Questions for Money/Macro Historians of Thought about "doing nothing"

First I need a long Preamble, to try to explain better what my questions are, and why they matter.

I was reading Simon Wren-Lewis' good post on the centrality of monetary (and fiscal) policy (as opposed to price flexibility) on how long recessions last. (As I have said before, "Is the macroeconomy self-correcting?" is a daft question, because the answer depends on the monetary policy regime. For a bad enough monetary policy regime, the answer must be "Of course not!".)

Any answer to the question "Is the macroeconomy self-correcting?" has an implicit assumption about what it means for the central bank to "do nothing".

Simon says:

"So why do many economists (including occasionally some macroeconomists) get this wrong? Why are textbooks often quite unclear on this point? It could be just an unfortunate accident. We are so used to teaching about fixed money supply rules (or in my case Taylor rules), that we can take those rules for granted. But there is also a more interesting answer. To some economists with a particular point of view, the idea that getting policy right might be essential to whether the economy self-corrects from shocks is troubling. They prefer to think of a market economy as being ‘naturally’ self-correcting, and to think that government intervention only has a role to play if there is some serious ‘market imperfection’. The market imperfection in the case of business cycles is price rigidity."

I want to focus on that "unfortunate accident" possibility. Because Keynes in the General Theory assumed a fixed money supply, when he asked the question "Is the macroeconomy self-correcting?"

The General Theory was a very influential book. And that same assumption of a fixed money supply is usually carried through into textbook treatments of the ISLM model (though there is absolutely no reason why the conventional Ms = Mbar assumption in the ISLM model cannot be replaced by any money supply function you like).

And the textbook ISLMBP model, when it compares fixed vs "flexible" exchange rates, just naturally assumes that "flexible exchange rates" means "fixed money supply". But there is absolutely nothing natural about that assumption. All "flexible exchange rates" really means is that the central bank does not hold the exchange rate fixed. OK, but that leaves it with 1,001 different things it could do instead. Why pick "fix M" as the natural alternative to "fix S"?

Only recently, and usually in upper year courses, does it seem natural to model monetary policy as holding fixed, not M, but the parameters in the Taylor-like Rule. But even that is very questionable. A much more natural way of modelling monetary policy today would be to hold fixed the central bank's internal forecast of expected future inflation. Because that is what modern central banks actually do (or say they do, anyway).

Sorry. My Preamble turned out to be even longer than I thought it would. Time for my questions:

1. Why did Keynes in the General Theory interpret "monetary policy" as fixing/controlling the money supply? Why didn't he interpret it as fixing the price of gold instead? Wouldn't a fixed price of gold have been a more natural way of thinking about monetary policy, at the time?

2. What were Keynes' contemporaries and predecessors doing? IIRC (and I probably don't, which is why I'm asking historians of thought) a lot of them were more Wicksellian, and thought of "monetary policy" as the central bank setting a (nominal) rate of interest. Like Hayek, for example.

3. How the hell did we end up thinking about macro in a way that would have made perfect sense if Milton Friedman had been born 50 years earlier and had persuaded the world's central banks to adopt the k% money growth rate rule (with k=0)?

4. To put it another way: how did we end up interpreting a central bank's "doing nothing" the way that we did? "Doing nothing" can mean anything whatsoever. It's a language game.

Comments

[Roger Sparks]

"1. Why did Keynes in the General Theory interpret "monetary policy" as fixing/controlling the money supply? ........."

Does the word "fixing", as here used, mean "repairing" as when combined with "fixing/controlling", or does "fixing" mean "making solidly stable"?

As I understand Keynes, he would have recommended a controlled money supply that flexed with the needs of the economy.

Here is a condition that might not have been anticipated by Keynes: A steady increase in the supply of money that is the same percentage increase in each measuring period.

A steady annual increase is fully anticipated, becoming as solidly stable as if the money supply was unchanging.

[Nick Rowe]

Roger: I do not mean "repairing".

What does the central bank "doing nothing" mean? For Keynes General Theory it means "holding the money stock constant". And "doing something" means "changing the money stock".

[Pat Deutscher]

Don't forget Hawtrey in all this. I think much of Keynes on money is a response to Hawtrey who viewed central bank control over short-term interest rates as the key tool for monetary policy. Hawtrey did not view economies as self-equilibrating without active monetary management. He wrote of the 'inherent instability of credit' and that 'the trade cycle is an inherently monetary phenomenon'. When low interest rates didn't do the trick he (Hawtrey) argued that monetary expansion would eventually reignite an economy.

