One or two of the responses seemed to be arguing against the title of my post, rather than the contents. That's understandable, since titles are important. In this case, though, it probably detracted from the debate a great deal. The Bloomberg title people are good, and they usually get things right, but once in a while the title they choose doesn't quite capture the point I'm trying to make. This was one of those cases. The title they gave my post was "Economics Without Math Is Trendy, But It Doesn't Add Up." But actually, this wasn't what I was arguing. My point about non-mathy models wasn't that these are bad, useless, or inferior. It was that they're different from mathy models, and so comparing non-mathy models with mathy ones is an apples-to-oranges comparison. Both types of models have their uses, but you can't really compare one to the other. I make that pretty clear in the text of my post, but most of the people who responded tended to focus more on the title. Oh well. These things happen.
Anyway, on to some of the responses. The numbering here is arbitrary, corresponding to the order in which the tabs were open on my browser.
Response 1: Frances Coppola
First, here's a response by Frances Coppola, whom I cited in my post. After calling my post "arrogant," Frances claimed that I made the following claims about heterodox macroeconomics:
- it is not mathematical
- it is too vague to be any use for forecasting
- the models it uses are just revamped mainstream ones
- it over-fits data to its models, giving spurious results
- it hasn’t managed to produce any reliable results
This isn't a very good reading of my post.
1. I certainly said *some* heterodox macro is non-mathematical, which is true. But I gave some examples of mathematical heterodox models. In fact, it would be hard for me to make the penultimate claim in this list if I had made the first one, since you can't overfit data unless you have a mathematical model!
2. I didn't say heterodox macro was too vague to be used for forecasting, only that non-math models are too vague to be used for forecasting. Again, Frances has misread my post.
3. I certainly never claimed that heterodox models are just revamped mainstream ones; I have no idea where Frances got that.
4. I did say that SFC models, which tend to have a large number of parameters, are vulnerable to overfitting; this is obviously true, and is a well-known problem in all such many-equation models.
5. Regarding reliable results, I said only that agent-based modeling hasn't yet produced reliable results, which to the best of my knowledge is true. (Note that I also think mainstream models haven't yet produced any reliable results, at least in terms of policy analysis.)
So Frances appears not to have understood most of the points I was trying to make. That's a shame, and it's my fault; I should have explained my points better.
Frances also has a second post, giving her views about the virtues of non-mathy vs. mathy models. I'll think about that and maybe write something about it later.
Response 2: Steve Keen
Next, we have a response by Steve Keen. Steve, unlike others, did get the point I was making about mathy vs. non-mathy models (Thanks, Steve!), and had some good commentary on the subject:
There is indeed a wing of heterodox economics that is anti-mathematical. Known as "Critical Realism" and centred on the work of Tony Lawson at Cambridge UK, it attributes the failings of economics to the use of mathematics itself...
What Noah might not know is that many heterodox economists are critical of this approach as well. In response to a paper by Lawson that effectively defined "Neoclassical" economics as any economics that made use of mathematics (which would define me as a Neoclassical!), Jamie Morgan edited a book of replies to Lawson entitled What is Neoclassical Economics? (including a chapter by me). While the authors agreed with Lawson's primary point that economics has suffered from favouring apparent mathematical elegance above realism, several of us asserted that mathematical analysis is needed in economics, if only for the reason that Noah gave in his article[.]Steve also offers some useful criticism of Milton Friedman's ideas about how to evaluate a model's empirical success (I agree).
Steve also makes the useful point that linearization critically hampers many mainstream models (I agree).
Steve points out that non-mathy models can make qualitative forecasts. That's true. However, my point was that these are often a lot less actionable than quantitative forecasts. A non-mathy model might tell you that private-sector debt is dangerous, but it might not tell you how much of it is dangerous, or how dangerous. For that, you'd need some kind of mathy model. Steve definitely seems to get this point too, though, so I'm not disagreeing.
