Thursday, August 30, 2012

Economics: A Gaping Separation Between Theory and Operations

commentary by Roger Erickson



I've posted a longer essay on this topic, at Global Economic Intersection.

Since the two sites attract different readers & comments, the link is also posted here.

78 comments:

Dan Kervick said...

Completely off topic, but this just broke my heart:

http://www.kickstarter.com/projects/1158154396/american-winter-a-documentary-film?ref=card

I'm completely fed up with the incompetence and corruption and selfishness of our political leaders and the elites they protect.

Tom Hickey said...

Excellent, Roger. Shows why mainstream economics is not even wrong. It's absurd other than as an exercise in model-building.

marris said...

Come on Tom, have you read this essay?

An apt criticism is that it is *not even an exercise in model-building.* Most professional BS artist would be embarrassed to publish it.

I'm sure Roger's a nice guy, but he needs to put aside the holism Kool-Aid once in a while and spend more time thinking through his ideas on a clearer conceptual level. What question does he want to answer? What concepts are relevant to the question? For each, increasing larger set of concepts, how does the set of possible answers change?

Everytime time I think Roger is about to say something useful, he disappoints. For example, he mentions Automatic Stabilizers as being *really* important in complex systems.

OK, let's expand on that. What are we trying to stabilize (nominal values, "real" values)? How can we stabilize it?

The world central bank consensus on what stabilizers to use has been (1) unemployment benefits, and (2) inflation targeting by the CB.

Many people are currently unhappy with these stabilizers, and they think we need new ones. MMT offers one (level targeting of cash + outstanding government bonds).

Roger's essay, albeit really long and chock full of buzzwords, does not address these questions, present an answer, or even try to list the relevant ideas.

And it certainly does not present an adequate alternative to "neo-classical" economics or "equilibrium" analysis.

paul said...

Ah, marris, marris, marris…

Another poor soul that doesn't know what he doesn't know, unaware of the tyranny of system relationships.

Of course everything Roger writes is obscure to you.

Makes good sense to those that are aware and understand the binding of system relationships.

Theres a reason the Universe doesn't careen out of control every time a disruptive event takes place.

Matt Franko said...

Marris,

I dont think Roger is trying to "replace" the economists with an "alternative" academe... he is a biologist the way I look at Roger... and it looks to me like he is trying to explain what is going on to people outside the academe of economics....

Just speaking for myself (non-economist, non-semantic) when Roger says "automatic stabilizers are important in systems" those few words mean A LOT to me... my brain starts to think of many things and system parameters, etc..

I DO NOT NEED ANY MORE WORDS... I GOT IT RIGHT THERE... MORE WORDS ARE A WASTE OF OXYGEN TO ME...

I'm trying to communicate to you how non-semantics brains function so to speak...

At the bottom of Rogers post he mentions a book titled "The Tyranny of Words"

http://books.google.com/books?id=pTXfGAAACAAJ&dq=stuart+chase&source=bl&ots=FvgAzNrHup&sig=DljoCirU2ewI9eaRPTsD4x-yJAY&hl=en&sa=X&ei=yH4_UJXYAaj_6QGFqYHIBQ&ved=0CC8Q6AEwAA

This is contrasted with what Paul comments above and Ive seen Bill Mitchell coin "The tyranny of the Math"...

Many of us live under the tyranny of the math. I for instance do. I think Roger may also.

Many others do not. So I dont know if it is reasonable to expect someone who is under math to be ever able to meet the expectations of someone who lives under words...

Resp,

Tom Hickey said...

Look up "homeostasis." It's about automatic stabilization in an organism.

Neoclassical economics is based on physics (mechanical model). Systems theory is based on biology (organic model). Mechanical don't grow. Biological systems do. Mechanical systems do not exhibit complexity. Biological systems do.

The neoclassical assumption that an economy can be modeled as a mechanical system is wrong. The appropriate modeling is biological. The difference here is the difference between structure and function, statics and dynamics, maintenance and operations. It's the difference between an 18 and 19th century approach and a 20 and 21st century approach.

paul said...

Matt,

"when Roger says "automatic stabilizers are important in systems" those few words mean A LOT to me... my brain starts to think of many things and system parameters, etc."

That's about as succinct a description of the differences between the two groups as I have seen.

The semantic/abstract divide is probably responsible for most of the arguments that take place in our arena.

Until you made me aware of this simple thing I spent a lot of time wondering why my arguments seemed to be completely missing the target with some people (including my wife).

There are those of us that generally think in the abstract and must translate our thoughts into a complex language to communicate, and those that apparently see things in a more superficial way and so miss embedded relationships hidden in plain sight. This cohort of people may be classed as "memorizers" rather than "understanders" of patterns.

"money", "saving", "borrowing", "debt" are all abstract constructs the meaning of which depends on context.

If the context is based on semantics the argument becomes obscure.



In other words it's as simple as what Matt said.

Tom Hickey said...

Words and numbers are complimentary. Words are more appropriate for quality and numbers for quality. They are different tools for thinking and they complement each other. The problem is that people tend to be divided along these lines instead, when what is required is an integrated appraoch. In turn, this requires the ability to know what tool to use when and how to to use it skillfully to get the job at hand done, i.e, in context.

paul said...

Tom,

What you say is true except that re-arranging words can't change how a system behaves, only the perception of how it behaves.

How words are arranged when describing an abstract process can never be more than a cartoon or characature of the real thing.

So in the context of understanding real-world systems, words alone are inadequate.

paul said...

Tom.

My bad, I see that's basically what you just wrote.

The only thing I can add is that there is a cohort that doesn't "grok" the numbers part, at least not the underlying relationships.

Matt Franko said...

Tom,

imo You and Beowulf are kind of enigmas as you both seem to be able to "go both ways" (no pun intended ;)

That is something you may want to keep in mind when you are dealing with lesser mortals ;)

Paul,

At the risk of going "Larry Summers" here....

I once read something from a child behavioral scientist that IIRC: "little boys are about objects moving in 3D space, and little girls are about communicating"...

Dont know if this is sexist???

Rsp,

Tom Hickey said...

I once read something from a child behavioral scientist that IIRC: "little boys are about objects moving in 3D space, and little girls are about communicating"...

Similar to the idea I just expressed about quantity and quality, words and numbers being complimentary. One can look at as "the battle of the sexes," or else as yin-yang complimentarily within tao, which is the basis of energy (chi, qi, ki) flow. Life in the relative sphere is about that flow and being in harmony with it. For humans this requires cultivation due to the accumulation of stresses that disrupt the flow. Societies and their economies behave similarly.

marris said...

> Another poor soul that doesn't know what he doesn't know, unaware of the tyranny of system relationships...Theres a reason the Universe doesn't careen out of control every time a disruptive event takes place.

paul, I don't think your posts here add anything to Roger's already confused essay. System relationships? Why not focus on expressing one idea that you clearly understand, rather than hand-waving over multiple ideas that you many not?

I'm not sure what "universe careening out of control" means. We didn't cover that in my physics class. Your "analysis" is probably even less appropriate here, since we're discussing how a system (the current economy) is *not returning* to a point that we want to reach. The whole discussion is about pushing the economy, whether by discretionary acts or rules.

There is no need to hide behind hand-waving.

marris said...

> Neoclassical economics is based on physics (mechanical model). Systems theory is based on biology (organic model). Mechanical don't grow. Biological systems do. Mechanical systems do not exhibit complexity. Biological systems do...

I disagree with most of these statements. Mechanical and biological systems are superficially very different, but they are more similar then different. Most the "complexity" that you see in bio-systems is the result of evolution's billion year head start on mechanical systems.

Further, 21st century biology is not about homeostatis and hand-wavy cybernetics research. It's about *real* bioengineering, biomechanical, drug delivery, etc. That is the cutting edge. Our increased understanding of bio has come from a very mechanistic, often reductionist view. We now try to study the interaction of proteins on physical terms (often with a very "mechanical" computer to help us simulate what happens). On the other end of the spectrum, we have the application of statistical techniques to large data sets.

