Monday, June 29, 2020

Is fiat money to blame for the Iraq war, police brutality, and the war on drugs?

I often encounter memes claiming that fiat money is to blame for all sorts of government evils. Here is one example from Kraken bitcoin strategist spokesperson & bitcoin meme factory Pierre Rochard:

The rough idea behind this family of memes is that the Federal Reserve, the world's largest producer of "fiat" money (i.e. irredeemable banknotes), is responsible for financing all sorts of examples of government over-reach, say foreign invasions, police brutality, and the twin wars on terrorism and drugs. It does so by producing seigniorage, or profit, which it passes on to the state. Replace fiat-issuing central banks like the Fed with bitcoin or a gold standard, and seigniorage would cease to exist. With the government's purse strings having been cut, a relatively peaceful society would be the result.

This meme's premise is wrong. In practice, central bank seigniorage in both the U.S. and other developed nations is a very small part of overall  government revenues. And so even if fiat money were to be displaced, say by bitcoin or a gold standard, it wouldn't change the state's ability to fund the war on drugs and adventures in the Middle East.

Let's look at the U.S. Below are two charts showing how much income the Federal Reserve has contributed to the Federal government's overall receipts going back to 1950. (Beware. One chart relies on a regular axis, another a logarithmic axis. But they use the same data). The Fed's contribution has been steadily growing over time. In 2019, it sent about $53 billion to the Federal government.

You may be wondering how the Fed generated $53 billion in profit, or seigniorage, in 2019. Most of this income comes from issuing banknotes, or cash. For each $1 in banknotes that it issues to the public, the Fed holds an associated $1 of bonds in its vault. These bond have typically yielded 3-4% in interest. But the Fed only pays 0% interest to the owners of its banknotes. Which means that it gets to keep the entire 3-4% flow of bond interest for itself. It forwards this income to the Federal government at the end of the year.*

Seigniorage tends to grow over time. (But not always. Below I'll show how Sweden's seigniorage has been shrinking). The larger the quantity of banknotes that the public wants to own, the more interest-yielding bonds the Fed gets to hold, which means more seigniorage. In general, banknote demand increases with economic and population growth.

Interest rates are another big driver of seigniorage. If bond interest rates rise from 4% to 8%, the Fed earns more on the bonds it owns in its vault. Banknotes continue to yield 0% throughout, so the Fed keeps the entire windfall for itself (and ultimately for the Federal government).

By the way, a big driver of nominal interest rates is inflation. If inflation is expected to double, then bond owners will require twice the interest to compensate them for inflation risk. So inflation boosts seigniorage (because it boosts the interest rate that the Fed earns on the bonds in its vaults), and deflation hurts seigniorage (because it reduces interest rates). In the chart above, the one with the logarithmic axis, you can see how the Fed's seigniorage increased during the inflationary 1970s. It flatlined from the mid-1980s to the early early 2000s, which coincides with inflation subsiding.

US seigniorage is relatively small. In addition to enjoying revenues from the Federal Reserve, the U.S. Federal government also gets money from individual and corporate income taxes, social insurance and retirement receipts, excise taxes, duties, and more. Below I've charted the relative sizes of these contributions.

As you can see, the Fed's contribution (the grey line) is a rounding error.

Below is a chart showing what percentage of total government revenue is derived from the Fed.

In 2019 the Fed contributed just 1.5% of total U.S. Federal government receipts. This contribution has hovered between 1% to 3% over the last four decades. So the meme that fiat money abetted the Iraq War, the expansion of the police state, or the U.S.'s military industrial complex is mostly hyperbole.

What about other developed nations?

The Bank of Canada provided $1.2 billion in earnings to the Canadian Federal government in 2018. But the Federal government took in $313 billion in revenues that year, which means that the Bank contributed a tiny 0.4% fraction of total revenues. The reason for the big gap between the Bank of Canada's tiny 0.4% contribution and the Fed's 1.5% contribution is the global popularity of the US$100 bill. Canadian cash doesn't enjoy a big foreign market.

I mentioned Sweden earlier. Below is a chart of seigniorage earned by the Swedish central bank, the Riksbank.

Sweden is one of the only countries in the world where banknote ownership has been falling. This de-cashification is compounded by interest rates that have fallen close to 0%. Which means that the Riksbank's bond portfolio isn't earning as much as it used to. This combination has just decimated the Riksbank's seigniorage. In 2018 its seigniorage amounted to a paltry SEK 267 million (US$29 million). This is just 0.00003% of all Swedish central government receipts.

So in sum, central banks in places like the US, Canada, and Sweden are not a big source of government funding. If you want to stop governments from engaging in bad policies like the war on terror, the war on drugs, and foreign meddling, you've got to work within the system. Vote, send letters, go to protests. Sorry, but buying bitcoin or gold in the hope that it somehow defunds these activities by displacing the Fed is not a legitimate form of protest. It's a cop-out.

