Case closed: Milton Friedman would have favored monetary stimulus

In a recent post I argued that Milton Friedman would have been extremely critical of the Fed’s tight money policy since late 2008.  I cited a number of factors, including:

1.  He said in late 1997 that the ultra-low interest rates in Japan were actually a sign that money had been too tight.

2.  Late in his career he moved from money targeting toward other options like inflation targeting.  He even endorsed Hetzel’s proposal to target TIPS spreads.

3.  I forget to mention the interest on reserves policy, which is very similar to the 1936-37 policy of doubling reserve requirements.  Both programs only raised short term rates by about a 1/4 point, but Friedman (and Schwartz) understood that the 1937 policy was highly contractionary despite the tiny interest rate increase, because it sharply reduced the money multiplier.  He would have been a severe critic of the current IOR policy.

But there was one weakness in my argument, which I acknowledged.  Friedman was famous for favoring steady money supply growth, and M2 and MZM grew quite rapidly during 2008-09.  So would Friedman now have opposed stimulus, despite the fact that growth in these aggregates fell close to zero after mid-2009?  John Taylor thinks so.

But I have found an even more recent Friedman article that sharply undercuts the only plausible argument that Friedman would have been with the inflation hawks.  In 2003 he wrote a very interesting article on recent trends on monetary policy, and basically made peace with the new Keynesian inflation targeting approach:

To keep prices stable, the Fed must see to it that the quantity of money changes in such a way as to offset movements in velocity and output. Velocity is ordinarily very stable, fluctuating only mildly and rather randomly around a mild long-term trend from year to year. So long as that is the case, changes in prices (inflation or deflation) are dominated by what happens to the quantity of money per unit of output.

Prior to the 1980s, the Fed got into trouble because it generated wide fluctuations in monetary growth per unit of output. Far from promoting price stability, it was itself a major source of instability, as Chart 1 illustrates. Yet since the mid ’80s, it has managed to control the money supply in such a way as to offset changes not only in output but also in velocity. This sounds easy but it is not — because of the long time lag between changes in money and in prices. It takes something like two years for a change in monetary growth to affect significantly the behavior of prices.

The improvement in performance is all the more remarkable because velocity behaved atypically, rising sharply from 1990 to 1997 and then declining sharply — a veritable bubble in velocity. Chart 2 shows what happened. Velocity peaked in 1997 at nearly 20% above its trend value and then fell sharply, returning to its trend value in the second quarter of 2003.

The relatively low and stable inflation for this period documented in Chart 1 means that the Fed successfully offset both the decline in the demand for money (the rise in V) before 1973 and the subsequent increase in the demand for money. During the rise in velocity from 1988 to 1997, the Fed kept monetary growth down to 3.2% a year; during the subsequent decline in velocity, it boosted monetary growth to 7.5% a year.

Some economists have expressed concern that recent high rates of monetary growth have created a monetary overhang that threatens future inflation. The chart indicates that is not the case. Velocity is precisely back to trend. There is as yet no overhang to be concerned about. (Italics added.)

Note that Milton Friedman is criticizing “some economists” who have “expressed concerns that the high rate of money growth . . . threatens future inflation.”  Today those “some economists” are obviously monetarists, Austrians, and conservative Keynesians (but not all in those camps.)  And Friedman is telling his fellow conservatives (from the grave) that they are wrong, that this “is not the case.”

Friedman would have understood that the financial crisis was a special case that led to a rush for liquidity and safety, and a temporary fall in M2 velocity.  He would have seen the low interest rates and low TIPS spreads as indicators of tight money.  He would have favored temporarily allowing higher M2 growth to offset the low velocity, until the economy was back to normal.  Somehow modern conservatives seem to merely recall the bumper sticker message “stable money growth” but overlook the nuanced and highly sophisticated monetary analysis that made Milton Friedman an intellectual giant.

HT:  Jeffrey Hummel


Tags:

 
 
 

60 Responses to “Case closed: Milton Friedman would have favored monetary stimulus”

  1. Gravatar of Luis H Arroyo Luis H Arroyo
    16. September 2010 at 08:32

    Scott, good, as usual. A question:
    how is measured the velocity of circulation?

  2. Gravatar of Wonks Anonymous Wonks Anonymous
    16. September 2010 at 08:35

    I recall Henderson & Hummel making a similar argument about how the Greenspan Fed largely accommodated swings in V with M, resulting in overall stability. They had a bit of a running argument with George Selgin on the issue.

    Stephen Williamson is claiming that because the unemployed are stuck in houses, they can’t move to places with lower unemployment/higher vacancies, so the Fed can’t do much.
    http://newmonetarism.blogspot.com/2010/09/sectoral-reallocation-and-housing.html

  3. Gravatar of Greg Ransom Greg Ransom
    16. September 2010 at 08:36

    Late in his career Friedman endorsed free banking and the de-nationalization of money.

    You can look it up.

  4. Gravatar of Wonks Anonymous Wonks Anonymous
    16. September 2010 at 08:42

    Forgot to give a link on Henderson & Hummel:
    http://econlog.econlib.org/archives/2008/12/our_response_to.html

  5. Gravatar of Doug Bates Doug Bates
    16. September 2010 at 09:11

    Although I appreciate learning more about his views, I do not grant Milton Friedman the title of God — it does not matter to me what he said, but whether what he said on any occasion makes sense and can be supported by both observable facts and internal logic.

    So, for people to argue about whether Friedman would’ve supported a particular action today or not is irrelevant to my determination of whether the particular action makes sense.

    Economists can be weird in that they have a difficult time performing experiments to back up their claims, so instead they sometimes fall into habits of magical thinking and hero worship.

  6. Gravatar of Wonks Anonymous Wonks Anonymous
    16. September 2010 at 09:24

    Scott has been critical of some of M.F’s ideas, like “long and variable lags”. And (this is related) doesn’t share some of his (changing) views on good measures of monetary policy. Scott is referencing Friedman because he is arguing against inflation hawks, and many of them respect Friedman.

  7. Gravatar of Luis H Arroyo Luis H Arroyo
    16. September 2010 at 09:44

    Scott, i´ve just read the Friedman´s paper: I see how Velocity is estimated.
    Do you think that the differences in velocity between euro countries is a serious obstacle to reach a efective monetary policy?

