Seeing Lu Mountain
Brad DeLong has a post discussing a debate between Paul Krugman and Roger Farmer:
I find myself genuinely split here. When I look at the size of the housing bubble that triggered the Lesser Depression from which we are still suffering, it looks at least an order of magnitude too small to be a key cause. Spending on housing construction rose by 1%-age point of GDP for about three years–that is $500 billion. In 2008-9 real GDP fell relative to trend by 8%–that is $1.2 trillion–and has stayed down by what will by the end of this year be seven years–that is $8.5 trillion. And that is in the U.S. alone. There was a mispricing in financial markets. It lead to the excess expenditure of $500 billion of real physical assets–houses–that were not worth their societal resource cost. And each $1 of investment spending misallocated during the bubble has–so far–caused the creation of $17 of lost Okun gap.
(You can say that bad loans were far in excess of $500 billion. But most of the bad loans were not bad ex ante but only became bad ex post when the financial crisis, the crash, and the Lesser Depression came. You can say that low interest rates and easy credit led a great many who owned already-existing houses to take out loans that were ex ante bad. But that is offset by the fact that the excess houses built had value, just not $500 billion of value. I think those two factors more or less wash each other out. You can say that it was not the financial crisis but the destruction of $8 trillion of wealth revealed to be fictitious as house prices normalized that caused the Lesser Depression. But the creation of that $8 trillion of fictitious wealth had not caused a previous boom of like magnitude.)
To put it bluntly: Paul is wrong because the magnitude of the financial accelerator in this episode cries out for a model of multiple–or a continuous set of–equilibria. And so Roger seems to me to be more-or-less on the right track.
DeLong is certainly right that the housing bust is far too small, but it’s even worse than that. The vast majority of the housing bust occurred between January 2006 and April 2008, and RGDP actually rose during that period, while the unemployment rate stayed close to 5%. So it obviously wasn’t the housing bust. On the other hand you don’t need exotic theories like multiple equilibria—the Great Recession was caused by tight money. It’s that simple.
Or is it? Most economists think that explanation is crazy. They say interest rates were low and the Fed did QE. They dismiss the Bernanke/Sumner claim that interest rates and the money supply don’t show the stance of monetary policy. Almost no one believes the Bernanke/Sumner claim that NGDP growth and/or inflation are the right way to evaluate the stance of policy. Heck, even Bernanke no longer believes it.
And even if I convinced them that money was tight they’d ask what caused the tight money, or make philosophically unsupportable distinctions between “errors of omission” and “errors of commission.”
In previous blog posts I’ve pointed out that it’s always been this way. If in 1932 you had said that tight money caused the Great Depression, most people would have thought you were crazy. Today that’s the conventional wisdom. If in the 1970s you’d claimed that easy money caused the Great Inflation, almost everyone except a few monetarists would have said you were crazy. Now that’s the conventional wisdom. Even the Fed now thinks that it caused the Great Depression and the Great Inflation. (Bernanke said, “We did it.”)
The problem is that central banks tend to follow the conventional wisdom of economists. So when central banks screw up, the conventional wisdom of economists will never blame the central bank (at the time); that would be like blaming themselves. They’ll invent some ad hoc theory about mysterious “shocks.”
The other night at dinner my wife told me that the Chinese sometimes say, “If you cannot see the true shape of Lu Mountain, it’s because you are standing on Lu Mountain.”
In modern conventional macro, most people look at monetary policy from an interest rate perspective. That means they are part of the problem. They are looking for causes of the Great Recession, not understanding that they (or more precisely their mode of thinking) are the cause.
Only the small number of economists who observed Mount Lu from other peaks, such as Mount Monetarism or Mount NGDP Expectations, clearly saw the role of central bank policy. People like Robert Hetzel, David Beckworth, Tim Congdon, etc.
PS. Before anyone mentions the zero bound, consider two things:
1. The US was not at the zero bound between December 2007 and December 2008 when the bulk of the NGDP collapse occurred, using monthly NGDP estimates.
2. Do you personally support having the Fed use a policy instrument that freezes up exactly when you need it most desperately? Or might the problem be that they’ve chosen the wrong instrument?
PPS. Yes, not everyone on Mount Monetarism saw the problem, but as far as I know no one on Mount Interest Rate got it right. Perhaps Mount Interest Rate is Lu Mountain.
PPPS. To his credit, Brad DeLong thought the Fed should have promised to return NGDP to the old trend line. If they’d made that promise there would have been no Great Recession, just a little recession and some stagflation.
Tags:
23. May 2015 at 08:56
Scott,
I’m sure you’ve answered this before, but please humour me. This is with regard to
‘Do you personally support … wrong instrument?’
In the standard NK model at the ZLB, e.g. Woodford/Eggertsson or Werning,
there is an expected path of for interest rates that is *exactly* equivalent to any expected path for the money supply. The only slight difference between the two is that when, and only when, the ZLB is binding, any amount of money supply greater than the ZLB/satiation level is consistent with the same path of interest rates.
Talking about long term rates and credit spreads versus buying up non-safe assets does not alter this at all.
Therefore talking about money supply instruments versus interest rate instruments is essentially a distinction without a difference in NK models. So I’m curious about why you argue that changing the instrument is so important?
I agree with you that money has been far too tight over the last five years. But then, so does Mike Woodford. Crusading against the modelling choices in NK models seems to me to be completely tangential to the main issue of tight money.
23. May 2015 at 08:57
“So it obviously wasn’t the housing bust.”
-Maybe the decline wasn’t large enough to matter when housing prices only dropped 17%, but was large enough to matter when they dropped 30%.
“On the other hand you don’t need exotic theories like multiple equilibria””the Great Recession was caused by tight money. It’s that simple.”
-I don’t think it’s that simple. Tight money doesn’t explain the really weak real recovery (inflation should have been much closer to 0% during the recovery were it just a historically normal one in supply-side matters).
“If in 1932 you had said that tight money caused the Great Depression, most people would have thought you were crazy. Today that’s the conventional wisdom.”
-LOL. Have you cracked open any U.S. history textbook, lately? It’s still all “overproduction” and “underconsumption” and “wealth and income inequality” and “stock market crash”. You can weep if you like.
“And even if I convinced them that money was tight they’d ask what caused the tight money, or make philosophically unsupportable distinctions between “errors of omission” and “errors of commission.””
-I think the distinctions make sense if you have multiple targets, and Fed policy is only good at getting some of them to move in the direction you like.
“If in the 1970s you’d claimed that easy money caused the Great Inflation, almost everyone except a few monetarists would have said you were crazy. Now that’s the conventional wisdom.”
-It is (though Johnson’s and Nixon’s “guns and butter” fiscal policy and, on occasion, the oil price shocks [alone], are sometimes blamed among the less informed). However, the more I look at the Great Stagflation, the more I’m convinced that a substantial part of it was due to supply-side issues (independent of high NGDP growth, which also happened), at least, in the U.S. There are no excuses for Canada’s and the U.K.’s crazy loose monetary policy. I’ve abandoned the most vulgar form of monetarism in regards to the Great Stagflation, thankfully, and think monetary policy at the time was appropriate, if slightly too loose, in the U.S.
“So when central banks screw up, the conventional wisdom of economists will never blame the central bank (at the time); that would be like blaming themselves.”
-That explains quite a bit.
“Do you personally support having the Fed use a policy instrument that freezes up exactly when you need it most desperately?”
-Great point (and zinger).
23. May 2015 at 09:07
Also, I didn’t realize just how weird the Federal Funds Rate movements were between 9/15 and ZLB.
http://research.stlouisfed.org/fred2/graph/?g=1cfP
23. May 2015 at 09:22
Things are getting interesting in Europe;
http://www.eubusiness.com/news-eu/ecb-eurozone-bank.11e2
——-quote—–
The ECB had played its part in laying the foundations for recovery by cutting interest rates to record lows and pumping unprecedented amounts of liquidity into the economy via a series of unconventional policy measures, [Mario]Draghi said.
“Monetary policy is working its way through the economy. Growth is picking up. And inflation expectations have recovered from their trough,” he said.
