Well, it took embarrassingly long for me to see this but there’s really a very simple resolution to the quandary I posted Monday. The key point is this:
In a flexible price world, anybody can supply money at a social cost of zero. In a fixed price world, only the government can.
To be more precise: I currently hold $11. Suppose I agree to hold a twelfth. In a flexible-price world, I can get this dollar from the government (which prints it up at zero cost) or I can get it from my friend Jeeter, whereupon the price level adjusts and the rest of the world’s real balances (including Jeter’s) are restored to their original level. No social cost either way. In a fixed-price world, I can get this dollar from the government (which prints it up at zero cost) or I can get it from my friend Jeeter, whereupon prices don’t move and the rest of the world’s real balances are reduced. Zero social cost one way, positive social cost the other way.
This turns Monday’s question into a non-question. The question was, in essence: “If I demand money from a zero-cost supplier, there should be an efficiency gain. When the supplier is not the government, who gets the gain?” The answer is: If the supplier is not the government, he’s not a zero-cost supplier.
It also impacts on Friday’s problem: It seems to me that in a fixed-price world, you’d absolutely want the government to flood the economy with zero-cost money up to the point where people are satiated with real balances. If people hold money only because of a cash-in-advance constraint that’s no longer binding (i.e. in a liquidity trap) then there’s no point in printing any more.
I have not fully digested all the comments. My apologies to anyone who tried to explain this to me but failed to penetrate my thick head. Apologies also to others who said interesting things I haven’t yet digested. And many thanks to Jim Kahn who instantly saw the answer so clearly that he couldn’t figure out what the question was, and expressed it in language that forced me to see the light.
” I can get it from my friend Jeeter, whereupon the price level adjusts and the rest of the world’s real balances (including Jeter’s) are restored to their original level”
I don’t get it. Before his donation Jeter has $1. This gives him a claim to a non-zero fraction of the available goods. He gives the $1 to Steve. No matter how prices adjust Jeter’s share is zero. That is not his original level.
Ken B: “the rest of the world’s real balances (including Jeter’s)” means “the sum of everyone else’s real balances (including Jeter’s)”, not “each individual component of that sum”.
One of my favorite aspects of your blog is that you admit your mistakes. It means that if I think you’ve made a mistake, and you’re not admitting it, it’s probably me who’s not fully understood. This is really useful feedback for determining when to and not to dismiss a conflicting idea.
@Steve: Thanks for the clarification, but is it still true?
What if Jeter IS everyone else?
Ken B: Your example is odd, because now it seems that Jeeter is the proud holder of a 0/0 fraction of the money supply. Probably what’s going to happen at this point is that prices adjust down to zero (there is no commerce, in other words). That’s sad for the economy, Jeeter is still hurt and Landsburg’s conclusion still holds.
You know what, I’m actually confusing the two worlds. In a fixed price world, Jeeter is hurt because now he can’t buy anything. In a flexible price world, prices adjust down to zero and no commerce would take place, so I guess at that extreme…your question is interesting.
Ken B: Let M be the amount of money that everyone was holding to begin with, and P the original price level. When a dollar is taken out of circulation, M is replaced by M-1, and P adjusts so that the old value of M/P is restored. The only occasion on which P can’t do this is when M-1=0, which is the exception you’ve identified.
I thought there was a larger issue we were trying to get to the bottom of here, and that this fixed prices row was just a problem we ran into during our initial conceptualizations.
Remember: What kind of market failure is a liquidity trap?
Scott H.: My best current understanding is that the liquidity trap is not supposed to be a market failure. Instead, it is considered a bad thing because it renders ineffective certain policies that might otherwise alleviate certain *other* market failures.
I am about 85% sure I have that right.
@Steve: That calculation assumes I will not spend the $1 I get from Jeter. This is consistent with the idea of a “liquidity trap”. Do I have that right?
Ken B: Yes, it assumes you will not spend the $1, but this has nothing to do with liquidity traps. It only assumes that for some reason you’ve increased your demand for money, which is something that can happen in any model, classical, Keynesian, or otherwise.
Seems like a true liquidity trap requires expectations of deflation?
So, is anyone else left wondering what’s so special about the government, or am I the only one not getting this? Why can’t a counterfeiter print up a bunch of money and hand it out? Or is there an implied “only the government can /legally/ supply money at zero social cost”?
What are examples of market failures where the government would be the best choice to correct them? If we are in a time when there are no such failures, then it seems that a liquidity trap is fantastic occurrence whereby the government loses its ability to throw its weight around via spending. Could this be a possible upside, or am I missing something?
