The anti-Friedman

John Cochrane and Milton Friedman shared many similarities.  Both were strong supporters of free market policies.  Both were opposed to high inflation, and skeptical of Fed fine-tuning.  Both taught at Chicago.  Yet their views of monetary policy could hardly be more dissimilar.  Indeed Friedman’s views were closer to those of Paul Krugman than John Cochrane (and much closer to my views than either Krugman or Cochrane.)

Friedman was part of a long tradition in macroeconomics, which argued that nominal shocks have a real effect in the short run, but only lead to inflation in the long run.  The evidence supporting this approach is pretty overwhelming, partly due to the work of Friedman and Schwartz, but also reflecting many other researchers.  I’ve devoted much of my life to the Great Depression, and found lots of evidence that Friedman and Schwartz overlooked.  Others found evidence that countries began to recover once they left the gold standard.  And so on.  All mainstream discussion of macro starts with the premise that nominal shocks have real effects.  Except John Cochrane.  Or perhaps I should say that I don’t quite know where Cochrane stands on this issue.  Reading him you sometimes have the feeling that he thinks really severe nominal shocks might matter, but not small ones.  Or that nominal shocks matter a little, but not enough to be interesting.  So I’m not quite sure where he stands.  He seems to think the Fed was wise to prevent deflation in 2009 (actually they didn’t entirely prevent it, but he’s right that they quickly reversed it, unlike in 1929-33.)  But I’m not quite sure why he thinks the Fed was wise to prevent deflation.

In a recent piece discussing Michael Woodford, he makes some unhelpful disparaging remarks about “Phillips Curve” economics.  Of course there are lots of problems with Phillips Curve economics, but that has little or no bearing on whether nominal shocks (changes in NGDP) have real effects.  Phillips Curve economics fell in disrepute during the 1970s, mostly because it didn’t predict very well.  The problem is simple, the Phillips Curve assumes not just that nominal shocks have real effects, but also that changes in the rate of inflation are a good indicator of nominal shocks.  But of course they aren’t, inflation can reflect either demand (nominal) shocks or supply (real) shocks.  Never reason from a price change.  Only NGDP can reliably indentify nominal shocks.

Reading Cochrane one has the sense that discrediting the Phillips Curve is equivilent to discrediting the mainstream view that nominal shocks have real effects.  Not true, Milton Friedman strongly believed that nominal shocks have real effects, but was skeptical of Phillips Curve macroeconomics.

There’s nothing wrong with being outside the mainstream (in many ways I’m even further out there.)  But I see all sorts of other problems with Cochrane’s analysis.  Milton Friedman would be horrified by Cochrane’s equation of low rates and easy money, indeed Friedman argued (correctly) that ultra-low rates are usually a sign that money has been very tight.  Cochrane’s not the only one who makes this assumption, indeed in this recent post I argued that 99% of economists now seem to agree with him.  So now I’m the outlier.  Nonetheless Friedman and the Bernanke of 2003 and I are right that low rates don’t mean easy money, and 99% of economists including the Bernanke of 2012 are wrong.

Cochrane also claims that Woodford doesn’t believe that a promise of easier money in the future can have any effect on current NGDP.

Many people (and a few persistent commenters on this blog!) urge nominal GDP targeting by looking at a graph like this and saying “see, if the Fed had kept nominal GDP on trend, we wouldn’t have had  such a huge recession. Sure, part of it might have been more inflation, but surely part of a steady nominal GDP would have been less recession.” This is NOT what Mike is talking about.

Mike recognizes, as I do, that the Fed can do nothing more to raise nominal GDP today. Rates are at zero. The Fed has did what it could. The trend line was not achievable.

I hope and believe that Cochrane is wrong on this point.  I had read Woodford as arguing that current AD is very much impacted by expectations of future AD, and hence future monetary policy.  That level targeting would have kept us closer to the trend line.  It would be interesting to know what Woodford thinks of Cochrane’s claim.  Woodford’s no fool, and I very much doubt that he made the elementary error that Cochrane claims he made here:

But nobody has a Phillips curve in which raising expected inflation is a good thing.  It just gives you more inflation, with if anything less output and employment. Read Mike’s book!

My biggest problem with Cochrane’s whole approach is that it seems completely inconsistent with the EMH, which is supposed to be an important tenet of Chicago School economics.  I watch market reactions to monetary policy very closely, and in both the 1930s and the 2000s it is quite obvious that the markets are “market monetarist” and completely reject Cochrane-style monetary economics:

1.  Markets think QE is very important.  Cochrane focuses on interest rate responses to monetary announcements.  I’d be glad to discuss those, but quite frankly that’s the last place one should look.  Because the liquidity effect moves rates in the opposite direction from the income and inflation effects, the results are often ambiguous.  But the stock, commodity and forex markets are completely unambiguous.  Easy money makes the dollar depreciate.  Easy money makes commodity prices rise.  Easy money has a strongly powerful impact on stock prices when NGDP is depressed, and a much more ambiguous effect when NGDP is not depressed.  That’s because stocks are providing an indicator of the impact of monetary policy on real growth.  And the impact of easy money is much stronger when NGDP is depressed than when it’s not depressed.

