-300)
Also revealing is a footnote in Keynes’ biographical essay on Alfred Marshall:
Professor Planck, of Berlin, the famous originator of the Quantum Theory, once remarked to me that in early life he had thought of studying economics, but had found it too difficult! Professor Planck could easily master the whole corpus of mathematical economics in a few days. He did not mean that! But the amalgam of logic and intuition and the wide knowledge of facts, most of which are not precise, which is required for economic interpretation in its highest form is, quite truly, overwhelmingly difficult for those whose gift mainly consists in the power to imagine and pursue to their furthest points the implications and prior conditions of comparatively simple facts which are known with a high degree of precision.(Keynes, Essays in Biography 1951 158n)
I would argue that the difference between Keynes and those who espouse the modern view is that Keynes had an intuitive sense of the computational irreducibility of economic phenomena and of the problems it poses for economic analysis. His comments about the apple, especially, suggest the recognition of an inherent complexity of human decision that renders the traditional methods of the physical sciences ineffective.
Many observers of contemporary economics share Keynes’ intuitive sense and have raised protests from time to time. (See www.paecon.net for an example of a well-organized protest launched by students in France in 2000.)
In the past, the response to the protests has been a plea for patience. The assumption has been that if economists soldier on along the path of mathematically rigorous theory, eventually they will reach the Promised Land imagined by Von Neumann and Morgenstern, that is, “the field of real success: genuine prediction by theory.” The weight of authority of those issuing the pleas for patience has, in the past, successfully marginalized the critics.
The role of NKS in all this is its potential to cause a significant shift in the debate. If one accepts that the fundamental unit of analysis in economics is human decision and if one accepts that human decision is computationally irreducible, then it follows from NKS that the modern approach of formulating axiomatic-deductive models to generate falsifiable predictions is incorrect. Honest observers of contemporary economics have reason to move from a mere skepticism about its direction to a provisional conviction that it is indeed on the wrong track
Or, to switch metaphors, NKS identifies which branches of inquiry will never reach fruition. It provides the basis for a provisional conviction of something that has long been only a suspicion: that the reason for the lack of progress within in the discipline is that for the past fifty years, economists have been pruning off the wrong branches!
Posted by: Jason Cawley
I think the issue of practice and the difference between it and theory is fundamental in economics. The reason is that economies are themselves practical systems. They deal with problems of prediction every day. They do not deal with them quite as physics does or as positivist manifestos once promised, no doubt. But they do deal with them, because they must, or everything falls apart.
CS Peirce once defined belief as the willingness to stake much on a proposition. Economic practice is necessarily shot through with belief. Theoretical doubt is not good enough. You can't bet a trillion dollars - or the happiness of a hundred million people - on a theoretical doubt. If economic practice does not get axiomatic systems it will get something else, but predict it must.
If you look at the difficulties Keynes lists in your quote, every one of them confronts the economic actor at least as much as the economist. I also note in passing that computational irreducibility is not one of the factors he listed. That is, there have been various past understandings of why economics was different, in the degree of prediction it allowed. They were right that it is different, but it is at least possible the reason for the difference is not one they suspected before.
Before getting to all of that, though, I am going to complicate the picture by pointing out some mitigating circumstances in the history of economics as a discipline. First there is the fact that the period of its common sense also gave rise to numerous separate schools, some of them based frankly on refuted notions of cranks, and erected into whole ideologies. The iron law of wages, the exploitation theory, free banking etc. The common sense period of economics had its share of common nonsense.
Practice came in from another direction, in other words, and not a salutary one. Economists had cause to hype scientific rigor. There were quacks abroad claiming their opinions were as good as anyone else's, that economics didn't know any different, that if it did not know everything it knew nothing, that claimed to be scientific themselves, etc. Certainly there are other forms of rigor, in some German scholarship for example. But in the old common sense style, men as learned as Keynes and Hayek could sometimes talk right past each other.
I could go into the partial successes of calculus based methods in economics but it is probably clear enough. Instead I will point out the degree to which economic subsystems often are simple in their behavior and broadly predictable as a result. If you look at a quarterly GDP graph for a modest time period and don't deliberately pick the most unusual cases, complexity is not the leading characteristic you will see. A line is often a very good approximation. With a small noise term perhaps. Not, e.g. order of magnitude changes on a time scale of days or anything remotely like it.
