Scott Sumner  

What's the "sell by date" of Krugmanomics?

Bryan Caplan's Notes... Determinism and Motivated Reas...

In the 1990s and early 2000s Paul Krugman basically employed a classical approach to policy analysis, with opportunity costs playing a central role. (Note that this should not be confused with New Classical claims of the non-importance of demand shocks, a view Krugman quite sensibly avoided.)

In recent years he's argued from more of a traditional Keynesian perspective, with the concept of the liquidity trap almost always lurking somewhere in the background. Here's an example:

The way to deal with secular stagnation, if we believe in our models, is to raise the long-run neutral interest rate above zero. If we can do this via structural reform and/or self-financing infrastructure investment, fine. If not, raise the inflation target.

And how do we get to the higher target inflation rate, when monetary policy is having trouble getting traction? Fiscal policy! If you're really worried about secular stagnation, you should advocate a combination of a raised inflation target and a burst of fiscal stimulus to help the central bank get there.

I don't agree that we need to raise the inflation target, but I do agree that it would work. Indeed despite the fact that I am opposed to raising the inflation target, I so dislike current Fed policy that I would consider a 3% inflation target to be the "lesser of evils."

Another example of this liquidity trap economics occurred in a March 11 post, which I discuss over at TheMoneyIllusion. So the liquidity trap has become a central organizing principle of his recent policy analysis.

But here's the problem. Most experts, and also the financial markets, expect the US to escape the liquidity trap within a few months. That would almost instantly make Krugman's policy analysis look obsolete. Will he then stop making arguments that rely on the zero interest rate lower bound? (Now assumed to be more like a negative 1% interest rate lower bound.)

I believe Krugman will not abandon the Keynesian framework that he has recently invested so much effort in promoting. And my supporting evidence comes from the eurozone, which was not at the zero bound until quite recently. And yet Krugman applied zero bound arguments to the eurozone in the 2009-12 period, despite interest rates being above zero, and indeed even rising in 2011.

What would be Krugman's counterargument to the points I'm raising? Probably this, which immediately precedes the previously quoted passage:

But the assumption here is that the neutral rate will eventually rise, so that monetary policy can take over the job of achieving full employment. What if we have doubts about whether that will ever happen?

Well, that's the secular stagnation question. In fact, I'd define secular stagnation as a situation in which the neutral interest rate is normally, persistently below zero. And this raises a puzzle: If we worry about secular stagnation, should we then say that St. Augustine no longer applies, because better days are never coming?


Here Krugman doesn't discuss the actual policy rate, but rather the Wicksellian equilibrium rate, which is the interest rate necessary to generate macroeconomic equilibrium. Even when the Fed raises interest rates to 0.5%, or 1.0%, Krugman may continue to argue that the economy is so depressed that even a cut in rates to 0% would not do the job, and hence fiscal stimulus is needed. But I see several problems with this argument:

1. If that were true, then the interest rate increases expected later this year should drive us into a double-dip recession. Now that might happen, and did happen when the ECB raised rates in 2011, but as far as I can see the markets don't currently expect this outcome.

2. Recent data suggesting the actual lower bound is closer to minus 1% doesn't completely undercut the Keynesian argument, but it weakens it a bit. And it weakens the argument even more if interest rates rise to 1% or more. Then they would be at least 2% above the actual lower bound---pretty far from a liquidity trap.

3. However there is one problem that seems far more important that all the rest. And to understand this problem it will be helpful to back up and explain how my views have differed from Krugman's over the past 6 years. I've argued that monetary policy is still highly effective at the zero bound, which is of course the standard textbook view, or at least was until 2008. The Fed can do other things to boost aggregate demand; so fiscal stimulus is not needed.

Krugman had one advantage in this debate; most people (wrongly) envision monetary stimulus in terms of rate cuts. So they were skeptical of my claim that monetary policy was still highly effective, and indeed the preferred policy tool. But once rates rise above zero, Krugman will lose that advantage. When he calls for fiscal stimulus in a positive rate environment, people will ask why not try monetary stimulus first? Even worse, if you respond that the Fed won't do that because they are inflation hawks, your fiscal stimulus arguments collapse like a house of cards. After all, if the Fed is opposed to higher inflation, then they will simply negate the inflationary effects of more fiscal stimulus.

So I'm afraid the sell by date of Krugmanomics in the US is very short, perhaps 6 months. Then the debate will move on to other topics.

Suppose I'm wrong and rates don't rise later this year? And indeed, this is a possibility that I think is quite plausible, maybe a 30% chance. Would Krugman have been correct?

