26 Aug 2011

A Question for the Quasi-Monetarists

Economics, Federal Reserve 20 Comments

Oh man, now I know how Krugman feels about those dastardly Austrians gaining more influence… That’s what I think every time National Review and now The Economist start running Sumnerian agit-prop.

Joking aside, help me out here guys (especially Bill Woolsey if you’re tuned it). The Economist piece says: “For all its theoretical merits, a switch to NGDP targeting would throw up some new problems—and old ones. The Fed has not exactly sat on its hands since the financial crisis began in 2007, so it is far from clear it could easily reach the new goal.”

So please clarify for me: Are you guys saying that the Fed needed to expand its balance sheet more than Bernanke actually did? Or are you saying he went about it the wrong way? I.e. if the Fed had been targeting NGDP all along, then it wouldn’t have needed to expand its balance sheet as much as Bernanke did (to little avail)?

If the latter, then why does your approach give a bigger bang for the reserve buck? Is it purely because of expectations, or is it because you’d have the Fed acquiring different assets?

(I think I know the answers to the above questions–I read Sumner almost as much as I read Krugman–but I want to hear a true believer explain it.)

20 Responses to “A Question for the Quasi-Monetarists”

  1. Martin says:

    Well, I am not a true believer, I consider the argument to be basically sound. What I find to be convincing is the following:

    1. Future NGDP to a large extent determines present NGDP:

    Expectations matter.

    2. If the CB/Fed can credibly commit to deliver higher NGDP, then arbitrage will do most of the lifting.

    If you know that prices will be at x at point t+1, you will be willing to pay anything up to x* (time value of money) for said asset at any time between t and t+1. Profit opportunity.

    3. Can the CB/Fed create hyperinflation? yes. Therefore the Fed/CB can drive NGDP.

    The problem is with the credible commitment.

    Krugman’s 1998 paper essentially agrees, the problem of a liquidity trap is that it is actually an expectations trap. If you can overcome that, then there is no problem. Scott argues that the Fed can and his NGDP futures market idea is to make it so and to give the Fed something to target. Krugman, based on Japan, believes it can’t and therefore argues for fiscal stimulus.

    This is how I’d basically summarize it.

  2. Martin says:

    Personally, I think you should support Scott over the alternatives as under his scheme inflation is limited. That is, if you believe the argument is basically sound. You can still argue that the problem is structural, which means that at worst this policy will generate mostly inflation, but this is limited to the NGDP growth rate. The resulting mis-allocation due to inflation will therefore also be limited, however it will be rhetorical victory for the Austrians (that you cannot inflate your way out of a recession the mainstream considered to be nominal). So at worst, future policy would be more Austrian in orientation and at best the economy would be humming along.

  3. David Beckworth says:


    Here is a post that answers most of your questions:

    Note that a properly calibrated NGDP target would allow for productivity-driven deflation and thus allow the actual market interest rate to more closely track the natural interest rate.

    • David S. says:


      I wish I could say it’s good to see you here, but I can’t understand why you bother. You and Bill Woolsey have offered great commentary during this economic mess and do fine work in the field and yet you’re here bothering to try to explain these simple concepts to this supposed NYU econ PhD who’s had years to try to get it.

      Krugman has it right to ignore guys like this.

    • Rob R. says:


      In regards to:

      “Note that a properly calibrated NGDP target would allow for productivity-driven deflation and thus allow the actual market interest rate to more closely track the natural interest rate.”

      Wouldn’t that only be obtainable with a NGDP target of 0% ?

  4. David S. says:

    Bob, I think the fact that you’re asking this question after all this time’s revealing. You haven’t even paid attention to and/or simply haven’t understood what the quasi-monetarists have been saying. You’re supposed to be an economist and some of these are high profile economists that you’ve been blogging about for months or even years.

    Again, give it up. You’re no economist.

    • bobmurphy says:

      David S. wrote:

      Again, give it up. You’re no economist.

      I just figured it out! The “S.” is short for “Smith,” and you’re the Agent from the Matrix. You keep trying to break my will, but I’ll never give up!