  8. Gravatar of David Pearson David Pearson
    16. September 2010 at 10:52

    Scott,

    I read today a suggestion that the Fed revalue Treasury’s gold reserves (resulting in a credit to the Treasury’s account). I believe the revaluation would amount to $300b. It would be a one-time, permanent helicopter drop. It would also bring the U.S. in line with European practices on gold reserve valuation (the Bundesbank revalued theirs in 1997).

    Any thoughts on the idea?

  9. Gravatar of Morgan Warstler Morgan Warstler
    16. September 2010 at 11:06

    Scott, you are wrong.

    MF’s note you quote on this are much like Lacy Hunt said, and I told you to go look at…

    The ONLY offset you’d get from him is to keep inflation FLAT – meaning as V drops (and it is dropping), you add more M but ONLY for that V offset.

    He DOES NOT SAY, “hey lets target NGDP and do QE until inflation runs faster.”

    So, are you changing your policy that we ONLY need enough QE to offset the loss of V, to keep inflation running at under 25?

    One last note: we do not have a “temporary fall in M2 velocity” it is still falling back to the historical avg… like 1.6 or something.

  10. Gravatar of 123 – TheMoneyDemand Blog 123 - TheMoneyDemand Blog
    16. September 2010 at 11:49

    Scott,
    I think it is wrong to compare current IOR policy to 1937 reserve requirement policy in terms of money multiplier. In 1937 the primary cause of reduced money multiplier was the doubling of reserve requirements. During the current crisis, money multiplier decreased because of the perception that the current expansion of monetary base is temporary and not permanent, and because of the extremely elevated credit spreads during the early stages of the crisis. IOR has nothing to do with the money multiplier under the current policy framework. Imagine that congress abolishes IOR. In this case money multiplier will not change at all, as markets will now the Fed’s exit strategy will be faster than previously estimated, and the reverse helicopter drop will arrive much sooner than previously estimated.

    Friedman would have supported IOR policy, because already in 1959 he was concerned by the distortions caused by the artificial scarcity of reserves.

  11. Gravatar of Benjamin Cole Benjamin Cole
    16. September 2010 at 13:37

    Once again, Scott Sumner prove himself the most valuable economic commentator on the American scene today. This is just great stuff.

    To think we have the dithering weaklings and the Japan Wing of the Fed suffocating our economy money……for no good purpose.

    It is one thing to suffer for future benefit…it is another to suffer so that you can suffer some more….

  12. Gravatar of Jim Glass Jim Glass
    16. September 2010 at 15:17

    modern conservatives seem to merely recall the bumper sticker message “stable money growth” but overlook the nuanced and highly sophisticated monetary analysis that made Milton Friedman an intellectual giant.

    Same with Krugman, who gloated that Friedman had given up on targeting the quantity of money, and thus had “conceded” to Krugman’s critism that “monetarism was naive”.

    Oddly — apparently unbeknownst to Krugman — Bernanke, in describing how all central bankers are now Friedmanites, had posted on the Fed’s web site Friedman’s Eleven Key Monetarist Propositions.

    Targeting the quantity of money is not among them. Go figure.

  13. Gravatar of Benjamin Cole Benjamin Cole
    16. September 2010 at 15:33

    One correction:

    “But I have found an even more recent Friedman article that sharply undercuts the only plausible argument that Friedman would have been with the inflation hawks.”

    They are not “inflation hawks.” They are the “dithering Japan Wing of the Federal Reserve.” They are the “do-nothing crowd.” They are the “monetary quislings representing Deflation-land.”

    Anything but “inflation hawks.”

    And call yourself a “Growth Hawk” or something.

  14. Gravatar of scott sumner scott sumner
    16. September 2010 at 16:14

    Luis, V = NGDP/M

    Wonks Anonymous, People that make those arguments don’t seem to pay attention to empirical data. Unemployment is at recession levels in 95% of the country; it’s not some sort of localized problem. We’d benefit greatly from more AD, which would generate jobs in all sorts of industries.

    Greg, I like free banking as well.

    Doug, I agree.

    Wonks Anonymous, Yes, I was responding to Taylor’s post.

    Luis#2, No, variations in V are not a problem. The problem is variations in the business cycle from one country to another.

    David Pearson, I think they should sell off the gold at market prices, so I have no objection to revaluation. I doubt it would do much good unless the gold was monetized.

    Morgan, You said;

    “One last note: we do not have a “temporary fall in M2 velocity” it is still falling back to the historical avg… like 1.6 or something.”

    Fine, but that doesn’t change my argument. I don’t follow the rest of your comment.

    123, You said;

    “Imagine that congress abolishes IOR. In this case money multiplier will not change at all, as markets will now the Fed’s exit strategy will be faster than previously estimated, and the reverse helicopter drop will arrive much sooner than previously estimated.”

    I don’t follow. If exit is expected sooner that implies NGDP growth is expected to be higher. Standard theory suggests that when banks become more bullish because NGDP growth is expect to be higher, the multiplier should fall.

    I agree that Friedman would probably favor interest on reserves, but only during normal times.

    Thanks Benjamin.

    Jim Glass, Well Krugman may think the monetary theory of the Depression has been discredited, but he’s just about the only one.

    The Bernanke link didn’t work for me.

  15. Gravatar of Bonnie Bonnie
    16. September 2010 at 16:15

    This is sort of off topic, but I wanted to share some headlines I read this morning. The first is that the number of Americans living in poverty rose last year by the highest percentage since 1994 (I found this story on foxnews.com). The second is that the number of Americans smoking pot or using crystal meth rose by last year by the largest percentage (9%) since data started being collected, and I’m sorry I heard it on an ABC news radio broadcast this morning while taking my daughter to work and I don’t remember the date that was given for when data on illicit drug use were first collected.

    The reason I mention these is because a while back Scott had a blog post regarding the reduction in new coins being minted due to a glut of existing coins being returned to the banks. He mentioned how sad it is that people are being reduced to cracking open piggy banks and scrounging for change just to get by, and I completely agree with him. But it seems to be much worse than that as far as social effects of this recession, especially the figure on the increase in meth use since that one is extremely addictive and it isn’t a habit one can just get over.

