Scott Sumner  

Why the monetary policy pessimism?

Weekend Wanderings... The Perverse Death/Estate Tax...

I've recently noticed a lot of pessimism about what monetary policy can accomplish. For instance, if I discuss a higher inflation target, people will say, "what makes you think the Fed could hit a higher inflation target? After all, they have trouble hitting their current inflation target."

This is incorrect for lots of reasons, and I think some of the reasons are interesting enough to explore. But first, one obvious but boring flaw with the pessimist's argument:

1. The Fed targets their internal forecast of expected future inflation, and they've been successful in hitting that target in recent years. Now it's true that actual inflation has often fallen a bit short of their expectations, but that has no bearing on whether they could boost expected inflation. If bias in the Fed model results in their expected inflation estimate being 0.5% above the market expectation (say TIPS spreads) that gap would be equally likely to occur at a 4% inflation target as at a 2% target. Thus this sort of policy change would raise market inflation expectations from 1.5% to 3.5%.

The Fed is not cutting rates right now because they fear that it would lead to above 2% inflation going forward, not because they are out of ammo.

2. But I also think there is a more interesting mistake being made. People look at the path of interest rates, and wrongly conclude that they are looking at the path of monetary policy. They see that low rates have failed to boost inflation, and so conclude that monetary policy is not very effective.

But even in the (new) Keynesian model, interest rates do not represent the path of monetary policy. Rather what matters is the gap between the actual target rate and the Wicksellian equilibrium rate. In standard fiat money models, pegging the policy rate just 0.25% above or below the natural rate, and holding it there regardless of the condition of the economy, will quickly lead to hyperinflation or hyperdeflation.

Consider this simple example. The Fed cuts its target rate to 0.25% below equilibrium, and holds it there even if inflation expectations rise. Since the rate is now 0.25% below equilibrium, inflation expectations rise by a small amount, say 0.1%. But now rates are 0.35% below equilibrium in real terms, so policy is easier and thus inflation rises by even more, say 0.15%. But now real interest rates are 0.5% below equilibrium, so policy is even more expansionary and inflation expectations rise by even more, say 0.30%. But now real interest rates are 0.8% below equilibrium, and policy is becoming extremely expansionary. Rinse and repeat, and in a few years you have hyperinflation. If rates start out too high, you quickly end up with hyperdeflation

As far as I know, this has never happened, at least all the way to hyperinflation. Although to a lesser extent its what happened during the Great Inflation. And there's a reason we don't see this sort of extreme snowball effect---central banks aren't that dumb. (Actual hyperinflations occur for other reasons---chiefly monetizing the debt.)

In late 2015, the Fed raised its target rate by 0.25%, and announced that they expected to do 4 more rate increases in 2016. They probably slightly misjudged the future path of the Wicksellian equilibrium rates. In early 2016, the markets started warning the Fed that if they persevered with their plan to raise rates 4 times in 2016, the economy would enter a death spiral. Of course markets knew the Fed would not make that big a mistake, but there were seriously concerned about a smaller mistake. So you saw a combination of the market predicting 2 rate increases, and also predicting inflation would come in below the Fed's 2% target. Here's what markets were saying, translated into plain English:

1. We think that 4 rate increases would create a disastrous slump.

2. We think that the Fed will actually only do 2 rate increases, and even that will be too much.

3. We think fewer than 2 rates increases would be consistent with an expected inflation rate equal to the Fed's 2% target.

In other words, the markets expected the Fed to mostly "see the light", but not entirely.

The monetary policy pessimists tend to be unaware of all of this behind the scenes stuff going on with market expectations and Fed policy. There's a complex and delicate game being played, where the Fed tries to read the markets, as the markets try to read the Fed. If the Fed reads the markets pretty well, the interest rate will always be close to the Wicksellian equilibrium rate. In that case it won't look like interest rate changes "do anything", as they will occur in the fashion required to keep the macroeconomy stable. And if the Fed consistently keeps rates a tad too high, but otherwise closely shadows the equilibrium rate, it will seem like the Fed is "doing something" but is consistently unable to hit it's inflation target.

In fact, if the Fed did something wild and crazy, we'd see a dramatic market response, but we almost never see the Fed take an extreme position, perhaps with the exception of the Volcker disinflation, or the tight money policy of late 1929. The closest example in recent history occurred in October 2008, when the target rate (1.5% to 2.0%) rose far above the Wicksellian equilibrium rate. But that did not occur as a result of the Fed raising the target interest rate, but rather not cutting it as required to keep market inflation expectations close to 2%.

To conclude, there is no reason to be pessimistic about the efficacy of monetary policy, unless it is constrained by legal restrictions (say an exchange rate peg, or limits on what sort of assets the central bank can purchase.)

And remember, even if restricted to conventional interest rate policy, the Fed has 5 quarter point rate cuts in its quiver, before reaching Swiss IOR levels.

Comments and Sharing

COMMENTS (10 to date)
Brian Donohue writes:


People don't like monetary policy because they don't understand it. "Hard money, sound money" sounds right and prudent.

Politicians don't like monetary policy because they don't get to wear hard hats and cut ribbons.

Jeff writes:
If the Fed reads the markets pretty well, the interest rate will always be close to the Wicksellian equilibrium rate. In that case it won't look like interest rate changes "do anything", as they will occur in the fashion required to keep the macroeconomy stable.