RF: Yes, there’s an implied “legally”.
@ScottF not sure what you mean by “throw its weight around by spending”. The government can spend as much or as little as it likes at any time. During a liquidity trap, however, spending doesn’t push up interest rates, so government decisions about spending have *more* effect on GDP than during “normal” times. In the sense of “have an effect on the economy”, the government can “throw its weight around” better during the liquidity trap than at other times.
This is why, for example, austerity measures are proving such a big disaster in the UK and Europe – government spending cuts fail to result in private investment through lower interest rates, but they do result in lower sales. Similarly, the GOP’s stand on debt, spending and deficits is particularly badly timed – sure, cut the deficit, great idea – but not during a liquidity trap when it will depress the economy further. It’s why Keynesians are advocating increased spending during the liquidity trap – there’d be no crowding out of private enterprise, but there’d be an immediate influx of demand, making the economy better. If the spending were enough, the economy could actually recover. The extra spending could be on actually useful things – common or almost-common goods, for example, such as infrastructure, education, etc. Useful stuff the private sector won’t pay for (to identify your market failure) because they are common goods.
@Steve :
I’m confused about this :
Suppose we enter a liquidity trap, in year 0. Interest rates hit 0%, the federal reserve can’t boost the economy by lowering interest rates. Expectations are for deflation, people set their prices, fail to invest, etc, etc, etc.
In year 1, things are worse – expectations remain low, people set their prices low, fail to invest, etc, etc, etc.
Now, imagine two hypothetical universes, one called “Europe” and the other called “Australia”.
Just before the end of year 1, in Australia, the government realises there’s a liquidity trap, and announces a massive program of government spending. This comes as a surprise, but a welcome one. In Year 2, businesses raise prices a little, but government spends.
In year 3, confidence is restored, private investment rolls out, interest rates pop up, and the government starts to quietly roll back their spending programme. It will pay back all the debt it incurred in year 2 over the next two years.
In year 4, because of the dip in govt spending, interest rates drop, but not to zero, and private investment keeps rolling along.
In year 5, the economy is still rolling along fine.
In Europe, however, the government announces an austerity programme. This comes as a surprise, an unwelcome one. Year 2 is like year 1.
Year 3 is like year 2 also, little private investment, interest rates still 0%, and so forth. Year 4 and 5 are similar.
Now, if there’s no market failure in either case, why, in the hypothetical universe called “Europe” are people so much worse off than in the other called “Australia”?? In year 5, everything is identical, except that in “Europe”, nobody trades or invests because nobody is trading or investing, and in “Australia”, everyone trades and invests because everyone is trading and investing. And yes, the Europans are worse off than the Australans, because there’s less trade in their universe – and trade benefits trader and tradee (perhaps not at the margin, but not everyone is at the margin, so there are benefits to trade.)
Mike H: Part of it may be that at this point the public’s reaction to fiscal stimulus is affected by its perception of the prevailing fiscal balance. Are Australia’s books in better shape than Europe’s? I don’t know. I don’t think austerity programs are viewed as an economic elixir but rather a necessary lesser of two evils, reflecting a view that at some point accumulating more debt for ineffective spending risks doing more harm than good (e.g. “at what cost smoothing”?). Even in the short run people are starting to ask “can we afford it”? Larry Lindsey has called this “commmon sense Ricardian Equivalence” to help explain why sometimes stimulus works and sometimes it doesn’t (as opposed to the hyper-rational all-or-nothing version of RE).
As I may have mentioned here once before, the debate over the Keynesian suggestion that the disappointing stimulus results to date just prove that we should’ve done (and should do) more remind me of the old gambling “strategy” called a martingale: keep doubling down and you’re guaranteed to win…unless you run out of money first. So even assuming we all agreed that in theory more and more stimulus would be the right thing to do, in practice at some point we may get tapped out.
@Mike H
Gains from trade are but one advantage to your “Australian” solution. If stimulus works (to boost aggregate demand and restore full employment), and ongoing economic activity results in increased standards of living (people produce more and newer/better things, and more efficiently, including new capital goods/process innovations), then it is certainly better to have the benefits of an increased standard of living *sooner* rather than *later.* It seems to me this is a good response to the apostles of austerity who say “the poison needs to work its way out of the system; get out of the way, let it crash and rebuild from there.” It seems to me that, in the conditional we are considering (stimulus works), the costs (a larger deficit, i.e. a larger number in a government spreadsheet) are easily exceeded by the gains (from trade, and from improved standards of living arising sooner rather than later).