    When I think about it a bit I wonder what these people who sit in the FOMC meetings arguing really have on their minds. And then I wonder what might happen if most people understood what was happening at the Fed, or understood that it is just a big cluster of intellectuals arguing about what should be done, with nothing being accomplished at all. Apparently leadership quaities are in very short supply and we are all paying for it dearly in the immeasurable cost of human tragedy that is growing by recordbreaking amounts.

    What value does the Fed(zilla) bring to the table again? Perhaps someone can help me out and explain why we need it in terms of practice rather than theory. It might be a great thing when managed correctly, but then again so is nuclear technnology. When something goes haywire with it, however, it sure makes one heck of a mess.

  16. Gravatar of Jim Glass Jim Glass
    16. September 2010 at 17:08

    The Bernanke link didn’t work for me.

    https://federalreserve.gov/boarddocs/speeches/2003/20031024/default.htm

    Norton identifies the Fed web site as dangerous, at least to me. Perhaps they are on to something.

    Nevertheless, if one recklessly clicks through the warning page, Ben is there.

  17. Gravatar of 123 – TheMoneyDemand Blog 123 - TheMoneyDemand Blog
    16. September 2010 at 17:09

    Scott, you said:
    “I don’t follow. If exit is expected sooner that implies NGDP growth is expected to be higher.”
    If IOR is abolished, the Fed will not change the preferred expected future path of NGDP. And NGDP expectations will not change when IOR is abolished, because the Fed will promise to start withdrawing monetary base sooner than markets were expecting before.

    The main problem with the Fed is the degree to which excess monetary base is perceived to be temporary. The Fed should nudge these perceptions by switching to level targeting.

    You said:
    “I agree that Friedman would probably favor interest on reserves, but only during normal times.”
    I think Friedman would probably agree with Bernanke that 0.25 IOR is not a big difference to 0.00 IOR, and he would agree with Bernanke that the other policy options Bernanke described in his Jackon Hole speech are better.

  18. Gravatar of marcus nunes marcus nunes
    16. September 2010 at 18:51

    “When will they ever learn?”
    According to The Economist, America hasn´t rebounded (as usual)from recession and the prospects are bleak because…past recessions have been caused by tight money and when policy is loosened, demand rebounds. But not this recession, which was the result of a financial crisis. And the mantra is…recoveries in this case are normally weak and slow!!!
    http://www.economist.com/node/17039121
    @Benjamin Cole: You may be a bit optimistic about “us” gaining ground. Maybe the Economist doesn´t have a high regard for its blogger (Ryan Avent?)

  19. Gravatar of Mark A. Sadowski Mark A. Sadowski
    16. September 2010 at 19:34

    Scott,
    In the interest of being nuanced (like that really matters) I’ll both agree with you and disagree with you at the same time. Go figure out which is which.

    1) You said:
    “Both programs only raised short term rates by about a 1/4 point, but Friedman (and Schwartz) understood that the 1937 policy was highly contractionary despite the tiny interest rate increase, because it sharply reduced the money multiplier. He would have been a severe critic of the current IOR policy.”

    Could this have something to do with the interest elasticity demand of money gets flatter the closer we get to a zero interest rate? In other words, small differences at the zero lower bound could make a BIG difference?

    2) You said:
    “But there was one weakness in my argument, which I acknowledged. Friedman was famous for favoring steady money supply growth, and M2 and MZM grew quite rapidly during 2008-09.”

    Monetary aggregates were growing rapidly before the Great Recession. They’ve been growing more slowly ever since. John Taylor may have had a Rule named after him but we needn’t get taken in by his tomfoolery. Just look at the numbers. They never lie.

    P.S. I’m still having fun at Rowan. On Wednesday we ended class by looking at a graph from FRED of the MB. (I have a “smart” classroom.) I told them that next class I’d explain to them it’s true meaning (hint: IOER).

  20. Gravatar of Benjamin Cole Benjamin Cole
    16. September 2010 at 20:35

    Marcus Nunes:

    Well, obviously there is non consensus that an aggressive QR program is needed. Yet I see more and more talk about QE, and increasing comfort levels. It is no longer an exotic topic. After all, BofA, Merrill and Goldman all say more QE is coming.

    I still contend that many in the QE are not approaching the topic from the right stance, PR-wise.

    We talk about “easing” and allow the other side the moniker “inflation hawks.” Well, who is for high inflation?

    QE’ers need to call themselves “Growth Hawks,” and talk about the need fro an “innovative and aggressive” Fed policy. Don’t use the words “easing.”

    As I have said before, the non-QE’ers should be defined as the “Japan Wing” or “dithering do-nothings.”

  21. Gravatar of Morgan Warstler Morgan Warstler
    16. September 2010 at 23:05

    Scott,

    You recently here told me you didn’t really care about Velocity… I found this odd PRECISELY because V is the reason MF gives for having ANYTHING to do with your silly QE.

    Meaning his reasoning has nothing to do with your idea that we need to promise MORE inflation / QE in order to hit some old trend that gives you a boner (crass, I know, but MF needs defended from your idiocy).

    He’d be far more likely to say:

    1. “Scott, your old trend line is too aggressive use the one from 2001-2003, BEFORE the Fed screwed up, see we’re right on trend!”

    2. “Scott, since we’re right on NDGP trend, we only need enough QE to offset the fall in V, meaning if V stops falling – we don’t need any, and either way we just need enough to keep inflation at under 2%!”

    3. And most importantly he’d say, “Scott, let’s wait until after we get a far more conservative fiscal policy in place, before we even consider artificially pumping the economy, the LAST THING serious economists do is aid and abet Democrats… unless they are Bill Clinton doing EXACTLY what we tell them to do.”

    So please stop taking MF’s name in vain… when Krugman won the NP, Friedman rolled over in his grave. It would serve your own edification, to think about the parts where MF disagrees with you, and really go to town on those so you can fix your thinking.

  22. Gravatar of johnleemk johnleemk
    17. September 2010 at 02:13

    Benjamin,

    Inflation hawks are referred to as such because they are tough on inflation — want to wage war against it, if you will. Inflation doves (Janet Yellen is a typical example) make peace with inflation. A growth hawk would be someone who doesn’t want growth.

