• istock_money_falling_from_sky

    Debate: Is QE Ever a Good Thing?

    A Debate Between NGDP-Targeting and Monetary-Base Targeting Supporters  helicopter-money40 Glossary CPI: Consumer price index: Consumer goods cost more dollars when the CPI rises.

    NGDP: Nominal gross domestic product. The combined dollar value of all investment, consumption, and government spending.

    Money supply: The total amount of money in an economy. All money, including deposits at banks, cash, traveler’s checks, and bank reserves on deposit at the central bank are counted.

    Fractional reserve banking: When a bank lends out money that has been deposited at that bank. The bank lends out some of a deposit and keeps some as reserve, and the ratio between loans and reserves is called the reserve ratio. If the market-wide reserve ratio is 10%, then 90 cents can be lent out by the bank when it gets $1, and then those 90 cents can be deposited in another bank, who can lend out 81 cents, and as this process continues until $10 in total deposits are produced from the original dollar deposited.

    Monetary base: The total amount of deposits that commercial banks have in their central bank’s reserves and amount of cash in the hands of the public. The central bank can increase the monetary base, introducing some new amount of dollars, which will then be turned into a greater amount of dollars on deposit through fractional reserve banking.

    QE: Quantitative easing. A method of expansionary monetary policy wherein the central bank expands the monetary base while purchasing financial assets from commercial banks.

    Velocity of Money: The number of times a given dollar changes hand over a given period of time. In other words, the speed of circulation of money.

    Quantity Theory of Money: A macroeconomic identity M*V=P*Y where M is money supply, V is velocity of money, P is the general price level and Y is total real output. Assuming V is constant in the long-term, Y is determined by fixed stock of labor and capital, and money has no impact on real variables, an increase in M should have a proportional impact on P. Note that both sides of the equation equal NGDP.

    Hayekian Triangle: A graphic representation of the structure of production. hayekian-triangle-stages-of-production


    Andrew Criscione Austrians don’t agree with Friedman’s criticism of the Fed: He said it wasn’t expansionary enough, the Austrians say it was too expansionary. Whether the CPI is increasing or decreasing, Austrians are against expansion. If the CPI is decreasing, Chicagoans are for expansion. The problem is: The “free market” supporters flock to Friedman, the rest flock to Keynes, but they’re both wrong. Keynes is, if anything, less wrong, since he argued for Bretton Woods over a fiat dollar: “John Maynard Keynes and Harry Dexter White would lead the global return to fixed exchange rates”). Whereas the Floating Point gang was wholly composed of monetarists, including Friedman himself. Read David Stockman’s book “The Great Deformation” if you want an in-depth history of the Chicago School’s impact over the last 40 years. All the economic problems that have come from fiat money can be laid at the foot of the Chicago School.

    Zachary Woodman Not all Austrians may not agree with the policy prescriptions, but for the most part (there are exceptions at the Mises Institute, but almost all academic Austrians I know of think this ) they agree with the narrative Friedman gave that the Fed was responsible for the bank failures in 1930-33. Additionally, Austrians are not all monolithic in their policy prescriptions. Many Austrians (Horwitz and Selgin , for example) do believe in NGDP targeting which would call for expansion in some times to keep NGDP stable.I know your first response is going to be “that’ll re-inflate the bubble.” You forget that when NGDP starts growing at too fast a rate (as it did 2005-2008), there would be a massive tightening of money supply to help restrain the bubble. Sumner and the other NGDP adherents argue there should have been for significantly less expansion in the years leading up to the financial crisis, which Austrians are definitely in agreement with .

    Andrew Criscione The Old-School Austrians at the Mises Institute (and Garrison is a faculty member there) are all against QE. But there is definitely another, more inflationary strain of thought out there among the free banking branch of the Austrian School . I consider myself a free banker in that I think there would be lots of fractional lending without a state. But I disagree with Selgin and Horwitz’s prescriptions for the Fed and think the older Austrian strain of thought got it right . Remember that even Friedman didn’t advocate printing so much money as to target stable nominal GDP: He just wanted stable prices.

    Zachary Woodman The problem with the view of the old Austrians at the Mises Institute is simply a failure to recognize the current institutional reality . Sure, it might be the case that for the long-term complete inaction by the Fed to allow for the reallocation of resources is better. However, adopting such a policy does absolutely nothing to stop the Fed from causing bubbles in the distant future. In this ways, both schools of thought really have the same criticism of each other; the Mises Institute says the free bankers are causing a future bubble, and the free bankers are saying the Mises Institute isn’t doing enough to restrain the Fed in the long term which will result in bubbles in the future . Let’s do a pro-con chart of adopting NGDP:   Pros: Ties the hand of the fed permanently . Restrains bubbles in the future by cutting off expansionary monetary policy in the early part of the bubble . Mimics Free Banking in the way Selgin lays out .   Con: Might risk causing a bubble in the short term.   Obviously I have a bit of confirmation bias , I can’t think of many other cons other than that one, I’m sure you probably have another.

    Andrew Criscione If QE were several times bigger, that’s a pretty bit “might”, but sure.

    Zachary Woodman Offsetting the bubble by tying the hands of the fed when NGDP growth accelerates is the reason for my “might.”

    Andrew Criscione Your first two pros apply to both money supply targeting and monetary-base targeting, though . And NGDP trails the monetary base by a considerable timeframe. If we just start cranking the printing presses full bore until the NGDP increases back from its recessionary low, a LOT is going to be printed until the economy “recovers”, which it won’t according to the ABCT .   Japan is probably the best example of NGDP targeting gone wrong. They’ve been printing full bore for decades, and it’s causing stagnation. It’s the 1970s without the rising prices.   The Austrians view this as a price fixing problem: Monetary expansion fixes the interest rate below what it otherwise would have been, the economy won’t recover until rates go back up. If you keep the rate fixed low artificially, the savings gap won’t be bridged, supply won’t match demand. Until savings recover, there will be no recovery, and savings won’t recover until the interest rate rises and the monetary expansion stops.

    Zachary Woodman The biggest problem with your point about the pros applying to Money Supply targeting is that the second and third pros do not apply to it. If your expansion in the money supply is causing a bubble, the stable expansion of money supply just continues doing it no matter what as you continue growing the money supply . Whereas when the bubble begins inflating in NGDP targeting, there is a mechanism to offset it .   Also, Selgin’s theory of free banking does not say that money supply is stable so it doesn’t have the third benefit and I’d add another con list for money supply targeting: velocity is not stable in the short-term, as the post-Keynsians like to point out. NGDP targets not just M, but M*V, so it’s immune from the post-Keynesian criticisms.   Does NGDP  targeting trail growth in monetary base that much? I would argue that it doesn’t. Right now the reason for growth in monetary base and stagnation in money supply is that the fed is paying interest on reserves, I’ll side with Sumner and Christensen in saying they should stop doing that. But assuming that the fed stops doing that, why should monetary base trail behind growth in NGDP that much?   As for your points about Japan, they’ve kept the level of NGDP stable  (a growth rate of about 0%, which is also what Hayek advocated). Sumner and company are advocating a higher growth rate.

