Dear Shiozawa sensei and ResearchGate community,

I could not agree more with you when you state that all data-first theorist like Hoover, Hendry, Juselius, Johansen, Spanos are deeply influenced by F53. In the end, all of them follow a marshallian approach. According to the four aspects of scientific research, they start from (3) and end up in (1). Regarding (3), it is necessary to recall that "Data-First" theorist do not transform or curate data since they are "market processes" and, according to Hendry (2011), are subject to three kinds of unpredictabilities: intrinsic, instance and extrinsic. In other words they "let the data speak for themselves".

However, I don’t think the vast majority of economics are influenced by F53 Positivism or Popperian Falsificationism in strict sense inasmuch as RBC and DSGE models (the widespread models in Economics), whose predictive power is not good, have not being ruled out. Professor Mário Amorim Lopes explanation about popperian epistemological approach on social sciences was really clear and contundent. For instance, these models were not able to predict 2007/08 Financial Crisis, they were not able to survive falsifications, albeit they are still used for the vast majority of Central Banks in several countries. Kirman (2010) stated that “The Economic Crisis is a Crisis for Economic Theory”.

Now the question is, are DSGE models the best theory available? Are there other theories which are able to predict Economic Crisis? Kirman (2010) supports the idea that Shiozawa sensei stands for (so do I): viewing an economy as a complex adaptive system, a set of interdependent elements (agents) organized in networks (without a central control) which produce emerging aggregates and have the properties of adaptation and self-organization. In that sense, to overcome DSGE scenarios with representative agents, rational expectations, walrasian law (markets empying) and stochastic trend; it is necessary to build models that explain and predict economies with contagion, interaction, interdependence, networks and trust.

So far, we have identified that it is necessary to construct models which consider Economic Crises as inherent to the evolution of the complex system. But can we identify the evolution of the system? This responsibility lies in two different hypotheses: i) Former economic theories that have been ignored like the Financial Instability Hypothesis by Hyman Minsky; and ii) Approaches from other disciplines such as: Econophysics (see Jovanovic, F. y Schinkus, C., 2013; Rickles, D., 2008 and Sornette, D., y Zhou, W., 2007).

Allow me to discuss some ideas on Econophysics (I am deeply interested in this field). First of all it is necessary to recall that Financial Markets Data present certain stylized facts: i) Fat-tailed distributions (Instance Unpredictability, Hendry (2013) – Taleb’s Black Swan); ii) Volatility; iii) Autocorrelations (memory); iv) Leptokurtosis and v) Clustering. According to this, the normal distribution, martingales and random-walks which are the battlehorses of Eficient Market Hypothesis by Fama and therefore DSGE models, does not shed a light on Financial Market Data. On the other hand, Econophysics put forward the use of “Truncated Levy-Pareto” distributions which address all those stylized facts stated above. These distributions are bell-shaped like Gaussian distributions but unlike these ones, they assign bigger probability to the events in the center and the tails of the distribution (Economic Crises). (Jovanovic, F. y Schinkus, 2013).

In that sense, given that Econophysics view the economies as a Complex Adaptative System and provides a good explanation on Economic Crises, why DSGE models are still used? I think he answer to this question responds to interests (professor Karlsson emphasized on it above) and the arrogance of most Orthodox Economists. They are reluctant to ruling out DSGE models and accept developments coming from other disciplines outside economics. I do agree with Moisés Naím when he states that “while there may be budding intentions to appeal to other disciplines in order to enrich their theories (especially psychology and neuroscience), the reality is that economists almost exclusively study—and cite—each other”.  (http://www.theatlantic.com/business/archive/2015/04/economists-still-think-economics-is-the-best/390063/)

To sum up, I think neither Friedmanian Positivism nor Popperian Falsificationism is followed in strict sense by the vast majority of economists. The current bulk of models do not care about predictions, they just follow the “discipline of equilibrium” (Representative Agents, Walras law and Rational Expectations). I exposed the example of Economic Crises and Financial Markets inasmuch as it is the most important falsification of DSGE models, but there are other falsifications in other fields like Economic Growth and Development (my dissertation states about it but unfortunately it is in Spanish and I have not translated it to English yet, my apologies).

Thanks a lot for sharing your valuable concepts on this related topics

Édgar    

REFERENCES

1. Hendry,D. (2011). "Unpredictability in Economic Analyis, Econometric Modelling and Forecasting," Economics Series Working Papers 551, University of Oxford, Department of Economics.

2. Jovanovic, F. y Schinkus, C. (2013a). Towards a transdisciplinary econophysics, Journal of Economic Methodology. Volume 20, pp. 164-183

3. _______________________ (2013b). Econophysics: A new challenge for financial economics? Cambridge University Press, 319-352.

4. Kirman, A. (2010). The economic crisis is a crisis for economic theory. CESifo Economic Studies 56: 483-535.

5. Rickles, D. (2008). Econophysics and the complexity of financial markets, Handbook of the philosophy of science, Volume 10, pp. 133-152.

6. Sornette, D., y Zhou, W. (2007). Self-organizing ising model of financial markets. The European Physical Journal B, 55(2), 175-181

Hello Shiozawa sensei. It is nice to read you once again and thanks for bringing up such interesting matters.

I have got some ideas on Methodology of Economics and I would like to discuss them with you and with ResearchGate community.

First of all, It is necessary to recall that, according to Hoover (1984) and Juselius (2010), there are two approaches in terms of Methodology of Economics: i) General-to-specific (GETS), “Theory First” or walrasian approach and ii) Specific-to-general, “Data First” or “marshallian” approach.

The first one is characterized by presenting a “totalizing” hypothesis; therefore, it is already determined and it is not necessary to consider other hypotheses. In that sense, the researcher starts from the mathematical formulation of theoretical model and then embeds a stochastic component. The next step is to transform and adjust the data in order to support the initial and absolute hypothesis.

On the other hand, the second approach is more “archeological” insofar as the theory is constructed by parts. The researcher draws from the statistical model suitable to data and then it is reduced to several theoretical alternatives, thus this methodology “let the data speak for themselves” (Juselius, 2010).

Regarding to the statements above, Friedman, 1953 (F53) stated  that Economic Theory "is to be judged by its predictive power for the class of phenomena which it is intended to explain” (F53, p. 8). This thoughts totally fit on "popperian positivism" which states that, insofar as a certain theory has a good adjustment to the data it can be considered as science.

According to this, one could link Friedman’s Positivism with the second approach inasmuch as he considered that models had to be the closest possible to reality instead of mere abstractions and generalities; thus relevance must not be sacrified in order to gain mathematical elegance (mathiness). Morover, he used to agree with Marshall when he stated that economic theory “is not a body of concrete truth, but an engine for the discovery of concrete truth” (Keynes, 1956). In that sense, when he formulated his Permanent Income Hypothesis of Consumption, he drew from a big set of historical data on consumption, savings, etc. and then contrasted this data with his model, obtaining a good explanatory power. This model won over other models with good prections, such as Modigliani (1943) and Duesenberry (1943), since it was more parsimonious and fertile than its competitors.  

On the other hand, the bulk of theories derivated from Lucas (1976) and Kydland & Prescott (1977), such as RBC and further DSGE models, could be linked with the walrasian equilibrium approach inasmuch as, predictive power does not really matter. In 1988, during his welcoming speech for freshmen in Chicago, he stated that "economics are storytellers, operating much of the time in worlds of make believe". In that sense, researchers who follow this approach (the vast majority of mainstream economics) start with a really sophisticated mathematical model and then develop "virtual worlds" which fits the model under a walrasian equilibrium (calibration).

To sum up, regarding Methodology of Economics we have a path which forks in two approaches: marshallian and walrasian, each one with their respective "prophets": Friedman (1953) and, Lucas (1976) & Kydland-Prescott (1977). Personally, I think the same way as Shiozawa sensei, "economic science should take back a more suitable balance between theory and empirics". However I find really interesting this debate about what should comes first in economic research, theory  or data.

Greetings! Shiozawa sensei.

 REFERENCES

Duesenberry, J. (1949). Income, saving and the theory of consumer behavior, Harvard University Press.

Friedman, M. (1953). The methodology of positive economic, in Essays in positive economics, Univ. of Chicago Press.

Hoover, K. (1984). Two Types of Monetarism, Journal of Economic Literature, vol. 22, Iss.1, pages 58-76.

Juselius, K. (2010). On the role of theory and evidence in macroeconomics, University of Copenhagen Department of Economics Discussion Paper No. 10-12

Keynes, J.M., ‘Alfred Marshall’, Essays in Biography (Meridian Books, New York, 1956), p. 70.

Kydland, Finn E.; Prescott, Edward C. (1977), «Rules Rather Than Discretion: The Inconsistency of Optimal Plans», Journal of Political Economy 85 (3): 473–491

Lucas, Robert (1976), «Econometric Policy Evaluation: A Critique», en Brunner, K.; Meltzer, A., The Phillips Curve and Labor Markets, Carnegie-Rochester Conference Series on Public Policy, 1, New York: American Elsevier, pp. 19–46

Modigliani, F. (1949). Fluctuations in the saving-income ratio: a problem in economic forecasting, National Bureau of Economic Research.

