No: Money Is Not Debt

August 27th, 2016

A quick note in response to recent twitter thread, and to a widespread usage that I find to be deeply problematic.

You constantly hear very smart thinkers about money saying that money is debt. I strongly disagree. It’s not a useful way to think about money. Quite the contrary.

Balance-sheet assets designating the value of claims may have offsetting liabilities/debts on other balance sheets. (Not all do; that’s why, for instance, U.S. households have positive net worth — assets minus liabilities, credit minus debt — of $88 trillion.) But in any case, the asset being “held” is credit, not debt. A holder of a Target gift card is holding Target credit, not debt — the “claim” side of the tally stick.

“Holding debt” is handy and ubiquitous (Wall Street) shorthand, but it’s conceptually incoherent. You can’t own an obligation; it’s not an asset. Money is not “debt.” Exactly the opposite.

I think this “money is debt” confution cripples our our conversations, our collective thinking, and our collective understanding.

Semi-aside: a dollar-bill is best thought of as a handy, exchangeable physical token representing a balance-sheet asset. Sure, that asset has an offsetting nominal “liability” on the government balance sheet (which we devoutly hope will never be “paid off”). That’s immaterial; the dollar bill represents credit, an asset. It’s incoherent to suggest that when you have a dollar bill in your pocket, you are holding “debt.”

Even if the asset you’re holding will someday be redeemed by the original issuer, the thing you’re holding holding is an asset, which you can transfer to someone else’s balance sheet in exchange for work, or real stuff or…some other financial instrument, be it a Euro bill a bond, a title to land, or whatever (which is also a credit, or asset). They’re all financial instruments (with various rights designated). Claims. Credits.

  1. David Andolfatto
    August 27th, 2016 at 08:34 | #1


    I’m not sure why you have your knickers all tied up in a knot here.

    First, I never say “money is debt.” First, I define what I consider to be money (an object that circulates widely as a payment instrument). Second, I remark that many debt instruments appear to fit this definition.

    The demand deposit *liabilities* of chartered banks constituting an important example. Note “liabilities” = debt.

    So, what are you on about here?

  2. August 27th, 2016 at 09:18 | #2

    I usually take any theory of what money “is” with a grain of salt (even my own wild speculations). Until there is widespread agreement on what money *does* then most definitions are just working definitions. Instead of saying “money is debt” one should say something like “assuming a debt model of money”.

  3. August 27th, 2016 at 11:07 | #3

    Steve, as I am sure you know, in a modern economy “money” a unit of account that appears on accounting records.

    Eric Tymoigne has recently posted a series on Money and Banking at New Economic Perspectives and also a first draft of forthcoming book at Scribd, with a link at NEP. There is a link to the series and the book in the navigation bar at the extreme right. Here is the link, too.

    Understanding these operations shows in detail what “money” actually “is.”

    “The dollar bill in your pocket” is a tax credit, as are settlement balances in payments system, both being liabilities of the Fed. US Treasuries are also accumulated tax credits that drain excess settlement balances created by fiscal deficits. US Treasuries are liabilities of the US Treasury. They must be converted to settlement balances (or cash) for settlement of tax obligations. which is easily done in the highly liquid market for Treasuries. (IRS Offers New Cash Payment Option
    IR-2016-56, April 6, 2016)

    The tax credits that are issued when government spends by directing the Fed as its fiscal agent to credit bank accounts are extinguished when federal taxes, fees, fines are paid (reflux)

    Similarly, dollars credit to bank accounts are bank liabilities created when banks extend credit by making loans. Those deposit credits are extinguished as loans are paid down with interest. (reflux)

    The government’s (consolidating Treasury and Fed) and banks’ liabilities are assets of the counterparties. while they have intermediate uses financially and economically that appear in stocks and flows, they eventually flow back (reflux) to the issuers through payment of obligations holders of those credits to the issuers. The reflux mechanism is based on legal mechanisms that create constant demand for the unit of account in which the obligation is denominated since those obligated must obtain those credits since they cannot issue them themselves, or else be in violation.

    Arguing over whether something is a credit or a debt, an asset or a liability just depends on which side is being emphasized. There is no credit without a corresponding debt or asset (LHS entry) without a RHS entry (liability or equity). IIRC, coins are booked as asset-equity, and the amount of sovereignty is the different between the asset (face value) and the liability (production costs), the costs being subtracted from face value to yield equity.

  4. August 27th, 2016 at 11:56 | #4

    Why then is money often called “debt”? Because money begins and ends with the issuer, be it the government or a bank.

