Archive for the Category Monetary Policy


Mian and Sufi on monetary policy

Unfortunately I’m not able to keep up with all the new blogs, but I’m told that Atif Mian and Amir Sufi are brilliant new bloggers who have great ideas on debt.   Ryan Avent showed that their expertise on monetary economics is much weaker.  Here he discusses a graph they present that showed PCE core inflation has fallen below a 2% trend line since 1999:

Unfortunately, I think they’ve got this wrong in a few ways. First, the Fed doesn’t target core inflation. It targets headline inflation, but it uses core as an indicator because past core inflation is a better predictor of future headline inflation than past headline inflation. So here’s something interesting: take a look at what happens when you track headline inflation (as measured by the price index for personal consumption expenditures) since 2000.

Finally everything is clear: the Great Recession was a necessity engineered by the Fed in order to disinflate back to the 2% trend. I’m kidding, of course. In fact, this is the wrong period to consider entirely, because the Fed didn’t adopt an official inflation target until January of 2012.

Why did Mian and Sufi do a study of actual inflation compared to target inflation?  Here’s why:

The chart above plots the implied core PCE index if inflation had met its 2% target (red line), and the actual core PCE index (blue line) starting from 1999. The blue line is consistently below the red line, the gap has only diverged further since the Great Recession. The cumulative effect is that today the price level is 4.7% below what it should have been had the Fed achieved its long-run target…

What we are witnessing is the limit of what monetary policy alone can do. Sometimes there is a tendency to assume that the Fed can “target” any inflation rate it wishes, or that it can target the overall price level – the so-called nominal GDP targeting. The evidence suggests that the Fed may not be so omnipotent.

So their study was aimed at establishing whether the Fed is able to hit its 2% inflation target.  They found it was not, on the basis that core PCE fell 4.7% below trend over 14 years.  (Put aside the fact that in the 1970s, before they were inflation targeting, inflation often overshot the target by 4.7% in less than one year.)

If you are a sweet, naive, trusting, honest, idealistic soul, you will naturally assume this story has a happy ending.  Ryan corrected the data error.  With the correct data the study shows the Fed in fact hit its target almost perfectly in the long run. Great news!!  So Mian and Sufi will naturally change their conclusion to fit the actual outcome of the study that they themselves thought provided a window into whether the Fed was able to successfully target inflation.  Just as Paul Krugman changed his mind about MM after the results were in from what Paul Krugman himself said would be a 2013 test of MM. They will print a retraction, and change their forthcoming book to show the conclusion that is in fact correct.  The Fed can target inflation at 2%.  Because we are all scientists, aren’t we?  We learn from the results of our tests.  I very much hope you are right, but just in case they do not change their conclusions regarding the ability of the Fed to hit a 2% inflation target, let me explain why.

Most economists are not interested in finding the truth; they are interested in using ideas to advance their career.  Empirical studies become swords in the battle, to be used when effective and thrown away when they are found useless.  I sincerely hope Mian and Sufi are not like most economists.  (And to be fair, there have been times when I slip into bad habits as well, so perhaps I should not be throwing stones here.)

Ryan Avent was actually pretty kind to Mian and Sufi, as he overlooked this:

What we are witnessing is the limit of what monetary policy alone can do. Sometimes there is a tendency to assume that the Fed can “target” any inflation rate it wishes, or that it can target the overall price level – the so-called nominal GDP targeting. The evidence suggests that the Fed may not be so omnipotent.

First of all, price level targeting is not at all NGDP targeting.  I apologize if this sounds snarky, but they really need to get up to speed on the 2014 blogosphere debate over monetary policy.  After everything that has happened you simply cannot still be arguing that monetary policy is ineffective at the zero bound, and use as “evidence” the fact that low interest rates have failed to push the rate of inflation higher. That would be like claiming that the fact that; “more police patrol high crime areas proves that police patrols are ineffective.”