    Andrew Criscione The Fed can cause a bubble in one industry while the nominal GDP is shrinking due to a bubble popping in another industry: A bubble doesn’t necessarily have to increase NGDP. How much of NGDP growth is due to the bubble and how much is due to the economy as a whole? It’s the pretense of knowledge to think we can know these things.   If the Fed was handing out trillions directly to college students (not exactly a hypothetical), it would cause a bubble. But it doesn’t necessarily cause NGDP to rise, especially if there’s a depression currently going on from the popping of a previous bubble. The price level and NGDP aren’t monolithic.

    Zachary Woodman Your point about bubbles being micro-level distortions that do not follow macro-level phenomena like NGDP and the price level is well-taken, but all that does is reinforce the one con that it might re-inflate the bubble which I’ve already admitted . That is definitely a risk of it, but that is a risk of any sort of central banking regime with any policy. The only way to avoid that problem is to revert to a free-banking system, which I am heartily in support of. However, political reality is that that won’t be the case .   Regarding the Mises Institute’s proposed reforms, all they want is taking steps in the direction of free banking. I would love it if we *could* do that, but the reality is we can’t. First of all there’s the political reality that nobody in power is willing to do that, but they are willing to adopt policy rules like NGDP targeting . Second of all, there’s the reality that investors in the economy have heavily embedded in their expectations based on the very institutions (like FDIC) that they are proposing being abolished; even if it might be in the long-term interests of everyone to do so, in the short term interrupting those expectations will only serve to create regime uncertainty  and that is definitely to be avoided. On the other hand, trying policy rules like NGDP targeting are considerably less discretionary .

    Andrew Criscione NGDP targeting would be worse than both the monetary-base targeting being pushed by the Mises Institute AND total-money-supply targeting pushed by the Chicago School/monetarists/Friedmanites, I think , since NGDP-targeting would involve expanding the monetary base in any instance with a steady price level but falling NGDP. It would also be worse than the status quo, since it would involve much more QE during a depression, enough to keep NGDP stable ,But I don’t think the depression ends until QE stops, so we’re stuck with a permanent depression/stagnation/deformation, which is more-or-less what happened in Japan.But again, there’s no good way to run the Fed, just like there’s no good way to run a state monopoly shoe store.   On one side are the people saying the Fed is doing a good job, on the other side are those saying the Fed isn’t printing enough money. The old-school Austrians get left out of the debate. Mises went into the basement of the Austrian national treasury and told the central bankers: “Turn the printing presses off!” I just don’t see him going: “Crank the printing presses up higher!”   If the Fed adopted NGDP targeting, I suspect they would adopt it during a depression as an excuse to print an ungodly amount of money , and then give up on it when it called for tighter monetary policy during the boom . The layman walks away from the conversation thinking: “The free market types all want more money printing”, whereas they should be thinking the opposite. Inflation is just as bad as taxation, and for the same reason. The great tragedy of the last 40 years is that fiat money and monetary expansion/inflation have been identified as free market phenomena thanks to the Chicago School .   I don’t see how any monetary expansion could be justified in the Hayekian Triangle framework . It is simply price fixing of the interest rate, and price fixing is never justified under any circumstances. All price fixing causes shortages, and fixing the interest rate is no different. If the Fed is expanding the money supply, it is fixing the interest rate too low. Expanding the money supply pushes rates low, and nothing can change this fact. Austrian Business Cycle Theory says that monetary expansion fixes the interest rate too low, and this causes a distortion/shortage. Whether the NGDP and CPI are rising or falling doesn’t enter in to the Hayekian Triangle.

    Zachary Woodman NGDP targeting is not more expansionary than money supply fixing . At the very least it is on net the same because during a period of malinvestment (like the 2005-2008) it forces the fed to stop artificially reducing interest rates with expansionary monetary policy as NGDP skyrockets . That’s part of the reason why Hayek himself supported it .   Let me quote Horwitz on why Austrians should support NGDP targeting: One thing policymakers can do is ensure that, when enough time has passed, market participants will return to an institutional environment conducive to the market process . This requires addressing two major problems moving forward: monetary instability and moral hazard…In our view, monetary stability means continuously adjusting the supply of money to offset changes in velocity . Given the current monetary regime, where such adjustments are in the hands of the central bank, they should be made as mechanical as possible . Discretionary monetary policy unnecessarily introduces instability into the system with little or no offsetting benefit. Instead, the Fed should commit to a policy rule . Given our monetary equilibrium view, we hold that the Fed should adopt a nominal income target. Although nominal income targeting would require price adjustments in response to changes in aggregate supply, these particular price changes convey important information about relative scarcity over time and would be much less costly than requiring all other prices to change as would be the case under a price-level targeting regime … Under a nominal income targeting regime, monetary policy would have the best chance to maintain our goal of monetary equilibrium , at least to the extent that central bankers can accurately estimate and commit to follow an aggregate measure of output . As imperfect as this solution would be, we believe it is superior to the alternatives available in the world of the second best, and certainly an improvement over the status quo of the Fed’s pure discretion in monetary policy and beyond.   What’s missing from your analysis is a discussion of changes in velocity I mentioned earlier and an understanding of monetary equilibrium.

    Andrew Criscione You said earlier:  “NGDP targeting is not more expansionary than money supply fixing.” Monetary base fixing by definition means there is no expansion . NGDP targeting means enormous monetary base expansions during a depression, the likes of which would make Bernanke blush. The Fed’s QE was pocket change compared to the amount that would have needed to be printed to keep NGDP from dropping even a penny. Here is a good Austrian criticism of NGDP targeting.   Now, is monetary-base targeting necessarily a good answer? No. I think it’s less bad.   When the Fed lowers rates i.e. expands the monetary base, this eliminates the natural risk signal to investors and causes bubbles. This is true whether the NGDP is increasing, decreasing, or constant. Stop blowing bubbles, and they’ll stop popping, but if you keep blowing them, they’re going to keep popping and causing problems. It’s just as easy to blow a bubble during a downturn as it is during an upturn. ABCT doesn’t magically stop existing during depressions.   Whenever the Fed expands the money supply, it lowers interest rates below what they would otherwise have been. This causes a distortion. It might not necessarily cause NGDP to increase, but there will be a very dangerous bubble, even if NGDP is decreasing.   The interest rate matches supply and demand for savings/loans. When it is fixed too low, there is a shortage of savings. Until it is allowed to rise, until the money printing stops, the depression will continue . This is why the Depression of 1920 lasted one year, when the interest rate was allowed to rise, unlike during the early Great Depression. The market monetarists or monetary equilibrium theorists can’t explain this, only the Old-School Austrians can. If you look at the savings rate in 2007, it was the lowest since the Great Depression, and thanks to QE, nothing has changed since. Until the interest rate recovers to market levels, there will continue to be a shortage of savings and a credit crunch.   Why would a different set of economic laws apply during the “contraction” period?  I just don’t buy that ABTC has nothing to say about the crunch: Half of Garrison’s Mises Institute lecture on this theory is on the crunch and how the economy recovers by removing the shortage of savings via an increased interest rate. I don’t see how lowering the interest rate again through printing would do anything but prevent the recovery from occurring.

