Why Krugman Drives Me Nuts

[Dave Backus @ NYU]

Paul Krugman expresses frequent frustration with modern economics, so
perhaps it’s no surprise that economists are just as frustrated with
him. To be honest, he drives me nuts. I’m frustrated that so much
talent and ability has turned into the Glenn Beck of the left. It’s
his choice, but it seems like a waste to me. And to be fair, there
are two Paul Krugmans. I admire the brilliant expositor of economic
ideas. I’m less enchanted with the political partisan who can’t
resist using cheap debating tricks in what could be a useful exchange
of ideas. Even worse, he sometimes mixes the two, disguising politics
as economic analysis. There’s a market for this kind of thing, for
sure -- my mother, for example -- and cheap shots are always good
fun. But you don’t need a Nobel-Prize-winning economist to win a food
fight.

Here’s a recent example. It’s worse than usual, in the sense that
cheap tricks dominate. But you be the judge. I include his post in
italics and add a summary (“what Paul says”) and my own thoughts
(“comments”).

October 2, 2010, 9:33 AM
How The Other Half Thinks

Ezra Klein has written in, asking for a post laying out the difference
between the more or less Keynesian model Brad DeLong and I work with
and the models others have been using – and how their predictions
differ. It’s a good request, although the truth is that the other side
in this debate doesn’t necessarily agree on a single model, or even
use models at all. Still, I think it is possible to describe the
general views of the other guys — and to see how off their predictions
have been.

[What Paul says: “Brad and I” are “more or less Keynesian,” but “the
other side ... doesn’t even have models.”]

[Comment: If by “the other side” Paul means economists rather than
politicians, the truth is almost exactly the opposite. As far as I
can tell, he has no model. Or if he does, he’s only willing to say
it’s “more or less Keynesian.” In another post, he mentions the IS/LM
model, a textbook standard from the 1970s that I’ll come back to
shortly. So he adopts a classic debating trick and switches to
offense: “The other side ... doesn’t agree on a model or even use
models at all.” There’s not an ounce of truth to it, but it works in
a debate if you talk fast enough. The truth is that he’s ignoring
four decades of influential work, including at least two Nobel
Prizes. If he has a complaint, it has to be that he doesn’t like
these models, not that there aren’t any.]

So: first of all, the other side in this debate generally adheres,
more or less, to something like what Keynes called the “classical
theory” of employment, in which employment and output are basically
determined by the supply side. Casey Mulligan has been most explicit
here, coming up with increasingly, um, creative stories about how what
we’re seeing is a choice by workers to work less; but the whole
Kocherlakota structural unemployment thing is similar in its
implications. Oh, and the Cochrane-Fama thing about how a dollar of
government spending necessarily displaces a dollar of private spending
is basically a classical view, although there doesn’t seem to be a
model behind it, just a misunderstanding of what accounting identities
mean.

[What Paul says: “The other side” is what Keynes criticized in 1936
and does not include the influential work mentioned above. Some of
them claim unemployment is “a choice by workers.” On Cochrane-Fama:
he lost me. Identities?]

[Comment: Paul seems to me to have turned into a curmudgeon before
his time. He abandoned the academic world some time ago, and what he
sees there now he doesn’t much like. So he uses another cheap
debating trick: define the opposition in a misleading way that’s easy
to take on. He takes the classical model of the 1920s, sprinkles in
enough random links to modern commentary to turn it into a presentable
straw man, and slaps it around with obvious delight.

There is an issue here worth discussing: the role of supply (workers’
decisions) and demand (employers’ decisions) in employment outcomes.
If the classical model he has in mind leaves no room for demand, IS/LM
leaves no room for supply. Surely both matter. There’s a New Yorker
cartoon whose caption is something like: “Oh, I get it, supply AND
demand.” The rest is jargon. I’m with Lucas on this one: adjectives
don’t add much to our understanding of unemployment (involuntary,
structural, natural, etc). The real question is what we can and
should do about it. Paul favors more “stimulus,” but if the federal
government increased its spending by (say) 2 trillion dollars, what
would that do to unemployment? How would we pay for it? Would it
trigger a panic among investors in US government debt? Is it worth
it? None of these are simple questions, and we do readers a
disservice by treating them that way.

On Mulligan: one of my colleagues quips that he gives the phrase
“taking a Mulligan” new meaning. See, cheap shots are good fun,
whether they’re fair or not.]

