Insights from a Diagram-Machine


I’ve spent the last month or so tinkering with and observing a diagram-machine representing the workings of the U.S. monetary system. In the process, I can see that I’ve bored a lot of people beyond their capacity with the tedium of the tinkering. I apologize for that, and I’ll hereby discontinue the torture. Nevertheless, I’d like to share a few insights the tinkering revealed—at least to me—that made the exercise worthwhile.

1. Money-creation is a response to what the American people decide they want to produce and consume.

This, it seems to me, is a crucial insight because it reverses the way we habitually think about and visualize “money.” The habitual frame is that money exists first, then we decide what we want to spend it on—and then we determine if there is enough of it available for us to get what we want. While this is certainly true for the individual family or business, the diagram-machine revealed very clearly that, for society as a whole, this is a false framing. Actually, it works the other way around: We, as individuals, decide we need new shoes, and the private banks—through a process of accepting Promissory Notes in exchange for bank-dollars—create the “money” that will enable the shoes to be both manufactured and purchased. The Federal Reserve (FED) then issues the Reserves, as necessary, to back up those bank-dollars during the “clearing” process that happens at the Central Bank at the end of each business day. In other words, the amount of money in the system expands, as necessary, to meet the consumption decisions made by American society. (Of course, individuals, families, and businesses each have to strategize how they’ll earn or otherwise acquire some share of the money that’s created by this process—but that’s a separate issue from the question of whether there’s enough money available, or how it’s created.)

Or, on the other hand, we determine as a collective society that we need something big that will benefit our society as a whole (like an asteroid-blasting satellite laser-cannon) and the U.S. Treasury responds by issuing future Reserves—which it then trades for existing Reserves—enabling it to issue payments to the satellite and laser manufacturers who will build and deploy the city-saving weapon.

In each case, the money is created after and in response to an acknowledgement of the need. In each case, whether the need can, in fact, be met is NOT determined by “how much money is available” to meet the need, but rather by two other factors which will determine whether the required money shall be created—either by the private banks, or by the U.S. Treasury (in concert with the Federal Reserve):

  1. In the case of the shoes, the determining factor is whether the shoe production/consumption can be executed at a profit—meaning there is a buyer who is willing and able to pay more than the cost of production. If there is no demonstrated buyer who meets these requirements, the private banks will not accept the Promissory Note, the bank-dollars will not be issued, and the shoes will neither be produced or consumed.
  2. In the case of the collective good, the determining factor is whether the collective, democratic process can reach consensus that a satellite laser-cannon is a genuine need that will benefit American society. If that consensus cannot be reached, the Treasury will not issue the future Reserves and the satellite laser-cannon will not happen.

In the first case, the decision is a diffused, aggregate process guided solely by the principle of profit-making. In the second case, the decision is a political process of advocacy guided by the competition of interest-groups—many of which strategically confuse the debate by loudly (and bogusly) claiming “there is not enough money” available to go forward.

These are not really new or profound considerations, but it was interesting to observe how the diagram-machine made them operationally “visible.”

2. Treasury bonds (future Reserves) are different than corporate or municipal bonds

There is a profound difference between a treasury bond and a corporate, or municipal, bond—and tinkering with the diagram-machine made this very clear.

For a corporation or municipality to redeem its bond—and pay the promised interest—it must “earn” enough bank-dollars (claims on Reserves) to make the payments. The corporation must earn enough profits—and the municipality must collect enough local taxes or fees—to redeem the bond’s face value and pay the interest. This requires (a) a lot of work and effort on the part of employees and taxpayers, and (b) a lot of things must go “right” in the profit-making arena of private commerce (market demand, real-estate values, tax-base growth, etc.) which could go wrong. These factors create a substantial risk that when the corporate or municipal bond comes due, the dollars necessary for redemption and interest payment may not exist! Because of this risk, it is neither reasonable or logical to think of corporate or municipal bonds as guaranteed “future dollars”—because they might well prove not to be.

The operations of the diagram-machine made it easy to see that U.S. treasury bonds do not entail these risks. To redeem a treasury bond + interest: (a) nobody has to work and strategize to earn a profit, (b) taxes or fees do not have to be collected by the Treasury, and (c) nothing has to go “right”—except that the FED does what it was designed to do, and issues the new Reserves + interest promised by the treasury bond. And the only reason the FED might be prevented from doing that is if there were a willful and/or catastrophic undermining of America’s national sovereignty. Short of the collapse of America as a functioning nation, then, it is highly logical and reasonable to think of treasury bonds as “future Reserves”—because, operationally, that is precisely what they are.

