Recent Posts

Thursday, July 9, 2020

What Is MMT? (Short Version)

One of the unusual things about Modern Monetary Theory is that many non-MMTers have a hard time understanding what it is. Although it would be reasonable to argue that this reflects communications issue by MMT proponents, I am not entirely comfortable with that characterisation. I had no issues picking up the theory, and the same could be said for the relatively large number of activists who have come from non-economics backgrounds. Many of the complaints about understanding MMT come from people who either work in finance or have a background in economics – which raises a question why they are having these difficulties.

Note: This is an unedited introductory section from my manuscript. It will be the second section, and designed to offer a minimal introduction to MMT. The rest of the book will give more details, but I want readers to at least have an idea of what MMT is in case they just jump to the MMT criticism chapter.

My argument why MMT is awkward to grasp for people with already fixed views about economics is that we need to draw a distinction between two concepts of MMT: a narrow version (or core MMT), and a broad version. (This is my preferred wording, but is probably non-standard.) Most MMT primers on the internet focus on the definition of narrow MMT, while many of the interesting questions in economics (which show up in financial macro analysis) end up in broad MMT. It should come as no surprise that if you are interested in broad MMT, watching an online video about narrow MMT will leave a lot of questions unanswered. However, based on comments of MMT critics, this is exactly what many of them have done.

Finally, in the interest of brevity, I am presenting this overview from my perspective, and others may feel that there are better ways of presenting the material (or possibly that I am mis-characterising MMT). Chapter 4 gives a longer description of MMT and stays much closer to the source materials listed therein.

Narrow MMT

The narrow version of MMT uses as its base case a fiat currency (and not an alleged barter economy, or pegged currency). For a fiat currency, government money is a public monopoly – and should be analysed as such (and policy should be set according to this analysis). The economic reason why governments issue money is to provision themselves easily, instead of demanding goods and services in-kind (e.g., demand that citizens work for the government without pay). (To be perfectly clear about the preceding sentence, I specified that this is the economic reason for current arrangements, this says nothing about the history of the adoption of money – which is part of broad MMT. I specify this, as commodity money proponents often rely on ahistorical founding myths for money, which is a topic of little interest for people interested in analysing modern economies.)

Fiat currencies mainly exist as electronic entries, and have no inherent value (unlike gold, which has uses in industry and jewelry). (Banknotes and coins are material and thus might have some inherent value, but that value has no necessary relationship to the nominal value struck on the token). To provision itself using money, that money must have value in trading in the non-government sector. The government needs to think like a monopolist and set its policy to ensure that government money has value in exchange.

One leg of the monetary monopolist view of government leads to what is known as Functional Finance (an Old Keynesian school of thought, heavily associated with Abba Lerner.) These core principles attach most of interest around MMT (particularly from a bond market perspective).
  • The role of taxes is to create a demand for money to meet tax payment obligations – which are denominated in the domestic currency. There is a real need to pay taxes – since there is a real cost to going to jail for tax avoidance – and so money has a real value. This explains why private actors rationally offer real goods and resources in exchange for electronic entries.
  • Since the central government is a monopolist issuer of its money, and that money is not pegged to any external instrument, there is no danger of the central government running out of money. (To be clear, this does not apply to sub-sovereigns, nor to central governments that do not control their currency, such as euro area sovereigns.)
  • This means that the central government is not finance constrained like private entities such as households, firms, and sub-sovereigns. (The preferred phrasing is that the central government is the issuer of the currency, all those other entities are users of the currency.) In particular, government bonds are not needed as financing instruments: their economic function is to remove “money” from the banking system, so as to allow for a risk-free yield curve with positive interest rates. (This allows the possibility of using interest rates to influence the economy – a policy lever that MMTers feel is much weaker than conventional analysis suggests.)
The previous leg of argument is well-known and captures most arguments about MMT in popular forums (and even some academic reviews by mainstream economists that cannot be bothered to enquire further into MMT). However, the next leg is more important from the perspective of defining MMT as being distinct from Functional Finance.

In addition to using taxes to ensure that the currency has some value, governments help determine the value of the currency by the prices they set in their dealings with the private sector. This does not fit in with conventional economists, who rely on models in which the government is a price-taker in every single market. In fact, conventional governments have deliberately moved policy so that the government acts as a price taker. From the MMT perspective, this is a huge blind spot, and helps explain the inflationary outcomes in the 1970s – government policies exacerbated the drift in the price level. The Old Keynesians – like the rest of the mainstream – refused to accept the role of governmental provision in setting the value of the government’s unit of account.

More specifically, a key observation is that the labour market is the most important domestic market. Conventional economists argue that citizens must be kept unemployed to discipline workers, which hopefully moderates inflation. However, Bill Mitchell and Warren Mosler independently argued that the government can enact a policy similar to commodity price support programmes. That is, have a programme that bids for all unused labour at a fixed price. This is known as a Job Guarantee. Other prices in the economy are then set relative to this fixed price. (For example, wages for some workers are effectively set as a markup over the Job Guarantee wage, and then output prices are set as a markup over labour costs.) The theoretical concept of determining the price level is not normally discussed, but is a fundamental issue within economic theory. More details are found in this article. (Economists previously suggested and experimented with such programmes, but never with the explicit aim of using it to determine the price level.)