    When the government issues money by spending it creates “money” as its liability and the asset of the holder. (M1 increases). When bank extends credit in making a loan, it credits a deposit account or passes cash through the window. (M1 increases)

    Money is destroyed when obligation created by government on others are met (M1 decreases) and it is also destroyed as loans are paid down (M1 decreases).

    Why is this important? Because there seems to be so much confusion and misunderstanding about it.

    For example, when banks extend credit they create money by crediting a deposit account. In doing this, they don’t “lend out reserves,” nor do they “lend out deposits.” Extending bank credit is not constrained by either reserve balances or funds on deposit at the bank. Nor do banks lend out capital. They lend against capital, that is, assume default risk.

    While money is a credit to an asset account or cash balances of users, it originates with the issuers as a liability and refluxes to extinguish a corresponding liability, e.g. tax bill or loan.

    I think that the kerfuffle now is over the term “debt-free” money, when the users of the term really mean interest-free money.

    Of course, commercial loans are never going to be interest-free.

    Government issuance does not depend operationally on getting money by borrowing from non-government since government simply issues its money.

    Excess reserves are drained into interest-paying Treasuries. Total non-governemtn net financial assets is not affected by the balance of cash, settlement balances in the payments system and tsys. There is no need operationally to issue Treasury debt to offset deficits. It is a political choice that constitutes a subsidy to savers. The interest rate can figure into monetary policy to support the relative value of the currency and to deter conversion to other currencies (or to real assets like metals where applicable}.

  5. August 28th, 2016 at 06:46 | #5

    First, fully acknowledging myself as a typical internet econocrank, going on about what money is (and isn’t). But we really are all fascinated and flummoxed by this, have been for centuries. So I claim the imprimatur of tradition.

    David, I go on so about this particular picayune usage because as @noahpinion has said so well, economics terminology is a dumpster fire. I’d say: more like the sub-basement of Fukushima Three.

    As a largely self-taught student of the discipline, I’ve flailed and wrestled with that reality for a decade and a half, constantly struggling to figure out what economists are actually saying with their words. It’s often, even mostly, deeply unclear. These are necessarily conceptual struggles. Hell, think “Wealth.” “Capital.” “Saving(s).” And of course money.

    The excellent work by Eric Tymoigne that Tom points to (I’ve been glancing through it of late, though I know most of it) is a very fine effort on this front.

    To take two examples in the comments here. No disapprobation intended: we’re in this mess together.

    Money: “an object that circulates widely as a payment instrument.”

    Instantly, I’m stumped and flummoxed. I can’t figure out what “object” means here. (Really, good faith: I spent a couple of minutes thinking hard, trying to impute the meaning(s) you genuinely intended.) If someone as savvy and thoughtful as David Andolfatto is stuck with that kind of usage, the discipline has some serious work to do.

    Tom: “in a modern economy “money” [is] a unit of account that appears on accounting records.”

    I totally get and agree with the gist and intent of this. But for all those poor souls who haven’t been studying and untangling this stuff for years, who can’t “read into” that statement:

    Wait. Are we talking “the” unit of account, the arbitrary measurement unit? “The dollar”? Or “a” unit of account? “A dollar?” (Don’t even get me started on “medium” of account. I have no idea what “medium” means there — and not for lack of trying.)

    Using the same term for two completely different concepts, operating on two completely different conceptual levels, is deeply confusing to eager seekers like me. And not just me: You can watch very knowledgeable and thoughtful people talking past each other constantly based on this exact unit-of-account category confution.

    (Would “unit of exchange” be a clearer and more precise term for the latter concept?)

    I’m going to stop here because there’s no way I’ll douse this dumpster fire in a blog comment. But to confirm my econocrankery, I’ll end with my definition of money:

    The value of balance-sheet assets, designated in (necessarily arbitrary) units of account.

    The confusion of this conceptual entity with “currency” and currency-like financial instruments, in my opinion, creates much of the muddle in our discussions. Category error. Roughly: Currency, a particular type of financial instrument, is one thing that embodies money. (This despite the most common vernacular usage, where currency clearly “is” money.)*

    When somebody clever added arbitrary units of account to tallies (primitive proto-balance sheets) tens of thousands of years ago, they created the social-accounting technology that we call “money.” The first currency — coinage — wasn’t invented until 700 BCE.

    @ericlonners, there’s the “etymology” of money. (The etymology of the English word “money” being something quite different.) We’re still working on the semantics.

