Recent Posts

Tuesday, September 1, 2020

The Monetary Monopoly Model (Re-Write)

The determination of the price level is an important claim within Warren Mosler’s MMT White Paper. From a theoretical perspective, the theory of price level determination is a key feature that is distinguishes MMT from other approaches to economic theory. This theory is related to the Job Guarantee framework (Section 3.3). The argument is that one could come up with the suggestion for a Job Guarantee policy from any number of directions, the emphasis within MMT is the link to stabilising inflation.

NOTE: This is an unedited draft of a section of my MMT primer. It is a re-written version of an earlier article.

Verbally, the concept is that the government fixes a key price, which it can do courtesy of its monopoly on (base) money creation. Other prices within the economy are set relative to this price. This section will outline a simplified Monetary Monopoly Model, which is based on descriptions elsewhere, as well as the article “Monopoly Money: The State as a Price Setter” by Pavlina R. Tcherneva. I have simplified the discussion to allow the elimination of the use of equations. It would be a straightforward exercise to attach equations to what I describe herein.

The model I outline is not meant to be one that could be fit to real-world data. Rather, it is the simplest possible canonical model that captures the price-setting mechanism. If one wanted to create a more realistic model, the next step would be a model of the Job Guarantee. Given that we have not had a full implementation of a Job Guarantee in a developed country, the mathematical structure of such a model is obviously open to debate, with no easy way to resolve disputes.

Backstory: Introduction of a New Currency

To visualise what is happening in this simple model, imagine that a country wants to introduce a brand-new currency, conveniently labelled the “dollar.” No (local currency) dollars were previously in existence, although we assume that there is a private sector that already engaged in commerce. For example, the commerce may have been an exchange economy, or perhaps the locals used the Polish Zloty.

Why would a country do such a thing? Colonial powers did exactly this, to requisition local labour and goods. The Tcherneva article and the book Understanding Modern Money by L. Randall Wray offers a history of the imposition of new currencies in Africa (using brutal means). The use of money as a means of intermediation as part of the requisitioning process has organisational advantages over direct requisition (e.g., the difference between drafting soldiers versus paying volunteers).

Model Assumptions

The model assumptions are limited.
  • The model is of the “dollar economy,” and the government has a monopoly on the creation of these dollars, and the initial stock of dollars in the hands of the “private sector” is zero.
  • The government imposes a fixed lump sum tax on the populace, and the threat of sanctions means that the tax will be paid. For simplicity, we assume that there is no counterfeiting, tax evasion, fraud, or bankruptcies. (This is standard for a mathematical economic model, but in a verbal model, this clause eliminates many “what if?” objections.)
  • The government wants labourers to show up at a workplace, and it pays a fixed $1 per hour. (This price was an arbitrary decision by the government – there are no pre-existing dollar prices to compare this to.)
  • The government does not lend to the private sector.
  • The time period of the model is weekly, and the lump sum tax to be paid is $1000 by the end of the week.
  • The model does not consider other transactions in the economy.

The First Week

Since we assume that the $1000 lump sum tax is paid, and that the initial stock of dollars outside the government sector is zero, the populace must provide at least 1000 hours of labour. Otherwise, the populace in aggregate is short dollars, and since it cannot borrow dollars from the government, and we assume no defaults that must labour income must be earned.
  • The simplest case is that exactly 1000 hours are worked, and so the private sector ends up with a balance of $0. The next week starts with exactly the same state as the first week, and the previous logic repeats.
  • Alternatively, the populace would provide more than 1000 hours of labour. The excess ends up saved in the form of government-issued dollars. This pre-existing stock of assets needs to be accounted for in the next week. That logic will be returned to later.
The previous statements all had to be true, under the previous assumptions. However, we cannot be sure about the exact mechanism by which the private sector provides those labour hours. Two example possibilities are as follows.
  • Each household sends representatives to work to meet its own share of the lump sum tax and turns over labour earnings immediately to meet the tax obligation.
  • Labourers show up to earn dollars in excess of their own share of the tax bill, and they exchange the excess to others in some fashion. For example, if there was a pre-existing zloty economy, they would trade dollars for zloty, and thus the dollars are an exchanged asset within the unmodelled zloty economy. The zloty wage rates versus the government-paid dollar wage rate would be an obvious fundamental ratio driving the exchange rate.
However, having the private sector immediately surrender all their dollars each week partly defeats the purpose of introducing a currency. The hope is that more hours would be worked, to create a stock of dollars that would then be exchanged within the private sector.