Fiat money central banks can debase their currencies if they chose to.  So far as I know none of them really deny this.  You can’t be a serious monetary economist in 2014 and claim that a fiat money central bank can’t debase its currency.  You can claim there are political barriers.  Savers would be upset.  Or foreign countries would complain if you depreciated the yen in the forex markets. Or you’d have to buy so many assets that the central bank would be uncomfortable with the size of its balance sheet.  Don’t get me wrong, I think even those arguments are wrong, but they are defensible.  But in 2014 one simply CANNOT any longer argue that the fact that low rates and QE have been accompanied by low inflation proves central banks are out of ammo.  Too much has happened, the debate has moved on.

If you don’t trust me and the MMs, read Ben Bernanke, Michael Woodford, Christina Romer, Bennett McCallum, Milton Friedman, etc, etc.

PS.  I promised a friend that I would do a April Fools post.  I apologize, but I simply couldn’t think of one that was plausible.  Nobody would believe it if I claimed to have converted to MMT.  And then I tried to come up with whacky news stories, a la The Onion.  Like a story that the IRS ruled that Bitcoins were property, so if you used Bitcoins to buy a Coke from a vending machine, you had to file a tax form for the capital gains on the difference between the 90 cents you paid for the Bitcoin, and the $1.25 is was worth when you bought the Coke. Or that the IRS offered me $10,000 per year to divorce my wife, continue living with her, and continue to tell all our friends that we were married (establishing common law marriage.) But every time I thought up a whacky story like those two, I realized it was true.  I feel I live a sort of Groundhog Day film, where every day is April 1st.  Where very day someone repeats that low interest rates show easy money doesn’t “work.” Where high interest rate show that money was “tight” during the German hyperinflation.  I can’t keep up with reality.

Update:  Marcus Nunes also has a good post on Mian and Sufi.

Academics are rapidly catching up to market monetarists

Market monetarists have consistently argued that:

1.  QE is not risky, stopping QE is risky.

2.  Abenomics would modestly boost NGDP and RGDP.

3.  NGDP targeting wouldn’t just stabilize the labor market, it would also reduce volatility in the credit markets.

Here’s The Economist:

Jitters about market bubbles are one reason the Federal Reserve is dialling down its bond buying. A new study by Gabriel Chodorow-Reich of Harvard University shows that since 2013 Federal Reserve committee members, including Mr Bernanke, have cited concerns over increased financial-sector risk-taking as a reason to mute QE. Their anxiety is understandable: central bankers are still scarred by the lessons of the mid-2000s when banks “searched for yield” amid low interest rates, financing riskier projects and pumping up leverage to improve profits. After five years of shedding risk since the crisis that followed, some fret they could flip back into risk-seeking mode.

But those worries wither under closer scrutiny, as Mr Chodorow-Reich shows. He starts his hunt for a link between QE and risk with banks and life insurers, examining market reactions to 14 Federal Reserve policy announcements between 2008 and 2013. Using minute-by-minute data, and isolating small windows either side of each statement, Mr Chodorow-Reich measures market perceptions of risk. He finds that QE extensions are associated with a drop in the costs of default insurance that protects against a bank or insurer going bust. Markets, then, are not worried about QE, even if the central bank is.

And this:

And bold monetary policy has a big upside, suggests a new paper on Japan’s “Abenomics” by Joshua Hausman of the University of Michigan and Johannes Wieland of the University of California, San Diego. Japan’s monetary boost is huge, including a new 2% inflation target, unlimited asset purchases and a doubling of the money supply. Many worried, however, that it would not work. Japan’s slump is decades old and QE had already been tried. Between 2001 and 2006 the Bank of Japan boosted the cash that lenders held from ¥5 trillion ($46 billion) to almost ¥35 trillion using QE. Yet not much happened. Although inflation nudged above zero, policymakers increased rates too soon. By the time Shinzo Abe took office in December 2012 prices were falling by 0.1% a year and the economy was drifting sideways.

But Abenomics has lifted Japan’s GDP by up to 1.7%, according to Messrs Hausman and Wieland: up to a percentage point of that may be due to monetary policy. The market effects are clear: stock indices jumped and the exchange rate depreciated sharply when the policy was announced (see left-hand chart). The effects on broad money, which rises with bank lending, have been much stronger than with previous QE attempts (see right-hand chart).

