DEBT-FREE MONEY AND BANANA REPUBLICS, PART TWO

By L. Randall Wray

My previous blog sparked a lot of discussion, especially over at Naked Capitalism. I do pity Yves Smith! There’s enough nonsense in the commentary to populate a large nation.

As I have argued, it is very hard to figure out what the debt-free money folks want as they are confused on the accounting, vague on the terminology, and rarely provide details on their proposal. However, a reader has directed me to a fine published article that has mostly got the accounting right, lays out a detailed proposal, and contrasts the proposal against alternatives.

I’ll get to that in a minute. First let me quickly respond to comments on the first piece. I’ll limit my response to two complaints that have been made about Part One of this series.

1. Responses to comments on Part 1

The biggest complaint was that I did not take advantage of a teachable moment that the radio program producer had offered for me to explain MMT to the hosts and audience. Instead I just made fun of debt-free money supporters and insulted the producer.

The critics fail to notice that the producer wrote to me to come on the show to talk about debt-free money. There was no invitation to discuss MMT. Producers can and usually do perform a “background” before inviting a guest. I suppose the producer found that I had written pieces AGAINST debt-free money but still wanted me to discuss the topic. I honestly told him I do not understand what the advocates are proposing and hence would not be a good guest. I introduced the banana money as my best (humorous) guess at what they want–which was in the last piece I had written criticizing the debt-free money proposal. The banana is a stand-in for all forms of debt-free money, which must take a “real” form rather than a “financial” form—for reasons I discussed and will expand upon later. Rather than being offended or deterred the producer ran with the idea and created an entire alternative history of the USA based on banana money. I continued to try to get out of going on his show to discuss a topic I do not find appealing, but eventually agreed to come on to say that they can achieve everything they want through ZIRP, which I indicated would take a minute. At that point he invited me to talk about MMT. I was not offensive and he took no offense. Of that I am sure. The exchange was all in good humor. We had a number of cordial exchanges after that. And I assure you every word I posted was in the exchange; I only deleted identifiers.

I see nothing unfair about the banana analogy. Many of the debt-free proponents refer to money backed by “real wealth”, goods and services, precious metals. They fantasize about the good old medieval days, when gold was money and men hacked up dragons as they rescued damstrels in distress—as depicted, I think, on-screen in Game of Thrones (if I’ve mischaracterized the program it is unintentional as I stopped watching TV when they cancelled the double-header line-up of Melrose Place and Ally McBeal). Me? If I were to go back to a utopian past, it would be the primitive communism of tribal society, as depicted in The Gods Must Be Crazy, before the Coke Bottle Money was dropped from Friedmanian helicopters, destroying an idyllic way of life.

The second most popular complaint was about my use of the word “debt”. But the commentators apparently did not notice that the topic of “debt-free money” was introduced by the producer. He used the term, just as all other debt-free money types do, apparently seeing our current money as debt money.  I’m agnostic. My point is that we use double entry book-keeping, and if “money” (however defined) is someone’s financial asset then it is another’s liability. Call it a “credit” (from the point of the view of one holding it), or a “debit” from the other’s point of view; or a debt; or a liability. What debt-free monetary cranks insist is that the money they want the government to create will show up only on the holder’s balance sheet as an asset, with no liability on anyone’s balance sheet. That is what I object to. Some argue that the Treasury, itself, treats coins as “equity”, not “debt”. Fine. Equity is on the liability of the balance sheet. Twist and mangle the language all you want. But at least do the balance sheets correctly. More on that below.

Calgacus had an excellent response on a blog site explaining the use of the term debt. I hope she/he will not mind if I provide a long quote. This is extremely useful not only for the clear explication of the term, but also for links to early expositions of the views now taken by MMT. In particular, Calgacus responds to comments about my use of the cloakroom token (taken directly from G.F. Knapp) as an example of a debt token—a commentator argued that this is not a debt because the cloakroom doesn’t own the coat. And to the claim that coins issued by government are not debts because the taxpayer is the one with debts, not the government that issues the coin. And to the claim that bank deposits are not debts of the bank, because it is the borrower who owes the bank. Here’s Calgacus’s argument:

“Debt” is a word in English – and in every human language. Even nonhuman social animals have some grasp of it. Wray uses the word in the standard very general dictionary meaning of a social, moral obligation. Here is the full definition from the #1 on google online dictionary:

1. something that is owed or that one is bound to pay to or perform for another:
2. a liability or obligation to pay or render something
3. the condition of being under such an obligation:
4. Theology. an offense requiring reparation; a sin; a trespass.

Basically, 4 ways of saying the same thing.

” A cloakroom is not issuing a debt-token.”

