A liberating exit

Jebus Christ, I didn’t realize it’s that bad, sounds almost Lysenko-like, really.

The fact that the debate is being carried on in terms of orthodoxy vs heterodoxy is evidence of a perverse incentive system for innovation in the careers of young scholars in the field. Economists have a few tropes that they favor in discussions like this, and it would be better for them if their images of open competition, contestibility or creative destruction predominated on this point, rather than the panglossian tendency to think that everything is efficiently working out for the best and that success is ipso facto evidence of rightness.

#4: Am I just really missing Ingrid’s point, or is this post not about methods rather than political orientation / policy implications?

Actually, 5 is right. There is more than one issue here, though.

“Since sociology is a branch of biology, its ultimate expression must be mathematical.”

That makes my head all kinds of hurt.

Kieran:

Isn’t it Colin McEvedy in the introduction to the
“Penguin Atlas of Ancient History”? But he says it’s *history* which must ultimately have a mathematical expression, not sociology. Or were you joking?

The first of notsneaky (radek’s?) arguments holds, sort of, but it’s extremely limited. Boiled down, it says that mathematics provides a kind of conceptual book-keeping. This is not to be sneezed at (it’s one of the reasons I like game theory), but it certainly doesn’t provide reason for rejecting non-mathematical economists’ work _tout court._ The second seems to me to be extremely dubious. Nothing about the actual economy makes it ‘mathematical’ – some features of it are mathematically tractable, others, that may be equally or more interesting ain’t.

Which suggest that the statement that economics is by its nature mathematical is a tautology, and one covering over a set of power relations. We won, so tough isn’t an intellectually serious argument. Nor is the “I don’t quite understand why the field (and the field is composed of people, individuals, after all) should have an obligation to bend its strutures and methadology to the whim of everyone who thinks they know “the true path”” much cop. It turns what critics like Ingrid are saying topsy turvy – Ingrid et al. specifically disclaim any monopoly on the truth; all that they are asking for is that mathematical economists be prepared to admit that there are limits to what they do, and alternative ways of doing things that can get at interesting features of economic and social life. To put it differently, there is someone in this argument who is apparently convinced that they possess the One True Path, but it sure ain’t Ingrid. There’s something deeply weird about the fetishization of mathematics in notsneaky’s post (and his subsequent bit about respect only for mathematically sophisticated heterodox economists). Math is great and important and all, but it’s one conceptual tool among many. It doesn’t provide the keys to the kingdom.

I’m intrigued by this discussion, but its lack of concreteness makes it hard for me, as a non-economist, to follow. Poking around in Ingrid’s downloadable publications (why aren’t all her publications downloadable?), I came across her tutorial on the “capability approach”: a way of thinking about utility, happiness, or “goods and services”. If economics is the study of flows or interactions of “goods and services”, it seems reasonable to have a clear idea of just exactly what those things are. So the study of the “capability approach” is a concrete example of an area of work that is concerned neither with predictive mathematical models, nor with quantitative measurements, but with the foundations of economics. Is it “heterodox”?

Is there a really a better description of modern economy (ok, political economy) than “The New Industrial State.” I had to force myself through the macro grad sequence-mostly RBC stuff- and I don’t think RBC has 1% of the understanding of the modern economy as New Industrial State. This shows that Mathematical models can help good economics, but a sharp enough mind can do without it. And lots of mathemaical economics obscures more than illuminates. How many times have you worked through a super-fancy model only to realize that all the machinery obscured some really dubious assumptions or very simplistic decisions. All the time, and since economists don’t value writing, in general the writing is too horrible to explain the model.

notsneaky’s comment is bunk – and I say that as an economist whose work does fit into the normal tools.

The mathematical argument might be valid if the mathematics was up to expressing anything but the most simple ideas. Human life (including the economy) is too complicated for mathematics. The better part of economics is taking a simple idea and making it seem complicated by writing in mathematical expressions. And of course, there is the old problem with deductive reasoning. You can get any result you want. Just choose your premises carefully.

(radek’s?)

Yes.

And seriously, what’s up with all these ridiculous demands that “defenders of orthodoxy” show themselves? What, is someone compiliing lists for after the revolution comes? Nobody seems to have a problems with the “defenders of heterodoxy” – commentators at least – remaining anonymous while directing vague and not so vague insults at their opponents (we’re sexists because we do math, we’re paid shills of corporate interests, the only reason we do what we do is because we want to get tenure, we hold secret meetings in caverns where we decide which heterodox economist we’re gonna hunt down next and roast on a spit, etc.)

