Credit where credit is due
Since I’ve criticized Krugman for his recent posts on monetary history, let me praise him for some recent positive statements he’s made about Milton Friedman. For example:
Friedman “” like Solow, in most of his work “” was in the habit of writing crisp papers with very clear morals. So while you can, on a careful read, see from S-S [Solow-Samuelson] why you should not in fact trust the Phillips curve to be stable, people didn’t actually get that until Friedman and Phelps laid out the point with stark clarity. Credit where credit is due.
But this is what I’d really like to focus on:
What caught me in the Waldmann piece, however, was the brief discussion of the Pigou effect, which supposedly refuted the notion of a liquidity trap. The what effect? Well, Pigou claimed that even if interest rates are up against the zero lower bound, falling prices will be expansionary, because the rising real value of the monetary base will make people wealthier. This is also often taken to mean that expansionary monetary policy also works, because it increases money holdings and thereby increases wealth and hence consumption.
And that’s where I came in (pdf). Looking at Japan in 1998, my gut reaction was similar to those of today’s market monetarists: I was sure that the Bank of Japan could reflate the economy if it were only willing to try. IS-LM said no, but I thought this had to be missing something, basically the Pigou effect: surely if the BoJ just printed enough money, it would burn a hole in peoples’ pockets, and reflation would follow.
But what I did was a little different from what the MMs have done this time around: I set out to prove my instincts right with a little model, a minimal thing that included actual intertemporal decisions instead of using the quasi-static IS-LM framework. [If you have no idea what I’m talking about, you have only yourself to blame — I warned you in the headline]. And to my considerable surprise, the model told me the opposite of my preconception: there was no Pigou effect. Consumption was tied down in the current period by the Euler equation, so if you couldn’t move the real interest rate, nothing happened.
One way to say this “” which Waldmann sort of says “” is that even a helicopter drop of money has no effect in a world of Ricardian equivalence, since you know that the government will eventually have to tax the windfall away.
Let’s review a few facts:
1. I published an article claiming temporary currency injections have almost no effect on the price level 5 years before Krugman did. His 1998 article is 100 times better, but it’s not like we market monetarists don’t get his point.
2. I have several blog posts pointing out that (for Ricardian reasons) even “helicopter drops” don’t really solve the problem of liquidity traps, at least if you believe the liquidity trap is a problem. (I don’t.) I’ve cited Japan as an example. So (in my case) he’s preaching to the converted in the last paragraph. I believe David Beckworth disagrees with me on this point.
3. I don’t rely at all on the Pigou effect, and as far as I know other MMs don’t either. I rely on the expected hot potato effect, plus asset price changes. Because money’s expected to be neutral in the long run, a injection that’s expected to be permanent will raise future expected NGDP, which is expansionary even in a Woodfordian model. And no, it’s not all mysterious hand-waving, unless you believe the trillions of dollars being earned in the Japanese stock market this year is monopoly money.
4. In the past, Japan did not try to reflate during those episodes where Krugman claimed they did. I think independent observers who have looked at the issue now agree I was right and Krugman was wrong.
5. Point 4 was strengthened when the BOJ was pushed into a 1% inflation target a couple years ago, and then was dragged kicking and screaming into a 2% target more recently. No one with a deep understanding of Japan seriously believes they were trying to create inflation back in 2006 when inflation was zero and they raised interest rates and cut the monetary base by 20%. Thus proving the point of my 1993 paper that temporary currency injections are not inflationary.
6. If rates are zero and expected to be zero forever then Krugman’s right. But in that case buy up the entire national debt with currency, and start in on the debts of other countries too. Call that fiscal policy if you like; it’s still monetary policy to me.
I think we differ because we envision QE much differently. He sees it as a lever that might fail because it might not credibly change expectations of future monetary policy. I see the policy itself as NGDPLT. You credibly promise to do whatever it takes. Then you figure out how much base money the public and banks want to hold at that expected NGDP growth rate, and do enough QE so that the market for base money is in equilibrium (plus cut IOR to zero.). If you don’t know how, create an NGDP futures market.
NGDPLT is the policy, QE is the tool to make it work. If you don’t want 5% NGDP growth (and the tapering talk strongly suggests the Fed does not) then there’s no point in even doing QE.
PS. I can see how someone might have assumed I relied on the Pigou effect in the past, as I’ve argued that QE probably has a small effect above and beyond the expectations channel. Cash and bonds are not perfect substitutes, even at zero rates. But I’ve never emphasized this channel, and in any case it’s based on a (very weak) hot potato effect, not the Pigou effect.
PPS. Let me end on a positive note. I do believe the 1960s monetarists would have benefited from absorbing the message in Krugman’s 1998 paper. But I also think it unfortunate that (without saying so) his writing style sometimes leaves his readers with the impression that this 1998 paper taught him that monetary policy was ineffective at the zero bound. Krugman continued to favor monetary stimulus (in Japan) over fiscal stimulus in the late 1990s, and changed his view a few years later.
PPPS. Krugman ends as follows:
. . . modern Friedmanites are a very small group with no real constituency.
Maybe, but we’re coming on strong.
HT: Edward
Tags:
11. August 2013 at 12:58
I don’t know how Krugman can read this and continue to disagree with you. Let’s hope he does.
11. August 2013 at 14:12
The real balance effect is that a lower price level results in more real expenditure on output.
The lower price level increases real money balances.
