I love microfoundations. Just not yours.

Those of you who love to follow blog debates about macro methodology - and I'm assuming that's pretty much all of you - should check out the debate between Tony Yates and Simon Wren-Lewis. The debate started on Twitter, and then Yates wrote a long post about why microfoundations are the bee's knees (I think that means "good" in British).

Yates essentially recites the standard catechism: the Lucas Critique is really important, the '70s inflation proved it, SVaRs are a sort-of-acceptable alternative, New Keynesian models are OK except Calvo pricing is suspect, etc. etc. If you want to know how the average Freshwater-y DSGE-slinging macroeconomist thinks about his place in the cosmos, read Yates' post. Yates also tosses in this gem:

There’s something irksome about defending micro-founded macro from the attack that it is ‘without merit’.  A voice inside me says:  if they aren’t doing macro, by which I mean, generating new empirical or theoretical work themselves, who are they to go about proclaiming whether something has merit or not, or how macro should be done?  [I'm not singling out Adam here.  Lots are at it.]  And why should anyone care what they say?

So the only people qualified to judge the value of an activity are the people being paid by the government to do it, eh? How convenient. Snark snark.

(OK OK I'm being mean with selective quoting. Yates dutifully follows up with this quote: "[T]here’s the risk that we come to seem like a cult bent on disengaging, concerned to interact with those outside the cult only so far as is necessary to squeeze them for the money we need to continue playing with our toys.")

Anyway, it was kind of cute that Yates singled out Calvo pricing as an unrealistic, kludgey, hold-your-nose sort of microfoundation. Oh, it certainly is that. But it's hardly unique in that respect! In the Lagos-Wright (2005) model that is the basis of "New Monetarist" macro, for example, people exchange goods with anonymous counterparties with whom they are randomly matched and will never meet again. That's every bit as unrealistic as Calvo pricing, but you don't hear Freshwater guys like Yates kvetching about that, or any of the other equally unrealistic mechanisms in RBC-type models. Why does Calvo pricing receive special, negative attention? My instinctive guess is that it's because unlike many other equally silly microfoundations, Calvo pricing tends to justify a role for government intervention.

But I digress. On to Simon Wren-Lewis' response, which was really quite excellent, although - in typical New Keynesian form - he goes out of his way to be deferential to his hard-charging Freshwater counterpary. He prefers an "eclectic" approach, where microfounded DSGE models are used alongside other types of models. This is pretty much what the Fed does (and the Bank of England, and the Bank of Japan, etc.). Wren-Lewis describes a situation in which such an approach might be better than relying solely on DSGE models:

Suppose in the real world some consumers are credit constrained, while others are infinitely lived intertemporal optimisers. A microfoundation modeller assumes that all consumers are the latter. An eclectic modeller, on finding that consumption shows excess sensitivity to changes in current income, adds a term in current income into their aggregate consumption function... 

Now a microfoundation modeller might respond that the right thing to do in these circumstances is to microfound these credit constrained consumers. But that just misses the point...[E]ven the best available microfounded model will be misspecified, and an eclectic approach that uses information provided by the data alongside some theory may pick up these misspecifications, and therefore do better. 

Another response might be that we know for sure that the eclectic model will be wrong, because...it will fail the Lucas critique[.]...But we also know that the microfounded model will be wrong, because it will not have the right microfoundations. The eclectic model may be subject to the Lucas critique, but it may also - by taking more account of the data than the microfounded model - avoid some of the specification errors of the microfounded model. There is no way of knowing which errors matter more. (emphasis mine)

Wren-Lewis absolutely nails it. In the comments to Wren-Lewis' post, I wrote "YES YES A THOUSAND TIMES YES". In a follow-up post, Yates caricatures my comment as "NO NO GET RID OF ALL THE MOTHER&&&&&&G MICROFOUNDATIONS WHILE YOU ARE AT IT." But let us ignore that particular flerp-o'-derp for now, and focus on why Wren-Lewis is so very very right.

The Lucas Critique was important and right. If you don't take into account how your policy will change the workings of your model, you can't know what effect your policy will have. So if you use a model that doesn't satisfy the Lucas Critique (not that anyone knows what really satisfies the Lucas Critique!), you're going to introduce some errors into your policy recommendation. Potentially some very big errors.

But some macro guys seem to walk around thinking that Lucas Critique errors are the only kind of errors a model could possibly make! This is, of course, not the case. If your microfoundations are wrong, then you will introduce misspecification errors. As Wren-Lewis points out, those errors might also be very big!

Brad DeLong made a similar point in another excellent post a couple of weeks back:

[It's] dangerous to presume that you understood something because you had “microfoundations” when those microfoundations were wrong. After all, Ptolemy had microfoundations: Mercury moved more rapidly than Saturn because the Angel of Mercury left his wings more rapidly than the Angel of Saturn[.]

In fact, I encourage everyone to read that post in full.

But back to Yates. Yates says I just want to get rid of all the microfoundations. But that is precisely, exactly, 180 degrees wrong! I think microfoundations are a great idea! I think they're the dog's bollocks! I think that macro time-series data is so uninformative that microfoundations are our only hope for really figuring out the macroeconomy. I think Robert Lucas was 100% on the right track when he called for us to use microfounded models.

But that's precisely why I want us to get the microfoundations right. Many of microfoundations we use now (not all, but many) are just wrong. Obviously, clearly wrong. Lots of microeconomists I talk to agree with me about that. And lately I've been talking to some pretty prominent macroeconomists who agree as well. 

So I applaud the macroeconomists who are working on trying to develop models with better microfoundations (here is a good example). Hopefully the humble stuff I'm doing in finance can lead to some better microfoundations too. And in the meantime I'm also happy to sit here and toss bombs at people who think the microfoundations we have are good enough!


Updates:

Just to hammer the point home, see this excellent post by Antonio Fatas, about what macro models need to include that they (mostly) currently omit.

Also, read this additional DeLong post about the tragedy of misspecified microfoundations.