  23. Gravatar of Lee Kelly Lee Kelly
    17. September 2010 at 03:46

    Hey! That’s my argument about IOR being like an increase in reserve requirements. I wrote it here a couple of months ago. Give it back!

    TheMoneDemandBlog, 0.25% does not make a big difference, but it does make a difference. I don’t think anyone believes that IOR by itself can explain the sudden increase in demand for reserves, and perhaps reserves should pay interest during normal times. I think the issue is that positive IOR was introduced at precisely the wrong moment — just when nothing more contractionary was needed … bam, there it was. Maybe the absence of IOR would have had little effect on NGDP, but such could mean a big difference in the livelihood of tens of thousands of unemployed.

    In the long run it might not be so bad. Perhaps IOR is a good policy for normal times but would never be implemented except during a crisis. Something good may come out of this.

  24. Gravatar of Bill Woolsey Bill Woolsey
    17. September 2010 at 04:13

    Unemployment and real growth targeting rightly have a bad name.

    I see little option other than to simply advocate a stable growth path for money expenditures. This will prevent accelerating inflation in the long run. It doesn’t guarantee low (or lower) unemployment or more real growth.

    Money expenditures are well below the trend growth path of the Great Moderation. The Fed should commit to raising them. Perhaps they should be raised to the past growth path. My view is that we should switch to a new growth path consistent with a stable price level. But they need to be substantially higher now.

  25. Gravatar of 123 – TheMoneyDemand Blog 123 - TheMoneyDemand Blog
    17. September 2010 at 04:50

    Lee Kelly,
    IOR was introduced when the deflationary pressure was caused by exploding credit spreads. Before the introduction of IOR, reserves were artificially scare and holders of reserves were earning huge convenience yields. When Bernanke introduced IOR, he flooded the system with reserves, scarcity yield on reserves has disappeared. The net effect of increased IOR and decreased scarcity yield was stimulative. And because IOR allowed the Fed to expand its balance sheet without any cooperation from the Treasury, Bernanke started fighting the explosion of the credit spreads. Alas, the whole fight took him half a year, while I guess Greenspan would have finished in one week.
    IOR is a good policy both in good and in bad times. It is especially useful during the crisis, as it allows earlier deployment of QE.

  26. Gravatar of Lee Kelly Lee Kelly
    17. September 2010 at 05:39

    MoneyDemandBlog,

    Sorry, but I do not understand your argument. This is more likely my fault than yours, because I just do not know enough about these matters.

  27. Gravatar of 123 – TheMoneyDemand Blog 123 - TheMoneyDemand Blog
    17. September 2010 at 05:50

    Lee Kelly,
    let me simplify a bit:
    There are two modes of central banking:
    1. No IOR – central bank restricts the quantity of reserves, these reserves are very scarce, their scarcity value makes financial institutions happy to hold them even though they earn zero interest rate.
    2. IOR – central bank creates lots of reserves, they have no marginal scarcity value. Financial institutions are happy to hold reserves only because they are receiving competitive rates of interest.

    A switch from one system to another does not mean that demand for reserves has increased, whereas in 1937 demand for reserves was caused by the changed reserve requirements.

  28. Gravatar of W. Peden W. Peden
    17. September 2010 at 06:04

    By the way Scott, I’m not sure if you were aware of this article, but in it Tim Congdon explicitly endorses NGDP targeting-

    http://blogs.telegraph.co.uk/finance/jeremywarner/100006205/congdon-just-keep-money-growing-and-the-economy-will-expand/

    “But it needs monetary policy to be hospitable. And here we reach the heart of Professor Congdon’s contention. Stop worrying about inflation, stop worrying about banking capital ratios, or whether banks are providing sufficient credit to the economy, and stop worrying too about the effect on demand of cutting the deficit. The only important thing is that money supply expands at a rate compatible with the desired level of nominal GDP growth, or around 5-6 per cent.”

  29. Gravatar of Lee Kelly Lee Kelly
    17. September 2010 at 06:28

    MoneyDemandBlog,

    I thought banks demanded reserves, because … well, a minimum quantity of reserves is legally mandated. Of course, banks also demand reserves to satisfy depositor withdrawals, but normally this is less than what is required. The federal funds rate rises when the difference between the reserve ratio needed to satisfy withdrawals and the legal mimimum increases. In other words, since banks are unable to increase the multiplier by reducing their reserve ratio below the legal mimimum, and so increase the money supply to meet money demand, the Fed must step in and increase the quantity of reserves in proportion.

    Anyway, all else being equal, IOR would encourage banks to hold more reserves and therefore suppress the multiplier. Perhaps all else is not equal: maybe the Fed is offsetting the IOR policy by adjusting the quantity of reserves proportionally. In this case, the Fed would reduce the quantity of reserves when eliminating IOR and thus offset any expansion with an equal contraction. This might be the case, or perhaps these have occurred to some degree of other. I do not know how to assess such claims, but it seems plausible that IOR has been contractionary overall, even if only a little.

  30. Gravatar of 123 – TheMoneyDemand Blog 123 - TheMoneyDemand Blog
    17. September 2010 at 07:11

    Lee Kelly,
    It depends on what you mean by “all else being equal”. If you are keeping fed funds rate constant, IOR means that the Fed must increase the quantity of reserves to hit the fed funds rate target. As Milton Friedman argued in 1959, artificial scarcity of reserves is harmful, so if we keep fed funds rate constant, IOR is stimulative.

    If you are keeping the quantity of reserves fixed, yes, IOR is contractionary. But since the Fed was using Fed funds rate and not the quantity of reserves as the main policy variable in 2008, IOR was expansionary in 2008, and this was the reason Bernanke started the IOR policy.

  31. Gravatar of Lee Kelly Lee Kelly
    17. September 2010 at 07:54

    MoneyDemandBlog,

    Right: the fed funds rate cannot remain constant with positive IOR unless the supply of reserves is increased. If the Fed is targeting a fed funds rate, then IOR is likely to be offset by corresponding changes in reserve supply. But I don’t see how this make IOR expansionary rather than neutral. And it seems misleading to call IOR expansionary when what you really mean is IOR in conjunction with fed funds targeting requires an increase in reserve supply.

