The monetary economics of Scott Sumner

Here is my latest column, on the monetary proposals of Scott Sumner.  You probably know Sumner from his blog TheMoneyIllusion and in my view he has become possibly the most astute commentator on monetary policy at this time.  Excerpt:

The Fed has already taken some unconventional monetary measures to
stimulate the economy, but they haven’t been entirely effective.
Professor Sumner says the central bank needs to take a different
approach: it should make a credible commitment to spurring and
maintaining a higher level of inflation, promising to use newly created
money to buy many kinds of financial assets if necessary. And it should
even pay negative interest on bank reserves, as the Swedish central
bank has started to do. In essence, negative interest rates are a
penalty placed on banks that sit on their money instead of lending it.

to the chagrin of Professor Sumner, the Fed has been practicing the
opposite policy recently, by paying positive interest on bank reserves
– essentially, inducing banks to hoard money.

The Fed’s balance
sheet need not swell to accomplish these aims. Once people believe that
inflation is coming, they will be willing to spend more money.

other words, if the Fed announces a sufficient willingness to undergo
extreme measures to create price inflation, it may not actually have to
do so. Professor Sumner’s views differ from the monetarism of Milton Friedman by emphasizing expectations rather than any particular measure of the money supply.

There are more excellent posts on Scott's blog than I am able to link to.  Read through it all, if you have any interest in these topics. 

One thing I learned from a systematic reread of Sumner is that he isn't quite the advocate of quantitative easing that I had thought.  All things considered, he seems to favor QE over doing nothing, but he also thinks that a truly credible commitment to future inflation can get us there without much painful-for-the-Fed's-balance-sheet QE being required.

While I think there is a very good chance Sumner is correct, my reread of his blog also gave me a better sense of, if he is wrong, why he is wrong or maybe incomplete is a better word.

In very general terms, think of our government, or central bank, as being able to do some good things by creating credibility, the rule of law being one example.  In this particular case the Fed could use its credibility to guarantee two to three percent price inflation annually or more exactly some target for nominal GDP growth.

One point is that bureaucracies tend to hoard credibility rather than to spend it.  That still could mean Sumner's advice is correct and this is simply why the Fed doesn't follow it.  There is, however, a deeper worry.  One possibility is that a weakened Fed cannot today precommit to delivering on two to three percent.  Let's say that Congress gets upset along the way, for whatever reason.  The Fed has then put its credibility on the line, including for the longer future, and that credibility is utterly refuted.  Ouch.  More technically, combine the two ideas of self-fulfilling prophecies and nested games.

Maybe the Fed is too risk-averse but there's also the possibility that the Fed is prudent in its unwillingness to stick its neck out.  Maybe the Fed has credibility only as long as it doesn't try to spend it (try modeling that).  This would bring us into the literature on creative ambiguity and signaling.

Another possibility is that, instead of Congress intervening, markets simply don't respond.  Sumner's theory makes sense to me, but how certain can we be?  The Fed again is putting a lot of longer-term credibility on the line.  Maybe the best the Fed can do is a kind of "inch-along" promise, which probably won't be very effective, as we are observing.

Perhaps the key question is just how credible a central bank can be, relative to its (possibly unjustified) risk aversion.

I now read Sumner much more as a "theorist of credibility," and thus as an implicit game theorist, than I used to. 


Why should there be more lending and borrowing when the economy is monstrously overindebted and overcapacity?

What we need to do is flush out all the excessive lending, not encourage more of it.

I now read Sumner every day also. He makes me think in new ways. I think his ideas of targeted NGDP have merit, but it is unclear if the Fed can accomplish this on it s own. it is also unclear if it can be done after so much damage has been done to the economy. Just what exactly will induce people to start spending again with jobs still disappearing? Do we really even want people to stop saving and start spending again so soon?


If you knew the world history of central banking, you'd agree that any central bank is as credible as this guy

This doesn't much impact the main point, but...

...but he also thinks that a truly credible commitment to future inflation can get us there without much painful-for-the-Fed's-balance-sheet QE being required.

Why do we care at all about the Fed's balance sheet? The Fed basically has more money than God.

The guys who can't shoot straight still won't be able to hit a different target any better than the last one.

If I am reading this correctly. You are equating monetary growth with real GDP growth. If that is your intent, it cannot work. Money is simply a unit of measure increasing the availability of the unit of measure cannot make the underlying activity any better.

@George: If I am reading this correctly. You are equating monetary growth with real GDP growth. If that is your intent...

Based on several months of reading Sumner's blog: no, that's not his intent. Sumner believes that 5% NGDP growth will generally yield about 2% inflation and 3% real growth.

