I hate getting into debates where different bloggers go back and forth forever and never reach any conclusion. I am not blogging to get into debates, however, I must admit that Steven Williamson’s recent posts on NGDP level targeting have provoked me quite a bit.
In his first post Williamson makes a number of claims, which I find highly flawed. However, Scott Sumner has already at length addressed most of these issues in a reply to Williamson so I don’t want to get into that (and as you guessed I am fully in agreement with Scott). However, Williamson’s reply to Scott is not less flawed than his initial post. Again I don’t want to go through the whole thing. However, one statement that Williamson makes I think is a very common mistake and I therefore think a comment is in order. Here is Williamson:
“The key problem under the current circumstances is that you can’t just announce an arbitrary NGDP target and hit it with wishful thinking. The Fed needs some tools, and in spite of what Ben Bernanke says, it doesn’t have them.”
This is a very odd comment coming from somebody who calls himself a (New) Monetarist. It is at the core of monetarism in the sense of Friedman, Brunner, Meltzer, Cagan, Schwartz, Warburton and Yeager etc. that nominal GDP is determined by the central bank and no monetarist has ever acknowledged that there is a liquidity trap. Williamson claims that he does not agree with everything Friedman said, but I wonder what Friedman said he agrees with. If you don’t believe that NGDP is determined by the central bank then it makes absolutely no sense to call yourself a monetarist.
Furthermore, if you don’t think that the Fed can hit an NGDP target how could you think it could hit an inflation target? Both changes in NGDP and in prices are monetary phenomena.
Anyway, let’s get back to the question whether the central bank can hit an NGDP target and what instruments could be used to hit that target.
The simplest way to do it is actually to use the exchange rate channel. Let’s assume that the Federal Reserve wants to increase the US NGDP level by 15% and that it wants to do it by the end of 2013.
Scott has suggested using NGDP futures to hit the NGDP target, but let’s assume that is too complicated to understand for the critics and the Fed. Instead the Fed will survey professional forecasters about their expectations for the level of NGDP by the end of 2013. The Fed will then announce that as long as the “consensus” forecast for NGDP is below the target the Fed will step up monetary easing. The Fed will do the survey once a month.
Let’s start out with the first announcement under this new regime. Initially the forecasters are skeptical and forecast NGDP to be 12% below target. As a consequence the Fed announces a Swiss style exchange target. It simply announces that it will intervene in the FX market buying unlimited amounts of foreign currency until the US dollar has weakened 20% in nominal effective terms (and yes, the Fed has the instruments to do that – it has the printing press to print dollars). I am pretty sure that Williamson would agree that that directly would increase US NGDP (if not I would love to see his model…).
The following month the forecasters will likely have moved their forecasts for NGDP closer to the target level. But we might still have too low a level of forecasted NGDP. Therefore, the Fed will the following month announce a further “devaluation” by lets say 5%. The process will continue until the forecasted level for NGDP equals the target level. If the consensus forecast starts to overshoot the target the Fed will simply announce that it will reverse the process and revalue the dollar.
Therefore there is certainly no reason to argue 1) that the Fed can not hit any NGDP target 2) that the Fed does not have an instrument. The exchange rate channel can easily do the job. Furthermore, if the Fed announces this policy then it is very likely that the market will be doing most of the lifting. The dollar would automatically appreciate and depreciate until the market expectations are equal to the NGDP target.
If you have heard all this before then it is because this a variation of Irving Fisher’s compensated dollar plan and Lars E. O. Svensson’s foolproof way out of a liquidity trap. And yes, I have previously suggested this for small open economies, but the Fed could easily use the same method to hit a given NGDP target.
Update: I should note that the example above is exactly that – an example. I use the example to illustrate that a central bank can always increase NGDP and that the exchange rate channel is an effective tool to achieve this goal. However, the numbers mentioned in my post are purely “fictional” and again it example rather than a policy recommendation. That said, I am pretty that if the Fed did exactly as what I suggest above the US would very fast bee out of this crisis. The same goes for the ECB.
Update II: Marcus Nunes and Bill Woolsey also comment on Williamson. Nick Rowe comments on David Adolfatto’s anti-NGDP targeting post(s).
cthorm
/ July 4, 2012Hey Lars –
“no monetarist has ever acknowledged that there is a liquidity trap.”
I’d be interested in hearing your take on that, though I suspect we’re in agreement. I think that a liquidity trap exists if and only if central bankers believe it to exist; its a mind over matter sort of thing. If central bankers bothered to read Lars Svensson (or yourself, Scott, McCallum, Selgin…) then the liquidity trap idea becomes a sort of bizarre numerology. This interpretation makes the ISLM approach taught in schools seem particularly insidious.