As far as I know, the only guys interested in modeling their systems at the level of a biosphere are ecologist and zookeepers. And I wish them luck. The rest of the world wants mechanical formulas when they can get them, decent intuitive heuristics when they can't.

paul said...

marris:

Sorry you see my comments as "hand-waving". Most of us hanging out here understand the concepts we are basing our arguments on and none of us think it is particularly complicated or beyond the comprehension of others.

More likely we are somewhat surprised that more people don't see these relationships.

Your criticisms are just as vague to me as my "hand-waving" comment is to you.

Why don't you try and be more succint with your criticism or questions so that we can find some common ground?

Or would you rather just dismiss outright what we think we see that you don't?

We haven't been at all secretive about how we think the economic system is arranged. Seems like you could come up with some pretty pointed questions if you wanted to.

paul said...

"…Mechanical and biological systems are superficially very different, but they are more similar then different"

This statement rings generally true to me, thermodynamics and hydraulics are part of the mechanical engineering curriculum. I don'y think the point Tom was making is in general contrary to this reality.

Mechanical (machinery) systems in general don't replicate themselves. Robotics may change this.

Don't you think the logic behind thermodynamics is applicable to an economic system?

paul said...

"we're discussing how a system (the current economy) is *not returning* to a point that we want to reach"

Why should the economy "return" to some point we want it to be on it's own? We have control (so to speak) of the input. The economy doesn't have any "natural" equilibrium. It goes where we push it (through spending).

Tom Hickey said...

marris: I disagree with most of these statements. Mechanical and biological systems are superficially very different, but they are more similar then different. Most the "complexity" that you see in bio-systems is the result of evolution's billion year head start on mechanical systems.

The basis of 19th century scientific paradigm of "unified science" building from the bottom up. Understanding of complexity and emergence killed it.

Matt Franko said...

I think our "feedback mechanism" is for instance viewing the video that Dan posted to at the top of this thread... rsp,

marris said...

> Don't you think the logic behind thermodynamics is applicable to an economic system?

Not really. Thermodynamics studies particles that basically bounce around. You don't need a choice theory or any telelogical analysis isn't really necessary to study them. You do with people. There's no real evidence that any thermodynamic rule has proven useful in understandng economic phenomena, although it is always possible to dump economic data into a table and do the same mathematical operations you would do on thermo system data. I don't think you vet anything that econometrics already doesn't have.

I think you misunderstand my criticism of returning to equilibrium. I'm not saying the economy does or does not return. I'm saying that you *comment* about the universe not careening is inappropriate here. We are talking about a situation where the economy *did careen* from a high employment state to a low employment one. Talking about the universe *not careening* on every shock is not helpful.

Since you claim to understand systems, can you point out the most complicated system that you ever built.and what principles you needed to build it? Or maybe give an example of the last time you applied one of these principles to a system and walked away with some unique insight?

marris said...

> The basis of 19th century scientific paradigm of "unified science" building from the bottom up. Understanding of complexity and emergence killed it.

Tom, by emergence do you mean the development of high-level simplicity from low-level complexity?

That is how most scientists understand it. For example, you may have some organism made of lots of particles. Each of those particles follow the laws of physics, but it is impractical to make predictions about the organism from this level of detail. The number of calculations required are too many, the interactions are chaotic (too dependent on the very precise measurement of the initial state), etc.

Rather than trying to understand the organism through particle physics, we study it on a more coarse level. We study the digestive system as a multi-stage pipeline for processing food. We study the impact of high blood pressure on it's arteries by mechanical engineering approximations of the blood cells' shear stress on the vessels, etc. We study the impact of diseases by simple models of virus interaction with cell walls, etc.

The other way to think about emergence is as a barrier to current understanding. A very rich set of chess games can arise from a simple board configuration and simple rules. But that just makes the problem more fun to study! Some work has been done on understanding how exactly today's computer programs beat chess masters (where exactly the masters go wrong). It's not that the masters fail to see some of the configurations that the computer sees (the computer search depth is not that deep yet). It's that the master's evaluation of some configurations (his emergent science tools) are poor. He cannot see the value of a particular configuration, even though the computer's rules can compute it.

Tom Hickey said...

Chaos Theory is the study of fascinating yet deterministic [ergodic] systems, whereas complex systems are not deterministic [ergodic]....

Dynamic networks – or complex systems – are very different to the systems studied in Chaos Theory.  They contain a number of constituent parts (“agents”) that interact with and adapt to each other over time.  Perhaps the most important feature of complex systems, which is a key differentiator from chaotic systems, is the concept of emergence.  Emergence “breaks” the idea of determinism because it means the outcome of some interaction can be inherently unpredictable.  In large systems, macro features often emerge that cannot be traced back to any particular event or agent....

A useful way of distinguishing between chaotic and complex systems is to illustrate how uncertainty arises in each type of system.  In chaotic systems, uncertainty is due to the practical inability to know the initial conditions of a system.  If the initial conditions were either ‘x’ or ‘a tiny (immeasurable) variation on x’, we know that when non-linearity kicks in, this slight variation will eventually make a large difference in the outcome.  But in complex systems, uncertainty is inherent in the system because of the concept of emergence: even if we could measure the initial conditions today to an infinite degree of precision (for the serious geeks: I am ignoring Heisenberg’s uncertainty principle here), we still cannot determine the future.  However, as I noted in my article Patterns Amid Complexity, this does not mean the future is random because patterns are an important – and often repeating – feature of complex systems....

To conclude, it is tempting to believe that Chaos is a highly complex type of complex system.  But it isn’t – Chaos theorists have revealed some excellent insights, like sensitivity to initial conditions, but Chaos Theory is still a study of deterministic systems.  To understand non-deterministic systems, like social systems, it’s necessary to look at complex systems and Complexity theory.


Chaos Versus Complexity - Greg Fisher

Roger Erickson said...

MARRIS, please start here: "For every insurmountable task, there is a solution, and that solution will involve another level of indirection."

To find truly novel solutions, we have to start looking at many more indirect paths. That means looking at many other model systems, many of which have been nearly studied to death ... but not brought in to expand the thinking of economists.

See my comments here.
http://mikenormaneconomics.blogspot.com/2012/08/houston-we-have-distribution-problem.html

RE: the dichotomy of how males/females think (i.e., spatially distant vs very intimate realtionships) ... that's a valid generalization, but one that barely scratches the surface of the resilient diversity in a given human population. Bottom line is that, for all traits & tendencies, there's far more diversity within each category than there is between the median of the two supposed forms. (Try telling that to the Westboro Church Luddites.)

Same thing holds for the word/math/art/intuitive addicts. A fair bit of philosophy is dedicated to the residual problem humans have in UTILIZING INEXPRESSIBLE THOUGHTS, especially when they're valid or important concepts or recognized patterns. That's what art, poetry etc are about. Zen focuses on many of those limits of expression, while "western" philosophy leans more to eventually finding a way to crudely verbalize and name new concepts. In the English language, our semantics is always a mess - because that's the price we pay for continually re-defining some of our modern languages into a lean subset able to fluidly migrate between different situations. Compared to other languages, English is almost a pidgin language, with very loose & sloppy rules about clarity. English is agile exactly because it DOES vary so much with context, which we tend to forget, and hence get personally sloppy with. English CAN be very quick & efficient, but ONLY after discussants "get on the same page."

RE: Homeostasis, that's a very short-term concept in most system sciences, yet considered long-term in economics. It's a semantic problem between professional fields. Shewhart's phrase was that data is meaningless w/o context, and without enough interaction across fields, a LOT of data is thrown about uselessly. Even in this day of the internet, the fastest way for Marris & I to see eye to eye would be at least a phone conversation, if not a face to face discussion. We still have no communication form that fills the full bandwidth of face to face interactions. That in itself summarizes the problem of keeping scaling populations fully organized, let alone increasing their agility. [What happens in physiologies is that different cell types don't need to be in full communication, but they do need to rely upon what range of behaviors each other will stay within - tolerance limits.]