But he went on: “This is by no means the end of our challenges, and a cyclical recovery alone does not solve all of Europe’s problems.”
The emerging recovery “does not eliminate the debt overhang that affects parts of” the single currency area, the central bank chief argued.
“It does not eliminate the high level of structural unemployment that haunts too many countries. And it does not eliminate the need for perfecting the institutional set-up of our monetary union,” Draghi insisted.
“But what the cyclical recovery does achieve is to provide near perfect conditions for governments to engage more systematically in the structural reforms that will anchor the return to growth.”
Monetary policy could steer the economy back to its potential, but it was structural reform that could raise that potential.
——-endquote——
That’s from the EU equivalent of the Fed’s Jackson Hole Conference.
23. May 2015 at 09:24
Them’s fightin’ words, Mario;
http://www.thenewstribune.com/2015/05/23/3805553/what-will-get-europes-economy.html
———-quote——-
Several economists challenged Draghi’s verbal pressure on governments, saying non-elected central bankers had no business making political proposals. Paul De Grauwe from the London School of Economics “” a former member of the Belgian parliament “” said the central bank put its legitimacy at risk.
He said restrictions on layoffs existed “because people want them. … And in democracies, people will obtain protection from their governments.”
“And when a central bank then comes out and says we should do structural reforms, it really says we should break down this system of protections. And in doing so the central bank sets itself outside the democratic process.
“The danger is that people will reject such a central bank.”
———endquote——
23. May 2015 at 09:41
There is a deeper debate being had between Farmer and Krugman.
Farmer is using Keynes to argue a drop in animal spirits has lead to our new sad unemployment equilibrium. This is a back door DEATH BLOW to PK on Fiscal Stimulus.
23. May 2015 at 10:09
I dunno, isn’t it easy to make equivalent arguments even when reasoning in terms of interest rates by pointing to the fact that even at 0% rates can be above the wicksellian equilibrium (as recently argued by the NY Fed here: http://libertystreeteconomics.newyorkfed.org/2015/05/why-are-interest-rates-so-low.html ).
I’m not an economist but I find it easier to follow the reasoning down to the micro foundations when thinking in terms of interest rates.
It is easy to see that even at zero, if inflation is not sufficiently high, real rates can be above marginal yields on capital investment, that real returns on money and government bonds can be above the risk adjusted returns on private stores of value, that government money having a -2% real return when private stores of value have -5% is a 3% subsidy to the act of pulling investment out of the private economy and parking wealth as government backed paper promises that can only be fulfilled through taxes or through future monetary manipulations that funnels value away from private investors and towards fiat holders and government bond holders instead of being fulfilled by investment that grows the economic pie and that allows society to pay liabilities by working and creating new wealth instead of by sharing dwindling output.
A business that wants to get a loan doesn’t reason in terms of percentage of ngdp it is borrowing and how much ngdp it will have to pay later. Thinking in terms of interest rates allows us to follow the causal chain starting from aggregate circumstances and going down to the reasoning of individuals or vice versa.
Even when thinking in terms of interest rates, the ZLB makes it clear that when you hit zero, the only option to avoid distorting the real markets is to raise the inflation target, producing a price path not unlike that of an NGDP target.
What’s wrong in thinking in terms of interest rates this way?
23. May 2015 at 10:12
“Do you personally support having the Fed use a policy instrument that freezes up exactly when you need it most desperately?”
No of course not, if there’s a better instrument, please provide it and explain why it’s not helicopter drops. If you think it’s QE, explain how that overcomes the ZLB problem. Hint: a target is not an instrument.
23. May 2015 at 10:30
Britonomist, how does QE not overcome the ZLB “problem”?
23. May 2015 at 11:35
Joe, You said:
“In the standard NK model at the ZLB, e.g. Woodford/Eggertsson or Werning,
there is an expected path of for interest rates that is *exactly* equivalent to any expected path for the money supply. The only slight difference between the two is that when, and only when, the ZLB is binding, any amount of money supply greater than the ZLB/satiation level is consistent with the same path of interest rates.
Talking about long term rates and credit spreads versus buying up non-safe assets does not alter this at all.
Therefore talking about money supply instruments versus interest rate instruments is essentially a distinction without a difference in NK models. So I’m curious about why you argue that changing the instrument is so important?”
Good question and I could write a dissertation on this subject. First of all, the term ‘instrument’ is ambiguous. In the Keynesian model you could argue the fed funds rate is the instrument, or that the base is the instrument and is adjusted to target the fed funds rate, the short run target. Let’s put that aside for the moment.
Any path of a monetary policy instrument makes everything else endogenous. Thus if you target the interest rate, it implies a certain path for M1, M2, the exchange rate, TIPS spreads and the price of zinc. If you target M2 it implies a certain path for interest rates, exchange rates, M1, the price of zinc, TIPS spreads, etc. And the same is true for any other target. So by saying that a monetary supply target will be associated with a certain interest rate path, and that this interest rate path is equivalent, is true but not really helpful. The Fed could adjust monetary policy to try to keep the Walrasian equilibrium price of zinc close to the actual market price of zinc. That works in the model, but I think that Woodford would agree that it wouldn’t work well in practice. I believe that interest rate targeting can work OK at positive interest rates, but at the zero bound it’s worse than zinc price targeting.
If you think of conventional monetary policy as adjusting the base to keep the fed funds rate at the target value, then I propose instead adjusting the base to keep NGDP expectations at the target value. Preferably with a futures market, but even the Fed’s internal NGDP forecast would be superior to a fed funds target. There is no zero bound problem with NGDP forecasts. This has many advantages, but communication is one of the most important. Monetary policy relies on good communication of the future path of policy, and that’s hard to do with interest rates that have run up against the zero bound. That makes interest rates a poor instrument/short term target. How do you communicate with interest rates when sometimes higher rates mean easier money and sometimes they mean tighter money?
You said:
“I agree with you that money has been far too tight over the last five years. But then, so does Mike Woodford.”
When I started my “crusade” (as you put it) against Fed policy in late 2008, I did not see any NKs in support of my position. None. That tells me there is something wrong with the NK model. Why didn’t they see what I saw? Instead I saw lots of NKs saying the Fed was out of ammo and that we needed to rely on fiscal policy. We now know that this was incorrect, that the Fed was far from being out of ammo. Surely there must be something wrong with a model that would give such unreliable policy advice.
If NKs had agreed with me in 2008 and early 2009 I never would have gone into blogging.
23. May 2015 at 11:43
E. Harding.
1. I was talking about the fall in home construction, not prices.
2. The slow recovery is both weak AD and weak AS.
3. I was talking about conventional wisdom among monetary economists, not historians.
You said:
“I think the distinctions make sense if you have multiple targets, and Fed policy is only good at getting some of them to move in the direction you like.”
The Fed can have multiple policy goals, but only one AD target, so I don’t see your point.
You said:
“However, the more I look at the Great Stagflation, the more I’m convinced that a substantial part of it was due to supply-side issues (independent of high NGDP growth, which also happened), at least, in the U.S.”
Then you need to look again, from 1971 to 1981 NGDP growth was 11%, RGDP growth was over 3% and inflation around 8%. It was a 100% monetary inflation, with AS playing essentially no role.
So you consider 11% NGDP growth appropriate? Please explain.
23. May 2015 at 12:01
Thanks Patrick, Interesting debate.
Morgan, But I doubt Krugman sees it that way.
Benoit, I really need to do a blog post on that, as it’s so confusing. Saying rates are too high, relative to the Walrasian equilibrium, does not imply the central bank should lower them.
You said:
“Even when thinking in terms of interest rates, the ZLB makes it clear that when you hit zero, the only option to avoid distorting the real markets is to raise the inflation target, producing a price path not unlike that of an NGDP target.
What’s wrong in thinking in terms of interest rates this way?”
I’m not sure you need to raise the inflation target. With faster NGDP and RGDP growth, the Wicksellian equilibrium rate will rise.
The other problem is that this approach makes it seem like a central bank promise to hold rates at zero for 30 years is expansionary, but it probably is not.