2.  QE has an obvious and powerful effect on markets, which tends to refute both Woodford’s and Cochrane’s view of monetary economics.

But Cochrane understands there’s a problem here, and is thus forced to develop a rather ungainly theory that seems (to me) to be completely inconsistent with the EMH:

On many occasions Fed announcements, coupled with no actions, do move markets. Monika Piazzesi and I once looked at high-frequency data and came to the same conclusion.

But these what do we make of this fact? They certainly do not mean that the Fed can talk down rates at its pleasure. Mike briefly acknowledges one possibility: Markets do not interpret these announcements as changes in policy, or “intentions” but instead simply inform the markets of the Fed’s deteriorating economic forecasts.   If the Fed gets news, or forms an opinion, that the economy will be weak, then future interest rates will be lower even if the bank follows the same old Taylor rule.  We can see this reaction even if the central bank has no influence at all over market interest rates (as in Gene Fama’s latest) but has a decent forecasting shop. A coming recession means that interest rates will fall no matter what the Fed does about it, so long term rates fall now. Mike has a long section on open mouth operations that don’t work, or go the wrong way, and pages of advice for central bankers on how to move markets the way they want.

Here’s where my critique of Cochrane would probably diverge from people like Krugman.  I’m horrified by Cochrane’s apparent rejection of the EMH.  How can someone seriously argue that markets respond strongly to monetary announcements because they think the Fed sees the business cycle better than the markets do?  Markets are far smarter than the Fed.  Indeed the Fed often doesn’t realize they have a problem until the market tells them (as in late 2008.)  But it’s even worse, as the data completely refute Cochrane’s hypothesis.

It’s best to focus on the big announcements, the ones that were unanticipated and significantly moved markets, such as January 2001, September 2007 and December 2007.  Days when the stock market moves 2% to 5% within minutes of the announcement.  When bonds and commodities and forex markets are strongly impacted.  And in all three of the big surprises I just cited, the market response (other than short term rates) was exactly the opposite of what Cochrane’s model would predict:

1.  In January 2001 there was a bigger rate cut than expected (expansionary shock.)  Cochrane’s hypothesis implies it should have showed the markets that the Fed was worried about recession.  That’s true, but the markets were already much more worried about recession.  They took the expansionary surprise as “good news” and stocks soared immediately by 5%.  Long term bond yields also soared, as investors marked up their estimates of future NGDP growth.

2.  In September 2007 there was again a bigger rate cut than expected (expansionary shock.)  Same qualitative market reactions.

3.  In December 2007 there was a smaller rate cut than expected (contractionary shock), and both stocks and long term bond yields plunged.

Often Fed announcements are close to expectations, and the market reaction is muted.  But when there is a big surprise, the market reaction is exactly the opposite of what the Cochrane explanation would predict.  Markets are Friedmanites, not Cochranites.  Markets favor monetary stimulus when recent NGDP growth has put the line far below trend.  That’s not to say markets always favor easy money, high inflation hurt stocks in the 1970s.  Rather that they favor easy money when Milton Friedman would have favored easy money.  David Glasner has documented this pattern in a study that examines the time-varying correlation between TIPS spreads and equity prices.

After spending much of my life studying the Great Depression, none of this came as a surprise to me.  The big market reactions to falling NGDP growth in recent years are a just a pale echo of the even bigger market reaction to the much bigger fall in NGDP during the early 1930s.  And the positive market responses to hints of monetary stimulus were even larger, as the stakes were much higher.  The biggest 2 day rally in US stock market history occurred in February 1932.  It reflected a Hoover proposal to reduce the gold backing of money (a lower gold ratio), and (the following day) strong support in Congress for the proposal.  This allowed the Fed to undertake QE.  Was the market reaction a coincidence?  As far as I can tell both Woodford and Cochrane would claim it was coincidence, after all, rates were near zero so QE obviously could not have any impact.  Except the markets don’t believe these interest rate-oriented models, and neither do I.

In the near future I’ll do a post explaining why Woodford and Krugman are wrong about QE.  Right now I’m too jet-lagged to tackle the Woodford paper.

HT:  Bill Woolsey, and several commenters.

Update:  A commenter named RPLong asks the following:

Reading Cochrane one has the sense that discrediting the Phillips Curve is equivilent to discrediting the mainstream view that nominal shocks have real effects.  Not true, Milton Friedman strongly believed that nominal shocks have real effects, but was skeptical of Phillips Curve macroeconomics.

This paragraph sums up the whole post to me. It’s fair to disagree with Cochrane, but are you sure you’re being fair to his point of view? You are basically saying “It SEEMS like Cochrane is making Argument X, which I can thoroughly defeat as follows…”

Are you sure you’re not just straw-manning Cochrane?

No I’m not at all sure.  That’s why I hope that Cochrane will clarify his views on nominal shocks.   I’ve defended Cochrane when his views were attacked by Krugman and DeLong.  But even there I had to defend him by pointing out that his views were not clearly explained.  He seemed at one point to suggest fiscal policy could have no effect, then later said it could only work by affecting velocity.  It seems to me he has the same ambiguity in his comments about nominal shocks.  At times he implies that nominal shocks may have some real effects, at other times he seems skeptical.


Tags:

 
 
 

34 Responses to “The anti-Friedman”

  1. Gravatar of Saturos Saturos
    5. September 2012 at 06:43

    Scott, when and why did you first begin to believe in the efficiency of markets at capturing social knowledge?

  2. Gravatar of Nick Rowe Nick Rowe
    5. September 2012 at 06:48

    Not at all bad for a jet-lagged Scott!

  3. Gravatar of RPLong RPLong
    5. September 2012 at 06:48

    Reading Cochrane one has the sense that discrediting the Phillips Curve is equivilent to discrediting the mainstream view that nominal shocks have real effects. Not true, Milton Friedman strongly believed that nominal shocks have real effects, but was skeptical of Phillips Curve macroeconomics.

    This paragraph sums up the whole post to me. It’s fair to disagree with Cochrane, but are you sure you’re being fair to his point of view? You are basically saying “It SEEMS like Cochrane is making Argument X, which I can thoroughly defeat as follows…”

    Are you sure you’re not just straw-manning Cochrane?

  4. Gravatar of Becky Hargrove Becky Hargrove
    5. September 2012 at 07:11

    Glad you’re back! Good post.

  5. Gravatar of Saturos Saturos
    5. September 2012 at 07:30

    Uh, yeah, Michelangelo, another good sculpture there I see…

  6. Gravatar of DanC DanC
    5. September 2012 at 07:33

    What do you think of Fama’s recent work?

    “I find that The Federal funds rate, FF, moves strongly toward the Fed’s target, TF, but other rates show little day-to-day convergence to TF. When the Fed changes TF, it moves toward existing short rates. This suggests a passive Fed that follows the market, but it is also consistent with an active Fed that controls rates and rates adjust to reflect predictable changes in TF. When TF changes, short rates move toward the new TF. This is consistent with a Fed that controls short rates or a Fed that has no control but is an informed investor whose signals affect rates.
    http://papers.ssrn.com/sol3/papers.cfm?abstract_id=2124039

  7. Gravatar of Matt Rognlie Matt Rognlie
    5. September 2012 at 07:50

    I just don’t think Cochrane’s characterization of Woodford as saying that the Fed “can do nothing more to raise nominal GDP today” is accurate. Woodford believes that a commitment to a (higher) NGDP target path will raise nominal GDP today, because the looser monetary policy of the Fed in the future impacts expectations today.

    Now, Woodford is more skeptical of the power of an NGDP target than you are; in particular, although he thinks it would help, he doesn’t think a level target completely eliminates the problem of the zero lower bound. I agree with him, and I suspect that we have similar reasons for this belief. In Woodford’s model, one can prove that the zero lower bound will still pose a barrier in some situations even when there is an NGDP level target, and the central bank will not immediately be able to move the economy to its target. There are various reasons why his model might not be a complete description of reality, but I think it is a pretty good benchmark, and the burden is on you NGDP zen masters to come up with a modification of the model (or alternative model) where the zero lower bound is no longer a barrier. : )

    The spike in stock markets in 1932 (which is a good historical anecdote by the way, thanks) is easily interpreted in Woodford’s model, because decreasing the gold backing of currency was an implicit commitment to easier policy in the future. He tends to think of “policy” more as a contingent commitment to a path for interest rates, while you think of policy as a contingent commitment to a path for NGDP, but formally these are two equivalent ways of describing the same phenomenon. (For what it’s worth, I think that due to various institutional changes, the NGDP view is a nicer way to think about Depression-era monetary policy, while the interest rate view is a nicer way to think about modern monetary policy.)

  8. Gravatar of Morgan Warstler Morgan Warstler
    5. September 2012 at 08:12

    Scott, I don’t think you are being truly logical here.

    And it goes to something I can never get you to really talk about.

    There is a way that we could get Cochrane to like NGDLPT immediately:

    1. Set NGDPLT target at say 3%.

    2. Start it today. OOPS! we’re over the target!

    3. Time to raise rates!

    —-

    The point here is that NGDPLT is capable of being PERFECTLY COMFY with a view that right this minute money is too loose.

