Scott Sumner  

What's my core message?

PRINT
The Shining City on a Hill: Co... Cutting Corporate Tax Rates in...

I am currently working on the final chapter of a book manuscript, tentatively entitled "The Money Illusion: Market Monetarism and the Great Recession." I am trying to identify my core message. What is the essence of my critique of mainstream macroeconomics? And why should anyone believe me? I'll offer a few thoughts, but I'd be very interested in your outside perspective. BTW, one thing is very clear to me----NGDP targeting is not at all a part of my core message, it's totally compatible with mainstream macro.

It seems to me that market monetarism has two components, the market part and the monetarism part. In my view, monetarism is the school of thought that says shifts in the supply and demand for money drive the most important macro phenomena, including key nominal variables like inflation and NGDP growth, as well as business cycle movements in RGDP and unemployment.

More importantly, monetarism argues that other schools of thought reify various contingent epiphenomena, confusing side effects with core mechanisms. Thus non-monetarists are inclined to look at phenomena like inflation through the lens of changes in interest rates, bank credit, and/or the Phillips Curve.

To a monetarist, those epiphenomena are the side effects of changes in the supply and demand for money, in an economy with sticky wages and prices. But they are not the core mechanism. Increases in the money supply and/or decreases in money demand are inflationary even if they don't move interest rates at all, and even if they don't result in product or labor market tightness. In two recent posts, I explain this idea with a parable of an island economy lacking a financial system, where prices are flexible and the economy is always at full employment.

So that's the core of the "monetarism" part of market monetarism. But what about the "market" part of the theory? I believe that the flaw in modern macro is that the efficient markets hypothesis is not deeply embedded into all of our models. Thus when there is a policy initiative such as QE, mainstream economists take a "wait and see" attitude. They say that after observing a year or two of macro data, we will have a better idea as to the policy's effectiveness. A market monetarist says that within 5 minutes we'll know everything that we will ever know about the effectiveness of the policy move. Inflation, RGDP and NGDP futures will immediately adjust to reflect the optimal forecast of the effect of the policy initiative. If those markets don't exist, then other proxies such as TIPS spreads, exchange rates, commodity prices and stock prices will tell us all that we can know about the effectiveness of the policy. The future performance of the economy will be affected by that policy, but also a myriad of other factors. Waiting and observing the future course of events won't tell us anything that we don't already know.

Market monetarists see market driven regimes for "targeting the forecast" as a sort of "end of (macroeconomic) history". They are the final stage in the long process of discovering an optimal policy rule. How can any policy ever be better than "the policy stance expected to reach the policy goal?" And how can any macro model's forecast ever reliably beat the market forecast? Not occasionally, but reliably.

And of course we argue that market forecasts of the goal variable are the most useful measure of the stance of monetary policy. Other economists look at a wide variety of epiphenomena, especially interest rates. But the response of interest rates is dependent on any number of contingent factors, and can't possible serve as a reliable indicator of easy and tight money. In the end, the only useful definition of easy and tight money is relative to the policy stance expected to achieve the policy goal---is money too easy or too tight? And again, it's market expectations that will ultimately provide the optimal forecast.

Armed with this market monetarist perspective, we re-interpret macroeconomic history, trying to zero in on the core mechanism, and not be distracted by the various side effects of monetary shocks. What are some of those distracting side effects?

1. Changes in interest rates (due to sticky prices)

3. Shocks to the financial system (due to sticky nominal debt.)

4. Shocks to the labor market (due to sticky nominal hourly wages.)

These side effects are important, but the core message of MM is that these side effects are just that, side effects. They do not drive the process. That's why one can find examples of inflation that cannot be explained by conventional models. A good example is 1933-34, when (wholesale) prices rose by 20% after a monetary shock that produced almost no change in either interest rates or the money supply. Instead, a sharp devaluation in the dollar dramatically reduced money demand (by increasing the future expected money supply), creating rapid inflation despite 25% unemployment, and despite much of the banking system being shutdown.

So why should anyone believe MMs like me? After all, from a perspective of 64,000 feet up I'm an almost complete nobody, who spent his career teaching at an average business school. What distinguishes me from 100 other monetary ranks, all making grand claims to have reinvented macro, attached to policy nostrums that can supposedly cure all our ills?

When I was a teenager I was impressed by bold, heterodox thinkers. "Yeah, how could those pyramids have been built without the assistance of aliens from outer space." As an adult, I've come to appreciate the efficiency of intellectual markets. It's very unlikely that any heterodox thinker that I read about will actually have offered an alternative theory that will survive the test of time. Objectively speaking, I'm far more likely to be just another monetary crank, not the savior of macroeconomics.