Porque la ideologia neoliberal revstida de ciencia es la dominante a nivel mundial. Porque los pseudocientíficos ortodoxos son fundamentalmente los ganadores de los Nibel y el mundo hace mucho tiempo giró hacia la derecha del espectro político.

Édgar Hernando Sánchez Cuevas

Welcome back to my question page and thank you for posting a long answer. I am also happy to have a chance once more to discuss with you. 

I have to apologize that my response was not quick. I was reading Juselius's paper On the Role of theory and Evidence in Macroeconomics (2010). In this post, I well only talk about what I felt by reading Juselius's paper.   

My general impression is as follows. Jeslius's analysis or criticism on Ireland's paper (2004a) is detailed and well discussed and I believe her arguments are right. I am not well versed in econometric method and have to confess that I cannot understand the details of her arguments. However, her style of arguments, problem setting and efforts to seek alternatives impress me a strong influence of Milton Friedman or his predictionist methodology. This does not imply that her examination was useless. If it is put in a plausible total framework how science and economic science develops, I think it includes many teaching. It must be more persuasive than my arguments for economists who still believe in the relevance of DSGE model and RBC theory and I hope it will have strong effects on them. 

Let me first explain my image on how a science develops. We have to think it develops in a spiral which expands in time (imagine a logarithmic spiral). At the antipodes of this spiral, there lie theory and empirical studies. Then we should distinguish at least four aspects of scientific research: 

 All these four aspects go round and round and science evolves. If we distinguish two aspects theory and empirical studies, the scientific development is a co-evolution of theory and empirical works.   

All four aspects are important but Juselius is only interested in (2) and (3). She compares two aspects by the terms "theory-first" and "data-first" approaches. She mentions on other aspects but only sporadically. What is more serious is that Juselius seems to be identifying theory and models. Her concept of model is something which can be easily replaced by another. This signifies that she is not really thinking what a theory is. Theory is the framework in which we think of our world (economy in the case of economics). A theory is displaced only by a new theory (Nothing can defeat a theory but a theory; Hanson's thesis) and new theory emerges within the old theory (Neurath's ship metaphor). There is a firm idea among many economists that we can freely replace models when they do not fit to empirical data. This idea is what Friedmanian predictionism dictated to us. 

The second dissatisfaction with Juselius is the shortness of the memory. Of course, I know she is much younger than me. It seems she started her research life about thirty years ago (in 1980’s). I do not blame her in person. What I blame is the short memory of economists as a whole. When REH (Rational Expectation Hypothesis) appeared, there was a tremendous debate about this hypothesis. There is no mention on it and she brings IKE (Incomplete Knowledge Economics) but I do not understand why she goes back to years before the rattling arrival of RAH. She talks about Cobb-Douglas production function but do not question the relevance of this concept. Her main interest whether it fits to the data. She does not remind of the long history debate on conceptual relevance of this function. 

Friedman proposed to ignore the reality of assumptions. This proposition leads to ignore all relevance problems of a theory. A theory became an entity that can be examined only by questioning if it predicts well (or if this prediction is useful for our purpose). If Juselius really seek a new economic thinking, she must rethink more deeply. 

Yoshinori Shiozawa's question is well taken, and Édgar Hernando Sánchez Cuevas's response is a pertinent reflection upon the methodology of economics from the standpoint of theory. I fear, however, that this reflection does not give the answer to Sjiozawa's question; and I am afraid that the reasons that many economists are deeply influenced by Milton Friedman's economic theories - or at least say (or even think) that they are - is that so many economists are bought and paid for by people whose interests are served by Friedman's views. Those people are (a) politicians - Augusto Pinochet comes to mind as an example - and (b) those whose clients the politicians are, namely large corporations, and most particularly large financial institutions. Friedman was undoubtedly an important economic theorist, but there are many of those, and quite many opposed, as theorists, to Friedman's ideas (or those of the so-called "Chicago School"  more generally) It's the Chicago School approach to running economies that is most congenial to the perceived interests of the powerful, and few high-profile economists, who "advise" governments, are uncompromised. Economic "advice" and opinion is definitely for sale. In fairness to the profession, it may even be the case that the majority of economists are not much persuaded by, or at any rate, are not apostles of, Friedman and the Chicago School, but those are not the economists that one hears about or who are recruited, and very well paid, to "advise" the powerful (and to run central banks, issue think-tank "reports", publish in high-profile journals, and so on). For many young economists, Friedman is the sign-post to career advancement as opposed to the advancement of science. This has all been well explained and documented by many people, including, not least, William K. Black, as for example here:

http://neweconomicperspectives.org/2014/01/economist-treats-milton-friedman-jesus-asks-disciples-preach-gospel.html

I realize that this answer is not framed in the terms that Shiozawa was looking for, but I believe that it's the right answer to the question he actually asked.

Dear Ramiro restrepo u,

I do not read Spanish. This is what I guessed that you wanted to say.

• Economists are under the strong influence of Milton Firedman, because ideology of neoliberal-revisionism of science is dominant in the world level and because orthodox economics is a pseudo-science. The world has turned to the right of the political spectrum since long time ago.

Here is my opinion:

You are right to detect this political aspect, but it does not help us to redress economics to a right way. If you really want a good economics, you should enter into the battle field of economic theory, and win the battle. Please search and find a new way to construct a new economics. Political denounces will not help you very much. Your enemy has much more political power than you.

Dear Yoshinori Shiozawa, I agree emphatically with your response to Ramiro. But his brief comment, and my longer comment in much the same vein, were, I think, appropriate responses to the question you posed. You did not ask how economics might be put on a better path, and by the way, it is not only "orthodox economics" (which is not only that of the Chicago School) that has degenerated into pseudo-science, sometimes referred to as "voodoo economics", but also much of non-orthodox economics: the part that is convenient to be used as an instrument for political ends. Here I speak of theory or methodology. But in addition to theoretical weakness, there is the character of professional practice which has become largely compromised for the reasons mentioned by Ramiro and myself (which is hardly news). However, there are various on-going attempts to put economics onto a better path - one that is more "scientific" - and to reform the profession; and these two things go hand in hand. Economics as a "science" is a nascent science, with competing schools of thought and theortical controversy. Historical study reveals that this is characteristic of any science in its formative stages. Sadly, serious "scientific" economics has not been allowed to develop in a studied and principled manner but has always been coopted for political or commercial ends. Those who are dedicated to serious theoretical development in economics must pay close attention to economic history and must have a just view of "classical" economic theories, such as those of Smith, Mill, Ricardo, the Physiocrats, Marx, etc. and of more modern theories such as those of Keynes, the Austrian School and, yes, even the Chicago School. To understand is, of course, not necessarily to agree. But agreement or disagreement without understanding is worthless. Few economists today have such a background. And then beyond that, a serious economic theorist must either critically develop some point of view or propose a new theoretical framework, supported by reasons. Since you now pose the question of how economics might be "redressed" to a better path, with an emphasis on theoretical or methodological matters rather than professional fraud, I suggest looking at the following, if you haven't already done so:

(1) Some critiques of the Chicago School by Henry H. C. Liu, who has a blog and also writes for the Asia Times on line: "Milton Friedman and the Money Matters Controversy" (2008): http://henryckliu.com/public_html/page167.html and the similar, but not identical, "Friedman's misplaced monument" (2008): http://www.atimes.com/atimes/Global_Economy/JI05Dj03.html and also "Monetarism enters bankruptcy" (2009) http://www.atimes.com/atimes/Global_Economy/KA06Dj04.html

(2) The efforts of some "heterodox" economists (including William K. Black, linked in my previous comment), who are developing what they call "Modern Monetary Theory", for instance: "Modern Monetary Theory – An Introduction: Part 1" http://neweconomicperspectives.org/2013/04/modern-monetary-theory-overview-part-1.html; "Modern Monetary Theory – An Introduction: Part 2" http://neweconomicperspectives.org/2013/04/modern-monetary-theory-overview-part-2.html; "Modern Monetary Theory – An Introduction: Part 3" http://neweconomicperspectives.org/2013/04/modern-monetary-theory-overview-part-3.html; and the more comprehensive Modern Monetary Theory Primer http://neweconomicperspectives.org/modern-monetary-theory-primer.html

(3) Various articles and books by the economist Michael Hudson - strong in economic history - for whom you can begin here: http://michael-hudson.com/

I am not suggesting that these people have the best theories or methodologies, but only that they are making serious efforts to do what you asked about. And there are indeed plenty of others, but you have to hunt them down because they are strenuously resisted by the establishment.

I hope that this may help in trying to answer your new question.

Dear Mikael M Karlsson

I have just printed William K. Black's paper and am going to read it now. Please wait until I will read it. Your second post includes much more readings. It must take much time to read.

Shiozawa

The simple answer to your question is that there is nothing better yet been developed. Until it does happen, the economists will always stick to and teach according to the older and previously acceptable ways. That way they "keep their asses covered" to use the Americanism.

But wait a moment! There is something new coming over the horizon! At last a new, scientific, logical and easy to understand theory is available and free for use. Please review my attached writings and in particular my "Consequential  Macroeconomics"  that hopefully will at last set the course of the Ship of State under better guidance and help the next generation to stop making the same silly mistakes that our past love of orthodoxy will never allow to change.