Yes, the Friedman thermostat. Too many people don't understand this. Suppose you're in a car with a skilled driver who's trying to maintain a steady speed on a hilly road. Going downhill, he lets off the gas, going uphill he puts the pedal to the metal. If you're observing only the pedal position and speedometer, you'll conclude that pedal position has no effect on speed.

Matthew Waters writes:

From the pessimists' point of view, there are two arguments:

1. Certainly in EU and Japan, little capacity for conventional interest rate policy or something close to it. If negative IOR is pushed too far, it will only result in more cash withdrawals.

2. Ineffectiveness of asset purchases tie into the first issue. With both high negative IOR and asset purchases, monetary policy may only increase employment among printing press manufacturers.

Expectations are a big deal in practice, but in theory expectation still don't feel as concrete as printing money to lower interest rates. The Fed has worked to increase expectations, but the market ALSO expected TBTF to prevent such a negative Wicksellian rates from happening again.

In other words, it's hard to imagine a 2008 scenario, without bailouts, where QE2 would be good enough for market expectations to stay at 2% inflation.

Now, there are concrete policies below the zero-bound. My preferred policy is simply to eliminate or vastly reduce cash withdrawals as the Wicksellian rate goes below the zero bound. But either this proposal or Kimball's proposal runs into near-insurmountable political issues. "The Money Illusion" is too fixed in the mind of most people.

Past these cash limitation mechanisms, the proposals get stranger, such as NGDP futures (as OMO's themselves rather than informing other policies) or the Fed buying stock market indexes. Anyway, I can see the case for pessimism, especially for the ECB. Unlike the US, bailouts of peripheral countries and their banks have always been somewhat in doubt.

James Alexander writes:

You have it the wrong way around, I think.

The targeting of inflation projections is the heart of the matter. It is an easily understandable mistake coming from a psychological inflation-phobia and one that public or political pressure on a central bank can rectify. Hence, your tireless blogging.

The mysterious, unobservable Wicksellian interest rate is an academic toy by comparison and just a fig leaf for the inflation-phobia.

Have a good holiday. Stuff often happens when you are away so don't be too long.

Scott Sumner writes:

Brian, Good point.

Jeff, Good analogy.

Matthew, You said:

"In other words, it's hard to imagine a 2008 scenario, without bailouts, where QE2 would be good enough for market expectations to stay at 2% inflation."

Here's where we disagree. With sound monetary policy there would be no 2008. And if there were, I doubt QE would even be necessary. In other words, 2008 was the result of the Fed's tight money policy. With an expansionary policy that does not happen.

We don't need more "concrete steps", we need a new policy regime. Monetary policy only seems ineffective, because central banks have the wrong regime in place.

Thanks James, and yes, inflation targeting is part of the problem, but not all.

Matthew Waters writes:

Yes, I would say I'm on the board of the "People of the Concrete Steppes."

It's simple to see a situation where concrete steps matter. If the Central Bank is legally limited to conventional interest rate policy, then at 0% rates money becomes something like the gold standard. You could have rapid deflation, depending on the circumstances of the market.

If only QE of government bonds and negative IOR down to ~-0.25% are added, then a natural rate of -2% would probably exhaust QE and negative IOR.

In 2008, the Fed ended up doing many things past QE and negative IOR. The commercial paper facility, Bear Stearns and AIG bailouts and expanding collateral for the discount window. These precedents feed into expectations like QE announcements do. Furthermore, you have the automatic Keynesian mechanisms in the US which you have less of in the EU.

Either Kimball's proposal or using NGDP futures for open market operations are interesting concrete steps, but I have to think the Fed has to settle on something for negative natural rates.

Benjamin Cole writes:

There is some "theo-monetarism" involved in the constant genuflection to tight money, and perhaps saluting what people think is a right-wing escutcheon. I think also some people erroneously conflate "easy money" with federal spending. Others say any inflation is theft etc.

But is there more?

What do multinationals want?

My guess is they want a strong dollar, so as to make easier imports into the US economy (the world's largest consumer market). You get a higher exchange rate by tight money.

I suppose a strong dollar also finances overseas expansions, by lowering the cost of overseas facilities and labor.

Never underestimate vulgar Marxist analysis---although Marxist medicine is poison

James Alexander writes:

I was right about one thing: stuff happens while you are away. You probably know by now, though.

Gianni writes:

To be sure, Claudio Bprio has highlighted the perils of overreliance on the wicksellian neutral rate which may be biased by the building of previous large unbalances

Maurizio writes:

Great post.

But I have a doubt. You are basically saying that the Fed does not really have the power to set rates, because if it sets them too far from the Wicksellian rates for too long, you end up with either iperdeflation or iperinflation.

Ok. Now suppose someone replies as follows: "that is not *necessarily* true. Suppose the Fed keeps rates much lower than Wicksellian rates for 10 years, but during that same period there is some technological innovation (a positive supply shock) that prevents prices from rising. That is a counterexample: you get no visible inflation but a deviation of actual rates from the Wicksellian rates."

Does this reply make sense? If not, why? Is it using a different definition of Wicksellian rates?

Comments for this entry have been closed
Return to top