In the same vein, and for the same reasons, while it is logical to view corporate and municipal bonds truly as a “debt,” it is illogical and inaccurate to view treasury bonds in the same way. Corporations and municipalities do not have the legal authority or mandate to create—out of thin air—that which they must use to repay their “debt.” The FED, in contrast, is designed specifically to do just that. When the Treasury issues its future Reserves, implicit within that act is the lawful promise by the U.S. sovereign government that when the future Reserves come due, the FED will make them “real”—and the FED can legally do that by simply crediting bank accounts with keystrokes. In that sense, future Reserves are Reserves, whereas corporate or municipal bonds are a debt that must be “worked off.”

I’ve written about this before—conceptually—so it was interesting to see how the operations of the diagram-machine reinforced the “truth” that treasury bonds do NOT represent a “debt” the U.S. government (or the American people) have to work, somehow, to pay off.

In this regard, the machine’s operations over and over led to the question: “Is this what they mean when they talk about ‘printing money’?” In each case, the implicit answer was, “No—it’s just the machine operating as it was designed to operate.”

3. Operating the machine-diagram creates a lot of “inflation fuel”!!

The biggest surprise (although, I suppose it shouldn’t have been) was to observe how rapidly the diagram-machine built up dollars in the spending accounts that fed directly into private commerce—a build-up that could, it seemed, create inflationary pressures. The most interesting aspect of this was to see which Production/Consumption Chamber—Private or Public—produced the greatest dollar build-up (and why).

Was money-creation by private banks responsible for the build-up? The machine-diagram made clear (for the first time in my mind, at least) that while private banking creates a lot of money, its very process inherently destroys most of the money it creates in an on-going, daily procession. When bank-loans are paid off, the principal (the bank-dollars borrowed) is, obviously, paid back to the bank. But what does the bank do with those bank-dollars? It simply cancels them—and is happy to do so, because what it is cancelling are claims on its Reserves (“real money”) at the Central Bank. The only bank-dollars that remain after a bank-loan is paid off are the interest dollars promised by the Promissory Notes. So private banking creates a lot of money, but in net-aggregate (at any point in time) the only money it’s really creating are the net profits of private commerce. That’s still a lot of money—but it’s only a fraction of what it appears the banks are generating.

On the other hand, looking at the direct spending by the Treasury to facilitate the building and deployment of an asteroid-blasting satellite laser-cannon, what we observed was that ALL of the dollars issued and spent were deposited in the private spending accounts of American businesses and citizens—and stayed there! The build-up of spending dollars directed at private consumption—dollars competing to buy goods and services in private commerce—was dramatic. And there was nothing that cancelled them out (except taxes, which we’ll get to in a moment).

This observation seemed like a red flag to me:  Did it give credence to the alarmist warnings that significantly increasing federal spending to address collective needs (as proposed by MMT) will lead to strong inflation pressures that could adversely affect private commerce? Did it support the arguments that federal spending should be minimized? —that government efforts to improve society should be checked? —that private enterprise should instead be given freer rein in the hopes it will provide the solutions to better the lives of American citizens? And what about the arguments being put forward by many—including myself—that government spending will have to dramatically expand to meet the new and unique challenges of climate-change and a robotic workforce?

To answer these questions, I put the diagram-machine through a few more operations to observe the options for removing the excess “fuel” generated by government spending for collective benefits. This led to some additional insights.

4. Tax-dollars really are spent by the Treasury—but they aren’t necessary to fund the Treasury’s spending.

Tax collections are obviously the first strategy for removing potential inflation-dollars from the spending accounts of private commerce. MMT enthusiasts (again, myself included) are fond of saying that tax-dollars aren’t used for federal spending—that the promissory note a dollar represents is simply cancelled when it is used to pay taxes; i.e. taxes “destroy” or “drain” money. When you observe the federal tax operation in the diagram-machine, however, a somewhat different perspective pops up.

Yes, when the taxes are paid, bank-dollars are debited (or “drained”) from private spending accounts—but the same operation also adds Reserves to the Treasury’s spending account. When the Treasury spends, those added Reserves get “spent”—along with the new Reserves generated by the Treasury’s operation of issuing future Reserves. The point isn’t that tax-dollars get “destroyed,” but rather that they reduce the amount of future Reserves necessary to undertake any particular spending for the collective benefit. The further point is that there is no inherent constraint on the number of future Reserves the Treasury must issue to achieve that particular spending goal. In other words, the spending could just as easily happen if zero tax-dollars were in the Treasury’s spending account.