Dear Yoshinori Shiozawa, Yes, reading does take time, particularly when the issues are large and complex. I hope that the references are helpful. By the way, the excellent paper by Akerlof (and Romer) referred to in Black's article (that you are now reading) is available in its entirely on line at:  http://pages.stern.nyu.edu/~promer/Looting.pdf  That paper does not propose a new methodology (theory), but it is extremely critical, in specific and well-reasoned ways, of the application of a more familiar methodology. Akerlof and Romer are excellent economists. William K. Black is an economist but also a white-collar criminologist and often applies his economic skills in revealing fakery, fraud and what he calls "economic illiteracy".

Dear Mikael M. Karisson

I have just read Black's article. It was interesting and revealing to me. The story of Edwin Gray, the Chairman of the Federal Home Loan Bank Board of the time was encouraging. We had  such a person. I am happy to know that not all people in the top are "bought and paid for by people whose interests are served by Friedman's views". We have some honest people whom the money cannot buy.

I also peeped at some other papers and posts in website. Hudson's paper read at Beijing University, October 10 this year, showed me a new reading on Marx's 3 Volume Capital and  Surplus Value. It detected how Marx was mislead and hints a new reading and a new understanding of the present economics system, i.e. the relations between FIRE economy and Industrial Economy.  Henry H. C. Liu's story on Chicago School was also a good reading. Lieu knows to combine various different topics into a coherent story.

It seems Hudson worked in Wall Street and Lieu is working there now. There are these kinds of people who are working in the center of Greedy Capitalism but are keeping a healthy spirit.

What you and Ramiro wanted to point is right and it is necessary that many people should know (including me). However, what I wanted to say (as you understand well) is the problem of the economists who are not proponents of Neo-Liberalism or Milton Friedman style free market fundamentalism. They are, I believe, a majority of economists who are not adhered to Friedmanian  economic thought. If they are questioned if they agree with Friedman's market liberalism, many of them would answer that they are against such a crude market liberalism. What I question is that even those people are under the strong influence of Friedman's methodology and cannot find out another possibility in economics.

As this is a paradigm change for most of mainstream economists, a deep theory debate is necessary. Without this serious theory arguments, there will be no new development. 

George Soros stated an Institute for New Economic Thinking. Many prominent economists are now affiliated to this Institute. I wonder if a really new thinking is developed around this institute. 

Excellent comment by Édgar. Let me just add that according to Popper the right epistemological approach to social sciences, including economics, is comparable to Alfred Marshall's approach to a certain extent: you start by devising your theory (inspired by observation or even empirical data) and subject it to rational criticism. Thereafter, you should use data to try to falsify your theory. You'll never be able to prove it right, but as long as it survives falsification it will remain valid, or at least not proved wrong — Einstein's relativity theory is still.. a theory.

In response to Edgar Hernando and Mario Amorim, I want to point out that the oppositional framework of "theory first" and "data first" does not have proper space for theoretical works. If it is replaced by the dichotomy of　"general to specific" and "specific to general", the same point can be questioned. 

Why do many people believe that theory is general and data is specific? This fact reveals a common specific vision concerning a theory and data. Many economists identify theory as a general model (with undetermined coefficients) and think that its structural coefficients are specified by data. At the center of this vision lies the concept of model. Theory tells how to construct a model with many undetermined coefficients and empirics determines these coefficients. If you make a regression or a calibration, the story is the same. Here lies already a very special idea about theory and how it works. Theory is a simple selection of a model among many candidates and empirics determine coefficients of the model. This is almost what Milton Friedman had in mind when he talked about theory and prediction. 

For him,

• "theory is, in general, a complex mixture of two elements. In part, it is a "language" designed to promote "systematic and organized method of reasoning. In part, it is a body of substantive hypothesis designed to abstract essential features of complex reality." "Viewed as a language, theory has no substantive content. Its function is to serve as a filing system for organizing empirical material and facilitate our understanding of it." (Section II, Positive Economics) 
• "Viewed as a body of substantive hypothesis, theory is to be judged by its predictive power for the class of phenomena which it is intended to "explain."" (ibid.) 

Although Friedman concedes that theory is a language designed to promote "systematic and organized method of reasoning," it is after all "a filing system." Many possible models are filed in order that one of them is drawn out to be tested in empirical examination. In the same way, in the brain of majority of economists, several models are filed. If one model did not show a good performance, another model should be drawn out. 

Of course, this is a caricature, but this is the rough idea that occupies the heads of almost all economists. There is no proper place for theory. More precisely, a name exists but theory has no function to play. Theory is but still-lives which are classified on a shelf. In my previous post (the third post in this question page), I contended that there are four aspects in scientific research and theory and empirical work must co-evolve. In the mind of Friedman, the second aspect (i.e. testing predictive power of a hypothesis) comes in the forefront and other three aspects are almost forgotten. A dichotomic question "theory first" or "data first" lies in this very particular problematic of scientific research. 

Then, how does theory work function? Let me cite an example. I believe that many people agree that an economy is a complex system. If we really think that an economy is a complex system, the above standard procedure of “build a model and test/reject it” is too naive. Testability of a model or a hypothesis depends largely on the working of the system. Test/reject paradigm is only applicable to a simple system or to a system with a special structure that enables application of test/reject paradigm. 

In 1948, Warren Wiever published in American Scientist a paper with a title Science and Complexity. This was the first manifest for whole complexity sciences which came to develop about 20 to 30 years later. In this paper, Wiever distinguished three types of problems of sciences. 

One is the "problems of simplicity". It treats a system of a few variables. A typical example is Kepler's elliptic orbit. It is a two body system and we got an exact solution (a solution which satisfies the given system of differential equations.). For example, earth moves in an elliptic orbit with the sun as one of its centers. When a moon was added to this system, it is a three body system. Poincare proved that we have no global analytic function by which to describe the movement s of three bodies. It is possible to compute by a computer for example, and we can get a precise solution as far as we want but it is not exact in the sense that it expresses a solution to a system of differential equations. Even if there are various kinds of complications, a system was tractable as far as it is relatively a small system. Almost all results which have a predictive power before the 20th century contained only a few variables. An exception was the general theory of Newtonian mechanics. Analytical mechanics developed extremely in the 18th and 19th centuries but it was a kind of "algebraic theory" (in the sense Hayek used this term). We know how the system works in principle but there were no means to compute a solution for a large system. 

At the second half of the 19th century, a new physics emerged: thermodynamics and a kinetic theory of gases. In the first half of the 20th century, we obtained statistical mechanics and quantum mechanics. Wiever classified these sciences that treat "problems of disorganized complexity." The system contained enormous number of particles (for example, a mol of hydrogen gase contains about 6×1023 molecules.) This system became tractable by the development of probability theory in mathematics and the success of statistical mechanics ignited a rapid development of probability theory thereafter. Statistical mechanics was one of greatest achievements of 20th century physics but it was not a full-fledged complexity science. Mutual interaction of the particles was rare and limited and most of events can be treated to be independent of other events. 

At the middle point of the 20th century, Wiener, the then science department head of Rockfeller foundation, thought that the second half of the 20th century should be the era of sciences which challenge "problems of organic complexity. In this category, Wiever included biochemistry or molecular biology, economics and others. The target of these sciences was a system of intermediate size with relatively strong internal interactions. 

Wiever's manifest was too ahead of the time. Only a few scientists and scholars notice it as having an important message. Among such people, we count Hayek and Jane Jacobs. Inspired of this paper, Hayek wrote "the theory of complex phenomena" (1964). In this article, Hayek used the term "algebraic theory" which was borrowed from Watkins. 

Topics in complexity now spread into too large an area of sciences and it is difficult to describe even a rough map of the complexity sciences. However, those who have little knowledge of problems of complexity are required to obtain a basic knowledge on it. For example, it is advised to read the First Chapter of Gottinger's Coping with Complexity (1983, 2013). It is a short summery of what complexity implies for sciences and for economics in particular. 

Friedmanian paradigm of test/reject of a hypothesis is an image of scientific activity before the arrival of complexity argument. Much earlier than the proliferation complexity sciences, this kind of naive philosophy of test/reject was rejected by W.V.O. Quine (see “Duhem-Quine thesis” in Wikipedia). It is not evident that a model can predict anything if we really know the complex structure of an economy. Theory inquires this kind of questions. Selecting and building a model is a very small part of theoretical activity. 

Some people classify Hayek and Friedman in the same category, because their political thoughts have something in common. However, Hayek was much more serious in examining the possibility of economic inquiry. Friedman's "model" of science is too simplistic. Leave a defunct Friedman alone. The trouble is that there are still so many economists who think in the old framework of Friedman. This makes the real crisis of present day economics.

Dear Shiozawa sensei and ResearchGate community,

I could not agree more with you when you state that all data-first theorist like Hoover, Hendry, Juselius, Johansen, Spanos are deeply influenced by F53. In the end, all of them follow a marshallian approach. According to the four aspects of scientific research, they start from (3) and end up in (1). Regarding (3), it is necessary to recall that "Data-First" theorist do not transform or curate data since they are "market processes" and, according to Hendry (2011), are subject to three kinds of unpredictabilities: intrinsic, instance and extrinsic. In other words they "let the data speak for themselves".