One effect that those with higher real balances use them to purchase financial assets. The higher prices of the financial assets and lower yields motivate people to sell existing or new financial assets and purchase currently produced goods and services. In a simple model, this amounts to higher bond prices, lower nominal and real interest rates, and more investment and consumption.
The Pigou effect is another aspect of the real balance effect. The increase in real balances is an increase in real wealth. Wealthier households spending more on consumer goods and services.
From a Wicksellian perspective, the first effect pushes the market interest rate down towards the natural interest rate. The second effect, the Pigou effect, involves a decrease in saving supply, and so an increase in the natural interest rate to towards the market interest rate.
Open market operations don’t create a Pigou effect because they substitute one asset for another.
Helicopter drops do create a Pigou effect.
In a simple model where the only asset is one type of bond, then the first effect can’t work once the market rate has been pushed down to zero. However, in the real world, where there are a variety of assets, and the yields have not all been driven down to zero, then the real balance effect can push down the yields of these other assets.
With open market operations, it is necessary to purchase these other assets.
The Pigou effect doesn’t work with inside money. This is money issued by banks. It is an asset to those holding it but a liability to the issuers. A lower price level makes those holding money wealthier, but it makes those issuing it poorer. Still, there is a real balance effect from money to other assets.
Since market monetarists favor a target growth path for nominal GDP, no part of the real balance effect is of much interest, including the Pigou effect.
From our perspective, the issue is solely the effectiveness of open market purchases. In my view, if nominal GDP is on target and expected to remain on target, then purchasing assets with a yield less than the yield paid on reserves is pointless. The yield on reserves can be reduced to slightly less than zero while maintaining redeemability with government issued hand-to-hand currency. Purchasing more of such assets with yields that low does no good. Other assets, with higher yields need to be purchased.
That is, if you want to keep to the target. To me, that is the right thing to do.
11. August 2013 at 14:16
If by “inflationary” it is meant rising prices from what they otherwise would have been, then even temporary monetary inflation is inflationary. The data from 2006 Japan does not “prove” nor “disprove” this fact. This fact is not capable of being proven by observable data, since the counter-factual world from which the price inflation in the factual world is comparably higher, is itself unobservable.
In other words, observable economic data can only tell us what “is”. It cannot tell us in what way this “is” is different from what could have been an “is” had different economic actions taken place instead.
The 2006 Japan data only shows the prices that prevailed given the actions that took place. We cannot reason from this price change that monetary inflation is therefore not always “inflationary.”
What we can know about 2006 Japan is that the prices that prevailed with the temporary monetary inflation, are higher than they otherwise would have been had there been no monetary inflation, or less monetary inflation. In other words, if the BOJ did not temporarily inflate, then prices would have otherwise been lower, even if they did not rise, or even fell, in temporal terms.
This knowledge derives from what we know of marginal utility. Ceteris paribus, an additional unit of a homogeneous good (in this case money), will be given a lower value, than would be given to that homogeneous supply without that additional unit. This manifests in more dollars being exchanged against the same supply of real goods. It is like sheep falling in value as their supply increases, which manifests as more sheep traded against cows.
It is literally impossible for the supply of money to increase from what it otherwise would have been, even temporarily, without there being a new valuation of that higher supply, and lower valuation per unit of good (dollar), which manifests in more dollars traded against the same supply of goods.
This is universally true. It is true even if prices fall over time in the observable sense. What is true for the scenario of temporal deflation despite temporary monetary inflation, e.g. 2006 Japan, is that prices in the counter-factual world of no or less temporary monetary inflation, would have otherwise been lower.
The law of marginal utility as it applies to money and real goods is not proven by observing prices along with changes in the money supply. It is proven by what we can know of economic action via self-reflective activity of action itself.
11. August 2013 at 14:20
Bill, good points – I’ve argued along different lines that the real balances effect shouldn’t be discarded even under Ricardian Equivalence here: http://ashokarao.com/2013/08/11/an-inquiry-into-the-efficacy-of-ricardos-helicopter/
When he says “modern Friedmanites” I don’t imagine he’s speaking of MMs, but perhaps who knows.
11. August 2013 at 14:31
Even if commercial banks immediately “hoard” every dollar of new money created by the central bank, prices would still be higher than they otherwise would have been, namely, the prices of the assets the central bank buys. This is true even if right after or as soon as the central bank inflates, the observable price of the asset the central bank falls from one moment in time to the next, such as the price of a bond. In this case, the bond price would have otherwise been lower. So instead of observing a price trend of 100 to 99 to 98, etc, we would have observed 100 to 98 to 97, etc.
A problem arises when one reasons from a price change, instead of from economic law. Reasoning from a price change would lead one to make a causal argument that because the central bank inflated, they caused the price trend to go from 100 to 99 to 98 with the monetary inflation, instead of, say, 100 to 100 to 99 without the monetary inflation.
A market monetarist might then falsely conclude that “prices of bonds fall when the central bank adds to monetary inflation.” Then when the opposite event occurs, say after bond prices fall after a central bank accounces it will “taper”, the market monetarist is stuck in having to explain that “it isn’t always true”, or that “most of the time” it’s true, or that it is “generally true.”
For economists on the other hand, there is do difficulty at all in explaining the data. If bond prices rise, fall, or stay the same, any and all outcomes directly follow from the economic law of marginal utility and counter-factual analysis.
