Monetarism is dead; long live (quasi) monetarism
In this post Brad Delong treats me like a monetarist. That ideology that died in 2006, when Milton Friedman passed away. Friedman believed that the stance of monetary policy could be characterized by the growth rate of M2. That the Fed should stabilize M2 growth, and that current monetary policy affects future aggregate demand. That monetarism is dead.
Quasi-monetarists like me look at the world very differently. Monetary aggregates are neither good indicators of the stance of monetary policy, nor good policy targets. Rather than assume current changes in M affect future AD with long and variable lags, I assume current changes in the expected future path of M affect current AD, with almost no lag at all. That is, I’m a Woodfordian on the role of policy expectations, and a Friedmanite in that I believe that changes in nominal aggregates are ultimately caused by changes in the supply and demand for the medium of account. (Other quasi-monetarists like Nick Rowe would use the medium of exchange.)
Brad DeLong worries about what happens if an increase in the supply of money leads to lower interest rates, which produces an offsetting fall in the velocity of circulation. He cites John Hicks as an inspiration. But Hicks lived in a world where currencies were pegged to gold, or were expected to be pegged in the future. I used to think the main problem with the gold standard was that it constrained the central bank from printing more dollars than they could back with gold. In that sort of policy environment the danger Brad mentions is quite serious. Now I think the bigger problem was the fact that the gold peg anchored the future expected price level. To see why this is so important, we need to understand that modern policy is not really about either interest rates or (current changes in) the money supply. It is about changes in the expected future path of prices and NGDP, which can be signaled by changes in interest rate targets or the money supply (as with QE.)
Yes, a doubling of the monetary base might have no impact on the price level. Paul Krugman showed that in 1998 and I showed that in 1993. But the reason it would fail (we both agree) is not because rates are stuck at zero, but rather because the money supply increase is expected to be temporary. A money supply increase that is expected to be permanent will raise future expected NGDP, and Woodford showed that future expected AD is the most powerful determinant of current AD. And this is true regardless of whether rates are zero or positive when the monetary injection first occurs. A doubling of the money supply expected to be withdrawn 2 months later will have almost no effect, regardless of short term rates, and a permanent doubling of the money supply will have a huge effect, regardless of the level of short term rates.
We need to stop thinking of current changes in short term rates and/or the monetary base as causal factors, and start thinking of them as signaling devices. Thus the question is not; (as DeLong often implies) “Just how much base money would it take to boost AD?” The question is; “If the central bank promises to target 6% NGDP growth, how much base money do people want to hold?”
Because most economists wrongly assume that low rates and a bloated base mean easy money, they despair at how much easier money must be for significant stimulus to occur. But this is looking at things backward. Money has been extremely tight in the only metric that matters—relative to what is needed to produce on-target NGDP growth expectations. That’s why rates are low and the base is bloated. If the Fed promised to target a much higher future NGDP trajectory, and do level targeting (making up for undershoots), then the demand for base money would probably be far lower than today, and nominal rates might be higher. As Friedman said about Japan, ultra-low rates aren’t a sign of easy money; they are a sign that money has been too tight.
A promise to raise the expected future growth of NGDP is equivalent to a promise to raise the expected future money supply. But it is not a promise to raise the current money supply, nor is it a promise to lower current rates, nor even (as Woodford asserts) to keep short term rates at the zero bound for a period longer than markets currently expect. All those variables respond endogenously.
Friedmanite monetarism died because it wasn’t consistent with good right-wing economics. The long and variable lags could not explain why markets often failed to respond to policy shocks that Friedman and Schwartz thought were important. An X% money supply rule is clumsy central planning, clearly inferior to a policy of having markets determine the monetary base setting most likely to produce on-target inflation. Don’t believe me? Well then explain why late in his life Friedman endorsed Robert Hetzel’s 1989 proposal to have the Fed stabilize the TIPS spread. I’m simply doing the sort of monetary economics Friedman would have done had he been born in 1955, and if his IQ had been 30 points lower, and if he had been able to stand on the shoulders of giants like Irving Fisher and . . . well, Milton Friedman.
Part 2. Only Krugman understands the zero rate bound, and even he doesn’t really understand it.
Paul Krugman once had a post claiming he was the only person who understood liquidity traps. I know the feeling of exasperation. I often feel the same way. (Of course this isn’t actually true. I’m only joking.)
If I am right that the zero bound is not a problem, why do central banks seem to flounder in that situation? I’m not quite sure. I’ve gradually become convinced that the BOJ actually likes zero to -1% inflation—they sure act that way. Krugman doesn’t agree. I still believe the US will eventually exit the liquidity trap, and not end up like Japan. But I can’t be certain. Nick Rowe has as good an explanation as any for the seeming paralysis of US policy at the zero bound. He argues that nominal rates were the Fed’s way of signaling future policy intentions to the public. It’s not that the fed funds rate actually matters all that much for long term investments. And we know that when the Fed eases aggressively then long term rates often go in the opposite direction from short rates (January 2001, September 2007.) No, the Fed is saying “by cutting our fed funds target, we signal that we will provide enough base money over time to raise the long run NGDP growth rate.” That is, if they cut rates in order to change policy expectations. Roughly 80% of rate adjustments are merely reflecting ongoing changes in the Walrasian equilibrium rate, and aren’t intended to move expectations.
Nick says that when the nominal rate hits zero, the Fed is (temporarily) mute. They don’t know how to communicate policy intentions to the markets. Eventually if things get bad enough they develop other languages; quantitative easing, inflation targeting, level targeting, exchange rate depreciation, lower IOR, etc. As I pointed out earlier (and as Jim Hamilton also argued) QE2 did not do anything significant in a mechanical sense. But it did convey to markets a renewed Fed determination to speed up NGDP growth. And it worked; NGDP growth expectations rose significantly following each important Fed speech during September/October 2010, speeches hinting at QE2.
