Recent Posts

Thursday, October 21, 2021

Money Printing (Sigh)

Note: This is an unedited draft section from my inflation text. To what extent it depends upon observed behaviour, those figures will appear in another section (that needs to be written).

In a lot of financial market commentary – and even in the output of some academics – one might often see discussion of “money printing” in the context of inflation. Although this section explains why the concept is not meaningful, these references do serve a purpose: they are a signal that the reader can stop taking the person invoking “money printing” seriously.

In reality, what happens is that governments spend, and this spending may or may not have inflationary effects. Given the uncertainty about the inflationary impact of government spending, this text will not attempt to offer a definitive take on that subject.

Money is Printed – What Did You Expect?

The first thing to note is that all government issued banknotes are printed – they are pieces of paper (or plastic, as is the case for modern Canadian banknotes) with a denomination and other pieces of information (serial numbers, etc.) printed on it. They are not grown on money trees, nor mined in money mines.

As discussed elsewhere in this book, countries with fiat currencies have encountered any number of inflation regimes – deflation, price stability, and inflation rates ranging from low and steady all the way to hyperinflation. In all cases, their money was “printed.” As such, “money printing” by itself cannot tell us what inflation regime a country will be in.

Money Is Not Dropped from Freaking Helicopters

Based on discussions I have seen, it was (and possibly still is) possible to get a graduate degree in economics from a mainstream department and believe that governments pay for things by dropping money from helicopters or handing out wheelbarrows full of banknotes. This is somewhat misleading, at least for any modern developed country. I am no longer young, yet I have never been handed banknotes by a government employee (other than the corner case where the government runs retail-like establishments, like post offices). This is not atypical – the banknotes one normally encounters were withdrawn by an individual or a business from a bank (or a counterfeiter).

Governments spend via issuing cheques (“checks” in American) or by sending electronic transfer notices. If they print anything, it is the cheques. The cheques are not money, they are instruments that result in banks creating deposits (“bank money”) when cashed. This is perhaps being too finicky, but if one wants to dig into government finances, this is the correct framing. This is commonly discussed in Modern Monetary Theory (MMT), as well as my book Understanding Government Finance.

Money Circulates … Until It Doesn’t

When discussing money, one property that some commentators focus on is that it appears to be “conserved” (as the term is used in physics) – money is normally passed from entity to entity without being destroyed. Even if one person does not want to hold some money, they spend it, and it will be passed on to someone else. The phrase “hot potato” might be invoked, with the backstory that if people are worried about inflation, they get rid of money as soon as possible, and so it feeds faster transactions – increasing inflationary pressures.

The problem with this story is that money is destroyed.

  • If households do not want to hold banknotes, they will either return them to banks, or spend them at retail establishments. Retailers will then ship the banknotes back to the bank – they normally want to start the day with a target amount of cash. If the banks end up with too many banknotes, they ship them back to the government, at which point they effectively drop out of money supply measures. (There is also the continuous replacement of old and damaged banknotes and coins.)

  • Bank money is destroyed whenever someone repays a bank loan. They had a deposit (which counts as bank money in the wider money aggregates), which disappears (along with their bank loan liability).

  • Tax payments remove deposits of banks held at the central bank by private banks (often called “reserves” due to the influence of old American textbooks, but are “government money”) as the tax payment is transmitted to the governments. (Some critics of MMT make a big deal about the fact that taxpayers will meet tax obligations via a cheque or transfer from a private bank, being oblivious that such an act is just an instruction to their bank to make the transfer with government money. If the bank cannot meet that obligation, the payment attempt will fail – and deposit insurance would only cover a limited amount of a large payment.)

  • Governments issue bonds and bills, which also remove private banks settlement balances at the central bank when the bonds are paid for. Since government bond/bill issuance is often close to the fiscal deficit, this issuance tends to cancel out net spending by the central government.

There is a small technical issue with respect to the final two points. Some MMT critics point out that governments might hold deposits at private banks, and so the transfers just result in transfers of private bank deposits. However, this is avoided by many governments (such as the Government of Canada), and exposure to private banks is normally limited, as they can fail – and are reliant on the government bailing them out, which is harder to do if the government mainly deals with that private bank.

The fact that the circulation of money can be broken has a result: money holdings are generally the result of voluntary decisions by the private sector (with an important exception discussed later). If we take as a baseline the Canadian model for government finance before the Pandemic of 2020 (which is discussed in Understanding Government Finance), the only “government money” in existence are the private sector holdings of banknotes and coins. Those are held by households as well as banks and retail-facing businesses to handle cash transactions, and the amount held is obviously a private sector decision.