[Roger Sparks]

I think that the dilemma is that "changing the money stock" in a consistently predictable fashion is the equivalent to "holding the money stock constant".

The beneficiaries would be different between the two money-stock managements.

[Nick Rowe]

Pat: I was indeed trying to remember Hawtrey, and failing. So thanks. And if Hawtrey viewed monetary policy as interest rate policy, he would be totally right that the economy is not self-equilibrating. Though if the central bank pegged both the nominal interest rate and the nominal price of gold (letting reserves adjust endogenously), then it *might* be self-equilibrating *within a limited range of shocks* (until it runs out of reserves).

[Majromax]

If I remember correctly, weren't central bankers of the era very concerned about net gold inflows to or outflows from a nation? That's more consistent with an implicit target of the quantity of money/circulating gold; a target of a fixed nominal price of gold would lead to unconcern about gold flows.

The confusing part may be that central banks did not adjust the nominal price of gold to ensure stability; they instead adjusted things like gold reserve ratios to ensure the markets did not clear.

[Nick Rowe]

Majro: IIRC, they held the price of gold fixed, and adjusted the nominal interest rate to prevent excessively large gold inflows or (especially) outflows.

[Lord]

By this time weren't central banks already seeing through the price of gold to the price level in the economy? They certainly had seen the price of gold change as they resumed the gold standard after the war, as France.

[Nick Rowe]

Lord: and there was a big debate in post WW1 UK about whether the Bank of England should return to the pre-war price of gold, and suffer deflation. But IIRC there is no mention of the price of gold as a policy variable in the General Theory.

[JKH]

"Why did Keynes in the General Theory interpret "monetary policy" as fixing/controlling the money supply? Why didn't he interpret it as fixing the price of gold instead? Wouldn't a fixed price of gold have been a more n6atural way of thinking about monetary policy, at the time?"

Because liquidity preference was defined in terms of money and not gold?

[Andy Harless]

Surely it would have been awkward to define "monetary policy" as fixing the price of gold when this was something that hardly ever changed. Changing the gold peg would have been considered a regime change, not just an easing or tightening of policy. And there were clearly things going on within a given peg that most people would have considered to be policy changes.

You could imagine defining monetary policy as fixing the quantity of gold reserves, but that also seems odd. I even have trouble thinking about how that would work: easing policy = selling gold while holding the peg, which would require cutting interest rates to make gold more attractive, so as to create the demand required to absorb the new gold, and the additional demand for credit from the central bank at the lower interest rate would be expected to increase the monetary base by more than the gold sale decreased it. More straightforward just to use interest rates or the money stock directly as the measure of policy.

Of course one could also imagine a "fixed but often changing exchange rate regime" where exchange rate changes would be the normal way of easing or tightening policy, but if any such regime existed as of Keynes' time, it surely wasn't common.

[Ray Lopez]

Now consider this post with the assumption that money is largely neutral (say 96.7%). What changes? That's the exercise for the smart reader.

[Frank Restly]

Just a shot in the dark here:

"Why did Keynes in the General Theory interpret monetary policy as fixing/controlling the money supply?"

By 1936, fractional reserve banking was pretty well the norm. And so you might argue that because central banks were granted the authority to set reserve requirements, this power enabled the central bank to regulate the expansion of credit. If we accept that a central bank (or any other bank) is never in a position to simultaneously redeem all outstanding deposits for gold at the "official" exchange rate, then it makes sense to say that a central bank "controlled" the money supply not by buying and selling gold, but rather by limiting / encouraging the expansion of credit.

[Nick Rowe]

JKH: "Because liquidity preference was defined in terms of money and not gold?"

Dunno. I would have thought it would be the same, under the gold standard. Gold and paper money are equally easy to store.

Andy: many of Keynes' contemporaries were talking about monetary policy as setting a rate of interest. Which gives you a very different answer on whether the economy is "self-correcting".

Ray: that's too easy. If different monetary policies don't have different real effects, it doesn't matter.

Compare the real exchange rates for two countries with a fixed vs a flexible exchange rate monetary policy. E.g. France & Germany vs UK & Germany, You see much more variance with the latter.

Frank: you don't need reserves requirements for monetary policy to affect borrowing and lending.

[Pat Deutscher]

By the time of the General Theory the gold standard was pretty much toast, certainly in the UK which suspended convertibility in 1931; the Americans in 1933 and the French in 1936. There was probably a consensus that getting back to normal involved getting back to gold. The restoration of gold after WW1 had been painful and was short-lived.