Once you start to look at things this way, everything makes much more sense. Many of my commenters insist that monetizing deficits is the one surefire stimulus option. Not so, as the Japanese have discovered over the past 17 years. If the money supply increase is temporary, the future expected price level will not rise. In that case all the deficit spending in the world won’t create inflation. Yes, most examples throughout history of monetizing the debt have produced inflation, but that’s because no other central bank has been as obsessively masochistic as the BOJ. They reduced the monetary base by 20% in 2006, even though the the price level had not increased in the previous 12 years!
This is not to say that monetizing deficits won’t “work” on most occasions. I’m pretty sure if the Fed began dropping money out of helicopters most people would be able to quickly infer what they were trying to “communicate” about future policy. But why bother, when all they need to do is say they want higher NGDP, or a higher price level? So far the Fed refuses to set any sort of aggressive nominal target, thus it’s not surprising that we’ve been limping along.
The “new monetarism” of the 21st century can’t be found by reading the blogs of people who don’t think nominal shocks are important. Rather, it will be based on a more sophisticated understanding of the role of expectations. Eventually macro and finance will merge. “Easy money” won’t be low interest rates, and it won’t be increases in the current money supply. Easy money will be above-target NGDP futures prices, and tight money will be below target NGDP futures prices. Then and only then will we be able to tear down the confusing Tower of Babel called 20th century macroeconomics, and all start speaking the same language. Then and only then will we be able to focus on the real problems, which are the . . . “real” problems.
PS. A commenter asked me to comment on this post by Mark Thoma, which he thought was directed at my views on QE2. I’m not sure it was, as Mark never mentions my name. But he does link to one of my posts. All I can say is that his argument has no bearing on my claim that QE2 is working, because my argument was not based on how QE2 affected the economy, but rather how it affected market expectations. Because expectations respond immediately to rumors of QE2, there is no policy lag to worry about, and no identification problem. I did also discuss movements in the actual economy, but made it very clear that those changes would be of interest to others, not to me. I had all the information I needed by November 3rd, 2010.
PPS. I now realize that I never really answer Brad’s questions. Financial assets play no role in my model. In my view an increase in the expected future money supply (relative to demand) raises expected future NGDP. That raises the current price of real assets and flexible price goods. It also raises nominal spending, by boosting velocity. With sticky wages, this raises current output. Add interest rates if you wish, it adds nothing to the transmission mechanism in my view. Interest rates aren’t causal factors; they reflect what’s going on with the economy. Disinflation and low output produce low rates, and vice versa.
Yes, fiscal stimulus can “work” if the Fed wants it to work, by boosting V. But if the Fed wants it to work, why not just do the job with monetary policy?
Tags:
13. February 2011 at 17:36
And as I just said in a previous post – there is one person on the Fed who gets this. We are that close. Why does Bernanke or Obama not just say the same thing he does. I really find it almost heart breaking. All of those lives destroyed – because the Fed (as a whole) is terrified by the deflationists.
http://www.chicagofed.org/webpages/publications/speeches/2010/10_16_boston_speech.cfm
13. February 2011 at 17:54
Great post. A bit off-topic… I find myself vigorously nodding my head reading your stuff. But I wonder if it is my exposure to the graduate money curriculum that leads to such vigorous nodding. I hope not, because, no offense, I don’t think this stuff is rocket science and because I want this stuff to be widely understood. Woodford will never be widely understood, but your stuff, and the way you present it, could be.
All of which is to ask: have you thought of writing a textbook or general audience book on the subject of money?
13. February 2011 at 18:17
I like this explanation a lot. Talking about money as a medium of account (combining the medium-of-exchange and the standard-unit information elements of money) makes sense because that is how money is the basic economic signal, how economic agents “speak” to transact. Similarly, the smothering of the policy signal that has been relied on for monetary policy will create a dilemma whose effects on the action of the institution bureaucratic inertia, including problems of framing analysis, is sufficient to explain.
13. February 2011 at 18:37
Scott,
It seems to me that expectations only work to the extent that there are credible tools that will be used that have some impact on NGDP either through the provision of credit or an increase in asset value. It may just be me, but I still don’t understand your views on how this mechanism works.
13. February 2011 at 18:38
…ultimately caused by changes in the supply and demand for the medium of account. (Other quasi-monetarists like Nick Rowe would use the medium of exchange.)
Whoa, ignorant Austrian alert: Can someone explain this distinction? I know what a unit of account is, and I know what a medium of exchange is. But what’s a medium of account? (And it that means the same as “unit of account,” then what’s the difference between supply/demand for it versus medium of exchange?)
13. February 2011 at 18:47
JimP, Thanks, That’s a wonderful article. He didn’t beat around the bush–said we need both more inflation and more real growth to achieve the Fed’s mandate. It’s as simple as that.
Will, Thanks, I greatly appreciate that. I often feel guilty when I think about your comment a year ago that it was time to move on. I wish we had a model explaining why nominal shocks seem to have real effects for such a long time, when wage and price stickiness are not that persistent. My fallback position is that the 99 week UI may be making wages stickier in the low end of the labor market (where most of the unemployment is) and that faster growth would cause Congress to go back to 26 weeks more quickly than otherwise. More AD would lead to more AS. But I recognize that’s a pretty ad hoc argument.
I really wish I could move on from this easy money crusade, as I’m a Chicago inflation hawk at heart. But I just can’t.
In my work on the Depression I found that in some cases the most effective way of figuring out what caused the problem is by observing what produced the turnaround. Which is why I have my fingers crossed on QE2. I believe the policy fight is mostly over. Not it’s about winning the battle of ideas as to what went wrong.
It’s also interesting that lots of younger people like my blog. I don’t recall any middle-aged or older macro people saying good things about my monetary ideas. Except quasi-monetarists obviously. (Mankiw likes my EMH stuff, and the George Mason people like my blog–but they’re mostly microeconomists.)
You asked:
“All of which is to ask: have you thought of writing a textbook or general audience book on the subject of money?”
