Former Fed Governor Tarullo doesn’t much care for models based on inflation.

I linked to this piece in the Links post the other day, but I wanted to dive into it a little more thoroughly, as it pertains to the the Fed Chair, Monetary Policy, etc.

I’ll talk a little about my questions regarding internal Fed politics (maybe too much), and then look at former Fed Governor Tarullo’s white paper “Monetary policy without a working theory of inflation”.  He offers some really good insight into the policy and projection making practices at the fed.

Daniel Tarullo didn’t leave much of an impression on me as a Fed Governor from 2009 – 2017 (announcing his intention to resign after Trump was elected).  I end up using this phrase a lot but, I don’t understand his resignation decision.

On the one hand, if it was politically motivated, why?  If he thought Trump was going to be bad, wouldn’t five more years on the most powerful economic team in the world be a good place to make a difference?  Or did he think that Trump would be better at nominating Fed Governors than Obama had been, and so decided to vacate his seat?

On the other hand, maybe it wasn’t politically motivated.  In that case, did Tarullo decide that his job was done?  He had helped pump up the balance sheet from $2 ish trillion in 2009 to over $4, and now he was going to let somebody else clean up.

In fact, by the time I left the Fed after more than eight years, I had come to believe that my lack of prior involvement in monetary policy had proven in one respect to be something of an advantage for my participation in FOMC deliberations. Coming fresh to a place where much of the discourse was accepted or assumed by the very smart economists on the FOMC and the Fed staff helped me to see where some of that received wisdom was not holding-up well in the circumstances we were facing.

This is an important idea.  Tarullo et al. in 2008-2010 were throwing everything at the wall to see what would stick.  So much of what the FOMC (and all central banks) was doing at the time involved approaches for which there was no precedence.  To some extent, his argument that a lack of formal economic training going into his time at the Fed was a benefit makes sense.  He wasn’t beholden to specific economic theories, he asked questions that other governors probably skipped.  But what about now?  Isn’t the normalization process completely unconventional also?  Why does that require academic macroeconomic royalty (John Taylor)?

Questions regarding who should lead the Fed aside, Tarullo brings up some really interesting points about making the projections on which monetary policy is based.  The paper is a good read, but here are some key pieces:

  • We accept the Fed’s Statement of Economic Projections (SEP) as fairly straightforward economic theory, but Tarullo points out that many of the variables that go into the SEP are “unobservable”.  Basically, many of the variables are assumptions (otherwise known as “smart people guessing”).
  • But not all of them!  Some are simply “projections” that are changed to help build the Central Bank’s narrative … like projected long term inflation.  From page 6, body of the text and footnote:

The fourth longer-term variable is the inflation rate, but that is not really a variable at all, because all members of the FOMC are expected to, and do, project that inflation over the longer term will be at the target rate of 2%. (6)  […] Each participant’s estimate of r* can be calculated by simply subtracting the 2% inflation target from the value given by the participant for the longer-term federal funds rate.

(6) There is no inherent reason why this need be so, of course. A member of the FOMC could predict inflation in the longer term above or below 2%. But this prediction would sit a bit uneasily with the FOMC commitment to the 2% target. And, as I learned when I once penciled in a number other than 2% in the first version of one of my SEP  submissions (before the formal FOMC adoption of the 2% target), the result of such a deviation was a visit from the staff of the Monetary Affairs Division, who explained the awkwardness that such a reported submission would have for FOMC communications and for the (then) Chairman. Because, as explained in the next, I had little confidence in the ability of anyone on the FOMC to predict where any of the key variables would be in the medium- to longer-term, and because I didn’t want to complicate further an already challenging communication task for the Chairman, I changed my actual submitted number to 2%

The staff at the Monetary Affairs Division (MAD!) made him change his answer!  “Excuse me, Dr. Tarullo, we understand that you don’t think long term inflation is going to be 2%.  You do understand that we build models around here.  Good models.  Sophisticated models.  Bernanke likes them.  Bernanke likes you.  It would be a shame to ruin that friendship if we have to explain that one of our Governors thinks our default number is wrong.”

Turns out this goes along with my throw away idea breaking down the Taylor Rule the other day (which I’ve now linked to twice … I may have to do a more sophisticated breakdown).  Taylor talks about how he wants a more rigorous, model-based system to determine interest rates, but if that model is built with MAD checking everybody’s work, what’s the point?  It just ends up being a marketing tool for making it look like the Fed is doing “the thing the model says” instead of “whatever they think is best”.

Tarullo goes on to point out critical deficiencies in the understanding of inflation, taking the Phillips Curve to task and basically pointing out that everybody “knows” why inflation does what it does, but everybody knows something different.

I will continue to be uneasy with placing too much confidence in the efficacy of inflation expectations in making monetary policy. Thus I have been hesitant to rely on “well-anchored inflation expectations” to help pull inflation back up to 2%, as some have argued will be the case. We just don’t know enough to accept or reject that proposition. And, although we do not currently face a high risk of rapidly increasing inflation, in different circumstances we will have to be hesitant to rely too much on well-anchored expectations from keeping inflation from rising well above its target.

Great piece.  I highly recommend it.

One thought on “Former Fed Governor Tarullo doesn’t much care for models based on inflation.”

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 )

Twitter picture

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

Facebook photo

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

Connecting to %s

%d bloggers like this: