Krugman on helicopter drops
Every so often a Keynesian commenter will ask me about “helicopter drops.” This term refers to money-financed deficit spending. People will usually say something to the effect that “everyone agrees” a helicopter drop would boost aggregate demand. In fact, not everyone does agree. More specifically Paul Krugman and I don’t agree. At least we don’t believe that helicopter drops solve any fundamental problems created by “expectations traps.” Thus the Bank of Japan ran large deficits in the 1990s and early 2000s, much of which were financed by printing money. But the Japanese public did not expect the monetary injections to be permanent, mostly because the BOJ promised to avoid inflation, and they could only do so by (implicitly) promising to pull the excess money out of circulation when the economy started to recover. Sure enough, the BOJ reduced the monetary base in Japan by 20% in 2006 (when the Japanese economy was showing signs of recovery.) A couple years later deflation returned.
Paul Krugman recently expressed the same view:
And what about money-financed deficits? This sounds at first like something you can just do, not a commitment issue. But here’s what you should ask: how do you know whether a deficit is money-financed? As long as you’re in a liquidity trap, it doesn’t matter at all what it says on the zero-interest pieces of paper you issue; they may say that they’re dollar bills or they may say that they’re T-bills, but they’re completely fungible at the margin. What matters is what happens after you emerge from the liquidity trap, and that depends on how the central bank acts: does it withdraw the monetary base it created when the economy was depressed, or does it let it stay out there and cause some inflation?
In other words, you don’t know whether a deficit was really money-financed or not until the zero bound is no longer binding, and hence the effectiveness of allegedly money-financed spending depends on expectations “” again.
The only area where we disagree is the plausibility of expectations traps. I don’t think they are at all likely to occur in the real world. If a central bank promised to inflate, it would be believed. Krugman thinks expectations traps could be a problem, although he also argues that this is no reason not to try monetary stimulus:
You can see where I’m going here. Menu B is, if you like, safer for the Fed than Menu A, because it is defined in terms of actions rather than results; the Fed can point to what it is doing, rather than announce a target for long-term rates or inflation that it might fail to hit. So Menu B serves institutional objectives better. Unfortunately, it doesn’t do the job for the economy. To be fair, we don’t know that Menu A would, in fact, be sufficient. But Bernanke the Younger “” BB before he was assimilated by the Fedborg “” would have said that this was no reason not to try.
This is like a Greek tragedy. Most economists are clueless about what needs to be done. Bernanke probably does understand, but (as the recent Fed minutes showed) cannot persuade his colleagues. And so he watches all this unfold with what must be a horrible feeling in the pit of his stomach.
Tags:
6. April 2012 at 18:41
Money printed to cover deficits will reflux to its issuer just as fast as any other money. As long as the government has positive net worth, and as long as at least one channel of reflux is open, the Law of Reflux says that helicopter drops will have no effect on the value of money (or on aggregate demand, whatever that means)
6. April 2012 at 19:00
Hey Scott, off topic but have you seen this?
http://www.youtube.com/watch?feature=player_embedded&v=TypWtZYLCF8
hehe, I wonder what the Market Monetarist version is? 🙂
6. April 2012 at 19:04
At least we don’t believe that helicopter drops solve any fundamental problems created by “expectations traps.” Thus the Bank of Japan ran large deficits in the 1990s and early 2000s, much of which were financed by printing money. But the Japanese public did not expect the monetary injections to be permanent, mostly because the BOJ promised to avoid inflation, and they could only do so by (implicitly) promising to pull the excess money out of circulation when the economy started to recover. Sure enough, the BOJ reduced the monetary base in Japan by 20% in 2006 (when the Japanese economy was showing signs of recovery.) A couple years later deflation returned.
The BOJ reduced inflating in 2006 for the same reason the Fed reduced inflating in 2006: To avoid what they feared as runaway consumer price inflation after years of monetary easing, and thus overreaching their price inflation target. It wasn’t because the Japanese public knew and were “correct” that the BOJ promised to avoid inflation and then made good on it by taking the money out of circulation. John Q Public doesn’t abstain from spending money the BOJ prints solely because they believe the BOJ promises low inflation in the future.
If you were given $1 million of inflation money, would you abstain from spending it on the basis that you expected price inflation to be “only” 1% next year instead of 10%?
The majority of the public go by past prices and then naively extrapolate those prices forward. They don’t anticipate future prices by watching central bank minutes and announcements.
6. April 2012 at 20:57
Scott wrote:
“This is like a Greek tragedy. Most economists are clueless about what needs to be done. Bernanke probably does understand, but (as the recent Fed minutes showed) cannot persuade his colleagues. And so he watches all this unfold with what must be a horrible feeling in the pit of his stomach.”
“Drop, drop- in our sleep, upon the heart
sorrow falls, memory’s pain,
and to us, though against our very will,
even in our own despite,
comes wisdom,
by the awful grace of God”
AESCHYLUS (Agamemnon 179-183)
6. April 2012 at 21:56
Brito,
I actually took the time to watch your link. I’m still nursing my ribs. Thank You! 🙂
7. April 2012 at 01:35
Krugman’s “It’s baaaack” paper seems to get very little attention among monetary policymakers these days even though it shows this principle need to commit to not undo inflationary stimulus when the economy turns around. However, perhaps more irritating, one man that always seems get forgotten in discussions about monetary policy at the ZLB is Lars Svensson even though he not only showed a similar result to Paul (http://www.nber.org/papers/w10195) but he’s not been afraid to enact “unorthodox policies” whilst actually being the deputy governor of the Swedish central bank, having implemented a negative central bank interest rate.