  32. Gravatar of 123 – TheMoneyDemand Blog 123 - TheMoneyDemand Blog
    17. September 2010 at 08:13

    Lee Kelly,
    absence of IOR and resulting artificial scarcity of reserves functions as a tax that causes inefficient distortions. The reversal of these distortions is slightly expansionary. The main reason IOR is expansionary is that it facilitates QE and credit easing without cooperation with the Treasury.

  33. Gravatar of Lee Kelly Lee Kelly
    17. September 2010 at 08:38

    MoneyDemandBlog,

    I don’t know what you mean by the “artificial scarcity of reserves”. Since the supply of reserves is limited by what Fed officials can input on a computer or run off the printing press — which is practically unlimited — any scarcity is artificial by design. Reserves are only an economic good so long as the Fed maintains an artificial scarcity to some degree or other. What would a non-artificial scarcity be like? The notion makes little sense to me except, maybe, in the context of free banking.

    I can understand that IOR increases the fed funds rate, because the borrowing bank will need to compensate the lending bank for the interest they would have earned by holding the reserves. To bring the fed funds rate back to its targeted level the Fed will need to increase the supply of reserves. But so far as the multiplier is concerned and the broader money supply, this should be neutral. Why this situation would facilitate QE without the cooperation of the Treasury is also beyond me.

  34. Gravatar of Benjamin Cole Benjamin Cole
    17. September 2010 at 08:40

    JohnLeeMK:

    Yes, I see what you mean.

    Still, when you hear the words “Growth Hawks” do you not think of someone fighting for (not against) growth?

    “Growth Dove” does not have the right ring. In American politics, never allow the words “dove” or “weak” or “soft” be applied to your position.

    Thus QE should be “aggressive” or “innovative” not “easing.” Unfortunately, the word “easing” is part of QE.

    Well, I will keep trying to come up with handy catchphrases for QE.

  35. Gravatar of Benjamin Cole Benjamin Cole
    17. September 2010 at 08:49

    Maybe Sumner has turned the tide.

    This is how Reuters played today’s CPI report. Amazing change in coverage from a couple months back. QE is on the table…being tossed around seriously!

    Growth Hawks, keep sending e-mails, posting in blogs keeping the issue on the front burner. Sumner: Keep posting.

    TREASURIES-Bond prices up on Fed easing hopes
    Published September 17, 2010 | Reuters

    By Ellen Freilich

    NEW YORK (Reuters) – U.S. government securitiesprices rose Friday after a report that core U.S. consumer prices were flat in August appeared to offer the Federal Reserve even more leeway to ease monetary policy further.

    A rise of 0.1 percent in the core Consumer Price Index had been forecast.

    “This just gives the Fed room to do quantitative easing,”said John Canally, economist and investment strategist at LPLFinancial in Boston. “This should put downward pressure on bondyields.”

    The benchmark 10-year Treasury note was up 9/32in price, its yield falling to 2.73 percent from 2.76 percentThursday.

    The 30-year bond , which fell 30/32 Thursday,was up 15/32 Friday after the CPI was released. Its yieldeased to 3.90 percent from 3.93 percent on Thursday.

    The tendency toward soft numbers across almost every major component of the August CPI “leans in the direction of supporting further QE (quantitative easing) action,” said AlanRuskin, global head of G10 FX strategy at Deutsche Bank in NewYork.

    A Fed move toward more quantitative easing would reflectits mandate to promote maximum employment, stable prices, andmoderate long-term interest rates, he said.

    Though the Fed has played down deflation risks, if theprice trends evident in August persist into the November andDecember Federal Open Market Committee (FOMC) meetings, theywill likely be used to support the case for a strategy to avert the risk of entrenched deflation, Ruskin said.

  36. Gravatar of 123 – TheMoneyDemand Blog 123 - TheMoneyDemand Blog
    17. September 2010 at 10:31

    Lee Kelly,
    With IOR reserves are an economic good because the Fed is paying competitive rate of interest, and the only reason marginal unit of reserves has any value is IOR. Without IOR, reserves are valuable because the supply is restricted, and as a result the banking system is less efficient.

    Before IOR Bernanke had to ask Treasury to extend Treasury’s Supplementary Financing Program when he wanted to increase credit easing, as he could not grow the size of Fed’s balance sheet without the help of this program.

  37. Gravatar of JimP JimP
    17. September 2010 at 17:30

    I think one can say that Tim Duy considers Bernanke to be utterly contemptible.

    http://economistsview.typepad.com/timduy/2010/09/the-fair.html

  38. Gravatar of Mark A. Sadowski Mark A. Sadowski
    17. September 2010 at 17:30

    “Could this have something to do with the interest elasticity demand of money gets flatter the closer we get to a zero interest rate?”

    should read

    “Could this have something to do with the fact that the fact that interest elasticity demand of money gets flatter the closer we get to a zero interest rate?”

    Anyone who doesn’t get this doesn’t follow money. It’s similar to the MRS in micro.

  39. Gravatar of Mark A. Sadowski Mark A. Sadowski
    17. September 2010 at 17:37

    “Could this have something to do with the fact that the fact that interest elasticity demand of money gets flatter the closer we get to a zero interest rate?”

    Should read

    “Could this have something to do with the fact that the interest elasticity demand of money gets flatter the closer we get to a zero interest rate?”

    Silly me. I’ll get it right one of these days.

  40. Gravatar of Morgan Warstler Morgan Warstler
    17. September 2010 at 23:22

    Tim Duy is retarded.

  41. Gravatar of Keith Keith
    18. September 2010 at 04:18

    Scott, thanks for the reference.

    If you every experiment with simulation models, the monetary behavior of the link below probably fits Friedman’s descriptions reasonably well.

    http://forio.com/simulate/simulation/keubanks/eubanks-growth-model

    It is a generic growth model designed to explore how growth in population, money supply and government spending might affect capital accumulation, income and employment. Just note that government spending can be set to grow with or independent of income.

  42. Gravatar of scott sumner scott sumner
    18. September 2010 at 05:35

    Bonnie, My hunch is that they aren’t thinking about the side effects that you mention. That’s why we need to give them a much more explicit mandate.

    Jim Glass, That’s interesting, but I agree that on logical grounds money targeting is not an integral part of monetarism. On the other hand, from a historical perspective it certainly has been an integral part of monetarism.