However, I don’t think the vast majority of economics are influenced by F53 Positivism or Popperian Falsificationism in strict sense inasmuch as RBC and DSGE models (the widespread models in Economics), whose predictive power is not good, have not being ruled out. Professor Mário Amorim Lopes explanation about popperian epistemological approach on social sciences was really clear and contundent. For instance, these models were not able to predict 2007/08 Financial Crisis, they were not able to survive falsifications, albeit they are still used for the vast majority of Central Banks in several countries. Kirman (2010) stated that “The Economic Crisis is a Crisis for Economic Theory”.

Now the question is, are DSGE models the best theory available? Are there other theories which are able to predict Economic Crisis? Kirman (2010) supports the idea that Shiozawa sensei stands for (so do I): viewing an economy as a complex adaptive system, a set of interdependent elements (agents) organized in networks (without a central control) which produce emerging aggregates and have the properties of adaptation and self-organization. In that sense, to overcome DSGE scenarios with representative agents, rational expectations, walrasian law (markets empying) and stochastic trend; it is necessary to build models that explain and predict economies with contagion, interaction, interdependence, networks and trust.

So far, we have identified that it is necessary to construct models which consider Economic Crises as inherent to the evolution of the complex system. But can we identify the evolution of the system? This responsibility lies in two different hypotheses: i) Former economic theories that have been ignored like the Financial Instability Hypothesis by Hyman Minsky; and ii) Approaches from other disciplines such as: Econophysics (see Jovanovic, F. y Schinkus, C., 2013; Rickles, D., 2008 and Sornette, D., y Zhou, W., 2007).

Allow me to discuss some ideas on Econophysics (I am deeply interested in this field). First of all it is necessary to recall that Financial Markets Data present certain stylized facts: i) Fat-tailed distributions (Instance Unpredictability, Hendry (2013) – Taleb’s Black Swan); ii) Volatility; iii) Autocorrelations (memory); iv) Leptokurtosis and v) Clustering. According to this, the normal distribution, martingales and random-walks which are the battlehorses of Eficient Market Hypothesis by Fama and therefore DSGE models, does not shed a light on Financial Market Data. On the other hand, Econophysics put forward the use of “Truncated Levy-Pareto” distributions which address all those stylized facts stated above. These distributions are bell-shaped like Gaussian distributions but unlike these ones, they assign bigger probability to the events in the center and the tails of the distribution (Economic Crises). (Jovanovic, F. y Schinkus, 2013).

In that sense, given that Econophysics view the economies as a Complex Adaptative System and provides a good explanation on Economic Crises, why DSGE models are still used? I think he answer to this question responds to interests (professor Karlsson emphasized on it above) and the arrogance of most Orthodox Economists. They are reluctant to ruling out DSGE models and accept developments coming from other disciplines outside economics. I do agree with Moisés Naím when he states that “while there may be budding intentions to appeal to other disciplines in order to enrich their theories (especially psychology and neuroscience), the reality is that economists almost exclusively study—and cite—each other”.  (http://www.theatlantic.com/business/archive/2015/04/economists-still-think-economics-is-the-best/390063/)

To sum up, I think neither Friedmanian Positivism nor Popperian Falsificationism is followed in strict sense by the vast majority of economists. The current bulk of models do not care about predictions, they just follow the “discipline of equilibrium” (Representative Agents, Walras law and Rational Expectations). I exposed the example of Economic Crises and Financial Markets inasmuch as it is the most important falsification of DSGE models, but there are other falsifications in other fields like Economic Growth and Development (my dissertation states about it but unfortunately it is in Spanish and I have not translated it to English yet, my apologies).

Thanks a lot for sharing your valuable concepts on this related topics

Édgar    

REFERENCES

1. Hendry,D. (2011). "Unpredictability in Economic Analyis, Econometric Modelling and Forecasting," Economics Series Working Papers 551, University of Oxford, Department of Economics.

2. Jovanovic, F. y Schinkus, C. (2013a). Towards a transdisciplinary econophysics, Journal of Economic Methodology. Volume 20, pp. 164-183

3. _______________________ (2013b). Econophysics: A new challenge for financial economics? Cambridge University Press, 319-352.

4. Kirman, A. (2010). The economic crisis is a crisis for economic theory. CESifo Economic Studies 56: 483-535.

5. Rickles, D. (2008). Econophysics and the complexity of financial markets, Handbook of the philosophy of science, Volume 10, pp. 133-152.

6. Sornette, D., y Zhou, W. (2007). Self-organizing ising model of financial markets. The European Physical Journal B, 55(2), 175-181

Dear Édgar,

thank you for your long and powerful post. In many points I agree with you.

You are right in pointing that "[t]he current bulk of models do not care about predictions, they just follow the 'discipline of equilibrium'." It is necessary to criticize this 'discipline of equilibrium.' I agree with you that the vast majority of economist who is working on RBC and DSGE models do not follow Friedman's positivism or Popper's falsificationism. They are loose minded scientists who do not like to think deeply and seriously. As you put it, they are only following the 'discipline of equilibrium' and what was done by someone. 

In this sense they are not practicing Friedman's positivism or Popper's falsificationism in strict sense. These "philosophies of science" have some intrinsic deficiency and reflect an image of scientific development of one time. They are criticized and overcome by later philosophers such as Hanson, Kuhn and Feyerabend. They are based on dogmas of empiricism (See also Two Dogmas of Empiricism in Wikipedia). There is no pre-established criterion (extrinsic to the science) by which to determine what is true and what is false. Truth value of a proposition in a science is determined through continued examinations and discussions by scientists in the discipline. Majority of economists pretends to believe in Friedmanian positivism only to escape from this heavy burden of continued examinations and discussions. 

This is one of the reason why we should reveal that (1) they are explicitly or implicitly following Friedmanian positivism and show that (2) their "philosophy of science" (or "philosphie spontanée des savants" if I use the term of Louis Althusser) is out of date and functions as an epistemological obstruction for the further development of economic science. 

As for econophysics, I am not following its recent development but Stanley and Mategna's Introduction to Econophysics was a holy sutra to me when it was first published in 2000. As for Efficient Market Hypothesis by Fama, it is wiser to distinguish what Fama established and what he has imposed on his discovery. The term Efficient Market is misplaced, because what he discovered is informational efficiency of the stock market and not the economic efficiency of the market mechanism. His discovery does not exclude a collapse of stock market. The scientific core of EMF is simply this: the expected distribution of tomorrow's price is a random distribution with the mean equal to today's price. Fama believed that this distribution was log normal but it was revealed to be truncated Levy distribution by the later research by Stanley and others. Here is a progress of a science and we should not deny Fama's contribution, even if it is a very crude one and his neoliberal ideology is supported by his discovery. As I have point it, what Fama discovered is informational efficiency (namely the semi-strong hypothesis) and not the efficiency of the market mechanism. Many economists including Fama himself are confusing these two.

Dear David Harold Chester,

I have looked into your book. I cannot say I have read it, because it is volumious and I had yet no enough time to read it through.

I learned several points even from Chapter 0 Motive. Notably the name of Henry George was totally new to me, although it seems that he is a very important thinker in social thought (I have read the item Henry George in Wikipedia). It is quite strange that in Japan people mention very rarely this important philosopher of economic thought. The fact I did not know the name of this person is one of the evidence. I found an item Henry George in Wikipedia Japan and one Japanese translation of George's book: Progress and Poverty. When I searched "Henry George" in Google Scholar in Japanese, there were only 87 hits.

Shiozawa

Porque Friedman cree que la economía es una ciencia positiva, exacta y tiene el mismo estatuto científico de la física. Los economistas que siguen a Friedman no tienen ética intelectual, son facilistas y no se cuetionan la debilidad epistemológica de la economíaortodoxa.

As explained in a 1979 Journal of Economic Literature, Friedman is not a positivist (despite the title of his book). He is a classic instrumentalist. And his methodology has nothing to do with Popper who by the way is not a "falsificationist". That was something Imre Lakatos tied on Popper for his own dishonest purposes.

Friedman defiende la economía positiva en su ensayo La economía positiva, que fue criticado por Joan Robinson cuando planteó su famoso dilema entre el realismo de los supuestos y su con cordancia con la realidad.

Dear Lawrence,

I made a rapid reading over some of your papers. I am not sure if I understand you correctly. There are several points that I agree with you, although it seems we are opposed in the deep place. 

You point in various place that Friedman's methodology is widely accepted by many economists. Here are three excerpts:

• Milton Friedman's essay 'The methodology of positive economics' (1953) is considered authoritative by almost every textbook writer who wishes to discuss the methodology of economics. (A Critique of Friedman's Critics, revised final draft, p.1)
•  most economic theorists claim they do not believe in instrumentalism --- but somehow actions speak louder than words. (Methodology for a New Microeconomics, 2015, p.128)
•  there is a very large proportion of economics profession which still thinks Friedman's methodology is alive and well.(Boland on Friedman's Methodology: A Summation, revised final draft, p.1) 

 That most economists are in the realm of Friedman's methodology, or accepting his instrumentalism, is really a serious problem for economics. You are right when you say

•  Methodologists must face the facts and explain the continued widespread acceptance of Friedmans's methodology. (Boland on Friedman, p.1)

 and

•  People have been attacking Friedman and his methodology for over thirty years and, as I continue claim, noting much has bee accomplished. (ibidem.)