11. August 2013 at 14:49
Saturos, If he’d been reading me over the years he wouldn’t have done all those other posts that misrepresented MM. I think he’s too busy to follow this stuff closely.
Bill. Good post.
Ashok, If MMs are not modern Friedmanites, who are? He’s said in the past that I’m in some ways the heir to Milton Friedman, so I’m pretty sure he’s talking about MMs.
11. August 2013 at 14:51
I remember a few years ago Krugman stated the Fed should adopt a 5% inflation target until full employment returns. How does that not make him even more aggressive towards monetary expansion and stimulus than the “modern Friedmanites”?
12. August 2013 at 00:00
Central banks are also banks.
But Friedman didn’t think so, so he could believe in both Ricardian equivalence and Pigou without contradiction, if he wanted to.
12. August 2013 at 01:21
Scott,
“NGDPLT is the policy, QE is the tool to make it work.
I rely on the expected hot potato effect, plus asset price changes.”
Much better! B+
But could be improved.
1. You need to stop referring to QE. It’s all the same policy TOOL…OMO…. whether you call it setting the funds rates, QE, TARP, TWIST, or anything else. You just alienate the no-nothing right wing Austrophilliacs who think QE is some kind of hocus-pocus. Call it OMO.
2. You’re getting close on the mechanism, but drop HPE. It’s predicatively accurate, but it’s not causal …and your skeptics don’t buy it (the “pushing on a string” crowd). You agree Bill Gates doesn’t buy a Ferrari because he has more money. People hold the amount of money they need. At the other side of OMO are economic players who have been induced by the the higher price of financial assets to exchange financial assets for goods and services (marginal increase in AD). If not the Fed couldn’t execute the trade.
3. Instead of railing about and having to explain IOR, start talking about MB – ER. It will further simplify your argument.
12. August 2013 at 03:56
Scott,
what do you make of today’s Japan GDP numbers – real GDP slowed more than expected but NGDP picked up from 0.7% or so to 2.9%. I think that’s the “real” news as headlines focus on the RGDP.
Consumption and exports are up but investment is down (and I don’t expect that to pick up before there is clarity on the forthcoming tax regime). But so far, so good.
Best M
12. August 2013 at 04:24
Correction: my mistake. Japan NGDP up 0.7% y/y from 0.6% y/y… got overexcited and confused annual with y/y data. Apologies.
M
12. August 2013 at 04:47
Does anyone have the name of/a link to the 1993 paper on temporary currency injections that Professor Sumner mentions in this post?
12. August 2013 at 05:12
Brad DeLong on his personal conversations with Milton Friedman:
“Why Did Milton Friedman Think a Modern Economy Needed Heavy-Handed Government Regulation in the Liquidity Services Industry and Nowhere Else?”
http://delong.typepad.com/sdj/2013/08/why-did-milton-friedman-think-a-modern-economy-needed-heavy-handed-government-regulation-in-the-liquidity-services-industry-a.html
12. August 2013 at 05:16
Nick Rowe:
“The strong monetarist dark force of eventual macroeconomic self-equilibration”
http://worthwhile.typepad.com/worthwhile_canadian_initi/2013/08/the-strong-monetarist-dark-force-of-macroeconomic-self-equilibration.html
12. August 2013 at 05:18
Betting markets:
Odds of Taper in September 2013 Soar
http://www.zerohedge.com/news/2013-08-11/sept-taper-odds-soar
12. August 2013 at 05:41
Josh Brown:
“Taper’s a lock, folks. And it couldn’t come soon enough.”
http://www.thereformedbroker.com/2013/08/12/oy-vey-2
12. August 2013 at 06:30
wufwugy, Good question.
dtoh, NGDP rises because of the HPE. Bill Gates buys the Ferrari because wages are sticky.
Thanks Mikio.
Thanks for the links Travis. “Regulation” is probably not the right term, but I get DeLong’s point. Sticky wages and prices is the answer.
12. August 2013 at 06:40
Prof. Sumner,
Why is it wrong to view sticky wages and prices as a “market failure” problem and stable NGDP growth an effective government intervention solution to the problem?
12. August 2013 at 07:23
Prof. Sumner,
Emergency! Rand Paul just wrote an op-ed in National Review that really really really really mischaracterizes Friedman (found it via Pethokoukis).
http://www.nationalreview.com/article/355500/milton-friedman-and-restraint-rand-paul
Market Monetarists NEED to respond to this historical argument about clearinghouse associations, etc.
12. August 2013 at 07:39
I struggle with how much of this is semantics, and how much real. Krugman would need to look at global asset markets, and argue that a) markets do not react to monetary action OR b) markets do not influence AD. Empirically, this is a hard argument.
Theoretically, Krugman’s problem is that he is still operating from a comparative statics perspective. In this case it is possible that (theoretically) he’s correct if we do not allow dynamic response. Let’s say the Fed drops $500 billion in QE and says “we’re done – no more, ever – we sure hope that’s enough to get this thing started!”. Let’s say the market responds with hundreds of talking heads yelling “that’s it! the Fed’s out of ammo!” and people start to believe it because the Fed refuses to disagree (because it’s worried about a political response). In that case, Krugman could be right – there might be a moderate hot potato effect, but the increase in wealth would need to be large enough to (at one fell swoop) obliterate the debt overhang problem that’s causing a balance sheet recession. Otherwise, the effects would wear off rather quickly.
The difference is if the Fed says – or the markets believe that the Fed means (quietly or not) – “And if that ain’t enough, there’s plenty more where that came from!”