Inflation expectations explain the difference. Abenomics was announced not as a temporary boost but a permanent change in policy. People quickly anticipated that prices would begin rising by 2% a year, instead of remaining flat. Long-run inflation predictions have risen too. This means borrowing looks more manageable and gives shoppers confidence to spend more. Nonetheless, Japan’s economy remains weak. Reinforcing the commitment to monetary boldness would give it a boost, the researchers say.

As far as I can tell, Japanese annual inflation rates (CPI) rose above US inflation last November, for the first time in nearly 40 years (excluding a period of a few months after the Japanese instituted a national sales tax in 1997.)  However the plan is still flawed as it does not represent level targeting.  Japan needs to do more to assure that it doesn’t slip back into deflation.  And why not do more?  All the good things predicted have happened, and none of the bad things critics worried about (like higher Japanese government bond yields.)  It was a free bento box. When the world offers free lunches, go to Jiro’s in Tokyo and pig out.

There are even more radical options. Kevin Sheedy of the London School of Economics reckons that gains may be made from replacing an inflation target with a nominal-GDP (NGDP) target. Typically central banks focus on inflation as this helps stabilise the value of pay, which might otherwise be eroded by rising prices. But wages are not the only rigid contracts workers face—their debts are fixed too. This means that a GDP shock which lowers incomes can cause a big jump in their debt burden.

A central bank focused on NGDP would help, Mr Sheedy argues. Take a supply shock, which tends to lower GDP without causing prices to fall. A central bank focused on prices might not respond at all due to the absence of inflation. An NGDP targeter would be bolder, stimulating the economy to generate inflation and keep the value of debt and GDP aligned. Hard-wiring a shift like this into the monetary system will take a lot of persuasion. But household debt-to-income ratios were much lower when inflation targeting was set up in the early 1990s. In today’s high-debt world, an NGDP target looks more attractive.

How influential these studies will be remains to be seen.

A few comments:

All three papers were published by the Brookings Institution—a respected think tank.

Ben Bernanke recently joined the Brookings Institution.

Even though MM is winning, we aren’t given credit.  But that’s always the way things work.  What matters is that the ideas get adopted. If not being associated with a fringe groups of academics at small institutions helps the ideas gain respect, then I hope we MMs remain completely anonymous.

PS.  Congratulations to Bentley’s women basketball team for winning the national championship (division 2.)  And for the University of Wisconsin (my alma mater) making the Final 4.

PPS.  I have a multiplier post that comments on Nick Rowe, over at Econlog.



What do the VSP think about the current economy?

This is a sincere question.  Do the Very Serious People think the US economy is deeply depressed, or nearing full employment?  I see conflicting evidence:

1. As far as I can see the policy of tapering has overwhelming support among the VSP, suggesting they think it’s time for the economy to stand on its own two feet, without a crutch from the Fed. Is that correct?

2. As far as I can see most of the press and pundits, and moderate politicians, seem to support an extension of the emergency unemployment benefits? Recall that the proposal would y currently extend them to 73 weeks, which is far more generous than the emergency benefits under Bill Clinton in early 1993, when unemployment was even higher. The call for an emergency extended unemployment benefits program would imply emergency level unemployment rates, where monetary tightening (when inflation is also below target) would be utter madness.

I’m not interested in arguing these two perspectives.  Both are defensible.  Rather I’m trying to get a sense of where the VSP opinion is on this issue. I honestly cannot tell.  Do they think the economy is deeply depressed?  Or not?

PS.  I’m a bit less sure on point two, but a number of GOP lawmakers recently indicated that they would support extended unemployment benefits, which suggests that the center of gravity political was for extension–that’s certainly the overwhelmingly popular view among press and pundits.

PPS.  Totally off topic, but this comment by Greg Mankiw intrigued me:

In this case, the issue is the reduction in capital taxes during the George W. Bush administration.  Paul [Krugman] says that the goal here was “defending the oligarchy’s interests.”

Really? As Paul well knows, there is a large literature in economics suggesting that an optimal tax system imposes much lower taxes on capital income than on wage income (or consumption).

