Sunday, April 15, 2018

A Teachable Moment: The Importance of Meta-Learning

Today’s New York Times has a fine article by Manil Suri about math education and the development of reasoning skills.  Its concluding point is that, while the general contribution of the first to the second is weaker than you might think, math instruction can be improved by bringing the math-reasoning tests themselves into the classroom.  I’m pretty confident that Suri is right, since I’ve seen positive results from doing something similar in economics and related areas.

When preparing to introduce a new topic in econ, for instance, I’ll often start by taking stock of what lots of people without an economics background think they know about it.  This might mean looking at surveys or some excerpts from news or other websites.  It often involves drawing out this information from the class itself.  For instance, I’ll divide students up into groups of five or so in which they can say to each other what they believe, or even suspect, about the topic, and then have the groups report in a general way what these views were.  (I try to use methods that don’t identify potentially mistaken concepts with specific people, to avoid any sense I’m trying to belittle anyone.)  Then we will go on to learn about the question, keeping in mind the misconceptions we’ve found and trying to locate the points at which “pop economics” veers off from the real stuff.*

There are many reasons for doing this.  One is frankly political: a lot of the political babble in this country is framed by erroneous economic thinking, such as nearly all the fretting over “the national debt”.  (Every time I bring this up in the context of the income accounting identities I see expanding eyeballs all across the classroom.)  Another is pedagogical: if you don’t put effort into deconstructing pre-existing beliefs as well as developing new knowledge, what you will see on papers and exams is a weird mishmash of the two.  It took me too many years to figure this out.  But a third is the insight Suri also came to, that using an external point of reference to step outside oneself and observe one’s own learning process provides a powerful boost to learning of all sorts.  The misunderstandings of pop econ provide a baseline from which students can measure their progress; they illuminate what they are learning and how.

The name for this is meta-learning (or deutero-learning in cybernetic-speak).  It is foregrounded by activities that help students get outside the technique or concept immediately in front of them and see their learning of it as the object of attention.  Like all forms of learning, it is best approached inductively and in context: rather than give lectures on meta-learning, provide exercises that call attention to it in situ.  I incorporated material to support meta-learning in my textbooks, more in the second (macro) than the first (micro), since I was learning (and meta-learning!) as I went along.

I’ve come to think that explicit incorporation of meta-learning may be the single most important innovation to transform teaching.  For those of you who have this as a day (or night) job, give it a try.

*Just to be clear, “real” economics is not mean “sanctified by the mainstream”, just conceptual approaches that can be supported by careful reasoning and empirical data.  Some mainstream econ is rather closer to the pop variety than to legitimate analysis.

3 comments:

AXEC / E.K-H said...

New insight from Meta-Learning: delete economics
Comment on Peter Dorman on ‘A Teachable Moment: The Importance of Meta-Learning’

Peter Dorman summarizes his insights: “There are many reasons for doing this. One is frankly political: a lot of the political babble in this country is framed by erroneous economic thinking, … Another is pedagogical: if you don’t put effort into deconstructing pre-existing beliefs as well as developing new knowledge, what you will see on papers and exams is a weird mishmash of the two. It took me too many years to figure this out.”#1

Current economics fits the description of a weird mishmash. The four main approaches ― Walrasianism, Keynesianism, Marxianism, Austrianism ― are mutually contradictory, axiomatically false, materially/formally inconsistent and all got the pivotal economic concept profit wrong.

So, the first thing to do is to deconstruct pre-existing beliefs. Standard economics is given with this set of neo-Walrasian hard-core propositions a.k.a. axioms:
HC1. There exist economic agents.
HC2. Agents have preferences over outcomes
HC3. Agents independently optimize subject to constraints.
HC4. Choices are made in interrelated markets.
HC5. Agents have full relevant knowledge.
HC6. Observable economic outcomes are coordinated, so they must be discussed with reference to equilibrium states. (Weintraub, 1985)

The methodological blunder that suffices to make this axiom set unacceptable is that HC3, HC5, and HC6 are NONENTITIES like angels, unicorns, the Easter Bunny, or Superman.

First insight from Meta-Learning: all theories/models that contain just one NONENTITY are a priori false. Never ever apply any of the propositions of the set HC1/HC6.