The real effect of the tax-dollars becomes clear when the diagram-machine’s operation is completed by the Treasury’s spending: The Treasury’s Reserve account is debited, the Reserve accounts of private banks are credited, and bank-dollars are added to the spending accounts of private commerce. The net increase of spending fuel in those accounts depends on how much was “drained” out earlier by the tax payments. Whichever way you look at it, the net result of tax payments is a reduction of the potential “inflation-fuel” generated by federal spending for collective goods and services—the tax-payments, in other words, don’t “pay for” the spending, they reduce the net build-up of potential inflation-dollars created by the spending.

5. Inflation is still the problem

Despite the “fiscal space” created by taxation, a very high ratio of future Reserves to tax-dollars—as many of the spending programs advocated by MMT will require—still generates a lot of potential “inflation-fuel.” The question that arises, then, is how high can that ratio go—i.e. how many goods and services benefiting the collective good can be produced—before a level of problem-inflation is induced in private commerce?

Apparently, a formula or evaluation mechanism to answer that question has yet to be devised. In the meantime, it seems to be an argument between the alarmists and the optimists. Alarmists postulate horrific things will happen if the level of future Reserves issued exceeds a certain percentage of GDP. But this seems to have less to do with inflation than the hazy (and fundamentally incorrect) assumption that GDP somehow represents the amount of dollars available to “redeem” future Reserves: if future Reserves exceed that number of dollars, the thinking seems to go, it will be impossible to “redeem” them, thereby resulting in default and catastrophe for the U.S. economy.

As an optimistic architect, I look at the issue from what I call an “enabling strategy” for federal spending. By this, I mean that federal spending should be targeted toward the creation of goods and services that ENABLE private commerce to produce what American’s need—and ENABLES all American citizens to buy the needed things private commerce produces. What is a “needed” thing is negotiable, but it would certainly include healthy food and clean water, a safe and supportive dwelling, pre-K through college education, and life-time medical care. Designing, building, and deploying the “Enabling Structures” (including service structures) that accomplish this goal would—from my perspective—expand the markets of private commerce to absorb virtually all the potential inflation-dollars generated by the federal spending necessary to create them.

For me, personally, this feels like coming full circle from where I started, seven years ago, with my first MMT essay: Playing Monopolis Monopoly. The Enabling Structures I built in that essay, on the Monopoly game board itself—and actually photographed—I still consider my best explanation of what they are, what they are intended to do, and why they should be built. Full circle—but I think I understand the macro-economics of it much better now. It would be wonderful, from this point on, to focus exclusively on the strategy of Enabling Structures without having to endlessly search for explanations about why, as a collective society, we actually can afford to build them.

15 responses to “Insights from a Diagram-Machine

  1. I’m perplexed. As person who has spent a career in banking, being ever aware of loan-to-deposit ratios and ALCO meeting discussions, I cannot understand your comments per below. My archaic brain is molded to see that making loans is dependent upon taking in deposits, or other forms of funding. How can any other entity other than the sovereign issuer of a currency create additions to the money supply? What am I missing?

    “Actually, it works the other way around: We, as individuals, decide we need new shoes, and the private banks—through a process of accepting Promissory Notes in exchange for bank-dollars—create the “money” that will enable the shoes to be both manufactured and purchased. The Federal Reserve (FED) then issues the Reserves, as necessary, to back up those bank-dollars during the “clearing” process that happens at the Central Bank at the end of each business day.”

  2. Hi, I had an unanswered question in post #2. Why must the consumer pay 5% interest for a loan to the private bank when the Fed is issuing loans at 2%.
    Is there a reason to the middle man other than ensuring banking capitalists’ profit?

  3. Christopher Herbert

    Future reserves? The process of creating State money has been around for more than 3,000 years. The state dispenses currency then taxes it back in order to finance itself.
    Private banks do not create money, they make loans which creates deposits that are spent or consumed. The aggregate wealth is unchanged. Over time each dollar loaned comes back, plus interest payments. But no new money is created. Only Congress creates money. It creates dollars to pay its bills. Not debt. Congress has a monopoly over money creation and taxes. It spends money into existence, then taxes money out of existence. Spend into, tax out of existence. As to how much to spend, and how much to tax, depends upon values, resources and the willingness to save.

  4. Your entire diagrammatic scheme fits well with the metaphorical gift-certificate outline that I offered in comments to Part 2. The differences seem to be semantic and to some extent, emphasis.