      • Joseph Fetz says:

        Easy there, Bob. If you tempt him he may not let you ride on his yacht. And, you know what that means….

        No Cristal for you.

  5. Bill Woolsey says:

    At first pass, yes, base money should have increased more..

    Yes, it is true that if we had NGDP targeting all along, a more modest increase in base money would have been needed.

    Also, the Fed’s policy policy of paying a higher interest rate on reserves than the rate on other short and safe assets is a mistake. Right now, going back to no interest on reserves would be an improvement, though I think that the intereset rate on reserves should float and be a fixed amount below the interest rate on other short term and safe assets. That means that now, the Fed should be charging banks for holding reserves.

    Reducing the interest rate the Fed pays on reserves from the current .25% any amount, to zero, or slightly below, would reduce the increase in the quantity of base money necessary.

    I think most quasi-monetarists like QE2 (purchasing long term bonds) better than QE1 (lending money to banks and other financial institutions.) QE1 should have been more like QE2, just purchasing government bonds.

    The basic approach is to have an explicity target for the growth path of NGDP. At any future date, there is a level of NGDP that the Fed is committed to reaching. And it will buy or sell the amount of assets, and increase or decrease base money, whatever amount is needed to reach it. And it says it will do that.

    The quantity of base money depends one whatever it takes. Interest rates, the price level, and real output all depend on market forces.

    Most quasimonetarists (including me) believe that NGDP today very much depends on what people think NGDP will be in the future. But for people to believe that NGDP in the future will be on some kind of target growth path, there must be method of getting it there even if people have perverse expectations. The periodic adjustmetns in short and save interest rates (the fed funds rate) fails in some circumstances. A committment to quantitative easing, in heroic amounts if necessary, can provide what promising lower interest rates cannot.

    And remember, the other side of the coin with NGDP targeting is a commitment to sell assets or raise interest rates however much is needed if NGDP is too high. If people believe that will happen, then inflation won’t develop.

    • bobmurphy says:

      Thanks Bill (and David Beckworth). I was wondering how much of the approach was dependent on using actual NGDP futures markets, or if the Fed could just buy and sell Treasuries.

    • Yancey Ward says:

      Bill wrote:

      And remember, the other side of the coin with NGDP targeting is a commitment to sell assets or raise interest rates however much is needed if NGDP is too high. If people believe that will happen, then inflation won’t develop.

      So, I have seen Sumner write that targeting NGDP growth should be targeted at 5%, but I have never really understood where he gets this number, so I am at a loss to understand what makes the actual number “too high” or “too low”, but now I am even more confounded by your statement that if NGDP isn’t allowed to get “too high”, then inflation “won’t develop”. So, what counts as inflation, or excess inflation, or whatever, in this scenario?

  6. Bill Woolsey says:

    The way I see it, the futures contracts allow entrepreneurs (specifically futures contract speculators) to provide information about whether the current finanical conditiosn (whichever they think are imiportant) is consistent with the NGDP being on target in the future. The sort of convertibility schemes I favor would require the monetary authority (fed) to be adjust the quantity of base money so that it takes no position on the contract. Entrepreneurs, then, determine the quantity of money.

    • David S. says:


      I’ve seen this idea here and there and I hope you guys will really push hard for it. Why let this economic slump go to waste? Clearly fundamental changes are needed at the Fed, at the very least.

  7. Bill Woolsey says:

    Sumner gets 5% based upon the trend growth rate from 1984 to 2008.

    It is also equal to the trend growth in the productive capacity of the economy (3%)
    plus the Fed’s target for inflation, 2%.

    Why did we get the recession? Because NGDP is 14% below the growth path of 1984 to 2008. The Fed should have kept it on that growth path. And if it didn’t, it should get it back up ASAP.

    Sumner is open to a shift in the growth path to a slower rate, but not right now. Basically he says a planned disinflation should not occur in the midst of a banking crisis.

    I favor a 3% growth rate, rather than a 5% growth rate. That implies zero inflation rather than 2%.