Hawtrey, as I recall, thought that the value of gold could be and ought to be, stabilized through cooperation between the leading central banks. In general he thought that monetary policy should try to stabilize what he called the wealth value of money, which sort of meant the price level though he sometimes indicated the average wage level was the better measure. Hawtrey was not enamoured of gold but saw it as having a role if it increased confidence in a currency.
As to the stabilizing role of gold, his diagnosis of the 19th century trade cycle was based on the mechanics of gold reserves and the steps taken by central banks (moving bank rate) to protect reserves and maintain convertibility. That was stabilizing, but in the course of adjustment economies could get seriously out of whack.

[Frank Restly]

Nick,

"you don't need reserves requirements for monetary policy to affect borrowing and lending"

But at the time of publication for the General Theory, reserve requirements set by central banks were in effect - yes?

Was Keynes describing the objective of the central bank (control the money supply) rather than individual actions of that central bank (buy gold, sell gold, raise reserve requirements, lower reserve requirements, etc.)?

[Scott Sumner]

Nick, Good post, here's my 2 cents worth:

1. Keynes did implicitly assume a fixed price of gold. And you could argue that the gold standard anchored the money supply. Yes, the money supply was not absolutely fixed under the gold standard, but it was certainly constrained. I see this as being implicit in the General Theory. I don't think Keynes was aware of this, as fiat money was then viewed as am almost pathological case, not worthy of serious consideration. Keynes didn't think about the fact that there was no fiat money for the same reason fish don't think about the fact that they are wet. Keynes strongly opposed fiat money, which at the time was identified with the German hyperinflation. He preferred a flexible gold standard. But he said even a rigid gold standard was better than fiat money.

Nonetheless, even if I am right, you are correct that a fixed price of gold would have been a more sensible assumption that a fixed money supply. On the other hand Andy Harless's comment suggests a good reason to prefer money over the gold price.

2. In my view the profession's confusion over what it means for a central bank to "do nothing" helps explain why so few prominent economists are willing to entertain the hypothesis that the Fed caused the Great Recession with a tight money policy. Note I say "entertain" not "accept". I'm not surprised that few accept the hypothesis, I am disappointed that view seem willing to entertain the hypothesis. None that I've talked to have ever offered me a good reason why it's not a plausible hypothesis. I think it relates to their sense that the Fed didn't "do anything" to cause the recession. But what does that even mean?

[Nick Rowe]

Scott: thanks.

I fully agree with your second point of course. I am still trying to wrap my head around your first point. I don't think you are alone in saying Keynes implicitly assumed the gold standard in the GT. I'm not saying you are wrong on that; just trying to figure out where that implicit assumption would or would not have made a difference.

[Jim]

What I have to contribute is more or less semantics, but what is doing "nothing" when it comes to interest rate policy?

Is the setting of an interest rate peg by the CB while the economy crashes and burns around you doing "nothing"?

In a model where there is no money and just a Taylor Rule it might seem that way. The CB falls asleep and doesn't update the rule. They do nothing.

However, in actuality they have to conduct operations to defend their peg exchanging bonds for money or reserves. That doesn't seem like nothing to me. That being said, I don't think the economy will be necessarily self healing if they don't defend their peg.

[Scott Sumner]

Suppose the gold standard constraint gave central banks the ability to raise or lower M by 10%, but no more. In that case it might make sense to view stable M as a sort of baseline, and a higher or lower M as an expansionary or contractionary policy.

But yes, I suppose you could make a similar argument for something else.

[Oliver]

Hmm, can one not imagine an economy with no central bank? Only commercial banks, each with their own banks notes and maybe a separate state currency. After all, if there isn't a central bank, there's no danger of it doing anything.

[Nick Rowe]

Jim: However, in actuality they have to conduct operations to defend their peg exchanging bonds for money or reserves. That doesn't seem like nothing to me."

Given the way central banks are currently structured, and getting right down to "machine language", that seems to me to be the most defensible interpretation of "doing nothing". But they don't have to be structured that way.

It's a bit like comparing Cournot-Nash and Bertrand-Nash equilibria in duopoly theory. In both, each firm maximises profit assuming the rival will "do nothing" in response. But in Cournot that means do nothing with Q, and in Bertrand that means do nothing with P. Do firms really really deep down choose Q or P? Same question with central banks. It doesn't seem to have an answer.