My dream is that one day, Steve will write, “And thanks to Bob Murphy, who cleared up the confusion for all of us by framing the problem in such a way that, afterwards, the solution was obvious.”
Steven:
The “Doh” is bigger than you know. If you get a dollar from the government, the government will get a dollar’s worth of something (e.g., bonds) from you. (Same with Jeeter.) The government’s assets rise in step with its issuance of money, so the value of the dollar does not change. It’s just like when GM issues a new share of stock. GM gets new assets in exchange for that share, so GM’s assets rise in step with its issuance of shares and share price does not change.
There’s also the Law of Reflux to consider: As you add $1 to the circulation, someone else will probably retire $1 from circulation.
Mike Sproul: If I get an iPad from Apple, Apple gets something from me. Do you think it follows that if Apple doubles the output of its iPads, the price won’t have to fall?
Steve:
Naturally, if Apple produces more iPads, the price will fall. But if Apple issues $1 million of new shares, and it gets $1 million cash in return, then Apple’s assets rise in step with its issuance of stock, and the share price is unchanged. Similarly, if you deposit 100 oz. of silver in my bank, and if I give you 100 paper IOU’s in exchange, each redeemable for 1 oz. on demand, then each IOU is worth 1 oz. If you then deposit another 200 oz, and I issue another 200 IOU’s, then each IOU is still worth 1 oz., even though their quantity has tripled.
Mike Sproul:
What if there are three people who prefer IOUs to silver, and are therefore willing to give up 2 oz. of silver for 1 oz. worth of IOUs, whereas everyone else is indifferent between IOUs and silver?
Then if two 1-ounce IOUs are issued, they’ll trade (among the three heavy IOU demanders) for 2 ounces of silver each. If ten IOUs are issued, they’ll trade for 1 ounce of silver each.
More generally, if there’s a downward sloping demand for those IOUs, then the price will fall as more are issued. If the demand curve is flat, it won’t. Why should the demand for money be any flatter than the demand for iPads?
Steve:
You have to ask why the IOU’s were issued in the first place. Starting from a world where 100 IOU’s exist and trade for 1 oz. each, suppose someone wants 1 additional IOU and would pay 1.01 oz. for it if he had to. Assuming the printing/handling cost of IOU’s is zero, then someone (government, a bank, or an individual) will issue the new IOU and IOU’s will continue to trade for 1 oz. As long as the supply of IOU’s is flat, they will stay at 1 oz. even if the demand for them slopes down. (If the supply slopes up then price can rise, but since money can be computer blips, and since blips are not produced using scarce resources, I think it’s reasonable to say the supply curve is flat.)
You also have to consider the Law of Reflux. If that 1 extra IOU gets added to the circulation, but there was no additional demand for IOU’s, then somewhere an IOU will reflux to its issuer. Reflux works the same for goods as for money. If an extra silver spoon is manufactured and the world has enough spoons, then someone will find it profitable to melt 1 spoon. The spoon refluxes to bullion just as money refluxes to its issuer.
Mike Sproul:
As long as the supply of IOU’s is flat, they will stay at 1 oz. even if the demand for them slopes down. (If the supply slopes up then price can rise, but since money can be computer blips, and since blips are not produced using scarce resources, I think it’s reasonable to say the supply curve is flat.)
You are assuming a flat marginal cost curve. It does not follow that there is a flat supply curve unless you also assume competitive suppliers.
A qualification needs to be made to Mike Sproul’s claim that “If you get a dollar from the government, the government will get a dollar’s worth of something (e.g., bonds) from you..”. Govt in the sense of the “govt / central bank machine” may get absolutely nothing if a net increase in govt spending consists exclusively of automatic stabilisers, like unemployment benefits.
The (ludicrously convoluted) way this happens is thus. Govt borrows $X, gives $X of bonds to the private sector, and gives the cash to social security recipients. The central bank may then print money and buy back the bonds (QE), or do the same thing with a view to cutting interest rates. The net result is that the private sector has $X of cash it didn’t have before, and “govt / central bank machine” gets nothing in return.
Ralph:
It’s true that the government might get nothing in return, and if the government’s net worth is already negative, then this can cause the government’s money to lose value. But if the government’s net worth remains positive, then the money can hold its value. It’s covered by the ‘extra’ net worth.
