Scott Sumner  

I don't favor an interventionist monetary policy to fix recessions

PRINT
Krugman's Clever Misdirection ... A quick note on Luigi Einaudi...

Jose Romeu Robazzi recently left the following comment:

"IMHO the disagreement between keynesians and monetarists remain the same: both schools identify AD shortfall as a problem, and both schools recommend some sort of intervention (e.g. wealth transfers to sustain AD): monetarists suggest monetary policy (with its associated Cantillon and "hot potatoes" effects) and keynesians recommend outright government spending to make up for AD shortfall (maybe associated with some sort of wealth taxation and negative taxes for the poor)."
I believe this is a fairly common view, but I also think it's wrong. Speaking for myself, I do not support intervention to address demand shortfalls. Rather, I prefer a mechanical rule that would peg the price of NGDP futures contracts.

Now I suppose one could argue that even my proposal is "interventionist", and in a sense that's true, but it doesn't become more interventionist when there is a recession. Here's an analogy. Consider the old gold standard, where central banks bought and sold gold on demand, at a fixed price. Was that policy interventionist? In one sense yes, the central bank (or treasury) intervened in the international gold market. But would anyone serious claim that under the pre-1914 gold standard most governments "intervened to address AD shortfalls"? No, they simply pegged the price of a particular asset, and let the chips fall where they may. In contrast, I propose pegging the price of a particular asset, and letting the chips fall where they may. Oh wait, that sounds kind of similar, doesn't it?

Again, people are free to use terms like "intervention" in any way they wish, but in my opinion these terms do more to obfuscate than enlighten. For instance, in my view the Fed's mistake in late 2007 and early 2008 was not its failure to "intervene" to prevent a recession, but rather that the Fed did intervene and caused a sharp slowdown in AD, triggering a recession. How did it intervene? By bringing the growth in the monetary base that had been occurring for many years to a sudden stop. You may define what happened differently, and that's fine. But "intervention" is simply not a useful part of the discussion.

And I'd say the same about Jose's comments on Keynesians. Lots of Keynesians do favor wealth redistribution, but there isn't any necessary linkage. As I point out in this post, it's perfectly possible to be a conservative Keynesian and favor small government. You can simply use tax cuts as your preferred form of fiscal stimulus.

PS. George Selgin argues that it's more accurate to say that a true gold standard defines the currency unit, rather than pegs the price of gold. So my comparison is with a gold standard that involves a central bank or treasury that buys and sells gold.


Comments and Sharing






COMMENTS (19 to date)
Andrew_FL writes:
PS. George Selgin argues that it's more accurate to say that a true gold standard defines the currency unit, rather than pegs the price of gold. So my comparison is with a gold standard that involves a central bank or treasury that buys and sells gold.

He's right. The "price in dollars" of gold is no more fixed or pegged by a gold standard than the conversion factor between centimeters and inches is "fixed." Under a gold standard, a dollar is a unit of mass of gold. The "price" of gold is measured by its price in goods, that is, the inverse of "the price level."

Given that a central bank exists, there is no policy it can pursue which does not constitute some kind of intervention. The existence of a central bank is an intervention. Non intervention, in a case where gold is the base money of choice, would consist of the amount of monetary gold being determined by miners and private mints. I don't know what a non interventionist base money regime would look like today, however, and I'm not sure anyone does.

marcus nunes writes:

The Fed "gets what it wants". Therefore, it could "want" an NGDP Level Target. It could be done automatically, maybe through a NGDP futures contract. Let the market "guide" the Fed!
https://thefaintofheart.wordpress.com/2015/06/24/what-the-fed-wants-the-fed-gets/

Yaakov writes:

Thanks for another good post. I am slowly leaning more and more about monetary policy.

Jose Romeu Robazzi writes:

@Prof Sumner
First, thank you for honoring me by using one casual comment as basis for an entire post.

Yes, I did use the word "intervention" freely. Maybe I should have used "monetary authority action" and "fiscal action". I did no intend to use the term to make it sound negative.

The purpose of my comment was to be concise in defining what is the key difference that I see between the two schools of thought. Of course, in doing so, one simplifies things.

Last, I do believe that the Market Monetarist's model works best because my work experience leads me to believe that monetary policy tends to be broader in reach and can be reversed quickly, as may be needed, as opposed to fiscal policy, that history has shown hardly gets reversed, on average. In addition I also believe targeting NGDP is actually a very "neutral" policy in terms of relative prices, as opposed to fiscal spending, that inevitably will affect directly certain sectors chosen by the encumbent government.