Harry b, I don’t see how I’m any more damn anonymous than you. Unless of course your last name does happen to be “b”.

This is not to be sneezed at (it’s one of the reasons I like game theory), but it certainly doesn’t provide reason for rejecting non-mathematical economists’ work tout court.

Perhaps my frustration and annoyance with much of this debate is starting to show. Of course non-math econ shouldn’t be rejected tout court. As an undergrad I did history and I can appreciate a non-math approach. I’m not saying that mathematical economics is the only way to go. But I am saying it is the best way to go. JKG may be a great writer, and a genius to such a degree that he doesn’t need no stinkin’ equations to understand economics. I’m not that, and neither is your average practitioner. There’s been plenty of time when you think you have a good idea about how things function but when you sit down and try to actually write it down you realize it doesn’t fit together.

Here also I’m gonna open up another can of worms, and say that the main benefit of verbal expositions is that they suggest ideas, areas of potential research, and possible puzzles to be thought about. But at the end of the day you want to sit down, write a formal model and make sure the story actually clicks – and in the process you also eluminate the implicit assumptions that the writer is making, perhaps unconciously.

Marshall, Jevons, Schelling, Coase: imprecise, no equations, no truth

Exceptions! And there’s math (even if they hide it) in all three of those authors – and once you do it you see much more than when you just read the words (I think this is most obvious with Jevons). And sure, if there was no math progress, innovative research, etc. would still happen. But it would be a lot slower and half the time would be spent on asking each other “what do you really mean?” (which isn’t always a waste of time but often it is – jeez I’m getting so defensive here I want to qualify everything)

The mathematical argument might be valid if the mathematics was up to expressing anything but the most simple ideas.

I believe the problem with simplicity here isn’t mathematics.

Human life (including the economy) is too complicated for mathematics.

Human life is too complicated for language! You think human language adequately captures the complexity of human life?

This is actually a typical argument that has frequently been made in these discussions:
Step 1: But humans aren’t rational!!! Life is too complex!!!
Step 2: …
Step 3: Socialism!!!

(I’m exaggerating to make a point, you know)

As far as whether economics is “intrinsically” mathematical. It is in the sense that a lot of variables that we care about are numerical in nature, quantifiable. Some are less so. But of course that’s just accounting not mathematics.
Further, economics is a social science which means that it deals with human relations. And equations are convenient, precise ways of expressing relations between variables and also force you to put your cards on the table. And since, apparantly, life is “complex” you want some way to keep track of how all these variables affect each other, how one phenomenon leads to another, which then leads to another, which leads to another, which feeds back into the second, etc. etc. etc. This makes the written langugage very inadequate for the task. But if you write as a system of equations, or propositions, or relationships, monkey around with it a bit, then patterns emerge, things become clearer and you understand way more than if you had written thousands of confused sentences.
Put it this way, if I could somehow express this whole post in a series of equations, rather than this long winded comment, wouldn’t it be better?

Finally,
(and his subsequent bit about respect only for mathematically sophisticated heterodox economists)

See, this is how language fails us. You added the word “only”, which I did not include. If we were doing this in math this wouldn’t be a problem: MS/HE C ERbyNS =/> NMS/HE upside-down-U ERbyNS = empty set.
(The fact that my keyboard has all 26 letters of the alphabet and all the (mostly useless) punctuation marks, but is extremely lacking in standard mathematical symbols is prima facie evidence of the oppresion the nerds have endured, and are still enduring from the literati)

I’ve read McCloskey. Pretty much all of it. It’s good writing and there’s some good insight in it but a lot of claims she makes are just exaggerated or attack strawmen.

Oh yeah, this sort of reminds me (because he’s mentioned in some of the McCloskey books). An essentially non-mathematical economists – he definetly knows it, but uses it very sparingly – that I respect tremendously is the economic historian Joel Mokyr, though I guess one would not consider him heterodox.

“Mathematical work keeps one precise and consistent.” Whenever I hear this sort of thing I have to chuckle.