RE: "If the context is [dependent upon sloppy] semantics the argument becomes obscure."
That's about the best summary of our situation I've seen all year.

Roger Erickson said...

RE: looking for simple answers; Marris, when we don't yet know the answers, there's a std process to follow;
1) cast a wider net & re-sample the context being explored;
2) re-examine all the new inter-dependencies;
3) test many responses to trigger new info about outside & internal outcomes;
If you want simple suggestions, make one yourself, or retreat into religion?
Our biggest problem is that most people want simple, immediate answers, and will do anything in their power to avoid the hard work of actually thinking.

RE: mechanistic reduction of scaling problems;
Ok Marris, now I think I see where you're coming from. Solution is to classify the range of problems we have. Some are local & tactical problems (e.g., individual disease) which are amenable to reduction (e.g., pharmacology, surgery, etc). Some are very diverse problems more in the realm of strategy, policy or even national goals (ecology, national policy [energy policy], etc). Distributed policy problems, like many business tasks, can't be subjected to adequate experiments - and we must constantly try things based on insufficient data. Those experiments fall under the same category as combinatorial experiments, where you throw a lot of things together, and then very carefully sample & fractionate all the diverse results. Combinatorial experimentation requires sophisticated instrumentation, to quickly analyze the diversity of results. When it comes to economics & policy the fact is that we simply don't have adequate "self instrumentation."

Basically, my article was about a year or two of sampling economics literature, and finally recognizing a pattern - that one of their basic premises (equilibrium, or maintaining homeostasis) is only one of the short/medium/long term paradigms that must be fused to model & manage any adaptive system. I had to wade through a lot of semantics to finally realize the obvious. Why most people don't learn that during economics 101 is the next question. At the bottom of the rabbit hole, I blame the entire field for becoming too isolated from all the other fields of study that it's supposed to serve as a community member in. The startling gap between theory & operations in their OWN field (economics/banking) is bad enough, but the gap between economics & many other fields is even worse. If Economists only knew a fraction of what American's know ... the fields of orthodox economics wouldn't be so useless.

At the bottom of that next, nested rabbit hole, I blame our entire population (& all NAIC codes) & education system for letting us become a nation of isolated specialists, with declining situational awareness, parsing skills and coordination skills.

marris said...

> But in complex systems, uncertainty is inherent in the system because of the concept of emergence: even if we could measure the initial conditions today to an infinite degree of precision (for the serious geeks: I am ignoring Heisenberg’s uncertainty principle here), we still cannot determine the future.

Hmm... I would say that this person does not know what he's talking about.

Let me ask an alternate question. Is the "complex system" described here substrate independent? If we had the measuring tools, could we, in principle, copy the complicated state into a computer?

If so, then it's easy to "predict" behavior... just run the thing and see what happens. You could even tweak the params and rerun.

If the system is not substrate independent, then we're basically talking about chaotic difficulties again.

paul said...

"Not really. Thermodynamics studies particles that basically bounce around"

Hardly. Thermodynamics employs evaluation of systems in the context of being "closed" or "open" or "simple" or "isolated" among other things such as conservation of energy and the impossibilty of perpetual motion. The energy relationship is not necessarily relevant to the discussion but some of the others are.

Once you identify a particular system as closed the tyranny of the arithmetic is established re the quantity of net finacial assets and the universe of possibilities is narrowed considerably.

You don't think that an analysis of an economy as a simple closed system of units of spending is relevant?

"We are talking about a situation where the economy *did careen* from a high employment state to a low employment one. Talking about the universe *not careening* on every shock is not helpful".

The economy suffered a disruptive event and has since come to a new equilibrium, albeit one with high unemployment and much sufferiing. it has come to this particular equilibrium state due largely to the "automatic stabilizers" or feedback you seem to object to. Another disruptive event will send the economy to a new equilibrium, the outcome of which is not mysterious.

Likewise, when a disruptive event occurs in the Universe feedback in the system brings it to a new equilibrium, a position of stability, otherwise events could cascade into a massive disruption.

"Since you claim to understand systems…complicated…"

I haven't been claiming economics is complicated…just the opposite, The underlying system is incredibly simple. So simple that I can't really explain it in words, kind of like it's very difficult to prove that 2+2=4 always.

Start with spending=income and the sectoral balances identity and you pretty much have everything you need to manage the system within the goals of price stability and full employment.

Combine that with closed system laws re the flow of funds and you have the overarching dynamic of the domestic economy. Comes down to binary choices.

Given a system with a distribution of funds and the most likely direction of flow overall, what will it take to keep the flow going at a constant or gradually increasing rate?

Too fast or too slow tend to lead to undesirable equilibriums.

Tom Hickey said...

The basics

Complex system - Wikipedia

Tom Hickey said...

More in depth

Complex systems — wikipedia

The two entries rather confusingly have the same name. This one is more comprehensive and includes economics.

The earliest precursor to modern complex systems theory can be found in the classical political economy of the Scottish Enlightenment, later developed by theAustrian school of economics, which says that order in market systems is spontaneous (or emergent) in that it is the result of human action, but not the execution of any human design.[5][6]
Upon this the Austrian school developed from the 19th to the early 20th century the economic calculation problem, along with the concept of dispersed knowledge, which were to fuel debates against the then-dominant Keynesian economics. This debate would notably lead economists, politicians and other parties to explore the question of computational complexity.
A pioneer in the field, and inspired by Karl Popper's and Warren Weaver's works, Nobel prize economist and philosopher Friedrich Hayek dedicated much of his work, from early to the late 20th century, to the study of complex phenomena,[7] not constraining his work to human economies but venturing into other fields such as psychology,[8] biology and cybernetics. Gregory Bateson played a key role in establishing the connection between anthropology and systems theory; he recognized that the interactive parts of cultures function much like ecosystems.


Goes on to explain complexity economics.

marris said...

Yes, those are examples of high-level simplicity arising from low-level complexity. How does Paris get fed, etc, where there are so many individual decisions that contribute.

Economic science (at least our understanding of price signals) needed to develop to the point where such questions could be answered.

What does that have to do with "economics is built on physics and should be built on biology?"

The thing to take away from economic theory is that even big chaotic systems (like the huge number of particles interacting in a complex economy) can be studied at the higher level of choice, exchange, and incentive.

Now what's this about economics being built on physics?

BTW, do you agree that the "cannot predict" of non-chaotic systems objection is invalid? There may be some questions we can't answer about the system without actually running it. And to the extent that the questions are of the form "will the system *ever* do X, Y, and Z", running it will not be enough [you will need to deal with the halting problem].

But for economic systems, we don't need to consider *ever*. We can consider the finite problem of will we see X, Y, and Z over the next 100 years. Whether this can or cannot be answered *is* a question of measurement and computation resources.

marris said...

> BTW, do you agree that the "cannot predict" of non-chaotic systems objection is invalid?

Sorry, this should have read: Do you agree that the cannot predict argument at http://www.synthesisips.net is incorrect?

paul said...

"The thing to take away from economic theory"

There isn't much to take away from neoclassical economic theory. It's based on a false premise and generally ignores system constraints. What can we learn from a theory like that?

Matt Franko said...

Marris,

Speaking for myself, I think what we are doing with "math" or the thermodynamic analogy is perhaps just debunking the notion of "deficits are bad", "we're borrowing from children/Chinese", etc... trying to communicate via "math" that is false.

ie we need positive balances of USD NFAs in the non-govt sector to be able to settle transactions among ourselves, savings is "leakage", CAD is "leakage", etc..

Beyond that, trying to model the whole economy via some type of thermodynamic /math model I dont know that seems pretty complex....

rsp,

marris said...

Paul, I am simply showing how one of Tom's complex system examples (order in market systems...) conforms to my description of emergence (high-level simplicity from low-level complexity).

Of course, if you *don't* think that an economy [laissez faire, command, mixed, whatever] exhibits emergent order, then you wouldn't find a theory of how such order arises to be very useful! It wouldn't matter whether the theory had true premises or false ones, or whether it paid attention to system constraints or ignored them.