RPLong writes:

Prof. Sumner, in a long-lost comment I posted to The Money Illusion quite some time ago (I searched for it, but came up empty handed), I asked you if you thought all monetary policy was either expansionary or contractionary, never neutral.

According to my recollection, you answered in the affirmative.

There is an obvious similarity between the claim that monetary policy is never neutral and the claim that monetary policy is always an intervention. Only this time, you're on the other side of the issue.

No worries if you don't remember my comment, but can you help me better understand your position here? Do you still believe that all monetary policy is non-neutral, or have you decided that NGDPLT-ing is a neutral monetary policy?

Or, have I just got things garbled up in my head again?

Njnnja writes:

Prof Sumner, do you believe that Cantillon effect is large, or even material? I know you have mentioned that the Fed could buy zinc or bananas to get the price level up, and as long as it wasn't so big that it distorted the market then anything could be used. That seems to imply that as long as the market is large enough, the Cantilion effect is small and no real wealth is transferred when the Fed gives somebody 1 million dollars in exchange for 1 million dollars worth of bonds or MBS.

Which of course makes sense and reminds me of the old "What weighs more, a pound of bricks or a pound of feathers?" question. But doesn't the extra demand for a security class by the Fed change the price at least a little? And if the market is so large that the extra demand doesn't change the price by much, isn't that just saying that the market is so large that even a miniscule price change could have a big effect on the aggregate value of the holdings of all investors in that asset class (e.g., even if the price of Treasuries is only 1 bps higher with the extra demand from QE, multiplied by the trillions of outstanding securities, that is still hundreds of millions or even billions of dollars worth of additional value for the holders of those assets)?

George Selgin writes:

Scott, I'm glad to see this. I've also been defending your position (and my own, by implication) against the criticism that it is a kind of monetary activism or "central planning." My recent Alt-M post on "Driverless Money" was written in that spirit.

Many free-market types don't seem to appreciate the fact that there is at present no one unique way in which to convert the dollar from a discretion-based currency to a rule-based one. A revived gold standard is conceivable; but in what sense would that solution be more "free market" than your rule? Because the dollar was once gold-based? True, but that was long ago, and bygones should be bygones: there are no longer any contracts outstanding that rest on the former gold dollar. A new gold fix would just be one of many possible new rules, all of which ought to be judged on their merits--for there's no other basis by which to judge them.

Nor can "the market" decide. Free choice in currency is a fine thing. But such choice by itself will neither easily lead to the dollar's replacement nor fix the dollar itself, which will remain a dominant currency, subject to the Fed's manipulation, unless active steps are taken to make it subject to some automatic regulation.

James writes:

I think this post is the first time I've ever seen someone claim that a policy option which calls for intervening in a market is not interventionist because the intervention would be in progress at all times.

Ninnja: Your question is a good one. Most central banks openly state an intention to affect yields which implies that the people running the central banks think they can cause changes in the prices of bonds. I'll defer to them on this one.

George Selgin writes:

James, a fiat money is of course a creature of government; and it is correct to say that both its existence and its ability to continue to command purchasing power necessarily depend on intervention in some sense. But there are degrees of intervention. Surely a rule-based arrangement (of which revived gold convertibility would be an instance; Scott's plan another) is less "interventionist" than a discretionary one. And I think it is not correct to describe a rule-based system as one in which intervention is "in progress at all times." The fact that the stock of money would continually adjust, in accordance with a fixed rule, isn't the same as continuous intervention. Again, consider the gold standard. In that arrangement also the money stock was subject to continuous adjustment. Yet that didn't imply continuous "intervention." The only intervention consisted in the decision to define a dollar as a fixed quantity of gold, and to make paper dollars convertible equivalents to metallic ones.

Scott Sumner writes:

Andrew, Yes, there are many ways to define intervention, and I can see how any central bank role involves a certain amount of intervention.

Thanks Yaakov.

Marcus, I agree.

Jose, I did understand your comment was the normal way people look at things in casual conversation, so no criticism was intended.

RPLong, I don't recall exactly what I said, but I consider expansionary to be above target NGDP growth and vice versa. If you peg the price of NGDP futures contracts at the target price, it's neutral in my view. Perhaps I meant that in the real world, policy is almost always at least a tiny bit expansionary or contractionary.

Or perhaps it was a post where I said there's no clear meaning to a central bank "doing nothing" because that might mean a stable monetary base, or a stable interest rate, or a stable price of gold. I guess I can see how it might be seen as contradictory---some of this involves philosophical distinctions that aren't always clear.

Njnnja, I've done posts on Cantillon effects, which is a very complicated topic. During normal times they are small, because the Fed doesn't buy very much of anything (relatively speaking.) At the zero bound they are somewhat more important, but I still think they are less important than the direct effect of more money. I think the macroeconomic impact of Cantillon effects is small. Also, many people forget that central bank bond purchases can cause bond prices to fall (due to inflation. ) So it's complicated.