Why should we assume that Paul Krugman is aware of that literature?  You would think he would be aware of the literature—other progressive bloggers like Matt Yglesias and Brad DeLong obviously are.  But I don’t recall reading a single Krugman column that showed any awareness of the need for replacing our current tax system with a progressive consumption tax.  I don’t recall a single post pointing out that taxes on capital should be much lower than taxes on labor income, if not zero.  I don’t recall a single post pointing out that nominal tax rates on capital income are meaningless, and that real tax rates on (for instance) Treasury bonds are now well over 100%.  Or that corporate income is triple taxed, making nominal corporate tax rates utterly meaningless as indicators of progressivity.  Or that Warren Buffet was spouting utter nonsense when he claimed his tax rate was lower than that of his secretary.  If such Krugman posts exist then please show them to me, I’d love to read them.

Either he is unaware of the literature, or he is aware of it and is knowingly spouting misinformation. I’d prefer to be charitable and assume that he’s not aware of the literature.

Mark Sadowski on Cullen Roche

Here’s Mark Sadowski (much of the following is Mark quoting Cullen):

I want to focus on one particular claim by Cullen Roche which he repeats three times in various forms at two different posts.

Cullen Roche
“Depends on the banking system and the “money” you’re referring to. “Federal Reserve Notes” are not actually issued by the CB even though they’re a liability of the CB. They are purchased by the CB and printed up by the Bureau of Engraving. Reserve Banks purchase coin at face value from the US Mint (who is determining your “conversion” rate there – a branch of the US Tsy or the Fed?). So, in a strange sort of way, the Treasury acts as a banker to its own bank and charges that bank quite a bundle in fees to deal in “legal tender”. There is this circuitous/hybrid relationship here that I think gets bungled by a lot of people.

So, when Scott Sumner refers to reserve notes as “paper gold” he’s really not referencing the power of a Central Bank. He’s actually referring to the issuer, the US Treasury which makes Sumner an ironic/funny kind of way. And this is the problem with parts of Market Monetarism. If you don’t respect specific institutional arrangements you end up saying things that make no sense when you consider the actual design of the system.”

Cullen Roche:
“…1) I think you overstate the exchange rate concept given that the Fed buys coin and currency from the US Tsy and the fact that there’s deposit insurance via FDIC. US bank deposits are at a very low risk of ever trading below par. The “value” of commercial bank money has been backstopped for all intents and purposes and the Fed is far from the only cause of this effect…”

Cullen Roche
“I understand Nick’s point. The thing is, the “unit of account” in Sumner’s theory is not even created by the CB. It’s created by the Bureau of Engraving. All cash and coin is sold to the Fed at face value by the Bureau, a department of the US Tsy. The alpha bank, if we’re going to use such a term, is actually the US Treasury in the Market Monetarist model and they don’t know it because they don’t actually describe the reality of how money is created. Sumner’s “paper gold” is a creation of the US govt, not the Fed…”

For this point forward it’s mostly Mark’s reply:

In a word, this is wrong.

“The distribution of coins differs from that of currency in some respects. First, when the Fed receives currency from the Treasury, it pays only for the cost of printing the notes. However, coins are a direct obligation of the Treasury, so the Reserve Banks pay the Treasury the face value of the coins…”

The cost of printing the notes has averaged approximately 0.325% of their face value in the past 11 years:

Of course the Federal Reserve pays 100% of the face value for coins. What proportion of the value of the currency in circulation consists of coins?

“…The value of U.S. coins in circulation as of May 31, 2010, was approximately $40.4 billion, or about 4.3 percent of total currency and coin in circulation.”

So the Treasury only gets paid about 4.6% of the face value of the nation’s currency in circulation by the Federal Reserve, which is appropriate if the Treasury is really in the role of performing a service for the Fed.

As Cullen says:
“If you don’t respect specific institutional arrangements you end up saying things that make no sense when you consider the actual design of the system.”

Kind of “ironic/funny”, eh?

I just have one question for Mr. Roche.  If the Fed buys currency at par, how did the Fed earn its seignorage prior to 2008?

Could we have had a severe recession without the 2008 financial crisis?