I'm not sure I have a good answer to this dilemma. I suppose I could point out that there was a period when quite a bit of praise was lavished on my blogging (here, here, here and here), by reporters and other bloggers. They seemed to think that the way things were playing out somehow validated the arguments I had been making. But that success occurred at a pretty modest level; I certainly didn't convince the overall profession. In the end, all I can do is view myself as one tiny component of the "wisdom of crowds". Yes, markets are efficient, but only because each trader is willing to take a fresh independent look at the situation, and do their best to make accurate forecasts. And yes, the market for ideas does tend toward efficiency in the long run, but only because intellectuals are willing to continually probe weaknesses in existing theory, and seek better ones.

Robin Hanson and Scott Alexander recently reviewed a new book by Eliezer Yudkowsky, which wrestles with exactly this question. I haven't read the book yet, but was intrigued by Scott's summary of one of Eliezer's examples:

Eliezer spent a few years criticizing the Bank of Japan's macroeconomic policies, which he thought were stupid and costing Japan trillions of dollars in lost economic growth. Everyone told Eliezer he couldn't be right, because he was an amateur disagreeing with professionals. But after a few years, the Bank of Japan switched to Eliezer's preferred policies, the Japanese economy instantly improved, and now the consensus position is that the original policies were deeply flawed in exactly the way Eliezer thought they were. Doesn't that mean Japan left a trillion-dollar bill on the ground by refusing to implement policies that even an amateur could see were correct?
Of course other people that I view as monetary cranks, such as the MMTers, also see events confirming their worldviews. But I believe the following Alexander observation offers a bit of evidence beyond "he said, she said":
Why was Eliezer able to out-predict the Bank of Japan? Because the Bank's policies were set by a couple of Japanese central bankers who had no particular incentive to get things right, and no particular incentive to listen to smarter people correcting them. Eliezer wasn't alone in his prediction - he says that Japanese stocks were priced in ways that suggested most investors realized the Bank's policies were bad. Most of the smart people with skin in the game had come to the same realization Eliezer had.
If I have the serene confidence of a monetary crank, or a religious nut, it is founded on one bedrock principle---it's really hard to get rich. And it's hard to get rich because markets are pretty efficient. If markets reacted the wrong way to economic news, then it would be easy to profit on that market flaw. But it isn't.

At its core, market monetarism is about the view that the best estimate of the way that the world works is roughly the way that the markets believe it works. The specifics will always be a work in progress. Market participants will continually discover new perspectives on the economy, and incorporate those perspectives into their mental model of the economy. I hope that future market monetarists disprove some of my claims, and come up with better versions of the theory. Maybe they'll discover that markets believe that fiscal stimulus is effective.

However the basic MM framework for thinking about the economy is likely to survive:

1. Shocks to the supply and demand for money drive the nominal aggregates.

2. Unexpected movements in the nominal aggregates drive fluctuations in real output and employment. They also contribute to financial instability.

3. The market forecast of key macro variables provides the optimal way of understanding what's going on with the economy, predicting its future course, evaluating the stance of monetary policy, and indeed setting the policy instruments.

To be persuasive, a monetary crank needs to explain not just why they are right, but why the mainstream is wrong. I have tried to do this in dozens of posts, all under the theme of "cognitive illusions". The mainstream has incorrect views of the operation of the macroeconomy, because MM is counterintuitive and the mainstream view is intuitive. It seems like Fed changes in interest rates are the "real thing" and not just an epiphenomenon. It seems like the theory of "supply and demand" predicts that an overheating economy would cause inflation, even though it does nothing of the sort. (Excess demand occurs when prices are held too low; it is not a theory of why prices change.)

There's no objective reason to rely on the pronouncement of an intellectual mediocrity like me. I was just lucky, with a set of research interests the dovetailed almost perfectly with what was needed to make sense out of the 2008 crisis. But Eliezer Yudkowsky is another story. He is a formidable intellect who took a fresh look at all the arguments, and ended up in the market monetarist camp. If he and I ever diverge on some core epistemological point, by all means believe him, not me.

PS. If you follow the 4 links above, you'll find that all my good press occurred on September 13 and 14, 2012. That was right after the Fed announced monetary stimulus measures to offset the impact of the fiscal austerity expected in 2013. More than 350 Keynesians predicted the fiscal austerity would significantly slow growth. We said it would not. And bloggers like Paul Krugman and Mike Konczal said this policy experiment would be a test of market monetarism.

We won. Surely that counts for something?

Has it really been 5 years? As Springsteen would say, all I'm left with is boring stories of glory days. And the feeling that I got far too much credit, especially relative to other MMs.


Comments and Sharing


CATEGORIES: Macroeconomics , Money




COMMENTS (32 to date)
BJH writes:

Great post!

I also think about market monetarism by breaking it into "market" + "monetarism". On the "monetarism" side, I think one thing I would add that you (and others) have really emphasized, particularly compared to old monetarism, is that: not just "inflation is always and everywhere a monetary phenomenon" but also that "*depressions* are always and everywhere a monetary phenomenon".