Britonomist, The zero bound problem reflects both the policy goal, and the extent to which the goal is met. If you always meet the goal, you are at the zero bound for a much shorter period. BTW, it’s not really a “problem” unless you are targeting interest rates. If a country targets the price of gold, or a foreign exchange rate there is no zero bound “problem” even if you are at the zero bound. There is a zero bound “situation.”
If you don’t like the zero bound situation, then set the NGDP target path high enough so that rates stay about zero, as in the pre-2007 period. It’s that simple. It has nothing to do with choosing between interest rates and QE, if you are targeting interest rates you are far more likely to have to rely on QE than if you relying on an alternative instrument like NGDP futures prices or forex rates.
23. May 2015 at 12:21
Tight money came from changing inflation expectations. Inflation was running closer to 2.5% for a few years leading up to fall 2008. Now, arguably the commodities boom driven by rapid real growth in China and elsewhere was the driver. The fed perhaps reasonably ignored this inflation as a supply side problem. But lip service was still paid to a two percent inflation target. This was patently dishonest by mid 2008, Leading to a unmooring of 2% inflation expectations. When oil spiked again from a supply shock, the fed decided to act to reset inflation expectations back to 2% and bam we entered a hard recession.
A recession where again lip service was paid to 2% but the fed seemed to have an effective target of 1.5 or even 1. The adjustment to this has been painful.
Low real yields are a bit of a myth, people look at TIPS spreads as a clean measure of real return but TIPS is indexed by CPI-U which is about the worst possible choice of inflation measure…
23. May 2015 at 12:30
@Scott
” If a country targets the price of gold, or a foreign exchange rate there is no zero bound “problem” even if you are at the zero bound.”
A CB can target the price of gold or foreign currencies by directly buying gold or foreign currencies.
What must the CB buy and sell in order to target NGDP? Base money? Most people say buying and selling reserves is used to target the interest rate, which in turn determines the broader money supply. If the CB buys and sells all money, including broad money, then since NGDP is kM, this would be an instrument to target NGDP. But since the CB controls the base, not all money, this is not necessarily the case. When you consider base vs broad, you have NGDP = k*F(M0), i.e. a function of base money which results in the amount of broad money in the economy. If the function of base to broad money is complex and non linear, then why is it reasonable to assume that buying and selling base money can be used to target NGDP?
You must either have good reasons to think that there is a consistent positive relationship between base and broad money, even under the ZLB, in which case can you show this? Otherwise you think the Fed can easily increase the broad money supply other ways, how?
23. May 2015 at 14:13
Sumner wrote:
“The vast majority of the housing bust occurred between January 2006 and April 2008, and RGDP actually rose during that period, while the unemployment rate stayed close to 5%. So it obviously wasn’t the housing bust.”
That is a flawed argument. It presumes as absolute that should A cause B, then A and B must be concurrent.
The economy is much more complex than that. Major events can often have effects that last years into the future.
It is your crude general equilibrium thinking that is to blame here.
“On the other hand you don’t need exotic theories like multiple equilibria””the Great Recession was caused by tight money. It’s that simple.”
No it isn’t, because there is still the question of why so many people all over the country all about the same time decided to hold on to their money earnings a little longer then they used to, that had the effect of spending falling the way it did.
When you say “tight money caused the recession”, what you aee really saying is that an absence of something (I.e. more inflation of the money supply) that would have prevented the effect of another something (people holding on to their money earnings a little longer than before) from taking place, but ignoring and refusing to consider WHY that first something was even necessary at all, or, equivalently, why the second something took place at all.
You never attempted to answer why the sudden widespread increase in so-called cash preference.
I have said this many times before: There is a good explanation for why you never addressed this. It is because if you do address it, you will invariably have to look at what external cause to the market would have caused so many people who never speak to each other and don’t even know each other, to suddenly hoard so much money all around the same time. That will lead right back to the Fed and what it must have done wrong PRIOR to 2006-2007 that had the effect later on in the sudden rise in cash preference.
And then poof goes your entire intellectual investment, all that time, money, revealed as the very same malinvestment that you advocate with “stable inflation”!
That is why you will not touch this question. The answer is too painful to think about. So, pretend the sudden increase in cash hoarding needs no answer. Pretend the answer is unimportant, because we need to pretend that what is important is the there is a strong socialist institution there to save us from our own stupidity, our own narrow interests in cash holding.
Markets are the most right? Please. The market is stupid according to MM. There cannot be a free market, because if there is, then we’ll have another Hitler as total spending becomes then outcome of all individual preferences, rather than a savior socialist’s preference.
23. May 2015 at 14:48
@Britonomist: “NGDP = k*F(M0)” OK, good enough for this discussion.
“If the function of base to broad money is complex and non linear, then why is it reasonable to assume that buying and selling base money can be used to target NGDP? You must either have good reasons to think that there is a consistent positive relationship between base and broad money, even under the ZLB”
A “complex, non-linear” relationship is still “consistently positive”. All you need is the positive relationship. You don’t need to know the details of the actual function.
As to why you can use base money to target NGDP, it’s simply because you implement a negative feedback control loop. Measure current NGDP. If it’s below your target, add more base money. If above your target, take base money away. It’s a pretty simple idea.
This only fails if some condition (ZLB?) causes base money to no longer have a positive relationship with NGDP. But this is highly implausible in theory, and never seen empirically. If you increase the amount of base money by an order of magnitude (together with a clear expectation that the injection is permanent), the idea that NGDP would not alter at all (or might decline) seems too silly to be worth spending much time worrying about. Perhaps you could explain your skepticism.
23. May 2015 at 14:53
@MF: The idea that some Fed action in 2005 (or before) “caused” a rise in cash preference in 2008, is not plausible. You assert hypothetical connections like this all the time, but provide neither theoretical nor empirical justification, just your assertion. To be blunt, nobody believes your causal claims.
But, as even you have realized, it doesn’t matter. The Fed has the power to offset whatever changes in cash preferences occur, to alter money supply in order to balance with changed money demand. So there’s no reason that the US economy needs to suffer from AD shocks. (And there would be no benefit to allowing it to do so.)
23. May 2015 at 15:08
“Perhaps you could explain your skepticism.”
Firstly, because just because banks have more base money doesn’t mean they’ll expand more credit at the ZLB. Now, Sumner might say that it’s not about bank lending but portfolio re-balancing, but portfolio re-balancing is an m1 story not a base money story, so does that mean Scott actually thinks the Fed should use m1 rather than m0 – doesn’t that contradict numerous earlier statements? The whole thing is incoherent, I just want a model. And when there is all this uncertainty about QE, why not just use permanent helicopter injections instead? That seems more direct and more obvious, yet it’s opposed.
23. May 2015 at 15:29
@ssumner
“I was talking about the fall in home construction, not prices.”
-Home prices supposedly have a “wealth effect” and were very important for keeping millions credit-worthy during the housing boom.
“The slow recovery is both weak AD and weak AS.”
-Agreed. How much higher do you think RGDP would have been had NGDP targeting been applied from 2005?
“I was talking about conventional wisdom among monetary economists, not historians.”
-That’s a relief.
“The Fed can have multiple policy goals, but only one AD target, so I don’t see your point.”
-I don’t see your point here, either. Most likely because it’s probably too complex to be fully revealed in a single sentence. What’s the substantive difference between a “goal” and a “target”?
“So you consider 11% NGDP growth appropriate? Please explain.”
-Do you see what I see?:
http://research.stlouisfed.org/fred2/graph/?g=1cgN
Throughout the 1973-5 recession, the U.S. was basically on NGDP trend, 1961-1975. In 1975, it loosened, shifting to a much faster trend from 1975 to 1981. However, during the 1982-3 recovery, while inflation was falling, the U.S. was went back to the 1975-1980 trend. and the 1975-1980 trend was well in the midst of stagflation.
23. May 2015 at 16:15
“Bernanke/Sumner claim”? Good God, your nut-huggery is just pathetic.
23. May 2015 at 16:19
Excellent blogging.
But I am not sure there is a consensus among economists that constrains the Fed. In Japan the Bank of Japan is pursuing $80 billion a month of QE in an economy half the size of the US economy. They are doing so despite unemployment in the 3 percent range and inflation now above 2 percent and real growth of of 2 percent. Yet the Bank of Japan governor recently promised to do more, if necessary.