Yes, but the experience with my depression books suggests that book writing isn’t my forte. I guess blog writing is. I’d need a co-author.
Lorenzo, The earliest monetary theory was medium of account theory. Basically value theory for whatever commodity served as money.
13. February 2011 at 18:56
dtoh, The mechanism is the simple monetarist excess cash balance mechanism. If you have more supply of apples, the value of apples falls. If you have more supply of cash, the value of cash falls.
In the short run that may not work well, as cash and T-bills may be close substitutes. That’s why I emphasize expectations. In the long run cash and T-bills aren’t close substitutes. Cash is not a good substitute for T-bills earning 3%. In the long run more cash reduces the value of cash.
Bob, Today cash is both the medium of exchange and the medium of account. In 1929 cash was the main medium of exchange and gold was the medium of account. It’s not a right or wrong question, it’s which approach is more useful.
Medium of account is the physical object associated with the unit of account. Unit of account is a pure abstraction ‘dollar” “pound” “yen”. Medium of exchange is the physical object used in transactions. Suppose all prices in Bahamas were posted in Bahama money, but 90% of transactions were with US dollars. The US dollar would be the medium of exchange, Bahama currency would be the medium of account. And the Bahama dollar (abstract accounting sense) would be the unit of account.
I don’t even know whether the Bahamas has its own dollar.
13. February 2011 at 19:07
Scott
You wrote:
“In my work on the Depression I found that in some cases the most effective way of figuring out what caused the problem is by observing what produced the turnaround. Which is why I have my fingers crossed on QE2. I believe the policy fight is mostly over. Not it’s about winning the battle of ideas as to what went wrong.”
I got the sense the policy fight was won when Axel Weber bowed out of the race for ECB the same week Kevin Warsh resigned from the FOMC (Yippee!!!).
Any thoughts on Mario Draghi (a former student of Franco Modigliani and Robert Solow) as a possible ECB head?
13. February 2011 at 20:40
“If the Fed promised to target a much higher future NGDP trajectory, and do level targeting (making up for undershoots) . . . .” This is confusingly stated. With the kind of NGDP futures market you envisage, the Fed can always hit its target, since it is targeting the forecast. It can engage in open-market operations to keep the one-year (or eighteen-month, or two-year) forecast of NGDP growing at a steady rate; it need *never* significantly undershoot or overshoot this goal.
Of course NGDP may turn out to have undershot or overshot the forecast of one year (eighteen months, two years) ago. But the Sumnerized Fed pays no attention to NGDP; it is concerned exclusively with the market forecast of NGDP. (This–not actual NGDP–it what it aims to keep growing at a steady rate.)
So I think you should drop the talk of “under/overshooting.” Or have I misunderstood you?
13. February 2011 at 20:49
I like this Nick Rowe view.
This may be a dumb question, and may have a more ideological significance than economic, but shouldn’t the US produce more debt (fiscal policy) in order for the Fed to buy it? Yes, there is a bunch of it out there already to buy, but in the quantitative easing model the Fed is purchasing “other assets” — I’m not sure I’m comfortable with the potential for market distortion through the Fed e.g. buying CDOs (causing their overproduction) … nor the lack of political accountability of, say, Congress allocating money, and the Fed left to buy the debt that went to stuff Congress decided, instead of the Fed deciding itself to buy the aforementioned CDOs.
13. February 2011 at 22:26
Scott,
Thanks for the answer. So I guess a really basic question is…what is the mechanism by which the cash gets into the market. If the Fed buys T-Bills from a bank, it will effect the rate of exchange between T-Bills and cash, but initially the only other thing that happens is that the bank has less T-Bills and more deposits with the Fed. What is the mechanism which then drives the growth in NDGP. Is it entirely expectational (consumer and businesses start spending more on the expectation of higher growth) or do the banks buy other assets pushing up asset prices with a wealth effect or does it cause some expansion in credit.
14. February 2011 at 01:01
You’re really good at blowing away almost anything one knows about macroeconomics 🙂 Other than blogs and Woodford what other quasimonetarists should one read? 🙂
14. February 2011 at 04:26
Scott
This is pretty depressing!
http://wallstreetpit.com/59989-mit-symposium-highlights-macroeconomic-muddle
Even “Moneyillusion” is “attacked” and Bob Hall “throws the towel” saying that “every model agrees that more G will save the day”!
14. February 2011 at 04:32
Scott,
Your reasoning seems to be that the announcement of QE2 was equivalent to an announcement of an increase in “planned” (i.e. policy targeted) NGDP growth.
It suits your overall story, but it’s not clear to me how this is the case.
Aren’t there two stages to the rationale – expectations generally, and NGDP growth expectations specifically? Have you really proven this?
14. February 2011 at 04:45
Bob Murphy: Medium of Account vs unit of account. In England, Sterling is the medium of account; the pound sterling is the unit of account. In China, the remnimbi(?) is the medium of account; the yuan is the unit of account. Under the gold standard, gold is the medium of account; the ounce of gold is the unit of account. (I got these good examples from Paul Krugman).
Just a slightly picky distinction, of no real importance.
Medium of account vs medium of exchange. One day I am going to have to get my head clearer on this, but this is what I think:
1. In the long run, when prices are flexible, the supply and demand for the medium of account is what determines the price level. (That’s where I agree with Scott)
2. In the short run, when prices are sticky, it is an excess demand for the medium of exchange that causes recessions. (That’s where I disagree with Scott).
Man, but you are on a roll Scott! Don’t have anything useful to add. Thanks for the plugs.
14. February 2011 at 04:46
“Because expectations respond immediately to rumors of QE2, there is no policy lag to worry about, and no identification problem.”
No identification problem? An instant response certainly does not imply such a thing. If you are trying to identify the impact of monetary rumors, you need to assume the rumors themselves are random events, not caused by anything in the economy — and if that were true, why would anyone listen to the rumors?