Cloud writes:

I really think you should write more books because when you are writing a book, your blog is extra good ~

(maybe because you are doing some more self-reflection or consolidation of your thinkings, I guess)

Nick writes:

As a close reader of your work and blogs, I have difficulty squaring the role of the real economy and the general market monetarist view. You often discuss supply side policy and offer tangible policy improvements. At the same time, the Fed sets AD and monetary offsets curtail impacts of certain policies.

A discussion of how to think about the short term versus medium/long term may be helpful in this, as that's where I tend to cabin these tensions.

Garrett writes:

I recently finished reading Yudkowski's "Rationality: From AI to Zombies" ebook. It's quite long, but I would highly recommend it as a primer on useful ways of thinking and approaching problems.

SG writes:

Scott, this is a good post, I have a few thoughts. As someone who has been with you nearly from the beginning (circa 2010, when I was in law school) I'm extremely excited for your book.

If market monetarism is the answer, what is the question? I think the question is: why did we have a Great Recession, and why didn't we recover from the recession sooner?

The Fed failed, from the beginning, to understand the causes of the recession. In summer 2008, the Fed was far more concerned about excess inflation than about the slowing economy. In September 2008 they failed to lower interest rates after Lehman's collapse and in the face of dire market forecasts of recession and financial calamity. In October 2008, they adopted IOR for the express purpose of putting "a floor on the federal funds rate" https://www.frbsf.org/education/publications/doctor-econ/2013/march/federal-reserve-interest-balances-reserves/. In the years that followed they persistently failed to appreciate the depth of the recession. There was year after year of embarrassing predictions that the recovery was just around the corner. http://evansoltas.com/2012/08/01/the-fed-as-little-orphan-annie/. There was the infamous "taper tantrum" resulting from Bernanke's unexpected retreat from QE3. https://www.bloomberg.com/news/articles/2017-09-20/fed-eyes-tame-balance-sheet-taper-after-tantrum-error-timeline
And there was the rate hike in 2015 that ended up being almost a year premature (the Fed waited until Dec 2016 to hike rates again).

And outside the Fed the confusion was a million times worse. Democrats/Keynesians beat the drum for fiscal stimulus, and predicted doom from the fiscal cliff. Didn't happen. Republicans predicted hyperinflation and/or a debt crisis. Didn't happen. Right wing economists said we were in a new era of structural unemployment. Wrong. Bond guys like Bill Gross and the financial press claimed central banks couldn't actually generate higher inflation/NGDP growth. Also wrong (proven when the Swiss National Bank successfully capped the value of the Swiss Franc against the euro during the euro crisis).

If you're writing to a lay audience, I think the key to making your case is explaining very clearly that all alternative explanations of the Great Recession and its aftermath are flawed/incomplete/downright nuts. Once you've done that I think then you're able to lay the groundwork of market monetarism. Stable NGDP growth is a NECESSARY condition and SUFFICIENT condition for stable labor markets.

Lastly, Eliezer Yudkowsky is a smart guy, but I would also mention that Christy Romer, Paul Krugman, Jan Hatzius, Mike Woodford, Brad Delong, Larry Summers, future Federal Reserve chairman Evan Soltas (and I'm forgetting a bunch of others) all of whom endorsed NGDP targeting or something close to it in the years after you started blogging. I'd say the slow pace of your ideas catching on at the fed mostly reflects the incredible status quo bias of that institution (as well as its poor governance, as explained in Peter Conti Brown's book) rather than any failure on your part.

Philo writes:

Society as a whole is so complex and interconnected that causation on the macro level is very hard to discern. To find the cause of some condition--boom/bust, change in unemployment, change in inflation, etc.,* we have to know, for each candidate for the role of cause, what would have happened if it had been varied keeping everything else the same ("ceteris paribus"). But this last consideration is tricky: only factors that are independent of the candidate-cause can be kept the same, while those that are in some way dependent must be allowed to vary along with the candidate itself. And these judgments of dependence/independence require a very good model of the economy (perhaps mainly intuitive).

My impression is that you have better intuitions, or a better model, than most economists. I trust you because your arguments are persuasive. (And as for getting credit, undoubtedly you are at least the star publicist for Market Monetarism.)

A side point: Scott Alexander writes that "Japanese stocks were priced in ways that suggested most investors realized the Bank's policies were bad." I consider this an instance of the Fallacy of Division: the market "regarded" (figuratively speaking) the JOJ's policies as bad, but that does not imply that most market participants took such a view. (Strictly speaking, it does not imply that any market participants took that view, though of course in fact many did.) By the very nature of its subject matter macroeconomics must be especially prey to Fallacies of Division and Composition.

*By the way, what would be the general term to cover all the conditions you want to account for; is "macro nominal aggregates" good enough [is unemployment nominal?]?