Paul Krugman argues that it would have taken a dramatically different set of policies back in 2007 to prevent the Great Recession, and Brad DeLong argues that a much more modest set of initiatives might have sufficed.  There is much to agree with in both posts.  First here’s Krugman, quoted by DeLong:

What It Would Have Taken: “Think of it this way: what would a really effective set of policies be right now? First… aggressively reverse the fiscal austerity of the last few years…. Monetary policy should accommodate that boost; interest rates should not go up even if inflation goes somewhat above 2 percent. In fact, there’s an overwhelming prudential case for raising the inflation target…. Say for the sake of argument that the right policy is two years of fiscal expansion amounting to 3 percent of GDP each year, plus a permanent rise in the inflation target to 4 percent. These wouldn’t be radical moves in terms of Econ 101 — they are in fact pretty much what textbook models would suggest make sense given what we have learned about macroeconomic vulnerabilities. But they are completely outside the bounds of respectable discussion. That’s the sense in which we are “doomed” to long-term stagnation. We have met the enemy, and it’s not the economic fundamentals, it’s us.

And here’s DeLong’s reply (or more precisely the response of Brad’s Greek friends):

Thrasymakhos: Oh, Krugman’s 100% right about today…

Khremistokles: He is indeed. We are totally tracked…

Thrasymakhos: Very few members of congress or FOMC participants seem to spend any significant time talking to anybody who is not a plutocrat…

Khremistokles: But he is wrong about how aggressive and radical the needed policies back in 2007 were. As a share of GDP, the bad shopping-mall and office-tour debts of Houston, etc, in 1989 were as large as the bad mortgages of 2007…

Thrasymakhos: But back in 1989 the political power of the princes of finance was much less than in 2007…

That’s probably right, but I have trouble with DeLong’s implicit assumption is that the financial crisis caused the Great Recession.  DeLong points out that the recession of 1990-91 was far milder, despite equivalent bad debt (as a share of GDP.)  And that the 1990 crisis was handled better. Krugman’s comment points to one overlooked factor.  In 1990 we did have a de facto 4% inflation target.  The years leading up to 1990 saw Australian-level NGDP growth, if not more.  So even if lending standards tightened sharply in the wake of the 1989-90 crisis, there still would have been no possibility of hitting the zero bound.  Rates fell to about 3% in the recession, still a bit higher than in Australia this time around.  With no zero bound in prospect, there’d be no reason for markets to expect an NGDP collapse.  Elsewhere I’ve argued that growing realization of plunging NGDP tanked the asset markets in the second half of 2008.

Even if we had managed the 2007-08 subprime crisis very well from a regulatory/resolution perspective, there is no question that banks would have tightened lending standards sharply.  That effectively reduces the demand for credit. And of course house prices were plunging even before Lehman, and then we got a “secondary deflation” of house prices when NGDP plunged.  It’s quite plausible that the Wicksellian equilibrium natural rate would have fallen to zero in late 2008, even with a better resolution of the banks.  On the other hand if we’d gone into 2007 with Paul Volcker’s de facto 4% inflation target (a policy he now opposes), then the Great Recession would have been a 1990-style mild recession.

One area where I slightly disagree with Krugman is his focus on inflation.  A 5% NGDPLT target path would have been enough; we didn’t need 4% trend inflation.  Nor do we need fiscal stimulus. On the other hand the supply side fundamentals of the economy were so poor after 2008 (for reason I don’t fully understand) that 5% NGDP growth would have led to some unpleasant stagflation. So we might have gotten Krugman’s 4% inflation anyway.  Indeed if my preferred policy had been adopted, it would have been widely judged a failure, partly because (as DeLong correctly pointed out) almost nobody back in 2007 envisioned a recession as severe as the one we got.

People see bad outcomes, and have trouble envisioning it could have been much worse.  That’s one reason why my preferred policy was not politically feasible in 2008.  But thanks to the NGDP targeting boomlet, it will be somewhat more feasible next time around.  Next time people will be able to envision a worse alternative.

All stabilization policies eventually fail, just as all presidents are judged failures in their 6th year in office.  The trick is to have a modest failure like Clinton or Obama, not a serious failure like FDR or Nixon.  NGDPLT would have given us just that in 2008-09.

PS.  I have a new post on Jeremy Stein over at Econlog.

HT:  TravisV