Wednesday, October 12, 2011

My Journey Into Market Monetarism

When I started blogging in 2007 my writing focused on the Federal Reserve's failure to properly handle the productivity boom of 2001-2004 and how this failure contributed to the global housing boom.  This productivity boom--spawned by the opening up of Asia and the ongoing technological gains--increased economic capacity, put downward pressure on inflation, and implied a higher natural interest rate.  The Fed, however, responded to the fist two developments as if they were signalling falling aggregate demand rather than rapid increases in aggregate supply.  The Fed did this by failing to raise the federal funds rate when the natural interest rate rose and then kept it well below the natural rate level for several years.  Given the Fed's monetary superpower status, this sustained easing created a global liquidity boom  that was a key force behind the "global saving glut".  This view was what initially drove most of my blogging. 

By late 2008 my focus began to change.  I had been critical of the Fed for allowing too rapid growth in nominal spending during the first half of the decade, but by this time it seemed the Fed was erring in the opposite direction.  Nominal spending was falling fast and the Fed's seemed more focused on saving the financial system than in directly preventing the collapse of aggregate demand.  The Fed's introduction of interest payments on excess reserves in October, 2008 only served to confirm my fear that the Fed was too narrowly focused on financial stability.   This fear combined with what I was reading from Nick Rowe and Bill Woolsey (in the comments section initially) about the excess money demand problem and early posts from Scott Sumner about the Fed causing the financial crisis by failing to stabilize nominal spending in the first place convinced me that the Fed had committed a colossal policy mistake in 2008.  This failure to respond to the drop in nominal spending I later came to recognize as a passive tightening of monetary policy (something that is easy to show using an expanded equation of exchange).

As time went on, it also became apparent to me that the Fed was not forward looking enough.  For example, as early as mid-2008 breakeven inflation from TIPs was indicating an aggregate demand slowdown was ahead.  The Fed, however, at the time put more weight on backward-looking headline inflation measures as was evident in its decision to not cut the target federal funds rate in the September, 2008 FOMC meeting.  Modern macroeconomics and experience tells that one of the most effective ways the Fed can influence aggregate demand is by managing expectations. Shape nominal expectations properly and one can immediately affect aggregate demand.  As Scott Sumner likes to say, the implication of this insight is that monetary policy works with leads not lags.  

Now here we are in 2011 and the Fed has yet to, one, correct its passive tightening of the past three years and, two, properly shape aggregate demand expectations by adopting something like a nominal GDP level target.  It has been incredibly frustrating to watch the incredible amount of human suffering caused by these monetary policy failures.  Consequently, I have been blogging away at these issues along with like-minded folks such as Scott Sumner, Nick Rowe, Bill Woolsey, Josh Hendrickson, Marcus Nunes, Nicklas Blanchard, Kantoos, and David Glasner.  We all have been making the case that  the prolonged economic slump has been mostly due to passively tightened monetary policy that could easily be loosened, even at the interest rate zero bound.

Our collective efforts have been summarized in a recent paper by Lars Christensen, who labels us as a group Market Monetarists.  He argues that we are a burgeoning economic school born out of the Great Recession experience whose views have largely taken shape in the blogosphere.  What defines us, he says, is (1) our belief that this crisis has it origins in monetary policy failure rather than problems in the financial system, (2)  our emphasis on using market signals to determine the stance of monetary policy, (3) our view that monetary policy's influence on nominal spending is not constrained by the interest rate zero bound, and (4) our push for nominal GDP level targeting as way to get monetary policy back on track.

While I largely agree with Christensen's assessment of our views, there are some additional points worth noting.

First, though Market Monetarism has been largely a blogging phenomenon it has had important voices in other mediums.  Ramesh Ponnuru has been pushing the Market Monetarist view at the National Review and at Bloomberg while MKM Chief Economist Michael Darda has been promoting it in the MKM investment newletter and on interviews on CNBC and Bloomberg Radio.  And even within the blogging medium there are other prominent voices like that of Matthew Yglesias, Ryan Avent, and Brad DeLong who often are sympathetic to Market Monetarists views.

Second, Market Monetarists prescriptions are not all that different than those of prominent New Keynesians like Michael Woodford and Paul Krugman.  We all agree that when the zero bound is hit the monetary base and t-bills became perfect substitutes and so the Fed should buy longer-term treasuries or foreign exchange as part of a plan to hit some explicit nominal target.  A big difference, though, between New Keynesians and Market Monetarists is that where the former sees the move from t-bills to other assets as a discrete jump from conventional to unconventional monetary policy, Market Monetarist see it as simply moving down the list of assets that can affect money demand.  The zero bond for us really is not a big deal, but simply an artifact of monetary policy using a short-term interest rate as the targeted instrument.  We approach monetary policy with much  less angst than New Keynesians.