And there are lots of economic systems of that character. There have to be, companies and entrepeneurs could not invest successfully if they couldn't count on any stability in relative input and output prices, for example. What is characteristic of economic systems is that this predictability is always partial, with noise, only for most cases or periods, and with exceptions (like equity markets) where aggregate risk is "parked".
The previous common explanations for the pockets of unpredictability seen include - self reference paradoxes, expectations, subjective preferences, and my personal favorite, "imperfect information". Economies process information and iteratively search for useful combinations by trial and error on a massive scale. It is kind of obvious that imperfections of information, and correcting them however crudely, are nearly the whole point. What is worth noticing here is simply that economies have a complicated set of problems to solve. Hayek stressed this point I believe.
You propose that the critical distinction is -
between the modern view that economic theory must take the form of an axiomatic-deductive model that generates falsifiable predictions, and the older approach exemplified in the work of most leading economists from Adam Smith through J.M. Keynes.
To the extent that the modern view is based on expecting all systems to readily yield to deduction, there is something to that. It is worth pausing to ask why that view expects so much from deduction. It was believed that reducing a problem to logical relations was tantamount to solving it. Implicitly it was assumed that all logic problems - or even all math problems - are easy. Now this simply is not the case.
You might have a formal model of something and no slippage (by hypothesis - stipulate it) between the real system and your axiomatic model. And nevertheless find yourself unable to predict the system. What if you can't predict what your formal model will do, other than by just running it and watching? You can experiment with the model certainly. But prediction is a harder requirement than "able to model".
To predict, the system itself needs another characteristic - reducibility, or some sort of simplicity. You may of course find some things you can predict without being able to get all the details. Nowhere is it written that economies shall be predictable (in detail), even if we did succeed in capturing them with axiomatic models. The details of the system's history may still matter, in arbitrarily complicated ways. If they do, you just have to go look. But then you are beyond deduction.
So, some economic systems will prove simple enough to be reducible. Simple program explanations may model some of these better than calculus methods or statistical models. Others may prove intractable. They might be modeled statistically to some extent, though beyond some point with those we are just trying to put a number on our ignorance. Which we get either empirically, or by some rule of thumb that we don't really know applies in a given case.
Instead what NKS methods suggest is that we try to make simple program models that reproduce the intractability we see in the data. These will not make the system tractable - the leading characteristic of the real phenomenon does not go away because you model it. It can suggest to us formal experiments to learn about the system. It may explain to us why predicting a given system is hard, what details will resist prediction, what aggregate characteristics might still be predicted, and the like.
I hope this is interesting.
Posted by: Gunnar Tomasson
As follow-up to Fiona Maclachlan’s thoughtful contribution, I thought I should give a bird’s-eye view of the conclusions of my own work over three decades in the field of the epistemology of science in general and of economics in particular insofar as they relate to certain key issues addressed by her.
1. Computational Systems.
First, as one who has spent a long time puzzling over questions of methodology in economics, I agree that NKS does have something "essentially new" to say with its distinction between computationally reducible and irreducible systems, and I'm excited about the prospect of witnessing the debate in economics moving forward in the right direction as NKS ideas gain acceptance.
Comment:
At first glance – for an NKS-neophyte such as myself – the distinction between “computationally reducible and irreducible systems” appears analogous to that between “closed” and “open” systems as exemplified in theoretical economics by Walrasian General Equilibrium Systems (closed) and non-Walrasian Systems (open).
In theoretical physics, the Laplacian construction of Newtonian Mechanics as applied to the Universe exemplifies a closed system – if a single photon could be added to or subtracted from its energy-mass, the Universe would exemplify an open system.
2. Analytical Economics and Common Sense
I’d argue, however, that the relevant debate in economics is not between theory and history, or between theory and application. Rather the issue is of the type of theory that is appropriate. Von Mises, for instance, takes the side of the theoretician against the Prussian Historical School, but he was also against those who wished to import methods directly from the physical sciences.
Where I would draw the distinction in economics is between the modern view that economic theory must take the form of an axiomatic-deductive model that generates falsifiable predictions, and the older approach exemplified in the work of most leading economists from Adam Smith through J.M. Keynes.
In the brief history of models of markets in the notes to the NKS book (1015R) this older view is labeled common sense, which is probably as good a term as any.