It most certainly is. To say it is not is to insist on an alternative meaning of “debt” and to avoid the standard general dictionary meaning, which is Wray’s usage. Alternative meanings involving money & interest are obviously not applicable. Money is credit/debt and obviously this view would be useless & unintelligible gibberish if the latter were defined in terms of the former.

“Nobody will accept this token for payment.” The cloakroom attendant will. Therefore it is a debt, a social, moral obligation, relationship between two moral agents. That is the point.

” the macro-economic substance of the act of issuing the coin is very different from what banks do.”

No, it is precisely the same thing, no more different than the US issuing dollars & the UK issuing pounds. Minsky’s “anybody can create money ….”

“By issuing the coin, the government allows a provider of goods or services to bring forward the settlement of their pre-existing tax debt to the government.”

This is not at all what happens. It could not happen that way, the way the rest of the story proceeds. Issuing of a debt in one direction must precede the settlement, the cancellation of the debt, which can only occur by a debt going the other way.

Here the coin recipient pays the coin to settle his subsequent, not pre-existing taxation. I can’t really understand what is being said here in a coherent way. If the coin is considered a receipt, it is a receipt for taxation-in-kind, taxation in real terms, like a government employee being “taxed” of his labor and given government currency in return. Taxation in kind or taxation in real terms is another word for government spending, which is the opposite of financial taxation – which is what “taxation” means nowadays. In any case, in any system, the coinholder of course relinquishes it, rather than merely keeping & showing it – that’s more like how titles of nobility operated, not coins!

“There is no pre-existing debt of the customer taking the loan. By giving the loan, the bank creates new debt (for which interest is to be paid, whether or not it is put to productive uses).” More errors, at least on what seems to me to be the plain meaning.

As above, there was no pre-existing debt in the government / tax case and the bank doesn’t create the new debt of the customer to the bank, the customer does. There are two credit-debt pairs being created in bank loans, but only one in government spending. That’s a difference between monetary and fiscal.

Basically, this is just Mitchell-Innes & his great predecessors. But only the MMTers – or circuitist / creditary economists like Ingham, Gardiner etc who contributed to the book on Alfred Mitchell Innes great papers seem to get things right. It is all so simple, so obvious, so natural, so easy, so entirely trivial…. That everyone makes a complete mess of it, by scorning the “trivial” chore of getting the trivialities right!

Thanks, Calgacus. Now let’s turn to a concrete proposal.

The Debt-free Stimulus Proposal

It is very difficult to get a handle on the debt-free money proposal because it is hard to find one with any details. However, here is an excellent analysis: “Stimulus Without Debt” by Laurence Seidman, Challenge, vol. 56, no. 6, November/December 2013, pp. 38–59. Now, I cannot be certain whether this is what most debt-free money proponents have in mind. However, Seidman lays out a concrete proposal and contrasts it with alternative methods of stimulus. He even compares his proposal with that of Adair Turner, Chairman of Britain’s Financial Services Authority, who has received a lot of attention for his “overt money financing” of government spending. I know Turner and have great respect for him as a serious critic of our runaway financial system. Many of the debt-free money proponents who have written to me have recommend Turner’s work. Hence, Seidman’s contrast of Turner’s proposal with his own is useful. Finally, as I said earlier, Seidman seems to have a good understanding of accounting.

I am going to use long quotes from Seidman’s piece as I would guess that most readers of this blog have not read the original. I’ll provide commentary along the way. For the most part, I find his analysis impeccable.

First, let’s look at Professor Seidman’s stimulus proposal. He is rightly concerned that fear of budget deficits and government debt hamper the ability to mount sufficient fiscal stimulus to counter a deep downturn, such as the one that followed the Global Financial Crisis. So what can we do next time to finance a stimulus without running up debt? In his example, he presumes the stimulus will take the form of a generous tax rebate for households, much like the one that President Bush pushed through earlier. I won’t go through the evidence he presents that this is an effective way to get income into the hands of consumers who will spend it, thereby stimulating demand. We’ll only concern ourselves with the question, “how can the government finance this without debt”. So here’s the proposal:

the stimulus-without-debt plan proposed here—a particular kind of monetary stimulus—is “a dual-mandate transfer” from the Federal Reserve (the Fed) to the U.S. Treasury. In a severe recession the Federal Reserve Open Market Committee (FOMC) would give a transfer to the Treasury in an amount decided by the FOMC that, in its judgment, would promote the Federal Reserve’s dual legislative mandate—enacted years ago by Congress—of promoting both high       employment and low inflation. It must be emphasized that the Federal Reserve would not be buying bonds from the Treasury; the Treasury would not be incurring debt—it would be receiving a transfer.