Third, Market Monetarist stress NGDP level targeting because doing so would forcefully shape expectations. Here is why.  Under such a monetary policy regime, the Fed would announce (1) its targeted growth path for NGDP and (2) commit to buying up as many securities as needed to reach it.  Knowing that the Fed would be willing to buy up trillion of dollars of assets if necessary to hit its target would cause the market itself to do much of the heavy lifting.  That is, the public would adjust their portfolios in anticipation of the Fed buying up more assets and in the process cause nominal spending to adjust largely on its own.  This would reduce the burden on the Fed and make it a less polarizing institution.

Finally, one critique of Market Monetarist is they lack an active research agenda and fail to take advantage of formal modeling methods like DSGE models.  While I cannot speak for all Market Monetarists, I can say that Josh Hendrickson and I have several research projects that formally evaluate the Market Monetarist  view.  For example, we have one paper where we make use of the search models developed in the New Monetarist's literature to formally develop a monetary theory of nominal income determination.  We also make use of structural VARs to examine the importance of nominal spending shocks in one paper and the portfolio channel of monetary policy in another paper.

With that said, Lars Christensen has done us a favor by documenting the rise of Market Monetarism.  It will be interesting to see what will be the long-run impact of the Market Monetarist bloggers .  I am glad to have been a part of the journey so far.

P.S. Lars Christensen is of the Market Monetarist persuasion too and now is blogging.


  1. David, that's a great summary.

    I've been seeing NGDP targeting getting greater and greater acceptance in the blogosphere lately. While I'm not a total convert (how would this work say in an emerging market economy with underdeveloped financial markets), I'm more than halfway convinced this is the future of monetary policy.

  2. Hishamh,

    Thanks. I agree how NGDP would work in an emerging economy does raise some tough questions. For example, under NGDP targeting, a negative supply shock would be ignored (NGDP growth target would stay the same, just with temporarily higher inflation rate and temporarily lower real GDP growth). While this would be fine in advanced economies where central banks have credibility, it is not clear how stabilizing it would be in an emerging economy that lacked inflation-fighting credibility.

    I am not sure there has been much, if any, formal work done on NGDP targeting in small open economies.

  3. There are only two papers that I know of, but I am not very familiar with work on small, open economies. The consensus in the two papers is that other rules would be preferable for such economies. Here are the links:

  4. Congratulations to David Beckworth for having an inquisitive mind free of dogmas or overpowering partisan sentiments.

    Keep blogging, keep up the buzz. Headway is being made.

  5. David, Thank you for your nice words! They are much appreciated.

    Hishamh, I hope to be able to address the issue of NGDP targeting in small open economies soon in my blog ( But the short answer in my view is that NGDP targeting would work fine in Emerging Markets and/or small open economies. However, instead of using NGDP I think one should use a measure of domestic spending instead. In small open economies like Iceland and Singapore net exports are high volatile by nature and that would make "normal" NGDP targeting harder in my view.

  6. My own journey: I was deeply ambivalent. I could see the world through a New Keynesian lens, but I thought it left something out. And I could also see the world through an older, monetarist lens, but it wasn't clear enough. Then I read Scott.

    But I'm still trying to reconcile the two visions, and clarify things.

  7. @Josh,



    I'm pretty sure the same problem would apply with large open economies too (I'm thinking of China and India here). Neither have developed financial markets, and have limited monetary policy tools as a result - interest rates for example don't have much traction.

    The other thing I'm trying to wrap my head around is what's the appropriate NGDP growth path to aim for? That path will change as economies transition from low income to middle income to high income.

  8. @Lars,

    I'm based in Malaysia. If you want any data support, feel free to drop me a line.

  9. David: As you note, the 'Market Monetarists' generally are not engaged in research. Until they are, and until the fruits of that research appear in respected outlets (far higher than, e.g., Applied Economics Letters), they'll only receive attention in the economics blogosphere. A massive literature exists on nominal income targeting (e.g., McCallum's 1999 Handbook of ME piece), and many non-trivial issues have been analyzed. As Hamilton emphasized in a 2009 exchange with Sumner, the recently christened MMs need to do much more than restate the MV = PQ identity (and the kind of light shed by your 'expansion' of that identity is not exactly what's called for).