So much as it annoys me, for investors what is relevant is the beliefs of central bankers, and in many countries these bankers seem to believe in liquidity traps. Now it makes me wonder as a consumer whether to bet on the current beliefs of central bankers persisting, or betting on a shift to a market monetarist perspective. Take a floating rate mortgage today, because the zero lower bound can last at least a decade, or take a fixed rate mortgage because the Fed will boost inflation/get NGDP targeting going and thus boost rates.
Lars Christensen
/ July 4, 2012Cthorm, that is a very good point. Obviously investors do general acknowledge that central bankers believe that there is a liquidity trap. And there is also a kind of a liquidity trap – an institutional or mental liquidity trap. Central banks seem to be paralyzed once interest rates get close to zero even though it is as you say just bizarre numerology. That is also why it is probably fully rational for investors to a the present to act as if we are in a debt-deflation scenario or a Bill Gross style “New Normal”. However, it would not have to be like that, but unfortunately there are more people on the FOMC that think like Williamson rather than like Scott or me.
David Eagle
/ July 4, 2012I both agree and disagree with Lars about the Federal Reserve being able to meet the NGDP Target. I agree with Lars that the Federal Reserve should be able to increase nominal spending by announcing a creditable, nonreversible NGDP target, and do quantitative easing to try to achieve that end. However, I disagree with Lars on other grounds. I do not believe that Federal Reserve has direct control over nominal spending (as measured by NGDP), economic agents have that control, not the Federal Reserve. By Lars’ definition, I am therefore not a monetarist. Also, Lars’ said that no monetarist has ever admitted the existence of a liquidity trap. Before 2008, I admit I thought the existence of a liquidity trap as preposterous, but I have since admitted the existence of a liquidity trap. Therefore, by Lars’ definition I am not a monetarist. So be it; I am what I am. I believe what I believe. However, but I am still a strong advocate of NGDP targeting, as strong an advocate as Lars or maybe even stronger.
My dad had a saying, “You can lead a horse to water, but you can’t make him drink.” That is my view with regard to the Federal Reserve. The Federal Reserve can put money through quantitative easing into the economic system, but the Fed cannot force economic agents to spend the money. Instead, the Fed must encourage the economic agents to spend the money. That is how the Fed has failed. The reason is that the Fed, like most other western central banks, has been thought to be targeting inflation. The “let bygones be bygones” property of inflation targeting actually encourages economic agents to sit on their money (i.e., a liquidity trap). On the other hand, if the Fed were to credibly target NGDP instead, then when NGDP is sharply below target, economic agents get the message that the Fed will be encouraging above normal inflation as nominal spending catches up to the NGDP target. Under such expectations, economic agents would be nuts to sit on their money and let the above inflation erode the value of their money. As such, the economic agents will be much more likely to spend that money, leading to an increase in nominal spending.
In summary, while I now recognize the existence of liquidity traps and I recognize that the Fed does not directly control nominal spending, I believe that the reason for the liquidity traps and the lack of the Fed’s influence on nominal spending is caused by the Fed (or other central banks) in essence targeting inflation, at least in the eye of economic agents. As I have said several times before and Lars has echoed, expectations is very important to the effectiveness of monetary policy.
Even prior to 2008, I have argued that inflation targeting leads to price indeterminacy. I now see that that price indeterminacy has manifested itself in a liquidity trap. NGDP targeting, on the other hand, does determine prices. One of my summer projects is to write a paper explaining my perspective.
Lars Christensen
/ July 4, 2012David, thanks for the comment.
Yes, you are most definitely as strong an advocate of NGDP targeting as me. In fact you are right – you might be a bigger advocate of NGDP targeting than me. And still consider as you know that I believe that you have provided the best theoretical foundation for NGDP targeting. Unfortunately I don’t think Williamson has studied your research.
Back to the liquidity trap. I think we have a slightly different definition of the liquidity trap. Yes, the inflation targeting regime means that we get into a situation which looks like a liquidity trap. However, I am sure that you would also agree that the Fed can influence NGDP and NGDP expectations via for example FX intervention as I have suggested. If you agree on that then there is no liquidity trap in the traditional sense.
Max
/ July 4, 2012I don’t think exchange rate intervention is a good idea for a large country. For one thing, it’s a hostile act given that other countries have exactly the same issue. And it can’t work without their cooperation, since they have the power to undo the intervention.
