Archive for the Category Monetary Policy


Responding to John Becker

John Becker asked me this question:


If the government massively cut spending and the economy boomed, do you think Paul Krugman would wave a white flag? If the Fed announced they were adopting NGDP level targeting and the markets tanked, would you admit defeat or start scrambling for other reasons why markets might react that way? I don’t think either of you would be wrong in either case, I’m just trying to make the point that economics theories are non-falsifiable by external events.

If the Fed announced a 5% NGDPLT and the markets tanked, then yes, I’d admit defeat.  I can’t speak for Krugman . . .

In fairness to Krugman, the proposed test of my theory is much more decisive.

If it were a 2% NGDPLT target I’d be much less confident, indeed I think markets might fall.

PS.  For some strange reason I like my new Econlog post, blaming Keynes and Hayek for the Great Recession.

Three widely held theories discredited in 5 minutes

I woke up this morning to find that my very useful commenters had provided me with information showing that 3 widely held economic theories were discredited last night between 12:45 and 12:50 am.  These theories are Keynesian economics, RBC theory, and the theory of beggar-thy-neighbor policies.  Yes, these theories have been totally discredited dozens of times before over the past few years (or decades if you wish), but piling on is fun!

Here’s the first comment, by Cameron:

BOJ announces it will expand its QE program, Nikkei rises 4.5%!

The liquidity trap theory is dead. Please please please follow in their footsteps ECB.

Of course Keynesian economics predicts the QE announcement would have no effect on stock prices, as the new money would just sit there as excess reserves.  You are just swapping one low interest government asset for another low interest government asset.  (In fairness, I don’t know what the BOJ is buying in this case, but we also see big market effects when central banks just buy government bonds.)

And yet the economic blogosphere is full of people telling us “we just don’t know” if QE has any effect.  Yesterday I saw an article indicating that the “markets” were uncertain whether QE had had any effect.

Of course the real business cycle people will say; “yes, monetary stimulus will raise nominal stock prices, but that effect merely represents inflation.”  On to comment number 2, from Steve:

BOJ expands monetary base to 80T yen per year

The central bank says it will expand annual bond purchases to 80 trillion yen a year, up from the current 50 trillion yen. It will also extend the duration of bonds it holds to about 7-10 years.

The impact on markets was swift: dollar-yen jumped 1 percent to 110.26, while the Nikkei surged over 4 percent to its highest levels since September 25.

So the Japanese stock market also soared in dollar terms, and if you look at the intraday data the huge spike was right after the announcement.  RBC theory says it’s just inflation, and that real values should not be affected.  And yet Noah Smith tells us that RBC theory is alive and well. And I’m sure it is, in the elite journals.

[As an aside, there are people who claim that some RBC models allow for nominal shocks.  Sorry, but the sine qua non of RBC theory is that nominal shocks don’t matter.  If both types of shocks matter it’s not an RBC model, it’s an NK model with both supply and demand shocks being important. When people say they don’t believe that RBC theory is correct, they mean that they think nominal shocks matter.  Everyone believes that real shocks also matter.  After all, the devastating earthquake/tsunami/shutdown of all nuclear power in Japan caused a  . . . oh wait; it didn’t cause a recession, or an increase in unemployment.  OK, everyone agrees that a real shock 10 times bigger than the Japanese tsunami could cause a recession. Say Godzilla destroying Tokyo.]

The last resort of the anti-QE crowd is to admit that it “works,” but only in a beggar-thy-neighbor way.  Stealing growth from other countries via currency depreciation and trade surpluses.  But macro is not a zero sum game.  Commenter Steve also provides this information:

I was wondering why the S&P500 futures abruptly spiked 15 points (0.75%) at 12:45am EDT. TheMoneyIllusion has the answer!

Steve, I’m sure that was a coincidence.  There must have been was lots of other news at 12:45 am that would have caused broad indices to spike sharply higher in the US.  And if it was Japan, it was surely only stocks in US companies with operations in Japan.

Seriously, the extent to which people go to try to deny the obvious is almost comical.  I think it has something to do with Bernanke’s comment that QE works in practice but not in theory.  That’s not true of course, it works in theory by raising the expected future money supply and expected future NGDP, but it gets at reasons why economic bloggers deny the obvious.  They live in a world of theories, of models, not reality.