    (I would never, by the way, say “oh that’s just semantics.” Because words/verbal language are the primary things we use to “think together” — along with accounting constructs, math, music, facial expressions, visual arts, dance, etc. If we’re not using words precisely, clearly, and in the same ways, we’re not thinking together (well). And, wildly social and verbal species that we are, we may not be thinking well individually, either.)

    * Though I suppose that depends on what the meaning of the word ‘is’ is… 😉

  6. August 28th, 2016 at 08:56 | #6

    “Are we talking “the” unit of account, the arbitrary measurement unit? ‘The dollar’? Or ‘a’ unit of account? ‘A dollar?'”

    Good point. Confusion arises from using the same term for both the currency (“the dollar’) and a unity of account (“a dollar”). China avoids this by calling the currency “renminbi” (RMB) and a unit of account (“yuan”). However, the currency market symbol for the renminbi is CNY rather than RMB. I guess they are confused too. As a result of this confusion there is no longer a distinction between renminbi and yuan. They have become interchangeable in usage.

    In addition, there is no distinction between money as an abstract unit of account and concrete money things (notes and coins.). “The currency” indicates the unit of account as denomination, and “currency” also means “cash” as a synonym for “currency in circulation.”

    The point this illustrates is saying what you mean and meaning what you say. Confucius called it “the rectification of names,” and pointed out that when meaning is unclear, confusion results and along with it all sorts of problems arise. Socrates points out that this is where sophists hide in The Sophist.

    I assume that this is what Steve is trying to say, and I agree with him, and not as a lay person but as a philosopher of language and logician. (I wrote my dissertation in Wittgenstein). Sorting this stuff out is what analytic philosophers do.

    Ordinary language is rich in meaning but at the expense of clarity and rigor. Science uses formal languages to specific meaning technically and to tie it tightly to reality through operational definition in achieving rigor.

    Using the same term for a variety of uses with related or overlapping but not univocal meaning invites confusion. Economic is full of this sort of thing. “Money” is one example. “Saving” is another. So is “investment.” This often results in category errors even among experts who don’t pay careful attention to how the meaning of the terms they use shift on the course of an exposition, for example.

  7. August 28th, 2016 at 09:13 | #7

    @Tom Hickey Eliezer Yudkowsky, in something I’m not bothering to google for, has a great recommendation. When interlocutors find themselves talking past each other and tangled in knots around some word or usage, ban that word from the conversation. Require each to use long form instead.

    That is not any kind of perfect solution, leaves us rather bereft, but still…

  8. August 28th, 2016 at 09:37 | #8

    @Asymptosis Great, if the parties agree, but that is often not the case in economics, it seems anyway. From what I can observe, when economists use ordinary language, there is a lot of imprecision if not outright confusion, and when they resort to formalization, the assumptions are often so restrictive that the models are not representative. Then the handwaving begins.

  9. August 28th, 2016 at 09:45 | #9

    @Tom Hickey

    I am not picking on economists here, either. We happen to be discussing economics, but this is has been an issue in philosophy almost form the outset of the discipline millennia ago. The Greeks were well aware of it and some addressed. The medieval scholastics had some good logicians, too. But the deep critique didn’t take place until the rise of analytic philosophy in the late 19th century when philosophical logic, philosophical (linguistic) analysis, and philosophy of language became sub-disciplines, and semiotics, general semantics, and linguistics spun off into their own disciplines.

    Philosophy of science and philosophy of social science (aka “foundations”) are fairly well developed sub-disciplines, too. Philosophy (foundations) of economics, not so much.

  10. August 28th, 2016 at 09:57 | #10

    @Tom Hickey and @Asyptosis — in the past when definitions and discussion of some concept became the subject of unresolvable debates it usually meant the concept was not useful (or even flawed). I think that’s why the heuristic to remove it from discourse can work.

    In my own wild speculations, money seems to be a physical manifestation of the chain rule in calculus. No amount of definitions based on ledgers or units of account can reach that definition — units and liabilities are only more or less proportional to money. Of course this may not be correct, but it is an existence proof of a definition of money that has nothing to do with any current definition. The closest is medium of exchange. But any particular “medium” is only proportional to “money”.

    Most definitions of money make assumptions about what money does: cause inflation, a government liability that solves a trust/coordination problem, a ruler defining the size of prices. As I said above, most definitions should be considered models of something that isn’t understood.