But the point of bringing up emergent systems in this post is because some of you think they *are* relevant, right? That some high-level simplicity can be utilized? If not, then why bring it up? Why not focus on Roger's essay, and the points, if any, that it wants to make?

paul said...

Matt,

Yes, and no one here has advocated doing that as far as I can tell.

To say that some complex systems behave similarly does not mean we should model all of the complexity.

First it isn't necessary.

Second, many of the variables in complex systems are inter-dependent so can't treally be pinned down.

I try to break the economy dynamic down to the simplest system that is affected by a known variable not subordinate to a bunch of unknowns.

marris said...

Sorry Roger, I just read your comments. Nothing really made sense until I got to this part:

"...my article was about a year or two of sampling economics literature, and finally recognizing a pattern - that one of their basic premises (equilibrium, or maintaining homeostasis) is only one of the short/medium/long term paradigms that must be fused to model & manage any adaptive system..."

This part started to make sense to me. You spent some time (a year or two) reading papers.

Then you learned something about equilibrium from the papers. So far, so good.

"... is only one of the short/medium/long term paradigms that must be fused to model & manage any adaptive system..."

OK, now we're really cooking with gas. You found that equilibrium analysis (by which I assume you mean comparative statics) is insufficient to capture and analyze some parts of an "adaptive system".

Now what, in the smallest number of concepts, is an example of *something else* you'd like to incorporate? What is the simplest example where comparative statics gives a wrong answer or a bad analysis and your (slightly improved) system gives a better one?

Remember, I'm asking for a simple list of additional concepts that will produce a slightly better analysis. So please don't bring in context, evolution, Hegelian mysticism, the phase of the moon, etc. That would not be *slightly better*.

paul said...

"But the point of bringing up emergent systems in this post is because some of you think they *are* relevant, right? That some high-level simplicity can be utilized? If not, then why bring it up? Why not focus on Roger's essay, and the points, if any, that it wants to make?"

Marris, I think Roger's posts are interesting and informative and I can connect with them in an abstract way, not as a tested solution to an immediate problem. I "see" familiar concepts hidden in his posts.

My education and past reading haven't taken me in those directions so I read with interest what Tom and Roger write and don't generally get involved in the discussions that ensue. I'm the student here. I have a lot of information to digest and a lot of thinking to do before I jump into those discussion.

On the other hand, I've thought a lot about the things I talk about and I'm confident with my understanding of the math and system concepts my arguments are based on, as well as my conclusions.

The same principles form the basis for MMT thinking. I've always (since college) known of these principles, introduction to MMT gave me a clue as to how to apply them.

It appears to me that your worldview and Rogers or Tom's may not be that far apart but semantics and communication problems are keeping you from converging on the root ideas.

I suggest a liitle less criticism and a little more discussion over these issues. YMMV.

Tom Hickey said...

marris, the fundamental difference between neoclassical economics and both Keynesianism (excluding the Keynesian-neoclassical synthesis) and Austrian economics is ergodicity v. non-ergodicity.

Tom Hickey said...

Sorry, this should have read: Do you agree that the cannot predict argument at http://www.synthesisips.net is incorrect?

As Greg Fisher points out in that article, complex systems are not absolutely unpredictable because they are not random but exhibit patterns. But because they are not deterministic (ergodic) they cannot be predicted with certainty even if the initial conditions are known because emergence is inherently unpredictable (ontological uncertainty). Chaotic systems are deterministic and the reason that it is very difficult to make predictions about them is due to lack of knowledge, that is, epistemological uncertainty.

Tom Hickey said...

paul, I think you may be somewhat overconfident about how accurate MMT predications can be in the real world where data is late and incomplete, and human action in society exhibits both complexity and reflexivity. The sectoral balance approach based on identities is a guide to policy flexibility, not a set of equations that will predict what will actually happen in the future. This is the case because changes in saving and spending desire are unpredictable and the govt fiscal balance is endogenous. It is only possible to forecast in a general way at the macro level, especially wrt putting a clock on it. But it is possible to formulate flexible policy options for achieving and maintaining balance of the sectoral balance at full employment using functional finance and the MMT JG as price anchor and buffer of employed. But how the fiscal balance will actually turn out due to changes in tax revenue and automatic stabilizers cannot be predicted accurately in advance with a high degree of certainty, as is possible in physical (ergodic) systems.

Tom Hickey said...

In think that basic point of Roger's post, boiled down to a paragraph, is that mainstream approach to economics is based on modeling with a view using independent variable to produce desired effects in dependent variables. The adaptive approach is to realize that this is a direct, fixed approach to a problem than requires an indirect, flexible approach. Even if an economic model has been correct in the past, it may not be in the future because social science is non-ergodic due to complexity and emergence.

So it is not possible to construct a deductive model that is applicable always and everywhere. This is an issue, for instance, in trying to export American-style capitalism on the belief that it is optimal for all situations across time and place, as it is with "mechanical" physical systems, which are ergodic

There is a reason that businesses, the military, software developers, etc. take an operational approach based on indirection. That reason is experience of the challenge of increasing complexity as systems grow.

Adam Smith is the most quoted economist. Is there any other science that looks backward to its beginnings for current information and development?

paul said...

Tom,

Yes, I'm aware of this and I don't mean to imply that I think I or MMT or anyone can predict the future or prognosticate with 100% accuracy or any high degree of confidence.

What I mean is that the universe of possible outcomes is much more manageable wrt if-then scenarios.For example one can assume with a high degree of confidence that any policy or event that results in a decline in the size of or in a demand-reducing distribution of the pool of funds available for spending it will likely cause an economic contraction. If it didn't, that would be news.

Semantics again or my natural-born ability to leave out some detail, but then we aren't writing books here.

I figure if my comments aren't clear a little probing like you have done here will clear things up.

Tom Hickey said...

Just clarifying the boundaries, paul. The sectoral balance approach is powerful but not omniscient. I would say it is a result of both epistemological uncertainty (bounded rationality) and ontological uncertainty (course of future events radically unknowable). this is why policy formulation has to be flexible enough to respond to the unknown unknown.

Ideally in the 2008 crisis, the policy response would have been immediate provide unlimited liquidity to the financial system while putting insolvent institutions in to resolution, providing fiscal stimulus as needed to sustain effective demand at full employment, and a buffer of employed to mop up the residual. Then the people in charge should have conducted an actual investigation into what happened and what to do about it, conducted by experts like Bill Black.

Instead, virtually everyone one in charge was clueless, and acted too late and with too little, when at all. Why? Bad models and rigid approach, along with insufficient built-in redundancy. The result has been massive failure that is still reverberating around the globe.

Anonymous said...

Other people are beginning to ponder about stability and minsky:

http://www.zerohedge.com/news/can-it-happen-again-again

Many have talked about it. More have eschewed it. But Minsky's hugely important insight in asking the question "Can 'It' Happen Again?" regarding the Depression remains critical reading for any- and every-one who opines day-in and day-out on how much we need or do not need Central Bank money-printing. As Bill Gross put it: "Minsky, originator of the commonsensical “stability leads to instability” thesis; the economist with naming rights for 2008’s “Minsky Moment”; the exposer of the financial fragility of modern capitalism; probably couldn’t imagine the liquidity trap qualities of zero-based money, because who could have conceived 30 or 40 years ago that interest rates could ever approach zero per cent for an extended period of time? Probably no one. Nor, more importantly I suppose, can Ben Bernanke, Mario Draghi or Mervyn King. In their historical models, credit is as credit does, expanding perpetually after brief periods of recessionary contraction, showering economic activity with liquid fertiliser for productive investment and inevitable growth." For a long-weekend, we present the full 30-year-old must-read paper.

Tom Hickey said...

One more key point to add to the discussion of complexity and emergence. The greatest and most far-reaching physical example of emergence from increasing complexity is consciousness as "an emergent property of matter." This was not only unpredictable — there was no one around to predict it — but also it has never been explained after the fact. There is as yet no scientific "theory of consciousness."