    I think Selgin’s view is that somethiing like 2% growth in NGDP is best, resulting in 1% deflation. (Maybe it is somewhere between that and 1% NGDP growth with 2% deflation.) However, Selgin proposes a free banking system and whatever NGDP that generates is what we get. He believes that there is reason to believe what we would get approximates what would be best. Oh, and what is best of NGDP really depends on the growth of the quantity of productive resources and the best deflation rate equals the rate of growth of total factor productivity.

    Anyway, for all of us, high and rising inflation is only a realistic possibility if NGDP grows at a high and rising rate. And I think all of us agree that realitistically, that would only be possible if the quantity of money were to grow at a high and rising rate. And so, the NGDP rules restrict the quantity of money from rising beyond the amount that will allow NGDP to grow at a slow stready rate, which will result in low inflation– 2% for Sumner, 0% for me, and between -1% and -2% for Selgin.

    On the other hand, if the demand to hold money begins to grow at a high and rising rate, then we have no problem with the quantity of money growing at a high and rising rate. This would not cause NGDP to grow at a high and rising rate, but rather would keep it growing at a slow and stable rate. And inflation would not be high or rising. (I think, realistically, we would all believe that such a high and rising increase in the demand to hold money would be odd and temporary and would be reversed in the future, so that the quantity of money would also have to be reduced as well.)

    The prospect that the productive capacity of the economy will begin to fall at a rising rate, isn’t worth worrying about much. And so, keeping NGDP growing on a stable growth path isn’t going to result in high and rising inflation.

    On the other hand, all of us agree that the productive capacity of the economy may grow more quickly or slowly sometimes. It might even fall. And when that happens, the least bad outcome is changes in inflation.

    For example, if the productive capacity of the economy grows only 1% per year for two years, then Sumner’s approach would have higher inflation for those years–4%. And the price level would be about 4% higher than it otherwise would be. (and continue to rise 2% from there.) And the opposite is true. If the productive capacity of the economy grew 4% for three years in a row, then the inflation rate would only be 1% per year for those three years, and the price level would be 3% lower than it otherwise would be. Though it would resume rising 2% per year.

    If the productive capacity of the economy grew 6%, then Sumner’s approach would result in 1% deflation that one year, then prices would go back to rising.

    For me, with a 3% rule, productive growth faster than 3% results in deflation and growth slower than 3% results in inflation. For example, suppose the productive capacity of the economy falls 1% rather than rises 3%. Then there would be 4% inflation that year, and the price level would be 4% higher. And if the productive capacity of the economy returned to growing 3%, then the price level would just stay at the 4% higher level.

    If, on the other hand, the productive capacity of the economy grew 5% one year, then there would be 2% deflation that year. And if the productive capacity of the economy went back to growing 3%, then the price level would just stay 2% lower.

    With Selgin, prices are usually dropping 1% (let’s say.) If the productive capacity of the economy grows 4% on year, then deflation would be 2%, and then prices would go on dropping 1% a year from that level. If, on the other hand, the productive capacity of the economy should fail to grow one year, there would be 2% inflation that year, but assuming the productive capacity of the economy resumed growing 3% per year, then prices would return to dropping 1% per year from that higher level.

    • bobmurphy says:

      Thanks for the thoughtful comments, Bill. In case it’s not clear, I think you quasi-monetarists are far more formidable than the standard New Keynesians. I wasn’t kidding when I said in a Mises.org article that I would much rather debate Krugman than Sumner (or you, for that matter).

      I guess I come back to my main complaint about Sumner: I don’t see how you guys can look at history and say, “The thing that makes this recession so special, is that the Fed isn’t inflating enough.”

    • Yancey Ward says:

      And, like you said, this all depends on the credibility of the central bank to enforce the latter condition of reining it all in if it gets too high (regardless of whatever growth rate is selected as the standard). I still don’t understand why this plan has added credibility over previous central bank policy, even if I grant that Sumner’s idea has more merit in other respects.