Surely there’s a lesson to be learned for other policymakers (cough – Bernanke, King) to be courageous, either implement policies that they’ve previously suggested but failed to deliver (Bernanke again) or look at alternatives that a broad base of highly intelligent experts, such as yourself and C. Romer, have been suggesting i.e. NGDP level targeting.
For this, I think Lars Svensson deserves a lot more credit than he often gets.
7. April 2012 at 03:43
Scott – I called it “Missing dish on Krugman´s menu”:
http://thefaintofheart.wordpress.com/2012/03/31/missing-dish-on-krugman%C2%B4s-menu/
7. April 2012 at 05:08
marcus said my piece
7. April 2012 at 06:35
If a central bank promised to inflate, it would be believed.
Just to get this straight, your talking “Austrian” here? Expecting the bank to “inflate” means expecting money injections? Or it means expecting a rise in the price level?
7. April 2012 at 07:56
“Sure enough, the BOJ reduced the monetary base in Japan by 20% in 2006 (when the Japanese economy was showing signs of recovery.)”
Can both the following statements be true?:
1) The Japan 1998-2006 monetary base doubling had no impact because it was seen as impermanent.
2) The Japan 2006 monetary base drop of 20% had an impact because it was a sign that the preceding base growth was impermanent.
To extend the analogy to the U.S., this would be like saying:
1) the Fed’s 2007-2012 $1.5tr reserve growth had no impact because of the 25bp IOR.
AND
2) reducing reserves by $300b in 2012 would have an impact.
7. April 2012 at 09:06
Krugman’s latest;
http://seattletimes.nwsource.com/html/opinion/2017926778_krugman08.html
———–quote———-
To be sure, more aggressive Fed policies to fight unemployment might lead to inflation above that 2 percent target. But remember that dual mandate: If the Fed refuses to take even the slightest risk on the inflation front, despite a disastrous performance on the employment front, it’s violating its own charter. And, beyond that, would a rise in inflation to 3 percent or even 4 percent be a terrible thing? On the contrary, it would almost surely help the economy.
How so? For one thing, large parts of the private sector continue to be crippled by the overhang of debt accumulated during the bubble years; this debt burden is arguably the main thing holding private spending back and perpetuating the slump. Modest inflation would, however, reduce that overhang “” by eroding the real value of that debt “” and help promote the private-sector recovery we need. Meanwhile, other parts of the private sector (like much of corporate America) are sitting on large hoards of cash; the prospect of moderate inflation would make letting the cash just sit there less attractive, acting as a spur to investment “” again, helping to promote overall recovery.
In short, far from fearing that more action against unemployment might lead to an uptick in inflation, the Fed should actually welcome that prospect. Which brings me back to those Republican attacks and their chilling effect on policy.
True, Bernanke likes to insist that he and his colleagues aren’t affected by politics. But that claim is hard to square with the Fed’s actions, or rather lack of action. As many observers have noted, the Fed’s own forecasts indicate that while things have been looking up a bit lately, it still expects low inflation and high unemployment for years to come. Given that prospect, more of the “quantitative easing” that is now the main tool of Fed policy should be a no-brainer. Yet the recently released minutes from a March 13 meeting show a Fed inclined to do nothing unless things take a turn for the worse.
So what’s going on? I think that Fed officials, whether they admit it to themselves or not, are feeling intimidated “” and that American workers are paying the price for their timidity.
————endquote———–
7. April 2012 at 09:34
CNN just hosted an interesting debate.
The guests were: Ken Rogoff, Diane Swonk, and Stephen Moore.
Moore argued that we’ve had 3 years of monetary crack cocaine, and it hadn’t worked.
Rogoff rebutted that the Fed had been too timid.
Swonk said that the Fed was doing the right thing.
Moore got a chance to respond and played the dystopian “$6 gasoline” and “$2000 gold” card — that if we had a little more inflation it would all go to hyperinflated commodities.
Swonk was shaking her head back and forth during Moore’s tirade.
Rogoff said it’s a slow grind and a little more inflation would take the edge off.
Nothing new here, except that it was a media program that actually presented monetary ease in a modestly favorable note.
7. April 2012 at 10:50
Yes, The Fed has to print more money, and make it obvious, and set up NGDP for three years running, and day it will hit those targets, no matter how much QE it takes.
Really, I have seen no meaningful objections to Market Monetarism, only sniveling and naysaying by gold nuts, Theo-Monetarists or currency fetishists, who insist we can accomplish nothing or hyperinflation.
Somehow the moderate inflation of 1982 to 2007, and attendant solid growth and rise in living standards, was a fluke.
7. April 2012 at 10:55
Benjamin Cole:
No, the Fed does not have to print more money, and it matters how much QE it takes.
Really, I have seen no meaningful justifications of Market Monetarism, only sniveling and naysaying by fiat nuts, Theo-Monetarists or currency fetishists, who insist we can accomplish stability by giving all power to create money to a group of monopoly counterfeiters.
Somehow the moderate price inflation of 1982 to 2007, which culminated in a spectacular crash, and attendant absence of real growth and stagnation in living standards, was a fluke.
7. April 2012 at 11:01
Mike, The Fed stopped redeeming dollars in 1933. That’s why money issue is now inflationary, and didn’t have much impact prior to 1933. Nick Rowe has a good recent post on this.
Brito, Cute.
MF, Japan hasn’t had inflation since the early 1990s.
Mark, Excellent.
Jason, I agree, although it actually gets more attention than it appears, because you commit to future money injections by committing to a future price level or NGDP target. So everyone calling for P or NGDP targeting is implicitly relying on the Sumner(1993)/Krugman (1998) expectations trap model.
Dan, No, I’m not talking Austrian, I talking about policy commitments. BTW, I hope you aren’t the same Dan Kervick who is making up false statements about me over at Naked Capitalism. Or the one who confused supply and demand shocks when criticizing Paul Krugman. Please say it ain’t so.