Once consciousness itself become sufficiently complex owing to the evolution of highly developed nervous systems, then discovery becomes possible. Discovery is not unique to humans, although rational discovery is.

No logic of discovery has been developed to enable an intelligent machine to make new discoveries. This is called "the technological singularity," or just "the singularity." Whether there ever will be is controversial although it doesn't seem to be logically impossible so it may come to pass through future discovery. Some project (Ray Kurzweil) this taking place in a relatively short period.

Since there is no logical of discovery, no direct path to making discoveries is yet known. Therefore, the need for indirection, which is the way that nature worked throughout the evolutionary process. While there were fantastic success stories — us — there were many more dead ends.

It is foolish and short-sighted of mainstream economists to declare themselves "orthodox" and everyone else "heterodox," since this severely limits the range of exploration of options and the testing of potentially promising alternatives simply because they don't fit the mold of the current paradigm and therefore are out of bounds. Exploration of options involves thinking "out of the box."

Tom Hickey said...

The Complexity Vision and the Teaching of Economics - David Colander

This work focuses on the implications of the complexity vision, such as that held by economists at the Santa Fe Institute, for the teaching of economics. This complexity vision suggests that answers to questions such as how do markets develop and how do they evolve need to be approached head on. Complexity economics is beginning to do just that. Most of the work in complexity is highly formal and technical; it seems far away from issues such as the teaching of economics. This book is different. The focus of this book is not on the grand theories, or technical aspects, of complexity. Instead it is on the teaching of economics. It asks the question: how would the teaching of economics change if complexity is taken seriously? An outstanding group of contributors, including Brian Arthur, Buz Brock and Duncan Foley, provide interesting and provocative answers to that question in a non-technical and highly accessible style. It is a book that should be read by all those teaching economics, as well as those who are interested in where the complexity revolution in science might be leading.

Reviewer:

Colander's compilation of articles is an excellent source of information for anyone looking to learn more about why the neoclassical paradigm is insufficient by itself to prepare undergraduate students for a complex, dynamical world (and economy). Well organized and well supported arguments make one wonder why programs haven't already been developed and implemented into undergraduate programs. This book is a great place to start for anyone new to the debate of neoclassical/traditional economics v. complexity/evolutionary economics.

Roger Erickson said...

Marris said:

"Then you learned something about equilibrium from the papers."

? What were you reading. I specifically stated that I learned NOTHING about equilibrium from economics papers. In fact, I learned that they were over-fixated on equilibrium as a inadequate paradigm for understanding adaptive systems. Talk about broken semantics!

"Now what, in the smallest number of concepts, is an example of *something else* you'd like to incorporate?"

First, it was implicit. "Add medium & long term process functions." Duh!

Plus, again, you're entirely misreading my input. I DO NOT WANT TO BE A BANKER OR AN ECONOMIST! In fact, I actively do not want to have to know. I only want to know why these fields have failed to maintain competence in their areas.

Citizens cannot possibly be, nor should they want to be, experts in all fields. The only scalable approach is to demand competency in other fields, and invent ways to measure & enforce such competency. Trust, but verify.

Hence my suggestions are NOT about the domain-specific details of economics or banking (anymore than I'd expect bankers to make suggestions about neurophysiology details). My suggestions are for them to be logically consistent in their education, training & practice of banking & education. Beyond, that, I do not want to have to know the details.

If we're not willing to let people in dedicated fields come up with their own solutions, we cannot succeed. We have to provide them with goals we as a group want to achieve, and them cede to them the freedom to achieve.

YOUR'RE ASKING FOR MICROMANAGEMENT!

General Patton & countless others have very specifically noted that your approach is suicidal for group achievement.
http://www.brainyquote.com/quotes/authors/g/george_s_patton.html

Also, various statisticians & philosophers have agreed that the best way to tune complex systems is to restrict interventions to causality branches ~5 layers deep in the causality cascade. (e.g., To find the truth, ask "why" 5 times.) That basically fits with Patton's suggestion, and the trust/delegation rule in biology (demand performance criteria, not micromanagement).

This all comes down to scaling by adjusting demanded tolerance limits among coordinating processes. The scalable solution is to broadcast the developmental needs, constantly update demands for altered tolerance limits, and delegate the freedom to innovate at will.

If YOU have a suggestion to make, make it. My suggestion to you is to avoid complaining about the lack of micromanagement. The only rational thing to do, ever, is to make a suggestion. (And to only suggest more suggestions is NOT a suggestion, it's whining & complaining.)

Anonymous said...

http://www.youtube.com/watch?v=5UwvaSLbIgc&list=SP47C8584198EB3982&index=0&feature=plcp

Erickson, Graham Hancock talking to Dr. Richard Dawkins (of the selfish gene) on altered consciousness and trying mind altering substances. I want to know if hickey and erickson and franko have tried these altered consciousness states as good scientists?

Roger Erickson said...

"I want to know if hickey and erickson and franko have tried these altered consciousness states as good scientists?"

A few. But given the unappreciated diversity of humans, some of us find that "mind-altering" drugs simply slow down the wanderings of an already hyper-active mind & subconscious. Personally, I didn't see the attraction. Plus, there were distracting side effects, from bad tastes & lingering smells, to headaches & painful sensory hyper-sensitivity.

I've heard they're most helpful for the typical engineering mind. :)

[And useless on English & political science majors. :( ]

But seriously, supposedly, there's a drug for every NAICS code. I think the Kyrds did a bong about that concept. :)

Sufi it to not say, most of us would give it a whirl.

Tom Hickey said...

"I want to know if hickey and erickson and franko have tried these altered consciousness states as good scientists?"

In the Sixties and early Seventies as a grad student I make such experiments and took it rather seriously as a research project. Definitely shifted my experience and understanding about philosophy and psychology, where I was already acquainted with writing and reports on altered states of consciousness historically and theoretically.

However, my experimentation led me to conclude that chemical means of altering consciousness put a strain on the physical system, including the nervous system, and so I decided to pursue natural means that had been described over the ages in religious, philosophical and spiritual literature.

I also felt that it was inefficient to try and re-invent the wheel, and that I should seek out masters in that field, as I would in any other. Productive choice. However, all that glitters is not gold and one needs to be discriminative.

As a result I have come to the conclusion that consciousness is primary and that the future of humanity lies in the direction of inner space rather than outer space.

This frontier is already opening widely in the West resulting from the impetus of the Sixties and Seventies, and the influx of some high-quality Astern teachers and proliferating translation of texts. But experience cannot be gained from study alone, since it involves developing inner awareness beyond ordinary limits. The requires supplementing knowledge of the mind with knowledge of the heart.

Anonymous said...

I was never interested in chemically enhancing my thoughts or brain, there was some curiosity there but never to overcome the propganda I was fed. The only honest reason I got in the scene (as someone who gives it to others but do not consume myself) is that it leads to sex with many beautiful young women that I couldn't achieve any other way. However, as I have aged, even that is not enough leverage. So now I am about to permanently exit the USA for south east asia my desire for that companionship is so strong. I feel this inability to let go of this path is certain to lead to my destruction. I am just so depressed and unhappy without young female companionship, to the point of losing my will to live. If there was a pill I could take to alter this and save my life, I would take it for that reason, but I don't think anything will help.

All this time I spend on these blogs and trading money, its an escape mechanism to get away from those thoughts and feelings and it has helped tremendously.

I really appreciate your openness.

http://news.yahoo.com/mitt-romneys-mexican-roots-033457860--abc-news-politics.html

In 1885, Romney's great-grandfather, Miles Park Romney, fled to Mexico to escape America's anti-polygamy laws. Along with a group of his fellow Mormons, Miles Park Romney, who had four wives and 30 children, settled in Chihuahua, Mexico, where polygamy was still legal.

Tom Hickey said...

@ Anonymous,

There is a teacher for everyone. Mantak Chia and Osho have specialized in integrating sex and spirituality. They attract a lot of good broad-minded folks. Check it out if you haven't already.