Comment:
In my view, the seemingly intractable nature of perennial methodological issues in the field is rooted in failure to recognize that Economics comprises both Closed- and Open-System features as reflected in what Bentham termed the Science and Art of Economics, respectively – what I call its Analytical and Common Sense aspects.
The most egregious example of such failure remains the construction of the Closed-System Say’s Law of Markets as an Open-System proposition, whereby a strictly axiomatic or Analytical conclusion is incongruously held to have been ‘refuted’ by Common Sense appeal to the Empirical phenomenon of non-clearing markets.
3. Models and Prediction
I recall Stephen Wolfram being quoted in the press saying that he runs his business using “common sense.” Certainly, there’s an overlap between arriving at business decisions and traditional economic analysis-- the determination of the profit maximizing price of Mathematica 5, for example, is a case of both.
The common sense approach does entail an element of rough prediction, so it extends beyond historical description. It can also make use of formal models as a check on the logical consistency of one’s conclusions. The decision to charge existing users of Mathematica less for an upgrade than what is paid by a new user may seem intuitively obvious, but the economist’s toolkit has the formal model of price discrimination to show exactly under what circumstances it is, in fact, the profit maximizing strategy. In the traditional approach formal models are used as tools to leverage and refine one’s economic intuition and judgment.
In the modern approach, on the other hand, the formal model is all there is to theory.
A typical case for the modern view can be found in the introductory chapter of Von Neumann and Morgenstern's Theory of Games and Economic Behavior (3rd ed. 1953 7-8). They argue that economists must start with the “very simplest facts of economic life and try to establish theories which explain them and which really conform to rigorous scientific standards.” They continue:
This preliminary stage is necessarily heuristic, i.e. the phase of transition from unmathematical plausibility considerations to the formal procedure of mathematics. The theory finally obtained must be mathematically rigorous and conceptually general. … Beyond this lies the field of real success: genuine prediction by theory. It is well known that all mathematized sciences have gone through these successive phases.
Comment:
The von Neumann-Morgenstern thesis is predicated on a fundamental misconception of the relationship in physics between “fact” and “theory”, on the one hand, and “theory” and “prediction”, on the other hand, compounded by failure to distinguish between the Closed- and Open-System aspects of Economics addressed above.
With respect to the latter, it is self-evident that data culled from a computationally irreducible system cannot in principle be transformed into a computationally reducible system for purposes of predicting the parent system’s dynamic behavior.
As for the former, the “facts” of solar system orbital mechanics are a function of “theory” as fashioned by Newton and Einstein, respectively, whereby Closed-System observational data was translated into two distinct mathematical “models”. In turn, successful use thereof for prediction attests to the Closed System’s structural stability irrespective of whatever causal interpretation may have come to be associated with either “model”.*
[*As emphasized by Newton up front in ‘Principia’, the variables of the gravitational equations presented therein for the Earth-Moon System were NOT to be construed as explanatory in any manner, shape, or form with respect to whatever physical factors might account for the observational data which served as input for his “model”.]
4. Economics: A Branch of Logic AND Common Sense
Against this view consider Keynes’ remarks on methodology contained in his correspondence with Roy Harrod regarding Harrod’s paper “Scope and Method in Economics”:
It seems to me that economics is a branch of logic, a way of thinking; and that you do not repel sufficiently firmly attempts … to turn it into a pseudo-natural-science. … I also want to emphasize strongly the point about economics being a moral science. … It is as though the fall of the apple to the ground depended on the apple’s motives, on whether it is worth while falling to the ground, and whether the ground wanted the apple to fall, and on mistaken calculations on the part of the apple as to how far it was from the centre of the earth. (Keynes, Collected Writings, 14:296-300)
Also revealing is a footnote in Keynes’ biographical essay on Alfred Marshall:
Professor Planck, of Berlin, the famous originator of the Quantum Theory, once remarked to me that in early life he had thought of studying economics, but had found it too difficult! Professor Planck could easily master the whole corpus of mathematical economics in a few days. He did not mean that! But the amalgam of logic and intuition and the wide knowledge of facts, most of which are not precise, which is required for economic interpretation in its highest form is, quite truly, overwhelmingly difficult for those whose gift mainly consists in the power to imagine and pursue to their furthest points the implications and prior conditions of comparatively simple facts which are known with a high degree of precision.(Keynes, Essays in Biography 1951 158n)
Comment:
“The economist has consoled himself for his barren results with the thought that he was forging tools which would eventually yield fruit,” Paul Samuelson wrote in ‘Foundations of Economic Analysis’ (1942). “The promise is always in the future; we are like highly trained athletes who never run a race, and in consequence grow stale. It is still too early to determine whether the innovations in thought of the last decade [the 1930s – insert] will have stemmed the unmistakable signs of decadence which were clearly present in economic thought prior to 1930.”