Once these basic principles are in place, the focus then switches to monetary operations – what exactly is happening in the financial system in the context of government finance? The reality is that each jurisdiction has different legal and regulatory structures, and so a researcher needs to dig into each one independently. (One common feature of American MMT critics is to invoke some law in the U.S., and assert that this somehow globally refutes MMT.) The digging keeps coming up with the same answers – although there are self-imposed legal restrictions on government financing, the underlying economic forces match the MMT story outlined earlier. This is distinct from mainstream Economics 101 textbooks, which invent toy models of money markets that bear zero resemblance to how interest rates are determined in practice.

Another part of operations analysis looks at the banking system. The MMT story is straightforward: the conventional banks act as regulated utilities, and so “bank money” is an extension of “government money.” (This causes concern for some MMT critics, who argue that banks somehow create their own money without it having any relationship to government money.)

Another characteristic of MMT writing is an emphasis on sector balances and stock-flow consistent models. Since accounting identities are true by definition, the theoretical value rests more on the interpretation. A key example is that government debt is an asset for other sectors of the economy. Although this should be obvious, this is different from the prevailing narrative that governmental debt is an evil to be avoided. From the modeling perspective, this emphasis is tied to the stock-flow consistent (SFC) modelling approach.

This largely wraps up what I call narrow MMT. Despite all the arguments about these topics, it is extremely hard to find a serious fundamental objection to them. Most arguments end up being about wording or politics, without any substantive theoretical differences.

For someone new to economics, the previous topics looks like a lot of ground to cover. Understanding Government Finance (Amazon affiliate link) discusses most of the topics that are of interest to bond market participants at much greater depth. However, anyone experienced with macroeconomics would realise that macroeconomics involves more topics than what was listed above. This moves us toward Broad MMT.

Broad MMT

When I write “MMT” without qualification, I am almost certainly referring to broad MMT. My guess is that academic MMT proponents agree with this perspective. However, when outsiders discuss “MMT,” they only include the academic writings of self-described MMTers and the contents of online primers, and that ends up being closer to what I describe as narrow MMT.

My view of broad MMT is wide – and possibly wider than other MMTers. My starting point is to look at what the post-Keynesian academic Marc Lavoie describes as “broad-tent post-Keynesian” thinking. (I use the description found in Section 1.4.4 of Post-Keynesian Economics: New Foundations.) For readers who are unfamiliar with post-Keynesian economics, a key point to keep in mind is that it is a group that split from the mainstream relatively soon in the post-World War II era, and was deliberately pushed to the fringes of academia by the 1980s. Correspondingly, post-Keynesian thought developed a large body of thought in parallel with mainstream neoclassical economics before the advent of MMT proper.

Broad-tent post-Keynesian economics consists of a large number of different groups – not all of whom who agree with each other. (For example, the group that Lavoie defines as narrow-tent post-Keynesians argue that the other groups do not qualify as post-Keynesian.) I define broad MMT to be the parts of broad-tent post-Keynesian economics that is consistent with the core analysis of narrow MMT. This means that it can include research by economists who are critical of MMT elsewhere. Defining the boundaries of what thinking is consistent with the core MMT beliefs is going to be a judgement call, and not everyone will agree where the boundaries are (or even whether defining such boundaries is a good idea).

The textbook Macroeconomics by William Mitchell, L. Randall Wray, and Martin Watts is the best starting point for understanding broad MMT. However, that text is an undergraduate textbook, and simplifies the academic literature. One needs to dig into monographs and articles for more specialised references. (This text provides a starting point for such references.)

The following topics are a non-exhaustive list of areas of debate that I have run across over the years (and is obviously biased towards my interests).
  • Labour market analysis, and the inflation process. Alternative counter-inflationary policies.
  • Business cycle theory.
  • Rejection of the assumptions and methodologies of neoclassical economics. (Arguably, too much of post-Keynesian discourse consists of complaining about neoclassical economics.)
  • Analysis of fiscal policy.
  • History of money (associated with the Chartalist school of thought, many people refer to MMT proponents as “neo-Chartalists”).
  • Legal and social analysis of the monetary and financial system.
  • Analysis of the Job Guarantee, and historical analysis of similar programmes across the globe.
  • Analysis of the Green New Deal proposal.
  • The challenges faced by developing countries, as well as the difficulties associated with non-floating currencies (e.g., the euro area).
  • Although my interest is exclusively with macroeconomics, the late Fred Lee at the University of Missouri Kansas City was a noted researcher in microeconomics. (UMKC is a stronghold for MMT academics.)
The rest of this text gives a limited introduction to some of these topics. However, it is clear that one would need a graduate level textbook to just introduce these topics. Broad MMT is an entire school of thought of economics, and not just a model like the Quantity Theory of Money, or the Fiscal Theory of the Price Level. Based on conversations with neoclassical economists, they assume MMT is just a single model like those “theories” that can somehow be slotted into the neoclassical body of thought, rather MMT aims to completely replace neoclassical theory in its entirety.