    Zachary Woodman To see what makes Friedman’s proposal inferior to NGDP targeting, let’s turn to the quantity theory of money. Friedman calls for a constant growth rate in M . However, let’s say there’s a sudden change in velocity (which often happens during the period of resource misallocation in the Austrian Business Cycle Theory). Friedman’s rule calls for still growing the rate of M  constantly regardless of what happens to V. If V accelerates during a speculative bubble, continuously growing M keeps interest rates low which exacerbates malinvestments. By contrast, when we target NGDP we are targeting not just one variable, but the equation, M*V . So if there’s a sudden spike in V (as happened in 2005-08) , the fed is forced to tighten its money supply , thereby allowing real interest rates to stick much closer to the natural rate of interest and helping to lessen the malinvestment. That cannot happen if you are simply targeting M.   This is exactly why Hayek supported NGDP targeting, and why Horwitz does. See again the paper I cited by Horwitz, where he says his monetary equilibrium view is “characterized as a desire to keep the MV side of the quantity equation constant , and allow P to move inversely to Y.” And also: “In our view, monetary stability means continuously adjusting the supply of money to offset changes in velocity. ” I would recommend you read that paper in the entirety, it also lays out the Rothbardian view you seem to be taking and that the types at the Mises Institute take and distinguishes his own “monetary equilibrium” view from that.   So there is little doubt that NGDP targeting has a major advantage of preventing bubbles as compared to Friedmanite money-supply targeting . However, I suspect the main area of our disagreement is the post-bust policy prescription. You argue that because NGDP is falling, the Fed with an NGDP-targeting rule would be expanding monetary policy , which is inherently bad as that would create other bubbles. That is a potential problem with NGDP targeting , but let me explain why I think you’re overstating it and are slightly misunderstanding Hayek’s theory.   First, let me answer this question which you ask: “Why would a different set of economic laws apply during the ‘contraction’ period?” It’s not that a different set of “economic laws” apply, the very same insights from the Hayekian triangle apply. The point is that it is a case of things going too far in the opposite direction.   It does not follow from Austrian Business Cycle theory, especially Hayek’s “The Structure of Production,” that central banks should never increase money supply . If interest rates are diverging from the natural rate in the positive direction (if it becomes artificially too high rather than too low) , which is the case when there are deflationary spirals (such as 1930-33 during the Great Depression), that can also create massive distortions and create underinvestment in capital .   The key point here is that a deflationary approach during the contraction period may create its own distortions. This can be illustrated by the hypotenuse on the Hayekian Triangle being stretched into the outward direction; it artificially creates incentives to invest in lower level capital goods rather than higher ones creating less than optimal resource allocation . Let me quote Hayek on this matter: “The moment there is any sign that the total income stream may actually shrink [during a post-bust deflationary crash], I should certainly not only try everything in my power to prevent it from dwindling, but I should announce beforehand that I would do so in the event the problem arose.” This is why Hayek agreed with Friedman on the great depression: “I agree with Milton Friedman that once the Crash had occurred, the Federal Reserve System pursued a silly deflationary policy. I am not only against inflation but I am also against deflation. So, once again, a badly programmed monetary policy prolonged the depression.”   However, you do point out the problem that the bank does not know how much to expand the money supply because the “natural” rate of interest is not directly observable. There is the danger that, as you point out, the Federal Reserve could act far too much in the opposite direction, expanding the money supply excessively and driving interest rates far below the natural rate creating more resource misallocation, creating an inflationary boom.   Obviously, the best solution is to leave it to the market with a free banking system. However, that is not politically possible so we are in the world of choosing a second-best . I argue that NGDP targeting is the second best for the reasons I lay out, though I readily admit it is far from perfect. But most of the imperfections are more problems inherent in the existence of a central bank than any particular problem with NGDP targets: don’t confuse your criticisms of the monetary regime with the specific monetary policy .   That leads us to this question: has monetary policy been too tight or too loose in the last few years? The view you are reacting to is Sumner’s, which is that we needed more QE and the Fed did not do enough. I disagree with Sumner, placing me closer to your view. I’ll let Horwitz comment: It is true that nominal GDP growth—which averaged 5.7 percent a year from 1986 to 2006—began falling in 2007 and turned negative in late 2008, suggesting that monetary policy was too tight. But if the last decade was marked by monetary expansion (as we argue above), the 5.7 percent benchmark overstates the sustainable long-run GDP growth rate. As such, the fall in nominal GDP growth might very well indicate the inevitable return to the sustainable long-run trend. Whether the decline in nominal GDP growth is evidence of monetary disequilibrium is harder to discern given that the prior years of growth took place in an inflationary environment, but the magnitude of the decline suggests that some increase in the monetary base over the past two years was likely the appropriate response. Even if the Fed acted in the right direction, it is not at all clear whether this expansion was too big or too small.   My own view is a combination of the two: the Fed likely expanded the monetary base too much, however that was offset by the fact that it was paying interest rates on reserves, distorting the transmission mechanism in monetary policy. So it was too tight in one sense (interest on reserves), but too lose in another (open market operations). On net, it was too tight  (though not nearly as much as Sumner thinks). Now that the Fed has expanded the base too much, we can’t put the toothpaste back in the tube (I suspect you’ll agree with me on this point) . My solution: we need to stop paying interest on reserves and adopt a NGDP targeting policy.   Though you might interpret that as reinflating another bubble, I think the long-term benefits outweigh those short-term costs . NGDP will rise relatively quickly as M1 money supply growth catches up with the monetary base, which will lead to some inflation, lower-than natural interest rates for a short time (most likely no more than a few months), and a limited amount of resource misallocation. But the minute the bubble forms the NGDP targeting will force the fed to stop the printing presses, emulating exactly what you think needs to happen. That sudden tightening of money supply will do exactly what NGDP targeting is designed to do: stop the bubble before it can get out of control.   You are likely going to say at this part that there is going to be a recession from the resource allocation. However, that is not the fault of NGDP targeting, that is the fault of the bad policies adopted before the target was taken. Even under your proposed policy of immediately stopping the presses, this short-term problem exists . Let’s adopt your policy say we just stop the presses without taking an NGDP target: the low interest rates, in your view, of the last few years have already led to malinvestment that needs to be worked out so a bust will inevitably happen. That is the same short-term cost that you say is the problem of NGDP targeting. The difference comes in the long term: now NGDP targeting has tied the hands of the fed significantly reducing its ability to cause problems in the future . In short, your proposed policies carries the same short term costs as NGDP targeting without the long-term benefits. In fact, NGDP targeting effectively implements the very policy you want: halting the presses if a bubble begins to arise. Please tell me if there’s anything here that is unclear, or you think I’m wrong on.

    Andrew Criscione I think the Old School Austrian suggestions to the central bank were better than Hayek and Friedman’s. Probably, it was Hayek’s biggest mistake apart from his invention of Obamacare in The Constitution of Liberty (which was later picked up by Heritage in the 80s and Romney in MA). Hayek and Friedman said to print giant amounts of money during recessions to keep the price level stable (Friedman) or the NGDP stable (Hayek). Note that Hayek is more inflationary than Friedman, since, when we reach the point of stable prices yet a declining NGDP, Friedman says to turn the presses off.   The Old School Austrian approach is to allow the recession to run its course: The rising interest rates and the crash is the cure to the boom. The boom is the disease, and the bust is the cure.   “Though you might interpret that as reinflating another bubble, I think the long-term benefits outweigh those short-term costs”   I don’t see any long-term benefits to expanding the money supply drastically (QE would have had to been bigger to make the NGDP not drop a penny if NGDP targeting were instituted in 2007, it seems to me like Horwitz is being misleading). Every expansion of the money supply distorts interest rates. Japan is the perfect instance of NGDP targeting failing, and now the only solution the neo-Austrian, market monetarist consensus can come up with is to print even more money and abandon the goal of a stable NGDP. The fact is that trying to stabilize NGDP is as silly as trying to stabilize the price of shoes, and it’s going to cause shortages just like the state would if it tried to stabilize the price of shoes. The shortage is a shortage of savings, it’s very real, and it’s the biggest problem in the economy today. There are problems with monetary-base targeting, namely that you first have to establish a money supply i.e. pivot, which is something the Fed currently doesn’t even bother doing, read Brendan Brown’s Mises Daily for more on that.

