Monday, September 19, 2016

The Next New Macro

Amidst all the artillery fire, it's easy to forget that the New Classical and New Keynesian "schools" of macroeconomics were once ...well...new.  (Hence the names.)  The New Classical school arose as a response to (a) real defects in conventional macro-econometric practice and (b) the failure of conventional macro theorists to make progress towards an actual explanation of the observed behavior of modern economies.  The New Keynesian school arose as a response, taking on board both the econometric critique (yay) and the proposed solution (boo, hiss) more or less in toto.  The New Keynesians also "fixed" the most obvious defect in the New Classical models -- the fact that depressions are ruled impossible by assumption -- but did so in a manner that largely failed to advance our actual understanding of the economy (1).

The seeds of disaster, in retrospect, lay in how easily New Classical-style models could be tweaked to get Keynesian behavior.  Failing that, the New Classical-style models could never have achieved the near-monopoly position they now hold in macroeconomics (2).  And the mess in macro did not come about because some economists committed to a modelling style which turns out to yield little insight.  The mess is a consequence of the universality of this commitment -- the macro-mono-culture, if you will.

It is the danger of mono-culture that should be uppermost in our minds as we try to figure out the next "New" macro.  So much so that convergence on anything purporting to be the Next New Macro any time in the next five years (at a minimum) should be regarded with alarm.

The fact is that, right now, we do not know the way forward, and no approach, no matter how promising (or how congenial to our pre-conceptions and policy preferences) should be allowed to dominate the field until it has proven itself empirically successful.

It's important to remember that the New Classicals didn't claim that their models were successful -- only that they were more "sound" on theoretical and econometric grounds.  They were quite clear at the outset that their approach was in its infancy, and that a great deal of work remained to be done before its value could be determined (3). Why, then, did the entire discipline latch on to the New Classical approach?

In a word: panic.

Whatever the flaws in the New Classicals' positive program, their negative critique of existing econometric practice was both true and devastating.   You can't just say, "I feel in my heart that A, B, and C cause D, so here's a regression," and claim to be doing social science.  And when you're constantly saying, "Did I say 'B'?  I meant X!  A, X, and C cause D.  Also, maybe the log of B.  Here's another regression!" your credibility does not improve (4).

So imagine the plight of the working macro-economist, circa 1978 or so.  The demand for policy advice is intense -- CPI inflation had spiked to 9% by the end of the year (compared with 4.7% at the end of 1968), and unemployment seemed to have bottomed out at a floor of around 6% (compared with 3.4% in 1968).  From the President on down, your bosses want to know what the hell is going on and -- scarier still -- how to fix it.  At the same time, the terrible truth has begun to sink in:  there's no particular reason to believe that your working models actually tell you anything about how the economy works, particularly under these conditions.  (It turns out that you've actually got a pretty good idea what to do about a demand-failure depression, but this isn't that.)  Similarly, in academe, economists are scrambling to produce policy-relevant results, which is difficult with your colleagues pointing out the gaping flaws in your econometric logic.  (They literally point and laugh.  Jerks.)

So with the desperation of a drowning man, macroeconomics latched on to the first floating object to drift within reach.  Which, regrettably, then carried us all out to sea.

So here's the lesson:  don't panic.  The basic Keynesian premise ("demand matters") is firmly established.  The key policy implications (monetary policy affects the real economy as well as prices; in a demand constrained economy, public spending can increase output) are pretty much conventional wisdom, buttressed with a slew of new evidence emerging from the recent unpleasantness.  Disputes over monetary and fiscal policy are, today, largely political rather than technical (5).   The world will probably muddle through more-or-less adequately long enough for us to verify that this research agenda or that one actually leads someplace worth going.

The good news is that we have more, more detailed, higher quality, and more diverse data available than ever before.  (Example: we've only been systematically estimating the number of job openings in the U.S. for about fifteen years.  Previously, we relied on weak proxies like the volume of help-wanted advertising.  Seriously.)  So the discipline is much better prepared to weed out models that just don't work than it was forty years ago.  We just need to get out of the bad habit, acquired back when data was scarce and unreliable, of claiming that it's "too soon" to abandon "promising" models, just because they're empirically false.

I have my own notions about the best way forward, of course, and I'll post about those soon.  But I could easily be wrong and so could anyone else.  I hope we bear that in mind.  Maybe we'll be able to retire the Robert E. Lucas Jr. Award For Derailing An Entire Discipline without ever bestowing it on a second recipient.  




---------

(1) Short version:  In New Classical-style models, prices are perfectly flexible and markets (including the labor market) always clear, so there's no such thing as involuntary unemployment.  To generate depressions -- you know, like in real life -- New Keynesian models introduce some form of inflexible pricing by assumption.  (Example:  "Calvo pricing," named after economist Guillermo Calvo, assumes that firms only alter their prices when granted permission by an imaginary magical being -- traditionally known as the Calvo Fairy -- who periodically bestows such permission at random.  No, seriously; I am not making the Calvo Fairy up.)