Steve then discusses overfitting of data, and points out that many mainstream models do this too. That's certainly true, although I think DSGE models tend to be a lot more parsimonious than SFC models or stuff like FRB/US. Actually, overfitting is one of the big criticisms of the most popular DSGE models in use at central banks.
Steve then addresses the idea that heterodox models are similar to mainstream ones. I never said they were, although I said there are some similarities between the FRB/US model and Wynne Godley-type SFC models. In fact, there are some similarities, though there are also differences. But in general, most heterodox models are very different from most mainstream models.
Steve also discusses my (admittedly too brief) mention of agent-based models, and has some good comments:
Largely speaking, this is true - if you want to use these models for macroeconomic forecasting. But they are useful for illustrating an issue that the mainstream avoids: "emergent properties". A population, even of very similar entities, can generate results that can't be extrapolated from the properties of any one entity taken in isolation...Neoclassical economists unintentionally proved this about isolated consumers as well, in what is known as the Sonnenschein-Mantel-Debreu theorem. But they have sidestepped its results ever since...Multi-agent modelling may not lead to a new policy-oriented theory of macroeconomics. But it acquaints those who do it with the phenomenon of emergent properties - that an aggregate does not function as a scaled-up version of the entities that comprise it. That's a lesson that Neoclassical economists still haven't absorbed.I think this is right. Agent-based models have so far served as a demonstration of the fragility of representative agent models. In the future, they might be much more than that.
So anyway, I'd say I pretty much agree with Steve's response. Good stuff.
Response 3: Ari Andricopolous
Ari has a response as well. His response comes in the form of a list of things that he thinks macro models should not include. The list is:
- Microfoundations
- Rationality
- Loanable funds
- Interest rate effects
- The financial sector
It's pretty clear that the last item on this list is misplaced, since Ari thinks one should include the financial sector in models.
Whether macro models should be microfounded is a big open question, but I'd like to note that by saying they shouldn't be, Ari is saying that agent-based models are bad. Agent-based models are as microfounded as they come.
As for rationality, I kind of disagree...humans observe and learn and adapt (OK, some more than others, I'll grant). Even though perfect rationality is probably pretty unrealistic, to insist that models totally ignore human observation, learning, and adaptation seems very dangerous for the realism of any model.
As for the loanable funds thing...yeah, OK, sure.
Response 4: Jo Michell
Jo Michell's response might have been the first to go up, but it's #4 on this list because...the numbering is arbitrary!
Jo, which I believe is short for "Jörmungandr", has a helpful diagram of the "schools" of macroeconomic thought. He also pushes back on the notion that "heterodox" is a useful classification at all:
The problem with ‘heterodox economics’ is that it is self-definition in terms of the other. It says ‘we are not them’ — but says nothing about what we are. This is because it includes everything outside of the mainstream, from reasonably well-defined and coherent schools of thought such as Post Keynesians, Marxists and Austrians, to much more nebulous and ill-defined discontents of all hues. To put it bluntly, a broad definition of ‘people who disagree with mainstream economics’ is going to include a lot of cranks. People will place the boundary between serious non-mainstream economists and cranks differently, depending on their perspective.
Another problem is that these schools of thought have fundamental differences. Aside from rejecting standard neoclassical economics, the Marxists and the Austrians don’t have a great deal in common.This is a good and useful point. My Bloomberg post really did bite off more than it could chew. My point was that there wasn't something better and more successful out there that by rights ought to already have displaced the (unsuccessful) mainstream approach. But in making that point, I touched on a number of different types of alternatives that aren't really closely connected. And I left out others (for example, Steve Keen's own work, and the Austrians).
Jo was, unfortunately, also tripped up by the title:
Noah seems to define heterodox economics as ‘non-mathematical’ economics. This is inaccurate. There is much formal modelling outside of the mainstream.Well, no, I don't define it that way, otherwise I wouldn't have discussed SFC models and agent-based models in my post.
Jo goes on to make some good points about mainstream models, and some of the problems they encounter.
Anyway, those are the four responses I've seen. Thanks to everyone who took the time to respond!
from Noahpinion http://ift.tt/2bCF7WT