David Pearson, I think you misunderstood my point (which is completely uncontroversial, AFAICT.) Everyone agrees that temporary currency injections are non-inflationary. In 2006 the BOJ reduced the monetary base by 20%, despite there being no inflation in Japan. This confirmed the expectations that the BOJ would do whatever was necessary to prevent inflation. The expectations of the Japanese public were completely rational.
It’s difficult to talk about the effect of any single action in isolation—what matters is the future expected path of policy.
Patrick, Krugman’s right, except that he confuses people like Dan Kervick by talking in terms of inflation rather than NGDP.
Steve, That’s one reason I am no longer a member of the Cato Institute.
7. April 2012 at 11:13
ssumner:
MF, Japan hasn’t had inflation since the early 1990s.
PRICE inflation, no. But price inflation is not necessary for real economic growth, and too many economists who look at prices and don’t understand their significance incorrectly infer that a lack of price inflation is somehow equivalent to depression/lost decades.
Japan hasn’t had price inflation since the early 1990s, and yet they have experienced steady real growth almost the entire time (up until the collapse of 2008).
Japan refutes the myth that real growth requires “gradual” price inflation.
7. April 2012 at 11:33
“This confirmed the expectations…”
One can say markets were sure something would happen; or one can say markets were impacted by that something happening. I’m not sure one can do both.
7. April 2012 at 11:59
Sorry, I should have been clearer in my earlier comment. I should have linked to Brad DeLong’s post: http://delong.typepad.com/sdj/2010/07/helicopter-drop-time-paul-krugman-gets-one-wrong.html
I.e., no great hoovering. I was talking about a permanent helicopter drop, which works, and a permanent OMO, which also works. Temporary monetary injections are never effective, whether they come in helicopters or asset purchases (“Why would people bid up asset prices if the central bank was expected to pull the money out of circulation in the near future?” – Scott Sumner). I have read your earlier posts, Scott, particularly this one: http://www.themoneyillusion.com/?p=7960 and this one: http://www.themoneyillusion.com/?p=9627 which taught me most everything I needed to know about “liquidity traps” (I also like this one http://www.themoneyillusion.com/?p=5763). Plus I *have* read your earlier statements on helicopter drops (http://www.themoneyillusion.com/?p=10888 ; http://www.themoneyillusion.com/?p=6119), which were my early favorites. So I don’t think I’ve misunderstood you. I am simply trying to explain how things look from the perspective of those who don’t read you, and what they’re missing. (Until they’re reached, policy won’t change, regardless of what Krugman and Bernanke know.) Helicopter drops look more effective because they don’t crowd anything out, whereas with an OMO, it looks like the injection won’t have velocity because it substitutes as a store of value. What’s being missed here, of course, is that:
a) extreme liquidity preference is a symptom of tight future (and hence current) monetary policy, tight expected NGDP, so that even entire helicopter drops will be hoarded (demand for savings will rise over the period) if future NGDP is expected to stay put
b) hoarded injections still add to nominal spending in the long run, provided they are not evacuated prematurely (and if expected to be effective, then the required injection may be less than zero)
Is that a correct statement of Market Monetarism, Scott? Or am I still too Keynesian? Please, I’m here to learn.
P.S. I always understood Krugman’s argument as being that the Fed Chairman never even would make the promise to inflate (he thinks like a Keynesian, inflation, not nominal income, like a MM or Friedmanite) because a credible Fed wouldn’t do that, and if it did then the markets wouldn’t know what to expect anymore, which the Fed doesn’t want, so the markets know it won’t. Ie. the markets believe that the Fed could never make such a promise, and their expectations are right, as usual.
P.P.S I see Bernanke is still a “good guy” for you, then. What would you say if you read this http://thefaintofheart.wordpress.com/2012/03/30/bernanke-is-truly-amazing/ Marcus Nunes post?
P.P.P.S. How do you get italics and hyperlinks on this interface? And Econlog has that great inline quote thing, too.
7. April 2012 at 12:05
Oh, yay, hyperlinks do work.
Responding to your reply to me on the earlier post, which I’ll do here because I’m afraid it would get ignored over there (over 100 comments):
“I don’t understand why you say a helicopter drop would increase money supply and demand equally, creating no HPE.”
For the same reason that a currency reform, or Hume’s thought experiment for that matter, increases quantity supplied and demanded of stocks of nominal money balances equally at all interest rates. Money receipts are either spent, boosting velocity, or held, adding to aggregate stock demand. In Hume’s case, wouldn’t both allocations double in nominal terms, prima facie? The boost to nominal spending comes from the nominal doubling of both stocks and flows.
“And the HPE assumes constant V, not a change in V.”
The HPE is the result of a collective readjustment of portfolios, when everyone finds themselves with a little too much money vis-a-vis other assets and tries to pass it onto others. The velocity of those particular dollars would skyrocket, wouldn’t it? (I do find it helpful to look at velocity as a “real” concept.) Indeed, the total quantity of money need not change at all, and yet nominal spending would still rise, so long as everyone had too many non-interest bearing assets in their portfolios (As when the Wicksellian rate rises – you yourself said this was how fiscal stimulus worked). An OMO will raise NGDP slightly more than an equivalent HD, ceteris paribus (assuming permanence, of course).
7. April 2012 at 12:07
“MF
Benjamin Cole:
No, the Fed does not have to print more money”
He’s right you know. If the right target was set and credible, the Fed would be evacuating money, not printing it (assuming it also ends IOR).
7. April 2012 at 12:18
Saturos:
He’s right you know. If the right target was set and credible, the Fed would be evacuating money, not printing it (assuming it also ends IOR).
That assumption is necessary, it’s not exactly an offhand “also” criteria.