(2) Well, in academic macroeconomics, anyway.  Professional forecasters still use Big Macro models, the large, complex systems which attempt to model the actual economy on a sector-by-sector basis.  (It turns out that Big Macro was not fundamentally "discredited" in the 1970's, as the New Classicals liked to claim, but was simply infeasible with the datasets and computers available circa 1975.  We're better at it now.)  And central banks tend to employ a mix of Big Macro, Paleo-Keynesian, and New Keynesian DSGE models.  (Though one might suspect that the function of the DSGE models at central banks is to merely prove that we can tweak a DSGE model to yield the same results as the Big Macro models and Paleo-Keynesian models.  Houdini lives!)

(3) Today, forty years later, we know the answer: nope, doesn't really work.  (And the claim of theoretical soundness was always...tenuous.)

(4) Note also that there was nothing especially "anti-Keynesian" about the critique.  It applied with equal force to Monetarism, the other theoretical school of the day.  The New Classical critique was about method, not theory.  The anti-Keynesian theoretical program slipped in quietly behind it, using the disrepute of mainstream econometrics to tar the reputation of mainstream Keynesianism.  But the two were, in fact, quite unrelated.

(5) Largely.  However.  Note to central banks: if your forecast for a variable (inflation, for example) is wrong in the same direction (e.g. too high) every quarter for more than five years...your model has technical problems.  Fix, please.




Thursday, September 15, 2016

The Microfoundations Hoax

Demolition work on the rotten edifice of "modern macroeconomics" continues apace.  The emperor, it turns out, is not merely without clothes.  Upon closer inspection, he appears to be simply an empty cardboard box with the words "Emperor Inside" scrawled across its surface in felt-tip pen.  Paul Romer's devastating critique really deserves to be the final word on the matter.   But even Paul leaves one stone unturned, an element of modern macro so transparently intellectually dishonest that it may properly be termed a hoax: its so-called "microfoundations."

No modern macro model is complete without a pean to the virtues of its own microfoundations.  It seems that the word "microfoundations" is not allowed to appear unaccompanied by at least one self-congratulatory adjective -- "careful," "well-specified," even (shudder) "rigorous."  But, as diligent readers of George Orwell will recall, war is not peace, freedom is not slavery, ignorance is not strength, and representative agent models are not rigorously microfounded.

But let's back up a step.  What is this "microfoundations" business anyway, and why should anyone not currently seeking a tenure-track appointment in econ care even a tiny bit?  Here's a short version of the very long story:

Forty years ago, the name of the game in macroeconomics wasn't theory at all; it was forecasting.  And it wasn't particularly successful.  In retrospect, the lack of success isn't surprising.  Models were typically estimated by running regressions on a handful aggregate data series representing the experience of a single country over a very short (and rather placid) period of time (1). Moreover, in macroeconomic data, everything is pretty highly correlated with everything else.  So you could put pretty much whatever you liked into your regressions and get a really good fit with in-sample data.  Then history would happen, new data would arrive to contradict the model's predictions, and you'd either re-estimate the model (and watch the coefficients bounce around more or less at random) or you'd declare the latest data to be some kind of special case and "adjust" for it.

So when critics denigrated the models of the early '70's as "ad hoc," they had a pretty serious point.

But what was the solution to all of this ad hoc-ery?  Where were we to look for the all-important virtue of discipline?   Ideally, in social science as in physical science, the source of discipline is data.  If you want to tell the difference between a true theory and a false one, you ask reality to settle the question.  But that was the heart of the problem: with so little data, all the models looked equally good in-sample, and no model looked especially good out-of-sample.  Discipline, if there was to be any, would have to come from theory instead.  And "microfoundations" was put forward as one form of theoretical discipline (2).

The idea certainly sounded good: rather than simply making up relationships between aggregate variables like interest rates, output, unemployment, and inflation, we should show how those relationships arise from the behavior of individuals.  Or, failing that, we should at least restrict the relationships in our macro models to those which are consistent with our understanding of individual behavior.  For surely our standard assumptions about individual behavior (basically: people do the best they can under the circumstances they find themselves in) must imply restrictions on how the system behaves in the aggregate.

Sadly, this intellectual bet was lost even before it was placed.  If we take Lucas (1976) as the beginning of the microfoundations movement, we may note with some puzzlement that the premise was mathematically proven false two years earlier, in Debreu (1974) and Mantel (1974).

It is sometimes said that modern macro suffers from too much math.  But the problem is not "too much," but rather that its use of math is strangely selective.  In particular, the idea that microfoundations per se can impose "discipline" on aggregate models represents deliberate ignorance of one of the most important results in mathematical economics.  Debreu, Mantel, and Hugo Sonnenschein had shown conclusively that, for any macro behavior you care to invent, there exists a set of classically well-behaved rational utility optimizing agents that will, collectively, exhibit the desired behavior.