    You just have to not have the same set of personal values about what the nature of the problem is…

    But a bunch of hard money guys could 100% support the NGDPLT above.

    —–

    Now here’s where I don’t think you are really exploring the logic of such a situation.

    What happens if we set it at 3% and have to raise rates today and ekep raising them until we are averaged out at 3%???

    Well according to you and DeKrugman et al, the economy will crater, right?

    Which drops us down below 3%, and suddenly it is time to print.

    EXCEPT everyone KNOWS that we’re gonna do some make up and at 3% Chuck Norris will come out and kick our ass, so people don’t get very excited.

    ——

    Here’s my pint:

    Ultimately accomplishing NGDPLT is a trade, the stakeholders, the hegemony, are going to have to be wooed.

    But for sure WOOING can get you NGDPLT.

    And in my mind, GETTING NGDPLT is the MAIN THING. Getting it running, getting everybody to KNOW WHAT IT IS, as soon as possible is MORE IMPORTANT than the target rate.

    —–

    I say this for a completely personal observation: you and Bill Woolsey are good buddies.

    You have been long time compatriots who have talking about this shit for long before anyone else.

    And he’s at 3% and you’re at 5%.

    But neither of you really care, both of you are more concerned with getting the rules changed in the game, which means to me that:

    BOTH OF YOU kind of believe inside that the target rate isn’t the most important.

    Tell me if I’m wrong.

    Since I’m right, then the question is WHY NOT go woo the opposition with a NGDPLT system that makes rates go up right away?

    The magic here isn’t getting back to some historic trend. The magic is getting the damn shocks to stop killing us, to get the Fed’s power dialed back, and get a sustainable growth going forward.

    It might take 10 YEARS to do it your way. It might take 1 to do it mine.

    I’m not saying this is right, I’m saying WHY isn’t there a calculation on this being made and discussed?

    If Cochrane CAN BE WON OVER, the question becomes why no discuss what it would take to win him?

  9. Gravatar of Cthorm Cthorm
    5. September 2012 at 08:34

    Except the markets don’t believe these interest rate-oriented models, and neither do I.

    I think most market actors believe these interest rate-oriented models, but they don’t act like it. Much to my dismay. Instead they try to throw together a bunch of other obfuscations to keep their rate-oriented view, like a focus on central bank balance sheets. I think more people are understanding that there is a much better, clearer model and it looks to NGDP.

  10. Gravatar of Tomasz Wegrzanowski Tomasz Wegrzanowski
    5. September 2012 at 08:46

    > which argued that nominal shocks have a real effect in the short run, but only lead to inflation in the long run. The evidence supporting this approach is pretty overwhelming

    And according to overwhelwing evidence, how many year is “short run” and “long run” approximately?

    Japanese short run lasts 20 years already and shows no indication of ever ending. Persistently suppressed demand ends up having pretty severe long term effects.

    All this might be true if you only consider one-off shocks (and then NGDP/inflation back to some reasonable %/year growth trend), or high inflation/NGDP growth (which is easy to adjust to), but it seems that persistently 0% NGDP growth, or persistent deflation causes a new shock every year – faster than economy can ever adjust.

    This never-ending tightening seems to be ECB’s economic policy for 21st century as well.

  11. Gravatar of Ritwik Ritwik
    5. September 2012 at 08:53

    Scott

    As I understand you, your preferred ‘mechanism’ of monetary policy transmission is asset prices. You pay lip service to the hot potato, but the hot potato is not a very ‘market monetarist’ phenomenon.

    You are right to include commodities, Fx and stock prices in the assets through which this might work, rather than just credit. But things get complicated there.

    Does an increase in commodity prices boost consumption and production? I’d imagine no, overall. Commodities that are traded as assets are input commodities – they boost producer’s profit margins but are problematic for finished goods manufacturers as well as final consumers.

    Does an increase in stock prices boost consumption and investment? It should by all theories, but the evidence seems to be weak, esp. for this recovery. Mostly because of

    1) The empirical evidence for Tobin’s q has always been flimsy.

    2) the low marginal propensity of spending of financial wealth holders (who are actually able to access their wealth).

    Does a decrease in the exchange rates boost spending? For the US, I’d say almost unequivocally yes.

    Of the three asset mechanisms, only Fx seems a surefire way. Commodities is most probably contractionary. And a lot of the effects of any such announcements are aided or constrained by the heterogeneity of agents – monetary policy cannot possibly be distributionally neutral.

    Does all this not give you any pause about the omnipotence of a central bank?

  12. Gravatar of Greg Ransom Greg Ransom
    5. September 2012 at 09:22

    Philips himself rejected “Philips Curve” economics because IT’S BAD ECONOMICS.

    The assumption that Friedman Instrumentalism & “Prediction” is fundamental is bogus — bad economics is bad economics out of the box, and everyone SHOULD HAVE KNOW IT.