 We should be aware of this history and contemplate why we were ineffective. I hope that many people join us in this question, but I will not argue this pint in this post. 

It seems that you are defending Friedman's methodology. Your arguments are well organized and your criticism against critics of Friedman's methodology is accurate as far as you discuss. I also agree with you that Friedman's purpose was to proclaim instrumentalism. All right. 

I am against his instrumentalism and deny that it is useful or effective as a methodology of economic science. More than that, instrumentalism is one of the main causes of the present crisis of economics (not the cause of economics crisis, if it may be indirectly). In my understanding, utility or effectiveness of a methodology lies on the effects it plays for the development of a science. In our case, this science is economics. Instrumentalism may have given some good effects in certain phase of economic science, but in my observation it is one of the most serious obstructions for the further development of economic science. 

Economics is in disarray in many aspects. However, most economists do not pay attention on this serious fact and continues to construct a new model and try to show that it fits to the data. If they succeed they claim that their model has a predictive power and useful in policy making. As you assert, they are only obeying the instrumentalism (or what they have understand as instrumentalism) and imagine that they are contributing to the wellbeing of the human kind. This is only an act of hypocrisy and yet they do not acknowledge what they are doing. This state of economics owes much to the instrumentalism.

What I cannot understand is why you want to defend Friedman's methodology. Do you defend him only because that many critics are misunderstanding Friedman's 1954 paper?

In your book (Methodology for a New Microeconomics), you examine the theoretical deficiency of equilibrium economics and you point that unless we have disequilibrium dynamics or out-of-equilibrium process theory, most of neoclassical theory has not firm foundations that permit its standard interpretation. You even write that 

• Followers of instrumentalism reject the complex theoretical questions that are posed by the need for the disequilibrium foundations of any equilibrium model used in policy economics. (Methodology for a New Microeconomics p.128)

 Don't you think that instrumentalism is an obstruction to the further development of economics?

If you were to read my 1979 JEL article to its very end, you will see that I am criticizing Milton. My point was that all the celebrated critiques were unfair and wrong, My critic that now everyone accepts was that his methodology is nothing more that 18th Century instrumentalism. I think that does not constitute a defense in any way. Criticizing failed critiques does not constitute a defense!

I suggest you look at Chapter 5 of my 2003 book.

Dear Lawrence,

I have read the paper in a draft stage of your 1979 JEL paper. I have seen the final part of the paper in JEL and confirmed there is no change in the published version.  

What you suggests as criticism of Milton Friedman seems this point:

• Any effective criticism of instrumentalism must at least explain the absence of refutations. (JEL, 1979, p.)

Do you think that this is a strong criticism against the wide spread instumentalism? I am rather doubtful. I want to know more positive "refutation" (argument) of yours  against instrumentalism that exists among many economists' heads.

A few lines earlier than the above cited phrase, you write

• in any effective criticism of his view the only issue possibly at stake is its truth or falsity of instrumentalism. (ibidem)

I cannot understand why you made this comment. A methodology can be useful or not or effective or not, but never be true or false. 

My simple point was that, given that Miltion's instrumental was being criticized, his ONLY defense was to invoke instrumentalism. I still do not think that is an adequate defense. At best it is circular.

Dear Yoshinori Shiozawa,

Thank you for your response of about two weeks ago to my earlier comment. I am pleased if you got some use out of my references to Black, Hudson and Lieu.

Now that the methodological discussion here has gotten quite serious and interesting, I would like to point you and your colleagues to an interesting book by Nancy Cartwright: Hunting Causes and Using Them: Approaches in Philosophy and Economics.

To my mind, Nancy Cartwright is one of the most interesting philosophers of science now writing. Her discussion is very pertinent to questions of methodology in economics.

All the best,

Mikael

Mikael,

I suggest that you read my review of Cartwright's book published in the Philosophy of the  Social Sciences. Unfortunately, Cartwright confuses econometric models with all of economics. There are no causes in economic models. We refer to exogenous variables instead.

Thanks, Lawrence, I will. By the way, what are you referring to under the label of "exogenous variables"?

Variables that economic models are designed to explain are called endogenous variables. Those variables that are not explained but play a role in the determination of the endogenous variables are called exogenous variables. Models usually consist of vary many of both kind. If you really do not know about this aspect of economic models, I suggest you look at Chapter 6 of my 1986 book which you can download free from my web page: www.sfu.ca/~boland.

My point before is that since there are many exogenous variables and all play a role, at best you can say they influence the determination of the endogenous variables. They are not the cause.

I have argued that Freedmanian predictionism and instrumentalism neglect the important role of theory works. Majority of economists are too busy in making "predictable models" and pay no attention to the inconsistency of theories. This must be one of the greatest poisons that Freedman had produced. 

I have posted a following question in May 8, 2015.

What's wrong with economics and what's right with it?https://www.researchgate.net/post/Whats_wrong_with_economics_and_whats_right_with_it#563b1fbb5dbbbdbaea8b457e 

Unfortunately this question attracted an interest of a very few persons (exactly 3 including myself as of November 5, 2015). This may also be an example of evidence that economists are not interested in theory questions.

Lawrence, I was being a bit ironical in asking what you meant by "exogenous variables". When you say "Those variables that are not explained but play a role in the determination of the endogenous variables are called exogenous variables. ... [A]t best you can say they influence the determination of the endogenous variables. They are not the cause"; you seem (as I suspected) to be working with a notion of "cause" (not to mention "effect" or "endogenous variables") that has been disowned by most of the best philosophers of science (including social science), although it will never lose its fans. Some of the philosophers of science whom I would mention particularly are Wesley Salmon, Nancy Cartwright, Donald Davidson and Fred Dretske - not to mention the best of them all, Aristotle. It's no wonder that you didn't understand Cartwright's book. Actually, how to understand causation and, correspondingly, the character of scientific explanation, are still lively fields of dispute, and will probably be so for a long time (you and I will not live to see any principled resolution), so I here express my opinion, not received dogma (as there is none). I admit to being no expert on economic models, and so I will, indeed, look at your book. Thank you for the reference and for making it available on your web site. However, I cannot imagine that models in economics are sui generis, i.e. uniquely unlike models in other sciences, such as physics, psychology, sociologiy, geology, or you name it. I can see in all this the basis of your deep methodological disagreement (as I believe it to be) with Sensei Yoshinori Shiozawa.

Dear Yoshinori Shiozawa, Thank you for your latest comment. I agree with you that predictionism and instrumentalism are ruinous ways of understanding scientific methodology and that Friedman can be credited with having pushed economics in a bad direction in just the way you suggest. However, he was actually a bit of a latecomer as fars as predictionism and instrumentalism go. Scientists and philosophers of science were persuaded of this idea of scientific methodology long before Friedman. It was perhaps the reigning methodological persuasion in many, if not most, of the sciences (I mean always to include the social sciences) during the first half of the 20th century and beyond, very characteristic of logical positivism (you can see it at work, and persuasively argued, by Bas Van Fraassen, who has been called by some "the last positivist"). It remains a lively persuasion, although, in my view, the best philosophers of science have long ago seen through it and moved on - or perhaps back - to better sorts of views. I applaud your recommendation to take theory seriously in economics, but what that means requires articulation, and the best minds are not agree upon how to do that - it's an ongoing project. Anyway, I am certainly going to look at the discussion thread you mention:

What's wrong with economics and what's right with it? https://www.researchgate.net/post/Whats_wrong_with_economics_and_whats_right_with_it#563b1fbb5dbbbdbaea8b457e

Thank you again for raising the issue.

Mikael, I do not think Shiozawa and I are in disagreement. Why do you think that? And as to Cartwright's book, I understood it well. It is economic model building that she does not understand. As I said, not all economics or economic model building involves econometrics but that is all she talks about. Perhaps you should look at my 2014 book on model building.

It is unfortunate that philosophers think they have a monopoly on clear thinking.

As to causes, economists do not talk about causes except when they want to engage in an ideological dispute as we saw in the 1970s and '80s about whether inflation is caused by too much money in the economy.

Lawrence, 

In the last post, you argued with regard to causes that

• economists do not talk about causes except when they want to engage in an ideological dispute as we saw in the 1970s and '80s about whether inflation is caused by too much money in the economy. 

I do not deny that there were many cases that the arguments of causality were used for the sake of ideology or for a predetermined policy recommendation. However, to proclaim that all arguments referring to causes are done with an (explicit or implicit) intention to engage in an ideological dispute seems to me also ideological. Here I mean by ideology a metaphysical or first-principle based argument which refutes any reasonable discussion. 

In macroeconomics, causes may be blurred, but if you examine human behavior, causal reasoning is often very useful. For example, you can examine what a manager does when such and such conditions are given. Neoclassical people argue that the manager equates the marginal cost with the product price. I will argue differently, because I believe it is not the price but the sales volume which determines the volume of production. This kind of discussion is I believe scientific. 

Yoshinori

First of all, I did not say all arguments. I said all arguments among economists.