But maybe there is some macro nuance I’ve missed…
14. February 2011 at 05:02
Scott: one word of warning. I don’t think your and Michael Woodford’s views are very compatible. Yes, Michael Woodford emphasises the effect of expected future AD on current AD. You agree there. But the causal mechanism may not be to your liking. It’s intertemporal substitution (present vs future consumption and investment) and the real interest rate that drives AD in his view of the world. It’s not the demand and supply of money. It’s the demand and supply of “bonds”.
14. February 2011 at 07:07
Scott –
Yes – the article was very clear. And I am sure Bernanke was glad it was written, and that he agrees with it.
But – as the article says:
begin quote
There are quite a number of academic studies of liquidity trap crises that find either price-level targeting or temporary above-average inflation to be nearly optimal policies;[1] and yet, central bankers and the public generally loathe the idea that even a temporarily higher inflation rate could be beneficial or be consistent with price stability over the longer term.
end quote
So – the problem is in the minds of the economics profession as a whole, and the mind of the public as a whole.
I don’t think Bernanke can address this directly. The deflationists in Congress are after the Fed enough now. I don’t think he thinks he can take that risk (unless a full blown crisis arrives again).
The task of changing the mind of the public is the task of the President. I sure do wish Obama would do it. I think what Bernanke is saying, with all his sad projections of low inflation and high unemployment from now to the end of time, is:
“Mr. President Obama – this is the best we on the Fed can do – unless you get off your damn ass”
14. February 2011 at 08:12
“Yes, but the experience with my depression books suggests that book writing isn’t my forte. I guess blog writing is. I’d need a co-author.”
Didn’t you write the book before you started blogging? I’d reckon you have gotten a great deal better at it than when you first started? On the other hand you could always go the Tyler-route (Short eBook on the great depression for Amazon).
14. February 2011 at 10:37
Nick,
You and Scott both have me a bit puzzled. Where does the demand for the medium of account come from in a fiat money world? The demand for the medium of exchange is obvious: you need it to undertake transactions in which money changes hands. But both of you seem to be saying there is some separate demand for the medium of account.
You don’t actually need to have any medium of account on hand to use it as such. If we price things in yen but pay for them in dollars, that doesn’t increase the demand for yen.
We can also rule out store of value and precautionary motives for holding yen, since yen-denominated T-bills are more suitable for those purposes.
So where does the demand for the medium of account, as opposed to the medium of exchange, come from?
14. February 2011 at 11:35
The liquidity trap does not seem to be trapping all markets in the United States to the same degree at the moment. Consider the $50B valuation for Facebook and the announced JPMorgan Media Fund. P/E ratios remain compressed for the bulk of U.S. industry, which requires constantly renewed credit to match recurring capital expenditures on property, plant & equipment. What happens when the liquidity trap doesn’t work as well on capital light firms as it does on capital intensive ones?
14. February 2011 at 12:53
Jeff: If the medium of account is not the medium of exchange, where does the demand for the medium of account come from?
Depends what the medium of account is. There’s no one right answer.
If it’s wheat, then it’s the demand for bread, etc.
If it’s Yen, then it’s Japanese demand for Yen as a medium of exchange, plus our demand for Japanese goods, etc.
14. February 2011 at 14:42
Lorenzo, The earliest monetary theory was medium of account theory. Basically value theory for whatever commodity served as money. So, what went wrong? 🙂
14. February 2011 at 15:11
Scott:
a) In your section on “monetarism is dead,” can you isolate at least one refutable null hypothesis that could be subjected to empirical testing?
b) Is it right to conclude from this section that, in your “model,” the entire policy apparatus of the Fed, including the NY Fed’s Open Market Desk, could be replaced by periodic distributions of clearly articulated press releases?
14. February 2011 at 16:57
Trying to separate the MOA and MOE functions brings to mind a variation on Leijonhufved’s blueback scheme.
Suppose the govt decrees two kinds of money, greenbacks and bluebacks. Only contracts specified in bluebacks can be enforced in court. However, bluebacks don’t circulate, only greenbacks do, and there is an official “exchange” rate posted by the central bank that says how many greenbacks equal a blueback. If your contract says you owe 100 bluebacks, and the official rate is 2 greenbacks per blueback, you meet your obligation by paying 200 greenbacks.
Bank deposits are contracts, so when you take deposit 200 greenbacks in the bank your account is credited with 100 bluebacks, or whatever the official rate is. Same thing happens when you make a withdrawal. Checks are contracts, so they’re written in bluebacks.
So here we have bluebacks as the medium of account and greenbacks and the medium of exchange. Note that this doesn’t actually require the physical existence of even a single blueback. They’re purely virtual. And since they don’t actually exist, there’s no demand for bluebacks, either. Money demand is purely demand for greenbacks. If it increases, the central bank can either print more greenbacks or it can increase the real value of existing greenbacks by lowering the green/blue exchange rate. Inflation (defined as an increase in blueback prices) is easily tamed by depreciating the greenback. You don’t have to raise interest rates or reduce the nominal stock of greenbacks.
What do you quasi-monetarists think of this?
14. February 2011 at 20:15
Mark, I don’t know enough to comment.
Philo, Technically you are correct, but most people think in terms of actual, not expectations. So I usually explain it that way.
Jason. Is that your question, or are you quoting Nick? I believe the Fed is buying T-bonds, not CDOs. I certainly favor them buying Treasuries, not CDOs.
dtoh, Normally the mechanism is that banks can earn more on T-bills than reserves. So they spend reserves on assets like T-bills, and the cash goes out into circulation. That’s why I argued that one approach is to pay negative interest on reserves. But if they set a higher NGDP target, nominal rates would probably rise, and if reserves did not earn any interest, the money would naturally flow out into circulation. In any case, that’s the long run transmission mechanism, and the short run effect is caused by expectations of the long run effect.
In actual practice, however, expectations are the main channel, not short run substitution between cash and T-bills. My description is more important in the long run than the short run.
