In the comment sections to various posts, I see lots of ideas being conflated. Here I’d like to separate out some distinction policy questions:
1. Should the value of the medium of account be stabilized in terms of gold or NGDP? Or should its quantity be fixed?
2. Should the monetary system be determined by the government or the private sector?
3. Should banking be regulated, or should we allow free banking?
In my view, these are three completely unrelated questions. Any of the 12 possible combinations are at least theoretically feasible.
I make this point because people will say things that make no sense to me, such as that they oppose NGDP targeting because they favor free banking. Or that they want to abolish the Fed because they favor a gold standard. Huh?
Here are some feasible systems:
The monetary authority (public or private) might do nothing more than define the US dollar as 1/2000 ounces of gold. Or it might define the dollar as 1/20,000,000,000,000th of NGDP, as measured by an NGDP futures contract. Or it might fix the monetary base at a constant level. None of those options are any more libertarian than the other two. In each case, you either artificially fix the value of the dollar or you artificially fix the quantity of dollars.
Any of those three systems could in principle be done by either the government or the private sector. So that makes six possible combinations.
And for each of those six possible combinations, you could either regulate the banking system (our current system), or allow a completely unregulated free banking system, even allowing banks to issue banknotes to be used as currency.
FWIW, my own preference is to have NGDP targeting, done by the government, combined with a 100% unregulated free banking system. But even if you don’t agree with me, try to keep the issues separate. The role of the government in monetary policy is one thing. The price or quantity of the medium of account is another thing. And the status of commercial banking regulation is a third thing.
READER COMMENTS
vince
Aug 19 2022 at 4:51pm
OK, I will play along:
1. Real GDP;
2. Private sector but only if the losses are not subsidized;
3. Regulated, because the private sector will find a way for public subsidization of its losses.
Scott Sumner
Aug 19 2022 at 5:04pm
It’s not possible to target RGDP in the long run.
Spencer Bradley Hall
Aug 19 2022 at 5:19pm
R-gDp is pre-determined by the 10mo roc in required reserves.
vince
Aug 19 2022 at 6:08pm
RGDP is just NGDP without inflation. Who needs inflation?
Scott Sumner
Aug 19 2022 at 6:27pm
Inflation is the only part of NGDP that the Fed can control in the long run.
vince
Aug 19 2022 at 6:38pm
It would be great if they could control it to zero.
Spencer Bradley Hall
Aug 19 2022 at 7:15pm
Inflation is easier to regulate because the lag is much longer than the lag for real output. The proxy for inflation is exactly 24 months (courtesy of “The Optimum Quantity of Money” – Dr. Milton Friedman.
It’s unfortunate that the G.6 debt and deposit turnover release was discontinued. The late William Bretz, of Juncture Recognition, used to be able to tell when the series was wrong. Bretz would call Ed Fry, the G.6’s manager in D.C. whenever he thought the data was suspect.
See 1963:
Analysis of bank debits as a business cycle indicator (richmond.edu)
I got the 10mo roc in RRs from The Bank Credit Analyst, from their debit/loan ratio.
As Dr. Richard G. Anderson (Ph.D. Economics, MIT), Emeritus, FRB-STL (world’s leading guru on bank reserves), stated: Sent: Thu 11/16/06 9:55 AM:
“Today, with bank reserves largely driven by bank payments (debits), your views on bank debits and legal reserves sound right!”
Thomas Lee Hutcheson
Aug 19 2022 at 7:16pm
To Vince:
For zero inflation to be optimal we’d need extremely flexible prices with no more inflexible downward than upward. In the real world with some downwardly inflexible prices, some amount of inflation can permit greater adjustment in relative price flexibility than zero inflation. Presumably that optimal inflation estimate is what a PL targeting central bank targets or an NGDP targeting central bank uses together with it’s estimate of real GDP potential to set its NGDP target.
vince
Aug 21 2022 at 12:43pm
“For zero inflation to be optimal we’d need extremely flexible prices with no more inflexible downward than upward.”
Sounds great. Flexible prices are a good thing, aren’t they?
Which came first, downward inflexibility or Fed inflation? Thanks to the Fed, a dollar in 1913–when the Fed took over–is now worth less than a nickel. How does that fit with the stable price mandate?
Spencer Bradley Hall
Aug 19 2022 at 7:00pm
You need a little inflation because of limited downward price flexibility and monopolistic elements in the economy.
It axiomatic that the smaller the degree of price competition in a market and the greater the degree of private unregulated monopoly power over prices and output, then the higher the amount of unit prices, the greater the tendency for restricted output and employment and the smaller the degree of downward price flexibility.