Moreover, it is easy to find economists since Alfred Marshall to be discussing comparative static analysis where one changes just one exogenous variable of an equilibrium model but holds all others fixed by means of the ceterus paribus assumption. The analysis then recalculates the solution to the equilibrium model and thereby suggests that the one changed variable causes the new equilibrium solutions values. But this is totally fake causality since the world cannot realisticly be assumed to hold still. In other words, this causality is a mathematical property of an equilibrium model and not something claimed to exist in the real world.

I, by the way, am not disputing your view about the determination of supply. But, it is just a different model than the neoclassical model.

LB

La economía dominante es epistemologicamente metafísica. No le interesa el realismo de los supuestos y concluye con pretensiones poder predictivas. Pero el mundo es más complejo y puesto que la economía está sujeta a decisiones individuales sin coordinación no puede predecirse el resultado económico. Pero el conocimiento económico que posibilita el marxismo, el keynesianismo y el poskeynesianismo permiten actuar para evitar consecuencias desastrosas para el funcionamiento económico.

Lawrence and Mikael, 

I have for the moment no time to read (and no money to buy) Cartwright's book: Hunting Causes and Using Them. I only peeped at the Amazon's "Look inside" pages. I have also read a review by Kevin D. Hoover (2009), which is openly accessible in WWW. 

I did not know the name of Cartwright and her works. I am happy to know that she came to discuss the proper problems of social sciences and economics in particular. Too many philosophies of science (I mean spontaneous philosophies of scientists, if I employ Louis Althusser's term) have incorrectly intruded into economics because of their relevance to physics. Many of philosophies of science have been generated at the time when the physics dealt only with problems of simplicity. (See for this terminology Warren Wiever's 1948 article Science and Complexity, available on WWW.) Economics has its object a system of organized complexity. It requires very different concepts and experiences in order to discuss or dress the philosophy of economic science. According to Hoover's review, Cartwright comprehends that economy is "a part of a complex world" and "each of the special science is [not] reducible to the more basic sciences--physics forming the bedrocks." We must welcome these philosophers who really want to know what is happening in economics, in stead of applying or forcing their view on economics. 

I am in favor of her defense of causal pluralism. She wrote in Economics and Philosophy, 26(1), March 2010 in defense of herself:

• There are a variety of different kinds of causal systems; methods for discovering causes differ across different kinds of systems as do the inferences that can be made from causal knowledge once discovered. As to causal models, these must have different forms depending on what they are to be used for and on what kinds of systems are under study. 

As for the details of discussions, I was not very much fascinated by topics she has chosen. 

Instead I have discovered something interesting in her co-authored book: Otto Neurath: philosophy between science and politics, by Nancy Cartwright and three others 1996 Cambridge University Press. In the Introduction, they wrote: 

• Knowledge has no foundations. The things we believe can only be checked against other beliefs; nothing is certain; and all is historically conditioned. These are lessons to be learned from Neurath and they are lessons that are commonly taught in Britain and America at this moment. But Neurath's attack on prevailing pictures of scientific knowledge went beyond this. For Neurath not only has science no foundations; it has no cast-iron method either. Falsificationism, confirmation, the hypothetico-deductive method are all pseudo-rational. Like the story of God in his heaven, they set a fairy-tale ideal; and the harm is not just an epishemological one, that the purity of our cognition will be threatened by false beliefs, but a problem of how we make our life in this world.(p.3) 

I thought this remark is interesting. Neurath is a person difficult to understand. The book above may give us a more accurate understanding of Neurath. According to the authors, there are five different uses of the boat metaphor (p.6)

Lawrence, 

Of course, the question is not whether my theory of manger’s supply behavior is right or not. You can imagine a manager who is requested to make a decision on today’s volume of production. He (or she) makes a decision (according to one or other rules).

This is an ordinary scene in any production site. This decision may result in equilibrium but also may raise a series of different volumes of each day production. What the manager decided becomes a fact together with other agents’ actions (in the economy as a whole) and becomes a given situation which will influence their future decisions. There are no harms in thinking that managers’ decisions are a part of causal series of events. It may not be causality as it is presumed in physics, but economics treats human behavior and we can use causality in different meaning from physics or philosophy of  (physical science). Economics is an independent science and cannot and must not be reduced to physics.   

The equilibrium concept may be inconsistent with causal arguments but why is it wrong to analyze time series of events that are always out of equilibrium? The assumption that the economy is in equilibrium is one of basic assumptions that mainstream economists ordinarily suppose but that is the very assumption whose scientific status we must inquire. This is the real problem of theoretical works.

Well, you are going to enjoy my forthcoming book on equilibrium model in economics which will be published by Oxford either late 2016 or early 2017. It discussing the many problems with equilibrium models. Causality is not one of them. You could also take a look at my 1986 book on my web page.

BTW: what does the falsificationism in what you quoted refer to?

LB

 Lawrence,

"falsificationism" is a word which appears in Cartwright's book. As I have  quoted the relevant part, you can discuss it by yourself.

I am in the same position as Neurath in thinking that there is "no cast-iron method" by which to discern or judge which proposition is right or not. Each science has its own "logic" by which to consider these questions but the "logic" itself evolves as the science develops.    

It may be in Carwright's book but that does not tell me what it means?

LB

Lawrence, I have read your paper:

• On the Futility of Criticizing the Neoclassical Maximization Hypothesis

https://www.researchgate.net/publication/2368836_On_the_Futility_of_Criticizing_the_Neoclassical_Maximization_Hypothesis

You have written in one page of your book The Foundations of Economics Method (pp.155-156):

• Methodology is not considered an urgent topic for neoclassical research programs. ...    ...  the subject of methodology is paradoxically considered ... a waste of time.

In the context of the citation, those who thus consider are mainly neoclassical economists, but I as opponent to neoclassical economics also think that your argument is a waste of time. You seem for me a most stubborn medieval philosopher in the scholastic tradition. You have a passion for a minuscule difference of meanings and for the correctness of names of thoughts or methodologies, but you are indifferent of the future or development of the economics. You have no intention to redress the present state of economics. You are only enjoying your discussion. It seems for me that you are close to the philosophers who discussed how many angels can dance on top of a needle.

What does it follow from your soliloquy? Do some neoclassical economists change the attitude in their investigation? Or, does it contribute to orient heterodox economists to advance their research? I do not think so.

In the Wikipedia you appears as an economist who is critical of the neoclassical research program:

• Wikipedia Lawrence A. Boland:
• Boland is critical of the neoclassical research program. He has attempted to draw out the unstated assumptions of neoclassical economics and submit them to methodological scrutiny. His key criticisms of traditional economics center on the problem of induction, methodological individualism, and the acquisition of knowledge.

Maybe you are subjectively critical, but I doubt if you are really critical of the neoclassical research program.

The title of your paper can be repeated exactly against you. You will be able to write a beautiful paper with the title:

• On the Futility of Criticizing the Neoclassical Research Program.

If you do not think so, please show us how you can contribute in any way to the economics in general, including neoclassical, heterodox and newly emerging (non-existent) economics.

Article On the Futility of Criticizing the Neoclassical Maximization...

Hi Yoshinori,

Your suggested alternative title "On the Futility of Criticizing the Neoclassical Research Program" is basically equivalent. There is no N/C economics without maximization.

Dear Lawrence,

This is the point that I want to discuss. Please give me two or three days.

Yoshinori

Dear Lawrence,

the maximization principle that you are interested and referring to is very different for the  maximization that a majority of economists assume. Economists are interested if people's behavior can be described by a maximization principle such as profit maximization and utility maximization under a budget constraint. f you change a part of the problem setting of any maximization formula, the behavior and the theory change drastically.

What matters is the concrete form of maximization: concrete objective functions and constraints. You are talking maximization in general and you are indifferent in the concrete formulation. 

For example, Ian Steedman considered what happens when  the constraint for consumers is not only the budget constraint and time is introduced as constraint into this formula. In this case, the standard non-satiety assumption no more holds, for the time constraint becomes much more important that for high-income people.

If we admit that time constraint is important, we have to reformulate our general equilibrium framework. The first question will be to ask if Arrow-Debreu competitive equilibrium framework still holds in the extended form.

Another example is H. S. Simons  satisfying behavior. As you argued in the   

  On the Futility of Criticizing the Neoclassical Maximization Hypothesis

paper, this can be reformulated as a kind of maximizing behavior if you introduce the computing cost. To find a exact maximum requires (even for a simple utility maximization under a budget constraint) enormous time and computing a more exact solution does not gives us more reward that can compensate  longer calculation. In a suitable case, you can formulate the people's behavior in a maximization formula. You may be satisfied by this new formula, because you are not  interested in the relationship between the real behavior of people and the mathematical formulations.

Sincere economists wonder how it is possible to give a concrete form of newly discovered behavior pattern. A possible form may be a maximization formula but it is also possible that the behavior is well described by some forms of routines. It does not matter if this behavior can be interpreted as a maximization. Such an interest only reveals your narrow philosophy that economics should be constructed from the basis of individuals maximizing  behavior. What you are contending is but an expression of you hidden ideology.

If a routine can be described with a small number of parameters, (whether it can be interpreted as maximization or not) we can advance to investigate what happens in the economy.