Bogdan, Woodford’s not a quasi-monetarist, he’s new Keynesian. I’m not sure where you’ll get good quasi-monetarism, other than blogs. I don’t know of any books or articles.
Marcus, Eh tu, Bob Hall?
JKH, We pretty much know that QE2 raised inflation expectations, from the TIPS markets. The boost in stock prices, commodity prices, and the fall in the dollar suggest that real growth expectations may have also risen. Also, theory suggests that the only way a demand shock (and QE is demand) can boost inflation expectations is by shifting the AD to the right. And that means more NGDP.
There might be flaws in my story, but that’s one aspect I’m pretty sure about. It’s really just a question of how much.
Nick, I just talked to my wife for 20 minutes about renmimbi and yuan, and I still don’t understand, but I’ll try.
The word ‘yuan’ is not equivalent to any Western word, but is close to ‘dollar.’ The Chinese call all dollars, yen, and euros ‘yuan’ with the prefix of the country in front. Much as Americans say “dollar” and “Canadian dollar” and “Australian dollar” etc. But the pound, mark and franc had different words, which sound slightly like their name in the home country. This is not quite right, but the Chinese would say something like “mei yuan” for the dollar, and “ing pound” for the UK pound. Where the prefix mei signifies America (literally beautiful country.) And the prefix ing signifies England. I have a horrible ear so any Chinese reader will see I am botching these terms, but you get the idea.
At first this sounded very arbitrary to me, but then I realized that Americans call France by the name “France,” and Deutschland by the name “Germany.” So we are equally inconsistent about following local custom.
Yuan is roughly a unit of money, and renminbi is a unit of Chinese money. Thus in everyday life the Chinese often say “yuan” just as Americans often say “dollar” but if working at the foreign exchange trading desk a Chinese would say “renminbi” just as an American working at a foreign exchange desk would say “US dollar.”
Interestingly I had a bit of trouble explaining to my wife why the use of ‘yuan’ for dollar seemed weird to me, as the Chinese are so used to looking at money this way that they take it for granted. Similarly, few Americans ever bother to stop and think how weird it is to call Deutschland by the name “Germany.” It’d be like me saying to Nick, “I know you prefer to be called Nick, but from now on I’ll call you Arthur.”
You said;
“1. In the long run, when prices are flexible, the supply and demand for the medium of account is what determines the price level. (That’s where I agree with Scott)
2. In the short run, when prices are sticky, it is an excess demand for the medium of exchange that causes recessions. (That’s where I disagree with Scott).”
If you agree with point one, then here’s how I’d make the argument. Tight money causes expectations of a lower future price level and lower future NGDP. That expectation causes a fall in the current price of flexible goods and assets, and a also fall in current velocity. The fall in current prices of flexible price goods and assets raises real wages and unemployment. But if most prices are sticky, real wages may not change much. Instead it is the fall in velocity impacting current NGDP, and with monopolistically competitive firms that means lower output.
But I’m still open to your interpretation, both may be right.
Humble microeconomist, I see your point but I think you are missing something. The rumors are discrete events, often related to things like speeches. The backdrop that leads to these rumors is continuous, stretched out over time. So over months the economy weakens, and I agree that makes QE2 more likely. But even so, a market response to a Fed speech is responding to the new information in that speech, not the longer term economic backdrop that led to the policy change. So I think identification is possible, although not always easy. If I didn’t explain this well maybe someone else could chip in. I think it’s an important point so I want to know if I’m wrong.
Nick, You are right that Woodford uses a different mechanism. But the more I think about it, any good model of business cycles must have the property that a shock expected to produce 8% RGDP growth next year (i.e. fast growth) will also raise growth this year. Just as any good model must imply investment is procyclical. I firmly believe that.
more to come tomorrow . . .
14. February 2011 at 20:42
Scott,
You wrote:
“Mark, I don’t know enough to comment.”
Well, read up. More AD in Europe will benefit everybody. I don’t think I can take Draghi’s candidacy to Kantoos because NGDP level tageting friendly Germans seem satisfied with a huge reduction below trend in Eurozone NGDP.
This is a planetwide crisis. I expect you of all people to take an interest.
14. February 2011 at 23:33
[…] マãƒã‚¿ãƒªã‚ºãƒ ã¯æ»ã‚“ã 。quasi-マãƒã‚¿ãƒªã‚ºãƒ ã¯æ°¸é ã«ã€‚by Scott Sumner　 // 日銀ã£ã¦ç•°å¸¸ãªãƒžã‚¾ã ãã†ã§ã™…。スコット・サムナーã®ãƒ–ãƒã‚°ã‹ã‚‰ã€”Monetarism is dead; long live (quasi) monetarism“（February 13th, 2011）　 […]
15. February 2011 at 01:30
Scott,
Thanks for the clarification. So as I proposed before if instead of manipulating rates on reserves which works through a long term mechanism (in essence a very crude and dull policy tool), the Fed manipulated capital reserve ratios, it could get much more leverage and the mechanism would work much more quickly….(e.g. if you raise the capital reserve ratio on Fed deposits to 50% and set a negative reserve ratio on new loans to small businesses you would have a near instant expansion of money and credit). Having this more effective tool would also result in much more control over expectations as well.
15. February 2011 at 04:55
The identification discussion was quite clear, thank you. I had not appreciated the discrete nature of the event, and it does go a long way towards reliable identification.
But I don’t think it goes all the way. The new information in the speech is about the Fed’s observations of the economy as well as its actions. Perhaps that can be dismissed as trivial by someone who knows the Fed better than I. But if market participants believes policy surprises to reflect surprising judgments based in information to which they do not have direct access, then there will be no way to identify policy impacts per se from the market’s response.
15. February 2011 at 06:01
Scott,
“It’s also interesting that lots of younger people like my blog. I don’t recall any middle-aged or older macro people saying good things about my monetary ideas. Except quasi-monetarists obviously. (Mankiw likes my EMH stuff, and the George Mason people like my blog-but they’re mostly microeconomists.)”