Mantak Chia, Chiangmai, Thailnd. He has centers all over the world, has written many books, and made a lot of YouTube vids, too.

Osho Meditation and Retreat Center, Pune, India (Osho is no longer in the body, but his movement and meditation center is going strong.

Also, Harbin Hot Springs, Middletown, CA.

marris said...

> "What were you reading. I specifically stated that I learned NOTHING about equilibrium from economics papers."

Uhm... maybe I should have written that you learned something about *the role* of equilibrium analysis in economics. Did you at least learn that much?

> First, it was implicit. "Add medium & long term process functions." Duh!

Wow, I think we've finally hit a concrete idea here.

Now my understanding of process functions is that they are used to describe paths that systems take from one equilibrium to another. [http://en.wikipedia.org/wiki/Process_function].

Isn't this what comparative statics already does?

You start with one equilbrium state. You shock something. You find the new equilibrium.

The analysis *finds* the process function (or at least one step of the path) for a given shock? We could do something similar for a given sequence of shocks.

Are you suggesting we need a "shock path" ? I don't see how that would be useful. That sounds like vanilla dynamic equilibrium analysis to me, something which I've seen neoclassicals and Austrians do. [Some Keynesians have probably also done it].

Remember, I'm really excited about this discussion, because it's the first understandable chain of ideas that I've seen in your articles:

- You ready some papers
- You learned something about the role of equilibrium in economics
- You found something lacking about the ideas that economists use... something to do with process functions.

But as I've described above, we already have something that *looks* like process analysis.


Oh, I'm not asking for any kind of micromanagement. I'm asking if there are any concrete management ideas here. For example, if you said: I believe the MMT policy of

- Level targeting cash+GovertnmentBonds (nominal, or nominalish rule)
- Level targeting of employment (real rule)

is good, but you have no idea why it works, then that's fine for me. Most guys who have nominal target rules don't know why they work.

But I had hoped to find something that I could learn from your essay.

vimothy said...

Tom,

Chaos Theory is the study of fascinating yet deterministic [ergodic] systems, whereas complex systems are not deterministic [ergodic]....

Not sure why you've inserted the word ergodic here.

Ergodic is not synonymous with deterministic. (Stochastic processes can be ergodic).

the fundamental difference between neoclassical economics and... Keynesianism... is ergodicity v. non-ergodicity.

It's not the case that neoclassical economics is only concerned with deterministic systems.

And it's also not the case that post Keynesians are only concerned with systems that are not deterministic (i.e. are stochastic). For example, all of the models in Godley Lavoie are completely deterministic.

Tom Hickey said...

For a technical explanation of the difference between ergodic and nonergodic stochastic processes one should read my book, The Keynes Solution: The Path To Global Economic Prosperity [Davidson (2009)] . For our discussion here we merely need note that, in essence, the ergodic axiom imposes the condition that the future is already predetermined by existing parameters (market fundamentals). Consequently the future can be reliably forecasted by analyzing past and current market data to obtain the probability distribution governing future events. In other words, if future events are assumed to be generated by an ergodic stochastic process (to use the language of mathematical statisticians), then the future is predetermined and can be discovered today by the proper statistical probability analysis of past and today’s data regarding market “fundamentals”. If the system is nonergodic, calculated past and current probability distributions do not provide any statistically reliable estimates regarding the probability of future events. 

The ergodic axiom: Davidson versus Stiglitz and Lucas

vimothy said...

Note the following sentence:

Consequently the future can be reliably forecasted by analyzing past and current market data to obtain the probability distribution governing future events.

Davidson is saying that we can take a sample of past realisations of a process and use that data to obtain the distribution. In other words, the process is "random" (i.e., not deterministic), but its distribution is knowable.

Tom Hickey said...

As I read that in context, Davidson is stating the position that he is opposing there, not affirming.

Roger Erickson said...

Marris said:

> Uhm... maybe I should have written
> that you learned something about
> *the role* of equilibrium
> analysis in economics. Did you at
> least learn that much?

It's like we live in different worlds. The whole essay was about realizing that economists are trained to obsess about equilibria where there ARE no equilibria.
Nothing to learn here. Move along, folks.

> Now my understanding of process
> functions is that they are used
> to describe paths that systems
> take from one equilibrium to
> another.

? You must live in a very stochastic world. The rest of us liver in a very dynamic, unpredictable world. You want a recipe for every day of your life? Try surfing instead, where nothing can be predicted, but ANYTHING can be adapted to.

> Isn't this what comparative
> statics already does?

No.

> You start with one equilbrium state.

NO. You DON'T. Does a stone skipping over a water surface bounce between equilibrium states? Imagine the stone was a sentient & adaptable surfer. She considers the equilibrium state of each impact as only a step in a path. There are enough unpredictable things that can occur during the next touchdown to make any expectations irrelevant. Plus, the touchdown periods are so transient as to not matter to long term planning of where the surfer wants to go. Gen. Patton would say "We need to get to shore. Find a way."

> You shock something. You find the
> new equilibrium.

No. You adapt to unpredictable events, then make sure you have enough resiliency to adapt to another set of unpredictable events. We estimate Survival as = F[s(a), m(b), l(c)] (short, medium, long term events), and make new adjustments based on any surprises. Most required adaptations are automatically delegated to semi-autonomous functions, and THERE IS NEVER ANY EQUILIBRIUM, only continued adaptation on a reverse-entropy path.

> The analysis *finds* the process
> function (or at least one step of
> the path) for a given shock? We
> could do something similar for a
> given sequence of shocks.

You sound like an engineer. We never have enough data to engineer responses to the future. Formal engineering only applies to things we have adequate data for. Biology, including human culture, is always discovering, and defining solved puzzles as engineering ... as it moves on to emerging surprises.

> Are you suggesting we need a
> "shock path" ?

Nature provides with enough already. I was presuming familiarity with total adaptation, where we consider our entire nation as a tribe of "Context Nomads," ready to use all available resiliency to adapt ASAP to constantly emerging options.

Roger Erickson said...

Marris says:

> Remember, I'm really excited
> about this discussion, because
> it's the first understandable '
> chain of ideas that I've seen in
> your articles:

You & I still think differently. You seem to want a recipe for a slowly changing, fairly static world. I'm used to thinking of our ability to use vast amounts of latent resiliency to surf an entirely unpredictable world. Maybe you haven't experienced the "Fog of Reality." [google "fog of war" & Clausewitz]

> - You learned something about the
> role of equilibrium in economics

How many times do I have to say that I learned NOTHING about equilibrium in economics - and that I instead realized that economists have a false belief that there is such a thing?

>- You found something lacking
> about the ideas that economists
> use... something to do with
> process functions.

Sigh. If that's the way you like to imagine things.

> But as I've described above, we
> already have something that '
> *looks* like process analysis.

Then, conveniently, it seems that you don't have to think?

> I'm not asking for any kind of
> micromanagement. I'm asking if
> there are any concrete management
> ideas here.

No management ideas. Only adaptation ideas.

> For example, if you said: [some
> fixed recipe would work] I would
> understand

I'd never say any simplistic recipe is guaranteed to work. Even monetary operations only defines SOME the variables we currently know we have to work with. I'd only advise we know of all the inter-dependent variables that are in our long adaptation polynomial, and be prepared for the polynomial order of every strong/weak variable weighting to be a probability function (i.e., can surprise us, and constantly drift too). Then, I'd advise us to never be timid about changing anything in our own responses, because our goal is to keep on surfing, NOT to establish a new arbitrarily defined equilibrium.

As an aside, so much of policy & economics is sidetracked into discussion of what's "equal" vs "fair." For adaptation, that's usually irrelevant. Take a squad of Marines in a battle. Their concern is never over distributing ammo equally or fairly - it's on agile re-distribution in whatever moment-to-moment way guarantees squad survival. Equality/fairness discussions are a sure sign of a bored population who's lost situational awareness - sitting cluelessly in the road of national options, about to get run over by unpredictable fate.