Now, some sixty years later, the continued “attempt [by mainstream and monetarist scholars] to turn [economics] into a pseudo-natural-science” remains the single most telling indicator of intellectual decadence within the economics profession – an attempt which has been doomed to failure since its beginning upon the advent of neo-classical economics after mid-19th century for reasons indicated above.
On the basis of my own research during the past thirty years, I am persuaded that a veritable renaissance of Economics is within reach, provided only that young scholars (a) grasp the distinction between computationally reducible and irreducible systems, and (b) draw the appropriate conclusions therefrom insofar as the respective places of Logic and Common Sense in a Future Economics are concerned.
5. A Future Economics
I would argue that the difference between Keynes and those who espouse the modern view is that Keynes had an intuitive sense of the computational irreducibility of economic phenomena and of the problems it poses for economic analysis. His comments about the apple, especially, suggest the recognition of an inherent complexity of human decision that renders the traditional methods of the physical sciences ineffective.
Many observers of contemporary economics share Keynes’ intuitive sense and have raised protests from time to time. (See www.paecon.net for an example of a well-organized protest launched by students in France in 2000.)
In the past, the response to the protests has been a plea for patience. The assumption has been that if economists soldier on along the path of mathematically rigorous theory, eventually they will reach the Promised Land imagined by Von Neumann and Morgenstern, that is, “the field of real success: genuine prediction by
theory.” The weight of authority of those issuing the pleas for patience has, in the past, successfully marginalized the critics.
The role of NKS in all this is its potential to cause a significant shift in the debate. If one accepts that the fundamental unit of analysis in economics is human decision and if one accepts that human decision is computationally irreducible, then it follows from NKS that the modern approach of formulating axiomatic-deductive models to generate falsifiable predictions is incorrect. Honest observers of contemporary economics have reason to move from a mere skepticism about its direction to a provisional conviction that it is indeed on the wrong track
Or, to switch metaphors, NKS identifies which branches of inquiry will never reach fruition. It provides the basis for a provisional conviction of something that has long been only a suspicion: that the reason for the lack of progress within in the discipline is that for the past fifty years, economists have been pruning off the wrong branches!
Comment:
What are the appropriate places of Logic and Common Sense in a Future Economics?
The brief answer, I submit, is that Monetary Economics must be reconstructed from the ground up to reflect the Logic of the monetary aspects of cooperative productive activity viewed as the pooling and conversion of the individual Factor Endowments of A, B, and C into Final Output in which A, B, and C will share as agreed up front and formalized through Credit arrangements whereby A, B, and C all acquire claims to Final Output which are commensurate with their Factor Contributions thereto.*
[* THIS IS SAY'S LAW, although Jean Baptiste Say never put it this way – and, without going into details, the underlying concept of cooperative productive activity is readily shown to accord with the concept of “income” set forth by Keynes in ‘A Treatise on Money’ (Ch. 9):
"Income. - We propose to mean identically the same thing by the three
expressions:
(1) the community's money income;
(2) the earnings of the factors of production; and
(3) the cost of production;
and we reserve the term _profits_ for the difference between the cost of production of the current output and its actual sale-proceeds, so that profits are NOT part of the community's income as thus defined."
Finally, "...now, at greater distance,” John Hicks wrote in 1973, “we find (I believe) that the 'General Theory' loses stature, while the 'Treatise', in spite of its eccentricities, grows. The 'General Theory' is a brilliant squeezing of dynamic economics into static habits of thoughts. The 'Treatise' is more genuinely dynamic, and therefore more human."]
Everything else is Common Sense.
Gunnar
Posted by: alphasun
[If one accepts that the fundamental unit of analysis in economics is human decision and if one accepts that human decision is computationally irreducible [/B]
This is an appealing axiom, but I suspect that, bearing in mind that economics concerns itself with large numbers of humans, there will be an eventual accumulation of discoveries in a number of fields that will improve the predictability aspect. I remember being struck by the imprecision of the curves in economic textbooks my student friends doing economics used to show me, and the unreliability of stock predictions or those of Marxism is notorious, but that does not mean that the field is irreducible in mathematical or theoretical terms.