Lars Christensen
/ July 4, 2012Max, thanks for your comment. I have written a long answer: http://marketmonetarist.com/2012/07/04/is-monetary-easing-devaluation-a-hostile-act/
cthorm
/ July 4, 2012A nearly identical effect could be had without targeting a bilateral exchange rate. Targeting the dollar index, or the price of Brent in $, or the Russell 2000 index, or TIPS, would do much the same without creating animosity with a specific country. As far as I know no one is angry with Latvia. In any case it’s not about trade balance, domestic assets and liabilities are more important by far.
Benjamin Cole
/ July 4, 2012Williamson ties himself into knots trying (to what end?) refute Market Monetarism.
Milton Friedman advised Japan’s central bank to buy bonds, and keep buying until they saw economic growth, and then keep buying until they got some real inflation.
Williamson seems to be saying that central banks cannot cause inflation.
If the Fed buys Treasuries, the sellers can do two things: Invest in other assets (good) or spend the money (good).
If the Fed buys $4 billion in Treasuries in next four years, I suspect the USA would have a good recovery going, and some inflation.
If nothing would happen, as Williamson contends, then we will have eliminated $4 trillion in public debt without any cost. Our nation will be less indebted, and our children relieved of a financial burden.
I guess Williamson would be against that too, for some inexplicable reason.
Lars Christensen
/ July 4, 2012Benjamin, you are so very right. From a public finance perspective one could really hope we where in a true liquidity trap where money printing will not cause inflation. If there indeed is no relationship between the amount of money the central bank prints and inflation then the logic conclusion must be to abolish taxes altogether and finance public expenditure via money printing. I had written about that some time ago – see here: http://marketmonetarist.com/2011/11/07/taxes-and-the-liquidity-trap/
Max
/ July 4, 2012It’s not correct to say that debt monetization eliminates the public debt. It may or may not be a cheaper method of financing (this depends on the future course of interest rates), but the public debt still exists. The only way to eliminate the public debt (besides default) is the reverse of how it was created, by running a budget surplus.
I’m baffled by the idea that if somebody believes that bank reserves and treasuries are close substitutes, they must also believe that bank reserves and *anything* (e.g. nuclear submarines) must be close substitutes.
Saturos
/ July 4, 2012David Eagle, if the public isn’t spending money then they’re hoarding it, and the central bank can always give them more money than they want to hoard, at least in the long run (in the short run, too, if actual helicopter drops were legal, or trend inflation high enough).
Max, see this: http://marketmonetarist.com/2012/03/05/bring-on-the-currency-war/
Saturos
/ July 4, 2012I guess if you asked most macroeconomists they would admit that printing enough money would lead to hyperinflation. Which gives rise to Scott’s question: Why not somewhere in between, like, say, 5% NGDP growth? He points out that it’s basically cognitive dissonance. And I think that in turn is due to how Keynesians don’t like to think about budget constraints, and how they think of private spending as being constrained by interest rates alone. They forget that it also depends on nominal permanent income, which is ultimately determined by the quantity of money. Or as Scott says, why is our NGDP today 15 trillion, and not 10 trillion or 150 trillion or 16 trillion? QTM. and QED.
Saturos
/ July 4, 2012Help, my comments are getting blocked again just like on Glasner’s blog! They’re not rude, I swear!
Saturos
/ July 4, 2012Oh, phew, never mind. Sometimes my comments are getting entirely swallowed up on WordPress blogs. It’s even worse on Blogger – I haven’t been able to post a single comment on Evan Soltas’ blog.
Lars Christensen
/ July 4, 2012Don’t worry mate – you will always have access to post comments here. But you are right both WordPress and Blogger from time to time eat the comments. I have not be able to post a comment on Evan’s blog either.
I know Scott basically is unable to post comments on any blog.
dwb
/ July 5, 2012yes, SW thinks that QE (along with any other Fed policy) is ineffective in the current environment. Thats why i personally am trying to refrain from commenting on his blog these days, i find his views highly peculiar, insular, and impervious to persuasion. Thankfully, i know even Bullard does not agree (which SW readily admits).
His view on QE and the “tools” of the Fed are entirely premised on the fact that now the Fed pays Interest on excess reserves, and he readily admits that if the Fed changes that policy they might be able to effect bank lending. He readily admits that “the market” believes in QE, but does not seem to think about the wealth effect of rising stock/bond prices (which in a world of underfunded state, local, corporate, and personal pension and retirement plans, i think is significant).
Of course, dissecting this to get to the bottom of his views requires decrypting numerous blog posts, which i don’t have much time for these days ( and i do mean decrypting: sometimes he seems to talk in strange code i do not understand and has very MMT-like views, and it takes me a while to find the key false assumption).