Working to hold nominal GDP growth steady may result in temporary increases in inflation -- though it will also reduce inflation "below target" in periods of high real GDP growth (due to rapidly growing productivity, etc.). (i.e. below what a pure inflation target would yield)

To provide some context: the economy was doing a reasonable job of adjusting to the subprime problems in early 2008, and commodity prices were even increasing based on a worldwide boom (increased demand from China and India). If I have the timeline right: the Fed made a significant rate cut in early 2008 ... but didn't take similar action in the summer or fall as money got tighter again.

Both in the run up to Lehman's failure and right afterwards, people got nervous and the demand for money increased. Sumner believes that a strong stance by the Fed to accommodate the increased demand would have prevented things from falling apart. Yes, of course there were still financial problems, but those need not pull down GDP so sharply. One data point: the huge stock market drop in 1987 had minimal impact on "Main Street".

E. Barandiaran,

Why don't you try finishing reading his post before making wrong inferences?

You could also post a response at his blog. He reads them, and answers them. Talking to Tyler about it seems the wrong way to go about clearing up your confusion.

Although I agree with the inflation solution to reduce debt to GDP levels, I believe Sumner is missing a key point.

Money isn't neutral and in creating higher price expectations, some prices will increase more than others. This will be especially true because a credible 2-3% inflation target should raise interest rates from current levels and constrain price increases in interest rate sensitive sectors of the economy such as housing and autos, which are precisely the sectors the FED is trying to support.

Thanks Tyler, And an apology for skipping over the comments, because I am running late today (partly because I answer almost all comments on my own blog.) Here's my initial reaction. I think most good economists don't like the "mistake explanation" of systematic behavior. But I really think the world's central banks are making a mistake. They are afraid to jump in the water because they think it is colder than it really is. Most "responsible opinion" in the 1930s thought the UK and US devaluations would lead to armageddon. And yet things turned out OK.

I think they are currently too afraid of targeting the forecast, and level targeting, and prefer to stumble along in their old familiar ways out of fear of the unknown. If they tried it I think they'd like it.

You said:

"Maybe the Fed is too risk-averse but there's also the possibility that the Fed is prudent in its unwillingness to stick its neck out. Maybe the Fed has credibility only as long as it doesn't try to spend it (try modeling that). This would bring us into the literature on creative ambiguity and signaling."

This is a good point, and I admit that I hadn't thought of it, (if I understand you correctly.) Right now a lot of the Fed's power and prestige comes from its mystique, that fact that monetary policy is very complex. If they went my way with an explicit 5% NGDP target, or some inflation target, then it would immediately become obvious to everyone that we are dealing with a political issue, and Congress would say "we should set the policy goal." This is especially so if they relied on market expectations to determine the money supply and interest rate paths over time. No institution wants to become irrelevant in that way. As an example, the Fed became irrelevant between March 1933 and January 1934 when FDR ran monetary policy out of the White House by tinkering with the price of gold. Would Congress start tinkering with the NGDP target? If the Fed understands this then they may understand that once they start down the road I laid out, then politicization of monetary policy is the obvious endgame. In that case they would have no long run crediblity.

On the other had I think they might be able to convince Congress that a temporary emergency CPI or NGDP target was needed. But if it worked, would people ask "why don't you continue it?"

Is this the "Nixon was right" line of reasoning, inflation is worth it in order to get growth, and we can control wages and prices to make sure we get the benefits of inflation without the disadvantages of inflation?

If the problem is the exchange rates from China's taking its tax collections and buying dollars because it can't spend the taxes fast enough on infrastructure and public services, then the solution is for the US to hike taxes and spend on repair, rebuilding, and adding to our infrastructure, and increasing spending on universal health care. The higher taxes will force Americans to cut consumption and that will cut imports, and that will limit the dollars China will be able to buy as they are able to spend more of their tax collection inside China.

I have to agree with Matthew here, the way towards long term recovery is paying off our debts so we can build new credit when we need it.

Doc Merlin:

Personal saving was positive for all but two quarters in the last decade. While some households have greatly increased their debts, which is why household debt was high relative to income, other households have been lending to them and accumulating assets.

While personal saving perhaps should have been even higher, so that households were better preparing for retirement, the notion that all households are in debt is false. To say that "we" should decrease our debts is a claim that some households should pay back other households. It is possible to repay debt and expand consumption expenditures.

As for corporate finance, to say that firms are too "leveraged" is to say that firms should finance their operations by selling stock and not bonds. That is to say that households should by the stock of firms rather than bonds (and profitable firms that retain earnings should buy the stock of other firms rather than their bonds.)

Households as a whole can save more without firms going into debt, by firms financing their expansion by stock sales.

Total debt can decrease, even as the banking system holds more of that debt, and finances more of it with monetary liabilities.

There is no contradiction between an expansion in the quantity of money or a return of nominal income to its previous growth path, and a reduction debt.

For every borrower there is a lender. Net debt is always zero.

Everything is very open and very clear explanation of issues. was truly information. Your website is very useful. Thanks for sharing.

Comments for this post are closed