Put another way, the classical assumptions about individual behavior impose no limits whatsoever on the behavior of aggregate models.  Oops.

The specious pretense, then, that one's preferred models are "better" microfounded than the competition (when, in fact, all models are equally micro-foundable) is part one of the microfoundations hoax.  But it gets better.  (Or worse, depending.)

The models which preen themselves most ostentatiously in their "rigorous" microfoundations are invariably based on so-called "representative agents."  Now, every economist, at some point in their first year of graduate school, learns the mathematically necessary conditions for the existence of a representative agent corresponding to a collection of individual agents.  In the lingo of the field, we say that a representative agent exists only if (a) all of the individual agents have identical preferences; and (b) if those preferences are [jargon] quasi-homothetic [/jargon].  "Quasi-homothetic" is a fancy way of saying that 10,000 households whose resources added together equal those of Bill Gates will buy exactly what Bill Gates will buy.

Neither of these conditions are remotely plausible, and nobody believes that they are true, including macroeconomists. And if either of those conditions fails to hold, an economy which behaves as if it posessed a representative agent cannot be derived from classical microeconomic foundations.

Let that sink in for a moment: of all the macro models that have been floated over the last century or so, only the so-called microfounded models are completely and demonstrably incompatible with classical microfoundations.  And this is (or should be) obvious to anyone who didn't sleep through the first year of graduate microeconomic theory.

So when I call "microfoundations" a hoax, I'm not kidding around.  The only question is, what proportion of macroeconomists have perpetrated this hoax upon themselves, and what proportion has known this all along.








(1) Bear in mind that the entire apparatus for gathering and reporting economic statistics in the U.S. was basically created in 1947.  Pity the macroeconomist circa 1965, trying to understand the most complex social system in history based on n < 20 observations.  Yikes.

(2) "Rational expectations" was another.  The "rational expectations revolution" business probably deserves a separate post.  For now, just know that the name is a kind of mathematical pun, and that neither "rational" nor "expectations" means what you probably think they mean.  Aren't we clever?


Friday, September 9, 2016

Houdini's Straightjacket

Consider the escape artist.  He dons handcuffs, a straightjacket, leg irons, a blindfold, and a skin-tight leotard made entirely of SuperGlue (tm).  His assistants seal him inside a steamer trunk, weld the locks shut, and sink the whole mess to the bottom of a shark-infested lagoon.  Then, in dazzling display of skill, grit, and showmanship, he frees himself and emerges, alive, unharmed, and not at all eaten by sharks.  Crowd goes wild.

Now, escape artistry may be a fine form of entertainment, but it probably wouldn't be anyone's first choice as a model for the conduct of social science.  Yet, bizarrely, it has become the prevailing paradigm in macroeconomics.

How so?

Consider the macroeconomist.  She constructs a rigorously micro-founded model, grounded purely in representative agents solving intertemporal dynamic optimization problems in a context of strict rational expectations.  Then, in a dazzling display of mathematical sophistication, theoretical acuity, and showmanship (some things never change), she derives results and policy implications that are exactly what the IS-LM model has been telling us all along.  Crowd -- such as it is -- goes wild.

And let's be clear:  not even the most enthusiastic players of the macroeconomics game imagine that representative agents or rational expectations are, in any sense, empirical realities.  They are conventions, "rules of the game."  That is, they are arbitrary difficulties we impose on ourselves in order to demonstrate our superior cleverness in being able to escape them.

They are, in a word, Houdini's straightjacket.

Of course, this would all be good, clean fun, except for one thing:  Harry Houdini drowned in a straightjacket. (1)

Similarly, even defenders of "modern" DSGE models, including those of the New Keynesian (NK) variety, seem to agree that their modelling approach made arriving at sound policy recommendations unnecessarily difficult.  For example, George Evans at the University of Oregon writes:
[T]he profession as a whole seemed to many of us slow to appreciate the implications of the NK model for policy during and following the financial crisis ... because many macro economists using NK models in 2007-8 did not fully appreciate the Keynesian mechanisms present in the model.
Now, if after thirty years of study economists failed to "fully appreciate the Keynesian mechanisms present in the model," one might wonder exactly what such models have to recommend themselves.  What is the advantage of an intellectually demanding and mathematically complex modelling approach that makes it harder to actually get the job done?

The answer, I suspect, is that "intellectually demanding and mathematically complex" has become an end in itself -- that modern macro has become an arena within which to show off technical virtuosity for its own sake.  And the harder we make it look, the tighter the straightjacket, the cleverer we appear when (after long, painful struggle) we finally emerge.

Which is fine as long as the goal is entertainment.  But if the goal is, you know, looking after the economic welfare of seven billion human beings, the whole enterprise begins to look more than a little bit self-indulgent.


(1) Well, in the movies.  In real life, he died in a hospital of peritonitis.  Dammit.  But why let the facts get in the way of a perfectly good analogy?