PS.  Cameron also provides this nugget:

Also debunked is the belief that consensus is more important than policy stance. The vote was 5-4 in favor of expansion.

Make that 4 theories debunked.  Not a bad day’s work for my comment section!  I don’t have links, but all comments are quoted in full, and are listed among the first 8 comments of the previous post. They are easy to find.

PPS.  This is another reason why we need a NGDP futures market.  Hypermind has started one, but it’s still small, and appears inefficient in my view.  The market forecast 3.9% NGDP right before yesterday’s announcement, which seemed too low to me.  The actual figure was 4.9%.  So there are still $100 bills on the sidewalk.  The prizes at Hypermind will grow sharply in a few days, so when I make the announcement get in there and start picking them up.

Maybe the economics profession doesn’t want NGDP and RGDP futures markets because deep down they know that Keynesian and RBC models would be totally discredited almost immediately. I can’t think of any other rational explanation.

Will it matter when the Fed has “traction?”

People have made all sorts of arguments against “monetary offset,” but there’s only one that actually makes much sense.  The argument is that the Fed does not like doing “unconventional policies” like QE, because they feel “uncomfortable” with a large balance sheet.  (Put aside the fact that QE is perfectly conventional monetary policy–open market operations—and that there is no reason at all to feel uncomfortable with a large balance sheet.  The Fed is effectively part of the Federal government.)

Nonetheless, there is a sort of plausibility to the theory; Fed officials will occasionally say they would cut interest rates further if they could.  But what is the implication of this theory?  It seems to me that this theory implies that Fed policy should become much more aggressive when the Fed is no longer hamstrung by the zero bound.  When they can stimulate without adding to the balance sheet.  But this raises an interesting paradox—the Fed is conventionally viewed as being “stimulative” when they cut rates.  Thus the Fed should want to cut rates as soon as they can do so, which means right after they raise them!

Of course I’m half-kidding.  More realistically the implication is that once the Fed stops doing the “uncomfortable” QE, there will be a long period of zero rates before they raise them.  And perhaps there will be, but right now the Fed suggests it will be raising interest rates in less than a year.

Here’s a graph from a Marcus Nunes post:

Screen Shot 2014-10-30 at 6.15.21 PMNGDP had been rising at about 5% per year in the 17 years before the recession, and it’s been rising about 4% per year in the “recovery.”  Because wages and prices are flexible in the long run, the real economy has been recovering despite the lack of any demand stimulus.  We have fallen from 10% to 5.9% unemployment.  But most people think the economy is still in the doldrums, and needs more stimulus.  President Obama just instructed the Department of Labor to increase unemployment compensation benefits (without any authorization from Congress of course–why do you think would Congress be involved in spending decisions?) This was done because unemployment is at emergency levels, requiring extra-legal remedies.

Fortunately the Fed is no longer doing the “uncomfortable” QE policy, which adds to the balance sheet.  So if you believe the fiscal policy advocates, the Fed should be raring to go with stimulus. How do they do that?  By promising to hold rates near zero for a really long time, or until the labor market is really strong.  But instead, they are suggesting that they will probably raise interest rates soon.  There will be no attempt to get back to the old trend line; the new one seems just fine.

Let’s consider an analogy.  A bicycle rider has a “policy” of maintaining a steady speed of 15 miles per hour.  Then he hits a long patch of ice, and slows to 10 miles per hour, perhaps due to a lack of traction, perhaps because he decided to go slower.  How can we tell the reason?  How about this, let’s put a strong headwind in his face, and see if the speed slows even more.  But now he petals harder and keeps maintaining the 10 miles per hour speed.  That suggests it’s not a lack of traction. But the pessimists insist it must be a lack of traction, why else would he have slowed right when he hit the ice?  Then the bicycle final comes to the end of the ice.  The lack of traction proponents expect him to suddenly speed up, exhilarated by the sudden traction of rubber on asphalt.  Oddly, however, the bike keeps plodding along at 10 miles an hour.  Nothing seems to have changed even though the ice patch is long past.

[In case it's not clear, the headwinds were the 2013 austerity, and the end of the asphalt was the end of the liquidity trap.]