Arnold Kling writes:

I think that your core message is that money matters but the usual monetary indicators do not. That is, the central bank controls aggregate demand, but neither interest rates nor money-supply measures are useful for identifying the stance of monetary policy.

The indicator that market monetarists want to use is one that is forward-looking and reliably reflects the expected path of aggregate demand. A market in NGDP futures would appear to have these properties. Absent that, alternatives include . . .

Mark writes:

I have to take issue with your description of ‘intellectual markets.’ They’re not really markets, are they. There’s no real financial punishment or reward for being right or wrong, respectively. How many academic economists who were egregiously wrong about the financial crisis lost tenure?

In many fields (including my own) there are actually strong incentives to be publicly wrong about things. To be overly optimistic, or overly pessimistic, or to prefer simplistic models, or complex models; there are fads and trends that pervade research just like clothing and fashion, and many of the gatekeepers - the publishers and funders - are non-academics who are more interested in what sounds exciting than what’s true. This habitual deference to academic consensus is just not justifiable.

Mark writes:

Ioannidis’s article “Why Most Published Research Findings are False” is worth reading, especially for anyone who thinks academic consensus within a field invariably tends toward the truth. It’s not about economics, but a lot of it is generalizable.

Thaomas writes:
Shifts in the supply and demand for money drive the most important macro phenomena, including key nominal variables like inflation and NGDP growth, as well as business cycle movements in RGDP and unemployment.

Since the demand for money is the supply of everything that can be purchased with money, how is this not a tautology?

Increases in the money supply and/or decreases in money demand are inflationary even if they don't move interest rates at all, and even if they don't result in product or labor market tightness.

If starting from equilibrium in all markets, there is a decrease in the demand for money (= increase in the demand for everything that money can buy), how can this not move interest rates and labor market tightness? [I see that given technology and preferences a new equilibrium in which all real variables are the same and the price level is higher and the real money supply is lower], but how is that equilibrium achieved?

“A market monetarist says that within 5 minutes we'll know everything that we will ever know about the effectiveness of the policy move.

Why not 4 minutes? 4 nanoseconds? This seems to imply that all (relevant) market participants lean within five minutes of the new policy and in the same period come to believe that it will be maintained. It seems to me that only by observing the economy (for a period of time that cannot be specified in advance) can we and market participants know what the new policy is and how likely it is to be maintained. For example the Fed announced a 2% inflation target. Only little by little did we learn that the target was for a soft inflation ceiling and perhaps a soft floor of 0%.

Market monetarists see market driven regimes for "targeting the forecast" as a sort of "end of (macroeconomic) history". They are the final stage in the long process of discovering an optimal policy rule. How can any policy ever be better than "the policy stance expected to reach the policy goal?" And how can any macro model's forecast ever reliably beat the market forecast? Not occasionally, but reliably.

This seems like a reasonable recommendation, reasonable because we assume that the monetary authority’s forecasting model is not totally wrong and will improve in the course of carrying out the policy of targeting the forecast. But is assumes that the monetary authority has real autonomy to use its instruments and that everyone know that it does. These are patently false assumptions. In practice monetary authorities are constrained in what they can do. Take the Fed, it has been under enormous pressure to return nominal short term interest rates to so undefined level of “normality,” even when this is inconsistent with its stated target of keeping inflation at 2%.

These side effects [interest rates, financial system shocks, labor market shocks] are important, but the core message of MM is that these side effects are just that, side effects. They do not drive the process.

Why could there not be a change in collective risk premium or time preference or expectation of the rate of technical change or changes in assumptions about financial system solvency or increases or decreases in the supply of labor that would “drive” macroeconomic variables. And should a monetary authority not try to adjust the levels of its instruments to maintain it’s target (or possibly if it thought that the changes are permanent and not consistent with the model that it used in setting its target optimally, to change the target.)

Let’s take 1933 34. What would conventional monetary policy (trying to maintain “stable” prices and “full” employment) have done? What would MM policy (targeting NGDP) have done? You work it out, but it looks to me like both would have done more or less the same thing, goosed the nominal money supply, though I guess MM, being unconcerned with inflation per se, would have goosed it more.

And I think that shows the way to being persuasive. Do multiple period by period alternative histories, comparing conventional policy, MM policy, and actual policy. Presumably the results of this will be to show that MM policy would have been better than both actual and conventional policy.

Steve F writes:

Your core message is not reinventing macro. It is remembering macro. The Fed forgot rational expectations in 2008, and you gained popularity because the things you said were bringing back rational expectations to the Fed in ways they had lost it.

John Hawkins writes:

Scott, the section where you outline different forms of price stickiness got me thinking...

"
1. Changes in interest rates (due to sticky prices)

3. Shocks to the financial system (due to sticky nominal debt.)

4. Shocks to the labor market (due to sticky nominal hourly wages.)
"

My thoughts were, are there some sort of indices we have to measure how sticky an economy is at a specific time?