What’s NGDP growth right now? If it’s anything less than “target of 5%”, then the Fed would have to be creating more money, all else equal. If the Fed promised to target 5% NGDP, then unless the IOR policy is abandoned, the Fed will have to create more money.
7. April 2012 at 12:26
Another thing. It’s interesting how you suggested I *Google* your earlier posts – not use the search box at the top of the page. (Clearly you’ve been having as much trouble with it as I have.)
7. April 2012 at 12:30
Major Freedom:
I’m never offhand.
“If the Fed promised to target 5% NGDP, then unless the IOR policy is abandoned, the Fed will have to create more money.”
If the IOR wasn’t abandoned, the Fed could print money until Ron Paul declared revolution, and it still wouldn’t work. Nobody would think the Fed was serious about hitting higher NGDP levels as long as it still paid IOR – unless Congress actually wrote it into their mandate (Not this decade).
7. April 2012 at 12:34
Wouldn’t money financed deficit spending be effective if it were truly money financed? That is, that the central bank printed the money and gave it to the Treasury to spend; not that the Treasury borrows the printed money. Once the Treasury borrows it and the central bank has the bonds or notes to sell back into the market, then no one has confidence that the money will stay in circulation, so it just sits in the bank.
Thought experiment:
A: the Fed GIVES everyone a million dollars in cash. Would we have a debt overhang left in a week? Wouldn’t the price level rise pretty quickly to a new level?
B: the Fed gives everyone a million dollars in cash but in exchange takes a note for a million dollars that it can turn in at any time. The interest rate on the note is less than 1%, maybe even 0%. Most people would put the money in the bank and wait for the Fed to call back and ask for the million dollars. Nothing changes.
The Cash financed debt in option A is key because it’s clear to everyone that the Fed is “committed” to this monetary expansion because it has no way to recall the money. Frankly, $10,000 per American equals $3 trillion, so clearly an amount that small or probably even less than that would be a pretty amazing jolt to the price level and have a serious effect on AD, NGDP, the debt overhang and the price level.
The Fed can create inflation if it wants. So while Krugman is hoping the Fed will make a commitment, we’re stuck with a Fed that not only won’t commit to some inflation or NGDP level targeting, we’ve got a Fed that doesn’t even want those things to begin with.
7. April 2012 at 12:42
“The Fed stopped redeeming dollars in 1933.”
They stopped redeeming for gold in 1933. They still redeem for bonds, loans, etc. Sometimes the fed even sells used furniture for dollars, in which case they redeem for furniture.
7. April 2012 at 12:44
Bill,
The Fed boosts NGDP all the time using option (B). Or at least, it used to. (But it doesn’t lend cash, it pays people for bonds (or “bills”). It doesn’t “turn them in”, it sells them)
The key thing is what the public expects future NGDP to be. Then the required quantity of money is endogenous. You seem to think its the other way around. As Lars Christensen showed, under effective policy, V will rise to offset undershoots in M.
7. April 2012 at 12:56
General P.S.
Just reread Brad DeLong’s “How I Learned to Love the Liquidity Trap”. (This is the view I’m trying to explain the way out of, based on what I’ve learned here) You’d hardly think a Keynesian needed help on how to stop worrying and start living, when it comes to “liquidity traps” – it’s their raison d’etre. As Scott has explained: http://www.themoneyillusion.com/?p=220.
There, I think I’ve blurted out what all my favorite posts are now.
7. April 2012 at 12:57
Nope – just checked the list – not by a long shot. Scott’s a fine writer. (You should totally do a book.)
7. April 2012 at 12:58
Dan Kervick,
I too am hoping you are not the one who wrote the Naked Capitalism article. But if you are, just remember that lies are no way to make a better world. If you did not already see it, I left a comment for you shortly after the comment you left with the Naked Capitalism link.
7. April 2012 at 12:59
Whoops, forgot the link (won’t make that mistake again – though if it inspires more posts like this, maybe I should!):
http://delong.typepad.com/sdj/2011/11/the-sorrow-and-pity-of-the-liquidity-trap.html
7. April 2012 at 13:21
Aah, I see what was confusing in my HD explanation. Of course the people who receive the HD amounts don’t increase their demand by the amount that’s supplied to them. But then neither do we demand all of the money balances we initially receive, without any monetary injections. Monetary supply and demand is an equilibrium condition which holds in the aggregate, not for individual rounds of spending. K is a fraction, supply always exceeds demand *for each individual*. But in the aggregate, at the end of the circular flow, all money is held – and then ultimately spent again. So when you draw the supply and demand for monetary balances, you should really draw several demand curves, one for each round of spending. Equilibrium means the *series* of quantities demanded equals the aggregate supply.
7. April 2012 at 14:35
@MF
Japan hasn’t had price inflation since the early 1990s, and yet they have experienced steady real growth almost the entire time (up until the collapse of 2008).
Still peddling this fish, eh? It’s not smelling any better with age.
Just reminding you of what you’ve seen before, via the IMF:
~~~~
Japan versus All Advanced Economies, GDP per capita, 1992-2008
By PPP, 2008 dollars
__________ 1992 ___ 2008 __ Change _ growth rate
Adv Econs… 30,790 ….. 37,921 .. +23.2% … 1.31%
Japan ………… 31,721 ….. 33,997 ….+ 7.2% … 0.43%
~~~~
“they have experienced steady slight, well below average real growth almost the entire time”. Fixed it for you.
BTW, I enjoyed your comment about how Austrianism uses irrefutable principles, which can never be falsified by making mistaken predictions because they do not enable predictions to be made about the future.
They apparently can’t predict the past either.