Steve:
I am assuming competitive suppliers. But let me back up a little. If Apple issues new shares, and gets equal-valued assets in return, then you’d agree that the share price doesn’t change, right? That’s not a controversial position among accountants and financial economists. People can argue about whether the assets really moved in step with the number of shares, or whether the firm put the new assets to good use, but the simple proposition is that if a firm’s assets are worth $60 million, and that firm has issued 1 million shares, then each share is worth $60. And if the assets rise to $120 million, while the number of shares rises to 2 million, then the shares still are worth $60.
By extension, if a bank starts with 100 oz. backing 100 IOU’s, and the next day the bank has 200 oz. backing 200 IOU’s, then each IOU remains at 1 oz. If the IOU’s fell in value to .99 oz., then that’s an arbitrage opportunity.
Mike Sproul: I certainly agree with you re the issuance of equity (or for that matter debt)/. I’m glad you agree with me re the issuance of iPads.
The question, then, is whether money, for the purposes we’re discussing, is more like equity/debt or more like iPads.
If there is a transactions demand for money, then it seems to me that that’s enough to make it more like an iPad.
But I do also agree that the demand for money is irrelevant if the supply is flat, as would be the case under competition. For better or for worse, though, money is not competitively supplied in our economy.
iceman: “As I may have mentioned here once before, the debate over the Keynesian suggestion that the disappointing stimulus results to date just prove that we should’ve done (and should do) more remind me of the old gambling “strategy” called a martingale: keep doubling down and you’re guaranteed to win…unless you run out of money first. So even assuming we all agreed that in theory more and more stimulus would be the right thing to do, in practice at some point we may get tapped out.”
Doesn’t that underscore the importance of going off of the gold standard during the Great Depression? The graph that has appeared on Delong’s site a couple of times and elsewhere, as well, of six major economies shows a dramatic change for the better when they went off of the gold standard. A fiat currency means never getting tapped out.
Mike Sproul said:
The “Doh” is bigger than you know. If you get a dollar from the government, the government will get a dollar’s worth of something (e.g., bonds) from you. (Same with Jeeter.)
This is false in a fiat money, floating exchange rate regime, and the refutation doesn’t require bringing in QE1 or 2 as Ralph Musgrave did. I checked your web site and saw that you subscribe to a strange monetary doctrine which claims that fiat money doesn’t exist (and even proves it logically)!
On the contrary, since 1971 in the USA (and in very many other countries) fiat money very much does exist. It is backed by nothing. It is simply a credit against federal taxes, which the US government may levy in various ways (as we all know and are reminded every April 15th). As Warren Mosler (and others such as L. Randall Wray) has explained, the government spends by marking up entries in electronic spreadsheets. Government borrowing is the shifting of dollars from reserve accounts to securities accounts at the Fed. There are some $14 trillion in securities accounts at the Fed (the national debt). This represents the net financial assets the economy has left after the Fed added to our bank accounts when the Treasury spent, and subtracted from our accounts when the IRS taxed. This is exactly equal to the economy’s total net savings of dollars.
When Congress spends more than it taxes, dollar savings in the economy goes up by exactly the amount of the difference. To the penny. The government’s assets don’t rise with its issuance of money, on the contrary, the net financial assets of the private economy rise.
So, the US government has created some $14 trillion in money for which it has received nothing in return. It can, if it wants, tax all of that back, but of course the economy would be destroyed and we’d return to a primitive existence based on barter of real goods and services.
This is not original research on my part, but is instead simply a brief restatement of a part of modern monetary theory. Much more can be found at the site I linked.
Draco:
As you observed, I believe there is no such thing as fiat money. In a nutshell, governments issue paper money backed by metal and bonds and convertible into metal. Then one day they suspend metallic convertibility. The metal and bonds are still there in the vault, and metallic convertibility can be (and often is) restored at some future date, but people make the mistake of thinking that inconvertible=unbacked. That is about the whole reason that people believe that such a thing as fiat money exists.
Steve:
I’m glad we’ve at least reached a common starting point. (I’ve taught price theory for 31 years, so you and I are not likely to disagree on the fundamentals of price theory or finance.)
I suppose you’d also agree that IF money were competitively issued, then your original statements about how “the price level adjusts” would not be correct. If every dollar is backed by 1 oz of silver (or assets worth 1 oz.) then there is no price adjustment, and if you get your twelfth dollar from the government, then the government will get a dollar’s worth of assets from you and the dollar will hold its value.
But down to the brass tacks of monopoly/competitive issuance of money: Your ideas about monopoly issuance imply that paper money should sell at a premium relative to the assets held by its issuer. If a bank held various assets worth 100 oz of silver, as backing for 100 paper dollars, then the bank’s monopoly power would supposedly result in those dollars being worth something more than 1 oz. each. Say 1.20 oz. each. I don’t see any evidence of that kind of monopoly premium anywhere. Every bank that I’ve ever heard of looks more like the case where each dollar is worth 1 oz.