    A non pathological science wouldn’t have required to the human destruction of the policies of the 1960s & 1970s to do science right from the beginning.

    Scott writes,

    “Phillips Curve economics fell in disrepute during the 1970s, mostly because it didn’t predict very well.”

  13. Gravatar of Patrick R. Sullivan Patrick R. Sullivan
    5. September 2012 at 09:30

    It’s Cochrane who writes as if he suffers jet lag. I didn’t read any further than this;

    ‘Second, drop money from helicopters, i.e. “coordinated monetary-fiscal policy.” Basically, the Treasury borrows money, writes checks to voters (“helicpoters”), and the Fed buys the debt.’

    Last week he denied to me that helicopter drops were monetary policy, but only fiscal policy. Now it’s a ‘coordinated’ policy.

    But, he’s got the transmission mixed up. Should be that the Treasury issues bonds (borrows), the Fed buys the bonds with bookkeeping entries (creates more money), the Treasury then withdraws currency based on those entries, and finally showers that from helicopters (and I doubt they can only target voters).

    What would prevent the recipients who happened to be under the helicopters from spending the new money on goods and services? Low interest rates?

  14. Gravatar of Doug M Doug M
    5. September 2012 at 09:31

    Low rates and easy money…

    You have said several times that low rates are a sign that money was tight rather than that it is currently easy. I will give you that if we are talking about nominal rates. What if we are talking about real rates. I would say that negative real rates are unambiguously easy.

    As for the rest of the takedown of Cochrane.

    Cochrane says that QE does not work, and indeed this flies in the face of Freidman’s theory. Cochrane suggests V is not stable any increase in the money supply is counteracted by a decrease in velocity. However, I have always thought that Freidman’s assertion that V is stable, as pretty darn suspicious.

  15. Gravatar of Assorted Links « azmytheconomics Assorted Links « azmytheconomics
    5. September 2012 at 09:35

    […] Woodford’s new macro paper. Angus, Sumner, and Cochrane […]

  16. Gravatar of Razer Razer
    5. September 2012 at 09:49

    So you’re comparing different schemes of inflationists? Will the central planners ever cease to long for more planning?

  17. Gravatar of Can the Fed commit? « azmytheconomics Can the Fed commit? « azmytheconomics
    5. September 2012 at 11:04

    […] Further reading: Sumner comments. […]

  18. Gravatar of K K
    5. September 2012 at 11:26

    Scott: “Markets think QE is very important.”

    There’s a whole section of Woodford’s paper dedicated to showing that those market responses are due to expectations of future rates policy, and that QE itself actually has nothing to do with it. He even goes all in on Wallace neutrality, arguing that the even the nature of assets purchased is irrelevant. While I expected his argument against “pure QE” (he says that Japan shows that it can be unwound in a few months so it cannot have any consequences for post rate hike policy), he is the only economist other than Williamson that I know of who has taken such a strong position on actual, real-life balance sheet neutrality. Fun, but while he’s right that the Wallace result is resistant to a wide variety of agent preferences and frictions, I’m quite sure it’s not resistant agent balance sheet constraints or divergent expectations. If you don’t assume Wallace, that doesn’t mean you get the Market Monetarist results, of course. It just means it’s more complicated than both sides would prefer.

    And Cochrane is *really* out to lunch if he thinks Woodford thinks the Fed can’t raise NGDP today (unless he means literally “today!”). The whole point of the paper is that the Fed need to use forward guidance to stimulate the economy. As in stimulate it *now*.

    “As far as I can tell both Woodford and Cochrane would claim it was coincidence, after all, rates were near zero so QE obviously could not have any impact.”

    No. Woodford says it’s via the expectations channel. Read the paper. In this case, keeping the gold peg would have demanded high future interest rates. Unpegging allowed future rates (and expectations thereof) to drop. That was the stimulus. There is nothing weird here for standard theory.

  19. Gravatar of Major_Freedom Major_Freedom
    5. September 2012 at 11:56

    As usual, the cart is being placed before the horse with “nominal demand” and “real effects” in market monetarist analysis.

    Nominal demand “shocks” do not take place in the market for no reason. These “shocks” are not the datum by which we judge the path of causality in economic phenomena. “Nominal demand” follows “real effects.” I spend my money on a sandwich now only because the real economy is such that allows the purchase and hence nominal demand for sandwiches. I cannot purchase what is not already available in real terms.

    Production of real goods and services (consumer and capital goods) precedes nominal spending on them. If there is to be a “nominal demand shock”, then the key question is not “What real effects will this nominal demand shock have?”, but rather “What real factors lead to this nominal demand shock?”.