Under these conditions, unless money expands at least at the rate prices are being pushed up, output cannot be sold and hence the workforce will be cut back.
vince
Aug 19 2022 at 7:03pm
“You need a little inflation because of limited downward price flexibility and monopolistic elements in the economy.”
What we need is less monopolistic elements and more downward price flexiblity. People expect inflation because we have inflation.
Scott Sumner
Aug 19 2022 at 11:00pm
“What we need is less monopolistic elements and more downward price flexiblity.”
We’ll never have enough. Thus we need a monetary system that can handle wage inflexibility.
Radford Neal
Aug 21 2022 at 6:03pm
There doesn’t seem to me to be anything “axiomatic” about any of this.
Perhaps there are workers who would be embarrassed to take a wage cut, seeing it as diminishing their self-worth. Or perhaps such workers are few. It’s an empirical question. If this is a big thing, inflation might help the labour market reach equilibrium.
Perhaps there are businesses for whom raising prices is costly (need to redo their marking materials, etc.). Or maybe this isn’t a big issue. It’s an empirical question. If it is a big issue, inflation will lead to the market not reaching equilibrium, with consequent shortages, since goods are underpriced.
You suggest that that inflation will reduce monopolistic pricing. Is the idea that the monopolists will just not be able to raise their prices as fast as inflation? I expect that in this situation, with a limit on how often they can raise prices, they would raise prices to more than the optimal (for them) point, then wait until they are a similar amount below the optimal point before raising them again, with the result being that the average price reflects the monopolistic situation just as before.
Scott Sumner
Aug 22 2022 at 12:19pm
“You suggest that that inflation will reduce monopolistic pricing.”
No, I do not suggest that. Find a quote where I say that.
“It’s an empirical question.”
Yes, it’s an empirical question that has been intensively studied and for which we have reliable answers.
vince
Aug 22 2022 at 12:32pm
“No, I do not suggest that. ”
Scott, the post apparently refers to a statement made by Spencer Bradley Hall.
On the empirical question, was downward inflexibility studied before the Fed began inflating?
Jim Glass
Aug 22 2022 at 8:12pm
Perhaps there are workers who would be embarrassed to take a wage cut, seeing it as diminishing their self-worth. Or perhaps such workers are few…
“Embarrassed”? “Few”? Put this question to a union: “You can all take a 10% pay cut or choose instead to lay off the 10% of your members with least seniority, all the rest of you keeping your full pay.” The union vote will go 90-10 … in favor of what? Why?
The same basic process occurs with non-union businesses. Give everyone a 10% pay cut, everyone’s morale will suffer, harming the entire firm — and the highest value employees will bolt to competitors offering the higher old (or better) wage. Which costs the company its best employees and leaves its worst, hurting morale further. To avoid all this, it instead lays off the least-productive 10% and keeps the 90% happy.
Wages don’t fall, unemployment rises. An empirical question answered long ago. Employees really, really don’t like taking wage cuts. And employers don’t like forcing them on employees.
On the empirical question, was downward inflexibility studied before the Fed began inflating?
Before it began inflating most recently? Like 100 years before then. Labor economists noticed the Great Depression!
vince
Aug 22 2022 at 8:44pm
“Before it began inflating most recently? Like 100 years before then. Labor economists noticed the Great Depression!”
The Fed started inflating when it opened its doors in 1913.
Eric Charles
Aug 19 2022 at 6:18pm
Your first post started with “I am interested in writing a critique of libertarian monetary ideas” but it wasn’t clear if you meant monetary ideas under minarchism or anarchism (my hunch is that there is a 40/60 divide among libertarians on this) . It seems plausible that one policy might work fine under statism but be a disaster under anarchism or vica versa.
“FWIW, my own preference is to have NGDP targeting, done by the government, combined with a 100% unregulated free banking system.”
I don’t see any philosophical objections (if it works and is stable, why not?) as a libertarian to this configuration if one believes the government should be involved. It’s just a matter of convincing statist gold advocates like Ron Paul that NGDP targeting is better.
Similarly, I don’t an anarchist like David Friedman objecting to this either if it was efficient absent government.
Roger Sparks
Aug 19 2022 at 8:16pm
Where you see three unrelated/distinct questions, I see only one possible system built in as many as twelve different ways.
I see a monetary system as fundamentally a political creation. A small government exercising little control will let the private sector exercise great control. But WTS, concentrated control within the private sector begins itself to represent government.