In any event, the government has accomplished its goal. It has gotten (at least) 1000 hours of labour effort, solely based on the enforcement of tax laws, while not having to offer anything of real value in exchange.

Second Week: Excess Money Case

The reasonable expectation is that more hours would be worked in the first week, to create a surplus of dollars for the private sector. For example, 1500 hours could be worked, to generate an excess of $500 at the end of the week. In that case, the situation is different. The government can only be guaranteed that 500 hours of labour will be provided – since the private sector can run down its existing balance. However, it is guaranteed that there will have been a cumulative 2000 hours worked by the end of the second week, with the excess of 2000 hours equalling the stock of dollars held by the private sector at the end of week two.

In order for the model to give a prediction about the number of hours worked, we need to add behavioural assumptions: how large a stock of dollars does the private sector want to hold at the end of the week? Since there is an infinite number of behavioural functions, this branches into an infinite number of models. For a simple example, Sam Levey has developed one example (URL:

Since the described situation is far removed from the situation in most real-world economies, there is no need to pursue the details of a behavioural function. Rather, they key characteristic of the model is that the government applies a carrot-and-stick approach to money: it creates money to requisition desired goods/services/labour at a fixed price, and it uses taxes to drive the demand for money.

Banks Do Not Affect the Story

As a lead-in to the discussion in Section 5.7 (discusses whether MMT ignores banking), the existence of a banking system will not affect the number of hours that have to be provided, even though “banks create money.” The reason why is straightforward: even if banks layered dollar-denominated deposits on top of government money, the payment of tax bills requires the banks to send government money to the government on behalf of depositors. (If a central bank existed, this would be a transfer of settlement balances at the central bank.) There is no way for the bank to get that government money without somebody providing the labour hours (that makes its way to the bank). Remember that we assume that there are no defaults, and the government does not lend dollars. (In the real world, central banks do lend dollars to banks.)

To be Interesting, the Price Set Must be Significant

We do not normally need a model of how a government can drive requisition using a tax and a fixed price. We want to have a model of the whole economy. To be interesting, we need to add other goods and services, as well as wage rates, to the mix. This can be done in any number of ways – and greatly increases the complexity of the model.

For the fixed price core of the model to be interesting, it needs to be a significant part of the aggregate economy. If it is, then we should see some form of model relationship between the fixed price and the other prices within the economy.

This is where the Job Guarantee comes in. The Job Guarantee wage is meant to be a living wage, and so it is directly competitive with all low wage jobs. As a result, the government has the hope of directly influencing the wage structure of the economy – which then feeds through to the prices of final output.

As such, the simplified Monetary Monopoly Model points in the direction of any number of possible Job Guarantee models. However, the key dynamic that should be common to all of these models is that a key price is being set as a policy variable.

Government as a Price Taker?

The obvious concern with applying the Monetary Monopoly Model to the real world is the following: what happens if the government is a price taker? That is, it lets the price of everything it requisitions to be set “in the market?”

From a theoretical perspective, if we look at the simplest models where the government is a price taker, the price level ends up being indeterminate. That is, any number could be used to satisfy the model equations. For example, the Real Business Cycle (RBC) found in Chapter 2 of Jordi Galí’s Monetary Policy, Inflation, and the Business Cycle. (The indeterminacy of the price level in RBC models is more worrisome to outsiders like Warren Mosler and myself than it is to neoclassicals.) This indeterminacy explains Warren Mosler’s comments in his White Paper about recognising the source of the price level. Neoclassicals have various methods to get around this indeterminacy, but they are largely mathematical kludges.