In the US PCE core is below 2 percent, unemployment is above a 5% and real growth is dead in the water. Yet Janet Yellen keeps talking about raising rates.
I think the problem lies inside the Fed, not inside the profession. BTW, many economists in the private sector, such as those working in real estate, are not for tight money.
I will concede that among a branch of economic punditry, it has become PC to bray about tight money.
23. May 2015 at 16:50
Geddis:
“The idea that some Fed action in 2005 (or before) “caused” a rise in cash preference in 2008, is not plausible.”
It is not only “plausible”, but the best theory we have about why it took place.
You are merely denying it through some unproved and unestablished hypothetical. You do this all the time, providing not a shred of theoretical justification or empirical evidence.
Regardless if the Fed “has the power” to print more money and temporarily prevent the effects of such widespread sudden increases in cash preference, that it itself caused, has no bearing on the truth of what caused the rise in cash preference!
Also, contrary to your claim, there is in fact a very good reason why you personally should “suffer” from a reduced income, and, in exactly the same way and for exactly the same reason, why you and someone else should both “suffer” reduced incomes, and, once again in the exactly the same way and for exactly the same reasons, why you and that someone else and any number of other people, should “suffer” reduced incomes.
There is no good reason whatsoever why it must be the case that if any individual or group of individuals experience a reduced income, that violence must be introduced against innocent people, so that any reduction in one or more people’s incomes is at least exactly matched by, or more than offset with an arbitrary number pulled put of your derriere by, an inflation induced rise in incomes of other not so random, special interest groups of people.
Just because a person or group of people have come to acquire some range of undefended and unprosecuted ability to wage violence against the innocent, it does not mean that the various “targets” that they could theoretically constrain themselves to suddenly take on this social good, or this higher ideal that somehow solves some hitherto unsolved dilemma.
If you or if a million other people produce or offer something that earns you fewer dollars this year than last year, there is absolutely no truth or reason why that should be either reversed by violence, or accompanied by others earning equivalently more money also through the use of violence.
Not only does it cause discoordination between people’s productive activity, but it is also based on violence against peaceful activity. Your ethics and economics are at a supremely lower level compared to what I advocate. You know it, I know it. You’re just upset that you cannot refute anything I say intellectually.
23. May 2015 at 16:58
Gravatar of Beefcake the Mighty Beefcake the Mighty
23. May 2015 at 16:15
“Bernanke/Sumner claim”? Good God, your nut-huggery is just pathetic. Indeed!
23. May 2015 at 17:23
Jon, You’ve done a nice job explaining why the Fed should not target inflation. NGDP growth gave the “right” signal.
Britonomist, The entire discussion of broad money is beside the point. Market monetarists do not assume any sort of stable relationship between the MB and M2, or NGDP.
You said:
“What must the CB buy and sell in order to target NGDP?”
Ideally you target the price of NGDP futures contract. Second best option is to target the central banks internal forecast of NGDP (Lars Svensson’s preferred choice.)
E. Harding, OK, but you need to clearly explain the hypothesis. It’s very different claiming that less housing construction directly reduced RGDP, and in contrast that a loss of wealth led to less AD, which could in principle be offset under NGDP targeting.
As I said I would have expected a small recession and some stagflation under NGDPLT. I can’t really say more than that. Maybe flat RGDP in 2009?
On targets & goals, recall that they use AD to impact both inflation and unemployment. So let’s say 2% inflation and 5% unemployment are the goals. They might aim for a target of 4% NGDP growth as most likely to hit the twin goals. But that’s a single target, as it affects both goal variables.
I don’t follow your last point. If the Fed wants two percent inflation they needed 5% NGDP growth during the 1970s. They had 11%. Six percent too much. It’s that simple.
BEEFCAKE THE MIGHTY, I can tell by your name that you are quite modest, but you seem to have missed the intended humor in my Sumner/Bernanke reference. I even said Bernanke no longer believed it, hoping that would be a clue. I guess not.
Ben, Polls show that the Fed did a bit more than the average economist wanted them to do, I have multiple blog posts on that point. But your Japanese example is a good counter; the BOJ seems to be doing more that the average Japanese economist favors, perhaps much more.
23. May 2015 at 17:30
Scott, apologies if I missed your joke, but you’ve kissed his ass on other occasions so I assumed the worst here.
23. May 2015 at 18:06
Scott, but I still don’t think futures contracts gets to the heart of the issue. Don’t you think that, even if the Fed couldn’t create an NGDP futures market, it could still set the path of NGDP easily? If there is no futures market, what does the CB buy and sell to target its internal forecast?
Honestly, if you just said “the CB can just inject new money into the economy by creating money and giving it to the government to fund a reduction in VAT or further spending, and keep doing this until NGDP target is hit” – all problems would be solved, but instead you insist on bizarre policies that avoid government altogether and seem to only work through complex indirect effects like portfolio re-balancing.
23. May 2015 at 18:27
“1. The US was not at the zero bound between December 2007 and December 2008 when the bulk of the NGDP collapse occurred, using monthly NGDP estimates.
2. Do you personally support having the Fed use a policy instrument that freezes up exactly when you need it most desperately? Or might the problem be that they’ve chosen the wrong instrument?”
1. If the fed used tools (hel-e drops) which generate less credit and financialization the gyrations in demand should be much more manageable.
2. Ngdp targeting using OMP’s will grow NGDP but the financial secotr growth component will be larger than under other tools which depend less on portfolio flows and credit to grow ngdp.
“CMA, This is a completely separate argument. The effect on interest rates might be different, but I doubt the difference would be significant–perhaps just a basis point. In any case there is never any justification for helicopter drops, which are an exceedingly costly and wasteful policy.”
At least you agree there would be a difference. If the rate is different so must the price be. I would say the difference is significant though because M expansions don’t expand loanable funds and because the demand for bonds doesnt increase in line with M expansions. Imagine the fed got rid of its bonds entirely? Also central bank emoney provision would need banks to pay interest on deposits to attract deposits.
Hel-e’s are less wasteful than OMP’s. OMP’s are a gift to asset holders in form of higher wealth and to the financial sector in the form of increased portfolio and credit activity. The gov does not need to tax back hel-e’s at a later date if done directly by central bank.
23. May 2015 at 18:27
Scott,
Sorry if the ‘crusade’ thing came on a bit strong! And thanks for the reply.
1) You are of course right on the endogeneity of every other asset price given the path of the monetary instrument, whatever that is.
2) For this reason, it doesn’t really make sense to say that interest rates ‘freeze up’ at the ZLB but that the monetary base instrument does not. All instruments freeze up to an identical extent, or not, at the ZLB, at least in the formal ratex equilibrium. It is quite misleading to claim otherwise. Communication and conceptual framing is of course a different matter.
3) Hopefully we agree that regardless of whether or not you think that NGDP forecast targeting is a good idea (I do! Have been very much persuaded by you on this!) it is an analytically distinct issue. That is, NGDP forecast is the (possibly intermediate) target; interest rate, base money supply or zinc prince is the instrument; the two are somewhat separate issues, though related in important ways.
4) I therefore read you as saying that interest rates (and their expected future path) versus base money (and its expected future path) is largely about semantics, ease of communication and conceptual framing of a model. The very last of these pertaining to how NKs think about monetary policy.
5) NKs failing to advocate loose money is, of course, a problem. But explaining to NKs that their model indicates that expected money supply growth between now and normalisation should be faster, is not analytically or logically different from explaining to them that the expected path of interest rates between now and normalisation should be lower. Any intelligent NK economist should accept either both or neither of these propositions. But not one and not the other. This would be a logical howler. It would mean that NK economists don’t just disagree with MMs, but *are too stupid to draw out the implications of their own models*. I’m not sure it makes sense to believe that lots of smart and thoughtful economists are committing basic fallacies – it would be a straw man. If NKs don’t believe in looser interest rates, I would be astounded if they believed in a faster monetary base growth rate.
So I think our disagreement boils down to whether or not the monetary base is a meaningfully better communicator of the short term stance of monetary policy. Is this fair? I can think of lots of reasons for why interest rate communication is not inferior.
