The Indictment of Dennis Hastert
Liveblogging World War I: May 28, 1915: The Lusitania

More on Ben Bernanke vs. John Taylor: In Which I Give Up

Over at Equitable Growth: I confess that I think it is time to stop trying to make sense of John Taylor's views on the Taylor rule. There simply seem to be too many gaps in logic, and too many assertions about the literature that I cannot understand.

In fact, Tony Yates reads John Taylor's attack on Ben Bernanke, and sounds almost... shrill:

Tony Yates: More on John Taylor vs Bernanke: "John Taylor... claims that the Taylor Rule emerged from ‘two decades of research on optimal policy’... READ MOARh

...It is important that John does not allow this to be read by WSJ readers... as ‘after two decades, we figured out that the Taylor Rule was optimal policy’. Because that would not be right.... Taylor Rules... fall short of optimal policy.... Pre-crisis research... ignored banking and finance. We have no idea, really, in model terms, how Taylor Rules do in a financial crisis.... Even if we leave aside the banking and finance problem... optimal Taylor Rule[s have]... very large coefficients on the term in inflation. Perhaps an order of magnitude greater than JT’s.... [A] huge disconnect between what the models were suggesting should happen, and what central banks were actually doing to tame inflation [and what John Taylor was saying they should do].... JT recommends... that rather than flipping to quantitative easing, central banks adopt a ‘fixed money growth rule’.... What does it even mean?... In a... temporary liquidity trap... the amount of money growth induced while at the zero bound is irrelevant... [unless] we make a credible commitment to ditch the Taylor Rule later. JT isn’t advocating that. What is he advocating? I am mystified...

I must confess I am having less and less success in even understanding what intellectual project John Taylor is currently engaged in.

I could understand it if John Taylor were calling for all of:

  1. no excess of high powered money above its long-run trend adjusted for the Taylor rule,
  2. no positive gap between the Federal Reserve's short term safe interest rate target and its Taylor rule value, and
  3. strict limits on bank leverage to make it very difficult for any overleverage to develop as a result of which a negative financial shop composite chain of bankruptcies and fear bankruptcies that would clog the credit channel.

The problem is the Taylor seems to be on board for (1) and (2) but not for (3). He seems to be for using monetary policy as a tool of prudential regulation, but not for using prudential regulation as a tool of prudential regulation.

And the considerations that he claims drive him to (1) and (2) make no sense to me.

And this is what drives Tony Yates into shrillness. Over on Twitter:

  • @Noahpinion: @RebelEconProf His support for the classic Taylor Rule as the perfect monetary policy is not supported by research.
  • t0nyyates: @RebelEconProf @Noahpinion @EconomicsOne he gets pretty darned close so close your op ed reader won't know the difference in our eyes: and it wasn't just that. I think the surprise is because he seems to be using different intellectual standards for his op eds and his research
  • @t0nyyates: @RebelEconProf that doesn't account for it, though that's true. He's good at it. Just communicates the wrong ones. His work is good; but his op ed urgings are not, as Noah said, justified by it. In fact they are directly contradicted. In my blogs on him i explain why i think he contradicts his earlier intellectual self. His disavowall of countercyclical fiscal policy; his ignorance of the zero bound; his misapprehension of QE. Failure to see financial crisis invalidates earlier work; hubristic extrapolation from good simulation results to real life that was pure hot air conjecture. Invalidated because it shows models in which Taylor Rule tested were severely lacking. All the comfort we drew from good TR performance out the window, as we see models lack something profound.
  • @t0nyyates: @RebelEconProf The rule is not even operational as stated [Bernanke/McCallum's points on lags]; and the RE eq'm highly questionable. The truly optimal TR coefficients involved coefficients on inflation that were inexplicably huge anyway, and these were only tamed by ad-hoc penalties on interest rate variability. So TR lover wd really have said 'wtf is the Fed doing going anywhere near coefficient of 1.5?' and sensible person says 'ok, so the model is missing something, because who believes the microfounded loss function?' But by the same token sensible person doesn't put any weight on deviating from Taylor's 1.5-0.5 rule. And what about the other teeny weeny causal factors in the crisis? Like regulation; ratings shopping; uphill savings flow to Western safe assets...
  • @t0nyyates: @RebelEconProf If you buy my points about the TR, then John Taylor's thesis that deviating from it caused the crisis is absurd. I'm saying his op-ed's are not founded on research. Saying he wrote a book about it doesn't rebut that. He invokes that literature as anointing his rule in his response to Bernanke. I'm saying; T's writings do not follow logically from a sound reading of prior literature.
  • @t0nyyates: @RebelEconProf Even if the earlier writing was sound, the later op-ed is grossly misleading, because it conceals so much counter to it. Imagine new macro phd who didn't know this stuff, who goes back after reading his op ed to find out more. they'd be shocked. All you can say is that they summarise prior, false, internally contradictory, and partial writings of his. That doesn't make it defensible at all.

Suppose you say: "X--with a lot of Y added to it--is optimal in a model with no Z, therefore it is close to optimal in the world, which has Z." Then you are, I think, committing yourself to both:

  • Things that lead to Y are important in the world, and even though we cannot model them we need to allow for a large deviation between what the model says is optimal policy and what we should do.
  • Z is not important in the world.

In this case, X="Taylor Rule with Taylor's original coefficients", Y="lots of interest-rate smoothing by the central bank", and Z="the financial sector and the credit channel". It seems to me that if one is trying to make an argument--as opposed to succumbing to one of the species of the bestiary that Paul Romer calls "Mathiness"--you have to make strong arguments for including Y and for excluding Z from your thinking. And Taylor does not even make the first attempt at a gesture toward doing either of those.

I hereby give up: I hereby classify trying to make sense of any of Taylor's critiques of Bernanke as a dissipative activity, to be avoided permanently.