Here’s my claim.  The Fed promise to raise rates soon is not the sort of statement you’d expect from a central bank that for the past 5 years had been frustrated by an inability to cut rates.  (Nor is their other behavior consistent—such as the on and off QE.)  Rather it’s the behavior of a central bank that has resigned itself to pedaling along at a slower speed.  Ten miles per hour is the new normal.

I don’t want to sound dogmatic here.  Obviously monetary offset is not “true” in the sense that Newton’s laws of mechanics are true; the concept only applies in certain times and places.  Oh wait, that’s true of Newton’s laws too .  .  .

Opponents of monetary offset face two big problems.  In theory, the central bank should target some sort of nominal aggregate, and offset changes in demand shocks caused by fiscal stimulus. And in practice it seems like they do, as we saw in 2013, even at the zero bound.  So if monetary offset is not precisely true, surely it should be the default baseline assumption.  Instead, as far as I can tell 90% of economists have never even considered the idea.

PS. Totally off topic, I love this sentence from an article on why a million dollars no longer makes you rich:

Although it sounds like a lot of cash, $1 million of today’s money is only worth $42,011.33 of 1914 dollars, which is less than today’s median household income.

Someone should collect all these amusing claims in the media.  They could have added that today’s median income of $42,011 is only equal to $1764 in 1914 dollars, roughly equal to the per capita GDP (PPP) of Haiti.  I guess I was wrong, the American middle class really is struggling.

The problem with procyclical inflation

Here’s Charles Evans in the WSJ:

“We’ve averaged well under that 2% mark for the past six-and-a-half years,” Mr. Evans said. “With a symmetric inflation target, one could imagine moderately above-target inflation for a limited time as simply the flip side of our recent inflation experience–and hardly an event that would impose great costs on the economy.”

There’s no doubt in my mind that a policy of letting inflation run a bit above target during the next boom will not cause great hardship during the next boom

But a policy of running inflation below target when unemployment is high and above target when it is low makes the business cycle much worse, and does impose great hardship.  Some conclusions:

1.  A procyclical inflation policy violates the dual mandate.

2.  NGDP targeting would lead to countercyclical inflation (a good thing).  As Nick Rowe likes to say, you want to make it so that the public’s stupid belief that inflation is bad . . . is true.  Good supply-side policies would become anti-inflation policies.

3.  Discussions of “what should the Fed do now?” are meaningless and incoherent, unless embedded in a clearly specified long run policy regime, as are discussions of whether QE “increases inequality.”

Charles Evans is actually one of the best people at the Fed.  Then there is the other Charles:

Federal Reserve Bank of Philadelphia President Charles Plosser said Friday that inflation levels that have fallen persistently short of where the central bank wants them to be are not a significant issue to him right now.

It’s true that inflation levels are “a little bit low” relative to the Fed’s desire to have price pressures hit 2%, Mr. Plosser said at an appearance in New York. But, “for the most part, I’m not too concerned about that,” he said.

What he doesn’t say is that the reason the Fed has failed is partly due to the fact that he’s consistently been pressuring them to be more contractionary, even as they were already far too contractionary to hit their dual mandate. So Plosser’s telling us that the Fed is not doing its job, partly due to his consistently bad advice, but he doesn’t much care.

Fortunately, market monetarist ideas are gradually seeping into the media.  A few days ago we saw this at the Financial Times, now it’s Bloomberg’s turn:

Based on the gap between yields of government notes and TIPS, traders have scaled back estimates for average inflation through 2019 by a half-percentage point since June to 1.52 percent, Fed data compiled by Bloomberg show.

.  .  .

With the Fed’s preferred measure averaging 0.34 percentage point less than CPI in that span, traders are signaling prices based on that gauge may rise as little as 1.18 percent. Through August, the personal consumption expenditures deflator has fallen short of the Fed’s 2 percent goal for 28 straight months.

Fed officials “need to be paying attention to that because there’s a collective wisdom element to the TIPS market,” Mitchell Stapley, the chief investment officer for Cincinnati-based ClearArc Capital, which manages $7 billion, said in an Oct. 8 telephone interview.

Did QE worsen inequality? That’s not even a question

When people ask whether QE worsened inequality they think they are asking a coherent question. But that merely shows how poorly most people understand monetary economics.

