Note: My publication schedule is likely to remain erratic. I added in primers that were previously written, and the word count of my MMT manuscript has hit 29,000 words (and my target was 20,000-25,000 words). I am now mainly editing and cutting text, with only a few sections to be written.
J.W. Mason writes:
The big weakness is the book’s central theoretical claim that government has a monopoly in money creation, a theory known as “chartalism.” This claim has a serious problem: the existence of banks. It is true that government has a monopoly on currency. But most of the money we use in our daily lives is not coins and bills issued by the government, but ledger entries created by banks. Banks are money issuers every bit as much as the government. Government has tools to influence how much money is created by private banks, but its control isn’t absolute. And when its control is effective, that’s a function of the regulations and institutions of the financial system; it has nothing to do with the government monopoly on currency.The first thing to note that this quote comes in the context of a review of a popular book. Unsurprisingly, the book is missing a lot of theoretical content. My assumption is that J.W. Mason views this critique as applying to the MMT project as a whole (which it seems to be by context). I would note that J.W. Mason is not the first to make claims similar to this, so I my objective here is to use this quotation as an example, and not worry whether this is the most effective version of the critique.
Semantic Wrangling(Update: I have re-written the manuscript, and I think the following bit covers most of the debate.)
The simplest possible explanation for this debate is that Mason and MMT proponents are using “money” to refer to two different things. In particular, MMT proponents can be seen as referring to what is technically known as the monetary base (which consists of currency in circulation and settlement balances at the central bank), while Mason is referring to a broad monetary aggregate (e.g., M1 or M2).
I view the source of this difference as being the perspective taken on the analysis of the economy. The MMT perspective is viewing the economy from the point of view of the government, while the opposing view is taking the perspective of a member of the public.
From the perspective of the government, the government has a monopoly over the means of payment – currency in circulation, net positions in the wholesale payments system. Private “moneys” are subordinate, and layered on top of the monetary base. (This is exactly how private money is described in MMT primers, so this is not something that was forgotten about.)
Critics seem to be looking at this from the perspective of private sector actors, and thinking about how the private sector settles payments. From this perspective, bank deposits are the primary form of money, with banknotes being used for small transactions. In this case, the government does not appear to have a monopoly.
My concern with this private sector perspective is that it offers no useful information about the economy. Why don’t banks float multiple domestic currencies? Why do these private currencies have value, other than the circular argument that they are used in exchange? Why does a floating currency sovereign have no reason to fear default? We are back at the conventional view – which has very clear analytical defects.
In any event, the reader will have to accept that this is a semantic story. One can be distressed that MMT proponent define “money” in a way that does not match one’s intuitions. However, complaining about a cosmetic wording issue when people like myself have no problems understanding the meaning from context is not particularly impressive. Having written the book Abolish Money (From Economics)!, it is rather clear that I am not particularly impressed with semantic squabbling about the word “money.”
The rest of this article looks to see whether there are non-semantic issues.
Difference Between Government and Bank Money ArtificialThe first thing to immediately note is that MMT proponents hardly ignore banking in other contexts. Chapter 10 of Macroeconomics has the title "Money and Banking." Eric Tymoigne has a long sequence of primers on banking available at the New Economic Perspectives website: https://neweconomicperspectives.org/money-banking Correspondingly, it makes no sense to argue that MMTers are unaware of the existence of banking and money creation (as some other critics claim).
The next issue is that MMT researchers spent a good deal of time researching the details of monetary operations, both legal and institutional. The summary version I use is that banks are utilities -- their money creation powers are overseen by the state to achieve particular objectives. As Minsky argued, all private sector entities can be analysed like banks -- they want to issue liabilities that are used as money. The trick is getting other entities to accept those liabilities as money. The state set up the legal apparatus to give banks a special status, so that this private sector money creation can be concentrated under the eyes of regulators.