    At first blush this may seem the wrong way to go about it, due to fact that the Fed really can affect aggregate nominal demand unilaterally. The thinking goes: If the Fed can create any quantity of new dollars essentially at will, then aggregate nominal demand really should be the datum for analyzing causality in economic phenomena, and we really should be asking what real effects the Fed brings about by, say, letting NGDP growth fall to zero rather than having it grow at a constant rate every year.

    The problem with that thinking is that it does not address the real factors that would have changed nominal demand without the Fed’s change in money inflation, after which (or during which) the Fed creates fewer dollars or more dollars than they did before, such that nominal demand does not fall or rise from where it was before.

    In other words, no market monetarist is asking why nominal demand should fall so drastically in 2008 despite the fact that the Fed didn’t destroy dollars. Most MMs explain the drastic fall by saying that there was an increase in the demand for cash, and the Fed did not create enough new dollars to maintain NGDP. Yet this explanation is tautological. A fall in NGDP means the same thing as a rise in demand for money, all else equal. So this isn’t an answer to the question of why NGDP would fall (or why demand for money would rise) in 2008, despite the fact that the Fed was not destroying dollars.

    So the question remains for MMs to explain: Why would NGDP fall (or, equivalently, why would demand for money rise), despite the fact that the Fed was not destroying dollars? In other words, why was it necessary for the Fed to accelerate its creation of dollars in 2008 in order for NGDP to be kept from falling?

    The answer “There was an increase in the demand for money that was not “properly” responded to by the Fed, and we won’t explain why there was suddenly an increased demand for money” is not a sufficient, satisfactory answer.

    The wool is being pulled over the eyes of the readers of this blog. They are being constantly presented with a self-fulfilling prophecy that presupposes inflation as the answer. By presenting “the problem” as changed demand for money that has to be responded to by the Fed in such a way that nominal spending does not fall, then the “solution” that sees the Fed creating more dollars obviously follows.

    However, if we instead asked why the demand for money would suddenly rise, and tracing that back to prior causes, then “more inflation!” isn’t the obvious solution. Finding out the prior factors that led to the demand for money suddenly rising, opens up the possibility that the cause was prior inflation that changed the real structure of the economy in such a way that later on, even if the Fed does not destroy dollars, nominal demand suddenly falls.

    MMs actually believe they have it figured out when they treat the demand for money as an independent, separate value judgment of its own that is unconnected to the real state of the economy. People just want money for the sake of wanting more money, the same way people want more flat screen televisions for the sake of having them. Since the Fed has a monopoly on money, then MMs believe a rise in the demand for money should be responded to by more money creation, similar to how a rise in the demand for flat screen televisions should be responded to by television producers in producing more flat screen televisions.

    The problem with this is that it doesn’t take into account the fact that money is a medium of exchange that facilitates trading of goods and services for other goods and services. People don’t want to hold more money only because they derive utility from holding more money. There is a connection between money and real goods/services, and therefore between demand for money and demand for real goods/services.

    If I want to hold more money, then I do so not because I merely like to have a fatter wallet. I do so because I want to allocate my spending over a longer period of time than before. If cross-sectional spending pattern changes should be communicated to investors without state intervention, and MMs are willing to tolerate unemployment and idle resources at those individual firms that experience a fall in revenues, then why can’t inter-temporal spending pattern changes be communicated to investors without state intervention as well? Why won’t MMs tolerate unemployment and idle resources then?

    MMs are caught in a contradictory position. On the one hand, they thumb their noses at temporary unemployment and idle resources deriving from cross-sectional changes in preferences, and say that it is wrong for the state to interfere. On the other hand, they don’t thumb their noses at temporary unemployment and idle resources deriving from inter-temporal changes in preferences, and say that it is right for the state to interfere.

    If it is OK that my changed relative consumer preference leads to a reduced spending at McDonalds and causes temporary unemployment and idle resources there, then why is it not OK that my changed cash preference leads to reduced spending at McDonald’s and causes temporary unemployment and idle resources there? Why is inflation justified in the latter, but not the former?

  20. Gravatar of James Oswald James Oswald
    5. September 2012 at 12:05

    There are two issues to address:
    1. Whether or not to stabilize NGDP going forward.
    2. Whether or not to engage in catch up growth because of the 2008-2009 period of low NGDP growth.

    Only the second runs into time consistency problems, and only then if you assume that temporary NGDP growth higher than whatever the target is would be politically unpopular. I find that hard to believe, as I think fast NGDP growth would lower unemployment and bring in more tax revenue.

    The second point of disagreement between you and Cochrane is “can the Fed actually raise NGDP growth now”, which probably won’t be resolved quickly. Also, here is a post by Angus which also discusses this topic.