The difference is night and day. If the Fed does not say this, then it’s not targeting NGDP, and we sure hope that the one-time-boost in money supply (which Krugman might agree is permanent) has ENOUGH of a hot potato effect to get the economy into “escape velocity”. There’s no guarantee, and if people THINK it won’t have enough punch to do the trick, then it probably won’t do the trick.
If the Fed says “and there’s more where that came from”, then if there’s ANY HOT POTATO / WEALTH EFFECT AT ALL (even small), and the threat is credible, then that’s enough. You can show this in a model, and in this dynamic framework Krugman would need to argue there is ZERO WEALTH EFFECT for monetary policy targeting NGDP to fail. ZERO, because if it’s at all non-zero, then the Chuck Norris Theorem kicks in (so to speak).
In terms of no audience, was Krugman paying attention to Jackson Hole for the last couple years? Or, perhaps, he thinks Goldman Sachs created the idea of NGDP targeting all by itself?
BTW, nice tone to your post – written from a position of confidence and graciousness.
12. August 2013 at 07:47
How about Rand Paul’s take on Friedman?
http://www.nationalreview.com/article/355500/milton-friedman-and-restraint-rand-paul
12. August 2013 at 08:17
Wow. That Rand Paul op-ed is really something.
12. August 2013 at 08:33
Krugman used to write ‘crisp papers with very clear morals’ too, as in 1998;
http://web.mit.edu/krugman/www/japtrap.html
——-quote——
Because the traditional IS-LM framework is a static one, it cannot make any distinction between temporary and permanent policy changes. And partly as a result, it seems to indicate that a liquidity trap is something that can last indefinitely. But the framework here, rudimentary as it is, suggests a quite different view. In the flexible-price version of the model, even when money and bonds turn out to be perfect substitutes in period 1, money is still neutral – that is, an equiproportional increase in the money supply in all periods will still raise prices in the same proportion.
So what would a permanent increase in the money supply do in the case where prices are predetermined in period 1? Even if the economy is in a liquidity trap in the sense that the nominal interest rate is stuck at zero, the monetary expansion would raise the expected future price level P*, and hence reduce the real interest rate. A permanent as opposed to temporary monetary expansion would, in other words, be effective – because it would cause expectations of inflation.
Let us now bring this discussion back to earth, and to Japan in particular. Of course the Bank of Japan does not announce whether its changes in the monetary base are permanent or temporary. But we may argue that private actors view its actions as temporary, because they believe that the central bank is committed to price stability as a long-run goal. And that is why monetary policy is ineffective! Japan has been unable to get its economy moving precisely because the market regards the central bank as being responsible, and expects it to rein in the money supply if the price level starts to rise.
The way to make monetary policy effective, then, is for the central bank to credibly promise to be irresponsible – to make a persuasive case that it will permit inflation to occur, thereby producing the negative real interest rates the economy needs.
This sounds funny as well as perverse. Bear in mind, however, that the basic premise – that even a zero nominal interest rate is not enough to produce sufficient aggregate demand – is not hypothetical: it is a simple fact about Japan right now. Unless one can make a convincing case that structural reform or fiscal expansion will provide the necessary demand, the only way to expand the economy is to reduce the real interest rate; and the only way to do that is to create expectations of inflation.
——-endquote——-
12. August 2013 at 08:36
Aidan,
Seems to me that Rand Paul is being very tricky. Limiting the Friedman discussion to small-picture stuff.
Big-picture bottom line:
The Austrian position: tight money can never be bad
The Friedman position: tight money can result in gigantic disaster in some cases. And such gigantic disasters typically promote rather than discourage “big dysfunctional government.”
12. August 2013 at 08:54
It’s funny how much disagreement there is even over whether CBs are trying to inflate at a given point in time.
Everyone on the econ right thinks QE is inflationary and I have to keep explaining to them that the long-term expectations are more important — and there hasn’t been any spike in inflation!
12. August 2013 at 09:21
“The Friedman position: tight money can result in gigantic disaster in some cases. And such gigantic disasters typically promote rather than discourage “big dysfunctional government.”
This is what the Austrian nitwits will never understand. They’re to libertarians like Ralph Nader is to the Democrats.
12. August 2013 at 09:24
And the Austrians and Rand Paul miss a fantastic point. Clearinghouse associations are a private market way of creating EASIER money!
12. August 2013 at 09:32
Just a thought – you really should define the hot potato effect. You reference it a lot, and the notion you use seems to be “there’s some amount of money that people don’t like to hold”, and the rationale is “because they don’t”.
I think you really have some more logic underneath this you need to share, and I suspect it has something to do with intertemporal consumption. Money, a store of wealth, is really future consumption. A helicopter drop has three effects. The primary effect is that they increase current consumption because there’s diminishing marginal utility of consumption in each time period. The notion of diminishing marginal utility of consumption is indisputable, and yet it needs to be reconciled with a zero or negative inflation rate. All a zero rate means is that, GIVEN THE CURRENT LEVEL OF MONEY AND CONSUMPTION GOOD, people would rather consume more in the future. But if you increase the overall level of money permanently, it’s entirely consistent for the rate to remain zero or negative, and current consumption to increase (just not as much as future consumption). At a few hundred billion dollars in a 17 trillion dollar economy, this is small. The secondary effect is that if everyone does this, the value of money declines both immediately AND in the future, assuming non-infinite supply (cue SRAS / LRAS argument), and the rate changes, which shifts consumption from future to present (changing the real rate). The tertiary effect is expectations – EVEN IF the secondary effect is negative due to infinite SRAS (which Krugman might argue), the PRIMARY effect (which you call hot potato) still exists. Since this exists, the tertiary effect of a credible threat can dominate the secondary effect so long as the primary effect is non-negative, and so the tertiary effect will create the expectation of inflation in the short term even if SRAS is infinite unless LRAS is also infinite – that because of the intertemporal tradeoff (if value of money is lower in the future, I’ll spend a bit more today).