For example, with respect to the last one, innovations that bring just-in-time production styles to the service economy (like Uber) make prices less sticky for the labor market than they were before.

And for nominal debts, I have to imagine it would be easy to construct a "stickiness" index based on the weighted average life or duration of fixed payment/fixed rate loans. More concretely, an economy with more 30-year mortgages vs. 15-year mortgages is "more sticky", an economy with more 5-1 ARMs is "less sticky", an economy with entirely floating rate debts wouldn't be sticky at all, etc...

Is there any academic or professional research on this, such as is a stickier economy more susceptible to booms and busts, while a less sticky economy is generally more stable? It would seem to me we would have many good natural experiments for this.

You're point would be this is a secondary cause to the monetary changes, but it would be interesting to see how monetary changes are leveraged throughout the economy due to how "sticky" it is.

Eliezer Yudkowsky writes:

I blushed to read the Slate Star post. I think Scott Alexander may have accidentally exaggerated, in the course of summarizing a longer minibook, the amount of intellectual labor I did that wasn't simply "I can tell that Scott Sumner has the ball."

How could I tell?

- By having read econblogs long enough to watch different positions make divergent predictions about events like the effect of the sequester, and noticing the market monetarists made correct predictions and other people didn't. I think this is sometimes called "science".
- Like, a dozen different subtle alarms I'd have trouble explaining went off, that I use to identify correct contrarians.

If I had to frantically handwave at a couple of example cues:

- There's a certain tone correct contrarians often end up using that sounds like "God damn it would you listen to your own damn theory and experimental evidence even!"
- There's one way that a person sounds when they want you to be impressed with their complicated, difficult, exotic, counterintuitive theories of what is surely a terribly difficult problem; and a very different tone that correct contrarians use when they want to grab you by the throat and scream that their complicated impressive theories are totally straightforward and normal, god damn it.
- Correct contrarians often make desperate appeals to process-level and meta-level rules... correctly! This is a very helpful cue if, uh, you happen to personally know how to discriminate correct use of process-level and meta-level rules better than any adversary knows how to fake.

Not enough listeners can discriminate these cues for there to be a social advantage in speakers learning to fake them--indeed, the social advantages are usually pointed the other way--so they haven't been Goodharted into unreliability. Though I think in general you can't get by on meta alone; but fortunately in this case the object-level arguments were also, ahem, totally straightforward and normal.

But anyway, I don't think Yudkowsky is an independent source from Sumner. I mostly just figured that Sumner had the ball, and then actually believed in his conclusions when his arguments sounded straightforward to me.

I certainly didn't say anything about Japan or the ECB being silly in advance of the Money Illusion blog saying that Japan or the ECB were being silly.

Todd Kreider writes:
But after a few years, the Bank of Japan switched to Eliezer's preferred policies, the Japanese economy instantly improved, and now the consensus position is that the original policies were deeply flawed in exactly the way Eliezer thought they were.

Or more recently when growth has averaged about 1%? What policy did he advocate that economists missed since there have been differing views among economists on Japan for years? Japan's growth has oscillated for years so there was no period of "instantly improve" apart from a couple of brief moments in 2012, (then back to 0%), early 2013 (then back to 0% in 2014) and 2015 (then a slide to 0.5% in the last half of the year) with 1.6% growth from late 2016 through Sep 2017.

Let's see how long 1.6% will be sustained.


https://tradingeconomics.com/japan/gdp-growth-annual

Scott Sumner writes:

BJH, Thanks, but I'm not sure my views on depressions are much different from Milton Friedman. We both think they are often caused by tight money, but real shocks can also cause them (as in Venezuela today.)

Thanks Cloud.

Nick, Yes, that's a tricky subject, which I do occasionally post on. In general, short run real shocks tend to be monetary, and long run changes in RGDP growth tend to be non-monetary. But that's just a generalization, which doesn't hold in every case.

Garrett, Thanks for the tip.

Philo, No, unemployment is a real variable.

I believe that talking about what "the market believes" is a useful fiction.

Arnold, Good observation. I mostly agree, but just to be clear when I say "money matters" I mean something a bit different from what Friedman meant. He focused on shifts in the money supply, I put more equal weight on money demand shifts.

Alternatives include TIPS spreads, commodity prices, stock prices, exchange rates. Note that some of these are highly imperfect, and they are all somewhat imperfect. Most useful when taken as a group, and evaluated without an agenda.

Mark, There is some truth in what you say, but I'd prefer to say that intellectual markets do impose some discipline, but they are much less efficient than financial markets.

Take something like cold fusion. I know nothing about the subject, but put a lot of weight on the skepticism of conventional physicists. They aren't always right, but they usually weed out the cranks pretty effectively.