Concluding Remarks

Online MMT debates are inherently frustrating. In most cases, critics have a limited grasp on what narrow MMT says, but then they immediately veer into discussing topics that fit within broader MMT. Meanwhile, it is an extremely difficult task to find a conventional critic that cites even a single academic article on MMT, which would be required to discuss broad MMT.

(c) Brian Romanchuk 2020


  1. "There is a real need to pay taxes.......and so money has a real value. This explains why private actors rationally offer real goods and resources in exchange for electronic entries."

    There is a need to pay taxes because private individuals generate income. "Private actors rationally offer real goods and resources in exchange for electronic entries" because they want to generate an income so that they can purchase other goods and services, not to pay taxes.

    Henry Rech

    1. They could use gold or copper coins, which have inherent value.

    2. ... which might be described as a form of bartering?

      For people with an interest in economics (which should be every educated person), this is the best intro to MMT I've seen.

    3. Thanks.

      Is using gold/copper coins bartering? Not exactly. People get used to the convenience of money, and are fairly happy to use coins at face value, without worrying about the metallic content. This means it’s monetary exchange not barter. Pre-Confederation Canada operated with fairly dodgy copper tokens issued by practically any firm. Even so, people did pay some attention to metal content, hence you had Gresham’s Law.

    4. BTW Brian that second anonymous is not me.

      I can't see the relevance of your comment on coins.

      Henry Rech

    5. Your argument is that people want money to do transactions. Sure. But what determines the price level in that currency? For a commodity-backed currency, it’s value is tied to the real cost of production or value for alternate uses. For electronic entries, we can multiply values arbitrarily without real effects.

    6. Brian,

      I still don't understand the relevance of your response.

      Money has "value" for a host of reasons, the least of which it is used to pay taxes.

      MMT emphasizes that money is used for paying the taxes because it adds weight to the claim that taxes are not needed to fund government expenditure. MMT has it the other way round - government expenditure funds taxes.

      Money is used for a multitude of transactions other than for paying taxes.

      Henry Rech

    7. It’s the question of the determination of the price level. For an electronic entry currency, there are no relative prices to tell us whether a wage should be $50,000/year or $500,000/year.


  2. That largely fits my view. I often describe MMT as a synthesis of many post-keynesian strands of thought. I have read many things that you may consider broad MMT - I consider them MMT compatible or consistent.
    That is not to say MMT has brought into its synthesis all relevant PK thought or other relevant thought. As long as it is not inherently in conflict with "narrow" MMT I do not see an issue.

    Also what Bill Mitchell calls the MMT project could probably be considered part of Broad MMT. I have referred to it as Applied MMT.

    1. I could change my terminology, but the narrow/broad distinction is very useful for understanding *arguments* about MMT. If I tell someone they read a “narrow MMT” primer, but the topic they are complaining about is in “broad MMT,” they should more easily see why the primer does not cover it.

      How broad is broad? Narrow MMT is the area where there is a consensus; you step outside that, you might be a broad-tent post-Keynesian, but not a “MMTer.” However, as we widen topics, people will disagree more, and whether one is “broad-tent post-Keynesian” or “MMTer” may not predict your views on the subject.

  3. "In addition to using taxes to ensure that the currency has some value, governments help determine the value of the currency by the prices they set in their dealings with the private sector."

    I certainly agree that 'value of currency' is linked to government spending into the private sector. More specifically, value should be linked to the source of governments ability to spend. We would have one model based on government obtaining money from taxation and a second model based on government obtaining money by borrowing from itself (which is the same as 'printing money'). A third model would also be needed, based on government obtaining money by borrowing from the private sector.

    Real versions of all he three models actively coexist simultaneously in the real world. Coexistence and the resulting interplay make for interesting macroeconomic conversation.

  4. Trying to understand the political (or not) nature of MMT. If we were to implement MMT+Job Bank up to the point where we achieved a realistic inflation threshold...but then we found that real wages had actually fallen as a result...would the MMT supporters still support it?

    1. That’s an unusual hypothetical. Why would real wages fall?

    2. I don't know that they would fall, but i could certainly envision scenarios where they fall or remain stagnant. It's hard to know in advance, the economy is a complex system.

      I suspect that despite being "apolitical," MMT proponents have a side-agenda of wanting to push up real wages...which is why i ask the question. MMT would lose all credibility if it were implemented but then had to be modified to achieve unstated political goals. It'd be better to state the goal up-front even if it means being political.

    3. MMT is a theory. The Job Guarantee is tied to the theory, but the details of the implementation are not set in stone. The argument is that the theory is itself apolitical, although a hard money advocate might disagree, but that is more a reflection of hard money ideology than MMT.

      Most MMT proponents are pretty clear about their political goals. I’m vague, but I don’t fit on a political spectrum most people are familiar with (I’m a Canadian prairie populist). Most proponents want a relatively high JG wage, which would have the expected effect of flattening the wage distribution. So the question in response to your query is: whose wages fall in real terms?

      In any event, it’s not as if MMTers would take over and all policy decisions will be handled by them. Elected politicians will implement the programmes, and they likely will change what they want to do over time.


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.