1) Essentially similar to above. If you think that financial market participants will respond meaningfully differently to money supply communication than to interest rate communication you are assuming that they are collectively making a giant irrational logical howler. It means that financial market participants do not have anything close to rational expectations. This requires jettisoning EMH wholesale, which I expect you aren’t comfortable with doing.
2) Credit spreads matter. Regardless of whether you think that the Fed should have attempted to offset the impact of financial market shocks on demand (again, very much in agreement with you) they clearly impact the monetary policy transmission mechanism. Policymakers need to think about how to offset shocks to firms’ external financing premia with monetary policy if they want to meet their intermediate target (which should be the NGDP forecast). It’s very very easy to think about how to do this with interest rates. It’s almost impossible to think about how to do this with money supply.
3) Things are very different at the ZLB, regardless of your instrument. Even if managing the monetary base is preferred, monetary policy now needs to either a) appeal to some kind of financial market imperfection to shift the prices of assets that are not short term debt; or b) manage the expected future monetary base. Whereas during normal times it is sufficient to manage the current monetary base and only buy short term safe debt. Clearly the role of expectations becomes much more important *whatever the instrument*. A money supply instrument occludes this important fact.
None of this is to disagree with you that money has been far too tight since 2007. But I don’t think the NK model is remotely responsible for this.
23. May 2015 at 22:37
Perhaps this is the graph that will get the right to understand why they need market monetarism. Look what happened the last time the Fed got monetary policy badly wrong. Heck, look at 2008!
24. May 2015 at 01:55
Scott: well, polls, but if a substantial fraction of the respondents work for the Fed…and how were the polls worded…
24. May 2015 at 05:35
Beefcake, You said:
“Scott, apologies if I missed your joke, but you’ve kissed his ass on other occasions so I assumed the worst here.”
No apology needed, after all it’s widely known that TheMoneyIllusion is famous for praising the Bernanke Fed’s monetary policy over the 5 years from 2009-14. I just kept saying over and over again what a swell job they were doing.
(That’s also a joke.)
Seriously, I think Bernanke’s a fine economist, just as I think lots of other people are fine economists. My Sumner/Bernanke comment was for shock value, as most people assume I am saying something really heterodox.
Britonomist, You said:
“If there is no futures market, what does the CB buy and sell to target its internal forecast?”
Buy debt. It’s that simple. In your next paragraph you suggest a grotesquely inefficient policy of having the Fed finance government spending that would not otherwise occur. Why do that when you can simply buy debt?
I think lots of people get all mixed up here because they don’t understand what’s going on at the zero bound. They think the zero bound is easy money that has failed, whereas it’s actually tight money. If you want to avoid the zero bound you need easier money, i.e. a higher NGDP growth target. It’s a choice between a big balance sheet for the central bank or a higher NGDP growth target—pick your poison. What bothers me is when some commenters demand a low NGDP target, and also complain about a big central bank balance sheet. It’s like a fat guy complaining about his weight, but insisting he doesn’t want to start exercising or to stop eating a pint of Ben and Jerry’s each day. He wants surgery. Fiscal assistance is like surgery
CMA, You keep making completely unsupported arguments. You don’t even seem to understand why heli-drops are extremely wasteful policy. I’d take you more seriously if you seemed to understand their massive waste, but argued it was worth it to prevent “financialization.” BTW, I do not believe OMOs lead to more financialization.
Joe, You said:
“For this reason, it doesn’t really make sense to say that interest rates ‘freeze up’ at the ZLB but that the monetary base instrument does not. All instruments freeze up to an identical extent, or not, at the ZLB, at least in the formal ratex equilibrium. It is quite misleading to claim otherwise. Communication and conceptual framing is of course a different matter.”
I don’t agree, and other economist like Lars Svensson also do not agree. I find this to be a very odd claim. In what sense does the price of gold “freeze up” when rates hit zero?” Yes, at some point further increases in the price of gold may not lead to further reductions in the interest rate, but that just means one of the transmission channels has frozen up, all the others still operate. So I don’t understand your point. What do you think would happen if the Fed pegged gold prices at $100,000/oz at the zero bound? Obviously we’d get immediate hyperinflation.
On the other hand, promising zero interest rates forever might lead to expectations of a Japanese-style malaise.
The only way your claim would make sense is if nominal interest rates were the only transmission mechanism for monetary policy. But Mishkin’s textbook lists something like 10 mechanisms, and he missed a whole bunch.
You said:
“Hopefully we agree that regardless of whether or not you think that NGDP forecast targeting is a good idea (I do! Have been very much persuaded by you on this!) it is an analytically distinct issue. That is, NGDP forecast is the (possibly intermediate) target; interest rate, base money supply or zinc prince is the instrument; the two are somewhat separate issues, though related in important ways.”
I strongly disagree. In one regime the MB is adjusted to keep gold prices stable. In another regime (i.e. NK) the MB is adjusted to keep the fed funds target at say 3%. In another regime (i.e. MM) the MB is adjusted to keep NGDP futures prices at a 5% premium. As I said, economists use the term ‘instrument’ in inconsistent and confusing ways. But NGDP futures play exactly the same role as fed funds rate, whatever term you use.
You said:
“I therefore read you as saying that interest rates (and their expected future path) versus base money (and its expected future path) is largely about semantics, ease of communication and conceptual framing of a model. The very last of these pertaining to how NKs think about monetary policy.”
This might be correct if you are talking about the entire future path of interest rates. But of course in that case there is no zero bound problem. And see my “malaise” comment above.
You said:
“But explaining to NKs that their model indicates that expected money supply growth between now and normalisation should be faster, is not analytically or logically different from explaining to them that the expected path of interest rates between now and normalisation should be lower.”
This is not true, and it exactly illustrates the problem I have with NK. The NK model does not say lower rates mean easier money, it says lower rates relative to the Wicksellian equilibrium rate mean easier money. But a highly expansionary monetary policy will produce an inflationary boom, and thus a higher Wicksellian equilibrium rate. I think lots of NKs were looking at the actual falling market rate in 2008, and assuming money was getting easier, whereas the NGDP data show it was getting tighter. If they had looked at NGDP expectations, they would have seen that money was tight.
You said:
“So I think our disagreement boils down to whether or not the monetary base is a meaningfully better communicator of the short term stance of monetary policy. Is this fair?”
No, For reasons I provided above. The role of the MB is exactly the same in both regimes, the fed funds rate is replaced by NGDP futures prices, which are much more informative.
I won’t respond to your final points, as they were based on a misreading about the way MMs think about the monetary base–I agree that it’s information value is often small, but of course from the market reactions to QE announcements we know it’s not zero.
24. May 2015 at 06:00
@Scott
“Buy debt. It’s that simple.”
Who’s debt? The governments? If you frame it as buying government debt – and emphasize that such holdings of debt are permanent (or allow government to default on the debt held to CB), then that’s essentially monetizing debt. I can definitely see how that would be effective, but it’s also essentially helicopter drops.
24. May 2015 at 06:31
[…] Scott Sumner. He distinguishes himself by being consistent. I remember a post he wrote early in 2009 saying he […]
24. May 2015 at 07:40
Scott,
Thanks for the reply.
I may have phrased my point slightly unclearly. My central point is the following. For a *given* path of expected interest rates between now and infinity the following are fixed in the standard NK model:
The exact prices of all assets when interest rates are not at the ZLB
The exact value of the money supply when interest rates are not at the ZLB
The exact price of all assets when interest rates are at the ZLB
A lower bound on the money supply when interest rates are at the ZLB
Svensson’s ‘foolproof’ argument is *completely* consistent with this, and he does agree with me. His point is that an exchange rate peg serves as a credible commitment to monetary easing in the future. The peg *can only work* in Svensson’s model if it binds after the economy leaves the ZLB. Otherwise it is irrelevant, and he emphasises this point. That is, it overcomes Krugman’s original time inconsistency problem. In rational expectations equilibrium, it is *exactly equivalent* to credibly lowering the expected path of future interest rates relative to their previous path. Svensson does not invalidate the central NK idea he just argues that FX pegging is a good way of solving the commitment problem associated.