    123, The whole point of abolishing IOR would be to raise NGDP growth, so I don’t understand your assertion that the Fed might do that w/o changing their desired growth path.

    Marcus, Yes, you are right. Actually The Economist has always been very weak on money. Those earlier recessions that they now concede were caused by tight money, were viewed by The Economist at the time as being to to real shocks (tech bubble bursting, etc.) Someday they’ll regard this one as also being due to tight money–but it will be too later.

    Mark, I agree with both of your assertions.

    Good luck at Rowan, We also looked at a Fred graph in class yesterday.

    Morgan, Friedman and I use different paths to reach the same destination.

    Lee Kelly, I find there is always someone who said it earlier. I talked about the analogy several times in 2009, and I think David Beckworth did in 2008.

    Bill, I basically agree, but wouldn’t switch to stable prices until we have futures targeting.

    123, You said;

    “When Bernanke introduced IOR, he flooded the system with reserves, scarcity yield on reserves has disappeared. The net effect of increased IOR and decreased scarcity yield was stimulative.”

    Actually, Bernanke flooded the economy with money before IOR. When IOR was introduced its net effect was contractionary. Even the Fed admitted that in 2009. Bernanke recently cited the removal of IOR as a possible stimulative step, and I am pretty sure markets would respond positively to a decision to eliminate IOR.

    No one claims IOR plus QE was necessarily contractionary, all along I have been claiming that IOR considered in isolation was contractionary.

    W. Peden, Thanks, that’s good to see.

    Benjamin, We’ll know the tide has turned when the stock market starts rising. Come to think of it the recent rise has been associated with the increased talk of QE. Let’s see what happens Tuesday.

    JimP, That’s a very good Tim Duy post.

    Keith, I always have trouble following those complicated flow charts.

  43. Gravatar of Morgan Warstler Morgan Warstler
    18. September 2010 at 06:45

    Scott, he’s driving 4 blocks headed in the same direction you are driving across the state.

    He’d stop QE, the moment it offsets V, so please do the math – when V goes from 1.8 to 1.7, how much more M do you need to keep PY from falling below trend (MINUS THE GREAT RECALCULATION)?

    http://econlog.econlib.org/archives/2009/09/i_deny_the_sign.html

    Because after all:

    http://economicedge.blogspot.com/2010/01/velocity-zero-yet-ripples-propagate.html

  44. Gravatar of 123 – TheMoneyDemand Blog 123 - TheMoneyDemand Blog
    18. September 2010 at 08:20

    Scott, you said:
    “The whole point of abolishing IOR would be to raise NGDP growth, so I don’t understand your assertion that the Fed might do that w/o changing their desired growth path.”
    Here you are assuming your conclusions. In fact, in 2006 Bank of England started IOR without any intention to change the desired growth path. In 2007, there were some proponents of IOR who were worried about inflationary scenarios. In April 2008, just after Bear Stearns, the Fed started thinking about accelerating the 2011 switch to IOR because it was worried about deflationary pressures. In 2010, there are three different cases for zero IOR:
    1. IOR could be temporary reduced in order to reduce fed funds rate and provide more stimulus to economy as suggested by Bernanke
    2. some members of congress would be happy to abolish IOR in order to reduce deficit projections – this would do nothing to the expected NGDP growth
    3. goldbug case – in order to prevent hyperinflation the Fed should abolish IOR and return its balance sheet to the 2006 state

    You said:
    “Actually, Bernanke flooded the economy with money before IOR. When IOR was introduced its net effect was contractionary.”
    There was a 15% increase in monetary base after Lehman before the IOR announcement. To the extent it drove actual fed funds rate below the intended rate, this increase was perceived as temporary and thus it was ineffective. The real monetary base flood came after IOR announcement. IOR announcement was expansionary, because it reassured markets that monetary base will not be withdrawn soon, and because it implied that monetary base will be increased further.

    “Even the Fed admitted that in 2009. Bernanke recently cited the removal of IOR as a possible stimulative step, and I am pretty sure markets would respond positively to a decision to eliminate IOR.”
    The only reason 0 IOR would be stimulative now is the fact that the current ffr target is a range of 0-25bps. In fact, two months ago, in my blog I supported 0bps IOR as one of the good policy options.

    “No one claims IOR plus QE was necessarily contractionary, all along I have been claiming that IOR considered in isolation was contractionary.”
    Even without QE IOR is stimulative if you are keeping intended fed funds rate constant, which you should do if you are examining the Fed in 2008. At this time ffr is a range, but the size of the monetary base is fixed, so the analysis is different now.

  45. Gravatar of scott sumner scott sumner
    19. September 2010 at 07:59

    Morgan, You said;

    “He’d stop QE, the moment it offsets V, so please do the math”

    When NGDP falls below trend, then ipso facto M has not risen enough to offset V.

    123, I don’t see how your first response relates to my question. Bernanke has said that eliminating IOR is one way the Fed could boost AD. Are you arguing that if he does in fact eliminate IOR, it will not be with the intention of boosting AD? Especially given the negative side effects he is worried about? That’s different from the long run efficiency goals that motivated IOR, which I agree had nothing to do with NGDP growth.

    The 15% figure on the base is slightly misleading, as ERs had already skyrocketed. The Fed was worried that this would make policy too easy, so in panic they decided to tighten up to prevent high inflation.

    I don’t understand your final comment, as the fed funds rate was bouncing all over the place in September-October 2008, the fed certainly wasn’t keeping it stable.

  46. Gravatar of 123 – TheMoneyDemand Blog 123 - TheMoneyDemand Blog
    19. September 2010 at 15:26

    “I don’t see how your first response relates to my question. Bernanke has said that eliminating IOR is one way the Fed could boost AD. Are you arguing that if he does in fact eliminate IOR, it will not be with the intention of boosting AD? Especially given the negative side effects he is worried about? That’s different from the long run efficiency goals that motivated IOR, which I agree had nothing to do with NGDP growth.”

    If Bernanke suspends IOR in 2010, it will boost AD by lowering fed funds rate to the bottom of the target range. But Bernanke’s recent comments are not applicable to 2008. In 2010 it is OK to assume that the size of the monetary base is fixed, in 2008 the appropriate assumption was that the intended fed funds rate is fixed.