Krugman’s all about the secondary effect (rate), and then argues that if SRAS is infinite, then the tertiary effect has no leverage, and therefore even a credible threat creates no response – people just hold money.
The fundamental question to ask Krugman is this:
Do you or do you not believe that there is diminishing marginal return to consumption in all time periods in an intertemporal model?
To be clear, Scott, the three assumptions necessary for your view of the world to work are:
1) Diminishing marginal return to consumption in a multiperiod model
2) LRAS is not infinite
3) A credible promise/target can be made
This can all be laid out mathematically, without even requiring a full macro model
12. August 2013 at 09:45
SG: I think the easiest way to think about HPE is in terms of changing expectations — to the extent you expect the money to be worth less, you want to spend it sooner. Here’s a paper.
http://citeseerx.ist.psu.edu/viewdoc/summary?doi=10.1.1.173.4685
12. August 2013 at 10:30
TD, I don’t think this is the same hot potato effect Scott is describing.
“Intuitively, as a tax on monetary exchange, inflation reduces the return to this activity; when the return falls, agents invest less; and this means in the models that buyers search less and end up spending their money more slowly.”
The authors get around this by introducing free market entry…
The “hot potato” effect they discuss is purely a single time period phenomenon. They and others concoct crazy models to justify why people should want to hold onto goods instead of money when the value of money is falling, because they don’t want to allow for things like sticky prices, non-indivisible goods, or discretized time periods (like, oh, the concept of a day).
Moreover, the “hot potato” effect they are discussing is a result of inflation, not necessarily a cause. They’re trying to figure out how it could possibly be that – if people think there will be inflation – the macroeconomy can see an increase in velocity. Indeed, their model is one purely of trade (not consumption).
That’s not Krugman’s problem – he’s not saying expected inflation won’t increase velocity. He’s saying that in the absence of expected inflation, giving people money doesn’t create inflation or expected inflation. They just hoard it, even if it’s permanent.
If the hot potato effect is that paper’s view, then Scott should stop using that term, and we need another term to describe diminishing marginal return to within period consumption.
12. August 2013 at 10:44
Travis, I think the existence of government fiat money is the intervention. I don’t see NGDPLT as more or less interventionist than other government run fiat money policy rules.
And let someone else respond to Rand Paul – – I’m too busy.
Thanks Statsguy, That analysis is so reasonable I don’t see how Krugman would disagree.
Patrick, Thanks, I’ll use that.
Statsguy, For your second comment, all you need is the first paragraph that Patrick quotes from Krugman 1998.
I think of the HPE this way. Start with a simple money and goods economy, nothing else. Then more money drives up goods prices very quickly, as that’s the only way for the public to get rid of excess cash balances. Now add financial assets. Now a monetary injection might not immediately push up goods prices, instead asset prices rise enough so that people are willing to hold the excess cash balances. But money is still expected to be neutral in the long run via the HPE because no one expects money to affect real asset prices forever. So expectations of future NGDP growth lead to more AD today.
12. August 2013 at 12:23
Scott, your write:
“But I’ve never emphasized this channel, and in any case it’s based on a (very weak) hot potato effect, not the Pigou effect.”
How many “hot potato” effects are there? Are there others that are less “weak?”
12. August 2013 at 13:04
@dtoh, you write:
“3. Instead of railing about and having to explain IOR, start talking about MB – ER. It will further simplify your argument.”
MB – ER = RR + physical cash in circulation, correct?
12. August 2013 at 15:36
Scott, you said;
dtoh, NGDP rises because of the HPE. Bill Gates buys the Ferrari because wages are sticky.
No Bill Gates DOESN’T buy the Ferarari specifically because wages are sticky. If they weren’t sticky, the price of the Ferrari would drop and then he WOULD buy it.
Here’s an extra credit problem. Write a short post on why the fact that wages are sticky whereas as financial asset prices are not sticky, a) causes monetary policy to be both necessary and effective, and b) invalidates Wallace.
12. August 2013 at 16:32
[…] Paul Krugman knew all this years ago; Patrick just sent me the following quote from […]
12. August 2013 at 16:36
Scott, you said;
“I think of the HPE this way. Start with a simple money and goods economy, nothing else. Then more money drives up goods prices very quickly, as that’s the only way for the public to get rid of excess cash balances. Now add financial assets. Now a monetary injection might not immediately push up goods prices, instead asset prices rise enough so that people are willing to hold the excess cash balances. But money is still expected to be neutral in the long run via the HPE because no one expects money to affect real asset prices forever. So expectations of future NGDP growth lead to more AD today.”
You’re getting close. Your analysis of a money and goods only economy is correct. You get into trouble when you add in financial assets.
WILLING to hold cash balances is not sufficient. You have to have an ultimate counter-party to OMO who WANTS the cash. Otherwise, as long as there is a possibility of holding ER, OMO will just result in a rise of ER.
