The excerpt from Ophelia Benson’s article which Chris posted below got me thinking about a few particularly egregious examples of the phenomenon I’ve seen over the years. The one which sticks out in my mind was of a teacher proudly boasting that he’d spent half of a class ignoring the subject matter that was meant to be discussed and instead talking about technical arcana which added nothing to our understanding of the subject, made the discussion incomprehensible to the layman, but fitted the students to carry on a discussion among people working in the same field, according to the rules of a trivial formal game.
Hang your head in shame, Brad DeLong, for the following piece of wilful obscurantism and “Bad Writing”
“On February 12 I taught Andrei Shleifer’s “Implementation Cycles” (Journal of Political Economy, 1986) paper to my advanced macroeconomics Ph.D. students class.
Once again, just has happened the week before, I didn’t get through the paper. I had thought it would be easy—that I would finish with plenty of time to sketch extensions and qualifications. After all, the class does run from 12 to 2. However, that was not enough time.
My recurrent problem is that I spend so much time in asides on the modeling strategy—”this term is in this definition because twenty minutes from now it will cancel that when we take a derivative to establish the first-order condition”, “note that even though we have started with a rather general and flexible setup in which firms have a number of different decisions to make, the setup has been carefully designed so that when push comes to shove there is only one economically interesting and non-obvious decision a firm ever has to make”, “note that if this condition is not satisfied, then the consumer’s utility is infinite and it is not clear in what sense we can say that the model even has an equilibrium at all,” that kind of thing. These asides on modeling strategy take up a surprisingly large amount of time. Yet I don’t think I can cut them out or even cut them down. After all, if I don’t teach them this, when will they ever learn it?”
Pretty egregious, huh? After all, the essential insights of this model are clear. Brad even summarises them himself in the same post:
- A lot of firms implementing their new technologies at once creates a boom.
- An aggregate demand externality makes it profitable to cluster the implementation of new technologies. A firm with a new technology wants to wait until a boom to implement it because its technology will be quickly copied—its edge is temporary—and it is more profitable to implement when the economy is booming and demand is high than when the economy is not booming and demand is low.
- Partly offsetting this is the fact that the interest rate is high when a boom is expected, and thus the cost of waiting until a boom to implement your technological innovation can be substantial. (In fact with log utility or a utility function less risk-averse than log utility, the interest rate effect always outweighs the aggregate demand externality effect).
- Thus the model can exhibit periodic “implementation cycles” in which technological advance is delayed until periodic booms.
- However, in the basic model households always want new technology to be implemented as fast as possible: these implementation-cycle equilibria reduce welfare.
- But this can be reversed: in a version of the model with fixed costs of implementation and with ‘standing on the shoulders of giants’ effects in discovery, it is very possible that technological progress is only possible if there are periodic large booms. “
So, if it is perfectly possible to summarise the conclusions of the Shleifer implementation-cycles model of the business cycle in a few digestible bullet-points, surely it is counterproductive and unforgivable to shroud the simple underlying points in all this “Theory”, isn’t it? After all, if something is intrinsically simple, it’s ludicrous to suggest that the conclusions have to be established through convoluted language, massive generalisations and strange constructs which make no sense on the face of them, isn’t it?
Well of course, no. If the history of economic thought teaches us anything, it teaches us that people who don’t use the mathematics always, sooner or later, end up saying something badly wrong about economics. Paul Krugman has an essay on the subject with which I profoundly disagree on a number of points, but which contains one highly important truth; there are important ideas in economics which are crystal clear if you understand the mathematics and bloody hard to get your head round if you don’t.
But, of course, I’m being silly here, or at least satirical. The use of mathematics in economics isn’t the sort of Theory we’re concerned with trying to stamp out; like the War on Drugs, the War on Theory isn’t meant to touch the recreational hobbies of nice people like us. Nobody would question the right of economists to use whatever mathematical toolkit they need in order to write economics, because unlike the Bad Writing crowd, they’re using mathematics precisely in order to ensure the rigour of their analysis, not to cover up a lack of such rigour.
Well, not quite. The position that mathematics in economics is a) the best way to do economics and b) the only rigorous way to do economics can be attacked on two separate fronts. On the one hand, we have the view of Alfred Marshall and his famous rules for the use of mathematics in economic theory:
“”(1) Use mathematics as a shorthand language, rather than as an engine of inquiry. (2) Keep to them till you have done. (3) Translate into English. (4) Then illustrate by examples that are important in real life. (5) Burn the mathematics. (6) If you can’t succeed in (4), burn (3). This last I [Marshall] did often.”
This is the view of the subject in which the use of the mathematical theoretical toolkit predisposes economists toward what Schumpter called the “Ricardian vice”, after David Ricardo’s habit of building policy advice on the basis of arguments which made extensive use of deductive reasoning:
““He then piled one simplifying assumption upon another until, having really settled everything by these assumptions, he was left with only a few aggregate variables between which, given these assumptions, he set up simpler one-way relations so that, in the end, the desired results emerged almost as tautologies.”
This is exactly what Brad means when he euphemistically refers to “modelling strategy” above; the careful selection of the assumptions, not so that they are gerrymandered to give a particular result (although this too can be done, it is usually recognised by economists as the crime it is, whereas “modelling strategy” is a perfectly reputable thing to teach students), but so that they will deliver a deductive argument which has a conclusion at all. The problem being that there is one too few degrees of freedom here; you can either select assumptions so as to be realistic descriptions of behaviour, or you can select them so as to make a model soluble or tractable. Assumptions will only possess both desiderata by the purest of chance. Marshall’s approach, while honoured much more in the breach than the observance, could be seen as an attempt to avoid the Ricardian Vice. It’s the equivalent of (part of) the Bad Writing critique in the arts subjects; an exhortation to keep economic analysis to subjects that the common man can understand, in language that the common man can understand.