7. April 2012 at 15:16
I too enjoyed the “irrefutable principles” comment. Sounds more like a religion than anything else
7. April 2012 at 19:58
Bernanke probably does understand, but (as the recent Fed minutes showed) cannot persuade his colleagues. And so he watches all this unfold with what must be a horrible feeling in the pit of his stomach.
Yes, that horrible feeling is the gun held by the most powerful bankers in the world. He knows who they are; they’re the guys that put him in his current position.
7. April 2012 at 20:05
oh come on bob, please expand on this. can you name some names?
8. April 2012 at 00:24
Brito,
I’d say, “Shit happens at 5% per annum, unless Bernanke gets constipated”.
8. April 2012 at 02:31
Also, the argument against temporary injections also applies to fiscal stimulus, as Scott explained in my previous link. So that’s the real reason why we should expect the multiplier to be zero – which of course was reflected in the market reaction, or lack thereof, after ARRA was announced. (I just finished watching the Kauffman forum.)
8. April 2012 at 02:44
And to avoid further misunderstandings, by HD, I mean actual drops to the public, not money-financed deficits – though I think they are the same in a mechanical sense. And of course, I was ignoring the effect on expectations – otherwise if you took that into account, the effect on V would dwarf delta M, creating hyperinflation. Again, Scott explained this already, in a post I linked to.
8. April 2012 at 05:02
“Most economists are clueless about what needs to be done.”
And in my book, that includes both Scott Sumner and Paul Krugman.
I have no doubt that Japan, and the US, Europe, UK etc could blow their present problems away by inflating – just exchange money for assets as generously as necessary. Eventually, no-one will believe any inflation target anyway. I thought even Krugman believed that a Japanese liquidity trap could overcome by enough intervention against the yen (which would have brought Japan into trade conflict with the US, but let’s set that aside for now). But should they? The mal-redistribution involved, and the exacerbation of moral hazard, which I believe got the US and the UK into their present mess, would be huge.
By all means, central banks should meet the additional demand for base money arising from the precautionary motive after a bust, but they can do this without changing their inflation target. Ultimately though, the real problem is mispriced assets after the boom (eg, in Japan’s case, Olympus), and it falls to the government to ensure that (a) this fact cannot be concealed and (b) that any structural change in the light of this information is not blocked. Of course this is unpopular work, and the government can be expected to foster the perception that managing real activity is a job for the central bank. But it is not, and central banks should tell governments to get lost, which is of course the purpose of their independence.
Scott Sumner and Paul Krugman have a macroeconomic hammer, which might well be able to drive in the screw, but the outcome would be so much better if the government would admit that it possesses a screwdriver and use that instead.
8. April 2012 at 05:28
Rebeleconomist:
Sumner favors keeping nominal GDP growing on a stable growth path. With such a target rule, inflation changes with the productive capacity of the economy. If productive capacity is depressed for whatever reason, then there is higher inflation. Creditors who lent when the expectation was a higher growth path of productive capacity receive a lower real return than they expected. Debtors who promised to make payments when productive capacity was expected to grow more quickly, end up paying less than anticipated.
However, this does not protect debtors against changes in asset prices. For example, if home prices are expected to rise, and a borrower and lending make an agreement to finance a house, and home prices fall, inflation doesn’ rise to “bailout” the debtor. Because new houses are being produced, and the falling house prices will involve lower demand for new homes, there will need to be a reallocation of resources way from new homes to other goods. Perhaps goods for exports. Since the productive capacity of the economy is set up to produce houses and not export goods, this shift in the composition of the demand for output is a reduction in productive capacity. What roughly happens is the reduced production of new houses times their reduced prices would be matched by increased prices of export goods times increased production. If the production of new homes falls faster than the production of export goods rises (because there is not enough capacity to expand the export goods so fast,) then the prices of export goods rises faster than the prices of new homes fall. A price index shows higher inflation and lower real output.
While this presumably dampens the decrease in the prices of houses, and there is some inflation, the resulting inflation is not in the prices of houses (they fall.) And so, the debtor is still in difficulty. Now, if the debtor produces goods for export, the boom in that sector would help them. Similarly, construction workers have even greater difficulty paying their home loans.
As the productive capacity for export goods expands, the prices of export goods fall. And as resources from house construction are absorbed into the rest of the economy, costs for producing houses rise as do their prices. But the tendency will be for real output to recover, and inflation to slow again.
8. April 2012 at 06:25
Saturos:
“If the Fed promised to target 5% NGDP, then unless the IOR policy is abandoned, the Fed will have to create more money.”
If the IOR wasn’t abandoned, the Fed could print money until Ron Paul declared revolution, and it still wouldn’t work.
You mean the Fed continuing to inflate the currency cannot EVER result in higher nominal spending?
Nobody would think the Fed was serious about hitting higher NGDP levels as long as it still paid IOR – unless Congress actually wrote it into their mandate (Not this decade).
Then how do you explain the recent increase in NGDP despite IOR being maintained the whole time?
Bernanke’s money is trickling throughout the economy, which is why you have been seeing nominal upticks in various economic indicators.
8. April 2012 at 06:42
“…..there is higher inflation. Creditors who lent when the expectation was a higher growth path of productive capacity receive a lower real return than they expected. Debtors who promised to make payments when productive capacity was expected to grow more quickly, end up paying less than anticipated.”
Exactly, Bill; that is the problem. The whole point of debt is that the creditor does not run such risk and the debtor does (apart from the extreme case of bankruptcy), and debt is supposed to be priced accordingly, relative to other assets. If you don’t like debt-financed house purchase, for example, then perhaps you should argue that house purchase must be financed by something like equity – ie (as far as my limited understanding goes) Islamic finance.
8. April 2012 at 07:04
MF, You said;
“Japan refutes the myth that real growth requires “gradual” price inflation.”
That’s right, the natural rate hypothesis. But at what rate of growth?