A gold peg at the ZLB is just another way of credibly committing but is also, inter alia, consistent with committing to a new (lower) expected path of interest rates (relative to the expected path of Wicksellian rates).
Put another way, in the standard NK model there is:
A commitment to a gold peg that binds after ZLB exit
A commitment to a FX peg that binds after ZLB exit
A commitment to lowering the expected path of future interest rates
A commitment to raising the expected growth of the money supply
A commitment to raising the future price level, inflation rate, NGDP level or growth rate
Which will be equivalent. These are just different ways of solving the same time inconsistency problem in the NK setup.
My claim isn’t at all about transmission mechanisms of monetary policy. In rational expectations equilibrium these *all lead to the same outcome*, because of arbitrage across all asset prices. It is true regardless of whether the exchange rate, interest rate, balance sheet, external finance premium or other channels (I forget Mishkin’s full taxonomy) dominate.
The Fed cannot leave the path of interest rates completely fixed, and also vary exchange rates separately. The two are not independently determined. With regard to promising zero interest rates forever, there are two possibilities, and only two possibilities in rational expectations. The first is that this is expansionary, if prior to the policy announcement the expected path of interest rates was not zero. The second is that a Japan style inflationary slump was expected anyway in which case nothing happens.
Do you agree with the above analysis? I think to disagree with it you have to jettison rational expectations in some way.
I agree that the instruments/intermediate target/ultimate target terminology is confusing and probably best left to one side.
24. May 2015 at 11:10
Britonomist, You said:
“Who’s debt? The governments? If you frame it as buying government debt – and emphasize that such holdings of debt are permanent (or allow government to default on the debt held to CB), then that’s essentially monetizing debt. I can definitely see how that would be effective, but it’s also essentially helicopter drops.”
Now you are way off the reservation. Words need to have clear meaning, otherwise discussion is impossible. If you use OMOs to stabilize the price level or NGDP, it is not “essentially helicopter drops.” That’s just flat out wrong.
Whether a given monetary injection is permanent depends on the demand for money over time. But hey, if you want to call an OMP a “helicopter drop” then you are free to do so. In that case we agree, and I get the policy I want.
And yes, buy government debt.
Joe, There are lots of points of confusion here, I’ll try to disentangle some of them (although I’m not sure I fully understand them myself).
Let’s start with an economy of money and goods, no financial sector and no interest rates. Would you agree that monetary policy affects goods prices without affecting interest rates? If so, then there are transmission mechanisms other than interest rates.
Now consider using an exchange rate instrument. Yes, a monetary policy that depreciates a currency might move interest rates to the appropriate level, according to NK models. But what if it didn’t? There are other channels by which monetary policy affects the economy. For instance, currency depreciation can create inflation via international goods arbitrage. If you don’t see that in NK models, maybe that’s because they assume sticky prices. So policy is not just working via interest rates, and it does not require any future commitments. If today the Japanese suddenly peg the yen at 1 million to the dollar, they’ll immediately get hyperinflation. You might argue they’d also immediately exit the zero bound. Fine, that just shows how easy it is to exit the zero bound. There are no “expectations traps,” and never have been any.
Krugman’s argument was that central banks would not be able to depreciate their currencies if they wanted to, but Japan and Switzerland showed that was wrong. (He agreed that a sufficiently large depreciation would be inflationary.)
You said:
“The Fed cannot leave the path of interest rates completely fixed, and also vary exchange rates separately. The two are not independently determined. With regard to promising zero interest rates forever, there are two possibilities, and only two possibilities in rational expectations. The first is that this is expansionary, if prior to the policy announcement the expected path of interest rates was not zero. The second is that a Japan style inflationary slump was expected anyway in which case nothing happens.”
There is a third possibility. No Japanese style slump was previously expected, but the promise of zero rates forever leads to that expectation. The Japanese scenario is the only one that doesn’t blow up, if you hold rates at zero forever. So if investors rationally expect the central bank to not blow up the economy, then the announcement of a zero rate policy will lead them to expect the central bank to gradually appreciate the currency against other currencies, and that leads to zero interest rates via the interest parity condition. And it leads to deflation via PPP.
I’m told that NK models assume monetary policy does not affect the Wicksellian natural rate. Is that true? Because if it is, that’s where I disagree with NK models. In the real world monetary policy obviously does affect the natural interest rate, because it affects the business cycle.
The bottom line is that while any policy I propose is consistent with a certain path of interest rates over time, there is almost no chance that the central bank could discover that path, without targeting something else, like NGDP futures prices. If they try to target rates directly they may do OK at positive rates (via the Taylor Principle), but they’ll fail miserably at the zero bound, because the future interest rate commitments they would need to make are far to complicated to model. The market can’t tell whether a promise of zero rates forever is easy money or promising to be like Japan. At that point you need to switch to something else, such as exchange rates or NGDP futures targeting. Or even QE.
Here’s a few questions:
1. Am I right that the NK model assumes monetary policy does not affect the natural rate of interest?
2. In Svensson’s model if the currency depreciation fails to affect future asset prices in the appropriate way, isn’t the policy impossible to adopt? Thus in my thought experiment where the BOJ depreciates the yen to 1 million to the dollar, if the yen were expected to fall back to 120 to the dollar in 3 years, then the current rate would not stick. If they tried to make it stick the BOJ would end up buying up the entire world’s stock of assets. But of course that exposes the central bank to massive price risk, and creates inflation via fiscal theory of the price level channels, doesn’t it? So then I suppose the future expectation for the exchange rate (at 120) does not stick. I always get the impression that when push comes to shove the NK models will grudgingly allow my thought experiments to work, but insist I’m somehow “cheating”–doing something so drastic I overcome the expectations trap.
3. I’ve always thought that if you look close enough, the NK zero bound problem is actually queasiness about large central bank balance sheets, disguised as a zero bound problem. Is that right?
24. May 2015 at 18:45
When I read this, I would say: why buy only governmetn debt? Or government bonds + mortgages ? QE should buy a certain fraction of all assets available in the economy, shouldn’t it?
25. May 2015 at 04:32
@Scott
Goal – Full Employment (or some permutation thereof)
Target – NGDP
Tool – Asset purchases/sales by the CB
Mechanism – Increase/decrease in the exchange of financial assets for goods and services.
Don’t know why you keep making this complicated.
25. May 2015 at 05:24
Jose, I don’t see any good arguments for having the Fed buy assets such as stocks.
dtoh,
Goal–Stable NGDP growth
Target–NGDP expectations
Tool–Adjustments in the monetary base
Mechanism–HPE
Is that complicated?
25. May 2015 at 05:59
Actually Scott, Jose’s question is very good. Why shouldn’t the Fed monetize everything in sight?
25. May 2015 at 06:23
@Scott
Ok, but then, there is no good reason to buy mortgages as well …
25. May 2015 at 06:52
@ssumner:
> Whether a given monetary injection is permanent depends on the demand for money over time.
I think this is a very worthwhile point that deserves a more detailed argument, because it depends on both “demand for money over time” and the Central Bank’s policy rule.
One likely problem with QE is/was that its effect was muted by the Fed making it clear that it was a temporary measure. Printing $10bn and giving it to the United Way, no-strings-attached, probably would have had a greater positive impact on the price level than $100bn of temporary QE.
This effect is obvious via monetary policy transmission, or how green pieces of paper turn into spending. Someone given money will spend it, but someone lent money will only spend it if they think the return is worthwhile. However, temporary QE acts as the Fed promising to take away the free lunch as soon as someone tries to eat it.
This is where level targeting (either NGDP or price level) is superior to inflation targeting: it commits the central bank to leaving the free lunch on the table for a while.
25. May 2015 at 07:45
“Now you are way off the reservation. Words need to have clear meaning, otherwise discussion is impossible. If you use OMOs to stabilize the price level or NGDP, it is not “essentially helicopter drops.” That’s just flat out wrong.”
I’m talking about OMO’s with the explicit aim of monetizing the debt. The ‘helicopter drop’ in this case comes from the government being able to increase or maintain their deficit more than otherwise, so essentially new money being directly spent into the economy.