  10. Glibfighter,

    I agree with you that we need to do more formal work and some of us trying as I mentioned. At the same time, how one influences public debate is changing and blogging can be a very powerful medium. The fact that NGDP targeting is now getting a wide hearing a(e.g being discussed in The Economist magazine, by Fed officials, etc.) is not because someone opened up a top economic journal and noticed, but because of Sumner's and others blogging. Times are a changing and one must adapt or become irrelevant.

  11. David: Forecasts of irrelevance are tricky. Before Ron Paul's recent presidential campaigns, there weren't all that many voters who gave thought to the gold standard, and fewer who were familiar with the work of academics (e.g., Eichengreen) who demonstrated the folly of such an exchange rate regime. So who's more popular now, Ron Paul or Barry Eichengreen?

    Do you think that important players at the Fed were uninformed about nominal GDP targeting before Sumner made his splash? Really? BTW, Sumner's recent musings about the recent Nobel prize in economics are telling. He has doubts about the whole Sims VAR modeling enterprise. Do you think the future history of monetary policy making will remember better Sumner or Sims?

  12. GlibFighter,

    You write:

    "A massive literature exists on nominal income targeting (e.g., McCallum's 1999 Handbook of ME piece), and many non-trivial issues have been analyzed. As Hamilton emphasized in a 2009 exchange with Sumner, the recently christened MMs need to do much more than restate the MV = PQ identity (and the kind of light shed by your 'expansion' of that identity is not exactly what's called for)."

    Are you suggesting that market monetarists haven't read this literature? If so, what evidence do you have?

    Similarly, the market monetarist approach, at least as I see it, goes far beyond MV = PY. There is a rich monetarist literature that market monetarists draw from as well as insights from contemporary monetary theory (e.g. Woodford). The use of MV = PY is simply a device for communicating an idea to a general audience. This is what blogging is intended to do.

    Also, I suspect -- and as David alluded to -- that the market monetarists are engaged in scholarly research, but notice the divergence in the group with respect to their ages. Some are young assistant professors. Some are long-tenured faculty. I suspect that the former group is doing scholarly work related in some way to these ideas. Do you know differently? In addition, Scott Sumner talks about his scholarly work all the time. Gauti Eggertson cites a number of Scott's paper in his work on the Great Depression.

    You are correct to say that market monetarist ideas will be less likely to catch on until they convince their colleagues with scholarly research. However, that does not mean that blogging isn't helpful in terms of having informal, public discussions with fellow economists and the general public. I think that it is wrong to suggest that they are not doing (or have not done) scholarly research and that they don't read the literature. (At the very least, they have read the literature -- as evident from reading their many posts over the last several years.)

  13. "A big difference, though, between New Keynesians and Market Monetarists is that where the former sees the move from t-bills to other assets as a discrete jump from conventional to unconventional monetary policy."

    Except that there never was such a jump, because the Fed always bought treasuries across the curve, so they were unwittingly practicing unconventional monetary policy at the wrong time.

    Can we put to bed this unrealistic idea that the liquidity trap is defined by t-bill rates once and for all?

  14. Glibfighter,

    Do you think that important players at the Fed were uninformed about nominal GDP targeting before Sumner made his splash?

    No, but that is not the point. The point is that NGDP targeting and more generally the notion that the Fed could have done more has been brought back into the national conversation thanks to Sumner. This greater public awareness in turn has put more pressure on policymakers.

    Do you think the future history of monetary policy making will remember better Sumner or Sims?

    Sim's time series work has been highly influential in my research and no doubt his methological contribution will be long remembered. Of course, the Nobel Prize will only enhance his reputation.

    But in terms of understanding the current crisis and furthering the public debate on it Sumner will be remembered as far more influential figure than Sims. It is not even close. Sims has not been anywhere near as publicly engaged as Sumners. Sumners has blogged, written numerous articles and Op-Eds, and has presented at various think tanks. The fact that you (who sounds like a "serious" person)read Sumner and are aware of his splash is a testament to his influence.

    One of my big disappointments over the past few years is that heavyweights like Sims and McCallum haven't engaged the public more about monetary policy. Only a few like Michael Woodford have made a concerted effort to engage the national conversation.

    Finally, Sumner's ideas about futures targeting is something he first championed and if ever adopted would revolutionize monetary policy. It would have far more of an impact on the profession than anything Sims ever contributed.

  15. Anonymous: No, I don't have 'proof' that MMs are not sufficiently familiar with the nominal income targeting literature I referred to. That said, amidst all the Jonestown-like cheerleading, I haven't noticed substantive recognition of various concerns that literature has established, e.g., Ball's (1999) paper showing that nominal income target rules are inefficient in the sense that they increase both output and inflation volatility. David refereed to MM research he's been doing. It will be interesting to see what the impact of that (and similar) work turns out to be.