David Pearson, I don’t follow your question.
Sauros, I basically agree with your first comment, you describe MM well. My views on Bernanke go back and forth. My biggest criticism is that he’s not more assertive. But I also understand how institutions work, and until we change the views of mainstream economists, it’s unlikely Fed policy will change.
Regarding velocity, suppose each dollar is spent once a week–then velocity is 52 (ignoring intermediate goods, etc.)
Now suppose V stays at 52, but the Fed doubles M. Then the hot potato effect takes over, as currency holding exceed money demand at current levels of NGDP. So NGDP must double to restore equilibrium. But this doesn’t require any speed up in V, indeed V will probably fall temporarily during the transition, as NGDP won’t double overnight (unless there is a currency reform.)
Bill, I agree with most of what you say. Regarding money financed deficits, it depends on expectations. But if, as you say, the cash was directly given out in government spending programs, then expectations of inflation might well rise.
Mike, The Fed sells bonds when the Fed wants to, not when the public wants it to. That was most obvious in 1979-82, but is always true when you look below the surface. If I bring a pile of cash to the Fed and ask for redemption, they are well within their rights to say “No.”
Sauros, Thanks, I appreciate the support.
Bob Murphy, Actually, President Obama put Bernanke in his current position. Bankers aren’t the problem—the problem is macroeconomists.
Rebeleconomist, You said;
“But should they? The mal-redistribution involved, and the exacerbation of moral hazard, which I believe got the US and the UK into their present mess, would be huge.”
You have this exactly backwards. The government created the recession with ultra tight money. This tight money policy greatly worsened the moral hazard problem. I am not calling for high inflation, but rather a stable monetary policy that never changes. I don’t want the Fed to fix problems (which you seem to assume.) I want then to refrain from causing problems.
8. April 2012 at 10:20
MF:
“Nobody would think the Fed was serious about hitting higher NGDP levels as long as it still paid IOR – unless Congress actually wrote it into their mandate (Not this decade).
Then how do you explain the recent increase in NGDP despite IOR being maintained the whole time?”
I meant relative to trend, relative to “the new normal” for NGDP levels. See eg. any Marcus Nunes post.
8. April 2012 at 10:26
“… currency holding exceed money demand at current levels of NGDP. So NGDP must double to restore equilibrium. But this doesn’t require any speed up in V, indeed V will probably fall temporarily during the transition, as NGDP won’t double overnight (unless there is a currency reform.)”
Okay Scott, unless I’m the one who’s horribly confused, I should be able to explain this to you.
Suppose velocity is 1. Now double M. Doesn’t matter how you do it, whether people wake up with twice as many coins or whether the government announces each coin is worth twice as much. Of course PY doubles in equilibrium. But is there ever any disequilibrium? No, the nominal volume of spending is simply doubled. All demand curves double in height; prices would have to double for the same reason they weren’t half as high before.
Now suppose velocity is 2. k is then 0.5. Before the drop, people hold half their receipts on average, and spend the rest. The spending is received as income, and spent again. And all holdings are eventually spent. We have PY = M + M/2 + M/4 + … + M/(2^n), n tends to infinity. The geometric series converges to 2 (setting M = 1). Nominal spending = PY = 2. And money demand = money supply, because people hold 0.5 + 0.25 + … + 1/(2^n) = 1. Now double the money stock, M = 2. And hold V constant. Of course in equilibrium, PY = 4. But when was there any disequilibrium? The M terms were simply twice as large. k is still 0.5. The series of quantities of money demanded is 1 + 0.5 + 0.25 + … + 1/(2^n) = 2, which equals the supply. Because the drop doubled everyone’s nominal income, everyone wants to hold twice as much money as soon as they receive it. So if there was equilibrium before, then there’s equilibrium after, with no disequilibrium in between. This kind of helicopter drop seems exactly like a currency reform to me.
But if the monetary injection for some reason does not immediately double the incomes of those who receive it, perhaps because it displaces other assets, then whilst aggregate nominal income is initially unchanged, and hence aggregate quantity demanded of money, the supply now outstrips it. As the excess is spent, it raises the nominal income of the recipients, increasing their willingness to hold the money. Eventually after n rounds of spending all the money is held and comes to rest.
In a sense, since no one holds all the money they receive, all spending is the result of “excess cash balances”. But so long as kPY in the aggregate equals M, we call it an equilibrium. An event which upsets this, say a rise in M which doesn’t instantly raise PY, gives us a temporary disequilibrium. But the increase in spending (by delta M) is also an increase in income, so PY then dynamically rises until kPY equals M once more.
So if for example the interest rate rises, and k falls from 1/2 to 1/3, but M stays fixed, then V goes from 2 to 3. Although kPY = kMV originally fell from {0.5M(2) = M} to {0.33M(2) = 2M/3 < M} which gives a disequilibrium; because of the increased size of each round of spending and income, V increases by a half, and MV = PY rises by 1/2 eventually. Then {0.33M(3) = M}, and we have equilibrium once more. So a disequilibrium in stocks will increase the size of the flows – but an increase in flows does not necessarily imply any disequilibrium in stocks.
So in your example, I would say: if it's a pure drop, then it's just like a currency reform, and nominal income doubles automatically (because everyone is going to receive twice as much M). This doubling naturally also doubles aggregate money demand, kPY, and there is never any disequilibrium. V is indeed unchanged. Whereas if you're talking about an OMO, where nominal income is not *immediately* affected, then kPY stays at 1 whilst M has jumped to 2. This is a disequilibrium. How does the increase in M affect the circular flow of income? If income rose straightaway, then people hoard twice as much and spend twice as much, and the circular flow and NGDP doubles. But since it doesn't, they spend the whole increase in money. So spending is higher by k (the dollars which would have been hoarded had income rose) times V = 1. NGDP rises to 105, not 104. I think it's wrong to assume NGDP only rises by enough that PY/52 = 2. Instead, since the fraction of *original* receipts hoarded falls from k to k/2 (because the entire injection is spent), overall k falls too, and PY has to rise slightly more to restore equilibrium. Only when k is held constant as a fraction of all receipts does income exactly double to equate demand and supply – i.e., when the original receipts add to income and are hoarded in proportion, which implies a real helicopter drop.