  8. Martin says:

    You’re surprised that the central planning of production leads to shortages? 😛

  9. Bob Roddis says:

    It’s not clear to me when central bank money dilution stopped creating Cantillon effects, including the surreptitious theft of purchasing power from unwitting victims and stopped distorting the price, investment and capital structure leading to unsustainable booms followed by tragic and unnecessary busts. Further, when did it stop violating basic common law and constitutional notions of private property and contract as it operates through a process of theft and fraud without due process, specific acts of the legislature or any form of redress for its victims?



  10. Bill Woolsey says:


    Starting in equilibrium and then tracing the effects of an increase in the quantity of money doesn’t really tell you much about the effects of alternative monetary regimes.

    Sumner is proposing to keep a 5% NGDP growth regime. We didn’t shift to that regime from a constant quantity of money, but rather from a much more inflationary regime in the seventies. We contracted into it.

    We aren’t creating an excess supply of money each day, pushing down market interest rates, and perhaps causing malinvestment.

    Because of current and expected future inflation, the nominal demand for credit grows faster, and so matches the growing nominal supply of credit. Market interest rates are stable, and higher by the expected inflation rate. Real interest rates are more or less the same.

    The banks pay higher interest on deposits as well as earn higher interest on loans. While the nominal value of the balance sheets is growing, the real value isn’t changing more than it would have anyway. Or not any kind of systematic way.

    The people who borrow the “new” money have to pay higher prices because those selling the goods already raise their prices because they correctly believe that they can and they also believe that if they don’t, they will lose money because they will have to pay higher wages and other costs or else not be able to get the resources they need.

    You are imagining that in an inflationary process, those selling goods keep their prices at the level of last year, until they see their sales rising extra fast. And so, the people who bought at the prices get a bargain. But that isn’t what happens when we have persistent 2% inflation. They plan to raise their prices because they expect the demand for the products to rise and that their costs will rise as well.

    And, as I explained above, when you think about credit markets, the growing quantity of credit matching the money creation is matched by a growing demand for credit. People have to borrow more money to pay for for the more expense stuff. Real interest rates aren’t lower. And nominal interest rates are higher.

    Now, is the pattern of demand going to be exactly the same as it would be under some other monetary regime? No. But there are very good reasons why the sort of patterns you describe won’t exist.

    If we have a 2% inflationary regime, and the central bank creates an excess supply of money, then it can easily result in the real market rate falling below the natural interest rate. This could result in malinvestment. But if the quantity of money grows with the nominal demand for money, then real market interest rates don’t fall and there is no particular reason to expect malinvestments.

    Now, if saving supply rises faster or investment demand grows slower, then the natural interest rate falls. The real market interest rate should fall to maintain coordination. And there are market forces that will result in real market rates falling.

    But to say that market interest rates fell and the quantity of money rose, and credit rose, so there must have been malinvestment? Not necessarily true.

    Again, thinking about an economy in equilibrium responding to an unexpected increase in the quantity of money, with the demand to hold money, the supply of saving, and the demand for investment all being constant may be a good start, but it is not sufficient to understand what happens with a regime of growing money expenditure on output.

  11. Nav says:

    Did you read the article Robert?

    “A central bank’s tools would be the same: adjusting short-term interest rates in normal times and reaching for unconventional implements such as quantitative easing—buying assets by creating money—when interest rates are, as now, close to zero. The difference lies in what signals the new target would send. At present a central bank targeting nominal GDP would be loosening policy much more aggressively. Mr Sumner reckons that the commitment to hit an NGDP target would itself boost the economy; but if it didn’t, the Fed would presumably have to do even more quantitative easing until the target was hit.”

    This comes from the article. If you need help what they are saying is that Mr. Sumner believes the change to NGDP target would alter expectations and boost the economy and the fed would have to do even more quantitative easing. I guess you could presume that they may try to buy other investments if they buy up the entire government and corporate bond market but the idea behind the article doesn’t imply that additional investments would be necessary.

    Does that clarify?

    What would be the implications of the fed buying the entire government and corporate bond market?