The way an OMO works is that it raises the real price of financial assets. This causes a marginal increase in the exchange of financial assets for real goods and services (consumption and investment), and this increased exchange requires money to effect the transaction so people WANT the cash.
(And yes, theoretically, if there were helicopter drops or if nominal rates on financial assets were negative then you would be right about HPE.).
12. August 2013 at 16:57
dtoh, yes yes and yes!!!
I’m fully with you especially when you say.. “it’s not causal …and your skeptics don’t buy it”. I generally agree with most of what Scott says but HPE is a a MAJOR turn off..
Especially given that his arguments generally fall back on vague empirical observations about Japan (which, given the difficulty of conducting a controlled macroeconomic experiment, aren’t particularly convincing) or how “laughable” it is to think that cbankers would have trouble inflating a fiat currency. I don’t find that laughable at all, I continue to believe that money/reserves/bonds are perfect substitutes at the zero bound and I have yet to hear a convincing argument from Scott on that.
Scott,
We’ve already debated this extensively, but there’s one more thing I’d be curious to ask you:
Assume the Fed were to credibly commit to an NGDP level target path, and aim to achieve it by “targeting the forecast” using Fedfunds as their tool (no futures).
I agree that hitting the zero bound would be less likely than in the current regime, but if we did hit the zero bound, would you still advocate:
(a) Growing the base beyond what’s needed to push Fedfunds down against IOR (i.e.: “QE”), or
(b) Let the expectations do their job and simply keep the base at a level where interest rates are 0% until the NGDPL forecast exceeds the target
And of course: why?
13. August 2013 at 05:01
Tom, There is only one, but it’s weaker at zero rates.
dtoh, If prices and wages were flexible then only prices would respond to monetary stimulus, and no more Ferraris would be sold.
You said:
“WILLING to hold cash balances is not sufficient. You have to have an ultimate counter-party to OMO who WANTS the cash. Otherwise, as long as there is a possibility of holding ER, OMO will just result in a rise of ER.”
Why would banks want to hold ERs if rates are positive? And if rates are zero, why would real asset prices rise?
DOB, So you seriously believe that BOJ could not depreciate the yen even if they wanted to? Just do it!! And BTW, they just did it, which proves you wrong.
On your question, I’d do QE so that there was less unemployment (compared to waiting, which might never work.)
Of course neither policy is optimal. If long term rates are zero under NGDPLT, then your target path is too low. Raise the target to 4% or 5%.
13. August 2013 at 19:53
Scott,
“So you seriously believe that BOJ could not depreciate the yen even if they wanted to?” ->
I believe CBs can do some things to devalue, but that negative nominal interest rates are the silver bullet.
I dislike asset purchases because they load up risk on the taxpayer’s balance sheet (the ultimate equity owner in the central bank) and the volumes needed to achieve anything meaningful can grow out of bounds very quickly because the signaling and supply/demand effects are quite small. I agree with you that they’re better than nothing if negative interest rates aren’t going to be implemented, but QE is strictly dominated by an interest rate commitment such as “we keep rate at 0% until NGDP forecast hits target path” (you disagree with that last part since you’ve said you’d still have QE on top).
I think we both believe in the supply/demand and the signaling channels in QE. However you seem to attribute importance to the increase in M0 because of this HPE thing. My belief is that if tomorrow the Fed drained about ~ $2tn of reserves (as much as they can while still keeping Fedfunds unchanged) by executing overnight repos on some of its inventory, then it would have absolutely no impact on the price level. I think you disagree because of HPE. Right?
“If long term rates are zero under NGDPLT, then your target path is too low. Raise the target to 4% or 5%.” ->
Unless otherwise specified, I mean short term rates when talking about the zero bound (I think most people do?). Raising the target just mean you hit the zero bound less often, but if the equilibrium overnight risk-free real rate swings low enough, you’ll still hit it once in a while so your system needs to be able to weather it.
13. August 2013 at 23:46
Scott, I’m a little confused. You’ve said there’s “just one HPE” when I asked, and you’ve said the HPE is “very weak” but only at “zero rates.”
http://www.themoneyillusion.com/?p=22948#comment-266884
So we’re at “zero rates” where we can expect the HPE to be “very weak.” Well, if that’s the case, then what’s left? The expectations channel, correct? So why not just have the Fed let the market know that it will provide any and all base money’s required: Fed deposits or paper reserve notes and coins. Why does the base money have to come BEFORE it’s needed? (i.e. excess reserves resulting form QE). This ties back to what David Beckworth wrote when pressed about the timing of when base money should be provided:
“My point is that the CB commits to a permanent increase in the monetary base. How it chronologically unfolds is besides the point. Yes, in an interest rate targeting regime the CB accommodates demand.”
“I have never argued (or meant to argue) that the monetary base has to increase first.”
http://macromarketmusings.blogspot.com/2013/08/a-permanent-expansion-of-monetary-base.html?showComment=1375901514604#c7523207141783165074
OK, so if you agree with David here, and HPE is “very weak” when we’re at “zero rates” then why not ditch the QE and the ER > 0, and just promise that base money will be made available when needed. And then… if we do that, how is that different than what we have been doing prior to 2008? Except that now there’s an explicit NGDPLT?
14. August 2013 at 00:21
“Why would banks want to hold ERs if rates are positive? And if rates are zero, why would real asset prices rise?”