The choice of a commodity backed money over a general economy backed money (like we have now) depends upon the mechanism used to create new money. We presently allow banks to create money with little supervision. Either commodity-based money or NGDP based money would require considerably tighter control by a monetary authority than we presently see exercised.
Tight control by any monetary authority precludes fragmented control. Fragmented control allows chaotic regionalization of monetary systems. Thinking worldwide, fragmentation is what we have today, with exchange rates forging a bridge between regions.
Adam M
Aug 19 2022 at 8:37pm
It makes no sense to say you favor free banking, but only when it decides to use the kind of currency you like: issued by government, targets NGDP, etc. I don’t think you would favor free banking if the market ended up settling on a gold standard. (This is basically what happened in America, until the government insidiously took us off it — very anti-free banking). What you mean is something like 99% free banking but with legal tender laws to enforce your favored currency.
Scott Sumner
Aug 19 2022 at 11:06pm
“It makes no sense to say you favor free banking, but only when it decides to use the kind of currency you like: issued by government, targets NGDP, etc. ”
Not at all. I believe banks should be able to issue any sort of currency they like.
You don’t seem to know very much about American banking and monetary history. I’d suggest researching the topic before commenting here. The US never had free banking—it’s always been regulated. And the “market” didn’t settle on the gold standard.
BTW, Canada had much freer banking than the US.
Adam M
Aug 19 2022 at 11:28pm
“I believe banks should be able to issue any sort of currency they like.”
This makes your bullets 1 and 2 moot. You can still hypothetically analyze the merits of different monetary systems regarding 1 and 2, but this is a separate concern.
Scott Sumner
Aug 20 2022 at 1:59pm
I have no idea what you are talking about. Of course it doesn’t make #1 and #2 moot points.
robc
Aug 20 2022 at 12:29am
In reverse order, because they arent completely independent, imo, and this order determines others…free banking, private sector, determined by each bank as they see fit.
I dont see how #2 and #1 can be anything else with free banking as #3. Otherwise, it isnt free.
And how did you get only 12 choices? There are way more than 3 options for #1: fiat, ngdp, gold, silver, mixed metal, SPY, VOO, etc, etc.
Scott Sumner
Aug 20 2022 at 2:00pm
I have no idea what you are trying to say.
robc
Aug 20 2022 at 3:43pm
Assume free banking in #3. Then, each bank issuing currency would decide #1 independently. Some might have a BoAbuck where 1unit equals 1/2000th of a ounce of gold. Chase bucks, on the other hand may set 1unit equal to 1/20th an ounce of silver, or whatever.
The argument is similar for #2.
Scott Sumner
Aug 21 2022 at 2:15pm
Network effects make that very unlikely.
Stéphane Couvreur
Aug 20 2022 at 1:06am
Hello Scott,
In 1, do you really mean the medium of account or the medium of payment instead? It would be useful to clarify what it means to stabilize the value or the quantity of one or the other.
It seems to me that 2 is not independent from 1, as there is no such thing as “the” medium of payment (or account) under a private – i.e. competitive – emission regime, with no legal tender. Any issuer can only regulate “a” medium: his own.
Best
Scott Sumner
Aug 20 2022 at 2:06pm
Take the gold standard as an example. Under that regime, gold is the medium of account. There could be 100 different issuers of banknotes (backed by gold), but there is only one medium of account–gold.
Of course it’s theoretically possible that you could have a regime with multiple media of account, but I suspect that’s almost never going to occur in a well functioning economy. If it were to occur, then each money issuer could have its own monetary policy. One might stabilize the nominal value of gold while another stabilized NGDP. A third might peg its currency to the euro.
The multiple media of account case is one that’s not worth wasting time thinking about.
robc
Aug 21 2022 at 10:35am
It is the only one worth thinking about, imo.
In a free banking system, even if most everyone settles on a single medium of account, you will always have oddball niche currencies using a different medium.
Scott Sumner
Aug 21 2022 at 2:13pm
Sure, we have Bitcoin, etc. But macroeconomists like me are interested in the dominant money that almost all goods, services, and wages are denominated in. That’s the one that really matters. The niche monies are mere curiosities.
robc
Aug 21 2022 at 4:10pm
I thought we were discussing what libertarians were interested in (re: end the fed) not what macroeconomists are interested in. Shouldnt curiosities make you curious?
Scott Sumner
Aug 22 2022 at 12:16pm
Exactly, I’m interested in what happens if we end the Fed. And Bitcoin is not the answer.
Stéphane Couvreur
Aug 25 2022 at 2:02pm
I realize only now that I had never come across the term “medium of account” until now.