    Zachary Woodman I think you’re really misunderstanding what you call the “new Austrians” are saying (they’re more accurately called Monetary Equilibrium theorists, but that doesn’t matter ). The debate between “old Austrians” (the Rothbardians at the Mises Institute) and “New Austrians” (Monetary Equilibrium theorists in academia like Horwitz,) isn’t over who’s more inflationary and who’s more deflationary, it’s over disagreements about the nature of money and banking itself . Most of the disagreements you claim modern Austrians have with older Austrians really don’t exist.   As Horwitz says: The monetary equilibrium approach should not be confused with the standard neoclassical position of price level stability. Both the Rothbardian and ME approaches recognize that prices—and, as a result, the aggregate price level—should fall in response to increases in productivity. Similarly, if goods become more scarce on average—as a result of a natural disaster, for example—the aggregate price level should increase to reflect this. ME theorists do not advocate stabilizing an aggregate price level. Rather, they suggest changing the money supply to offset changes in money demand and allowing the price level to move inversely to changes in productivity.   NGDP targeting allows for the price level and real output to drop when necessary. Don’t confuse real output for nominal output.   You say: “Note that Hayek is more inflationary than Friedman, since, when we reach the point of stable prices yet a declining NGDP, Friedman says to turn the presses off.”   You’re mischaracterizing the disagreement. It’s not a matter of Hayek being “more inflationary” at all. In the face of a massive short-term decrease in velocity (v) and NGDP targeting, that is when monetary equilibrium (ME) say we should be growing money supply . However, if V begins to accelerate Friedman would say keep the presses going no matter what and Hayek would say turn the presses off . And, again, in the face of productivity gains ME theorists say let the price level fall and Friedman would say keep growing Money Supply. In both those cases, Friedman would be far more inflationary than ME theorists.   It is not a matter of who’s more hawkish and who’s more doveish, and it extremely misleading to frame the debate in those terms. It is a matter of disagreement over what the monetary constitution should be.   “I don’t see any long-term benefits to expanding the money supply drastically.”   I’ll address your claim that NGDP targeting entails “expanding the money supply drastically” in a minute, but I think you misunderstood what I meant by “long-term benefit” to NGDP targeting. Namely it, completely restricts the Fed permanently from engaging in discretionary expansion. I thought we were in agreement on this, you said earlier that was also a benefit to Money Supply targeting and did not disagree that it is a benefit to NGDP targeting. Did I misunderstand you?   “The boom is the disease, and the bust is the cure.”   I’m not disagreeing with this at all. The bust is the cure, NGDP targeting allows for the bust to occur. (P is moving inversely to Y, so real GDP is certainly allowed to fall.) The difference is NGDP targeting has the added benefit of avoiding a deflationary spiral whereas just keeping monetary base constant (which is what the Mises Institute wants) does not. Sometimes, this is not necessary (eg. the twenties recession), sometimes it is (eg. Great Depression).   “QE would have had to been bigger to make the NGDP not drop a penny if NGDP targeting were instituted in 2007, it seems to me like Horwitz is being misleading.”   That is not necessarily the case. If the transmission mechanism hadn’t been distorted by paying interest on reserves, QE likely may not have needed to be as large as it was. This is what Scott Sumner means when he says that the Fed is not really doing QE; they didn’t allow for expansionary monetary policy to work because they distorted the transmission mechanism by paying interest on reserves. There is even good reason to believe that QE may have even been too large if the transmission mechanism would’ve been working (just look at how huge the growth in Monetary Base was relative to growth in M1). Just because NGDP fell does not mean QE needed to be bigger. Sumner makes that claim, but Horwitz and I completely disagree with him on that.   “Every expansion of the money supply distorts interest rates.”   That’s just simply not true. Every artificial expansion in money supply distorts interest rates. Likewise, an artificial contraction in money supply (like during the Great Depression) also distorts interest rates, in which case an expansion of money supply is necessary to offset the distortion.   Now you’re likely going to say that NGDP targeting calls for artificial expansion of money supply. That is somewhat true, but the point is that interest rates can never follow the real rate of interest perfectly so long as we have a central bank. The only point that ME theorists have is NGDP targeting is the best we can do to approximately follow it if we have a central bank. Do not confuse problems with a policy prescription that are really problems with the monetary regime.   “The fact is that trying to stabilize NGDP is as silly as trying to stabilize the price of shoes, and it’s going to cause shortages just like the state would if it tried to stabilize the price of shoes.”   That is definitely a bad metaphor. NGDP is not a price, it is an income level . You are again confusing the neoclassical argument that price level stability is needed with the Monetary Equilibirum perspective that we need to avoid monetary disequilibrium (ie., deviations from the natural rate of interest. ) NGDP targeting is not an instance of central planning any more than the Rothbardian prescription that we need to keep money supply absolutely level in the short term is. You ignore the fact that free banking theorists argue a completely free banking system would tend to stabilize NGDP . See George Selgin on this: How does free banking help? It does so, first, by allowing for a completely market-determined bank reserve ratio  and, second, by allowing commercial banks to issue their own currency to take the place of publicly-held central bank notes . To the extent that commercial banks are able to “capture’ the market for paper currency, the public’s preferred “currency ratio” (that is, it’s preferred mix of currency to bank deposit balances) ceases to influence the money multiplier, that is, the relationship between the stock of base money (B) and that of broad money (M). In the limit the multiplier, instead of having its usual, textbook formula of [(1 + c)/(r + c)], where r is the system reserve ratio and c is the currency ratio, becomes simply 1/r, making M = B(1/r); while the quantity of bank reserves, R, becomes equal to the stock of base money. The reserve ratio, in turn, will rise in proportion (though not necessarily in strict proportion) to the volume of gross bank clearings, that is, of payments, which will themselves depend on the velocity of money. As total payments increase, so does the demand for bank reserves. It follows that, for any given B (or, equivalently, any given nominal quantity of bank reserves) there will be a unique volume of payments consistent with equilibrium in the reserve market. Changes in V will tend, therefore, to give rise to such changes in r as will keep MV relatively stable.   Therefore NGDP targeting doesn’t try to control anything, it simply tries to emulate free banking to the best that we can with a central bank . The only response that I can see you making at this point is that Selgin is wrong about the nature of free banking . And Rothbardians typically do make this claim because his theory depends on Fractional Reserve banking which they think wouldn’t exist in a free banking system. (I contend there is very little doubt that they are wrong about that, but that’s a debate for another day.) The point is the core of the disagreement isn’t one of “inflationary” v. “deflationary” or planning v. not-planning, it’s a disagreement over the nature of banking in the first place. Rothbardians think that keeping money level completely stable emulates a full reserve system, Monetary Equilibrium theorists contend that NGDP targeting emulates free banking the best.  That is what is at odds here, not who’s more inflationary or who’s more deflationary.   I don’t necessarily disagree with all of the Mises Institute’s proposed reform. Indeed, I agree with these: Abolishing the payment of interest on bank reserves. Strict curtailment of lender of last resort function. Long-term abolition of deposit insurance.Fed withdrawal from creating liquidity in debt markets (no more eligible bills, repo-transactions, etc.). A legal attack on monopoly power in the credit card business which results often in payers of cash not enjoying a discount.”   So long as “long-term” abolition is understood as taking steps to abolish it in the future so as to avoid creation of regime uncertainty . This is implicitly endorsed by Horwitz in the article I linked to you earlier in his discussion of Moral Hazard, which he views along with Monetary Disequilibrium as the other key problem with banking today.   However, I don’t think the author has a good idea of what they mean when they say “set interest rates free.” NGDP targeting is how I would set interest rates free as it is the only way a central bank, given its existence, can approximate the neutral rate of interest in the way I’ve already described. But, again, this paragraph reveals the nature of the disagreement: Under the gold standard the pivot was a fixed price for gold alongside the widespread use of gold coins. And so the amount of high-powered money in the world grew in line with the above ground stock of yellow metal, which occurred at a glacial, but flexible pace. The demand for high-powered money was itself a fairly stable function of income and wealth. And so the system was well-anchored. Yes, there were imperfections, including the advent of fractional-reserve banking which meant that the demand for high-powered money became less stable.   Fractional reserve banking was not an “imperfection,” it allowed for the supply of high-powered money to better keep up with demand for it. That’s why it was created, and why it flourished in free-banking systems like in Scotland  and Canada . Most of the Rothbardian criticisms of fractional reserve banking are completely false . Rothbard and his disciples disagree with basically every other Austrian-oriented person on Fractional Reserve banking (Mises included). But you see here that the disagreement is not whether monetary policy should be expansionary or not, it is in what “setting interest rates free” means. For Rothbardians, this means keeping short-run money supply completely stable to emulate the “glacial” growth in supply of high-powered money under a full reserve system ; for Monetary Equilibrium theorists, this means NGDP targeting to emulate the mechanism Selgin describes that changes in M are proportional to changes in V under a fractional reserve free banking system.   In summary, Monetary Equilibirum theorists agree with Rothbardians that output and price levels should be allowed to drop when appropriate . They disagree about what policy prescriptions best allow for this because they are seeking to emulate two different banking systems.