    One semantic difference would be the idea of Treasury bonds as deferred reserves. I would see a Treasury bond owned-by-the- Fed as being equivalent in block size to a block of reserves held by the Fed. A Treasure bond owned by the private sector would represent a previous block of reserves borrowed for a second time by Government.

    A Treasury bond owned by Government would (in my scheme) not yet have been issued. It may as well be something created by the Fed itself as I suggested in my comment.

    Like you, I think your Diagram-Machine has yielded fruitful insight. Thanks for sharing it.

  5. Charles Silva


    Private banks cannot issue reserves, so they cannot increase the supply of High Powered Money (HPM). However, all deposits at private banks other than those dollars directly spent by the government can be traced back to loans. In that sense, loans create deposits. Deposits are used as ‘money’ in the economy, though they are a different type of money than reserves. For example, creating too many deposits may be an inflation risk whereas excess reserves has no inflationary effect. That is what people mean when they say banks create money: banks make loans, loans create deposits, and deposits are ‘money’.

    In reality, bank lending is not constrained by the quantity of reserves. The amount of reserves held by a bank is mostly dependent on how the bank can allocate its financial assets for maximum returns.

  6. Newton Finn

    While this non-economist has enjoyed and appreciated the exploration of this “diagram machine,” and looks forward to frequent visits to New Economic Perspectives, let me express a simple (what some would call “simplistic”) counterpoint. Frankly, I see little use in attempting to get non-economists (even highly-educated ones) to understand more than the fundamentals (what I like to call the axioms) of MMT, which are ably and entertainingly presented, for example, in Stephanie Kelton’s “Angry Birds” video. Once the concept of sovereign fiat money fully sinks in, with its liberating expansion of governmental agency tempered, inter alia, by the omnipresent danger of inflation, the non-economist knows all that he or she really needs to know about MMT. At this point of initial illumination, lay persons, having obtained a fairly good handle on what government can and can’t do in light of available or easily-obtainable resources, can concern themselves with specific governmental policies and programs they would like to see implemented. The essential role of economists would be to fine tune those policies and programs so as to steer clear of inflation or other structural problems. Those who realize that MMT reveals the possibility of a better world should (IMHO) focus on compelling descriptions of what that world might look like and the specific policies and programs that could get us there. This is a quite a different task (and I believe a more likely-successful one) than the attempt to make the laity cognizant of technical economic details, the proper domain of professionals.

  7. Creigh Gordon

    I have a slight quibble with #4. My “tax dollars” consist of an electronic balance deduction and a subsequent electronic balance addition to some Treasury account. Are they the same dollars in both cases, or are the Treasury’s added assets in fact newly created money? Realizing of course that there’s no practical difference between the two possibilities.

  8. This is for Dave (above) (I’m unable to link to his comment for some reason): I think you correctly answered the question in your comment. What puzzles me is that you seem to imply that banks should not make profits. The “ZEN” essays made clear, I believe, that private commerce, by its very nature, is driven by a profit model—and the banking operations of exchanging bank-dollars for Promissory Notes is very much a part of private commerce. The whole point of MMT, from my perspective, is that the profit-model of private commerce is very good (and even efficient) at producing goods and services that are profitable to produce—but is very bad (and inefficient) at producing things that are NOT profitable, but which we nevertheless very much need as a collective society. How are those things to be produced? That’s the question MMT is trying to explain.

  9. Creigh Gordon

    Dave, the Fed typically doesn’t loan money to banks, it typically buys Treasury securities (which adds newly created money to the banking system) until the interbank overnight rate (Fed Funds rate) hits 2% or whatever its target is. The Fed Funds rate between banks, which is a bank’s cost for borrowing money, is generally considered a risk free rate. But when a bank lends money to you or me, it is not a risk free proposition, and the bank adds a surcharge to account for their risk.

  10. Newton Finn: I believe I mostly agree with you! I simply became fascinated (perhaps temporarily obsessed) with thinking through and understanding how it all actually WORKS—i.e. why the “axioms,” as you call them, are true. Then I thought: maybe I can save a few other “non-professionals” the trouble of delving so deeply into the weeds. (Maybe I only succeeded in making the weeds thicker—I’m definitely not getting a lot of “aha!”s in response.) The last two paragraphs of the essay are the most important to me—and they are pretty much aligned with your comment. Having said that, I think the two challenges have to proceed hand-in-hand: making people aware of what “the world might look like,” as you put it, has to be partnered with making them understand that it’s actually possible to make it look that way.