  21. Gravatar of Patrick R. Sullivan Patrick R. Sullivan
    5. September 2012 at 12:11

    Off topic for this post, but we’ve discussed it before here;

    http://www.voxeu.org/article/income-taxation-us-households-facts-and-parametric-estimates

    ————quote————-
    One the one hand, a clear picture emerges from our findings. Effective tax rates on most households are relatively low (below 10%) and differ substantially from those at the top. For instance, married household around median income experience tax rates around 4%, while those at the top 1% face tax rates of around 23%. Furthermore, taxes paid are concentrated at the top. In a nutshell, the provisions in the law, in conjunction with the observed dispersion in income lead to the finding that the bulk of tax payments are concentrated in upper income households and that a large fraction of US households have effectively no tax liabilities. From this perspective, the answer to the question above is that there is substantial progressivity in the tax burden as measured by effective, average tax rates. Put differently and in plain terms, moving a hypothetical household along the income ladder implies substantial increases in average tax rates.
    On the other hand, tax rates at the top of the income distribution are essentially constant as income changes. Once high income levels are reached, effective tax rates do not change.
    ————endquote————

  22. Gravatar of PJ Glandon PJ Glandon
    5. September 2012 at 12:25

    Great response Scott.

  23. Gravatar of Woodford, Cochrane, Sumner: Geldpolitik-Lehren aus der Krise « Aus dem Hollerbusch Woodford, Cochrane, Sumner: Geldpolitik-Lehren aus der Krise « Aus dem Hollerbusch
    5. September 2012 at 13:00

    […] den letzten Jahren der führende Proponent eines Nominaleinkommensankers für die Geldpolitik war, antwortet ausführlich, und hebt dabei hervor, daß niedrige Zinsen meist auf eine vorangegangene strikte Geldpolitik […]

  24. Gravatar of dtoh dtoh
    5. September 2012 at 15:56

    AGAIN – If you focus on how monetary policy works through the financial asset price channel, the problem of the effectiveness of monetary policy at the ZLB entirely resolves itself.

    If you accept EMH, then an increase in financial asset prices has to cause market participants to exchange financial assets for real goods and services (spending on consumption and investment). This can happen either through the sale of financial assets currently held or through the issuance of new financial assets. No economist will disagree with this.

    It’s simply a question of correctly defining financial asset prices, and the correct definition is real (nominal minus inflation) risk adjusted expected annualized yields.

    Once you have the correct definition, then an increase in expected inflation (by definition) increases financial asset prices regardless of whether or not nominal interest rates are stuck at the ZLB.

    This is irrefutable and should be obvious.

  25. Gravatar of OneEyedMan OneEyedMan
    5. September 2012 at 17:35

    I see an essential point in Cochrane’s post that you don’t address. Why markets should believe the Fed in NGDP targeting when their is no costly commitment and any commitment here is likely to not be sub-game perfect. It seems like markets only started to believe the Fed was serious about fighting inflation when it at great cost induced a recession to fight inflation in the early 80’s. At this point isn’t any claim that we are going to keep rates too low (relative to what is optimal at that time) simply cheap talk?

  26. Gravatar of Tyler Cowen Tyler Cowen
    5. September 2012 at 18:12

    I view Cochrane’ rhetorical stance as somewhat different. I don’t think he sees the need to commit on a lot of the questions you are posing. Woodford is a foundationalist guy who wrote a foundationalist paper (a method you may or may not like, but for sure it is there in Woodford). Cochrane says he agrees with most of the paper, but there is no foundation given for arguing the nominal shocks matter in the relevant way. So why should he take that step of the argument with Woodford? He is just saying that Woodford has not established a key step in the argument and indeed it seems he has not.

  27. Gravatar of Morgan Warstler Morgan Warstler
    5. September 2012 at 20:37

    OneEyedMan repeats my point – and does it succinctly.

    There is MCSH TO GAIN from front loading the pain to convince everyone we’re serious about NGDPLT.

    Eating the medicine first, instead of the sugar, delivers real changed beliefs.

  28. Gravatar of Sisyphus Sisyphus
    5. September 2012 at 23:05

    Major Freedom – there was a large scale reduction in the money supply, just not in M1 (cash/currency). Credit can act as a form of money, too. Just ask all the people who partly lived off home equity lines or credit cards in the mid-2000’s. And credit availability was absolutely decimated in 2008. That’s what happens when you have a banking run, or perhaps more accurately in this case, a shadow banking run.

    Scott has elsewhere pointed out the three things the Fed did or failed to do that he sees as causing this reduction in available credit and then shadow banking run. First, the Fed raised interest rates in early 2008. Second, they failed to lower interest rates in the early fall 2008, when it was already clear the financial sector was seriously shaky. Third, they started paying interest on reserves held by the Fed for the first time right around the time Lehman collapsed.

    More than a trillion dollars is currently held in reserves at the Fed by banks instead of being lent out, seemingly because of the interest on reserves, although perhaps other regulatory strictures also drive that behavior (e.g. Basel). Not coincidentally, that is roughly the amount that the Fed added as the Great Recession peaked. Subsequent QEs have been far too small to return us to the trend of money supply growth (with credit included) we were on previously, so unsurprisingly we cannot return to the same NGDP trend. I’m not even sure we have returned to the same level of nominal money supply (again including credit) as we had in mid-2008.