Seems to me that banks hold a lot ERs. I guess because IOR is higher than the market rates on risk weighted basis. I think the point here is how much the economy is willing leverage. It is deleveraging and thus the rates are zero and there is (almost) no room for monetary stimulus. The helicopter drop is a fiscal tool as it is equivalent of lowering taxes.
14. August 2013 at 06:52
DOB, If they drained $2 trillion tomorrow then prices would fall.
Asset purchases don’t cause taxpayer risk, as they are buying government debt, which is a taxpayer liability.
Tom, I presume you are new here. I’ve been fighting this battle for 5 years. What’s left is for the Fed to set an NGDP target! Many commenters talk as if the Fed is trying to stimulate but just can’t get the hang of it. They are tapering! They don’t want more stimulus! What’s left is for the Fed to change it’s mind. Then QE works fine.
And I certainly agree with your last point. In a sensible system the base money is endogenous. You set the NGDP target, and the public tells you how much base money they want to hold. I’m all for that. But we don’t have a sensible system, the Fed uses QE to signal its target. That’s why it’s such a mess.
Jussi, You might want to look at my FAQs in the right column first, to get up to speed here. Leverage has no bearing on the effectiveness of monetary policy.
14. August 2013 at 09:16
Scott, THANK YOU! … That clears up a lot of confusion for me about where you stand. I wonder if David Beckworth would agree? I assumed that’s what he believes, but he moved on to other posts and I wasn’t able to confirm that.
So it sounds like you have a more fundamental issue with “dtoh” than just how to present the material. It sounds like dtoh is saying that HPE (but using his words for this: raising asset prices in general so investors exchange in to real goods and services) is a stronger channel than you are.
OK, so have I jumped the shark now? 😉
14. August 2013 at 10:08
Tom/Scott,
Sorry but Scott isn’t being consistent here:
(1) “In a sensible system the base money is endogenous. You set the NGDP target, and the public tells you how much base money they want to hold. I’m all for that. But we don’t have a sensible system, the Fed uses QE to signal its target. That’s why it’s such a mess.”
(2) “On your question, I’d do QE so that there was less unemployment (compared to waiting, which might never work.)”
These two conflict. The second one was precisely the answer to the same question i.e. if you had an explicit NGDP target, would you still increase base beyond the liquidity demanded at the zero bound, Scott’s answer was Yes. In (1) he says no..
On a different but related topic:
“DOB, If they drained $2 trillion tomorrow then prices would fall.”
Could you please explain the exact transmission mechanism? Yesterday, you/I/the banks/the private sector had “money”, today, the Fed replaced it by an asset who’s price is 1 but that’s not money (a 1-day repo). That asset can be turned into money in the market at 0% rate. How is *anyone’s* decision making affected by that transaction? And therefore how is the price level impacted by this?
“Asset purchases don’t cause taxpayer risk, as they are buying government debt, which is a taxpayer liability.”
Depends: when the SNB buys up EUR assets, that’s unnecessary risk to the taxpayer.
When the Fed buys (long term) treasuries, the interest rate risk is now born by the taxpayer (assuming the treasury department was managing its asset liability match properly, which it isn’t but that’s an entirely different story). The credit risk, as you point out, nets down but then situations where the fiscal revenues are insufficient to cover debt servicing costs get even uglier and become inflationary: the CB looses any ability to target NGDP or anything else at that point. I find it strange that you’d advocate for that.
14. August 2013 at 14:59
Scott, you’re going to HATE this (I have a feeling you don’t like looking at these kinds of things). But I have a nice tidy “proof” here that banks definitely don’t lend out reserves (I’ve included “cash” in my analysis). It’s a simplified world… blah blah… it’s all explained in the post. The “proof” is offered as an example at the bottom:
http://brown-blog-5.blogspot.com/2013/08/banking-example-11-all-possible-balance.html
If there’s a mistake, (somebody) let me know!
14. August 2013 at 16:37
And BTW, I realize that my example there had reserve requirements at 0%, which is why I phrased it as “loan out reserves.” This doesn’t matter terribly because as L increases then < 10% of this change in L in ER would convert to RR (or vice versa should L decrease) but overall reserve levels would not change. < 10% because some will not keep demand deposits but instead open savings deposits.
In any event, does that make a big difference?
14. August 2013 at 17:08
Tom,
“But I have a nice tidy “proof” here that banks definitely don’t lend out reserves”
Obviously, banks can only lend reserve to each other or to someone that has an account at the Fed.. No one else can hold reserves.
14. August 2013 at 17:34
Tom and DOB, I think there is confusion over semantics. Yes, once reserves leave the banking system it is called cash. What I meant is that banks can reduce their reserve holding either individually or in aggregate. Indeed if every bank shut down the base would immediately become 100% cash. That’s all I meant.
And sorry Tom, but I rarely have time to read other people’s blog posts. Bring your idea over here, and make it very short. If you can’t summarize it in one paragraph, that’s means you need to re-work the model.
Tom, I don’t know if you jumped the shark because I have no idea what that phrase means.
DOB, I don’t even know what it means to increase the base beyond the demand at the zero bound. The supply and demand for base money will be in equilibrium wherever you set the base. And I see no conflict between the two statements.
There are many transmission mechanisms for QE. When the Fed does QE the most noticeable effect is higher asset prices. That means if the reverse QE asset prices will fall. The fall in commodity prices and the fall in the price of foreign exchange tends to reduce inflation. There are other mechanisms as well, such as those described by Woodford (lower expected future AD.)