It gets worse for the users of mathematical theory in economics. They’re also under attack from the other flank from Econophysics. The econophysics crowd tend toward the belief that the problem with the Ricardian Vice is not the use of mathematical and deductive reasoning from assumptions per se but rather the use of particular forms of mathematical argument and insufficiently complicated assumptions. It’s always possible to fend off criticism from the likes of Robert Kuttner when they trot out the Marshallian/ Schumpeterian critique that they’re only doing so because they’re too thick to hack the maths. It’s rather more difficult to pull off that act with someone likeDidier Sornette, who predicts earthquakes for fun, or Barkley Rosser Jr who understands Goedel’s theorem and uses it, and who is one of the few people who I would trust to write a paragraph including the words “Chaos Theory”.
Faced with the twin critiques, that they use a dumbed-down, linear mathematical toolkit which is woefully inadequate to the task of accurately modelling economies, and that they use this toolkit excessively in an obscurantist fashion (linear mathematics can get really quite complicated!), how is it the case that the particular form of mathematical economics practised by Andrei Shleifer and Brad DeLong has come to dominate the methodology of the field? Surely there is some terrible organisational pathology of the academy, that it lets such Bad Writing persist in economics? The Post-Autistic Economics Network, the Bad Writing equivalent for critics of mathematical economics, certainly seem to think so; do they have a point?
Basically, no. The level of mathematics used in most printed academic journal articles is, I have come to conclude, about right. The point is this; economics is, as Deirdre McCloskey points out regularly, a form of rhetoric. At its heart, it is and has always been about the construction of a certain kind of argument, which is meant to be persuasive over human action. I state this without argument, in the knowledge that many people at work in the field believe that they are involved in a project of genuine scientific enquiry. I feel no argument of mine is ever going to carry the day on this issue, so if anyone wants to make the case for economics as a science, I’ll simply respond thus: “Sir, I gracefully concede that you yourself and your department are engaged in a value-neutral quest for scientific facts about the allocation of resources under conditions of scarcity. I apologise for having suggested otherwise. But would you at least grant me that the description ‘A form of rhetoric … the construction of arguments aimed to be persuasive over human action’ is a decent description of what all those other bastards are up to?”
And the point I’m trying to make is this; in the construction of arguments of this kind, there are certain kinds of mistake which it is fearfully easy to make. It’s easy to spot particular benefits and miss the fact that their counterparts are costs elsewhere in the system. To come up with arguments which, if true, would imply that people systematically allowed others to impoverish them without changing their behaviour. To miss the fact that your model requires the build-up of debts forever that never get repaid. Etc, etc. The bestiary of really bad economic commentary is full of all sorts of logical howlers. And the good thing about building mathematical models is that, in general, it acts as a form of double-entry book-keeping, to make sure that, if you’ve followed the rules of the game, your economic argument will not have any of these most common and most egregious flaws. It doesn’t mean that it won’t be bad or misleading for other reasons, of course, but it does mean that you’ll at least be saying something that makes sense, if only to other experts.
And I think that there is probably a generalisation here which can be extended to other fields; typically, the formal language of a discipline (its jargon) has, among its other functions, the function of making it more difficult to make the characteristic mistakes of that discipline. In economics, it’s politically convenient adding-up errors. In literary criticism …. well, I don’t know enough about criticism to be sure, but if I know properly the little bit I do know, one of the things that at least some of them are all about is careful analysis of the implicit assumptions of common language. And it strikes me as not on the face of it unreasonable to suggest that the most common mistake in this kind of analysis would be to make arguments which unconsciously rely on an unanalysed implicit assumption, and that one way to avoid this common mistake would be to adopt a formal use of language which made it more difficult to rely on the common meanings of words. So the defence of Bad Writing on the grounds that “some subjects can only be written about in unclear terms” actually encapsulates an important truth about the subject; it’s probably possible to write about the implicit assumptions of everyday terms without falling into exactly the same kind of mistake yourself, but it might take a hell of a guy to do it. Just as it is possible to write in a sensible and apolitical way about economic matters, but it takes a hell of a guy to do it. Furthermore, it’s much more difficult to write economics in a manner comprehensible to laymen (and check by hand that you’re not making the mistakes) than to write in the mathematical style (when the maths basically does half of your checking for you). So the progress of the subject at anything like its current rate depends on the ability of professionals to use the formal language when talking to each other, and to only use Good Writing when expressing ideas to a non-specialist audience which have already been judged as worthy of the extra effort.
All of which suggests to me that, as a criticism of professionals writing for professionals, the Bad Writing crowd are protesting far too much; there is a genuine place for formal language in subjects which have characteristic mistakes. I think that the sensible position in both the maths-in-economics debate and the theory-in-criticism debate is somewhere in the middle of this exchange between Krugman and James Galbraith; obscurantist theory is over-used, it’s overused specifically as a means of trying to shut outsiders out of the debate, but for a’ that, it doesn’t mean that the theory is worthless and it particularly doesn’t mean that you can ignore the theory altogether and still assume that your contribution to the discussion will be valid except by purest luck. Bad writing is writing that’s inappropriate for its content, and to assume that the same pristine clarity should mark out a postcard from Ibiza, an actuarial report and a discussion of subtext and metonymy in Buffy the Vampire Slayer is, well, bad.
 Yes yes yes, Rosser is not really an econophysicist, but he certainly associates with people who are, and he shares the same status of being an outside critic of neoclassical economics on the grounds of insufficient mathematical sophistication.