Why would that be. I am old but intrigued. Once I’ve found out that your ideas offer no solution either, I’d move on of course. But so far, in a couple of months following this blog (and taking into account indignation about the current stage of world economic history -that is, the OECD area- being intuitively needlessly problematic) I can only say that my curiosity has grown. Not enough to refute my belief that macro economics is merely a sophisticated tool for politicians (because whatever you expect will be different once everyone expects the same) to explain why they were not to blame, or, deserve credit.
15. February 2011 at 06:05
Mark,
More AD will benefit only those whose bankruptcy would otherwise make someone else happy. the EU is in a shakeout phase (in many industries) and German firms are on a roll..
15. February 2011 at 06:57
JimP, Yes, I fear you are right. More and more I believe we won’t be able to have any coherent debate on demand policy until we start actually talking about demand (NGDP) and stop talking about inflation and output. Just look at the absolutely inane debate going on in the UK, where inflation is 4% and output growth is -0.5%.
Martin, My depression book won’t work well on Kindle. A better idea would be to do a short book on the current economic crisis.
Jeff, Most of the demand for the medium of exchange in the US comes from criminals and tax evaders, who hold huge quantities of $100 bills. During normal times (before 2008) most of the monetary base is $100 bills. Those are rarely used as media of exchange.
Michael, Yes, the liquidity trap idea is usually applied to T-bills.
Lorenzo, You said;
“So, what went wrong?”
Good question. As economies got more sophisticated, large financial sectors developed. Then economists noticed that (because of sticky prices) interest rates immediately respond to monetary shocks. They falsely assumed that the subsequent movements in AD were caused by the change in interest rates. That was the serpent’s apple, the moment macro was expelled from the Garden of Eden (of microfoundations for monetary economics.)
mb, Regarding your first question, I’m not sure precisely which hypothesis you want tested. It’s easy to come up with lots of tests that monetary shocks are important.
Press releases by themselves won’t do the job. In the long run the expectations have to be about something. If the Fed promises easier money in the future (which is how I see QE2) then they must deliver at some point (through the trading desk) or future promises won’t be believed.
On question one I’ll try to provide specific tests if you tell me more specifically what hypothesis you want tested. Is it the idea that expectations of future policy are important?
Jeff, I have trouble understanding how a purely virtual unit of account can have value. The closest example to yours that I can envision is the Chilean system, but in that case the blueback was defined as have stable value against a basket of goods. In that case the basket of goods become the de facto medium of account.
Also, you mentioned debts being denominated in bluebacks, but it’s important to know whether ordinary wages and prices are also denominated in bluebacks.
Mark, You said;
“Well, read up.”
Do you know how many requests I get to “read up” every day? If only I didn’t get so many comments to answer I might have time. 🙂
dtoh, You may be right, I don’t really know enough about the proposal to comment, as I haven’t studied capital ratios. In general, I see them as part of the regulatory scheme, not the monetary policy scheme, although I do understand they can be used either way. I might add than we don’t face any dilemma in terms of how to do monetary expansion, even Bernanke admits that. We lack the desire.
Humble microeconomist, Good point. I generally assume the Fed has access to the same information as the rest of us, but I agree that others might see the issue differently, and hence interpret Fed actions differently. I will say however that a Fed speech promising easier money would usually result from a perception in the Fed that the economy was too weak. So if markets thought that the Fed had inside information about the economy, stocks should fall on a speech promising easier money, not rise. Thus at least in the case of QE2, I am pretty confident the market reactions were responses to expectations of easier money.
Rien, I can’t say, but there’s an old joke about scientific progress occurring one funeral at a time. (Meaning older scientists don’t change their minds, but younger scientists are open to new ideas.) More likely they simply don’t find my ideas to be interesting. It doesn’t worry me. I’m much happier having Cowen/Rowe/Avent, etc like my blog than a bunch of academics who pay little attention to current public policy issues. My blog is about the current crisis.
I’m happy to have my blog appreciated by other thinkers that I admire.
15. February 2011 at 08:17
With regard to testable nulls:
How do I test:
1) “that current AD is affected by the EXPECTED future path of M with almost no lag at all” (and how is M defined in this context because you are unclear about the role of the aggregates in your discussions)
2) “that changes in nominal aggregates are ultimately caused by changes in the supply and demand for the medium of account” and, as in the above, what is meant here by “nominal aggregates” and “medium of account” so that someone interested in empirical testing could confront these hypotheses to the data.
15. February 2011 at 08:31
Looking over the blogosphere, I think the past week or so, when you were more explicit about the failure of Fiscal, the DeKrugman get that you could do real harm to their endeavors, and I expect their attacks to get stepped up.
15. February 2011 at 10:23
Scott:
The virtual blueback is the medium of account because the government decrees it, same as any other fiat money. The MMT people, at least, all believe this. To answer your second question, yes, wages and prices are all denominated in bluebacks. Everything is.
The point I’m making is that the demand for money has nothing to do with the unit of account function, only with the medium of exchange function. It’s a fun feature of the blueback scheme that it also includes an easy way to do helicopter drops and vacuums.
Assuming you can measure medium of exchange shocks, you can perfectly offset them by simply changing the exchange rate. Since the medium of exchange is not the medium of account, you don’t have to go through a painful process of adjusting every price in the economy to get back to equilibrium.
Of course you can come up with good reasons why the scheme is impractical, but the point is just to clarify what a money demand shock is, and what it does.
15. February 2011 at 10:51
Rien,
You wrote:
“More AD will benefit only those whose bankruptcy would otherwise make someone else happy. the EU is in a shakeout phase (in many industries) and German firms are on a roll.”
More AD will prevent Germany from inevitably picking up the tab for losses on the Eurozone periphery, so it’s also in their interest.