  29. Gravatar of Sisyphus Sisyphus
    5. September 2012 at 23:05

    Re Morgan Wastler’s suggested compromise, I wonder if he goes far enough? After all, the Taylor Rule proponents are in many ways similar to the NGDP proponents. Both want predictable, non-discretionary monetary policy. If a compromise between all the economists who want to eliminate the Fed’s unpredictable discretionary policy were reached, it could probably go a long way toward becoming a plausible policy. Maybe a mean between the Taylor Rule output (3% inflation target) and 6% targeted NGDP rule, and no paying interest on reserves? An end to Fed discretion could be pretty convincing politically, too, especially for the many Republicans who are now skeptical of the Fed.

  30. Gravatar of Browsing Catharsis – 09.06.12 « Increasing Marginal Utility Browsing Catharsis – 09.06.12 « Increasing Marginal Utility
    6. September 2012 at 04:07

    […] Sumner against John Cochrane. […]

  31. Gravatar of ssumner ssumner
    6. September 2012 at 04:36

    Tyler, I’ll take your word for it that Woodford didn’t establish why nominal shocks matter, but I think that’s normal operating procedure. Elsewhere he has used sticky wage models that do show why nominal shocks matter. AFAIK it’s pretty standard in macro to simply rely on the assumption that nominal shocks matter, and not go back to square one on every assumption.

    If nominal shocks don’t matter, it’s not clear why the Fed would matter, and thus it’s not clear why Cochrane would even both discussing such an unimportant institution. Granted that nominal shocks might matter in some non-standard way (i.e. affect some real variables, but not real GDP) but that seems implausible.

    I’d add that Cochrane does more that just claim Woodford hasn’t proved his point, he also suggests NGDP targeting is a bad idea, but doesn’t really explain why. To do so he needs to explain how nominal shocks affect RGDP. But he doesn’t, other than to make derogatory remarks about the Phillips Curve. Are we to infer from that that he thinks nominal shocks don’t matter? If so, no one outside the RBC community will take his argument seriously. Even Austrians think nominal shocks matter. It’s fine if that’s the argument he wants to make, but he could have done so in one sentence. If he thinks nominal shocks do matter, then he needs to tell us why he thinks NGDP can’t deliever a better P/Y path over time.

    One-eyed Man, I’ll address the commitment problem in a new post, relatively soon. It’s a completely, 100%, phony issue.

  32. Gravatar of Wonks Anonymous Wonks Anonymous
    6. September 2012 at 06:41

    Speaking of Williamson, K, here is his post discussing Woodford’s paper. He does indeed think that QE matters, but announcements affect expectations of future rates.

  33. Gravatar of Matt Matt
    6. September 2012 at 12:22

    If you want to re-define the Phillips Curve as unemployment – real gdp + gdp deflator, that is fine, but then it makes hidden our long run ideas (LR neutrality Etc). 

    The whole edifice assumes that expected price changes have no impact and that unexpected price changes have a short run impact. This was friedman and phelps’ key insight, draw from their superior micro foundations. Why bury that?

    That might be a bad framework / assumption, but it is standard. Throwing stones at Cochrane because he uses the Phillips curve which _everyone_ is taught is childish, and reflects poorly on you.

    Your comments about identification of policy shocks do not prove anything except how difficult this question is – the equity market may have rallied / USD weakened because the Fed by being easier than expected showed that they actually GOT how bad it was. That is the market was worried about the economy, and more worried that the fed didn’t get it.

    When you contribute one tenth as much as Milton Friedman, then you can preen about your smarts. This exegesis you do is ugly. Reminds me of the post-Keynesians.

    This is one of the most childish posts i have seen on this blog. Shame on you.

    Welcome to the Krugman club

  34. Gravatar of TheMoneyIllusion » Credibility is not now and never has been a problem for fiat money central banks TheMoneyIllusion » Credibility is not now and never has been a problem for fiat money central banks
    6. September 2012 at 17:20

    […] Matt Rognlie made this comment on my John Cochrane post: Woodford is more skeptical of the power of an NGDP target than you are; in particular, although he thinks it would help, he doesn’t think a level target completely eliminates the problem of the zero lower bound. I agree with him, and I suspect that we have similar reasons for this belief. In Woodford’s model, one can prove that the zero lower bound will still pose a barrier in some situations even when there is an NGDP level target, and the central bank will not immediately be able to move the economy to its target. There are various reasons why his model might not be a complete description of reality, but I think it is a pretty good benchmark, and the burden is on you NGDP zen masters to come up with a modification of the model (or alternative model) where the zero lower bound is no longer a barrier. : ) […]

Leave a Reply