Even bank deposits (bank money) are voluntary. If households and businesses do not want to hold bank deposits as a financial asset, they can allocate towards other fixed income instruments, often held in money market funds. If depositors rotate towards money market funds, banks will face liquidity outflows that need to be funded somehow. Although they can raise short-term funding in a variety of ways, one theoretically simple solution is that they issue commercial paper to the money market funds that have inflows. The result is that bank deposits are destroyed – which are part of narrow money aggregates like M1 – and replaced with short-term fixed income instruments. (The reason deposits are destroyed by commercial paper issuance is that the issuance results in deposits flowing back to the issuing bank. Since a bank does not hold a deposit in itself, the deposit disappears.)

In a bid to save the concept of “money” in economics, desperate economists invented wider monetary aggregates (such as “M3”) which include money market instruments. Although this means that “M3 money” is not destroyed by such transfers. “M3 money” includes a lot of instruments that hardly resemble what most people consider “money.” Furthermore, these wide money measures are basically just an arbitrary set of short maturity fixed income assets, and the proportion of such instruments within portfolios are largely determined by investor portfolio allocation decisions. For such measures, “money growth” can have very little to do with governmental policies, and so do discussions of “money printing” do not apply.

The Exception: Deposits at the Central Bank

The exception to money being voluntarily held are deposits held at the central bank (“reserves”). The quantity of reserves in existence is largely under the control of the central bank. The reason is straightforward: the only way private banks can destroy them is to either transmit funds to the government for taxes or for auctioned bonds, or to withdraw banknotes or coins. The former is the result of fiscal policy and can be counter-acted by the central bank compensating for those flows, and the latter (withdrawing banknotes and coins) presents operational risks if done in size (bank robberies).

This is not like the case of private banks since under current institutional arrangements, central banks do not borrow from the private sector. This means that there are no loans to be repaid with central bank deposits, nor central bank issued commercial paper to absorb deposits.

Why this exception matters is the fad among central bankers to increase the size of their balance sheets – typically referred to as Quantitative Easing (QE). The central bank buys bonds (mainly government bonds, but not always), and the payment for the bond creates a deposit for a private bank at the central bank. Since the private banking sector can generally only redistribute these deposits and not destroy them (as noted above), this shows up a growth in monetary aggregates since they include those deposits.

Japan was the first country to experiment with this policy in the modern era, but the Bank of Japan was followed by Occidental countries after the Financial Crisis (and/or the Pandemic of 2020). The immediate effect of these policies was commentators predicting imminent hyperinflation (spoiler: they were wrong), and then an ongoing debate by more level-headed onlookers about the effects. I am in the camp that the effects were minor – at best changing some spread relationships between various fixed income instruments. The reasoning behind this was straightforward: why should I care whether banks hold deposits at the central bank instead of government bonds as their liquidity reserve? Since my objective is to avoid going into theoretical disputes in this book, I will just let the reader look at the charts comparing money growth to inflation that appear later in this chapter.

Effect of Voluntary Money Holding

We can now circle back to my main point: if money holdings are voluntary, the quantity of such holdings may not in any sense “cause” inflation, instead, money growth just reflects inflation (or more accurately, nominal income growth).

The reason is straightforward: it is entirely plausible that the amount of money private sector entities want to hold is related to their nominal income. If your nominal income is double what it was some years before, it would not surprise anyone that you would have a tendency to hold twice as much money on average. The doubling of your money holdings was the result of higher nominal income (which includes the effect of inflation), not the cause of it.

This is not the text to discuss whether this argument is correct. However, it offers an explanation why money growth is correlated to nominal income growth – as can be observed in the figures later in this chapter.

What About Government Spending?

The other key point to observe is that “money printing” is a useless description of how governments operate. Instead, if one wants to go down that route, one needs to take seriously how government financial operations are done. One way of summarising MMT points is to say that governments mainly interact with the private sector via spending and taxation, everything else (including interest rate policy) is a second order effect. I do not want to derail this text with a discussion of whether MMT is correct, but I will just take the MMT position as given for now.

If the government is not “printing money,” but rather spending, can it be inflationary? It is generally hard to find economists that disagree with that premise, although there are massive differences of opinion as to the mechanisms, and the quantitative effects.

My view is easy to state, and fits the observed data: what governments spend on matters, and so the inflationary impact of fiscal policy depends on the details of that policy. Saying “the relationship between inflation and fiscal policy is complicated” is not a satisfying answer – but being unsatisfying is not the same thing as “wrong.” That discussion is also being deferred to another text, but the reader is free to look at fiscal policy variables and attempt to relate them to inflation of they want to disagree. If they can spot a relationship, they then need to ask: why did everybody else miss this?