Also, I’m much less sanguine than you about Germany’s “recovery” (mind you I’m over 5000 kilometers away). Yes, labor markets are getting tight, but machinery and equipment investment, construction, and even exports (50% of the German economy) are all below peak (13.3%, 3.1% and 1.4% respectively) as is real GDP (1.4%). German growth is being driven by consumption and as a consequence labor productivity is still down nearly three years later. For example, unit labor costs in manufacturing are up 18% over 2006-2010 and output is down 4%. Everybody (especially in Germany) is pointing to the jobs numbers and the unemployment rate but in my opinion that doesn’t tell the whole story.
And, Eurostat released the fourth quarter numbers for the EU this morning and it’s not a pretty picture. Real GDP in the Eurozone grew at a 1.2% annual rate, the same as the previous quarter. Germany, France, Italy and Spain were up by just 1.6%, 1.2%, 0.4% and 0.8% respectively. Portugal and Greece were both negative (and it’s the 10th quarter in a row for Greece). In the past year Eurozone real GDP only grew 2.0%. (I suppose you’ll blame it on the weather):
http://epp.eurostat.ec.europa.eu/cache/ITY_PUBLIC/2-15022011-AP/EN/2-15022011-AP-EN.PDF
All this German cockiness is looking a bit silly for my tastes. In my opinion basing future ECB policy on their ill founded self perceptions of success will turn out very badly in the end for everybody.
P.S. Turning my attentions to the Far East, I notice Japan’s real GDP fell 1.2% at an annual rate in the fourth quarter. So apparently the recovery is faltering there as well.
15. February 2011 at 13:31
Scott,
You said “I might add than we don’t face any dilemma in terms of how to do monetary expansion, even Bernanke admits that. We lack the desire.”
Which is another beauty of using capital reserve ratios instead of asset purchases. The lack of desire comes from a fear of the political backlash to “printing money” and facilitating the growth of government debt. If the Fed was instead making a technical tweak to reserve ratios to “force the bad bankers to provide more credit to cash starved business and consumers”, there would be no backlash and the problem of desire would go away.
Bottom line…. setting capital reserve ratios is a tool which not only works better than the current policy tools but also eliminates the political problems associated with those tools.
15. February 2011 at 14:19
I don’t see how you can plug variable money lags into the equation of exchange. The equation of exchange is an algebraic way of stating a truism;
that the product of the unit prices, and quantities of goods and services exchanged,
is equal (for the same time period),
to the product of the volume, and transactions velocity of money.
If even on the first day, money flows exceed output, then you have inflation.
15. February 2011 at 15:19
Dtoh,
I agree that capital reserves are much less politically difficult than asset purchases (though I think lowering IOR would be easier than either). However, aren’t capital reserve requirements a rather unpredictable and potentially very disruptive way of conducting monetary policy?
15. February 2011 at 16:02
The “monetary base” is not now, and has never been, a base for the expansion of new money & credit. That is just one monetary myth.
Another is the that idea that a single money figure can be used as a guide post for monetary policy, i.e., that velocity is a constant.
Add to that Milton Friedman’s income velocity – a contrived figure.
15. February 2011 at 16:17
Scott, I see you are adressing Woodford’s points.
How do you see his opinions on the role of interest bearing reserves? As I understand, he posits that it is an independent tool, aimed more to liquidity problems. After reading your blog since the beginning, I think I can figure out, but…
He asserts that the optimal policy is to set the interest rate on reserves equal to the policy target for the fed funds (I think this is weird). “We also consider optimal policy with regard to the payment of interest on reserves; in our model, this requires that the interest rate on reserves be kept near the target for the policy rate at all times”.
He also discusses reserve targeting and says: “Reserves should be supplied to the point at which the policy rate falls to the level of the interest rate paid on reserves, or,
in a formulation that is more to the point, interest should be paid on reserves at the central bank’s target for the policy rate.”
No central bank in the world pays interest on reserves, so I must conclude everybody’s monetary policy is not optimal (except for the Fed today, but it is not regular).
15. February 2011 at 16:23
“But if the Fed wants it to work, why not just do the job with monetary policy?”
That answer is obvious. The member banks don’t loan out savings or existing deposits. The utilization of bank credit to finance real-investment or government deficits, doesn’t constitute a utilization in savings, since financing is accomplished by the creation of new money.
I.e., lending by the commercial banks expands both the volume and the velocity of new money. I.e., lending by the cbs is inflationary.
However lending by the non-banks activates existing money that has been saved, and all of these savings originate from outside of these institutions. I.e., lending by the non-banks is not inflationary (matching savings with investment).
15. February 2011 at 17:37
Peden,
You ask, “aren’t capital reserve requirements a rather unpredictable and potentially very disruptive way of conducting monetary policy?”
I don’t think so. To the contrary, they give the Fed a predictable and precise way to control credit and the money supply.
Lowering the IOR might help, but it might just result in the banks reducing their overall assets. By manipulating the capital reserve ratio, the Fed could actually force the banks to expand credit and could do it almost instantaneously. Using current policy tools is like pushing on a rope compared to the control the Fed could have by manipulating capital reserve ratios.
15. February 2011 at 18:14
The need is to see the threat of inflation from a typical consumer’s view. I have an indexed wage, so a 10% annual inflation would reduce my real income by 10% (since the indexing will have about a years delay. Therefore, I am reducing my spending (and thus the velocity of money (a very litle bit), instead of increasing it.
By the way, the increase in monetary base suggests that we can expect 10-20% inflation before it is over.
16. February 2011 at 18:32
mb, In the US today the medium of account is the monetary base. Nominal aggregates are things like the price level and NGDP. I prefer NGDP. When interest on reserves is zero, there is a tendency for the monetary base and NGDP to rise at roughly the same rate in the long run. The monetary base was about 7 billion in 1929, and about 120 times larger by 2007. Nominal GDP rose from about $100 billion to about $14 trillion, about 140 times larger. Those two facts are related.
When the Fed began paying interest on reserves the demand for base money soared, although it might have anyway once rates hit zero.