Although price level indeterminacy is not seen very often in the real world, one could argue that if the government runs its requisitioning strategy in a fashion that completely ignores its effect on the price structure, one might expect rapid inflation (e.g., unstable prices). This would be one high-level interpretation of the inflationary experience of the 1960s-1970s – albeit highly simplified.

Meanwhile, the reality is that most government spending does feature a fixed price element. Prices paid are fixed, although perhaps for shorter period. An ambitious modeller might attempt to show how having the government chase prices in the private sector – e.g., price taking behaviour – destabilises the model economy. In other words, doing exactly what conventional economists tell governments to do helps generate inflationary pressures in the economy.

Concluding Remarks

The Monetary Monopoly Model is the simplest model that captures key distinctive elements of MMT thinking. If one insists on using mathematical approaches, the differences between price-taking policies and price-setting policies needs to be examined.

Technical Digressions

  • I included the possibility of a pre-existing currency to make the back-story more plausible. Even in the historical colonial situation, state money is pushing out existing economic practices. However, the objective of the government in introducing a new currency is to drive out competing monies. Monopolies are not complete – things like Bitcoin exist – but competitor currencies in legal developed economies are generally insignificant. The other reason to add the digression in is to point out the “exchange rate” problem between the government-set price and prices in the private sector. For example, for a Job Guarantee model to be interesting, we need a private sector wage and the Job Guarantee wage. Since working in the private sector or working in the Job Guarantee are competing real economy transactions, there are behavioural constraints on the relative wage rates.
  • The model features lump sum taxes, which I normally reject as unrealistic. Simplicity is one excuse, and the other is that if we imposed a more realistic tax (e.g., an income tax), we need to have a better handle on the private sector economy, as well as some idea about the behaviour of the private sector within the model. To have an estimate of the amount of goods/labour it can requisition, the government needs at least a rough idea how large the tax take will be. If we are discussing real-world economies, we normally will have an idea of what taxes will be paid, unless they are completely novel (taxes on cannabis sales after they were legalised in Canada) or contingent on hard-to-predict events (capital gains taxes).
  • Price level indeterminacy arises in RBC models because the behavioural equations (derived on optimising behaviour assumptions) all imply constraints on relative prices: between the generic good prices and wages, and between spot and future goods prices. (There is also a relative balance sheet constraint between money and bond holdings.) However, the initial price level could be set to any level, and every other price just adjusts proportionally. One can point to three ways to eliminate this issue. The first method is the controversial Fiscal Theory of the Price Level, which creates other problems for the plausibility of the model. (The Fiscal Theory of the Price Level implies that the price level should immediately jump every time fiscal policy settings are tweaked, which obviously does not happen. Feeble attempts to explain away this discrepancy are incorrect.) The second is to impose a money demand component to the objective function that creates a demand for money, and since the initial stock of inherited money is fixed, forces the initial price to a particular value. The problem with this is that households essentially get a fixed amount of utility of holding a $10 bill, which does not match the reality that a fiat currency could be redenominated and nothing much would happen. The third obvious possibility is to invoke price stickiness – e.g., the Calvo price-setting mechanism. However, every treatment of the Calvo mechanism was purely forward-looking – there were not inherited fixed prices that would provide an initial scaling. The possibility of inherited fixed prices creates difficulties with finding a solution. There are more advanced ways of attacking this problem, but they are either somewhat unsatisfactory (argle-bargle about equilibria) or moving the model away from the canonical model roots.

References and Further Reading:

  • White Paper: Modern Monetary Theory (MMT). Warren Mosler. I worked with a copy dated 2019-11-11. URL: 
  • “Monopoly Money: The State as a Price Setter,” by Pavlina R. Tcherneva Oeconomicus, Volume V, Winter 2002
  • Understanding Modern Money: The Key to Full Employment and Price Stability by L. Randall Wray. Edward Elgar, 1998. ISBN: 978-1-84542-941-6
  • “Monopoly Money Redux,” by Sam Levey, July 17, 2020. URL:
  • Monetary Policy, Inflation, and the Business Cycle: An Introduction to the New Keynesian Framework, by Jordi Galí. Princeton University Press, 2008. ISBN: 978-0-691-13316-4.