> But I had hoped to find something
> that I could learn from your
> essay.

Per Walter Shewhart, what you learn from any data is the context you bring to it.

Roger Erickson said...

ps: Marris. In all sincerity, my suggestion is that you take up surfing. While you're practicing, ask yourself whether your net motor systems (smooth & skeletal muscles, autonomic & motor neural systems) is ever formally going between "equilibrium" states.

http://en.wikipedia.org/wiki/Motor_control

Navigation by an already complex physiology (your body) is actually a useful analogy for context navigation by a national population that is another order of magnitude more complex. No equilibria ever involved in either case though.

vimothy said...

As I read that in context, Davidson is stating the position that he is opposing there, not affirming.

Yes, but in order to do so, Davidson is describing what an ergodic process is. He means a kind of process where it's possible to discover the underlying probability distribution. Such a process is obviously stochastic, or there wouldn't be any need to talk of distributions.

So the type of process that Davidson is describing is not deterministic.

Tom Hickey said...

Per Walter Shewhart, what you learn from any data is the context you bring to it.

If one belives the context is an equilibrium-based on, then one is not going to be as adaptable as those who accept a non-equilibrium context.

Practically speaking in economics, is the optimal policy to do nothing and wait for equilibrium (presumptively at full employment), or do something. Some, notably the liquidationists would say that doing nothing and letting the system clear and reset is the optimal path. Neoliberals would says if we are going to do something positive, reduce the influence of government.

Contextualists would say to explore all options intelligently, adjust for feedback, and double down on the one's that are working out most successfully, and either revising or abandoning those that are not working out so well.

Tom Hickey said...

vimothy, I will grant you that I used the term "deterministic" unclearly. What I meant was a loose meaning of it rather than the technical distinction of deterministic v. stochastic or probabilistic.

As is well-known, Einstein objected to the probabilistic nature of quantum mechanics when he said, "God does not play dice." Here is very nice lecture by Stephen Hawking on it.

I thinking along the lines of the way Davidson uses "predetermined":

Contemporary neoclassical economists proceed under the assumption that as concerns the economy there exists a predetermined reality that can be fully described by “unchanging objective conditional probability functions”. 

This ergodic axiom assumes the economic future is already predetermined[4] .  The economy is governed by an existing ergodic stochastic process. One merely has to calculate probability distributions regarding future prices and output to draw significant and reliable statistical inferences [information] about the future

For our discussion here we merely need note that, in essence, the ergodic axiom imposes the condition that the future is already predetermined by existing parameters (market fundamentals).

In other words, if future events are assumed to be generated by an ergodic stochastic process (to use the language of mathematical statisticians), then the future is predetermined and can be discovered today by the proper statistical probability analysis of past and today’s data regarding market “fundamentals”.

Nothing Congress, the President of the United States, the United Nations, or environmentalists can do will alter the predetermined dates and time for future eclipses. For example, Congress cannot pass a law outlawing solar eclipses in order to provide more sunshine and thereby enhance crop production. In an ergodic world, all future events are already predetermined and beyond change by human action today.

What is this reflexivity? In a letter to the Editor published in the March 15-21, 1997 issue of The Economist Soros objects to Paul Samuelson insistence on requiring the ergodic axiom to make economics a science. Soros argues the ergodic hypothesis does not permit “the reflexive interaction between participants’ thinking and the actual state of affairs” that characterizes real world financial markets. In other words, the way people think about the market today can affect and alter the future path the market takes; the future is not predetermined. Soros’s concept of reflexivity, therefore, is the equivalent of Keynes’s rejection of the ergodic axiom[6]. Reflexivity means peoples thoughts and actions create the future, while mainstream economists presume the future has already been predetermined and can be discovered by analyzing today’s market fundamentals.

Tom Hickey said...

Oops, I neglected the link to the Hawking lecture. He is not only a brilliant scientist but also an excellent expositor.

Roger Erickson said...

"If one belives the context is an equilibrium-based on, then one is not going to be as adaptable as those who accept a non-equilibrium context."

Thanks, Tom. I was expecting more people to see that as the obvious extension of my essay.

Well said. It helps to explicitly say what you may at first assume to be implicit. There are too many audiences for any of us to know them all - so any data has to be explicitly restated in terms any given audience finds relevant.

paul said...

Roger,

I'm beginning to believe that for the semantic-based thinkers there is no implicit element of an argument.

It is what it is.

marris said...

Roger, I haven't surfed before, but I've controlled a sailboat. It is certainly about management. Manage the direction of the sail relative to the wind. The direction of the rudder. How much to let out the sail.

I could just as easily say I am adapting the direction of the sail to the wind and my destination.

Are there any concrete adaption rules in your essay? If not, are you saying the essay thesis is "we need to adapt... yep, that's what we need to do"?

I'm not sure why you've sprinkled repetitions of this through my explanation of comparative statics. Process functions (which you thought were not covered) are indeed part of comparative statics and dynamic equilibrium analysis.

Now you're trying to shift from talking about process functions to response functions? Do you think economics doesn't cover those either?

s. Before criticizing something (neoclassical models, MMT, whatever), it helps to talk to a knowledgeable person, like a professor, learn what they are saying, and reflect on it. It looks like you tried to read some really specific domain papers without grasping the higher level layout of the field. That's a bad move. I recommend studying some of the philosophical oriented papers instead, as well as working through an intro and intermediate text to see what ideas are already part of the field.

paul said...

"I recommend studying some of the philosophical oriented papers instead"

marris,

This may be good advice but as far as I'm concerned, the core of economics is a math/engineering problem and if that part is misunderstood there's not a whole lot any of these other disciplines (social science, finance, etc.) will be able to accomplish re the management of economies, other than by accident.

If your car is out of gas you can still make it go by pushing it, but why bother? Walking to your destination would be more efficient.

paul said...

New record.

Took six tries to prove I'm not a robot?

Am I number 1?

Roger Erickson said...

I should have said "Per Walter Shewhart, what you learn from any data is [based on] the context you bring to it."

Marris, we have a semantic problem. You talk of managing sailing when what you describe is actually managing your own responses to sailing conditions. At the same time you imply that managing economic parameters equates to managing the wind, waves & currents.

May I suggest a beginning course in semantics?

And, are you referring to "statics" or statistics? I can't tell.

"response functions? Do you think economics doesn't cover those either?"

Oh lord! I specifically stated that, as a bare minimum, all complex systems adapt through a dynamic mix of short, medium & long time-course adaptive processes. There is NO existing business or economics model that covers more than the rudiments of very short-term response planning. Medium & long term planning is still left to the realms of martyrs & statesmen.

My essay simply baldly stated that the economics emperor has no clothes other than a mirage, and that even the mirage itself was only intended as a short term delusion.

My overall view is in support of Warren Mosler's. The entire finance field is more trouble than it is worth. So are you.

Hence, I'm no longer motivated to try anything except Nebraska humor on you.

vimothy said...

Tom,

Deterministic has quite a specific technical meaning. That meaning is strongly suggested if you run it alongside terms like "ergodic". I suggest that it would improve the clarity of your message if you found a different word that doesn't have such a specific meaning already attached to it.

Incidentally, I'm not sure that Davidson is a great reference for "ergodicity". He seems to use it in a very expansive, catch-all sense.

A definition of ergodic by an academic statistician can be found here (with some references to post Keynesians):

"A stochastic process is ergodic when all invariant sets either have probability zero or probability one. What this means is that (almost) all trajectories generated by an ergodic process belong to a single invariant set, and they all wander from every part of that set to every other part — they are "metrically transitive". (Because: no smaller set with any probability is invariant.) From this follows Birkhoff's individual ergodic theorem, which is the basic strong law of large numbers for dependent data. If X is an ergodic process, then for any (integrable) function f, the average of f(Xt) along a sample path, the "time average" of f, converges to a unique value almost surely. So with probability 1, time averages converge to values characteristic of the ergodic process."

http://cscs.umich.edu/~crshalizi/weblog/679.html

Tom Hickey said...