Scott: that's about as plausible an answer as I can come up with.

Oliver: yes, we can imagine a country with no central bank (or we can just look at history). And that seems like a reasonable way to repose the question. Though some might argue (not sure I would) it's a bit like asking whether the market system would be self-correcting if the government did not enforce property rights.

[Pat Deutscher]

Scott Bringing back the history of thought part of this, I'm a little startled at the categorical way you say Keynes supported a (flexible) gold standard and opposed fiat money. Keynes was a very fluid thinker and his position on many issues evolved, perhaps even oscillated with changing times. But an advocate of a gold standard as anything other than a small part of an incremental step in creating a stable international monetary system? I don't see it.

Here is Keynes in the Tract on Monetary Reform: "The advocates of gold, as against a more scientific standard, base their cause on the double contention, that in practice gold has provided and will provide a reasonably stable standard of value, and that in practice, since governing authorities lack wisdom as often as not, a managed currency wil, sooner or later, come to grief. Conservatism and skepticism join arms -as they often do. Perhaps superstition comes in too; for gold still enjoys the prestige of its smell and color." That was in 1924 with the debate raging over whether to return the pound to gold and if so at what rate. I don't see any other bits of this section on 'Restoration of a Gold Standard' (third section of chapter 4) In the proposals he put forward, he advocated a policy of the Bank of England controlling but not pegging the price of god. It would announce its buying and selling prices for gold each thursday, along with the Bank rate.
Comes the General Theory, in one of the few bits where he refers to gold he says "it is intesting to notice that the characteristic which has been traditionally supposed to render gold especially suitable for use as the standard of value, namely, its inelasticity of supply, turns out to be precisely the characteristic which is at the bottom of the trouble" (the trouble being excess demand for money) P235-236 of the Collected Works version

In between, in the Treatise publshed in 1930 when Britain was still on the gold standard, I don't see much to suggest JMK was a strong proponent of gold. His chapters in fact start with a Latin phrase translated as the accursed hunger for gold and goes on to cite Freud. Of course he was always keen to find practical solutiions and much of his writing is about how to improve the working of the gold standard since that is what the world (or the UK) had. Lots of wonderful writing throughout but I can't spot anything that looks like anything more than practical support for gold: e.g. "even if an international gold standard does serve to keep slovenly countries up to the mark, it may also keep progressive countries below the standard of monetary management which they might otherwise attain." (P 268 of the CW edition, volume 2)
Sorry for the length of this.

[Nick Rowe]

Pat: no apologies for length needed, given the quality.

"... he advocated a policy of the Bank of England controlling but not pegging the price of god. It would announce its buying and selling prices for gold each thursday, along with the Bank rate."

Sounds very similar to the Irving Fisher plan. The price of gold is an instrument, not a target. And in some ways it is a great pity that plan was not adopted, because the Zero Lower Bound would not be a problem; just raise the price of gold, like Roosevelt did. And if Keynes had written the General Theory assuming the Bank of England followed that sort of monetary policy, it would have been a very different book. The Liquidity Trap would just be the consequence of a bad monetary policy, where the central bank failed to raise the price of gold when it needed to. Which brings us to the main point of my post.

[jonathan]

Isn't a gold standard roughly equivalent to a fixed money supply, since you're effectively tying your money supply to gold supply?

I realize that the stock of gold and the stock of money are not identical under a gold standard (you're fixing a commodity price, not using gold as money). But in practice, any fluctuation in gold supply would cause a shift in gold prices that would necessitate a proportional shift in money supply to maintain the peg.

[Nick Rowe]

jonathan: if the total stock of gold is fixed, and if all gold is used as money, and only gold is used as money, then yes. Or if we add paper money, but banks have 100% gold reserves, then yes too. But the gold exchange standard wasn't like that. Banks had less than 100% gold reserves, and the gold reserve ratio fluctuated. And there was an "industrial demand" for non-monetary gold.

If the industrial demand fell, or if the reserve ratio fell, the money supply would expand.

[Oliver]

@ Jonathan

I just came across this definition which I find very concise:
...The gold standard operates by fixing the price of currency at a certain value in terms of gold (or stated equivalently, defining the currency unit as representing a fixed quantity of gold). The amount of currency under a gold standard is therefore whatever quantity of currency is demanded at the fixed price. That is very different from saying that a gold standard operates by placing a limit on the amount of currency that can be created...

http://uneasymoney.com/2015/12/01/eric-rauschway-on-the-gold-standard-2/