(c) Brian Romanchuk 2020


  1. Excellent Brian,

    Gets you thinking about all sorts of stuff that keeps me awake at night.

    I think China understands all of that. Reading Australia's white paper on full employment from 1945 that Bill provided via his blog. Also shows clearly that the Australians were thinking about all of these things at that time. What it means for both domestic policy and the external sector. As you need a working model for both.

    The reason I think China is a better example of understanding MMT than Japan is that China have kept their main large exporters in public hands. Which means they can adapt to what they need to do more quickly and effectively. The exporters can act as one.

    For me the more I think about it if you understand MMT and the price level that is how you would set up your main large exporters. You can set both the quantity and the price. You would want to automate them as much as you can so that you are using least amount of workforce as possible. Leaving as many skills and real resources as you can for your domestic policies.

    When China learns/ steals everything they can from allowing western companies to take advantage of their cheap labour and move to running trade deficits. I'm convinced this is how they are going to set things up.

    China gets it. They have shown for years quietly that they get it. That they know how to set things up and understand fully MMT. More so than Japan who are under the influence of the US.

    Well that's my take on it anyways. China introduce policies that fully support their own Monetary Monopoly Model. They always choose the right policy When they need to. They get things done.

    As MMT'rs I know it is very difficult with the data that is available. I feel We haven't studied China enough. China for me shows what you can do and how you do it when you understand MMT. They are not finished yet not by a long shot. They have a fantastic handle on banking.

    1. China gets it, in addition to their common sense and self-respect, they were taught Keynesian economics by old-Keynesians who were being ignored by the ever-crazier US government and economic establishment. Who went to China to teach, like James Tobin. I have a book by a Chinese author somewhere on my hard drive, whose author and title I can't remember about this fruitful episode of education by Tobin & others. James Galbraith also spent some time advising China, and Mrs. Galbraith & Mrs. Wray are both Chinese. So until idiots trained at Harvard or Chicago start overruling the sane people, which may not happen, hope not, China will be fine.

  2. If the UK was by a miracle ever to vote for a left wing party that fully understands MMT and wanted to create a UK Monetary Monopoly Model.

    If they were to sit down and plan what they were going to do. You could bet that China has already implemented it.

  3. If I was the leader of the SNP and Scotland became independent. 10 years before I planned to set up a Monetary Monopoly Model in Scotland.10 years before I won independence.

    I would have had a full team of economists living in China.

  4. This comment has been removed by the author.

  5. (typo corrected)

    You can move your model a step closer to reality by assuming that a preexisting currency is used by the economy. Then government simply sets the value of uniquely new money as some ratio of old money. Presumably taxes could be paid with either new money or old money.

    With new money (which is nearly cost free to government) in hand, government can pay premium wages as needed to accomplish the goals of government.

    Now we can take a macroeconomic overview. Government has become a bigger employer within the economy. The macro economy has begun accepting elements of a second money system in exchange for real goods and services.

    it seems logical that people would rather work for government rather than private employers. Government could always pay better than private employers.

    1. A government cannot easily enforce an exchange rate with an existing currency by fiat. In countries where this is tried, parallel black market exchange rates pop up.

      Accepting tax payments in either currency destroys the premise of the government gaining a monopoly. Since my example discusses two weeks, I did not dwell on this.

      Yes, the model assumes that taxes are always paid. That is for simplicity. In a developing country, slippage between tax laws and reality is acceptable - and since the underground economy cannot normally be measured, invisible to standard national accounts.

  6. Oil is another great example of if you are a price setter or price taker.

    Many people now say China with their huge storage facilities are now setting the price for oil.

    U.S. and other producers prefer a market price at the inflated $50/$60 per barrel level, clearly China, India, Japan, and above all the EU who are all major consumers would prefer a price pegged between maybe $30 to $40/bbl since the difference between the two price bands represents a net transfer from consumers to producers approaching $1 trillion per year, and that is before the additional costs of refining crude oil to fuel in turn distributed and sold to consumers as energy services such as heat/cooling, power, mobility and so on.