And what if there were a premium of .20 oz? Issuers of rival moneys, including domestic banks and foreign governments, would work very hard to issue competing moneys in order to get a piece of that premium.
@Steve Landsburg:
“It seems to me that in a fixed-price world, you’d absolutely want the government to flood the economy with zero-cost money up to the point where people are satiated with real balances.”
Does this mean we want the short term real interest rate to be zero when costs to price change are high?
“In a flexible price world, anybody can supply money at a social cost of zero. In a fixed price world, only the government can.”
In a fixed price world where interest rates are near zero, anyone can create money at zero cost. Does this mean that fixed prices are incompatible with zero interest rates?
Min: I guess a liquidity trap could be defined as a situation where for practical purposes our ability to print fiat currency (to any good effect) is “tapped out”. I was referring more to fiscal stimulus (Keynes’ fallback if we are in fact liquidity trapped), where we’ve been racking up natl debt at pretty good clip lately.
Mike Sproul: Not sure I’m following you on the no-fiat-money thing (based solely on your comment above). Say we suspend convertibility and issue a bunch more paper notes without increasing the amount of metal in the vault, then if/when we restore convertibility each note has been devalued, no? This seems a lot like just issuing a bunch of “unbacked” notes.
Iceman:
Correct. Start with 100 notes backed by 100 oz (or 100 oz worth of various stuff) and then issue another 100 notes in exchange for another 100 oz. worth of stuff. Each note remains worth 1 oz. But if the bank threw away that last 100 oz. of stuff, or never got it in the first place, then 200 notes are laying claim to 100 oz, so each note is worth 1/2 oz.
Mike Sproul:
I can’t argue with you on the history of fiat currencies; you are substantially correct to the best of my knowledge in the case of the US (but not of course of the Euro). When the Federal Reserve system was set up in 1913, the US dollar was convertible, whereas now it’s not, after various changes to law since that time. 1971 was a true turning point, which created a pure fiat currency in the US. But the conclusion you draw, that the US dollar really is backed by some commodity, does not follow, and is not true under current US law. As you state, convertibility could in theory be restored at some future date – but I’d be willing to make a pretty big bet – in fiat money – against that!
One advantage of a pure fiat currency is that it’s easy (as Steve says “zero cost”) to issue more money when necessary. How to get that money into people’s hands though? What MMT states is that this can be done through fiscal policy, due to the nature of our system. So, it’s relatively easy to apply counter-cyclical stimulus under our system. The tragedy is that the Right opposes this because they mistakenly believe that the government can go bankrupt or that inflation can be created via stimulus in an economy with 16% effective unemployment (both wrong) whereas the Left doesn’t understand the mechanism either, and so resort to the typical emotional arguments (the government should do more to help people) which convince only the most gullible of the populace. The fact is that the US deficit currently needs to be much, much bigger – in essence, we are currently being taxed too highly for the level of government we have. But both parties are convinced of the opposite (see the Deficit Reduction Super Committee). So instead of restoring aggregate demand and getting the advantages of full productivity now, we’re going to have to suffer for years to come, enjoying less ourselves, and leaving less to our descendants – in large part because almost no one understands the workings of our monetary system.
Draco:
The old European currencies generally started out backed and convertible, then switched to backed but inconvertible. The creation of the European Central Bank just consolidated the old banks, and the euro is just the old currencies printed on different paper, so if the old currencies were backed but inconvertible, then so is the euro.
Suppose that in 100 years, people no longer need paper dollars. They will return them (as they can now) for the fed’s bonds. Once the fed runs out of bonds, and people still have dollars to redeem, then if the fed still refused to pay out gold for dollars, it would be an effective default by the fed and the dollar would be unbacked and worthless. But until that day, it is enough for the fed just to redeem dollars for bonds, only promising to pay out gold once the dollar is obsolete and the fed is liquidated.
I’ll leave it to others to discuss things like deficits, stimulus, and aggregate demand.
Mike Sproul: Not familiar with this view — so what causes the fed to run out of bonds (what are they backed by)? To me the notion of “fiat money” means people are willing to accept unbacked notes simply because they’re confident other people will continue to accept them. Your argument seems to be that the real reason this works is because there is still an implied redemption into gold, someday…even though in the meantime the govt can dramatically dilute the value of that claim, which makes this seem like semantics to some degree.