8. April 2012 at 15:14
[…] the leftish post-Keynesians who are upset about how central banks behave. Market monetarists like Scott Sumner and saltwater Keynesians like Paul Krugman constantly lament the bureaucratic caution of the […]
8. April 2012 at 21:06
Scott:
It is irrelevant whether convertibility is at the Fed’s option or at the customer’s option, as long as it happens.
9. April 2012 at 08:17
Rebel Economist:
The only way for a monetary system to provide a stable price level is for a monetary authority to manipulate the quantity of money to offset the effects of changes in productive capacity.
Debt contacts pre-existed monetary authorities seeking to stabilize the price level or inflation rate.
Your notion that debt contracts “should” be free from all risk is mistaken.
For example, debt contracts under the gold standard were not free from risk of changes in the supply or demand for gold. Changes in the supply of other goods, say a bad harvest leading to higher bread prices, did not “automatically” generate a decrease in the quantity of gold so that other prices would fall enough to offset the effects of the higher bread prices, leaving the overall purchasing power of the money and interest unchanged.
Similarly, if productivity grew less than anticipated, the demand for gold would rise less, and so the inflation rate would be higher (or deflation rate lower) resulting in lower than expected real interest rates.
If, instead of a gold standard, the quantity of money is on a fixed growth rate, the results of all of these changes in the same as a gold standard.
If, on the other hand, we have a nominal GDP target, the effects of these changes on debtors and creditors is the same.
Only if we go to a price level target are creditors insulated from changes due to supply shocks in other markets or changes in overall productivity.
What real, evolved credit markets do, is shift risk between the lender and borrower specific to the market. Somone lending money to the farmer bears no risk (short of bankrupcy) due to a bad harvest or reduced productivity by the farmer.
But a real, evolved credit agreement between the a lender and a farmer does not shield the creditor from having to pay higher prices for gasoline if there is a decrease in the supply of gasoline.
Either you are being naive, or you are being too sophisticated and doing imagining an economy with one good.
9. April 2012 at 09:29
I am certainly not capable of being sophisticated, Bill, but I am simplifying the economy to one good – the basket of consumption on which the price index is based, which I am assuming is also the product mix of the economy. I am also assuming that the debtor at least is participating in the production process. Then, if the authorities hold the price level, and, for example, output falls, the debtor loses because he gets to consume less of his output. Is there something wrong with that reasoning?
9. April 2012 at 10:43
ssumner:
“Japan refutes the myth that real growth requires “gradual” price inflation.”
That’s right, the natural rate hypothesis.
But at what rate of growth?
It depends on the innovativeness and productivity of the people.
If investors and entrepreneurs live and work in an economy with gradual price deflation, or zero price inflation, then they will simply change their pricing of current factors of production accordingly. After all, investors take into account price spreads, not aggregate price levels, when making business decisions.
Instead of expecting a price next year of $105, say, they will expect a price of $95, say. As a result, they will price current factors of production for $90 instead of $100 (these particular prices are for illustrative purposes only).
It is certainly not the case that investors, workers, and managers will all of a sudden become more productive in the real sense if they expected a price of $105 next year instead of $95, if they know that inflation will be lower in the case of $95 and higher in the case of $105. In other words, if people come to expect a gradual price deflation over time, if they are born into that system, if they get used to gradually falling prices, then they won’t even consider expecting a price of $105. They will expect a price of $95 because that reflects the rate of money production that leads to falling prices over time.
It would be like asking you if you will work harder next year if instead of making $X and the current rate of price inflation, you will make $2X living with double the rate of price inflation. Of course you don’t even consider double the rate of price inflation, which is why you’re not working at half your capacity knowing what you do about current price inflation and your current income.
Now, I know what you want to believe. You want to believe that gradual price inflation is somehow a secret source for boosting general productivity, because it somewhat coddles people, it allows people to make more nominal income and thus better cushion any unforeseeable “shocks”, and that it acts as a sort of grease in what would otherwise be a dry and brittle engine prone to preventative failures, where instead of some foreign country’s people spending less or more “rigid” money, thus affecting our precious national economy’s incomes here and causing strife in this country, you want nationalist citizens, or what you call “optimal currency area” citizens, to be “protected” from the volatility that world economic competition brings.
Of course if you took that patronizing thought to its logical conclusion, you’d have to support the Fed using its power of inflation to prevent individual US firms from going bankrupt due to the competition from other US firms. But you don’t do that, because you don’t want to use logic. You want to use nationalistic religion to act as a boundary for when good economic logic is allowed to turn into bad logic.
Use inflation to protect Americans from the economic chaos from abroad, but do not use inflation to protect Americans from the economic chaos from other Americans. We are to only be sacrificed to each other, not to foreigners. Why? Because foreigners are yucky and evil and do not deserve to gain such that we lose. They don’t pay US taxes. They don’t use US dollars.
You know, 21st century xenophobia shrouded in a veil of value free nationalistic market monetarism.
9. April 2012 at 10:48
Ahhh, I think I’ve finally seen where I’ve gone wrong. You said, “ignoring intermediate goods, etc.” My analysis is true for MV = PT (k as a fraction of receipts). But for MV = PY, yours is perhaps more accurate. It’s V(T) that rises, but the proportionality parameter v could rise with it. (It’s accruals vs. cash flows again, darn it!). But I still think that a pure helicopter drop is better described without reference to “excess cash balances”, as it doesn’t really involve any more monetary disequilibrium than before.