How does this differ from normal procedure? Seidman explains:

Standard fiscal-monetary stimulus works this way. To raise aggregate demand for goods and services through fiscal stimulus, Congress cuts taxes or raises government spending (transfers or purchases), and the Treasury borrows to finance the resulting deficit by selling U.S. Treasury bonds to the public, thereby increasing government (Treasury) debt held by the public. The Fed   then buys an equal amount of Treasury bonds from the public in the “open market,” so that the Fed, not the public, ends up holding the increase in Treasury debt. A crucial point is that the Fed’s action does not reverse the increase in Treasury debt: official Treasury debt increases by an amount equal to the deficit that accompanies the fiscal stimulus, whether or not the Fed buys Treasury bonds from the public. Standard fiscal-monetary stimulus entails “monetizing the debt,” not preventing debt.

The Fed is providing a “transfer” to the Treasury, not a “loan”. How does this affect the Fed’s balance sheet?

If the Fed buys a Treasury bond in the open market, it obtains an asset, but if the Fed gives the Treasury a transfer, it obtains no asset. According to conventional accounting, the Fed’s “net worth” or “capital”—defined as assets minus liabilities—would therefore be lower if the Fed gives the Treasury a transfer instead of buying Treasury bonds.

The Obama fiscal stimulus during the GFC amounted to about $400 billion a year for two years. Federal Government debt was increased by approximately the same amount, $800 billion. If the stimulus had been done through a Fed transfer, the Treasury’s debt would not have been increased. Instead, the Fed’s capital would have been reduced by $800 billion—equal to its transfer. On the Fed’s balance sheet, its liability to the Treasury (deposits) would rise by $400B each year, and its equity would fall by $400B each year. As the Treasury’s checks were deposited in household bank accounts, the Fed would debit the Treasury’s deposits and credit bank reserves by the same amount. As households drew down their deposits buying consumer goods, the deposits would shift from bank-to-bank and the Fed would shift reserves from bank-to-bank. (The reserves would remain at the higher level until either cash is withdrawn from the banks, or banks repay loans to the Fed. Note that the nonbank public decides how much cash to hold, which determines the ratio of reserves/Fed reserve notes.)

Why does a Fed transfer to the Treasury reduce the Fed’s net worth? Note that under normal operations, the Fed either lends (reserves to banks) or buys assets (government bonds from Treasury or from banks, or, recently, purchases of MBSs). Its assets go up by the same amount as its liabilities. If the assets earn more than the Fed pays out on its liabilities (reserves; note that Federal Reserve Notes are also Fed liabilities but don’t pay interest), then its net worth rises. The Fed distributes its profits to the Treasury and to its shareholding banks. Transfers, by contrast, increase reserve liabilities without increasing assets; the difference has to be made up by reducing equity. This reduces equity as well as profits since its earnings on assets have not changed but it pays more interest on reserves (unless for some reason the demand for Federal Reserve notes rises by an amount equal to the transfer—which is unlikely).  Lower profits mean the Fed distributes less profits to the Treasury, reducing Treasury’s revenue.

(If the total stimulus amounted to $800B and the interest rate on reserves were 1% then the Fed would have $8 billion less profits to turn over to the Treasury, all else equal. To avoid adding more Treasury debt, the Fed would have to transfer more. This is not a major consideration, but should be recognized: reducing Fed profits reduces Treasury “revenue”.)

If the Fed “transferred” more than its total net worth in its stimulus program, it would have negative equity.

Should we care about the Fed’s balance sheet? On one hand, any bank can operate with negative equity—many have done so and probably some of the biggest ones currently are right now, on rigorous assessment of the values of their assets and liabilities. Banking supervisors often adopt the “extend and pretend” approach—extending the life of insolvent banks while pretending they have positive net worth. We can certainly do that with our central bank, and the justification is probably far stronger. With insolvent banks, the biggest danger is that the incentives are aligned to “bet the bank”—take the riskiest bets imaginable, gambling that some might pay off while the downside is that the already insolvent bank fails. Shareholders have already lost, so who cares. But if the Fed is driven into insolvency while bailing-out the economy in a downturn, that can easily be justified as reasonable public policy.