You mentioned in your post (post number 23) that 

• There are two concepts that are not used interchangeably: equilibrium in the real world out our windows and equilibrium as a property of formal mathematical models.

in the question page:

Can dissipative structure concept usefully take place of equilibrium framework in economic analysis?https://www.researchgate.net/post/Can_dissipative_structure_concept_usefully_take_place_of_equilibrium_framework_in_economic_analysis

You know to distinguish these two but you are not interested in the relation between them. You are a philosopher of economics and not a scientist in economics. You have no program how to develop economics. You have no concrete problems that you want to solve. You are aloof of all such vagaries and prefer to contemplate if economics science is possible without a notion of maximization in general. It is the pure form of scholastic  speculation.  No sincere economists worry about it. 

That kind of philosophy of science is but an obstruction for the further development of economics. It is only a cynicism. 

Oh my, Yoshi. My philosophy of economics friends are all laughing at your claim that I am a philosopher of economics. In my education I took one undergraduate philosophy class (ethics) and one graduate seminar by Joseph Agassi. I see myself as first an economics teacher and second a critic of economic model building. BTW: what is an economic scientist?

I obviously failed to communicate correctly in my 1981 AER article. My point was simple: if one builds a neoclassical model to explain or predict something and it turns out that the prediction is false, the last thing that a N/C economist will blame is the maximization assumption.

Dear Lawrence,

is the article that I have read a revised version of AER 1981?  In the document information, it was indicated AMERICAN ECONOMIC REVIEW 71(5) · AUGUST 2001.

If this is such an old paper, I will abandon to discuss the contents of the paper. You may have changed your opinion too.

You state that    

• My point was simple: if one builds a neoclassical model to explain or predict something and it turns out that the prediction is false, the last thing that a N/C economist will blame is the maximization assumption.

Does this mean that you prefer to continue thinking in the neoclassical framework? I agree with you on the point that there are many bad aspects in the standard neoclassical economics. If you want to conserve the neoclassical framework as far as possible, you are right to say that maximization is the last assumption to criticize and in such a  research program, you are reasonable.

I have a totally different position vis-à-vis neoclassical framework. My understanding is that economics needs fundamental change of the theoretical framework, from neoclassical economics to some other economics. That economics has not yet a concrete form as a textbook like Marshall's  Principles but  we have already some core rudiments: evolutionary point of view and classical theory of values. As an analytical tool, we are expecting that agent-based simulations may contribute to our economics as mathematics contributed to neoclassical economics.

For the scope and implications of evolutionary view point, please see my Manifest:

• Evolutionary Economics in the 21st Century: A Manifesto

As for classical theory of values, see my paper:

• The Revival of Classical Theory of Values.

I admit that classical theory of values (i.e. cost of production theory of value) has its grave defects: it lacked theory of international values. However, a new theory was constructed as an extension to the classical theory of value. Please note that classical theory of value is now renovated and has become as powerful as neoclassical theory of prices. Combined with the market process framework, the new economics is much more realistic than neoclassical economics.   

Article Evolutionary Economics in the 21st Century: A Manifesto

Conference Paper The Revival of Classical Theory of Values

Yoshi, what is the AER 2001 article?

Lawrence,

the paper with the title 

On the Futility of Criticizing the Neoclassical Maximization Hypothesis

in the RG has such an inscription.  

I haven't contributed anything to this thread for a long time - I have simply been overwhelmed by other matters. i did read the things that I was directed to by Lawrence Boland and have various reactions that I hope that I'll have time to contribute before to long. Meanwhile, the discussion has continued spasmodically. In any case, I just ran across the following, which seems to me to give at least a partial response to  Yoshinori Shiozawa's original question, if anyone can remember that far back. http://www.counterpunch.org/2016/03/23/junk-economics-and-the-parasites-of-global-finance/print/

Dear Yoshinori Shiozawa and Colleagues, Here's something that perhaps throws some light on the discussion:  https://www.youtube.com/watch?v=szIGZVrSAyc  Best wishes, Mikael

Porque Milton Friedman representa el pensamiento conservador en economía. y los economistas, por falta de ética intelectual, no son críticos.

I have posed a new question:

Is econometrics a cherry-picking?

https://www.researchgate.net/post/Is_econometrics_a_cherry-picking?_tpcectx=profile_questions

This will be interesting for many followers of this question page. Our discussion changed the economics landscape very little. Even though, the state of economics is very bad and we should argue continuously our method or strategy of research as well as research works themselves.

Economics needs radical change and, for that purpose, it needs radical change of methodology and philosophy.  

I agree with this last statement. The stste of our current knowledge in the subject as very confused. However, my new book does allow us to view macroeconomics in a different and better way. It is included here on LINKIN references, so I don't need to send it to you. But what about microeconomics? Should this also be improved?

I found an argument against Friedman's methodology.

George H. Blackford

Stop Defending Milton Friedman’s Pseudo-science

On the pseudo-scientific nature of Friedman’s “as if” methodology

http://evonomics.com/economists-stop-defending-milton-friedmans-pseudo-science/

It contains not much economic arguments. It is based on a simple fact  from history of science: how Galileo Galilei found the law of falling bodies.  It elucidates how and why Friedman's contention is flawed and wrong. 

There are many criticism against Friedman's economic and political ideology.  However, it is indispensable to criticize his methodology. If not, we economists have a big chance to be unconsciously caught up in the way of thinking that is prevalent in economic discourses. At a time when we are required to create a new thinking, this is absolutely necessary  thing. 

The main problems is that many model builders today do not realize they are practicing Friedman's instrumentalism whenever they have no regard for the realism of their assumptions.

 Yes, Lawrence, that is the main problem. 

More than sixty years have passed since his book Essays on Positive Economics was published. It is now time to place his methodology in a right place. His methodology is so outdated but, as Lawrence Boland observes, many economists do not realize that they are practicing Friedman’s instrumentalism and his positivism.

Why did this state of affairs continue for so long period? One easy answer is that Friedman represents economists’ “spontaneous philosophy.” (Cf. L. Althusser, Philosophy and the Spontaneous Philosophy of Scientists) In order to escape from deeply rooted spontaneous philosophy, it is necessary to know the history of philosophy of science.

After the publication of Essays of Positive Economics, many important works in the philosophy of science(s) appeared. Friedman’s Essays methodology is based on naïve verificationism and falsificationism. In his mind, there are no Duhem-Quine thesis, no theory-ladenness, no scientific revolution, no methodological anarchism, no revolt against positivism. In sum, Friedman’s Essays is an expression of naïve philosophy before the "historical turn in philosophy of science."

Among many writers in philosophy of science, Thomas Kuhn must be the most notable person. In 1960’s his book Structure of Scientific Revolution was read by every sincere economists but he was forgotten. His name is retained by many of them but Kuhn’s most important teaching was forgotten by most of mainstream economists. This is a natural reaction or reflection, because they believe that economic science is in a right line and developing without experiencing major anomalies. Financial crisis of 2007-08 was an episode of big deviation. Bad performance of quantitative easing and negative interest rates does not teach them anything. They are just like French nobles after the decline of Napoleon. They had learned nothing and forgotten nothing.

Es que la mayoría de economistas se plegaron, sin esfuerzo intelectual, al individualismo metodológico. Les da pereza pensar y estudiar otras alternativas teóricas.

With respect to all of you in this thread, I appreciate what Boland claims,

"The main problems is (sic) that many model builders today do not realize they are practicing Friedman's instrumentalism whenever they have no regard for the realism of their assumptions."  

but I do wonder if all of the methodological and philosophical problems of and in economics are solved by appealing to realism in assumptions? Not that Boland claimed they would be, but the implication seems floating about in some of the responses.

What I see in much of this extended--and fascinating conversation--is that as soon as someone can find a reason to criticize Instrumentalists without placing users and the approach in the context of the problem-situation they're trying to resolve, then 'the' alternative seems to be clear: realism in model-building is lacking so we have to stat there.

When is an assumption realistic enough? There is probably some 'rule' or convention to help us determine that? My point being, it's the same problem all over again.

Just as Friedman tried to bypass the Gordian knot of realism-in-assumptions, replacing it with some equally difficult to measure concept like 'predictive success', it's hard to argue in reverse that starting from realism in assumptions is a 'better' (for whom and why) approach. It is problem-dependent.

'Predictive success' sounds appealing, but I agree it is also very questionable. What is a successful prediction? When is a theory judged to be predictively successful? Nine out of the last ten predictions? As Prof. Shiozawa suggests, the one failure might be a spectacular failure. 

Look to see what problems economists think they are solving and for whom. Economists do not work in a vacuum. Much of their research is driven by trying to solve problems in real-time and in time short of infinity. Therefore shortcuts i.e. conventions, are used. Much of their research is funded by government and NGO agencies wanting some 'results', the sooner the better.  No surprise, then, that methodological shortcuts (like instrumentalism) are used.

This is not a defense. It is a suggestion as to why a sizable portion of economists appear to be 'Friedmanites' (when they probably have not read any of Friedman's work). I do not believe the apparent success of Instrumentalism can any longer be attributed to Friedman's personality or influence in the profession. When the last few of us die off, there will be no one left to call people 'Friedmanites', but the same instrumentalism will be practiced because the problems and funders will be the same.