I work in a very technical social science department (political science) chock full of folks every much as testosterone driven as the economists one flight down (and so committed to bizarre game theoretic and economtric techniques without ever discussing their point). I find the rhetoric of ‘precision’ and ‘rigor’ and ‘consistency’ that my colleagues bandy about to be something of a howler. In everyday language we have tools for being more precise; we have, for instance, things called adjectives and adverbs. And they tend to make things more specific, particular, concrete … and less abstract. Mathematics simply allows you to make things abstract. But as Tom Schelling makes clear repeatedly we need to justify – in (gasp!) words – why we are abstracting from this or that, treating it as ‘incidental detail’ to be cast aside in the model building process. Mathematical abstractions may or may not be useful. As with verbal modifiers, it depends.

As for consistency it is easy enough to be consistent in plain prose. One simply has to pay careful attention to what one is writing. More economists should try it. (And,as a journal editor, more political scientists should too!)

The problem with mathematical techniques in social science is that they are a sometimes useful tool that folks somehow inflate into something more – what, I do not know. So let’s take a real live high-tech economic theorist like, say, Ariel Rubinstein. He thinks mathematical models in economics are virtually useless without (gasp!) verbal interpretations. He is right. And the uselessness goes all the way down to the basics – like, say, the notion of a strategy. He also thinks that the claim that economic models are “empirical” is a bit of a howler and I suspect he is correct.

Having done a fair bit of economics in different modes, I find I’m a bit torn. On the one hand, there are far too many exercises that amount to little more than PhD variation on mathematical models with no real relationship to anything. And while maths can help to make discussion more rigorous it can also obscure what’s really going on.

On the other hand, the core concerns of economics (prices, output volumes, incomes, hours of work) and so on are inherently quantitative. There are good reasons why economists are more keen on mathematics than other social scientists, and why applications of economic methods in other fields often turn out badly.

To take the example of feminist economics, the gender gap in wages is probably the biggest single issue in this field, and it’s virtually impossible to address this without fancy econometrics (having worked in a department where this was a cottage industry, I got to the point where if I saw another Heckman equation I would scream, but it was important work),

21: …all the (mostly useless) punctuation marks…”
If you don’t know what they’re for, couldn’t you just ask someone?

Particularly if, like me, you are coming from a leftwing point of view in political terms, this is important in disarming the claim that your arguments just show that you don’t understand economics.

I sympathise but I also find this problematic. It’s almost as if, as an atheist, I felt I had to enroll in a Catholic seminary, because if I didn’t nobody would take seriously my arguments against Theism.

Step 1: But humans aren’t rational Life is too complex
Step 2: …
Step 3: Capitalism

Isn’t that Hayek in a nutshell?

also, notsneaky, it was clear to all other readers here that Ingrid’s post and the letter she contributed to did not say no mathematical work was worthwhile, but just that some non-mathematical work was. Language can be amazingly precise if you care to read precisely.

Economics does seem to suffer from an unwarranted hubris (that’s probably a redundant expression) when it comes to the mathematical rigour of the discipline. Language is also a form of logical expression and maths is, after all, only another language.

Second, economics is by nature mathematical. Wages, unemployment levels, growth rates, income distributions. These are all numbers, equations and functions

This is unserious though. Wages, unemployment levels, growth rates and income distributions can all be talked about with nothing more sophisticated than GCSE maths. Galbraith’s books are full of numbers. When people talk about “mathematics” in economics (as opposed to econometrics, which IMO is not really the sort of thing that can be orthodox or heterodox), they mean a specific mathematical toolkit, and nine times out of ten, they mean something that has dynamic programming at its heart. It helps to make a model, but I would be much less confident that it helps to make the same kind of model of more or less everything, changing the names and parameters around but leaving the structure the same.

Proof of the pudding here would be someone like Barkley Rosser. He’s certainly no slouch at mathematics. But he’s also firmly in the heterodox camp, in the sense of “someone who can’t get his papers published in the top journals”. If this was just a matter of model-building, the need for rigour or “nerds versus literati”, then complexity would have been lapped up by the economists. Instead, they just basically said “ahhhh well chaos theory never amounted to much” and continued on.

Actually, thinking about it, econometricians ought to be a lot more embarrassed than they are at the length of time they took to get used to Bayesian methods, long after the rest of statistics had adopted the toolkit.

Tony Lawson is the name that comes to my mind for that particular charge (he writes a lot of very interesting stuff himself, but he stands as metonym in my mind for the gruesome fascination of a lot of the heterodox space with windy methodological debates).

Ingrid:”I really don’t know. I don’t know enough about “computational work” in economics to know how it’s being treated in the profession. Do you?”

I only know some such work exists.

Dynamic programming arises in Macro with the whole DSGE approach

you talk about this as if it was just a minor detail.

