Economists who rationalise their discipline’s value can be convincing, especially with prestige and mathiness on their side. But there’s no reason to keep believing them. The pejorative verb ‘rationalise’ itself warns of mathiness, reminding us that we often deceive each other by making prior convictions, biases and ideological positions look ‘rational’, a word that confuses truth with mathematical reasoning. To be rational is, simply, to think in ratios, like the ratios that govern the geometry of the stars. Yet when mathematical theory is the ultimate arbiter of truth, it becomes difficult to see the difference between science and pseudoscience. The result is people like the judge in Evangeline Adams’s trial, or the Son of Heaven in ancient China, who trust the mathematical exactitude of theories without considering their performance – that is, who confuse math with science, rationality with reality.
There is no longer any excuse for making the same mistake with economic theory. For more than a century, the public has been warned, and the way forward is clear. It’s time to stop wasting our money and recognise the high priests for what they really are: gifted social scientists who excel at producing mathematical explanations of economies, but who fail, like astrologers before them, at prophecy.--by Alan Jay Levinovitz "The New Astrology" @Aeon.
Levinovitz has a PhD from the Divinity school at Chicago. This is why I'll call him a 'theologian' in what follows. (In the piece he calls himself "a scholar of Chinese religion.") That will help distinguish what he does from what 'philosophers of economics' do, and also call attention to the significance of labels and labeling. In his Aeon piece, Levinovitz refers smoothly to 'economists' without distinguishing the many varieties of economists and economics; he talks of 'mathematical theory,' 'mathiness,' and 'mathematical reasoning' without distinguishing the many different kinds of mathematical theories and uses of mathematics in the varieties of economists and economics; and he talks of 'performance' without bothering to explore the particular ends that different kinds of economists set for their particular theories and models. And he conflates different kinds of non-performance: for example, within a single paragraph, he moves from (a) failed predictions of interest rate movements, (b) to failures to foresee economic crises, to (c) a purported failure of not succeeding at -- pace the confident claim of "the Nobel Laureate Robert E Lucas Jr" -- 'depression prevention.' What's especially odd about this list is that a depression was prevented (Stateside!) by the actions of former Fed Reserve Ben Bernanke (who was the leading expert on the great depression and who followed to a considerable degree the policy relevant morals of pre-existing research.) I return to the failures of (a-b) shortly--these are both failures of "prophecy," they need not involve failures of explanation or understanding (or abuses of mathematics).
Without mentioning it nor irony (nor historical self-consciousness), Levinovitz re-activates a familiar Meme: the hyper-mathematization of economics (recall this piece in the New York Times by Tyler Curtain and Alex Rosenberg; also recall my response). The modern version of the Meme goes back to Boulding's response to Samuelson Foundations [recall my treatment]--the book that self-consciously tried to generate a mathematical revolution in economics in the 1940s. (No, I am not claiming Samuelson invented mathematical economics as such, but he did develop the modern edifice and the disciplinary self-understanding that accompanied it.) I mention this, rather, to point out that when earlier generations of economists debated the question in public they were, in fact, very interesting on the issue. But by now the Meme gets recycled without distinctions that may make the conversation illuminating.
Much of the piece is about the abuse(s) of financial economics. But we get no analysis of the different uses and kinds of mathematics in it: there is a huge difference among (d) so-called, ‘Technical analysts at big banks...trying to find patterns," in stock-market movements, which is thoroughly discredited by professional economics (already since the 1960s), and, say, (e) portfolio allocation models, (f) rational pricing techniques (like Black-Scholes), and (g) bankers' risk-models (even though some of the same underlying principles are used in e-g). The character of the mathematics and the models used in (e-g) is, in turn, very different from the models and theories used in (a-b). And the reasons for their failure are different in kind from the failure of risk-models.
For example, I have studied the failure to forecast the economic crisis of 2009 by the Dutch planning agency. And the explanation for it is fairly simple: they did not (and perhaps could not) foresee the collapse in world trade (as a consequence of a liquidity crisis post Lehman)-that's no trivial matter in an economy in which the variance in growth is extremely dependent on trade volumes. It took the planning agency six months to catch up with reality. To solve this problem, they do not need to abolish the use of mathematics, but they need to get much more creative in having access to relevant trade-data in real time. (I am not suggesting there will not be other problems--I am a big defender of the significance of uncertainty, after all.)
This is not say that Levinovitz's piece is all rubbish. We learn quite a bit about the "the enchanting force of mathematics," especially, its role in facilitating the huge social investment in "li production in ancient China" that is, "the science of calendrics" (or astrology/astronomy). But Levinovitz's underlying point that 'mathematics' is excessive status enhancing is a familiar one. For, as I have noted, Mandeville already made the point about the social, non-epistemic role of mathematics in eighteenth century medicine (Mandeville was an insider, a trained and practicing physician): "The reason, therefore, why the mathematics are so highly recommended to all young students in general, is not so much the utility they are of in their studies, and to understand the business they are to follow, as that they are a modish science, the knowledge of which thought to be a fine accomplishment." In particular, "to ingratiate themselves with the Publick; and that they hope from it to be sooner trusted with sick People, than they would be, if it was known that thy had never applied themselves to that science." That is, the utility of math in eighteenth century medicine is a signaling device that makes mathematical medical doctors more attractive in the market place. (There are some fantastic observations about this in Mandeville's piece.) Levinovitz's piece is, in fact, framed with a series of complaints ("wasting our money") about the high value(s)/incomes of economists in various market places. (Again, he conflates these.)
Mandeville also anticipates Levinovitz's point that the use of mathematics (now quoting Levinovitz) "creates a high barrier to entry for those who want to participate in the professional dialogue." More recently, Sandra Peart and David Levy, two creative economists that often fuel my ideas about these matters, have investigated the ways in which economics and finance are represented as alchemy and subtly rely on selling themselves as alchemy (see here, and pay attention to the cartoons).
But it is by no means obvious why the use of mathematics "makes checking someone’s work excessively laborious." The problem here is the economists' history of lack of norms on sharing data (and other forms of transparency) and, more general, lack of incentives to check each other's results and publish dis-confirmations (this may be changing). It is certainly not obvious why the use of mathematics (an abstract, non empirical tool) would imbue "economic theory with unearned empirical authority." (I leave aside that at one point in the piece the evils of measurement and mathematics are conflated.)
In conclusion, one may be tempted to say that I have missed Levinovitz's point. He claims that that economists have become modern high priests, which because they do so bad at predicting future events,"fail at prophecy." Surely, I cannot object to his position? But I do because it also misunderstands the nature of prophecy. Prophecy is not prediction (recall this post). Here's an argument from authority: as Hobbes notes, the Hebrew Bible recognizes false prophets are also capable of accurate prediction--Balaam is the exemplar in this category. The previous sentence is a partial joke, of course. The more important point is that social prophecy is only tangentially related to prediction. The sooner we understand this, the more sophisticated our conversation about the possible roles of reflexive social science in our societies can be.