“Medium of payment”, yes: one possible definition is “An economic good which is commonly accepted [within a given community] as a means of payment in exchange for other goods.”
“Unit of account”, yes: one possible definition is “A given quantity of a given medium of payment, commonly used to express amounts to be paid in exchanges and contracts.”
But “medium of account”, I’m not sure what it means. I’ll have to mull it over.
Arqiduka
Aug 20 2022 at 4:59am
Off topic but not really,
I’d be interested in reading your thoughts on the efficient frontier of monetary systems, as a trade-off between being good and being simple.
I imagine you’d place NGDP targeting by way of prediction market at the top of both, but what about simpler systems? What would a less developed country think of?
Scott Sumner
Aug 20 2022 at 2:08pm
An LDC might be better off targeting nominal aggregate labor income. Especially if commodities are a big share of GDP.
Jim Glass
Aug 20 2022 at 11:20am
“The government isn’t the only entity allowed to issue money. Private citizens and businesses can too, and throughout U.S. history, they often have….The list of those who have issued private money in the United States is long…
“Studying various examples of private money that have arisen throughout U.S. history has taught economists much…”
Private Money in our Past, Present, and Future
Scott Sumner
Aug 20 2022 at 2:10pm
George Selgin and Larry White are good places to start when trying to understand these issues.
David Friedman
Aug 20 2022 at 3:08pm
I don’t know how you are defining free banking. I would have said that with free banking the monetary system is determined privately and how a dollar is defined is a market outcome, not something you get to choose.
Scott Sumner
Aug 21 2022 at 2:11pm
I see the banking system and the monetary system as being separate issues. Thus I could imagine free banking being combined with a gold standard. (I think many free banking proponents have discussed that option, so I’m not trying to redefine the term.)
But I can also imagine a system of free banking combined with fiat money. Again, I recall proponents of free banking discussing a system where the (fiat) monetary base is the medium of account.
In my view, the medium of account has already been “determined”. If we allowed the market to determine the medium of account, it’s likely that the US dollar would remain dominant. Network effects are very powerful. But I am also in favor of competition, so if an alternative displaces the dollar that’s fine.
But as long as the dollar is dominant, we need a “monetary policy”. One policy would be to legalize counterfeiting. Obviously that won’t happen. Another policy would be to allow competing private banknotes backed by US currency. That might happen in a free market.
artifex
Aug 21 2022 at 2:50am
If all laws, regulations and taxes were abolished, the Fed could keep running since it’s self-funded anyway. The Mint is funded by the Fed. The Fed could easily fund the BEP’s production of banknotes and arrange something to prevent counterfeiting, funding enforcement itself if necessary. It wouldn’t be that different from a private firm at that point, except one that would have been originally created by Congress and gotten into a position in which it can fund its activities only because of regulations favoring it over privately issued currencies.
Roger Sparks
Aug 21 2022 at 9:46am
Ownership-of-money, whether commodity money or contrived money, is the key into markets.
When governments or banks create keys and make new keys available to select elements of the economy, the utility of keys as a moderator of labor, resources, and skills is diminished or severely distorted.
After reading this post and comments, I leave thinking that the basic use of money has been obscured with details about system variations.
Carl
Aug 22 2022 at 11:50am
In your preferred system, what does theory predict will happen with bank issued currency? Do banks with currencies that are more stable than the legal tender expand until all non-government transactions are conducted in its currency? Or does Gresham’s Law dominate and legal tender crowds out the more stable bank-issued currency?
Scott Sumner
Aug 22 2022 at 12:15pm
The higher the nominal interest rate, the more profitable it would be to issue zero interest currency. Beyond that would be just speculation on my part.
John Hawkins
Aug 22 2022 at 9:52pm
Gresham’s law only occurs when parity is legally enforced between items the market views as non-fungible. So it depends on the legal arrangement between legal tender and bank-issued currency – in “free banking era” bank-issued notes didn’t necessarily all trade at par with each other, and multiple notes co-existed without crowding.
Carl
Aug 23 2022 at 10:51am
Thanks John. I hadn’t understood all the parameters of Gresham’s Law.
MarkLouis
Aug 24 2022 at 4:58pm
I think you left out a key question. Is your preferred target symmetric? Meaning do we make up for past over/under-shoots, or just some, or none? I personally think in the long-run that may be more consequential that what exact metric we target.
My vote is a symmetric NGDP target but with an inflation cap and deflation floor such that inflation cant move by more than say 5% in any given year.
Comments are closed.