Thaomas, For example, a monetary reform dramatically moves the money supply, but interest rates don't move at all.

You asked:

"Why could there not be a change in collective risk premium or time preference or expectation of the rate of technical change or changes in assumptions about financial system solvency or increases or decreases in the supply of labor that would “drive” macroeconomic variables."

There can be. I was referring to what drives the monetary transmission mechanism.

Steve, Good point.

John, Things like uber don't have all that much effect on the overall economy, which remains full of sticky wages. My sense is that wages were more flexible during the gold standard, as evidenced by the 1921 deflation (when wages fell sharply.) But even then they were somewhat sticky. It's not easy to do the test you mentioned, as there are so many complicating factors, but I presume some studies have been done.

Eliezer, Thanks for those very generous remarks. If you are still reading this, can you give me an example of "process-level and meta-level rules..."

Is that what I'd call "methodological approaches"?

(anyone familiar with his work can also comment.)

As soon as I finish my initial manuscript I'll start in on your new book, which I can use to adjust the final manuscript. Or perhaps I should start with the book Garrett recommended?

mbka writes:

Scott,

methinks the lady doth protest too much ;-) You have your place on the pedestal, you just don't want to believe it. I've rarely seen a better evangelist for a cause than you.

As to "I am trying to identify my core message. What is the essence of my critique of mainstream macroeconomics? "

I can't tell you what your core critique is but I can tell you what your key messages are, to me, as to how monetary macro works and why it is important. You claim this:

- to maintain stability in its internal workings, the economy needs a supply of money such that a unit of account buys a near constant fraction of GDP
- to err on the side of caution, a slight and constant inflationary increase is desired
- it is impossible to accurately predict these monetary needs from first principles because real GDP changes unpredictably
- but this does not matter because one only need to adjust money supply correctly to get near stable prices
- to achieve this, you suggest that it is better to use expectations than feedback
- therefore you propose to use prediction markets in monetary futures to get the best achievable estimate for the money supply needed to maintain a stable dollar value for a GDP fraction
- once the above is fixed, monetary macro is done. At first order, no other monetary interventions are necessary.

Rajat writes:

A few comments:

1) NGDP may not be fundamental to your approach, but from the perspective of 2008, or from an Australian perspective, it has been pretty important. It is easily forgotten now (in 2017) how important the distinction between demand and supply shocks is, and how NGDP is 'the real thing'. If w/NGDP rises, hours worked falls, and vice versa. This is a huge insight for Australian readers. In the last year, we've seen very strong jobs growth, as a consequence of a boost to the terms of trade from higher commodity prices. Many think it is preordained that this continues, but it won't if NGDP growth slows - as it already has with commodity prices flattening. Very important.

2) One of your key insights has been that central bankers like to appear to behave deliberatively, moving slowly in a given direction. This creates a bias towards inertia and policy error, given that money demand can change very rapidly. I think this is hugely significant. A central bank that moves quickly, and reverses direction in response to market movements looks flakey and unserious - central bankers are meant to know better than markets and almost 'stare markets down' where they throw tantrums rather than submit to them.

3) As for why you're right and not others, I think your success is attributable to (a) a clear and simple framework that more recent generations of economists have jettisoned and (b) a degree of intellectual honesty that few can afford themselves the luxury of indulging or displaying. Unlike many of the high-fliers, I suspect you didn't have a lot to lose when you embarked on your crusade. And also, you make an effort to try to be understood by laypeople. You're not always great at concrete steppes, but you're very good at simile and metaphor.

Rajat writes:

Continuing my point 3, while intellectual markets may be efficient in a very long run sense, the long run (especially in the social sciences) is much longer than it is in financial markets. In light of the overlooking of monetarist ideas from the 1930s to the 1960s, it should come as no surprise that monetarist explanations are not taken seriously for a few decades after 2008.

Rick McIntire writes:

I'm looking forward to reading your book. One reason I've been drawn to your blog posts is the fact that you are one of the few sources referencing IOR when discussing the "mystery" of low inflation.

What effect do you think IOR would have at the extremes? For example, what do you think would happen if the Fed pegged IOR to inflation while massively increasing the size of their balance sheet?

I am curious, also, as to how your book complements and/or diverges from Robert Hetzel's book on the topic. (as a side note, one of my great joys in life is trying to explain to my 5 year old son what the cover image of Hetzel's book means -- any help in that area would be appreciated)

Thank you for the work you are doing. It is a pleasure to read.

Todd Ramsey writes:

If you are taking requests, you have introduced me briefly to two concepts that I wish you would illustrate more often in your blog, or elaborate more completely upon in your book.

First, that manipulating interest rates may have been an appropriate Fed tool in the pre-Nixon gold holding era, but not in a fiat money world. It seems like many economists, the financial press, and the general public have not fully internalized the implications of a fiat money world and are holding on to an interest rate theory that is no longer relevant.