However, if you’re talking about just simple QE. Then I again need a model to show that buying T-Bills at the ZLB to target NGDP is equivalent to buying gold in order to target the price of gold.
25. May 2015 at 07:58
http://econpapers.repec.org/paper/imeimedps/08-e-04.htm
“This paper examines the effects of a money-financed fiscal expansion — a helicopter drop — when an economy is in a liquidity trap. It uses a textbook-style model calibrated to fit Japan’s economic slump and deflation as of 2003. According to the results, money-financed transfers totaling 9.4% of GDP end the output slump and guide the economy to a steady state with 2% inflation. By raising output and inflation, the policy also reduces the ratio of government debt to GDP. The policy’s long-run effects are the same as those of a bond-financed fiscal expansion, but money finance prevents a short-run rise in debt. “
25. May 2015 at 08:24
“I’m talking about OMO’s with the explicit aim of monetizing the debt. The ‘helicopter drop’ in this case comes from the government being able to increase or maintain their deficit more than otherwise, so essentially new money being directly spent into the economy.”
Whether it is the Federal Reserve’s “explicit aim” to monetize the debt should be irrelevant. What is relevant is the effect of its actions.
The Federal Reserve has been very profitable, even before QE was initiated; however, QE has made it even more profitable than ever:
https://ycharts.com/indicators/us_federal_reserve_banks_corporate_profits_yearly
Most of these profits are remitted to Treasury, reducing the federal deficit. Since 2009, the Federal Reserve has remitted about $400 billion. Would Congress have spent less had these profits not been remitted, or would it simply have incurred higher deficits? I see some degree of “helicopter drop” in this, but the magnitude is difficult to measure because the counterfactual over the short and medium term isn’t clear.
25. May 2015 at 08:45
1. The US was not at the zero bound between December 2007 and December 2008 when the bulk of the NGDP collapse occurred, using monthly NGDP estimates.
Yeah, everyone seems to forget the issue was not whether the Fed could engage in monetary stimulus at that point, they had decided not to! Look at those Fed minutes, they were focused on inflation as though they had a single mandate.
Britonomist — That’s not quite monetizing debt, because the Fed only needs to promise to hold the debt forever if necessary — they’re just promising to not unwind the purchases until inflation is high enough that the Fed identifies a need to sell them to tamp NGDP down into the target range. The distinction is important because of how it affects expectations — the mistake the Fed has been making is to insist all QE is temporary and will be rolled back as soon as possible, but it would be an equally large mistake to allow investors to think the Fed can ever be a spigot for government spending.
As Vivian points out, within those expectation bounds QE is a kind of win-win — for the period the Fed holds government debt the interest is returned to the government, essentially monetizing it; meanwhile, NGDP is forced into a healthier range for the economy as a whole.
25. May 2015 at 13:30
Then there is the “monetary policy desperately needs re-armanent” reason for the conventional wisdom to move along.
http://nextbigfuture.com/2015/05/hsbc-says-there-is-no-traditional.html
Change the framing of policy and the possibilities get expanded.
25. May 2015 at 15:44
@scott
Goal-Stable NGDP growth
Target-NGDP expectations
Tool-Adjustments in the monetary base
Mechanism-HPE
Is that complicated?
Not complicated….. just incorrect, which is why monetary policy is poorly understood.
25. May 2015 at 16:24
@scott
Also, do you think the goal of monetary policy is stable NGDP growth. I think not. There is some expectation of employment/unemployment levels from monetary policy.
26. May 2015 at 01:44
[…] themoneyillusion.com: Seeing Lu Mountain Scott Sumner quotes this Brad DeLong post. It’s an interesting discussion because I recently […]
26. May 2015 at 04:25
DeLong and Sumner miss the point that the housing bubble was first and foremost a credit bubble. Housing was just the real asset that enabled trillions in phony assets to be created, bought borrowed and sold.
The nature of credit leverage meant that with just a modest decline in asset prices the lending firms were insolvent. Most of 2008 was wasted hoping that the insolvency would be just a temporary condition. In some ways it was. The key solution was not TARP but the reversal of mark-to-market regulations. US stocks declined until March 2009 when it became known this policy would change. Stocks have tripled in value since then. Coincidence?
The Federal Reserve manifestly goofed in 2008. Their participation in the policy panic of fall 2008 is particularly egregious. And panic is what all the policymakers did. Henry Paulson in particular was pathetic and President Bush’s support of him is one of his signature failures of his administration (yea, I know there are many contenders for that prize).
Perfect monetary policy may have lessened the decline but it would not have prevented the financial crisis. There was going to be an economic reckoning for the dishonesty of the credit bubble. And it must be this way. For the history of the world shows that every institution that mocks economic reality eventually gets “called” and those holding a losing hand must show their cards.
26. May 2015 at 04:32
Beefcake, I’m not 100% sure it’s a bad idea, but when you look at if from a consolidated government balance sheet perspective, it’s exactly like the Treasury issuing additional bonds and using the money to buy up stocks and commodities and corporate debt.
The argument in favor is that stocks outperform T-bonds in the long run, the government becomes a sort of hedge fund. I just don’t see it as a monetary policy issue, it’s a broader issue.
Jose, I agree on the mortgage bonds.
Majromax, Yes, that’s right.
Britonomist, We are just going around in circles. I’ve explained why QE works, when it works and when it doesn’t work, and swatted down all the arguments against it. The critics engage in whack a mole. “It won’t be enough” “Then buy more” “What if they run out of stuff to buy?” “Then raise the NGDP target so the demand for base money is lower” “What if the central bank is too conservative to do so?”
This game goes on forever, it has no end.
Vivian, I agree. Economists have always recognized that money printing is an inflation tax, even in normal times. But they still find it useful to distinguish between monetary and fiscal, as you suggest.
Lorenzo, They keep trying to reinvent the wheel.
dtoh, Yes there’s a hope that employment will be more stable, also a hope there will be fewer financial crises and bailouts, but I think the term “goal’ refers to the variable you want to move on a specified path. There is no objective to make employment move on a completely smooth path—what if boomers suddenly start retiring?
26. May 2015 at 05:00
scott,
i think the hope (goal) is stable full employment. NGDP is the target for getting there. At least that’s how I’d look at it, but I suppose you could make a reasonable argument that a specific rate (e.g. 4.5%) of stable NGDP growth is the goal.
26. May 2015 at 05:07
“Jose, I don’t see any good arguments for having the Fed buy assets such as stocks.”
Farmer (and I disagree here)…
I’d prefer the NGDP Futures market be used as the transmission mechanism – only let SMB owners in…. but failing that, I think the Fed should buy / sell a random basket of US stock market stocks.
Then who cares if they retire them, just sign them over to shareholders?
Buying Govt Debt is just bad form.
26. May 2015 at 06:25
It was time for Greece to bend 5 years ago:
http://www.nytimes.com/2015/05/26/business/dealbook/with-money-drying-up-greece-is-all-but-bankrupt.html
Let’s hope they bend now.
26. May 2015 at 06:45
Two Italian economists bare a gift for Greeks (a parallel currency to overcome the ‘euro bound’);
http://www.voxeu.org/article/parallel-currency-greece-part-ii
‘There are alternatives to austerity as the only adjustment rule for Greece, a rule that while dramatically impoverishing the country is also proving ineffective in reviving credible growth prospects. Introducing a currency in parallel to the euro could help Greece change trajectory and move out of depression rapidly, provided that the currency is designed in a way that carries incentives for private-sector spending.
‘A variety of options have been proposed. These have been illustrated and discussed in this two-part column. All options require the state to transform future tax revenues into immediate spending capacity. Each option, however, has a very different expected impact on domestic aggregate demand and fiscal sustainability. We believe that only the tax credit certificates, used as a quasi-currency, would have enough power both to create new spending capacity and to generate enough resources to pay for themselves over a recovering economic cycle. Also, sufficient safeguards can be built on these certificates, which would protect the budget against the risk of fiscal under-performance in a way that would prevent new austerity measures from becoming necessary.’