As such, Professor Seidman recommends changing the way we do accounting:

For a household, firm, or governmental unit, it is important to worry about whether its “liabilities” (what it owes others) listed on its conventional accounting balance sheet are greater than its “assets” (what it owns or is owed by others). But there are at least two problems with using a conventional accounting balance sheet to evaluate the Federal Reserve in the same way it is used to evaluate a firm, household, or other governmental unit. First, Congress has given the Fed the power to create money by writing checks and standing ready to print and provide cash (Federal Reserve notes), a power not available to a firm, household, or other government  unit. Second, one of the large liabilities listed on the Fed’s conventional balance sheet—Federal Reserve notes—differs in an important way from the liabilities listed on the balance sheets of firms, households, and other governmental units…the power to create money surely gives the Fed an important tool for meeting its financial obligations not available to firms, households, and government units. A conventional accounting balance sheet alone is therefore inadequate to evaluate the financial position of the Fed.

Second, on the Fed’s conventional accounting balance sheet, the quantity of Federal Reserve notes outstanding is listed as a liability, and it is usually the largest liability on the Fed’s balance sheet. This made sense historically when the Fed promised to pay gold to holders of Federal Reserve notes if the holders requested gold. But this rationale no longer holds because the Fed no longer promises to pay holders of Federal Reserve notes gold or anything else. Thus, it is no longer obvious that Federal Reserve notes are a genuine liability of the Fed—or even if they are still a liability, whether they are as burdensome as other liabilities.

Despite these two problems with applying a conventional accounting balance sheet to the Fed, there will no doubt be concern about any plan that reduces the conventionally measured net worth or capital of the Fed. Advocates of the stimulus-without-debt plan should emphasize these two problems, object to the use of the conventional Fed balance sheet to pass judgment on the stimulus-without-debt plan, and call for new and better ways to evaluate the financial position of the Fed.

OK, accounting is a human invention, although it follows a logic. Congress can, if it chooses, throw logic to the wind and create special accounting for the Fed. It certainly wouldn’t be the first time a government has applied special accounting to itself—it is common in so-called Banana Republics (and maybe appropriate for banana monies!).

But would the Fed’s debt-free stimulus be legal? Seidman discusses the separation of powers that our founders thought important, with the separation further delineated by the creation of the Fed itself in 1913. Seidman downplays the power to create money given by the Constitution to Congress, focusing instead on the apparent intention of Congress to bestow that right on the Fed—something he believes Congress did in order to constrain itself from simply printing up money and causing inflation:

It was therefore a wise and crucial step for Congress, a century ago, to establish an independent central bank that would control the creation of money. Congress thereby gave up the power to cover its deficit by creating money. This has provided an important check against Congress’s setting government spending well above taxes in a normal economy when no stimulus is warranted, creating money to cover the difference, and thereby unilaterally injecting a combined fiscal-monetary stimulus that overheats the economy and generates inflation.

I would guess that this is the view of most economists and I’ll leave it up to our scholars of US legal history to comment (I find it to be a dubious interpretation). I’m also going to leave to the side the typical belief of economists that Congress is naturally hell-bent on ramping up inflation (again, I’m skeptical); as well as thee typical claim that the Fed is independent (nay, it is a creature of Congress and no more independent of government than are other agencies). What is important is Seidman’s recognition that the Fed’s “right” to create money might not give it the “right” to distribute tax rebates. If that is so, he believes Congress has made a lamentable mistake:

It was, however, unwise for Congress to apparently (if this is the judgment of legal scholars) prohibit the independent central bank from unilaterally deciding to give a dual-mandate transfer to the Treasury. The danger in prohibiting a dual-mandate transfer is that it prevents stimulus-without-debt in a recession or a weak recovery. If legal scholars judge that the current Federal Reserve Act in fact contains such a prohibition, then Congress should amend the Act to specifically authorize a dual-mandate transfer—a transfer that the FOMC judges would implement its dual mandate of high employment and low inflation.

Note that Seidman argues his proposal does respect the separation of powers intended by Congress, for he would have the Fed decide how big the tax rebate would be (hence, decide how much money to create, and when to do it), rather than letting Congress dictate how much, and when, the Fed would stimulate. This would be entirely within the Fed’s “dual mandate” to pursue high employment and price stability; it would ramp up the stimulus when unemployment is high, and cut it off when inflation rises. If this is illegal, he recommends changing the law (and presumably, the Constitution, if need be).

(This would expand the powers of the wise men and women who sit on the FOMC—from interest-rate setting to controlling fiscal stimulus. Well, why not–they’ve done such a “Heck-uv-a-job-Brownie” job so far, missing ten out of the last ten recessions and contributing to ten out of the last ten financial crises. The Fed always “fails upward”, gaining power and prestige when it screws up, so that its next screw up will be even more damaging. But I digress…)

Assessment of the Proposal

Seidman has provided us with a coherent proposal for debt-free stimulus. While he uses an example of a tax rebate, there is no reason why the finance method could not be used for a spending stimulus, such as Bernie Sanders’s infrastructure proposal. Instead of Treasury financing using tax revenues or bond sales, the Fed would provide transfers, reducing its net worth. Treasury can treat these as gifts, meaning it will not issue any debt. (Thanks, Aunt Janet!)