Second, what is your recommended mechanism by which the Fed targets nominal GDP? If memory serves, you and another economist have proposed the Treasury issue a security (he named it a "Tril") whose price is based upon a future nominal GDP. The Fed would buy and sell this security on the open market, thereby creating and removing money as appropriate to hit the target. The security would replace much of the funding the Treasury currently gets from Bills and Notes.

I apologize for my imperfect layperson understanding and presentation of these concepts, and that's why I wish you would present them more often and in different ways.

You are making the world a better place! Keep pushing!

Scott Sumner writes:

Todd, The Japanese working age population is falling very rapidly. 1.6% growth in Japan is like 3% in America.

Here's what's happened since Abe:

1. Deflation has ended.

2. Unemployment has fallen to the lowest level in decades.

3. NGDP has stopped falling and started rising.

4. The public debt is no longer rising rapidly as a share of GDP.

I call that success. But yes, the BOJ should have done even more--they have not yet hit their 2% inflation target.

mbka, Good summary.

Thanks Rajat. I would just add that changes in Australian NGDP minus commodity price changes might be a better indicator of their job market.

Thanks Rick. Do you mean the book with a tornado of cash on the cover?

I'll ask him when I see him next month.

Rick McIntire writes:

Yeah, the tornado of cash. He's with me up to a certain point but when it comes to my answers to his question of "where did the money go?" it becomes clear that I'm only demonstrating the woeful inadequacy of his dad's knowledge and understanding.

Todd Kreider writes:

Scott,

1) Abe was elected at the end of 2012. Here is the growth rate since then:

2013 2.0% starts out well...
2014 0.4% ... but just 1.2% growth in his first two years
2015 1.0% So 0.9% over three years. This is impressive? A jump start for Japan?
2016 0.9% ...0.9% for Abe's first four years.
2017 1.9% ...for the first three quarters

1) Scott Alexander wrote that "Eliezer spent a few years criticizing the Bank of Japan's macroeconomic policies,...", but I can't find anything written by him on BOJ policies prior to his essay (that Alexander linked to) a few weeks ago apart from a brief comment on MR in 2016 that he agreed with you about something.

2) You wrote: "Deflation has ended." As it has a few times in the past 17 years. Inflation was at about 0% for 2015 and 2016 and only at 0.7% in 2017. Many previous years show this hovering around zero pattern.

3) "Unemployment has fallen to the lowest level in decades."

Yes, but what is the evidence that the Abe administration played much of a roll, if any? The unemployment rate has steadily fallen from 5.5% in 2009 to 2.9% in 2017. One cannot tell when Abe was elected the second time based on the 8 year unemployment decline.

4) 1.6% GDP per capita growth in Japan is not the same as 3.0% per capita growth in the US, where per capita growth is what matters.

Jake writes:

Interesting post!

The way I have always seen your critique of the profession is that in a fiat money regime, a central bank has virtually unlimited control over nominal variables. It can create as much or as little NGDP/inflation/etc as it wants.

People want to talk about fiscal policy, ZLB, real shocks, etc, but none of that matters. The central bank does not attempt to "hit a target" but merely to choose one. And the best choice is that which is chosen by the market.

That doesn't mean the correct stance of monetary policy will make everything perfect all the time. But it will be the best we can reliably make it from a macro perspective.

Kurt Schuler writes:

It seems to me that your core message is that one important current of thought, derived from Friedman, does not adequately appreciate the importance of the demand for money, assuming that it would not change quickly, and that another important current, derived from Keynes, does not adequately appreciate the importance of either the supply of or the demand for money.

Nominal GDP targeting has not yet been implemented anywhere. Accordingly, you have the luxury of comparing an untested policy whose defects (if any) have not yet been revealed in practice with well-tested policies whose defects are a matter of record. Advocates of inflation targeting were in the same position when it was first widely discussed. Then it was implemented, and after some years of apparent success came the Great Recession. If you are plan to advocate nominal GDP targeting in your book, you should specify what results (if any) would lead you to revise your favorable opinion of it.

Brandon writes:

I would hardly call you a crank. For one, you are an established economist, who taught economics at a normal university and publishes in normal economics journals. Hardly the same as some random guy on the street proposing his theories about physics. And also, your ideas aren't that extreme. There are influential economists who have positive references to your work and it's not that fundamentally different from other economic theories. The meta level isn't giving us any hint on whether you're right or wrong.

Jeff writes:

You're making two arguments. The first is that the Fed should target the forecast, i.e., that whatever target it chooses, it should choose a policy such that the forecasted effect of that policy is to hit the target. This would appear to be obvious, but apparently it isn't, as the Fed refuses to do it.

Your second argument is that the correct target is an NGDP level path. That may or may not be optimal, but it's pretty hard for anyone to show empirically that some other target has a history of working so well that the NGDP level target is clearly inferior.