In the Kuhnian paradigm: there appears to be no "new school" of non-instrumentalist empirical economics (of which I am aware) that will push out the old school instrumentalism. It is its own argument. (This, of course,echoes Boland in 1979).

I have been fortunate enough to read Boland's forthcoming book in galleys. The right readers would help many others understand the foundational issues and problems. I am very sympathetic to the arguments all of you make--but the problem is not to convince each other, but to get others to notice the problem and to propose better performing admittedly 'instrumentalist' models. 

Most economists see themselves as too busy Doing Economics (i.e. producing quantitative, empirical results, e.g. maybe a 'better estimate of a multiplier value) to worry about the methodological niceties. Funding agencies of all sorts encourage this short-run approach. 

We either have to change what Doing Economics means (which is, I believe, Boland's lifetime attempt) or we have to change what research gets funded. Otherwise, I predict, the same laments will be being made by the always small band of methodologists hanging on to the fringe of the profession.

Speaking as if I were a practicing 'Doing Economics' person: My challenge: you don't like so-called Friedman 'instrumentalism', with what will you replace it? How long will that take? Will the predictive results be observably better? Compelling answers may get someone a nice NSF (US) or SSRC( Canada) grant!

Dear David L Hammes,

Welcome to our question page. You have honored us by posting your first answer in RG to this page.

Instrumentalism is one way to persuade the public to give support  to a branch of sciences. It may be appealing to practitioners  (including 'Doing Economics' people). However, I do not believe that it is a good methodology. It has, it seems to me, various bad side-effects that have influenced the Doing Economics since many years ago.

You have asked an alternative, something to Friedmanian 'instrumentalism'. Unfortunately, there will be no good method that is given from exterior of our science (i.e. economics). In this sense, I am for Feyerabend 's methodological anarchism. 

Innovation of economics starts from criticizing the economics we actually have. The vice of Friedmanian 'instrumentalism' lies in the fact that it abandons all criticisms related to "assumptions." Any theory contains a core system of assumptions (or a kind of axioms). When we argue neoclassical economics  or any other economic theory, criticism must be logical, empirical and synthetic. In this sense I believe Friedmanian 'instrumentalism'  is now an epistemological obstruction for the development of economics.  

Please read my last post (#31, the top answer in page 4) in the following question page:

Why microeconomic incentives do not always work?

https://www.researchgate.net/post/Why_microeconomic_incentives_do_not_always_work?pli=1&loginT=VDwL0Ke3DTxW794G0t_gNBYm3C2-OvRj&uid=m4UN0BO4QkaWOBSE1b9D70fYmrfFPuFl8VWh&cp=re68_x_p3&ch=reg&utm_medium=email&utm_source=researchgate&utm_campaign=re68&utm_term=re68_x&utm_content=re68_x_p3#view=584f57df217e20dfcf6bf814

As I explained in the Epilogue of my1997 book (p. 283), econometricians all follow Friedman's instrumentalism. Even those who ideologically disagree with Friedman.

Boland points the most serious problem of Friedman's methodology.

In recent years, examinations of deeper cause of economic development are going. However, the authors are simply questioning if they can get a good fit for variables they think deep cause. There is no study of development mechanisms.

In my understanding, the income level of a nation cannot be raised unless its potential set of production techniques are improved (in a simpler term productivity, see for example my paper on The New Theory of International Values). If geography, law enforcement, and others are related to economic development, they should first prove how these deep structure affects the change of production techniques.     

Let me explain my previous post a bit more in detail.

It was Rodrik and Others (2004) which started the deeper determinants of economic growth. After reviewing that there were three strands of thoughts: (1) geography, (2) integration (into the world market), and (3) institutions, they claimed that

• Growth theory has traditionally focused on physical and human capital accumulation, and, in its endogenous growth variant, on technological change. But accumulation and technological change are at best proximate causes of economic growth. No sooner have we ascertained the impact of these two on growth—and with some luck their respective roles also—that we want to ask: But why did some societies manage to accumulate and innovate more rapidly than others? The three-fold classification offered above—geography, integration, and institutions—allows us to organize our thoughts on the “deeper” determinants of economic growth.

I agree with Rodrik and others that physical and human capital accumulation was is misguiding in understanding economic development. It is necessary to investigate how economic growth occurs and how it is impeded. However, econometricians started to check which of three causes is more influential to the economic growth only by cross-section and panel data analysis. See for example, Osang (2006), Jacob and Osang (2008), Owen and others (2009), and Perryand Schönerwald (2009). This theme became very popular and it is now a theme of master thesis (Lauridsen, 2016). Hasan (2007) is in a sense a survey in this theme.

The search of deeper determinants by a regression or others reminds me a Japanese dictum. We say

•  When the wind blows, pail-makers make money.

This dictum does not tell a truth. It is rather a tacit aphorism. In order to understand the real implication of this saying, we should know why it is said that pail-makers make money. At the back of this saying, this chain of causation is implied:

• When the wind blows, people become blind. When blinds increase, more people learn Shamisen (see the note below). Cats become rarer and rats increase. More rats bite pails and the pail-makers can sell more.

The dictum implies that a too long chain of causation is nonsense and ridiculous. Search for deeper determinants is important, but the chain of causation is very long. If we find a deeper determinant of economic growth, we should examine how the causation works. If not, it is a research just like "when wind blows, pail-makers make money."

Cross-section or panel data analyses of deeper determinants are just like the dictum. Friedmanian methodology justifies this kind of ridiculousness.  

Note: Shaimsen is a Japanese string instrument similar to a Chinese fiddle with a resonator spanned by cat's skin. Singer with shamisen was a popular trade for blinds in Edo period (1603-1867).

Referenecs:

Hasan, L. (2007) Myths and realities of long-run developments: a look at deeper determinants, Pakistan Development Review 46(1): 19-44.

Jacob, J. and T. Osang (2008) Institutions, geography and trade: a panel data study, available at https://www.researchgate.net/profile/Thomas_Osang/publication/24130961_Institutions_Geography_and_Trade_A_Panel_Data_Study/links/0912f5128041616e4f000000.pdf.

Lauridsen, H. L. (2016) Geographical Determinants of Long Run Economic Growth, a master thesis

Osang, T. (2006) External and internal determinants of development, Federal Reserve Bank of Dallas Proceedings, 35-59.

Perry, N., and Schönerwald, C. (2009) Institutions, geography, and terms of trade in Latin America: A longitudinal econometric analysis (No. 2009-04). Working Paper, University of Utah, Department of Economics.

Rodrik, D,  A. Subramanian, and F. Trebbi (2004) Institutions Rule: The Primacy of Instittions Over Geography and Integration in Economics Development. Journal of Economic Growth 9(2): 131-165.

Article Institutions, Geography and Trade: A Panel Data Study

The production rates of a nation can be raised without any change to its money rates of distribution, simply by making its cost of production less. The cost of production depends on the rent being paid by tenants for access rights to land. But the holding of useful but unused land out of use for purposes of speculation in its rising value, is the cause of present high production costs. By taxing land values instead of produce and its associated money-flows (income and purchase taxes, etc) the amount of competition for access rights to land will reduce and more produce will develop as its costs become reduced and its demands be better satisfied.

Here is some relevant input, although only obliquely related to Friedman specifically:

https://www.youtube.com/watch?v=szIGZVrSAyc

https://www.youtube.com/watch?v=ZM0_7PVuVVg

Cheers,

Mikael

Pls. have a look at the following insightful book by Uskali Maki: The Methodology of Positive Economics: Reflections on the Milton Friedman Legacy:

http://www.cambridge.org/gb/academic/subjects/economics/history-economic-thought-and-methodology/methodology-positive-economics-reflections-milton-friedman-legacy?format=PB&isbn=9780521686860#EW4XQsB1Dd6ETVXO.97 

where a non-instrumentalist interpretation of F53 is offered, i.e., Maki claims that F53 has a realist spirit.

And when you are done, read my review of that book.

If you are interested, my review can be found at http://www.sfu.ca/~boland/.

After reading Larry Boland's review, you can also read mine in The Philosophy of the Social Sciences, Vol.41, No.3:

Abstract

Milton Friedman’s 1953 essay created controversy and consternation amongst economists. It provided a prescription, based on empirically generated predictive success, of how to do economics, yet many saw it as a concession of the search for truth and theoretical beauty within the discipline. This article reviews a 50th anniversary festschrift devoted to views of the essay. The purpose of the volume is to provide today’s reader with the essay, responses, and a guide to interpreting it. The volume is selective and several contributors have their own agendas, but the feeling of tumult the essay still engenders is nicely conveyed.

Thank you for drawing my attention to these two reviews. 

So-called "Popperian falsificationism" is an invention of Imre Lakatos and which Popper denies is his view. Friedman did not understand Popper. As to DSGE models, I suggest you read my recent book: Equilibrium Models in Economics: Purposes and Critical Limitations (New York: Oxford).

Thank you again. And I can faintly recommend you my new forthcoming book on Economics without Laws, which does not discuss Friedman but which at least tries to “interpret” economics as a humble science (or as science which needs to be humble) where the-good-old-debates between realism and instrumentalism seem to be a bit passe and thus new insights from phil. of science are necessary.

http://www.springer.com/it/book/9783319548609