    Andrew Criscione NGDP targeting obviously doesn’t eliminate bubbles, as we’ve seen in Japan : It simply makes the economy stuck in a permanent bust, where a new bubble is being inflated constantly while the old one is constantly popping, with very few periods of recovery . This is what we can look forward to if the US ever institutes nominal GDP targeting. NGDP targeting may be better than what the Fed is doing now during upturns, but it’s worse than what the Fed is doing now during downturns.   “NGDP targeting allows for the price level and real output to drop when necessary.”   The necessary time for the price level to drop is during a depression, this is how to cure the bubble. But NGDP targeting says to print an enormous amount of money, more than the Fed has ever printed in a depression, since all depressions have involved shrinking NGDP in the past, despite the Fed’s massive historical interventions. So even though NGDP targeters aren’t explicitly targeting interest rates or price levels, QE by its very nature raises the price level and lowers interest rates, which is the exact opposite of what the economy needs. Money velocity AND the money supply plummeted in 1920, when the Fed took no action, so this is what a “free banking” situation would look like during a depression: A truly free banking system would not have massive inflation to offset a crisis, the opposite occurs, as we’ve seen historically. Selgin is simply wrong on this.

    Zachary Woodman “QE by its very nature raises the price level and lowers interest rates”   Not true, NGDP targeting does not necessarily raise the price level in a depression . Let’s say there’s a fall in Y. All else equal, NGDP should fall and under NGDP targeting there will be some expansionary monetary policy, true. But does it necessarily follow that to offset the fall in Y, we must raise P? Absolutely not. NGDP targeting raises both P and Y. P might raise a little, but the more important thing is raising real output.  This is because we are raising M in reaction to changes in V, not P in changes to Y which is what the Taylor Rule does in targeting inflation. To argue that NGDP targeting necessarily results from inflation during a depression does two things: first, it relies on the discarded monetarist assumption that a change in M will necessarily lead to a proportional change in P in the short-term (though that might hold true for the long-term). That is not true because, especially during depressions, V and Y are changing drastically (which is something Austrians like Mises have pointed out in reaction to Monetarists). Second, it confuses the neo-classical urge to target P with the Monetary Equilibrium urge to target M*V (and therefore also P*Y).   So no, it does not follow from the monetary equilibrium idea that we must raise the price level in a depression . If, for an example in the 1920s, there was no deflationary spiral, NGDP targeting would have done very little additionally . Monetary policy would have been slightly looser, but not that incredibly loose as, though NGDP fell, it did not fall substantially.   “The necessary time for the price level to drop is during a depression, this is how to cure the bubble.”   Why?  It does not follow from anything in Austrian Business Cycle theory that deflation (ie. a sudden collapse in prices ) must occur during a bust. In fact, almost all Austrians  who aren’t Rothbard agree that if something like the thirties (and arguably 2008) occurs when you are having a massive drain in money supply for governmental reasons (in that case, the Fed’s monopolization of clearing houses) a rise in money supply may be necessary. In that case, a rise in money supply is not necessarily distortionary but what would occur in the absence of the central bank monopolizing control of money, as Selgin argues .   As I explained earlier, artificially tight money has the same distortionary effects as artificially loose money, just in the opposite direction. If the Hayekian triangle is “thinned” by the interest rate pushed too low (as it did 1925-1927), than that distorts investments in the direction of capital goods; but if it is “thickened” (as it did in 1930-33) by the hypotenuse being pushed in the outward direction, that distorts investments artificially away from capital goods. It is possible for actual interest rates to be above or below the natural rate, to ignore that is to fundamentally misunderstand Hayek’s “Prices and Production.”   “A truly free banking system would not have massive inflation to offset a crisis, the opposite occurs, as we’ve seen historically. Selgin is simply wrong on this.”   What historical examples do you have in mind? We did not have free banking in the twenties . To emphasize this point again, it does not follow at all from Selgin’s theory that there would be massive inflation in reaction to a crisis, just an offset in M to counteract the fall in V. And though money supply does fall for a short time during a crisis, it does eventually rise again as we get out of a recovery. There is a slight delay in the mechanism Selgin describes (which he admitted). Also, I would definitely (and you probably should too) defer to him on the history of Free Banking and how it handles crises as he literally wrote the book on it.   Finally, about Japan. Your perspective is completely factually off-base. First, they never explicitly adopted an NGDP targeting so it is awkward that you keep invoking it as an example of NGDP targeting’s failure . The reason they’re not invoking NGDP targeting is relevant is because decision makers were never able to make plans according to expectations on NGDP futures, a critical component of the NGDP target as Sumner and Christensen emphasize .   Second, what caused their lost decade and the crash in 1992, I think you agree, was likely way too loose monetary policy in the late eighties . Money supply growth exceeded 10% every year up until 1990. NGDP exploded at perhaps too fast a rate. That caused a bubble and a crash .   From a Monetary Equilibrium standpoint, the Japanese monetary policy during the nineties in reaction to the bubble was too tight. Interest rates were low, but that is because of expectations of deflation in the future. Instead, Japan kept money supply growth pretty slow, especially relative to the late eighties when it was 11.5% in 1987 alone (I’m pulling my numbers from the text book Economics: Public and Private Choice by Gwartney, Stroup, Sobel, and Macpherson edition 14e). If anything, they actually followed your policy prescription with tight money and turned the presses off allowing for money supply to reach -0.4% in 1992 and and growth below 3% throughout most of the decade. Meanwhile, they’ve experienced consistent deflation  since 1998 and NGDP has fallen consistently since the early 2000s, which not what Hayek would want with his stable NGDP rule nor what Sumner would want with some degree of NGDP growth. So no, Japan is not an example of NGDP targeting failing.   Instead, Japan is the perfect example of a central bank’s commitment to price stability failing. The reason for the relatively flat NGDP growth isn’t because they are targeting NGDP, but because they are targeting P. With Y pretty much flat or falling and the central bank attempting to make P stable, P*Y has also been flat. If anything, Japan is a poster child for why NGDP targeting should be taken rather than a Taylor Rule.   Also, I never claimed NGDP targeting eliminates bubbles . I merely claimed that it is the best we can do to reduce them by allowing interest rates to stick as closely (though far from perfectly) to the natural rate as possible in the way Horwitz describes. You are right that NGDP targeting may make monetary policy too loose at times, it is far from perfect, but it is the best a central bank can do .