In that sense, the proposal is, indeed, “debt-free”. Of course, it is not “debt-free” in a more general sense, because the Fed’s liabilities grow—first in the form of Treasury deposits and then as the Treasury draws those down, in the form of bank reserves. Further, some advocates of “debt-free spending” seem to mean spending financed in a manner that does not commit government to pay interest. However, Seidman’s proposal fails to meet that definition, too, since the Fed pays interest on reserves.

So it is neither debt-free nor interest-free.

As discussed in Part 1 of this series, many argue for use of “debt-free money” to finance government spending. The “money” created by the Fed in Seidman’s proposal also fails that definition since the Fed’s reserve money is the Fed’s liability. Unless we want to invent a quite narrow definition of “debt” to mean something different from “liabilities”, the Fed’s reserves are a “debt money”. We could call them  “liability money” and then explain that by “liability” we mean “it is not a debt”. (However, as George Lakoff warns us, if you tell someone NOT to think of an elephant that is the first thing they think of. Our debt-free money folks might consider that as they reframe their meme. Yet another reason to run with the banana money meme?)

Changing the terminology from “debt money” to “liability money” is of course possible. By the same token we could instead invent a definition of “debt” that excludes Treasury liabilities, too. Treasury liabilities such as bills and bonds are much like the Fed’s liabilities: both are presumably backed by the full faith and credit of the Congress and both pay interest. We could invent a new term to cover all such liabilities, replacing the usual term, which is debt. I’m open to suggestions from our wordsmiths. (How about “bananas”? That has the unfortunate disadvantage of bringing to mind bananas, but it does have the advantage that it directs attention away from “debt”. Saying that the government “is trillions of dollars in bananas” sounds so much better than saying it “has trillions of dollars of liabilities”—which sounds an awful lot like debt. Or we could just adopt the convention that if we use words like debt or liabilities, what we mean is bananas. What the bank means when it says I have an onerous mortgage debt is that I have a really big mortgage banana. I feel better already.)

To get closer to the goal of “debt-free money” proposals, Seidman could recommend that the Fed stop paying interest on reserves. In that case, while the Fed’s liabilities would rise with its stimulus, it would not pay interest. Banks would simply hold more reserves but would not receive interest on those reserves. Some of the debt-free money enthusiasts insist that government spending should not generate interest payments—especially to banks. That is easy enough to do in Seidman’s proposal—just return to the pre-GFC practice of the Fed by eliminating interest on reserves. (Some even think this will encourage banks to “lend out” their reserves to business, adding additional stimulus. That is confused, but I won’t go into it here.)

At this point we run into a fundamental problem: if the Fed doesn’t pay interest on reserves and the Fed’s stimulus creates excess reserves, then it will drive the fed funds rate toward zero. Indeed, this is precisely how central banks operate ZIRP (zero interest rate policy)—leaving excess reserves in the system is how you do a ZIRP.

How does a central bank keep the overnight interest rate at a target above zero (non-ZIRP)? It either pays interest on reserves (paying a rate approximately equal to the target) or it offers Treasury bonds in open market sales or REPOs. In normal times (before the GFC and QE), the central bank holds a limited supply of treasury debt to sell. This means it could run out of treasury debt before it could eliminate all the excess reserves it created by engaging in a Seidman-type “debt-free” stimulus policy. The only way to avoid a ZIRP in this case is either to return to paying interest on reserves, or to ask the Treasury to sell new bonds. (Admittedly, the Fed is now awash in treasuries, and thus facing the opposite problem; still it is paying interest on reserves so can maintain a positive rate even with massive excess reserves.)

Here is our “teaching moment”:

Debt-free stimulus, or more generally a debt-free government finance spending proposal, actually requires interest payment on debt, unless the central bank adopts a permanent policy of ZIRP.

Either the Fed or the Treasury must pay interest on debt to avoid ZIRP. We can have the Fed issue the debt rather than the Treasury, but it is still debt and it still pays interest. Or we have permanent ZIRP.

This is why I made the claim that all debt-free money proposals reduce to permanent ZIRP.

For a more detailed explanation of why this must be true, see Scott Fullwiler’s piece from last year.

That is probably a big enough lesson for today. Let that sink in. In Part 3 I will explain why I think there are other shortcomings in such proposals, especially misunderstanding over monetary and fiscal policy operations. It will be instructive on that count to compare Seidman’s proposal with Lord Turner’s.