But the first argument is more important, and I think the related observation that the Fed does not appear to be very interested in getting more and better measures of market expectations of the effects of its policies is telling. If there were a well-functioning market in NGDP futures, the Fed would have to explain why it pursues policies that the market doesn't think much of. And no self-respecting economist wants to argue that his or her forecast is better than the market's forecast.

E. Harding writes:

If Yudkowsky believes something, I see no obvious reason to think it more true. He often loses his bets. He's just a very smart man who can write well. Smart men believe all kinds of crazy things. I expect Yudkowsky to hold more correct beliefs than the average U.S. Senator, but that's not saying much. Caplan, on the other hand, is different. He wins all his bets.

Scott Sumner writes:

Todd Ramsey, I'd suggest you google one of my posts where I discuss the "guardrails" approach to NGDP futures targeting. That is my current favorite approach.

Todd Kreider, You dodged some of my key points. As I said, the RGDP numbers are misleading in a country where the working age population is falling at 1.4%/year.

If you read almost any expert on Japan, there's an almost universal agreement that the economy has picked up under Abe.

Take a look at the NGDP graph for Japan. Look at the public debt situation

As far as inflation, there's a big difference between 2006-08 when oil prices were soaring, and 2013-16 when they were plummeting. The underlying inflation fundamentals have gone from deflation to inflation.

Kurt, Here's one way of thinking about why NGDP targeting is not a part of my core message. If I favored the Fed's current dual mandate, my critique over the past decade would be almost identical. I would have proposed creating a futures market linked to the dual mandate (which would not be far from NGDP targeting) and I'd have made the same criticism for an overly contractionary policy during 2008-14.

Maybe I'll do a blog post on this.

As for what would refute the theory that NGDP targeting is best---suppose that the unemployment rate become more volatile than before---that would suggest there are problems with NGDP targeting. I would not worry about inflation volatility, as long as NGDP growth were stable.

Brandon, Fair enough, but would you agree that my theory that tight money caused the Great Recession is a bit . . . unusual?

Jeff, Good comment.

Todd Kreider writes:

Scott,

I did mention that it simply isn't correct to say as you did that 1.6% growth is equivalent to 3% growth in the U.S. The issue when comparing growth is GDP per capita. Also, in Japan there is a higher percentage working over 65 than in the US. At any rate, per capita grow is per capita growth and nothing misleading about comparing per capita RGDP growth of two countries.

My original point was Scott Alexander was wrong to write: "But after a few years, the Bank of Japan switched to Eliezer's preferred policies, the Japanese economy instantly improved" since 1) there is no evidence that Yudkowskly wrote anything on Japan prior to this year (he can comment here on when he did) and that 2) The economy only had one strong year after Abe was elected of 2.0% growth followed by 0.4%, 1.0% and 0.9% of sluggish growth. I added this year is much better trending 1.7% growth but years after Abe was elected and the year isn't over. There has not been a concensus among economists, Japanese or Western, on Abe's performance.

Keep in mind that like the UK and US, Japan's per capita growth had been 2% for five years prior to the Great Recession and like those countries has had much lower growth.

Oil is a very small part of Japan's economy. Oil prices have been stable for the past three years while Japan only momentarily had inflation in 2013 before returning to around 0% for the next three years.

Net debt is the important percentage for Japan, not total debt as you have written. It has been flat from the end of 2010 through 2017 so Abe did nothing to lower this. Please look at the graph linked below:

https://www.quandl.com/data/ODA/JPN_GGXWDN_NGDP-Japan-General-Government-Net-Debt-of-GDP

Lorenzo from Oz writes:

Market monetarism explains why Australia is still in the Great Moderation and has remained in it regardless of how its terms of trade have shifted.

Market monetarism is about not continuing the key mistake that Milton Friedman made in his famours AER Presidential Address when he (correctly) criticised relying on a statistical regularity because market agents would adjust but then proceeded to base his own recommendations on a statistical regularity which failed because market agents would adjust. It puts "market" all the way through monetarism.

In policy institutional terms, market monetarism is about blocking central banks from evading responsibility. To complete the job that monetarism started but did not complete. (See previous paragraph.)

Lorenzo from Oz writes:

Market monetarism explains why Australia is still in the Great Moderation and has remained in it regardless of how its terms of trade have shifted.

Market monetarism is about not continuing the key mistake that Milton Friedman made in his famous AER Presidential Address when he (correctly) criticised relying on a statistical regularity because market agents would adjust but then proceeded to base his own recommendations on a statistical regularity which failed because market agents would adjust. It puts "market" all the way through monetarism.

In policy institutional terms, market monetarism is about blocking central banks from evading responsibility. To complete the job that monetarism started but did not complete. (See previous paragraph.)

POST A COMMENT




Return to top