No legislation on the Taylor Rule please

John Taylor and Alan Blinder have been exchanging op-eds in the US press on a bill before Congress now that, if passed, will force the Fed to ‘describe’ its preferred rule for monetary policy.

I posted before here on why I thought this was a bad idea.

The recent exchanges are about whether the legislation would tie the hands of the Fed or not.  Blinder argues that it will, Taylor argues that it won’t, since all the Fed have to do is disclose the rule they will follow.

John Cochrane blogged some time ago sounding supportive of the idea.  Which I found puzzling, since he has written a few fascinating papers tearing apart the framework that produced central bank received wisdom that these monetary policy rules were a good idea.  For example, he questioned the notion of equilibrium that most people sweep aside along the way to admiring the stabilisation properties of Taylor Rules, and the stories people tell about why the rules do well.  And he even queried  whether people have the sign right in New Keynesian models when they think about the effect of monetary policy on inflation].

To recap earlier arguments, there are many reasons this is a terrible idea.

1.  Supposing the Taylor Rule was genuinely thought to be a great guide to policy.  In a few years, we might realise it isn’t.  And legislation is hard to change.

2.  The legislation sets a terrible precedent for interfering in the operational independence of the Fed.  The right way to delegate monetary policy is to be clear about what goal they are set.  [A little unfortunate that the Fed have been left to define their own goals, quantifying the previously unquantified definition of price stability].  And then the Fed ought to be left to decide itself how to achieve those goals, and be held to account based on performance.  Even if this were a well-designed interference, (it isn’t), the next one might be for nefarious political ends.

3.  The US Congress is not to be expected to reach fair-minded conclusions on anything these days.  The effort to tame the Fed is probably entirely a political thing stemming from the visceral libertarian feeling that the Fed’s activism – admirable in my view – is part of the problem in the US, rather than the solution.

4.  The Taylor Rule or any particular rule like it only ever performed well in a very narrow class of DSGE models.  That class of models now looks much less reliable after the crisis as a testing ground for monetary policy design.  In some senses, the crisis shows why it was a great idea not to have legislated in the days of the Great Moderation, when rules like this reigned supreme as explanations for why macroeconomic performance looked so good.

5.  John Taylor’s argument that the act of describing the rule won’t constrain the Fed must be incorrect.  If this were not the aim, or the likely outcome, there would be no point.  JT wants description to bring about more rule-following.

6.  The Fed do enough describing without legislation.  Continual reference to rules is made in their research outputs, the staff forecasts, research based work in the speeches of the PhD FOMC members, even the FOMC minutes themselves.  Why is legislation need to force this process of ‘describing’?

7.  JT’s support for the legislation is based on his view that deviating from the Taylor Rule was a major factor in causing the financial crisis.  This is a really weak argument.  In the models JT developed himself and in which the TR is a good policy, monetary policy just does not have such powerful and long-lasting real effects.  As I said above, we might argue that there’s good reason to ditch this model.  But if we do, we can’t keep the Taylor Rule or anything like it as the prescribed policy.  There’s a long learning process to go through that legislation won’t help.  Moreover, Bernanke pretty conclusively refuted the idea that monetary policy was anyhow loose relative to a sensible rule in the early 2000s.  JT’s own rule remember is not actually operational (point made by McCallum long ago) since it includes contemporaneous inflation and GDP, not available in real time.  If forecasts are inserted instead, policy looks fine.  And, there are some pretty strong pieces of circumstantial evidence that other policy failures were the cause!  This thesis is to be twinned with his comment on fiscal policy.  He sees the fiscal stimulus agreed at the beginning of Obama’s first term as a deviation from rule-based behaviour.  Despite the fact that in the same model that is the testing ground for the Taylor Rule, which JT built himself, a fiscal stimulus is an optimal response to a recession.

So, for all those reasons, legislation on monetary policy rules is a terrible idea.  My main hope is that this issue just sounds too technical for your average Republican mischief-maker looking to make some headlines.


This entry was posted in Uncategorized. Bookmark the permalink.

1 Response to No legislation on the Taylor Rule please

  1. Nick Rowe says:

    Yep. 4a: and if the Taylor Rule gets the natural rate of interest wrong (because it changes over time), we can get rather big consistent mistakes in hitting the inflation target. If the natural rate is 1% lower than you think it is, you will be targeting 2% lower inflation than you think you are, and the nominal interest rate will be 3% closer to the ZLB than you thought it would be.

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s