Scott Sumner  

Influence, target, control

PRINT
Gordon on Growth: Rates versus... Huffington Post's Stunning Ign...

I often get commenters complaining that the Fed should not be controlling interest rates. They think the market should set rates. In one sense I agree, but in another sense I wonder if the commenters are confused. I wonder if people think the Fed moves interest rates away from the market equilibrium. It does not, it influences the market equilibrium. Let's start with an analogy from the oil industry, and while doing so I want you to think about how inadequate the English language is in this area.

Consider two cases:

1. President Carter introduces price controls on oil, setting a price ceiling at $25/barrel, even as the market price in $34/barrel. A shortage results, with long gas lines.

2. The Saudi oil company Aramco adjusts oil production to keep market prices at $36/barrel. There is no shortage, no gas lines.

In ordinary English, these are two ways that governments could be said to "control" prices. But I think you'll agree that they are vastly different methods, and have vastly different implications for the economy. The term 'control' is being used in two very different ways.

I hope it's obvious that the Fed's control over interest rates is like the Aramco case, not the price ceiling case. There are at least three different senses in which the Fed could be said to control interest rates:

1. The Fed might fix interest rates, via usury laws.

2. The Fed might target interest rates (as the Fed actually does.)

3. The Fed might target some other variable like the money supply or exchange rates, and yet nonetheless Fed actions indirectly determine the level of interest rates.

Or I suppose the Fed could be abolished. But as long as the Fed exists, the meaning of the term 'control' is more ambiguous than you might think. Consider 3 policy options:

1. The Fed has a 2% inflation target, and sets a new fed funds targets every 6 weeks, adjusting rates as needed to target inflation. In that case is the Fed controlling inflation, interest rates, or both?

2. The Fed has a 2% inflation target, and sets a new fed funds targets every single day, adjusting them as needed to target inflation. In that case is the Fed controlling inflation, interest rates, or both?

3. The Fed has a 2% inflation target, and has no fed funds target. Instead it targets the TIPS spread, or a CPI futures contract. In that case is the Fed controlling inflation, interest rates, or both?

Case one and two are pretty similar, as we've merely shortened the time between meetings from 6 weeks to 1 day. And yet arguably cases #2 and #3 are much more similar to each other, than to case #1. That's because in both case #2 and #3, the interest rate changes each day, in a way that the Fed believes will stabilize expected inflation over time. The interest rate path is quite similar in those two cases.

Here's the problem. Fed policy affects all nominal variables, including nominal interest rates, nominal exchange rates, nominal GDP, the nominal prices of apples, oranges and Brazilian hot wax treatments. All nominal variables. In the 1970s, the Fed printed lots of money, causing almost all nominal variables to be higher than they would have otherwise been. But we usually don't think of the Fed as "controlling" those prices, for two unrelated reasons. One reason is that the Fed's actions don't impact real variables, in the long run. The other reason is that the Fed was not targeting most of those variables. Does it matter if they were? Not as much as you might think.

Consider two policies: In one case the Fed continually adjusts the target fed funds rate in order to keep expected inflation at 2%. In the other case the Fed continually adjusts the target nominal price of zinc, in order to keep expected inflation at 2%. If the Fed does its job in a competent way, then the path of zinc prices would be identical under either regime. Does it then make sense to talk about the Fed "controlling" zinc prices in one case but not another?

Yesterday I did a post criticizing Mike Konczal on the distinction between the Fed acting and not acting. Let me provide a better analogy over here, as some missed the point. It might make sense to distinguish between an "activist ship engine" that is fed lots of fuel and an idle engine. But the steering wheel of a ship is different. There is no single setting of the steering wheel that is more active than another. It doesn't require any more fuel to set steering at NNE than at NNW.

By analogy, it makes sense to talk about using fiscal policy, as the natural benchmark is a fiscal policy set on classical cost/benefit considerations, and the alternative is a deficit that would not otherwise occur, used to boost GDP in a recession. Deficits require costly future taxes with deadweight losses. In contrast, it's nonsensical to talk about "using monetary policy" unless your alternative is pure barter. There is no benchmark of not using monetary policy; rather there are merely different settings of various monetary indicators. (Even with no Fed, the private money issuers would have some sort of "policy".) If the Fed would like to see NGDP growth at X%, and a certain policy setting would achieve that, then any failure to achieve X% growth can be said to be "caused" by the Fed.

Konczal talked about a housing crisis causing a drop in AD, and the Fed not responding. But there is no evidence that that's what happened. If, after July 2007, the Fed had kept the monetary base growing at exactly the same rate as in the 5 previous years, there might well have been no recession. Or there might have been a Great Depression. (In my view there would have been no recession in 2008, but major problems later, but we can't ever know.) So it's meaningless to talk about a housing crash causing a drop in AD in 2008, without knowing the Fed's policy regime.

The Fed meets every 6 week to set policy. The only question is whether they set the right or wrong policy. They have no ability to "do nothing"; it's a meaningless concept for monetary policy.

PS. I just got this email from the Hypermind prediction market:

Bravo ssumner, Suite à la fermeture d'un concours sur Hypermind, votre compte de gains en Euro a été crédité de 10€

It's one of those good news/bad news situations. The good news is that I'm 10 euros richer. The bad news is that my ability to beat the market undercuts my belief in the EMH. Should I hope that I was right or wrong about the EMH?

(I shorted the one year NGDP contract early in 2015, when it was trading around 4.2%. NGDP growth ended up being 2.9%)


Comments and Sharing






COMMENTS (15 to date)
John Hall writes:

If you meant your PS to be kidding, you should have been more obvious. That's like saying I bought a put option yesterday on Amazon and it made a bunch of money, so therefore my belief in EMH is undercut. EMH doesn't mean that people never make profits!