Once you have a more or less classical view of unemployment, you
naturally have the classical theory of the interest rate, in which
it’s all about supply and demand for funds, and something like a
quantity theory of money, in which increases in the monetary base
lead, in a fairly short time, to equal proportional rises in the price
level. This led to the prediction that large fiscal deficits would
lead to soaring interest rates, and that the large rise in the
monetary base due to Fed expansion would lead to high inflation.

[What Paul says: “The other side” is also wrong on interest rates.]

[Comment: I’d like to leave 1920s economics for someone else, but if
we take the statement seriously, it applies not only to the classical
model, but to the IS/LM model Paul likes to use. His way out of this
box is to posit a “liquidity trap” -- more on that below.]

You can see the classical theory of interest and the soaring-rate
prediction clearly in Niall Ferguson’s remarks:

After all, $1.75 trillion is an awful lot of freshly minted treasuries
to land on the bond market at a time of recession, and I still don’t
quite know who is going to buy them … I predict, in the weeks and
months ahead, a very painful tug-of-war between our monetary policy
and our fiscal policy as the markets realize just what a vast quantity
of bonds are going to have to be absorbed by the financial system this
year. That will tend to drive the price of the bonds down, and drive
up interest rates

and, of course, in many WSJ op-eds, in analyses from Morgan Stanley,
and so on. Meanwhile, you can see the high-inflation prediction in
pieces by Meltzer and Laffer — with the latter helpfully titled, “Get
Ready for Inflation and Higher Interest Rates”.

[What Paul says: I can find people who are still in the 1920s.]

[Comment: More straw man tactics. My former colleague Niall Ferguson
is a wonderful guy and a natural provocateur. But he’s an historian,
not an economist. That said, the relation between government
borrowing and the interest rate is complicated: it’s not as simple as
Paul or his classical caricature suggest. For one thing, changes in
tax rates and spending are clearly different, and in both cases the
details matter: How is the money spent? How quickly? What is
taxed? Who pays? How quickly do we reverse it when the economy turns
around?

As for inflation, there has rarely been more uncertainty about it. We
know the link with the monetary base is weak, but it’s there. That’s
one of the reasons Bernanke has repeatedly mentioned his exit
strategy, to signal that the massive increase in the monetary base can
and will be quickly reversed when the time is right. Will the Fed get
the timing right? Time will tell.]

While the other side was making these predictions, people like me were
saying that classical economics was all wrong in a liquidity trap.
Government borrowing did not confront a fixed supply of funds: we
were in a paradox of thrift world, where desired savings (at full
employment) exceeded desired investment, and hence savings would
expand to meet the demand, and interest rates need not rise. As for
inflation, increases in the monetary base would have no effect in a
liquidity trap; deflation, not inflation, was the risk.

[What Paul says: This is a rare case in which his argument is
needlessly dense. If we cut through the jargon, it amounts to: the
classical model is stupid, but it’s especially stupid in current
circumstances.]

[Comment: I don’t feel like it’s my job to explain what Paul means --
he’s paid to do that himself. But let me take on the “liquidity
trap.” The issue is the role of monetary policy. In modern terms, we
would ask what the Fed does when the interest rate hits the “zero
lower bound.” (Interest rates on bonds can’t go below zero, or people
would hold cash instead.) The accepted answer is to print money, lots
of it, which will eventually produce inflation and push us away from
zero. The weasel words here are “lots” and “eventually,” but that’s a
start. And he’s right: the economy behaves differently at a zero
interest rate.]

So, how has it turned out? The 10-year bond rate is about 2.5 percent,
lower than it was when Ferguson made that prediction. Inflation keeps
falling. The attacks on Keynesianism now come down to “but
unemployment has stayed high!” which proves nothing — especially
because if you took a Keynesian view seriously, it suggested even
given what we knew in early 2009 that the stimulus was much too small
to restore full employment.

[What he says: Niall is still wrong.]

[Comment: So be it, he’s still a lot of fun in a debate.]

The point is that recent events have actually amounted to a fairly
clear test of Keynesian versus classical economics — and Keynesian
economics won, hands down.

[What he says: Mine is bigger than yours, so there!]

[Comment: Please, could we have a civil conversation without taking
our pants off? Maybe we’d all learn something.]

Link to blog post: http://krugman.blogs.nytimes.com/2010/10/02/how-the-other-half-thinks/

Other related links:
http://krugman.blogs.nytimes.com/2010/10/04/math-models-and-mystification/
http://andolfatto.blogspot.com/2010/09/tom-sargent-speaks.html