Let’s ask a different question:  Did Obama’s appointment of Ben Bernanke increase inequality? Any sensible listener would ask: “Compared to what?”  After all, most models are roughly linear, at least for very small changes (I’m rusty at math, so tell me if that is wrong.)  In other words, whatever impact monetary policy has on inequality, the impact of picking Bernanke over a more dovish alternative (Romer) would have been the opposite of Obama picking Bernanke over a more hawkish alternative (Summers.)   I can’t imagine anyone being able to make sense of the question “did Bernanke increase inequality” without knowing the counterfactual Fed chair.  And of course the same is true for Fed policies, is the counterfactual more or less contractionary than the actual policy?

Now some people will say; “the obvious implication is that the counterfactual was no QE, and that this was a more contractionary alternative.”  This is very likely how people think about it, but of course that assumption is wrong.  My preferred policy would have been far more expansionary, and hence would have involved far less QE.  Let’s break this down into 2 questions:

Does monetary stimulus increase inequality?

Does delivering monetary stimulus via QE affect inequality more than some other method?

I’ll take the second question first.  Suppose Bernanke did not do QE, but rather some equally effective stimulus method.  Perhaps slightly raising the inflation target, or going to level targeting. Would that make any difference for inequality?  I hope it’s obvious that it would not.  The mechanics of QE are totally uninteresting.  You are just swapping one Federal government interest bearing liability (reserves) for another federal government interest bearing liability (T-bonds.)  Any “Cantillon effects” are trivial.  I hope I don’t have to explain to people that this “money” did not “go into the stock market”:

a.  The money went into bank reserves, or currency.

b.  Money never goes into markets; there is no giant safe on Wall Street storing all the money invested in stocks.  Money goes through markets.  You buy, someone else sells.

If there were no QE, but equally fast NGDP growth produced by a higher inflation target, stocks would have done equally well.  Indeed stocks responded more strongly to forward guidance than QE3 in late 2012.

So now we can rephrase the QE question: “Did Bernanke’s monetary policy since 2009 worsen inequality?”  Now it’s much easier to see that we need a counterfactual.  You might prefer to describe that policy as 1.5% inflation, or perhaps 4% NGDP growth (my choice.)  Either way it’s a fairly contractionary policy.  And it’s no longer “obvious” what the counterfactual is, would it be 3% or 5% NGDP growth?  In my view 5% growth would have helped the unemployed and the rich more than the middle class with stable jobs (say teachers.)  So that has mixed effects on inequality, indeed so ambiguous that it’s probably not worth even thinking about, as the effect would be trivial compared to the net gain to America from a stronger economy.

If you think the alternative to QE was a more contractionary policy, say 3% NGDP growth, then it would hurt the rich and poor more than the middle class.  In order to favor that policy you’d have to hate the rich so much that you be willing to impoverish millions of poor people to screw the rich. But even someone who hates the rich as much as Paul Krugman favors QE.

Sorry, but “does QE increase inequality?” is a really, truly moronic question.  I apologize for wasting your time.

PS.  Here’s Buttonwood at the Economist:

This is at the heart of the matter. Even if the Fed does not increase rates next year, it will surely take a big economic shock to make it resume QE. The markets have relied on the central banks for so long, like a small child holding his dad’s hand when learning to ride the bike. It is time to let go of the hand now, but there will be a few bumps and bruises along the way.

This is truly a horrible metaphor, and helps explain how the developed world got so far off course. Taken literally, the counterfactual to “using monetary policy” is barter. Obviously that’s not what people mean when they say it’s time to stop using monetary policy.  Buttonwood probably means that we are propping up the economy with an excessively expansionary monetary policy.  But of course that’s confusing the tools (fed funds targets, the monetary base, etc.) with the actual policy itself (1.5% inflation, 4% NGDP growth, etc.)

By 2007 almost no serious economist in America believed that money was “easy” in the early 1930s, despite ultra-low interest rates and massive QE.  And now almost all serious economists believe monetary policy has been “easy” in recent years precisely because of ultra-low rates and massive QE.  This fact is appalling.  The intellectual decline in mainstream macroeconomics since 2007 is stunning–nothing like this regression has happened since the early 1970s, or perhaps the late 1930s.  And this time the worst mistakes are being made by those on the right.

By the way, the right metaphor is not training wheels, but rather which way do you want to steer the bicycle?  No serious pundit is advocating walking.