I am thinking of the empirical world of ratings and other marketing tools, as well as of the increasing body of commercial and historical data. With regard to individuals, their decisions can now be tracked more closely and will be even more so, with a consequent improvement in the data.
In biology and psychology, progress is being made in understanding brain function, suggesting that reliable psychological theories may help to predict human motivation and decision in quite a refined way, something that is already done by business experts and others already via common sense when they make commercial decisions, presumably with better than 50% success. This in turn implies that economic pheniomena have stable broad features (other than obvious ones such as those that make hot cakes sell as they do) that can be assessed by an acute observer. An eventual AI observer may do even better than a human one.
I would compare the problem to that of simulating weather patterns, and is the science of ecology not tackling analagous problems when and if it concerns itself with the development of complex environments and a given species within them?
Such a refinement of economic theory may even help towards a more equitable distribution of wealth, and at least is surely likely to provide ammunition to one side or the other on that issue. Hopefully, progress in technology and politics will make the whole field less interesting than unfortunately it is at present.
Posted by: Gunnar Tomasson
Addendum.
The twin concepts of Determinacy and Indeterminacy are counterparts in theoretical economics to the NKS concepts of Computational Reducibility and Irreducibility – an indeterminate theoretical proposition is one that may or may not be valid.
This has important implications in monetary economics for, as noted by David Hume in an essay ‘Of Public Debt’, in which he reasoned in terms of Commodity Money, his conclusions were not valid for Money which lacks intrinsic commodity value.
Modern money lacks intrinsic commodity value – that is to say, the Supply of Money, unlike that of Goods and Services, does not require Inputs of Factor Services, but is created with the proverbial stroke of a pen (or computer key).
In principle, therefore, the Supply of Money is indeterminate.
In turn, this implies that Computational Irreducibility is an intrinsic and inescapable attribute of real-world market economies – a fact obfuscated by Paul A. Samuelson in formulating the concept of “an operationally meaningful theorem” in ‘Foundations of Economic Analysis’ (1942).
“By a meaningful theorem,” Samuelson wrote, “I mean simply a hypothesis about empirical data which could conceivably be refuted, if only under ideal conditions. A meaningful theorem may be false. It may be valid but of trivial importance. Its validity may be indeterminate [sic], and practically difficult or impossible to determine. Thus, with existing data, it may be difficult or impossible to check upon the hypothesis that the demand for salt is of elasticity – 1.0. But it is meaningful because under ideal circumstances and experiment could be devised whereby one could hope to refute the hypothesis.” (Atheneum, New York, 1979, p. 4)
In the context, a “meaningful theorem” which is also “indeterminate” is an oxymoron on par with Reducible Computational Irreducibility!
Samuelson then proceeded to transform this oxymoron into the cornerstone of what, in due course, became orthodox mainstream economics as follows:
“In this study I attempt to show that there do exist meaningful theorems in diverse fields of economic affairs. They are not deduced from thin air or from a priori propositions of universal truth and vacuous applicability. The proceed almost wholly from two types of very general hypotheses [‘a priori propositions of universal truth and vacuous applicability’ – insert]. The first is that the conditions of equilibrium are equivalent to the maximization (minimization) of some magnitude….
“However, when we leave single economic units, the determination of unknowns is found to be unrelated to an extremum position. In even the simplest business cycle theories there is lacking symmetry in the conditions of equilibrium so that there is no possibility of directly reducing the problem to that of a maximum or minimum. Instead the dynamical properties of the system are specified, and the hypothesis [‘a priori proposition of universal truth and vacuous applicability’ – insert] is made that the system is in “stable” equilibrium or motion. By means of what I have called the Correspondence Principle between comparative statics and dynamics, definite operationally meaningful theorems can be derived from so simple a hypothesis. One interested only in fruitful statics must study dynamics.” (Op. cit., p. 5)
The fact that the Supply of Money does NOT require Inputs of Factor Services means that Samuelson’s hypothesis “that the conditions of equilibrium are equivalent to the maximization (minimization) of some magnitude” cannot IN PRINCIPLE apply to real-world market economies.