    Andrew Criscione If we’re going to have a Fed, we want it to do what it did in the Depression of 1920. You should really read this article by Jim Grant.  The money supply collapsed, the price level fell, the nominal GDP shrank, interest rates skyrocketed, and the economy rebounded quickly. This is the closest the Fed is ever going to get to a policy that mimics free banking. The Fed wasn’t engaging in Open Market Operations at all during this time.   I think you’ll agree that Japan has come closer than any other country in targeting NGDP, no other country has even come close. So while they may not have perfectly followed Hayek’s prescription (they printed a bit less) or Sumner’s (they printed waay less, he wants NGDP to rise), it’s the closest we get to see the policy in action.   I don’t know what the Fed’s intentions are, no one does, the important thing is how much they’re expanding the monetary base. People in Japan have the expectation that NGDP is going to be targeted, because that’s what the Fed has been doing, whether or not they’re talking about it. It might be unintentional or intentional, it doesn’t matter. The market has set its expectations, they’ve been doing it for so long, and there’s still obvious problems.

    Zachary Woodman No, is not an exemplar of NGDP targeting at all. The NGDP stability was coincidental, not a result of Central Banking policy.   Japan had the expectation of low inflation-to-deflation, which is why interest rates were so low, because the central bank was targeting the price level. The expectation was never focused on NGDP futures . What the central bank says does matter a lot for the formation of rational expectations, you seem to be assuming adaptive expectations if you think that just because NGDP was stable they would assume it would remain so.   As for the twenties depression, nobody had the expectations that the Fed really would act as the lender of last resort. It had only been in existence since 1913. After the Fed has acted as such with the establishment of the FDIC, and its last seventy years of operation, one can’t make policy based off of regime expectations  of eighty years ago.   Also, the question that needs to be answered is this: why did the Fed’s inaction  in the thirties result in a deflationary spiral and how do Rothbardians think that complete inaction would get us out of it? Horwitz and Hayek are right about the Great Depression.

    Andrew Criscione “It is possible for actual interest rates to be above or below the natural rate”   Expanding the money supply lowers the interest rate, that’s economics 101. The NGDP targeters think there should have been expansion in 2007, that QE wasn’t enough to counteract the Fed’s interest on deposits, as you put it because the Hayekian triangle was supposedly crunched up due to higher-than-market interest rates. But interest rates were near-zero, even despite massive expansion! How on earth is that higher than the market equilibrium rate? When does the market ever produce 0% interest rates !   “Why did the Fed’s inaction in the thirties result in a deflationary spiral and how the hell do Rothbardians think that complete inaction would get us out of it?”   You should read Rothbard’s America’s Great Depression.  Hoover’s price fixing and Smoot-Hawley tariffs were the main reason prices weren’t allowed to fall and the economy wasn’t able to restructure. The Fed then printed an insane amount of money in 1933, which caused another bubble that crashed in 1937. Modern university researchers agree with this view on Hoover.   The real question is: Why did doing the exact opposite of what NGDP-targeting supporters advocate work so well in 1920? I haven’t seen Horwitz answer that.

    Zachary Woodman “The NGDP targeters think there should have been more expansion, that QE wasn’t enough”   You’re, again, misunderstanding the NGDP targeters’s position. True, Sumner and the Market Monetarists think there should have been more QE. However, the Monetary Equilibrium theorists, as Horwitz says, argue that it is unclear whether QE was big enough or not. The point was that the transmission mechanism was distorted by paying interest rates on reserves (which the Mises Institute is also against). Had the Fed not been doing that beginning in 2008, the QE they put in may have been too large.   It’s not a question of “too large” or “too small” it’s a question of kind.   I agree with Rothbard that the other actions in the Great Depression (ie. the Tarriffs, the National Labor Relations Board, National Recovery Act etc.) completely distorted the market and made things worse . Friedman says that, Sumner says that, nobody relevant to this discussion disagrees. But those policies have nothing to do with the run on the banks and the destruction of money supply that occurred in 1930-33.

    Andrew Criscione You are saying that, because the money supply shrank during 1930-33, the depression happened. But correlation is not causation. The money supply shrinking is the cure. You can draw a correlation between medicine intake and illness, but that wouldn’t show its causation, the opposite in fact.   And I’m not talking about base money shrinking, just to be clear: Base money stays relatively constant as a pivot (or it used to before the Fed got rid of it). In a fractional system, a cluster of error means lots of bankruptcies, and most money is credit, so the bankruptcies shrank the larger money aggregates, but not the base money.   Two medicines are needed to cure a depression: Monetary inaction, and floating prices. In 1920 we had both. From 1930-33 we only had one.   The money supply shrank in depressions before 1913: Whenever there are a lot of bankruptcies, the money supply shrinks in a fractional reserve system unless a central bank is pumping out enormous amounts of paper by expanding the monetary base drastically. For the same reason, the nominal GDP shrank during pre-Fed depressions.   So I don’t see how a central bank pumping out lots of paper during a depression resembles free banking at all.   When the state floods the market with high-powered i.e. base money, whatever the distribution scheme is, prices are fixed below what they otherwise would have been. If the central bank shrinks the base money, i.e. the pivot, that would be bad, I agree with you there. That would cause a shortage on the other end, just flip the supply/demand curve.  But for the same reason that would be bad: Whenever the central bank expands the money supply, it fixes interest rates lower than they otherwise would have been i.e. the market rate. And I don’t have to explain to you what lower-than-market interest rates do to the Hayekian Triangle. My point is that, even during the “deflationary” 1930s and during 1920, the central bank didn’t contract the supply of the money pivot/base. The base stays relatively constant over time.   If we’re not going to go back to a gold standard, and I’m not sure myself whether that would necessarily be a good thing in and of itself , what we could do is simply make the monetary pivot the supply of cash. Print new cash for wear and tear but nothing else. Strictly tie high-powered money and the monetary base to the supply of cash. ]This way, since people don’t burn money and it’d be state policy not to either, we wouldn’t have to worry about monetary base deflation i.e. the Hayekian Triangle becoming convex. And if we follow Mises’ advice to “turn the presses off, and keep them off!”, with some modification for wear and tear on paper money I suppose, we’d never have to worry about inflation of the monetary base. Rather than try and do what Japan accidentally ended up doing plus or minus a few tenths of a % per year, which involves wildly fluctuating the base money supply, why not try to replicate what the market chose before we got into the central bank nightmare: A relatively stable monetary pivot/base, with free banking determining the larger money aggregates, interest rates, and nominal GDP.   A hokey metaphor I came up with:  A prankster sneaks into a farmer’s house and replaces half his eggs with fake eggs. Eggs are savings: The market thinks the savings are real, and acts as if they are. The farmer spends all his money on consumption and long-term investment, anticipating a huge egg harvest which won’t come. Nothing can ever justify those fake eggs being snuck in to the farmer’s hen house, since it always causes problems.