  11. After reflecting upon this post for a day or so, I am noticing that there is no accounting nor assignment of ownership for an inventory of reserves. I think this is an important but obscure component of MMT framework.

    I think that MMT theorist would relate (or should relate) reserves to loans taken out by a currency creating government. For the reasons you articulate, the money created by such loans remains in existence until the loan is repaid which, since the 1970’s, seldom happens (ignoring roll-over of loans).

    The ownership sequence of reserves (when we include the Fed as a money creating bank) is
    1. For the very brief period of money (reserves) creation– Fed ownership.
    2. For as long as it takes to spend the borrowed, newly created, money(reserves)– the currency-creating-government.
    3. For as long as the money(reserves) has subsequent continuing existence– private ownership.

    Accounting for reserves is complicated by a government borrowing event which results in the private sector lending reserves to government. Government already created the very same reserves at the time of initial borrowing from the Fed! We cannot theoretically create additional reserves again which only leaves us the option of considering that government is borrowing and using previously created reserves for the second time.

    That said, the bank accounts of the private sector may not reflect the source of the reserves loaned to government. This condition occurs when private banks loan reserves to government without assigning the risk of loss (admittedly very low) to any private entity. When government spends the proceeds from a private bank sourced borrowing event, the bank accounts of the private sector swell simultaneously. The net effect of this sequence is to give us two ways to count our money supply–the sum of reserves and the sum of private bank accounts. The comment by Charles Silva (above) relates to this two paths-of-accounting phenomena.

  12. point 3, “Was money-creation by private banks responsible for the build-up? The machine-diagram made clear (for the first time in my mind, at least) that while private banking creates a lot of money, its very process inherently destroys most of the money it creates in an on-going, daily procession.”

    Caution: the destruction is a slow process, taking years, for the destruction of the principle involved, for example, in a housing loan. Thus the build-up concept has to look at rates of creation/destruction by sectors. And looking at those, we see that with property inflation, there is a substantial build up of money from bank loans. The private sector still looks to me to be responsible for the majority of build up of money, not the government. Just a small fraction of the money created in a year by private sector banks is destroyed in that same year.

    ~5 million home resales per year, with price increases and fees part of the mortgage creation (money supply inflation) adds up to a lot of money.

  13. And to add to that 5 million home resales per year are all the commercial property and large multidwelling units, taller buildings…

  14. Roger C. Sparks: Boy, oh boy. Things really get confusing down here in the weeds! And I was trying to make it all simple, as in: The private sector does NOT loan Reserves to the Treasury. The Treasury issues something it, alone, can create: future Reserves (which we call “treasury bonds”—the essay explaining why it is accurate to call treasury bonds “future Reserves.”) Banks then trade existing Reserves to the Treasury for the future Reserves. There is no “borrowing” involved. The banks either want the future Reserves for their interest bearing component or, if they’re a Primary Dealer “working for the FED,” they trade the future Reserves they’ve just gotten to the FED for new Reserves the FED creates specifically for the trade. Net result, the Treasury has existing Reserves to spend, the Primary Dealer has its existing Reserves back, and the FED has a new cache of future Reserves on its balance sheet. Here’s the missing piece: at this point, the FED has ALREADY transformed the future Reserves into existing Reserves when it made the trade with the Primary Dealer—i.e. the Reserves the Primary Dealer takes possession of in its trade with the FED are the future Reserves transformed into existing Reserves. What happens when the treasury bond now held by the FED matures? The future Reserves are simply cancelled since they’re already been issued by the FED. Another way to think of it: the FED simply absorbs them into its infinite capacity to produce Reserves at will. It all amounts to the same thing. I’m sure that clarifies everything.

  15. Yes. Down in the weeds, the details are many and hard to separate. But I’ve heard that the devil is in the details.

    Money carries with it a chain of accountability. Accountability requires assignment of ownership. It follows that when the FED creates reserves (reserves are money, albeit a special class of money), we must assign an owner to them.

    Likewise, when Congress authorizes a project and instructs Treasury to fund it, the Treasury Bond (future Reserves) subsequently created has an initial owner.

    Once traded, neither of these products can be destroyed by the originator because the originator is no longer owner of the product he created. Destruction must await a second trade restoring product ownership to the originator.

    That said, can’t I destroy either a bond or money that I have in my possession? Yes, I can but I will suffer value-loss represented by the bond or money destroyed.

    Yes, indeed! It is hard to clarify everything.