I pointed this out to Samuelson in 1978 – and now give him the last word:
“…a scholar in economics who is fundamentally confused concerning the relationship of definition, tautology, logical implication, empirical hypothesis, and factual refutation may spend a lifetime shadow-boxing with reality.” (Op. cit., p. ix)
Gunnar
Posted by: John Gelles
I see political economy as common sense economics based on stated objectives, such as, mimicking "one-man, one vote" to try to achieve "one man, one income (above the poverty line)".
The objective might include, relative to gargantuan individual income, avoidance of an income ceiling but use of a restriction in law that prevented people (and firms) from "buying the law".
This view of "political economy" is within the positivist realm of my view of "economic science".
Economic science, on the other hand, can only embody a very general objective of seeking to understand and maybe model human economies we observe from time to time and place to place.
Thus economics as a formal science approaches human economies and insect economies with the same tools and non-judgmental attitude.
Because data expressed in material measure (excluding price) can be used in economic science -- especially in logistics -- I believe such formal study can be useful.
Because price data is distorted and unnatural, its use in economic science distorts all modeling and tends to accept and give license to much that is unfair and unnatural. Price data can, however, be very satisfactory for use in political economy (common sense economics).
Admittedly, we are not ready to wholly replace market prices with computed prices; but we are ready, I believe, to augment bank created money with debt-free managed money to reach economic objectives such as defense of freedom, full employment, protection of the environment, renewable energy, universal health care, reduced conflict, etc.
Prices that would flow from such augmentation of our money (to allow government to spend it into circulation) would be useful in gathering feedback and generally managing a national economy to reach its objectives.
A series of well managed national economies might improve the global economy.
Posted by: Fiona Maclachlan
alphasun raises the important question of whether the fundamental unit of analysis in economics is indeed human decision. Might it be possible to jump in and analyze some regularity in time series data, say, without inquiring into the decisions that generated it? Jason Cawley raises a similar point, I think, in his comments regarding GDP data.
My answer is yes, models can be constructed that will generate predictions that will be roughly right, most of the time. I could model tomorrow's gold price as today's gold price plus a normally distributed random variable with an expected value of zero and some specified variance, for example.
My model might be shown to predict gold prices as well as any other model, in which sense it's a "good" model. But is it an economic theory?
I'd say no, and that many of the more complicated models used in business, economics and finance for the purpose of forecasting future variables are essentially no different. They may appear scientific but they are really just extrapolating from the past, taking advantage of the fact some variables and relationships in economic life are
not subject to violent change.
We are led back to the question of causality and its role in scientific theory. My position with regard to the social sciences is expressed well by one of my teachers, Ludwig Lachmann (1906-1990):
With such meagre and unimpressive contributions to human progress to their credit, wherein lies the superiority of the social sciences? In the fact that they can go beyond mere description and correlation, and render the social world intelligible by reducing the phenomena of human action to that irreducible final cause: human choice. The natural sciences, after all, adopted their present-day methods after centuries spent in vain search for ultimate causes, not out of strength but out of despair. I can see no cogent reason why we, who are in a more fortunate position, should follow their lead. ("Economics as a Social Science" South African Journal of Economics, 18 (Sept. 1950), reprinted in Capital, Expectations, and the Market Process, Sheed, Andrews & McMeel, 1977.)
Posted by: Karl Smith
QUOTE
My answer is yes, models can be constructed that will generate predictions that will be roughly right, most of the time. I could model tomorrow's gold price as today's gold price plus a normally distributed random variable with an expected value of zero and some specified variance, for example.
My model might be shown to predict gold prices as well as any other model, in which sense it's a "good" model. But is it an economic theory?
ENDQUOTE
I'd say yes, it is.
If you just arrived at your model by fishing or by asking God then I would say no. However, if you took the Effecient Market Hypothesis serious combined with David Dickey's work and said, you know what "I think gold price is a random walk" then yes thats theory.
Gold may not be the best example because one might think naturally, "Why would gold prices change"
However, look at stock prices. Surely the causal observer doesn't say "Random walk with drift."
The causal observer thinks there are patterns or underlying predictible valuations, or mass hysteria but usually not random walk.
Forum Sponsored by Wolfram Research
© 2004-2008 Wolfram Research, Inc. | Powered by vBulletin 2.3.0 © 2000-2002 Jelsoft Enterprises, Ltd. |
Disclaimer
vB Easy Archive Final - Created by Xenon and modified/released by SkuZZy from the Job Openings