    “The 15% figure on the base is slightly misleading, as ERs had already skyrocketed. The Fed was worried that this would make policy too easy, so in panic they decided to tighten up to prevent high inflation.”
    I don’t think the phrase “tighten up” fits here, it should be reserved for those cases where the Fed shifts market expectations to tighter side. Yes, ERs had already skyrocketed, but markets did not think that this injection of monetary base is permanent. Markets thought that monetary base will shrink to the size that is consistent with market expectations of future path of fed funds rate. Skyrocketing ERs in September and early October did not cause the opportunity cost of reserves to fall, according to the market expectations of future fed funds rate.

    Some time ago you said that IOR might have pushed the economy to the other side of knife edge equilibrium. A quick look at the chart of 3 month LIBOR should persuade you that the economy was never close to that equilibrium. Those excess reserves were not really excess – the demand for liquidity caused by the failure of Lehman has dwarfed the regulatory reserve requirements. Bernanke should have printed much more reserves in late September than he actually did.

    “I don’t understand your final comment, as the fed funds rate was bouncing all over the place in September-October 2008, the fed certainly wasn’t keeping it stable.”
    While actual fed funds rate was bouncing around, intended fed funds rate was stable. What matters are the market expectations of the future fed funds rate, they tracked intended fed funds rate. IOR had no adverse impact on the expected path of future fed funds rate. IOR was inflationary in 2008, as it had a positive impact on the expected volatility of effective fed funds rate.

  47. Gravatar of Dude, Where’s My Central Bank? Dude, Where’s My Central Bank?
    19. September 2010 at 21:01

    [...] seems intent on letting such expectations steadily fall. I just don’t get it. And I bet that Milton Friedman wouldn’t get it either if he were alive today. The Fed cannot solve all of our problems, but it can stabilize [...]

  48. Gravatar of ssumner ssumner
    20. September 2010 at 05:49

    123, You said;

    “In 2010 it is OK to assume that the size of the monetary base is fixed, in 2008 the appropriate assumption was that the intended fed funds rate is fixed.”

    Here’s why I don’t agree. In 2008 the fed felt it absolutely had to injec tlots of funds into the banking suystem, to provide needed liquidity after Lehman failed. Tehy were going to do that no matter what. The pace at which they were going to reduce the ff target was a secondary issue. W/O IOR, the actual fed funds rate would have fallen even faster, which would have been good. Also, inflation expectations would have been higher, which is a completely separate channel from the fed funds rate. Thus the fed funds rate does not completely describe the stance of policy.

    You said;

    “Some time ago you said that IOR might have pushed the economy to the other side of knife edge equilibrium. A quick look at the chart of 3 month LIBOR should persuade you that the economy was never close to that equilibrium. Those excess reserves were not really excess – the demand for liquidity caused by the failure of Lehman has dwarfed the regulatory reserve requirements. Bernanke should have printed much more reserves in late September than he actually did.”

    Let’s suppose you are right, as we moved into October and November the crisis became less acute, the IOR may have been contractionary. Recall that the IOR didn’t start paying interest until October 8th. I believe that the next several months were crucial.

    Having said that, I certainly agree that the Fed should have been much more expansionary in September, before IOR was even implimented.

  49. Gravatar of 123 – TheMoneyDemand Blog 123 - TheMoneyDemand Blog
    20. September 2010 at 06:51

    Scott,
    you said:
    “In 2008 the fed felt it absolutely had to injec tlots of funds into the banking suystem, to provide needed liquidity after Lehman failed. Tehy were going to do that no matter what. ”
    Trichet also felt he had to lend lots of money to Greece and other PIGS. However he did that without expanding monetary base. Without IOR, the Fed would have used Treasury Supplementary Financing Program in order to lend money to banks.

    You said:
    “The pace at which they were going to reduce the ff target was a secondary issue. W/O IOR, the actual fed funds rate would have fallen even faster, which would have been good.”"
    W/O IOR, the effective ffr was sometimes below target, sometimes above target. Only after IOR the Fed has prevented the damaging spikes of actual fed funds rate. Before IOR the quantity of excess reserves was insufficient.

    You said:
    “Also, inflation expectations would have been higher, which is a completely separate channel from the fed funds rate. Thus the fed funds rate does not completely describe the stance of policy.”
    Inflation expectations were crashing because 3 month LIBOR was soaring. Only after IOR the Fed succeeded in bringing 3 month LIBOR down, albeit in a pace that was way too slow.

  50. Gravatar of 123 – TheMoneyDemand Blog 123 - TheMoneyDemand Blog
    20. September 2010 at 09:23

    You said:
    “Recall that the IOR didn’t start paying interest until October 8th. I believe that the next several months were crucial.

    Having said that, I certainly agree that the Fed should have been much more expansionary in September, before IOR was even implimented.”

    In the second half of September and the first week of October, the natural rate was falling down and short term rates were going up. Later in October and November, natural rate was falling further, short term interest rates were falling down, but the gap between them remained.

  51. Gravatar of ssumner ssumner
    21. September 2010 at 07:30

    123, I don’t think we’ll ever agree on the likely counterfactual. But in a sense nothing I am arguing really hangs on it. I feel I have a right to consider the IOR in isolation for analytical purposes, regardless of what policy combination the Fed would had actually done.

    Regarding the volatility of the IOR, it seems to me that your argument actually supports my point–they weren’t targeting the fed funds rate.

    The annoucement and implementation of the IOR was associated with falling inflation expectations.

  52. Gravatar of 123 – TheMoneyDemand Blog 123 - TheMoneyDemand Blog
    22. September 2010 at 04:12

    Scott, you said
    “I don’t think we’ll ever agree on the likely counterfactual.” and “Regarding the volatility of the IOR, it seems to me that your argument actually supports my point–they weren’t targeting the fed funds rate.”

    The first relevant counterfactual is counterfactual that was embedded in market expectations before the IOR annoucnement. Data from Treasury market and fed funds rate future market show that markets were expecting a path of short term interest rates that was too tight. Until December 2008 markets thought that the Fed is targeting fed funds rate. Only in December 2008 ZIRP was priced in, and only from that point it makes sense to talk about the benefits of 0bps IOR.
    The volatility of effective fed funds rate shows that the Fed has lost the control of monetary policy levers, it does not show that the Fed has abandoned fed funds rate target.

