Expect Less of Economics

|Peter Boettke|

That is the important message of Russ Roberts's WSJ column from this weekend, "Is the Dismal Science Really a Science?"

Russ's point is quite Hayekian as one might expect and challenges the pretense of knowledge that permeates much of modern economics.  In an essay by Pete, Chris and I entitled "High Priests and Lowly Philosophers: The Battle for the Soul of Economics" we argue that when you ask a discipline to produce results which it cannot possibly produce but demand that it pretend that it has in fact produced those results anyway you discredit the legitimate contributions of the discipline and may even destroy the intellectual creativity and scholarly "soul" of its practitioners.

Comments

[Daniel Kuehn]

It was a good article by Russ - but I think most critiques along these lines (ie - "when you ask a discipline to produce results which it cannot possibly produce") are vastly overstated. You don't question that scientific validity of meteorology or climatology because they can't forecast the weather more than a week in advance (and even then don't always get it right). It's a complex system - just like the economy. We do everything with understood and well documented error. That's to be expected in any science, particularly a science that is less experimental and more observational (although of course GMU and others are making great strides in experimental economics as well). What exactly were people expecting?

[Mario Rizzo]

Economics is not just one thing. Maybe it used to be. I don't know. But it has fractured. It is no longer defined by a subject area. It is defined by a certain method of analysis.

Some "economists" are only interested in deriving theorems in neat axiomatic systems. Some are econometric forecasters (these are in big trouble). The public still thinks economics is about a certain area of problems -- wealth, business cycles, welfare, etc. Sure, economists study these things as well as other things (like fertility). But the esssence of the discipline, as practiced in the top-tier schools, is to study whatever by the use of a certain method -- not the Austrian or Institutionalist method -- of axioms, theorems, maximization, often equilibrium, etc.

Does that method lead to expectations of precise results? Yes. Does it lead to expectations of precise "predictions" (= retrodiction)? Yes. Does it correctly lead to expectations of good FORECASTING. No. I don't think most economists would say they can do this. However, the public may expect this. Whose fault is that?

My upshot is I don't like "essentialist" reasoning in economics or about economics. Economics is not just one thing.

[Greg Ransom]

At a global level, economics is just one thing, like Darwinian biology, it identifies a design-like order and like Darwinian biology it provides a contingent causal explanation providing a mechanism that accounts for that order.

It's as simple as that.

The mistake is to define "economics" as what any particular guild of men do.

Mario writes:

"My upshot is I don't like "essentialist" reasoning in economics or about economics. Economics is not just one thing."

[Greg Ransom]

Essentialism reduces concepts to necessary and sufficient condition. Here's a example of an essentialist definition:

"Economics is what economists do, and what economists do is defined by the methods and topics they choose."

An example of an anti-essentialist notion of concepts is David Hull's notion of concepts as historical individuals, on analogy with biological species understood as historical individuals which cannot be defined in terms of necessary and sufficient conditions, (e.g. a "swan" is white, and lays eggs, etc.).

As an evolving historical individual, the species "swan" does not consist of parts which are necessarily white, etc.

The philosophical tradition in which Carnap, Russell, Nagel, Quine, Hempel, etc. work began from the "necessary and sufficient" conditions picture of language and concepts and knowledge and science -- they began without the ability to imagine the very different conceptual world of Darwinian, Hayekian, Wittgensteinian, biology / social theory / language theory -- populated by evolving populations, family-resemblance pattern perceptions, and overlapping patterns of behavior, etc.

Popper was caught with a foot in both worlds, and never really escaped from the Carnap / Russell world.

As an historical individual, we can chart an evolving historical individual of "economics" working on various evolving problem / solution matrices honing in on something like the problem / explanation achievement trackable in the work of Smith, Menger, Mises, Knight, Hayek and Friedman's popular work, and the implicit economics which lies as the background understanding of the supply-demand / price theory / information economics / equilibrium modeling of the freshman economics textbook.

[Peter Boettke]

Daniel,

It goes much deeper than forecasting let alone aggregation, as we discuss in our paper where we give examples beyond macroeconomics.

Pete

[Frederic Sautet]

Mario, what economics is about has evolved a lot in the last 200 or so years. The subject areas have been multiple. Yet, the method that you talk about (and which slowly emerged over time) seems now to unify economics (let's call it the rational choice framework), including non-heterodox schools. At least, this is the way it is perceived outside the discipline where most opponents to economics see themselves as against the rational choice approach. Is accepting this essentialism? Perhaps, but then what's the problem with that?

What bothers me in Russ's article is that it is a fine line between his take and the position that says that knowledge is impossible (i.e. economics cannot really tell us much at the end of the day). Is this what he really thinks? There may be some confusion here. The question for me is not so much whether economics is a science but whether the differences that exist among economists find their sources in extra-economic issues such as whether economics is an empirical science or not, the role of maths in models, the importance of physics as a model for social science, the source of funding for research, the fact that more than half the profession works directly or indirectly for a government organization, etc. In other words, disagreement does not disprove the absence of "laws" or "regularities" in economics, it just shows that other issues affect the way economists think and their incentives to think (just like other issues affected the way physicians thought in the 18th century when everyone denied the existence of germs).

[Greg Ransom]

Frederic, there is no single "rational choice framework", there are in fact alternative "methods" of explaining things while appealing to "rational choice" -- just as there are rival understandings of "methodological individualism".

Hayek in fundamental ways rejected the "rational choice framework" (see Vol. 1 of _Law, Legislation and Liberty_). Does that mean Hayek wasn't an economist?

[Rafe Champion]

On predictions in economics: Steve Keen wrote a book debunking economics a few years ago.
http://www.debunkingeconomics.com/
He also made some interesting predictions about the Australian economy and now he is paying for his sins.

"Sixteen months ago Mr Keen made a bet with Macquarie Group interest rate strategist Rory Robertson after claiming that house prices would dive by 40% when the GFC was at its worst.
Fortunately his predictions didn’t eventuate, and now Mr Keen will deliver on his promise to walk 224km from Canberra to the top of Australia’s highest mountain, Mt Kosciuszko. It remains to be seen whether he will wear a t-shirt saying “I was hopelessly wrong on home prices! Ask me how.”
Dr Keen was way off the mark. Australian home prices bottomed out by 5.5% from their peak in late 2008."

http://www.keenwalk.com.au/

[topills.com review]

does that method lead to expectations of precise results? Yes. Does it lead to expectations of precise "predictions" (= retrodiction)? Yes. Does it correctly lead to expectations of good FORECASTING. No. I don't think most economists would say they can do this. However, the public may expect this. Whose fault is that?