And macro as a whole has been on the decline

You talk about this as if it was just a minor detail.

I mean come on man. The discipline of economics has basically given up on the attempt to understand the aggregate system of production, distribution and exchange. What has the orthodox approach delivered, exactly, to counterbalance this monstrous surrender

48. I know what they’re for. That’s different than ’em being mostly useless.

So… You can’t to understand their uses; instead of learning what they might be, you dismiss “’em” out of hand? You’re a card, Radek!

I agree that these kinds of problems happen in all disciplines, but from the ones that I have a basic knowledge, economics is much, much, much worse than any other. In philosophy, for example, there are at least two broad streams (‘analytical’ and ‘continental’); but there are ample job & publication & recognition & funding possibilities for BOTH groups.

Well, if there aren’t also at least a half-dozen alternative streams (or sub-streams within those two main streams) whose embittered adherents can’t get a second look at a top school, then philosophy is a very unusual discipline indeed.

it may be a good solution for one individual (me), but what about (a) all the others,

Most of them, as I said, are doomed, unless they knuckle under and accept the orthodoxy. A few will follow your path, though, and find sanctuary as you did. You can do your part by encouraging them to do so. And perhaps one day a particularly gifted and industrious scholar (or a small coterie of them) will eventually succeed in getting a new, “heterodox” stream of thought established as mainstream.

and (b) the effects on the discipline (and consequentally its influence in society etc.) ?

There are three obvious possibilities: the current orthodoxy will be found in the long run to be worthwhile after all, and will thrive; or it will be found to be a dead end, and a new orthodoxy (perhaps yours, perhaps somebody else’s) will take over; or it will be found to be a dead end, and the entire field will decay into irrelevance, while the useful work is done elsewhere–in your department, perhaps.

But again, I offer this caution: before you let your perfectly understandable pot-shot-taking escalate into a genuine, vigorous campaign for “reform” of your former discipline from the outside, consider that they may have similar designs on your current discipline, and ask yourself just what kind of precedent you really want to set regarding external criticism of an established field…

Since economists freely attack all other social sciences already, and openly admit they plan on supplanting them, I don’t think anyone has to be too concerned with the reaction of economists to outside criticism.

Your suggestion that people who identify with one discipline should not criticize practices elsewhere lest they be criticized themselves is a recipe for mediocrity. Social science is a single project.

I strongly disagree. As I’ve already explained–and as everyone here seems to recognize–academic disciplines are very good at expunging methodologically flawed work, but not so good at fostering new and legitimate approaches. The boundaries between disciplines are one of the very few protections available to someone (like Ingrid, for example) who wants to buck the conventions of a particular discipline. If those come down, then there will be one narrow set of standards for, say, all of social science–and woe betide a future Ingrid with a new and different approach.

Since economists freely attack all other social sciences already, and openly admit they plan on supplanting them, I don’t think anyone has to be too concerned with the reaction of economists to outside criticism.

Criticism–of course. If Ingrid, or anyone else, wants to predict that today’s economics research is all destined for the dustbin of history, that’s fine too–healthy, even–and we’ll know soon enough anyway who’s right. As for reforming economics from the outside, though (as opposed to hoping to “supplant” them once they’ve decayed into irrelevance), bear in mind that if the economists tried to retaliate, they’d likely have the entirety of the hard sciences backing them up…

Dan, the point was not that certain streams were being ignored, but that ALL streams but one are being ignored. There’s no other social science for which this is true.

One person’s “ALL streams but one are being ignored” is another’s “ALL streams but one are welcomed”. It’s a question of which collection of research directions one chooses to lump together as one undifferentiated pile of garbage, and which collection of research directions one respectfully subdivides into equally valid but distinct approaches.

Dan: You seriously think that? Whenever hard scientists encounter economics, they think economists are insane. You can’t imagine the contempt that physicists have for economics.

the point was not that certain streams were being ignored, but that ALL streams but one are being ignored. There’s no other social science for which this is true.

But this is really a matter of definitions. If you emphasize the similarities between the various streams within the “mainstream” and lump’em all together then you can say that all but one stream get ignored. If you emphasize the differences within the “mainstream” then you can say that many different streams are represented. So it’s a matter of taxonomy, which is at least somewhat subjective.

Its not language versus math that’s the big problem its that we don’t have any math or words that effectively theorize how words or concepts are defined. Its all circular and inexplicable.

No effective theory of language, no effective social science.