    The second counterfactual that is perhaps even more important than the first one relates to the best case scenario. I guess in September 2008 Greenspan would have expanded monetary base via credit easing until 3 month Libor would have reached 1.0%. This way Lehman panic would have caused only a little bit more damage than LTCM has caused in 1998. I think that setting 3 month Libor to 1.0% would have required at least the doubling of monetary base, while the actual increase growth of monetary base was below 15% and was not sufficient to fill the increased money demand that was caused by Lehman panic. Now there were two ways to set 3 month Libor to 1.0%. The first option is to target 3 month Libor directly, like the Swiss central bank is doing, in this case fed funds rate target is formally abandoned and no IOR is necessary. The second option was to cut the Fed funds rate to 0.5% in September 2008 and use IOR to achieve the Fed funs target while using credit easing to achieve the 3 month LIBOR target. Both of these two options would have successfully prevented the Great Recession. It is a mistake to think that ZIRP (no IOR) with only 15% growth in monetary base would be sufficient to prevent that.

    You said:
    “The annoucement and implementation of the IOR was associated with falling inflation expectations.”
    Monetary policy was a bit less tight after IOR announcement, but it still was too tight.

  53. Gravatar of ssumner ssumner
    22. September 2010 at 16:28

    123, I think you are responding to views I don’t have. I don’t recall arguing for any specific increase in the monetary base, rather I argued for whatever it took to keep NGDP expectations on track. So I don’t really disagree with your assertion that it would have required doubling the monetary base, although I don’t agree either. I simply have no opinion on how much base money would have been necessary in the absense of IOR. I may have speculated that a smaller amount would have been necessary, but that would have been pure speculation, as the amount is extremely senstive to inflation expections, and those of course also depend on other types of Fed communication.

    October 1 to 10 was one of the largest stock market crashes in modern times. The policy was informally agreed to on October 3rd, announced on the 6th, and implimented on the 8th. Obviously there may be no connection, but given the market fell every single one of those days, often sharply, I doubt it was a positive influence.

    I have trouble understanding why the fed funds rate was important, given that the actual fed funds rate was not at the target.

  54. Gravatar of 123 – TheMoneyDemand Blog 123 - TheMoneyDemand Blog
    23. September 2010 at 04:17

    Scott, you said:
    “I think you are responding to views I don’t have. I don’t recall arguing for any specific increase in the monetary base, rather I argued for whatever it took to keep NGDP expectations on track. So I don’t really disagree with your assertion that it would have required doubling the monetary base, although I don’t agree either. I simply have no opinion on how much base money would have been necessary in the absense of IOR. I may have speculated that a smaller amount would have been necessary, but that would have been pure speculation, as the amount is extremely senstive to inflation expections, and those of course also depend on other types of Fed communication.”

    We have a very different ideas about relative sensitivity of NGDP expectations to the size of monetary base vs. IOR.

    You think that erratic effective fed funds rate is preferrable to stable effective ffr achieved by IOR. I think it is completely opposite. I believe the opportunity cost of reserves was extremely high in late September 2008 – 3 month LIBOR was 4,2% on October 02 2008 (elevated LIBOR was a reflection of high demand for monetary base). You believe that the opportunity cost of reserves was very low on on October 02 2008, as effective fed funds rate was 0,67%. The fact is that the Fed has lost control of monetary conditions in September 2008, and I view IOR as an attempt by the Fed to regain control and lower the opportunity cost of reserves by switching to a new improved IOR-based policy framework. Slowly it worked, from October 10, 2008, the opportunity cost of reserves started slowly descending (as measured by 3 month LIBOR).

    You said:
    “October 1 to 10 was one of the largest stock market crashes in modern times.”
    According to my analysis, the actual opportunity cost of reserves was increasing during this period, while the opportunity cost that was consistent with stable NGDP expectations was falling during this period. No wonder markets crashed. The crash stopped when new policy framework has started to reduce the opportunity cost of reserves.

    You said:
    “I have trouble understanding why the fed funds rate was important, given that the actual fed funds rate was not at the target.”
    Daily effective fed funds rate had absolutely no importance during that period. The expected future cost of money was important, and it was driven by expected future actions of FOMC plus the risk premium driven by the fear of erratic spikes of effective fed funds rate above the levels set by FOMC. The price of money was never below the FOMC target.

  55. Gravatar of Jon Jon
    26. September 2010 at 10:11

    The basic appeal of monetary aggregate targeting was always the SIMPLICITY of it. Computing a price index is difficult. There are many many independent transactions the details of which are off-book from the government’s perspective. So, the data simply isn’t easily available, and is a bit questionable.

    Meanwhile, we have a highly regulated banking system. Those banks already have a significant regulatory compliance burden and already disclose to the government the information necessary to construct the common monetary aggregates. Plus, the categories of accounting are simple and stable.

    That is or was the appeal. Its been a false promise.

  56. Gravatar of ssumner ssumner
    26. September 2010 at 12:07

    Jon, Yes, that about sums it up.

  57. Gravatar of What Would Milton Friedman Say About Monetary Policy Today? : Invest My Money What Would Milton Friedman Say About Monetary Policy Today? : Invest My Money
    26. October 2010 at 23:32

    [...] recent example.  Our article also draws from Scott Sumner who has made similar arguments  before  about Milton Friedman. [...]

  58. Gravatar of Debt Ceiling Part II | Really Important Things Debt Ceiling Part II | Really Important Things
    23. July 2011 at 20:42

    [...] in this belief by arguments that the paragon of free market capitalism himself, Milton Friedman, may well have been in favor of fiscal stimulus over the last few years, had he not passed [...]

  59. Gravatar of Free-market and inflation | INFORMATION Free-market and inflation | INFORMATION
    25. July 2011 at 22:56

    [...] of economics at Bentley University who identifies himself as a “neo-monetarist”, has argued that Friedman would have supported monetary stimulus. And he has argued, on neo-Friedmanite [...]

  60. Gravatar of NFA Trust Texas NFA Trust Texas
    8. October 2012 at 08:02

    The original contribution by Milton Friedman is dated December 17, 1997, and that rings a bell with me immediately: The timing mattered then and he was right in this regard, i.e. short-term disequilibrium.NFA Trust Texas

Leave a Reply