    China has taken advantage of this market discontinuity to fill strategic reserves now over 1.2 billion barrels, and a stream of literally hundreds of oil tankers is currently delivering to China oil bought at distressed prices which will be available for immediate re-sale. Meanwhile, China has also finalized a system of sanction-proof buyers able to take delivery of crude oil anonymously through the Shanghai Exchange, and this oil may then be refined by independent Chinese "teapot" refiners to be delivered/dumped into the global market at prices with which no other refiner can compete.

    That market power has moved permanently downstream to the buy-side in a structural market shift, and that in due course will move further downstream to energy services.The

  7. Brian, I'm interested in the points you make about the pre-existing currency - the zloty. Clearly if the populace is required to pay tax in the new currency, it will end up having less need for the zloty stored in their bank accounts or elsewhere.

    They will likely start exchanging their zloty for the new currency, and the central bank will eventually have all or most of the zloty.

    Now, imagine that the zloty was a shared currency between 2 parts of a state, one part of that state became independent and created the new currency that you envisage. In effect, after the zloty were exchanged for the new currency, the central bank of the new state would be holding a lot of foreign currency as reserves - which would likely be a good thing.

    Would you agree or disagree with those suggestions?

    1. I didn’t get into it, but in the case of a pre-existing currency, the government should pass laws that push the other currency into underground status, on top of tax obligation. Legal tender laws, things like minimum wage laws in the new currency (dollars, here) would push things in that direction.

      As for the case of a new state, I am not sure what happened in previous amicable splits (e.g., Czechoslovakia). However, I am living in a province which has had at least three independence referenda in my lifetime. The situation is that you have a lot of Quebeckers living in houses with Canadian dollar mortgages owed to banks that are de facto headquartered in Ontario. Forcing those mortgages to the new currency - I call them Queebucks - is going to blow up bank balance sheets, or not doing it will blow up households. It’s hard to see how to avoid at least a temporary peg system, at least until the debt cross-holdings are sorted out (banks can split off Quebec holdings into a new subsidiary, and then work on hedging things out). Admittedly, some people want to see capitalism burn - but that was a total disaster in opinion polls. As such, the pro-independence party was telling people that they could keep their Canadian passports and use the Canadian dollar - which relied on the goodwill of the rest of Canada, with no guarantee that such goodwill exists. As such, your point about holding reserves matters.

    2. Thanks for that Brian. I had the same legal obligations in mind.

      Was there any analysis done on what might happen to the exchange rate after independence?

    3. People came up with suggestions - including Warren Mosler in the last one, I believe - but the Parti Quebecois quashed any official discussion of the topic. The reason is straightforward: there was a long-running empirical regularity in polls. No matter what the level of support was for “separation,” the result dropped by 10% if the Canadian dollar was mentioned. With polls peaking at around 55% (referenda ended up below 50%), that was fatal.

      This sort of squishiness was widespread. The first referendum (working from memory, which was in the 1970s), they invented the term “sovereignty association” and used that instead of “independence” since “independence” polled 10% lower than “sovereignty.” (For whatever reason, the Quebecois interpretation of “souverain” is softer than the English interpretation of “sovereignty.”)

      After the last referendum, the Federal Liberals has enough of the Parti Quebecois word games and passed a Federal Act that said that the results would be ignored by the Federal Government unless the referendum was extremely clear as to what the consequences are.

    4. I think Warren Mosler's suggestion was essentially the same as he proposed for Italy exiting the Euro - which may yet come to pass. Personally I'm in favour of exiting the £sterling along the lines Warren suggest (and not joining the Euro thereafter) so you might guess where I live.

  8. This comment has been removed by a blog administrator.

  9. This comment has been removed by a blog administrator.

  10. This comment has been removed by a blog administrator.

  11. This comment has been removed by a blog administrator.


Note: Posts are manually moderated, with a varying delay. Some disappear.

The comment section here is largely dead. My Substack or Twitter are better places to have a conversation.

Given that this is largely a backup way to reach me, I am going to reject posts that annoy me. Please post lengthy essays elsewhere.