P.S. By the way, I am not saying that control of the nation's money supply hasn't been used to finance wars in the past. Obviously it has. Greenbacks helped pay for the Union's war against the Confederates. Henry VIII paid for his wars by dramatically reducing the supply of silver in the English coinage.

*The Fed enjoyed a big spike in seigniorage after the 2008 credit crisis. This is because it issued a bunch of deposits to bank (known as reserves) via quantitative easing. The Fed only had to pay 0.25% interest on these reserves, but the bonds that backed them were earning 2-3%. This QE-related income has declined as the Fed has unwound QE (since reversed) and long-term interest rates have declined.


  1. Not through seigniorage, but through Cantillon effects. Seignorage pre-dates fiat money. Well technically seigniorage could be said to be a very simple form a Cantillon effect, but this isn't what the issue is about. I'm baffled why you haven't familiarised yourself with the argument first before making a post.

    1. "I'm baffled why you..."

      Well, you get baffled by a lot of stuff, Peter ;)

      But to address your point, this post is specifically about the seigniorage claim. I could write another one addressing the Cantillon effects claim, if there is demand for it.

      For the record, I've written about Cantillon effects before:

  2. > But to address your point, this post is specifically about the seigniorage claim.

    I'm unaware of such a claim. Indeed as Pierre specifically says "inflationary", that should be a hint that it's not about seigniorage. Seigniorage could hypothetically exist in a deflationary system as well.

    1. The meme I'm attacking in my post specifically blames fiat money for facilitating wars, the police state, etc. And I've debunked that meme by showing that seigniorage is minuscule, and so it isn't responsible for funding the stuff for which it is blamed.

      I don't have a clue where you're going with your reply, but I'll bite.

      "Seigniorage could hypothetically exist in a deflationary system as well."

      As deflation increases, seigniorage falls to zero. Say that the price level is falling by around 2-3% per year. Then due to the Fisher effect, nominal interest rates on bonds would be ~0%. So the Fed's profit, or seigniorage would fall to zero (because it earns 0% on bonds while paying 0% on banknotes).

  3. But isnt fiat money "responsible" for the huge amount of State debt (as in Gov.debt+monetary base) without sustainability concerns?
    Isnt that debt a form of "revenue"?

    1. The amount of monetary base+gov.debt that could be issued in a gold standard would be inferior to the one with fiat money. (I said gold, but you could substitute with any fixed price anchor, I guess). So fiat money allows not only for seigniorage but as well for the "inflation tax" that depends on the expansion of monetary base+gov.debt.

    2. I don't see why governments would be able to issue more debt under on a CPI, or fiat, standard (say post-1969) than on a gold standard (pre-1969).

      In the US's case, I don't see much of an inflection point at 1971.

    3. The CPI standard is recent, no?
      And that chart has a distorted y axis, that will smooth the inflection.

    4. I used a logarithmic scale to make comparison over time possible.

    5. Yes, but the logarithmic scale allows to see if the growth rate has increased or not. But my point is about the level of debt. If a country (not the US, because the exorbitant privilege may distort) is on a fixed standard, it would have secular price level that would pin down the level of base+debt. If not, inflation target=2%, it will be able to keep on creating M+B to enable that inflation. What will it buy with that M+B? I don't know.. that's politics, but it may be the ones mentioned above.

  4. > JP

    In your calculations you did not take into account the issuer of the bonds. Interest past from the Fed to the government is not income for the government when it comes from the governments own coupon payments. I am thinking that in the classical case where all of the CB assets are government bonds the the seigniorage is less than small , it is zero.

    1. I suppose you could say that it's not income, it's just inter-governmental accounting.

      Let's put it differently. Say the Fed can either own 100% government bonds or 100% corporate bonds. If it holds 100% government bonds, then the government has to pay less interest to the private sector than the alternative.

      So yes, I still see both scenarios as generating seigniorage. Whether Fed holds all government bonds or all corporate bonds, real resources are being transferred to the government.

  5. I'd like to add an anecdote: when the central bank pulls a currency out of circulation, it records the full value of the currency that isn't returned to it by the cut off date. This could be a nice way to raise seniorage income from lost coins and banknotes, regardless of the interest rate.

  6. I think there are some questions about this post. At the beginning of 2009 the Federal Reserve had $826 billion in currency in public circulation and by the end of 2019 that amount had increased to $1,719 billion for an increase of $893 over those 11 years. During that period the Fed's securities held outright increased from $496 billion on 1/1/09 to $3,738 billion at the end of 2019 for a total increase of $3,242 billion.

    Most of the increase in the Fed's securities held outright were the result of the 2008-2009 market intervention programs and the follow on QE1,QE2, and QE3 programs. It looks like over 70% of what you call seigniorage has nothing to do with currency issued by the Fed.

    1. Good point. I discussed this briefly in the asterisk at the bottom of the post. Going back to the 1950s, regular banknote-based seigniorage seems to be in the range of 1-2%.