Kevin Erdmann writes:

I don't understand this:

"If, after July 2007, the Fed had kept the monetary base growing at exactly the same rate as in the 5 previous years, there might well have been no recession. Or there might have been a Great Depression. (In my view there would have been no recession in 2008, but major problems later, but we can't ever know.) So it's meaningless to talk about a housing crash causing a drop in AD in 2008, without knowing the Fed's policy regime."

The 5 year growth rate of the MB in July 2007 was 4%, and YOY it was almost down to 2% - both seem low. Are you saying this was too high or too low, or something else? And what major problems would have come later?

Scott Sumner writes:

John, Sorry, yes I was joking.

Kevin, The growth in the base between August 2007 and May 2008 was about zero. Had it been about 4% I believe the economy would have been much stronger, and the banking crisis far milder, probably implying no recession in 2008. We can't be sure. Regarding later problems, I believe that base velocity is somewhat unstable, and hence targeting the base at 4% would have led to economic instability at some point, but probably not in 2008.

Kevin Erdmann writes:

Gottcha. Thanks.

A writes:

On twitter, Konczal agreed that the Fed passively tightened in 2008, so, on this narrow issue, his disagreement with "the Fed tightened" seemed to be semantic. But his article linked to Barry Ritholz, who argued that the Fed obviously loosened policy in 2008. The evidence? Declining fed funds. If economists insist on identifying "active" policy by price changes, then laymen will continue to infer that price changes equal policy changes.

ThomasH writes:

"1.(&2&3) The Fed has a 2% inflation target, and sets a new fed funds targets every 6 weeks, adjusting rates as needed to target inflation."

No it does not. It has an inflation rate ceiling target. If the Fed had a 2% inflation target, the price level would sometime be above a 2% trend and sometimes below the 2% trend. Pleas do not go along with the confused way Fed policy is explained.

Having an inflation rate target might be a very good idea and would have mitigated the great recession if not prevented it. But it did not happen. Whether the Fed failed to drop interest rates farther, buy more LT assets or have a negative ST interest rate because it did not waht to breach the inflation rate ceiling or becaue it faced political pressure not to buy more "unconventional" assets/go below 0, we still do not understand.

ThomasH writes:

And the issue of whether to say the Fed does or doe not "control" interest rates is irrelevant to saying what the Fed should do to carry out its dual mandate.

Scott Sumner writes:

A, I wonder what he thinks of the stance of monetary policy during hyperinflation? That's just bizarre.

Thomas, You said:

"If the Fed had a 2% inflation target, the price level would sometime be above a 2% trend and sometimes below the 2% trend."

It is.

bill woolsey writes:

The Fed could seek to adjust the quantity of base money to meet the demand for base money consistent with nominal GDP remaining on a stab;e growth path.

The Fed would not be seeking to control nominal interest rates at all. They would change with the supply and demand for credit.

I don't doubt that errors would by the Fed would usually impact short term interest rates and correction of those errors would reverse those changes. However, that isn't the same thing as trying to keep nominal interest rates at some particular level.

And, in fact, the Fed has attempted interest rate smoothing. It intentionally adjusts short term interest rates in a series of steps. Expectations that it will do this is supposed to keep long term interest rates on the same path as would occur with more sudden adjustments.

Further, consider a situation where short term interest rates would likely fluctuate around an average level. Long term interest rates would remain relatively stable. The Fed keeps short term interest rates stable as well.

They don't have to follow that policy.

A policy of adjusting the quantity of money to the demand to hold it would allow fluctuations in short term interest rates to equililibrate temporary shifts in the supply or demand for credit.

The policy generates excess supplies and demands for money so that the needed liquidity effects are generated. Since the monetary disequilibrium is temporary and expected to be so, it has little impact on nominal GDP.

Khodge writes:

The English language is more than capable of expressing concepts as simple as "control." If there is a problem, it lies not in the language but in the users of that language. If it bothers you then you must let the persuasiveness of your words drive (or "control") the conversation.

Whoever controls the words wins the argument.

ThomasH writes:

Yes the PL is occasionally above trend, but recently it has been below more frequently than not And it was not apparent that the Fed was trying to do anything about it.

ChrisA writes:

Scott - Its interesting, very often the same people claiming that the Fed is unable to affect the "real" economy and that QE is "pushing on a string" follow on with complaints that the Fed is creating a "credit bubble" . Surely these are mutually contradictory claims? A certain writer at the FT is often guilty of this.

End. writes:

"The Fed could seek to adjust the quantity of base money to meet the demand for base money consistent with nominal GDP remaining on a stab;e growth path.

The Fed would not be seeking to control nominal interest rates at all. They would change with the supply and demand for credit."

But that is exactly what the FED does. Take a market, whatever market and two policies:

- You fix the supply and let the price adjusts.
- You fix the price and serve the demand at the given price.

In the end, the two policies are the same if the FED adjusts the price or the quantities so as to meet a target for inflation or NGDP.

Bill Woolsey writes:

End:

Your analysis makes perfect sense if the goal of the Fed is to fix short term interest rates.

It would fit better if the quantity of base money solely changed through discount lending. The quantity of base money would change as an unintended consequence of the demand for borrowing from the Fed.

But, the interest rate is not the price of base money.

I have some of it in my pocket today. And it has a price in terms of many different goods and services.

End. writes:

"But, the interest rate is not the price of base money.

I have some of it in my pocket today. And it has a price in terms of many different goods and services."

Money is an asset. Like every asset, it has a price in terms of many goods and services.The interest rate is the opportunity cost of holding money versus treasury bonds.

"Your analysis makes perfect sense if the goal of the Fed is to fix short term interest rates."

It makes perfect sense whatever the FED's target.

Comments for this entry have been closed
Return to top