  11. David Andolfatto
    August 28th, 2016 at 10:42 | #11


    I agree that it’s important to get terms right. One way to do this is by way of an explicit model economy. Consider the Wicksell triangle I discuss here:

    You should not be flummoxed by the term “an object that circulates widely as a medium of exchange” if you look at that example economy.

    In terms of the title of your post “money is not debt,” it is I who am flummoxed. Money is not like a lot of things. So what? You should begin with your definitions of money and debt, then state your proposition.

    As for myself, money may not be debt (because most debt does not serve as an exchange medium), but debt can be money. The demand deposit liabilities of modern day chartered banks constituting an important example. But also the bearer notes issued by chartered banks in (say) the U.S. free banking era 1836-63.

    Now, I’m not sure how much of what you and Eric are talking about relies on the notion of what economists call “outside money.” (Essentially, an object used in payments but which is never called or backed in any way by assets.) Here is an interesting example of fiat money in this sense:


  12. August 28th, 2016 at 11:16 | #12

    @Jason Smith

    The term “money has many uses. It is not appropriately defined exclusively in terms of a single “essence,” as some are wont to do, e.g., “real money is gold. This is a view held by a diverse set that includes the opposite extremes of Karl Marx and Austrian economists and the Libertarians the follow them.

    “Money” is a term of high level abstraction — a set containing a nested network of subsets. There is no issue in taking one of these meanings and using it in model-building as long as the definition is precise and the term is used consistently. The model can only be interpreted legitimately in terms of that postulated use, and if the model is interpreted as representational, then it is subject to evidence.

    Some hold that the language of economics is the language of science — mathematics. Others hold that the language of economics is the language of commerce — accounting. These are different versions of economics with different uses. Both have application if correctly applied.

    What I take issue with is that the methodological debate is forever settled in favor of the mainstream. That’s just dogmatism when there are good reasons to think that it is arbitrary.

    I support the foray into information transfer economics, and I wish you success in its pursuit. I hope that it gains traction since it seems to make a useful contribution to thinking about economics.

    Whatever approach is selected, it must conform to logic as a boundary condition. There is nothing illogical about construction a fresh approach to logic either. The fundamental rule of logic is that whatever set of rules one chooses to follow, the rules must be clearly delineated and followed consistently. This is true of “fuzzy logic” too.

    I am very much with Paul Feyerabend on scientific method. There is no overarching rule. This accords with Wittgenstein’s logical view that there are no absolute criteria in logic.

    Once a system of logic is adopted, the rules are absolute in that system and if changed, then another system is created. But Wittgenstein points out that there is no illogical in creating a logical system in which the rules change even arbitrarily, but that system would have limited use, e.g., a child’s game or a fantasy.

    This freedom accords with the liberal principle of freedom of thought, which entails freedom of expression, since language and knowledge are social. Discovery depends on freedom to explore.

    The arts and humanities are rich in meaning, which inhibits rigor. The logic, including math, and the sciences are rigorous in meaning which inhibits richness. They apply to different types of human endeavor.

    Moreover, there is no clear boundaries that separate different dimensions of reality. Reality is an integrated whole. Distinctions are intellectual. Some reflect (or approximate) actual boundaries in the world, others do not.

    This implies the epistemological principle of consilience, that all knowledge of reality must fit together and that evidence from different angles eventually converges. This is the human project. The sciences play an important role in this, but neither the dominant nor only role.

  13. August 28th, 2016 at 20:07 | #13

    I bought three things today. In each instance I paid with money, specifically my debit card. In essence I was trading a Bank of Montreal book entry to a merchant for goods. That book entry is a liability of the Bank of Montreal. It is their debt. So a Bank of Montreal book entry is both money, and debt.

  14. August 29th, 2016 at 06:22 | #14

    @David Andolfatto: First to say, I wasn’t really responding to you. More to this:

    JP Koning ‏@jp_koning Aug 27 @ericlonners @dandolfa But when too much money is created, marginal value of liquidity services falls to zero—money once again just a debt.

    …and similar, in this twitter convo and widely, easily googled, elsewhere.

    So I think my objection to this widespread usage is not unfounded.

    Also I did understand the gist of your phrase, and the constructs behind it. I was only confused by “object,” which you clarify in your post:

    “they can be paper or book-entry objects”

    Absent that additional explanation, the usage strikes me as obscure and confusing. I much prefer your use of “claims” therein. That cuts to the crux of what money is.

    I also like your usage of financial “instruments.” But you don’t address what I think is my key point: there are many kinds of financial instruments — from dollar bills to land titles — all of which embody or instantiate money. Like a gallon of gas instantiates energy.