I would still disagree with his advocacy of fiscal stimulus, as I prefer unconventional monetary stimulus. However I would agree that his argument would be more persuasive in the zero rate forever environment than in a positive rate scenario. The best example of a zero rate forever country is Japan, and yet even they are moving strongly in a market monetarist direction, with tight fiscal combined with monetary stimulus. So far all they have to show for it is the ending of 15 years of deflation, the lowest unemployment rate in decades, and a doubling of the stock market. In other words, all the Very Serious People say it's been a complete failure. (Oddly, the limited success of Abenomics is one area where Krugman and I agree. Although he favors fiscal stimulus, he also likes monetary stimulus.)

PS. Just to be clear I don't think Abenomics has been completely successful. As I've predicted all along, they are likely to fall short of their 2% inflation target unless they do much more, and of course I don't think inflation targeting is a good idea in the first place.


Comments and Sharing


COMMENTS (11 to date)
Brian Donohue writes:

I wonder if Krugman plays chess. If he does, he might be wanting to take back his last couple moves here.

ThomasH writes:

We will continue to "need" almost as much public investment when the Fed (probably far too soon) raises interest rates to 0.5% interest rate as we did at 0.0% because the value of activities with positive NPV at 0.5+% discount rate will be almost as at 0.0+%, especially since governments were not in fact investing as if the discount rate was 0.0+%.

Duncan Earley writes:

I have a dumb question (I am not an econ major)... Why cant the government print money and use it for fiscal policy? As long as they don't print too much as to cause excessive inflation. In other words why do they print money to buy mortgages (QE), but not print money to build bridges?

Trevor Adcock writes:

Duncan the Fed gives all its profits to Treasury so QE does fund bridges and social spending. If the Fed didn't buy assets and just gave the money to the Treasury it would have to rely on the Treasury to destroy any excess cash if monetary policy became to expansionary. This would require distortionary taxes and would undermine monetary policy independence.

It's better that the Fed just hold some safe assets that it can sell to contract the money supply if needed.

Nick writes:

I think Krugman is still in a good spot. He can stop favoring more 'stimulus' while still supporting more spending on stuff he feels has value (even if not always economic). He can still oppose new cutbacks even in inefficient programs because of fear of a new liquidity trap. Even if we get up to 1% it will be pretty clear that the fed will sweat about zero the next time they need to loosen.
As for responding to responsible people who just want to cut the deficit, Krugman claims to be very happy to raise more revenue by only taxing the very wealthy...

Brian Donohue writes:

Trevor, great response.

Two other thoughts:

1. If the Fed is targeting some kind of path for inflation/NGDP, it would take into account the impact of any fiscal stimulus in its calculations (monetary offset).

2. American taxpayers are furnishing various levels of government with $5.6 trillion in real (not printed) tax dollars this year. That's $19,000 per citizen. Many of us feel that this is quite enough. Yet today, 'austerity' means that governments in the US are only spending $6.1 trillion this year, a $500 billion deficit.

It's about the size of government.

Dan W. writes:


When you tout the doubling of the Japanese stock market are you not "reasoning from a price change"? More importantly, financial engineering, even when conducted by a central bank, is still just financial engineering. If one wishes to brag about the economic success of an economic policy should not the evidence be economic, and not financial? Consider the US stock market. How much of the increase in US equities is due to economic performance? How much of it is a result of increased expectations (ie PE) and stock buybacks, including those realized by buying stock with new debt? Aggressive monetary policy appears to have the primary consequence of juicing financial markets. That is not a bad thing but it can create a conflict of interest. For if it is perceived that an interest rate increase would "crash" the market would that cause the central bank to delay taking such action? And if this fear is real then who is in charge? Wall Street or Yellen?

Scott Sumner writes:

Thomas, That may be true, but it doesn't have any bearing on this post. If fiscal stimulus is not needed, the optimal budget deficit is smaller.

Duncan, It's much more efficient for the central bank to buy bonds than to buy goods.

Trevor's response is better.

Nick, I predict he will continue favoring fiscal stimulus.

Scott Sumner writes:

Dan, It's not reasoning from a price change because stocks rose on news of monetary stimulus.

I did point to economic indicators: inflation and jobs.

Stocks are an indicator of futures expectations about the economy, among other things.

J.V. Dubois writes:

"Krugman had one advantage in this debate; most people (wrongly) envision monetary stimulus in terms of rate cuts."