    Zachary Woodman Money supply only shrank by 5.6% during the Depression of 1920. By comparison, between 1929 and 1933, money supply shrank nearly 31%. The difference is there were a series of bank runs in 1930-33 that did not take place at all during the Great Depression . There was significantly less “contagion of fear,” to quote Milton Friedman and Anna Schwartz, in 1920 that did take place in 1930. The biggest difference between the two, as my professor Ivan Pongracic points out “was bad times in the farm belt, where the banks were especially weak and poorly diversified.”   The reason why banking panics in 1930 warranted action on the part of central banks whereas previous banking panics were resolved relatively quickly without a central bank (1893, 1907) was because the Fed had nationalize clearing houses (see Pongracic’s quote on Larry White in that article for that). I can’t seem to find data on NGDP during 1920, but I suspect it did not fall an incredible amount especially compared to the Great Depression. Monetary policy may have been slightly looser, and that may have been unnecessary in that case, but the lack of banking panics in 1920 makes it not comparable to banking panics like 1930 and what would have happened absence of Fed action in 2008 with the Central Bank in control of clearing houses.   With the given monetary regime, it is necessary most of the time for the bank to act with loose monetary policy in reaction to financial crises to avoid deflationary spirals like what happened in 1930. For that reason you are wrong when you say   “Two medicines are needed to cure a depression: Monetary inaction, and floating prices.”   Monetary inaction is sufficient if there aren’t massive bank closures, but monetary action is necessary so long as you have monopolization of clearing houses.   “My point is that, even during the ‘deflationary’ 1930s and during 1920, the central bank didn’t contract the supply of the money pivot/base, i.e. gold. The point is that the base stays relatively constant over time.”   Not true, the bank did restrict monetary base in 1936, which caused the onset of the secondary depression in 1937.   “So I don’t see how a central bank pumping out lots of paper during a depression resembles free banking at all. I know we’ve never had free banking, but to the degree that we’ve come close, the market responded in the exact opposite way to depressions, than Selgin is saying.”   Again, that wasn’t free banking in all those depressions (eg., 1836, 1872, 1893, and 1907). If you want to assess how free banking reacts to crises, turn to Scotland  and Canada. Again, Selgin (along with Larry White) literally wrote the book on that. That view is also fortified by Roger Garrison in the final section of this paper.   Furthermore, it might be true that there are short-time falls in money supply as there is some delay in the mechanism of providing liquidity in face of a fall in V.   “Why not try to replicate what the market chose before we got into the central bank nightmare: A relatively stable monetary pivot/base, with free banking determining the larger money aggregates, interest rates, and nominal GDP.”   First of all, because we are currently operating under a fundamentally different banking system. Clearinghouses can’t work the way they did prior to the Federal Reserve. I, again, turn over to George Selgin: In short, free up the banks all you like; today, in the U.S., they will continue to receive and pay fiat Federal Reserve dollars, so long as no steps are taken to actually demonetize such dollars. Banks might, of course, also offer notes and deposits denominated in other less popular but still well-established currencies; and a few might even offer gold accounts and notes. But such non-dollar bank monies will be but tiny sideshows compared to the main act.   And it will be a rare bank indeed that dares to enter the base-money-creation business, the rest remaining content to leave that business to central banks. It follows that, because it leaves the base regime largely unaltered, a move from regulated to free banking today would not serve to eradicate inflation or otherwise guarantee monetary stability. Such a move would have led to improved stability a century or more ago, because it would have entailed depriving central banks of their role as currency suppliers: so long as gold and silver were economies’ final settlement mediums, to deprive central banks of their paper currency monopolies was equivalent to reducing if not eliminating altogether any tendency for other banks to treat central bank paper (or other central bank liabilities) as a reserve medium, and hence as what might be termed “pseudo” base money. The strict dichotomy of bank- and base-regime that applies today did not, in other words, pertain to specie-based monetary systems.   Today, however, the strict dichotomy is quite valid; and this means that freedom of banking alone will no longer suffice to make our (or any) monetary system sound. Something else is needed, then. And that something must of course consist of a reform of the base regime itself. Broadly two alternatives exist for such reform. These are: (1) the restoration of a base medium consisting of some form of specie, or perhaps of some other commodity; and (2) reform of the existing fiat regime. Both options have advantages and disadvantages. A major advantage of the second is that it is likely to be less disruptive. This advantage isn’t itself decisive. But it does supply one important reason for not simply dismissing out-of-hand proposals for imposing strict rules upon fiat-money issuing authorities, including rules that call for targeting NGDP. Where people have become long accustomed to using fiat money, the scarcity of which necessarily depends on some sort of “central planning,” to suggest a better central plan, instead of merely insisting that people “ought” to use gold (or forcing them to use it when doing so may seriously disrupt their plans), doesn’t make one a pinko–not, at least, so long as one also insists that there be no barriers in the way of people switching to gold voluntarily. It’s easy enough to say, in hindsight at least, that Imperial Russian authorities screwed-up when they decided on a railroad gauge broader than that used elsewhere in Europe. But it doesn’t follow that ripping up the old tracks post-haste, or just neglecting them, is a good idea. With base monies likewise, there is such a thing as sunk costs.   Second of all, you are ignoring the concept of money demand and the distortionary impacts that comes from changes to purchasing power in monetary disequilibrium. What is really in contention here are you are taking a Rothbardian approach and I am taking a monetary equilibrium approach to monetary theory. I would recommend Steven Horwitz’s discussion of Rothbard in Chapter 5 of Micro Foundations and Macroeconomics: An Austrian Perspective starting around page 170. The short and dirty of it is that Rothbard defines inflation and deflation in such a way that is so heavily linked to his belief in full reserve banking that he doesn’t allow for the concept of money demand. In doing so, he argues just to keep money supply constant (as you say) at least in the short term. However, if money demand suddenly falls (ie., velocity increases) and money supply is held constant, Rothbard acts as if the Purchasing Power of Money can just quickly adjust. However, that can only happen *via changes in individual prices* that, in the process, undermines the discovery process on the microeconomic level and leads to highly distortionary effects. In short, if we just keep the monetary base and money supply constant and allow for the price level to fall, that price level carries with it changes in relative prices that distort economic investment as well.   In sum, here are three flaws with keeping monetary base and money supply constant: the distortionary effects of changes in the purchasing power in money when money demand changes, and the fact that our current monetary regime of holding clearinghouses creates the possibility of massive run on banks creating a deflationary spiral, and the fact that trying to reform our base regime is highly disruptive of the plans people have made, creating regime uncertainty.   In conclusion, let me more explicitly answer this question you raised earlier:   “The real question is: Why did doing the exact opposite of what NGDP-targeting supporters advocate work so well in 1920? I haven’t seen Horwitz answer that.”   First of all, see Horwitz’s discussion of that exact question here: “Put differently: the 1920-21 episode was, in fact, a severe, though not particularly long, recession. Allowing the money supply to fall isn’t painless. Allowing the money supply to fall in an environment of severe downward wage rigidity is VERY ‘not painless.’ The 1920-21 episode doesn’t demonstrate that deflation is harmless. It DOES demonstrate that if you have deflation, it will be less bad if you have nominal wages that are flexible downward. That recession is the best example of why Hoover’s wage policies were such a mistake. Hoover, as Secretary of Commerce, wanted to intervene in that recession but Harding held him off and we had ‘only’ a severe recession. The 1920-21 episode does not, in my view, show that deflation by itself is painless.”   You likely won’t be satisfied by that point, but we are in the world of counter-factuals and you have to admit that Steve Horwitz’s counterfactual is, at least, possible. This is a question that economic theory can answer to help us with a conception of this raw data since history is not giving the direct question, and Horwitz does have a point that sudden changes in the purchasing power of money do happen with Cantillon effects that follows directly from the Hayek-Mises emphasis on the epistemic value of relative prices in the discovery process by entrepreneurs. With that theory in mind, the take on the empirical episode of 1920 that Horwitz takes does seem, at the very least, plausible. Second of all, I would respond to that question with my previous point that there wasn’t systemic bank runs that were experience during the Great Contraction. Perhaps it is the case that deflation isn’t as horrible with less wage rigidity caused by interventionist policies such as those of Hoover and Roosevelt *and* the systemic bank failures. But it doesn’t follow from the 1920 depression that inflation is not horrible as there are key differences–both in the nature and extent of the crisis and the policies surrounding it–between it and the Great Depression.