    I’m suggesting that money is most usefully thought of (technically, as a term of art) in a different conceptual category than financial instruments. Cordoning off certain (currency-like) financial instruments and calling them money sows confusion, causes us to talk past each other.

    If you own (have claim to) twelve apartment buildings, with their value posted as assets on your balance sheet, do you have “money”? I’d say, certainly. That’s a quite standard vernacular usage, and I think that meaning and usage (as opposed to “dollar bills in your pocket,” for instance) is fruitfully tapped in defining our technical term of art.

    IMO the best measure of the “money stock” is aggregate household net worth. That’s the pool of claims that circulates. (And I think reversing Piketty’s r, thinking “velocity of wealth” instead, could yield deep and important theoretical and empirical insights. [Damn, I’m a monetarist.])

    Take on that concept and usage, provisionally. If we also call certain (currency-like) financial instruments money, it’s just like the “unit of account” problem discussed above. Two meanings for the same term — with the concepts existing on different conceptual levels, in different conceptual categories.

    Hope this helps or is at least interesting.

  15. August 29th, 2016 at 06:38 | #15

    @Jason Smith

    “when definitions and discussion of some concept became the subject of unresolvable debates it usually meant the concept was not useful (or even flawed).”

    Yes. The trick being that the concept and the usage/definition are iteratively intertwined. Again, we use words to think together — and maybe even to think. If your language has no word for “orange” (many/most didn’t, until recently) in what sense can you (or we) “think about” orange?

    I so wish I had the conceptual toolkit to understand your second paragraph. Multiple complex terms and concepts in there that would require long and deep study for me. Do you have any aspirations to unpack that for the likes of me? Would it even be possible?

    “Most definitions of money make assumptions about what money does”

    I’d go farther: they only define money by what it does. Which can never constitute a proper definition. What “is” is, and all that…

  16. August 29th, 2016 at 07:02 | #16

    @Tom Hickey

    All very nice thinking, thanks. Two comments, the first quoting my own tweet.

    >”Distinctions are intellectual. Some reflect (or approximate) actual boundaries in the world, others do not.”

    The visual system starts modeling when photons hit the retina. You literally can’t even see anything without a model. Empiricism ain’t easy.

    >”the epistemological principle of consilience, that all knowledge of reality must fit together”

    I’m most frustrated by the lack of mutual “inter-coherence” in the “glossary” of economics. I’m constantly struggling to understand what an economist’s use of Term X means in terms of Term Y and Term Z.

    This is where I find myself constantly running up against the “boundary condition” of logic.

  17. August 29th, 2016 at 07:10 | #17

    @David Andolfatto

    Oh to add: I don’t find the inside/outside money construct to be very useful. Metaphorically, this like the etymology/semantics distinction — where the money comes from vs what it is.


    I borrow money from a bank, add that newly-created “inside money,” those credits, to my account (assets to my balance sheet).

    I buy 2x4s.

    Now Home Depot is holding those credits, along with credits it received from government def spending for 2x4s (outside money).

    Are those particular credits that I transferred to Home Depot’s account still “inside money”?

    Inside and outside might be useful to distinguish flows into the nonbank private sector, but I prefer less-coded language: funds from bank lending versus funds from gov def spending.

  18. August 29th, 2016 at 07:32 | #18

    @JP Koning

    We’re talking past each other, I think, because there are multiple book entries at play. At least: The credit (asset) book entry in your BOM account, and the liability book entry in BOM’s “house” account.

    I think it’s most useful stated as: you transferred your credit book entry to the merchant. A cascade of other book-entry adjustments ensues.

    When you say “In essence I was trading a Bank of Montreal book entry to a merchant for goods,” you’re telescoping the change in your account with the change in BOM’s house account.

    If we’re telescoping like that, we could just as well go all the way up to the tippy top of payment resolution, and say “In essence I was transferring Fed reserves (or Fed liabilities!) from my account to the merchant’s account in exchange for goods.”

  19. August 31st, 2016 at 08:23 | #19

    “I think it’s most useful stated as: you transferred your credit book entry to the merchant. A cascade of other book-entry adjustments ensues.”

    I’ve go no problem if you want to change the wording.

    But I still stand by my offending tweet:

    “But when too much money is created, marginal value of liquidity services falls to zero—money once again just a [book entry].”

    …where I’ve substituted in the word ‘book entry’ for ‘debt.’

Comments are closed.