9. April 2012 at 16:21
Saturos, First of all, that’s way too much information—I completely lost your argument. But I think you are claiming that I believe PY will immediately rise with M. Not so, I agree there is disequilibrium in the short run. My point was that during the disequilibrium V falls. I thought you had claimed V rises during disequilibrium. Obviously if M doubles, and PY doesn’t immediately double, then V falls during the disequilibrium. I think we actually agree, but are talking past each other.
Mike, If it’s at the Fed’s discretion, then it doesn’t always happen. Obviously any currency with non-zero purchasing power can be “converted” into goods at market prices. It can also be converted into assets, like bonds. But the Fed doesn’t promise conversion at any fixed price, as they do under a gold standard.
9. April 2012 at 22:43
Scott,
No, I was actually wrong before, though I thought you had understood both my argument *and* the flaw in it when you posted on the money multiplier, and had slipped in a quick reply to me. But then I read too much. My mistake was in trying to represent MV = PY as a series of flows (not just holding V constant and solving for PY, but actually holding k constant as the marginal propensity to hoard as a fraction of income, and then calculating the geometric series for spending and deriving PY) without remembering that the full flow equation is MV = PT, where T = vY; and then you divide through by v to change transactions velocity into income velocity, which as you say is in fact better described as the “NGDP multiplier”. So an OMO probably would just double income, as that’s what’s required to get the economy to hold double the money.
But as for V falling in the disequilibrium, that’s only true if you insist on using the Equation of Exchange as an identity. But what I see is that as an *equation*, the sides can temporarily not balance, because PY here is *last* period’s income, whereas MV gives you *this* period’s total spending, which determines a higher PY for this period’s income – the injection raises MV this year, but k and V are determined by previous PY, which need not equal current MV … but I’m probably just confused again.
9. April 2012 at 23:01
Sorry, k and V stay constant, but quantity demanded is k times previous income, which gives a temporary disequilibrium. My point was that you don’t need to reduce V just because M increased – instead there is supposed to be a mismatch between *past* income and present spending. As V stays constant, PY rises over the period, from its level at the end of last period (disequilibrium with the new money supply), to its level at the end of this period (equilibrium with the new money supply). But it’s not really important, so forget it.
10. April 2012 at 06:58
Scott:
Bank A stands ready to redeem each of its currency units (dollars) for 1 oz. of silver at the customer’s option.
Bank B stands ready to use its assets to buy back its dollars whenever the market price of a dollar falls below 1 oz.
No difference, right?
Now change the silver peg to a CPI peg with 2% inflation and still, nothing changes.
10. April 2012 at 11:27
Saturos, You really need to treat MV=PY as an identity, otherwise it’s pretty useless.
Mike, I agree that a CPI peg is like a gold standard, money is endogenous. But I thought we were discussing the issue of whether monetary injections are inflationary. Obviously in that case we aren’t assuming any CPI peg, but rather something like the 1970s, where the Fed just let the CPI drift off unanchored.
It also makes a difference whether you are doing level targeting or growth rate targeting. If the Fed is growth rate targeting, they have some discretion.
10. April 2012 at 20:09
Scott:
Start with a normal gold standard. If the government prints $100 and spends it, helicopter style, then we probably agree that the $100 will reflux to the central bank in exchange for gold. Unless it’s Christmas time, in which case the public might use the $100 until after Christmas and then let it reflux.
Now switch to a CPI standard and let the central bank buy back the refluxing dollars with gold, bonds, or whatever. Nothing important changes from the gold standard case.
So when does a helicopter drop cause inflation? Answer: When the central bank is unable (or unwilling) to accept refluxing dollars at the par rate. That would normally be because its assets are not enough to buy back all its dollars at par.
What I take away from this is that the value of the dollar depends on the CPI (or gold) peg, and that peg depends on how much backing the bank has against its money. Conclusion: Backing is what matters. One more reason I like the backing theory better than the QT.
12. April 2012 at 11:44
Mike Sproul, You said;
“So when does a helicopter drop cause inflation? Answer: When the central bank is unable (or unwilling) to accept refluxing dollars at the par rate. That would normally be because its assets are not enough to buy back all its dollars at par.”
Might be, or it might be because Arthur Burns wants to help Richard Nixon get re-elected, so he simply refuses to take the money back, despite the resulting inflation.
I’ve never denied that some inflations are fiscal, indeed most hyperinflation have root causes in bankrupt governments. But that doesn’t explain why prices have risen so much in developed countries since 1965 (when most had modest national debts.) They rose as fast or faster in countries with no national debt (like Australia) as countries with huge debts, like Japan.
13. April 2012 at 10:26
Scott:
The backing theory explanation for inflation since 1965 centers on the peg maintained by the central bank. For example, suppose the dollar had been pegged to gold. If the central bank lowers the gold peg by 3% per year, then there will be 3% inflation even if the central bank’s assets are adequate to maintain the higher peg. (It’s a partial default by the central bank, even though it’s possible that the central bank could have afforded not to default.) Meanwhile, the public, seeing their currency inflating by 3%/year, will increase its nominal cash holdings by 3%/year. (Remember that the central bank passively provides dollars under a gold peg.) Note that quantity theorists will see P and M both rising at 3%/year, and claim that the data supports the quantity theory.
Now change the above story from a gold peg to a CPI peg, and let the central bank use OMO’s in the bond market instead of gold trades, and you have something closer to what really happened.
25. October 2012 at 22:01
[…] argument Davies is making here is a familiar one about the effectiveness of a permanent versus temporary expansion of the monetary base. It is […]