In any event , the division between "bank money" and "government money" is largely artificial. Although the private sector can set up rinky-dink payments systems like crypto-currencies, modern industrial capitalism is the story of large entities that send large payments to each other. Taking Canada as an example, that payment system at the time of writing is the Large Value Transfer System (LVTS). (The LVTS it is to be replaced by a real-time system.) This payment system is backstopped by the central bank (LVTS backgrounder, URL: https://www.bankofcanada.ca/wp-content/uploads/2010/11/large_value_transfer_system.pdf)
The Bank of Canada guarantees settlement of payments in the system in the extremely unlikely event that more than one participant fails during the LVTS operating day, and the amount owed by the failing participants exceeds the value of collateral pledged to the Bank of Canada.Banks make payments to and from the LVTS, and have a corresponding balance. Under normal conditions, banks kept their balances as close to zero as possible by the end of the day. (In 2020, the Bank of Canada hopped onto the Quantitative Easing bandwagon, and forced its balance to be negative at the end of day, which by implication left the private sector in aggregate with a positive balance. These settlement balances would be called "excess reserves" in American economics 101 textbook jargon, but that is a misnomer, since Canadian banks do not have deposit reserve requirement.)
Given the central nature of the wholesale payments system, other forms of money are of secondary importance. (Very few legitimate businesses undertake large transactions with banknotes.) Balances with the payments system are the unit of measurement (account), and there is no difference between "government money" and "bank money": a balance is a balance.
Despite feverish arguments you see on the internet, bank money creation is hardly a cost-free exercise. Bank money is generally created by extending loans, which is an activity that banks do under the watch of government regulators, which exposes the banks to credit risk. Meanwhile those deposits can go walkies via the payment system, and the bank needs to have the capacity to raise a countervailing cash flow the same day (except in the case where it is sitting on a large positive settlement balance, which is more common in the era of QE). Banks face very real constraints on money creation, and the government controls many of those constraints.
This is very different than the (consolidated) government that is a currency sovereign. It writes a cheque, and the only thing stopping the transfer are self-imposed constraints on intra-governmental accounting. External actors have little say in the matter.
This explains why governments keep bailing out the banking sector, and not the other way around. Given that most MMT primers (like The Deficit Myth) are discussing government policy and its constraints, it is clear that it makes no sense to treat bank money as equivalent to government money.
Paying Taxes with Private Money?Update: Added this section from my manuscript. Unfortunately, missing cross-reference text.
One argument that I ran into years ago made an argument to the effect that taxes are paid via cheque (or bank transfer), and not with “government money.” Furthermore, the means to make payment could have been borrowed from a bank – hence the money was created by the private sector. Does this contradict the Monetary Monopoly Model?
This logic is obviously weak. Even though the taxpayer used a cheque, the cheque is effectively an instrument that forces the taxpayer’s agent (the bank) to make a transfer to the government in the payments system. The bank will be forced to lose to liquid assets to meet the tax obligation. If it lent the taxpayer the money, it has replaced a liquid asset with an illiquid loan asset. Government regulators limit bank’s capacity to do such transformations.
As discussed in the linked article the Monetary Monopoly model does not preclude private currencies from circulating. The key is that government money – which is the unit of account in the wholesale payments system – is the money used to settle tax obligations.
Business CyclesIf the discussion is about the business cycle more generally, then yes, private sector credit growth matters. But it is archaic to argue that bank money creation is of primary importance; we live in a world where credit growth is increasingly in the hands of non-bank financing (also known as shadow banking). That said, this is a change of subject relative to most MMT primers. Given that the usual objective of fiscal policy is to be counter-cyclical, there needs to be other forces that generate the business cycle.
If the complaint is that there is no MMT magic solution to predicting the business cycle, I would argue that every other school of economic thought is in exactly the same boat. My interpretation of Keynesian business cycle theory is that the business cycle is inherently hard to predict, and so nobody should expect that any tractable model can predict the cycle.
(c) Brian Romanchuk 2020