Concluding Remarks

It is an article of faith among Monetarist and hard money economists that money growth drives inflation, and “money printing” matters. However, this is a misguided faith. “Money printing” is not a useful description of how government finances work, and the relationship between money growth and inflation is complex.

References and Further Reading

  • I will just plug my book Understanding Government Finance here; there are further references therein.

Email subscription: Go to 

(c) Brian Romanchuk 2021


  1. MMT seems to take gleeful satisfaction from the fact that inflationary pressures have remained weak in the face of massive increases in central bank liabilities.

    I think what has to be borne in mind is that for the last forty years or so globalization has melded the world economy into one giant labour market effectively.

    In the west we buy goods that have been made with very cheap labour operating in very competitive markets.

    The second aspect that needs to be considered is the secular trend in the mal-distribution of income and wealth which has resulted in less specific consumption and greater flow of funds into financial assets.

    Both these trends I believe have outweighed the effects that changes in the level of money might have had.

    Henry Rech

    1. That makes no sense to me. If massive changes result in no observable effect in inflation in a number of countries, that tells us that money supply has almost no causal effect on inflation. All those structural factors were in place for decades, money supply growth was wildly different at different points, yet inflation was mainly stable (albeit with obvious problems starting in 2020)

  2. "If depositors rotate towards money market funds, banks will face liquidity outflows that need to be funded somehow. "

    Is that the case? The default mechanism is that for a deposit to be transferred to any other target deposit taking institution, that institution has to replace the original depositor in the source deposit taking institution (ie the target lends to the source). If that doesn't happen then the payment can't be made in the first place.

    Over time a central clearing house may develop which allows target deposit institutions to swap these things between each other, but for the system to clear somebody has to end up with them net.

    For me paper is just a deposit on better terms. There's little functional difference. Which ownership tag is on the paper has to be recorded just like which ownership tag is on the deposit.

    1. 1) There’s no guarantee that the target bank will lend back to the original bank. In any event, the funds needs to be replaced somehow.
      2) Sure, they’re all short-term liabilities. But bank deposits are in M1, commercial paper isn’t, hence M1 falls.

    2. "There’s no guarantee that the target bank will lend back to the original bank."

      There is if they want to accept the payment for their customer.
      Otherwise the balance sheets won't balance.

      The only other option is to refuse to accept the transfer. In which case the payment system falls apart and they have an unhappy customer expecting a payment.

      There's no funds being replaced. All that happens is that the ownership tags change. Netting off and central banks may obfuscate all this, but that's what is happening.

      Remember we're talking the default situation - without the clearing facilities of a central bank.

      What central banks do is ensure that temporary funding is always there to make the balance sheets balance and the process go more smoothly. Hence the lender of last resort function.

  3. "But bank deposits are in M1, commercial paper isn’t, hence M1 falls."

    Yes. All because of the definition of M1, which includes bank deposits that don't move, but not commercial paper or public deposits.

    Because reasons.

    I was reading some Lombard Research from 2000 today that talked about government borrowing from the banking sector. Also technically correct, but practically irrelevant.

    Half the problem is Humpty Dumpty terms.

    1. I am discussing the issue of "money printing" that other people invoke, and what it means. I am certainly not endorsing the concept, nor wanting to find a way to rescue it.

  4. The point that you are missing throughout this presentation is that when new money is created, new financial wealth is also created. Hence, "If households do not want to hold banknotes," please do not divert to a discussion of how paper money is destroyed while ignoring that households are careful to preserve their wealth no matter how household holdings of paper money may evolve.

    Whether paid by paper money or electronic money, taxes paid to government represent a transfer of wealth. The wealth transferred is only lost if the surveying economist wishes to close his eyes to the wealth ownership concatenation.

    1. I am discussing the monetary aggregates are they are commonly defined, in reference to discussions about money supply growth. It seems clear to me that I am not endorsing the use of monetary aggregates in analysis.

  5. Yup- I agree with you. What governments spend on matters. A lot.
    It is way past the point, since about 2009, where Monetarist ideas might have had something to stand on. 'Misguided faith' seems the nicest way to describe them accurately at this point. The more reasonable monetarists realize this also and say things like if the central bank is paying interest on reserves then that's not really money or some such nonsense.
    I always love how Mike Norman emphasizes it in his podcast -'They're printin money! Printin money!'
    Loved your podcast with Christian and Patricia.

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

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

  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.

  12. 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.