George, Good comment, and I did enjoy your Driverless post. I should have linked to it, as it subconsciously probably triggered this one. I'm agnostic on totally laissez-faire money, but one problem I see for free market purists is network effects. For instance, if the Canadian government got 100% out of the money business, what would happen? Maybe gold-backed bank money, but maybe US dollar backed bank money. My hunch is the latter, despite all its flaws. Again, network effects, it's already very well established globally. I sense that you also feel the "free market" solution to money is not 100% clear.

James, I was making a more subtle point. I admitted that any regime involves a certain amount of intervention, but does it mean one must "intervene" to combat recessions? I say no.

ThomasH writes:

Actually, you do not need to be "interventionist" to be a Keynesian. If government just applies the rule of spending on activities that have NPV>0, then the lower discount rate and difference between market price and opportunity cost of activities into those activities will lead to an increase in "deficit spending" that would sure look "Keynesian" to most folks. The spending criteria should be the same in or out of a recession; only the data change. "Keynesianism" is just for the government NOT to alter its spending rule and start acting like a credit constrained household or a skittish business firm. Neither the zeroth, first, or second derivative of the deficit should be an argument in the tax and spend formula.

Levi Russell writes:

Scott,

I fail to see how a gov't solution to the problem of network effects is any better than one that would emerge from a freer market.

George,

Haven't you documented cases where private institutions provided the beneficial functions of a central bank? If Scott is correct that we don't know much about how we can transition to a market-based currency, isn't that a fruitful area of research?

This certainly isn't my area of expertise, but reading this blog and Alt-M it seems as though the bulk of the work done today is on second-best (gov't) solutions rather than looking back at history and determining how to make the transition to a market currency.

Not really a criticism, just an observation.

Andrew_FL writes:
if the Canadian government got 100% out of the money business, what would happen? Maybe gold-backed bank money, but maybe US dollar backed bank money.

Heck, Scott, its even possible Canadian banks could back their notes with the old Bank of Canada notes. Not like orphaned currencies haven't continued on without their central banks before. See the Somali shilling, for a great example.

Don Geddis writes:

@ThomasH: "NPV>0 ... would sure look "Keynesian" to most folks"

Not to folks who actually knew what the macro models were. Real Keynesian macro attempts to stabilize the growth path of NGDP (using adjustments in fiscal government spending to do so). That's simply not the same as maintaining NPV>0 projects. The fact that both rules cause government spending to be countercyclical, doesn't mean that they are the same rule, with the same effects. They're not.

RPLong writes:
Or perhaps it was a post where I said there's no clear meaning to a central bank "doing nothing" because that might mean a stable monetary base, or a stable interest rate, or a stable price of gold. I guess I can see how it might be seen as contradictory---some of this involves philosophical distinctions that aren't always clear.
Thanks. Indeed, I think this was the context of that old comment. This helps improve my understanding, so thank you.
George Selgin writes:

@ Levi Russell: "George,Haven't you documented cases where private institutions provided the beneficial functions of a central bank? If Scott is correct that we don't know much about how we can transition to a market-based currency, isn't that a fruitful area of research?"

Larry and I and others have shows than when you have competitive banknote issuers, you don't need a central bank. But that's when you have some non-central-bank-based reserve medium, like gold. Once you have a fiat reserve medium, you have to regulate the supply of that stuff somehow; it can't be regulated in the same way as gold is, because it isn't naturally scarce or reproducible only at marginal cost = marginal value.

ThomasH writes:

@ Don Geddis

Exactly what is the "real" Keynesian fiscal policy, then? How would one distinguish it from the NPV rule? Might the answer depend on what rule the monetary authority was following?

Scott Sumner writes:

Thomas, You are "reasoning from a price change". The fall in interest rates does not imply more projects are feasible. It all depends on why interest rates fell. Interest rates can fall because of less demand for new housing (tighter lending standards), which reduces the NPV of building new roads, sewers and schools. Indeed this seems to have happened.

Levi, I'm not sure it would be, but then I don't know what a free market solution would be. As I said, my hunch is that the "free market solution" in Canada would be to adopt the US dollar.

George Selgin writes:

I think it is very important for fans of the free market to come to grips with the fact that, with respect to some important economic problems today, there simply isn't any such thing as a clearly-defined "free market" solution. That's not the same as saying that there is market failure. It can also mean that only some deliberate government action can undo or limit the dame done by prior government action.

Comments for this entry have been closed
Return to top