It’s hard to get data on changes in the expected future path of monetary policy. When the government devalued the dollar in March-July 1933, the WPI rose 14% in 4 months, and the level of industrial output rose 57% in 4 months, indicating a quick response to monetary stimulus that was expected to lead to a higher future money supply. On the other hand the Fed’s open market purchases of 1932 were not expected to lead to a permanent increase, as we were still on gold. So prices and output didn’t rise.
It’s too soon to say whether QE2 has had an effect on prices and output, but some people think it has (see my post on QE2 after 3 months.) The effect is fairly modest, however. The US doesn’t usually do the sort of controlled experiments that would make it easy to identify the effect–1933 was an exception.
Morgan, You said;
“I expect their attacks to get stepped up.”
Let’s hope so, I can’t afford to lose my street cred with the right wing.
Jeff, If I understand you correctly, you are saying that greenbacks would no longer have macroeconomic effects. If so, I agree. But I worry that you understate the difficulty of stabilizing the value of bluebacks.
BTW, forget the Chilean system. Because you say all prices are in bluebacks, it is not relevant to your example.
dtoh, Good point about printing money being unpopular, which I why I think lower IOR would be better, or even NGDP targeting, which would require much less money to be printed.
flow5, No, usually on the first day that money rises, velocity falls.
You said;
“Another [myth] is the that idea that a single money figure can be used as a guide post for monetary policy, i.e., that velocity is a constant.”
This is correct, but it contradicts your assertion that more money means inflation.
Rafael, You said;
“How do you see his opinions on the role of interest bearing reserves? As I understand, he posits that it is an independent tool, aimed more to liquidity problems.”
I think that was the intention, but it backfired.
BTW, lots of central banks pay interest on reserves.
flow5, You last comment has nothing to do with monetary policy. That’s credit policy. I don’t favor easy money as a way of getting banks to lend more. Indeed I’d favor easier money even if our economy lacked a financial sector. I want to boost NGDP. The financial system is no more relevant than the system of fast food restaurants.
Dale, Since the Fed began paying IOR, the base is no longer a reliable inflation indicator.
17. February 2011 at 09:43
Scott,
I’m saying that greenbacks would have macro effects only through the real balances channel, which is how I interpret the views of people like Nick Rowe and Bill Woolsey. Price expectations could be anchored by some rule that would change the green/blue exchange rate to fix a general blue price index, so that part of your story would be gone.
But really, my only intent with this story was to show that there is no demand for the medium of account as such, only for the medium of exchange.
17. February 2011 at 09:44
Monetary or credit policy? (no, the member banks pay interest for something they already own). Velocity (rate-of-turnover), doesn’t drop as you say. In fact, it rarely ever drops, even in downturns (only income velocity drops). The G.6 proves that.
Even on the first day, where money flows begin to exceed REAL-output, then prices will mathematically adjust (based on the equation of exchange). It is self-evident from the equation: that an increase in the volume, & or velocity of money, will cause a rise in unit prices, if the volume of transactions increases less, and vice versa.
Thus, monetary lags are not “long & variable” as pontificated. However, production limits create an economic lag (the time period in the calculus).
17. February 2011 at 12:43
See here’s James Hamilton “when M1 goes up, the velocity of M1 goes down by an almost exactly offsetting amount”
Absolutely wrong. Economists need to get religion.
17. February 2011 at 20:06
I spent a couple of years at the university of Chicago earning my MBA (after finishing my course work and cumulative exams for a PhD in Physical Chemistry). My most important take-away: don’t listen to Chicago economists. Their models and world view are so distorted that their perspectives are meaningless. Come on – how can anyone believe that mass unemployment is voluntary?
17. February 2011 at 20:21
Jeff, Yes, there is no demand for the medium of account as such, but there often is a demand for the medium of account.
I guess I still don’t understand where you are going with all this. Do you claim it sheds some light on the dispute between Nick and I? I don’t see any implications.
flow5, The equation of exchange tells us nothing about causality, absolutely nothing. And I always define velocity as income velocity. If you have another velocity in mind you need to spell it out (and of course the equation of exchange doesn’t hold for transactions velocity.)
You said;
“Economists need to get religion.”
Jim Hamilton is a good economist precisely because he doesn’t rely on religious faith, he relies on evidence.
17. February 2011 at 20:49
Scott,
You wrote:
“Jim Hamilton is a good economist precisely because he doesn’t rely on religious faith, he relies on evidence.”
The two are not completely incompatible. I consider myself modestly religious and yet, perhaps oddly, I also consider myself extremely empirical/consequentionalist/utilitarian (but then I’m also Catholic, so go figure [Catholic Social Teaching?]).
In fact I was talking to my best friend (he’s an atheist) from high school last night (after a decade of him totally ignoring me, but that’s another story). I told him wouldn’t “Jesus was a utilitarian” make a nice T-shirt?
19. February 2011 at 05:35
Dismayed, The term “voluntary” is very misleading, as it implies that the unemployed are not suffering greatly. That poor word choice does not mean the content of their analysis is wrong. I think Chicago models provide very good explanations for the mass unemployment in places like Western Europe over the past 30 years—far better than Keynesian “involuntary” unemployment models. I don’t think anyone at Chicago would deny that unemployed workers would prefer to have the old jobs back–at least most of the time.
I also think it is simplistic to argue that they necessarily view all unemployment as voluntary. Consider minimum wage laws, for instance.
Having said all that, I disagree with those economists who deny the importance of nominal shocks. My teacher there in the 1970s (Robert Lucas) certainly understood the importance of nominal shocks–I can’t speak for the younger faculty.
Mark, But does religious faith inform your views of the economy? That’s what I meant. The Golden Rule is of course the essence of utilitarianism.
6. September 2011 at 14:55
[…] thankfully have a very excellent excuse for siding with quasi-monetarists like Scott Sumner and Niklas Blanchard for moving away from believing fiscal stimulus is the answer to economic […]