I agree about SNB risk, but it’s only “unnecessary” if there are better alternatives. There may be (higher NGDP targets, for instance) it’s a judgment call.
I don’t follow your last paragraph at all. The policy I advocate (stimulus) would dramatically reduce taxpayer fiscal risk.
14. August 2013 at 18:21
Scott, OK, assume no GSEs or foreign trade, ER > 0, and reserve requirements = 0%, and that Tsy spends every $ it gets.
F = Tsy debt held by Fed
B = Tsy debt held by banks
L = bank loans to non-banks
C = cash in circulation
Then w/o showing the rest of the accounting here:
bank reserves = F-C
non-bank deposits = L+B+F-C
non-bank stock of money
= bank deposits + cash
= L+B+F-C+C = L+B+F
So to make bank reserves decrease either F goes down or C goes up, but the non-bank’s stock of money is independent of C and it clearly goes down if F goes down. Thus, bank reserves decreasing means the non-bank’s stock of money either stays even or decreases. How could you call that “loaning out reserves?”
(The simplified “accounting” for this is shown in a balance sheet format in my post)
14. August 2013 at 18:47
@DOB, you write
“Obviously, banks can only lend reserve to each other or to someone that has an account at the Fed.. No one else can hold reserves.”
True. However Scott likes to point out that banks can loan out vault cash (a component of bank reserves). So technically, when a non-bank removes cash from the bank, it loses it’s reserve status, and thus cash is a valid escape path for reserves:
http://brown-blog-5.blogspot.com/2013/04/the-three-places-reserves-can-go.html
HOWEVER, if we do the accounting for all the players in the economy, it’s clear that this escape route for reserves does NOT increase the stock of money in the public’s hands. See my explanation for Scott above. The accounting is done in my post, but I’ll paste that again:
http://brown-blog-5.blogspot.com/2013/08/banking-example-11-all-possible-balance.html
It’s a very very simple “accounting” job there. 😉
14. August 2013 at 18:59
So Scott is correct to say that the base becomes cash, but again this, in and of itself, does not raise the stock of money in the public’s hands.
14. August 2013 at 20:17
Tom, I don’t recall saying that “lending out reserves” would increase the money stock.
14. August 2013 at 20:53
Edward:
“”The Friedman position: tight money can result in gigantic disaster in some cases. And such gigantic disasters typically promote rather than discourage “big dysfunctional government.””
“This is what the Austrian nitwits will never understand. They’re to libertarians like Ralph Nader is to the Democrats.”
Edward your ignorance abounds. Inflation is tool that encourages big government. Austrians cannot “understand” as true what is in fact false.
What you are talking about are different failures of inflationary regimes. You’re not talking about any failure of free market in money.
“And the Austrians and Rand Paul miss a fantastic point. Clearinghouse associations are a private market way of creating EASIER money!”
No you nitwit. Clearinghouses transfer existing money, they do not create new money.
14. August 2013 at 22:15
Scott: you write:
“Tom, I don’t recall saying that “lending out reserves” would increase the money stock.”
OK, great! I realized that wasn’t precisely what you said later. I guess I got the impression that’s what you meant, but it sounds like I was wrong. And as a matter of fact, as long as the LOANS are created at the same time that money is going out as cash, then it does increase the money stock… it’s just that it has nothing to do with the cash (or ER getting withdrawn as cash): it’s all about the loans, cash or no cash.
14. August 2013 at 22:20
In fact, I’ll partially reverse myself on that: as long as banks are “lending out” then it doesn’t matter what they’re lending out: bank deposits or cash: both increase the money stock! It’s all about the loans: deposits and cash can be swapped back and forth after the fact.
15. August 2013 at 03:33
DOB,
“I don’t even know what it means to increase the base beyond the demand at the zero bound. The supply and demand for base money will be in equilibrium wherever you set the base.”
I mean what is the minimum amount that the Fed needs to create to keep rates at 0%. Or more theoretically, what’s the limit of the quantity of base demanded as cost of funds converges to 0%/IOR.
“There are many transmission mechanisms for QE. When the Fed does QE the most noticeable effect is higher asset prices. That means if the reverse QE asset prices will fall.”
I was very specific in my example and said the Fed would drain the reserves by “executing overnight repos”. There is no way to “push the price” on these instruments, they’re too short and will always be priced at par.
“There are other mechanisms as well, such as those described by Woodford (lower expected future AD.)”
Does Woodford have a chapter on HPE? What does he call it? If not, could you summarize the HPE transmission mechanism in one paragraph? “If you can’t summarize it in one paragraph, that’s means you need to re-work the model.”
🙂
15. August 2013 at 11:03
DOB, I don’t think it matters (much) what the Fed buys. The important part of the OMO is not the the asset being purchased, it’s the fact that the base increased.
OK, in one paragraph:
When the Fed injects new base money, the stock of base money exceeds the demand at existing levels of NGDP and asset prices. The public tries to get rid of these excess cash balances. In the short run that raises asset prices and expected NGDP growth. In the long run that raises NGDP itself. Eventually the economy will reach a point where real asset prices have returned to equilbrium and NGDP has risen in proportion to the rise in the base.
Now I’m done here, bring the discussion to a newer post.
15. August 2013 at 18:50
http://en.wikipedia.org/wiki/Jumping_the_shark