This is key. And people not only wrongly think this, Krugman actively promotes such an idea. Like in this blogpost written just few days ago:®ion=Body

I am really surprised that nobody commented on this. According to Krugman the recession in 80ties was created by FED tight money policies. So far so good. But this is what is surprising to me:

"The first [81-82 reccession] was caused by tight money, imposed by the Fed to curb high inflation. As a result, there was plenty of room to cut rates (and also pent-up housing demand)."

So according to Krugman in 80ties FED first created the recession and only THEN lowered interest rates to pent-up demand. While in reality permanently lower interest rates were necessary product of tighter monetary policy (permanently lower inflation and NGDP). Which is very similar story to what we experienced recently.

It is very hard to have productive discussion where there is such a large difference in what actually defines tighter/easier monetary policy. That this question was never thoroughly solved and fact that Krugman can write things like this without any prominent keynesian blogger pointing his mistake is extremely frustrating.

Lauren writes:

Duncan Earley asked a great question:

Why cant the government print money and use it for fiscal policy? As long as they don't print too much as to cause excessive inflation. In other words why do they print money to buy mortgages (QE), but not print money to build bridges?
While Trevor and Scott both offered excellent answers, I interpreted Duncan's question as asking about something quite different and more basic--a question about the Fed's operating regulations. Why can't the Fed just buy whatever they want? What are their legal limitations? Why can't something we call "the government" just print money and use it however it wants?

The Fed can, in principle, print money and use it to buy anything it wants. The Fed can even buy your great-grandma's rocking chair if it wants. It could buy haircuts for us all. It could buy highways. It could buy your house. It could fund political candidates. It could give money directly as foreign aid in circumstances of cyclones, earthquakes, or war. It may have bought your mortgage a few years ago as part of a block deal. In principle, the Fed could come up with theories or aims or purchases to accomplish whatever goals it thinks are justified at any time, including even the personal goals of its members. Fiscal policy--the traditional domain of the Congress, the Treasury, and the President--surely could be one of the Fed's direct goals.

Instead, and based on years of preceding history, the Fed is limited by Congress and also limits itself. Those limitations are imposed in part by the Fed's own internal rules, in part by custom based on the history of how Central Banks such as the Fed have evolved, and in part by the Fed's Congressional Charter.

What it buys in accord with those limitations are generally limited by these rules or customs to a small range of certain assets that are, at least in the eyes of the public, a bit safer and a step removed from what historically has gotten out of hand. Those assets purchasable by the Fed historically include such things as gold, government bonds issued by one's own government--so, the U.S. Treasury or the carefully supervised bonds of designated member banks--or trusted foreign governments or international institutions' assets, etc. The idea is that if the Fed buys something and is authorized to print money to do so, the item should at least have some chance of being able to be sold later at the same or at least at a similar value. Not something like your 10-year-old car even if it's got another 10 years in it. Not a risky bond or stock holding in a private business, be it established or startup. And not even something like roads or highways, which can be subject to business failure, overpricing, or political pressures. The Fed is entrusted to buy with its printed money something that in principle it can hold on its books which has some clear and stable market value. Something we the public can look at in the Fed's books and potentially sell at the same or better price. Something stable, at least in the first instance, so that we know that the money we privileged it to print was not frivolously or overtly corruptly spent.

Historically, there have been many bad events which modern central banks like the Fed try to avoid citizens' thinking they could fall into repeating. Direct funding of private or government projects by a central bank often has led to corruption in the past. Direct buying of government loans--fiscal spending--by a central bank has led to hyperinflation in cases where central banks are expected or even required to fund a country's fiscal policies directly.

This might explain to you why, when the Fed extended what it would buy at face value and hold on its books to hard-to-value and possibly worthless assets such as mortgage securities in the aftermath of the 2008 Financial Crisis, there was an uproar--by the public and by economists alike. Was the Fed using its authority to buy safe assets that were going to be salable even at par in the future? At half their value? Was it using its authority to cover up helping some individuals or private businesses at the expense of others? How was that different from the Fed's speculating in your great-grandmother's rocking chair or buying the depreciated assets of other failed business sectors in the economy? Why should the Fed not instead buy millions of dollars' worth of failed tech startups? struggling farms, grocery stores, highway bonds, etc.? Where is the line drawn? The real issue is exactly the question you asked: Why can't the government print money and use it for fiscal policy? The answer is deeply about the essence of what central banks do when they print money, and how to separate what they do currently from other things they might do or might better not do to best help an economy.

Comments for this entry have been closed
Return to top