Krugman once introduced himself as follows: “most of what I and many others do is sorta-kinda neoclassical because it takes the maximization-and-equilibrium world as a starting point ...”. From Meta-Learning we know by now that Krugman has always been an idiot.

Of course, there are not only scientifically incompetent Neoclassicals but also Keynesians. Keynes defined the set of foundational macroeconomic propositions as follows: “Income = value of output = consumption + investment. Saving = income − consumption. Therefore saving = investment.” (p. 63)

Unfortunately, exactly at this point the lethal error slipped in because Keynes never came to grips with profit: “His Collected Writings show that he wrestled to solve the Profit Puzzle up till the semi-final versions of his GT but in the end he gave up and discarded the draft chapter dealing with it.” (Tómasson et al.)

Because Keynes got profit wrong the whole of Keynesianism and Post Keynesianism and New Keynesianism is scientifically worthless.

Second insight from Meta-Learning: all theories/models that contain Income = value of output or I=S are a priori false.

Now we could go on with Meta-Learning and deconstruct Marxianism and Austrianism but the Meta-Learner doubtlessly got the point. Deconstructing pre-existing beliefs means to flush both microeconomics and macroeconomics and the textbooks from Samuelson’s sorry effort to Mankiw and Rodrik down the drain. Or, as Joan Robinson put it: “Scrap the lot and start again.”#2

Egmont Kakarot-Handtke

#1 How to get rid of an obsolete theory
https://axecorg.blogspot.de/2016/08/how-to-get-rid-of-obsolete-theory.html

#2 How to restart economics
http://axecorg.blogspot.de/2016/01/how-to-restart-economics.html

Anonymous said...

Fine essay for teachers at any level, teach how to learn what is being taught as well as teach the material. I am always aware of this need.

AXEC / E.K-H said...

Knowledge is attainable ― even in economics
Comment on Lars Syll on ‘Sometimes we do not know because we cannot know’*

Lars Syll maintains: “To Keynes, the source of uncertainty was in the nature of the real ― nonergodic ― world. It had to do, not only ― or primarily ― with the epistemological fact of us not knowing the things that today are unknown, but rather with the much deeper and far-reaching ontological fact that there often is no firm basis on which we can form quantifiable probabilities and expectations at all. Sometimes we do not know because we cannot know.”

Yes, this is a well-known fact of life since the Stone-Age. There are three ways to deal with the annoying human condition: (i) to repeat the mantra, I know that I know nothing, ad nauseam, (ii) to senselessly speculate about the unknowable which is the business of mysticism/religion/philosophy/journalism, (iii) to put the gray matter between the ears to work which is the business of science: “The object of reasoning is to find out, from the consideration of what we already know, something else which we do not know.” (Peirce)

So, the growth of knowledge in economics has to start with what we know for sure:
• the profit theory, false since 200+ years,
• Walrasian microfoundations (in particular equilibrium), are false since 140+ years,
• Keynesian macrofoundations (in particular I=S/IS-LM), are false since 80+ years,
• Walrasianism, Keynesianism, Marxianism, Austrianism and all variations/derivatives thereof are axiomatically false and materially/formally inconsistent,
• economists are scientifically incompetent,
• economics is a cargo cult science,
• the Bank of Sweden Prize in Economic Sciences is a fraud,
• economic policy guidance has no sound scientific foundations since 200+ years,
• economists are a hazard to their fellow citizens,
• economics needs a Paradigm Shift,
• Heterodoxy is incapable of performing the Paradigm Shift because heterodox and pluralist economists are just as stupid and corrupt as orthodox economists.

So, yes, economists know nothing. But it is false to maintain, as Lars Syll does, that this is an ontological fact. No, it is sheer scientific incompetence. Just the opposite holds: “We must not believe those, who today, with philosophical bearing and deliberative tone, prophesy the fall of culture and accept the ignorabimus. For us, there is no ignorabimus, and in my opinion none whatever in natural science. In opposition to the foolish ignorabimus, our slogan shall be: We must know — we will know!” (Hilbert)

Egmont Kakarot-Handtke

* https://larspsyll.wordpress.com/2018/04/18/sometimes-we-do-not-know-because-we-cannot-know/