Interesting post by Cosma Shalizi, vimothy. Don't know how Post-Keynesians would respond, or if any have.

BTW, for those interested, here is Ergodicity and Non-Ergodicity in Economics by Ulrich Horst.

Roger Erickson said...

Pity that none of the conditions establishing EITHER ergodicity or non-ergodicity ever apply to real policy formation. We're always somewhere in between, in massively multi-variate internal & external conditions.

Ergo, one might imagine that economics in general is not relevant for policy formation - except for imagining ridiculously improbably limiting outcomes.

You'd think we'd pay more attention to operations, and less to theory that is always fascinating, but clearly not up to helping us make tough decisions.

Tom Hickey said...

Couple of posts on ergodicity and economics.

Rick Bookstaber on Ole Peters refutation of Karl Menger.

A Crack in the Foundation: An error that has wended its way through economics for 77 years. Based on Ole Peters, Menger 1934 revisited

Ole Peters explains ergodicity non-mathematically "for the rest of us."

Shaking the Foundation — Revisiting Basic Assumptions about Risk, Reward, and Optimal Portfolios — An Interview with Ole Peters

See also Santa Fe Institute.

marris said...

> Marris, we have a semantic problem. You talk of managing sailing when what you describe is actually managing your own responses to sailing conditions. At the same time you imply that managing economic parameters equates to managing the wind, waves & currents.

I don't know what "managing sailing" is.

"Sailing" is managing the position, orientation, and speed of a boat. Knowledge of how to do this is required to get the boat (and riders) from point A to point B in the face of uncontrolled wind and waves.


> May I suggest a beginning course in semantics?

May I suggest that the problem here is not so much a confusion of meaning so much as a lack of semantic content in your post(s)?

For example, it's not too difficult to combine words and phrases into syntactically correct forms which have no semantic content. [Example: http://pdos.csail.mit.edu/scigen/].


> And, are you referring to "statics" or statistics? I can't tell.

In response to your claim that process functions are not studied in economics, I pointed out that comparative statics and dynamic equlibrium analysis do exactly this.

In response to Tom's comments about complex vs. chaotic systems, I raised the distinction between probabilistic models and non-probabilistic ones. Maybe some of these statements can be applied to statistics? I don't think I needed statistics to make my points.


> There is NO existing business or economics model that covers more than the rudiments of very short-term response planning.

I'm not sure why you think this. For example, rules-based monetary policy covers targetting real and/or nominal values over very long stretches of time. So does MMT, to the extent that it is rules-based.

Thanks for your time. I haven't gotten much from your posts and this discussion, but I guess you should keep writing so long as someone does.

paul said...

""rules-based monetary policy covers targetting real and/or nominal values over very long stretches of time"

Monetary policy doesn't/can't change the level of net financial assets in the non-government directly. It can change the rate of interest payments on bonds but the ultimate transfer mechanism is through fiscal spending. Monetary policy is zero-sum - every gain by one entity is matched with a loss by another.

NGDP targeting may go beyond this reality to some extent but it still will not put spending in the hands of wage earners.

Monetary policy can move assets around among some cohort in the economy. It is not efficient at or even minutely effective at putting sustainable spending in the pockets of wage-earners, which I presume account for the vast majority of consumption.

"… So does MMT, to the extent that it is rules-based"

MMT is based on natural rules already in place in the universe of embedded systems, like entropy, conservation of energy, system relationships etc.

All MMT does is recognize these relationships that have always existed and put them to use.

There's nothing like using principles of which the outcome is predictable.

marris said...

> Monetary policy doesn't/can't change the level of net financial assets in the non-government directly...

paul, whether it does or does not may be an interesting topic, but I think this still holds up as an example of non-"short term response planning."

With respect to your particular claim, rather than use the term "NFA," I prefer to think in terms of X=cash+TotalTreasuryMoney... where TotalTreasuryMoney is a function of the face value and coupon sizes of the set of outstanding Treauries.

Simple MMT, as presented by Mosler and friends, is basically level targeting of a measure like this (maybe plus some real variable target, like unemployment rate... JG programs come into play here).

First, I think it is obvious that monetary policy *can* affect X. For example, the central bank (CB) can buy a zero coupon bond with a face value of $100 for $200. This would increase X by $100.

Second, as you state, other CB actions can also affect X. For example, the CB can drop money into private accounts. Or it can buy non-Treasury assets. Or it can hire people to do jumping jacks. Whatever.

Third, the CB actions may certainly be selected so as to target NGDP levels. Or to target the price levels. Or to target the price of oil.


> Monetary policy can move assets around among some cohort in the economy.

What it really does is replace non-cash items on the asset side of private balance sheets with cash. These assets can be thought of being "bought" by the CB.

The effectiveness of such actions is certainly debated. I think investment expenditure is probably more important than consumption, but that is also a debatable issue.


> MMT is based on natural rules already in place in the universe of embedded systems, like entropy, conservation of energy, system relationships etc.

It is always possible to bolt things like entropy onto any theory. Especially when you have a pending paper deadline. I'm fairly skeptical of the value of such approaches, particularly of their use in economics.

They may push the forecasting arms race further, but sooner or later, its proponents either incorporate teleology-inspired hacks of disappear into the void of consistent-but-inapplicable schools.

Roger Erickson said...

I rest my case.

Joseph E. Stiglitz "Much of what is taught in the graduate schools of economics is wrong."

http://www.youtube.com/watch?v=eCTIi2NNlr4

paul said...

"First, I think it is obvious that monetary policy *can* affect X. For example, the central bank (CB) can buy a zero coupon bond with a face value of $100 for $200. This would increase X by $100.

Second, as you state, other CB actions can also affect X. For example, the CB can drop money into private accounts. Or it can buy non-Treasury assets. Or it can hire people to do jumping jacks. Whatever"


This may or not be true, but to date the Fed has never done this. Further, any operation that adds net funds to to the non-government is by definition fiscal, not monetary, so if the Fed did this it would be straight fiscal policy, no monetary involved. The Fed is very limited in it's ability to do fiscal. Congress was explicitly given the power of the purse by the Constitution, th efact that Congresshas abdicated some of this responsibility does not change the fact that, as far as the Constitution is concerned the Fed doing fiscal is not allowed. But this is a useless side argument, so I will move on.

"It is always possible to bolt things like entropy onto any theory. Especially when you have a pending paper deadline. I'm fairly skeptical of the value of such approaches, particularly of their use in economics."

You have managed to completely misunderstand what I wrote, and as a result I am bewildered.

Nowhere did I imply that entropy was in any way "attached" or directly applicable to economics. What I wrote was that these things just "are", they are embedded system relationships that exist regardless of whether we acknowledge them or not. Similarly, because a monetary economic system has properties of a closed system, it will behave exactly like a closed system would be expected to behave within certain boundary conditions. In the case of a monetary economy, the unit of account and its quantity defines the closed system.

There is no policy, law, political action, economic postulate, whatever that can be true and at the same time violate the principles of a closed system.

The implications of closed-system properties are a little abstract maybe, but very simple, in fact, it is so simple that personally I can't fathom anyone not being able to understand it. But, as always, life is full f surprises.

My argument is for the extreme simplicity of the basic underlying constraints on any monetary economic system.

Most others involved in these discussions of MMT add complexity to the problem to the point that it is then impossible to understand.

As an engineer, I learned the tools of problem solving. If one can't define the problem it will be difficult if not impossible to solve it.

The most important thing one learns re problem solving is to break the problem down to manageable parts that are more easily analyzed, then analyze how the different parts work together and finally make adjustments on the pertinent parts as required to achieve the desired outcome. This becomes an iterative process that leads to the desired result.

Most internet discussions re economics fail to define the problem, therefore the problem is misunderstood from the outset. From there it is very unlikely that any discussion moving forward will have any meaningful outcome.

Economics doesn't get any simpler than the MMT view of it, but it is apparent that very few commenters really understand what MMT is based on.