    Andrew Criscione  You say: “In short, if we just keep the monetary base and money supply constant and allow for the price level to fall, that price level carries with it changes in relative prices that distort economic investment as well.”   You cannot “keep the monetary base and money supply constant”. You are conflating two opposite policies. If the monetary base is kept constant and the money is not debased by the government through monetary base expansion, the market will determine the larger money supply, including bank deposits, etc. If the larger money supply is fixed, as Milton Friedman argued for, then QE and monetary base contraction will have to alternately be engaged in order to fix the larger money supply at some number. If the nominal GDP is fixed, alternating between QE and monetary base deflationwill also have to occur. If interest rate targets are pursued, as in the Taylor Rule, alternating between QE and monetary base deflation will have to occur. Monetary base fixing, by definition, involves no QE or monetary base deflation, and is the only policy that involves no QE or monetary base deflation.   We no longer have a monetary base/pivot, but a pivot could be re-established, whether or not it involves gold. This would chiefly involve ceasing interest payments on Fed deposits, which we’re both in agreement would be a good thing. The question is: Once we establish a clear monetary pivot again, should we debase it?   Advocates of a privately or publically run gold standard, whether they supported free banking or full-reserve banking, were opposed to debasement of the monetary base by the state, when the monetary base was gold. “Debasement” is a curse word in the Austrian literature: The Mises Institute has compiled pretty much all of it, and a quick search for the word “debasement” on their website returns 186,000 results, check for yourself to see whether any of them are positive. I can’t see any of the older Austrians being in favor of this, they were fiercely opposed to debasement of the pivot/base when it still involved a strong or loose tie to gold, so why would removing that last vestige of fiscal/monetary restraint from the state make Austrians more eager for monetary base expansion than they were before?   As Mises put it in The Theory of Money and Credit: In theoretical investigation there is only one meaning that can rationally be attached to the expression inflation: an increase in the quantity of money (in the broader sense of the term, so as to include fiduciary media as well), that is not offset by a corresponding increase in the need for money (again in the broader sense of the term), so that a fall in the objective exchange value of money must occur.   So: How do we determine demand for money? Well, one thing is certain: Like demand for any other good, it cannot be determined by the state. If the market is allowed to determine the larger money aggregates, then supply and demand will meet. Note that Mises was friendlier than Rothbard to free banking. But if the state interferes with this process by debasing the monetary base (which also fiddles with interest rates, whether or not that’s what the Fed is trying to do), then this is inflation, in the sense that Mises defines it. And in this sense: Inflation always causes harmful distortion.   The Hayekian Triangle becomes distorted/concave when the money supply increases faster than demand for money due to state expansion of the monetary base: There is overinvestment in final-stage capital and early-stage capital, with a shortage of investment in mid-stage capital. This is because the artificially low interest rate that monetary base expansion produces leaves a gap between savings supplied and savings demanded: This gap is made up for through the newly created money arising from monetary base expansion.   In order to fix this distortion, investment in mid-stage capital must increase, and investment in early and late stage capital must decrease. They must increase and decrease, respectively, until the Triangle is flattened out. In order to incentivize investment in mid-stage capital, and discourage investment in early and late state capital, the interest rate must rise from its sub-market price-fixed level back to its market level. It was at its sub-market price-fixed level because of monetary base expansion by the central bank. In order for the interest rate to rise back to its market level, the central bank must stop engaging in the activities that are fixing its price below the market level: Monetary base expansion.   Can the Hayekian Triangle become convex, i.e. can interest rates overshoot market levels due to Fed contraction of the monetary base (or even not enough Fed expansion of the monetary base)? Well, this would mean that interest rates are higher than market levels. During a boom, interest rates are too low, and during a bust, they must recover to at least market levels. This means that interest rates need to rise at least a little during a bust. Interest rates are certainly NOT above their market levels if they are lower during the bust than they were during the boom, which is the case almost everywhere today thanks to monetary stimulus. Central banks almost universally lower interest rates during recessions, with a few historical exceptions like Warren Harding’s central bank. So the arguments for monetary expansion on the grounds of the Triangle being convex make no sense in this day and age.   Selgin and Horwitz say the monetary base should be expanded enough to counteract any fall in NGDP. The NGDP has fallen during nearly all recessions in modern times. But yet the Fed has always lowered interest rates during recessions, instead of allowing them to rise to their market level, and so we know that the Hayekian Triangle isn’t convex. This means that the Hayekian Triangle is concave and NGDP targeting will prevent it from snapping back into place during a correction.   Ideally, we should end the Fed. If that can’t be done, a stop-gap measure we both agree on would be to re-create a monetary pivot by stopping interest on Federal Reserve payments and other basic reform measures. The question, then, is what to do with this pivot of high-powered money that was once the gold base. If the government could magically produce new gold at will like Scrooge McDuck in this old cartoon, should it fiddle the dial up and down to try and hit some statistical target that has deep problems (there are dozens of different ways that various governments and economic agencies calculate NGDP, it ignores black market activity and has numerous other problems that all economists admit), or should it leave this mystical dial alone? This is essentially what the government will have control over, if we re-create a monetary pivot, and we base targeters just want the government to leave that pivot/base be, with no debasement, or if we must, with a debasement that mimics the historical rate of increase of physical gold (2-3% a year).