26. May 2015 at 06:56
@ Morgan,
A little off topic, but doesn’t this article touch on your concept of everything going digital?
http://www.forbes.com/sites/timworstall/2015/05/26/goldman-sachs-is-entirely-right-about-economic-growth
26. May 2015 at 08:07
DeLong: “Spending on housing construction rose by 1%-age point of GDP for about three years-that is $500 billion. In 2008-9 real GDP fell relative to trend by 8%-that is $1.2 trillion-and has stayed down by what will by the end of this year be seven years-that is $8.5 trillion.”
Some of the amplification here comes from leverage at the financial institutions and from derivatives. Most, if not all, of the MBS’ issued were removed from balance sheets in accordance with GAAP which hid their risks and also helped amplify their losses on those same balance sheets.
But these accounting malpractices were just a part of it. Tight money helped instill a terrible fear in banks that liquidity and credit would dry up.
26. May 2015 at 09:38
Tight money? In the first half of 2008 the CPI was increasing at a 7.5% annualized rate. The CPI in 2007 had increased 4% so not only was the Fed above its target but inflation was accelerating away from their target.
26. May 2015 at 10:39
Dan,
That’s the importance of supply-side vs. demand-side inflation, as well as core vs. overall inflation. If only oil prices go up and not the general price of goods, that’s not a monetary effect. That’s a supply-side effect.
Monetary policy CAN offset the increase in oil prices, but monetary prices decreases all prices. So let’s say oil goes up 20% while the rest of the economy’s prices go up 2%. You can theoretically put the whole economy in a 17.3% deflation to get oil back down to its original level. Does that really make sense?
So that’s one reason to support NGDPLT. I think overall CPI (NOT core) level targeting gets 90% there. Core CPI has less volatility, but overall CPI level targeting would have the volatility cancel out over the long-run.
26. May 2015 at 11:30
Scott,
I’m curious about your response to this recent Krugman post.
http://krugman.blogs.nytimes.com/2015/05/25/grexit-and-the-morning-after
This sentence particularly stuck out to me:
“But now the IMF is playing bad cop, declaring that it cannot release funds until Syriza toes the line on pensions and labor market reform. The latter is dubious economics “” the IMF’s own research doesn’t support enthusiasm about structural reforms, especially in the labor market.”
He didn’t link to their own research which showed this. I did some googling and the only IMF research from the IMF I could find found structural reform helped growth. Labor market reforms particularly increased overall economic growth, even though they could hurt protected interests. The labor market protections were there for a reason.
I honestly have no idea what research Krugman is referring to. IMF publishes a lot of stuff, so they probably published something which found structural reforms didn’t help.
More generally, it’s tough to know where Krugman is coming from. He approves of anti-austerity parties in Europe, including Syriza. But Syriza’s anti-austerity stance can only exist in continued monetary gifts from the Troika. Is he saying the Troika should write a blank check because monetary policy is ineffective? I don’t understand.
26. May 2015 at 13:05
Matt, Krugman’s rapidly abandoning the views he held in the 1990s, without telling us why. It’s all very strange.
27. May 2015 at 01:22
“Global unions Washington officer Peter Bakvis reports that in an analytical chapter prepared by the IMF for the April 2015 edition of its World Economic Output (WEO) report, to be released in full on 14 April, Fund researchers found no evidence that deregulatory labour market reforms could have a positive impact in increasing economies’ growth potential. As Peter comments, the finding is significant given that labour market deregulation has featured prominently in IMF loan conditions and policy advice for many countries, most notoriously so in several EU crisis countries.”
http://touchstoneblog.org.uk/2015/04/labour-market-deregulation-when-the-facts-change/
27. May 2015 at 03:46
@Scott, @Morgan Warstler
I believe that when a government agency buys private assets it is intervening further in the economy, and that may not be good. But let’s imagine a world where there is no government, but society has agreed to have a monetary authority. What assets should the monetary authority buy then, given that there is no government bonds to buy ? To me it is intuitive it should monetize a certain portion of all the assets available in the economy, proportional to market capitalization. If not, what then ?
27. May 2015 at 05:58
Why are you so considerate? – Viz.,
“The Great Recession was caused by tight money. It’s that simple”.
Bankrupt U Bernanke caused the Great-Recession all by himself. I have no sense of etiquette. He should be jailed.
You can either look at in Keynesian or monetarist terms, i.e., negative real rates of interest rates, or monetary flows. The last period of negative real rates prior to the boom was in Jan-Oct 1980. Then we got Aug 2002 thru Apr 2006 negative real rates (using the 1 year constant maturity and the CPI).
But that doesn’t explain the bust. The moron drained monetary flows (the proxy for inflation), beginning in Feb 2006 for 29 contiguous months (turning otherwise safe-assets into impaired assets (i.e. turning real-estate prices upside down and unsaleable).
Then as the 4th qtr. collapse was imminent Bankrupt U Bernanke pursued a contractionary money policy beginning in July. Money flows (the proxy for real-output), proceeded to fall by -160 percent.
Bankrupt U Bernanke and the 300 Ph.Ds. on the Fed’s research staff don’t know money from liquid assets. The 1966 S&L credit crisis is the paradigm (esp. apropos for those who want to eliminate cash). Then the FDIC provided unlimited transaction deposit insurance. And the unchecked “flight-to-safety” both pricked the commodity bubble and contracted the E-dollar market…
Yellen’s no different. Her contractionary money policy will prick the stock market bubble on 9/2/2015 (just like she collapsed the oil market).
Nostradamus
27. May 2015 at 06:19
You have cause and effect wrong. Roc’s in M*Vt = aggregate demand. N-gDp does not = AD (only in Keynesian economics). N-gDp is simply a proxy for roc’s in money flows in Irving Fisher’s “equation of exchange”.
Nostradamus
27. May 2015 at 12:55
Here’s who caused our recessions since the Great-Depression:
Period …………….. Chairman
Nov 1948 – Oct 1949… Thomas B. McCabe
July 1953 – May 1954… William M. Martin
Aug 1957 – Apr 1958… William M. Martin
Apr 1960 – Feb 1961… William M. Martin
Dec 1969 – Nov 1970… William M. Martin
Nov 1973 – Mar 1975… Arthur Burns
Jan 1980 – July 1980… Paul Volcker
July 1981- Nov 1982… Paul Volcker
July 1990 – Mar 1991… Alan Greenspan
Mar 2001 – Nov 2001… Alan Greenspan
Dec 2007 – Jun 2009… Ben Bernanke
I.e., all these recessions were both predictable and preventable.
27. May 2015 at 14:09
“The US was not at the zero bound between December 2007 and December 2008 when the bulk of the NGDP collapse occurred, using monthly NGDP estimates”
—————-
The roc in N-gDp), peaked in the 2nd qtr. of 2006 @ 12%. Bernanke let it fall to 8% by the 4th qtr. of 2007 (or by 33%). N-gDp fell to 6% in the 3rd qtr. of 2008 (another 25%). N-gDp then plummeted to a -2% in the 2nd qtr. of 2009 (another – 133%).
Roc’s in M*Vt = roc’s in N-gDp. The roc in money flows was already predestined to fall below zero in the 4th qtr. of 2008 as early as Dec. 2007. Bankrupt U Bernanke should have performed his helicopter drop way before the economy started sinking. Don’t sugar coat it. He is an idiot and should be jailed.
Nostradamus
29. May 2015 at 06:58
Albert Einstein quipped “Insanity is doing the same thing over and over again – expecting different results”
That’s what the Fed does. 1st qtr 2011, 1st qtr 2014, 1st qtr 2015.
Roc’s in M*Vt = roc’s in N-gDp (proxy for all transactions in Professor Irving Fisher’s “equation of exchange”).
The roc in the proxy for inflation fell by 2/3 (from Jan 2013 until Dec 2014), the roc in the proxy for real-output fell by 1/2 (from July 2014 until December 2014). Go figure.
Nostradamus
29. May 2015 at 08:51
Postkey, That’s really appalling, one more reason to abolish the IMF.
Jose, I’m not sure what you have in mind. Is it a profit maximizing monopoly? A cooperative?
29. May 2015 at 09:03
flow5,
Do you calculate the roc’s or is this data available online?
31. May 2015 at 09:20
I’m going to give it to Putin so he can fulfill Lenin’s prophecy.