Friday, October 25, 2019

A free market case for CBDC?

Central bank digital currency, or CBDC, is a form of highly-liquid digital debt that most governments have, till now, held back from issuing. But there is a growing push to change this. Free market economists are generally not big fans of CBDC. They see it as government encroachment on the banking sector.

In this post I'm going to push back on the free market consensus.

(This post was inspired after reading posts by Tyler Cowen and Scott Sumner).

Look, we're always going to have a government. Right? And that government is going to have to raise funds somehow in order to keep the lights on. The question is, how? Should it issue 30-day Treasury bills? Fifty-year bonds? Perpetual debt? Paper currency? Why not issue currency-ish debt instruments in digital form?

Let's start with a parable. Imagine a world in which the government has only ever issued 30-year bonds. But next month it wants to shift some of its borrowing from the 30-year bond range to the 10-year range. Government officials believe that this will reduce the government's interest costs and diversify government sources of funding.

Seems like a good idea, no?

But wait. The grocery industry has historically relied on funding itself with 10-year bonds. Till now, it hasn't had to compete with the government for the attention ten-year bond investors. Grocery store owners are furious over the impending decision. We could have difficulties funding ourselves! they fret. We might have to cut back on selling food!

Meanwhile, the restaurant industry in our imaginary world prefers to fund itself by issuing 30-year bonds. If the government raises more money in the 10-year end of the debt market and less in the 30-year end of the spectrum, restaurants will face less competition for investor attention. Go for it! say restaurant owners.

Which sector should the government choose to favor, grocery stores or restaurants? The choice seems entirely arbitrary. Government shouldn't be picking winners or losers, right? Civil servants should choose the most cost-effective form of financing.

The same argument goes for CBDC.

Bonds, bills, and CBDC are all just forms of transferable government debt.* But instead of having a fixed maturity like a bond, CBDC never matures. And whereas the interest rate on a bond is fixed and its price floats, the interest rate on CBDC is periodically adjusted while its price is fixed to $1. Either way, the government can use these instruments for funding projects and investments.

(For the rest of this post I'll use the terms CBDC and fixed-value floating-rate perpetual debt interchangeably.)

For whatever reason, modern governments choose not to fund themselves in the fixed-value floating-rate corner of the debt market.** No industry benefits more from this than banks. Individuals and businesses who want to buy fixed-value floating-rate perpetual debt have only one option available to them: bank-issued deposits. Regulations prevent all other industries from participating in this end of the debt market. So these non-banks have to turn to the 3-month to 30-year segment of the debt market where they must face the full brunt of government competition.

The presence of government competition means that non-banks' funding costs will be more onerous than otherwise. Conversely, banks' funding costs will be less onerous given a lack of government competition.

I don't see any compelling reason for why the government should avoid one end of the debt market and, in the process, favor the banking industry over other industries. I mean, if the government can cost-effectively issue CBDC in a way that reduces its overall interest obligations, then that's a win for taxpayers, no? It shouldn't go with an option that hurts taxpayers because it wants to help out a certain sector, should it?

The argument could be made that the banking industry is far more important than other industries because it does a lot of lending, and if lending slows then everyone loses. 

If the banking sector really deserves to be subsidized, why doesn't the government just pay the subsidy in a more transparent way, say by taking money directly from the pockets of individuals and non-banks and giving it to banks? 

Also, banks aren't the economy's only lenders. There are many non-bank lenders too. Sure, if a government were to issue CBDC, banks would now face more competition in the fixed-price floating-rate corner of the debt market, and perhaps would choose to lend less. But at the same time the government would be issuing less 30-year bonds, or 10-year bonds, or treasury bills. Non-bank lenders that issue debt in these ends of the debt market would face less competition than before, and might lend more.

In the end, it's a wash. One industry's loss is another's gain.

So let governments issue CBDC and compete for the attention of the fixed-price floating-rate investor, just like they already compete for the attention of the 30-year bond investor. This would remove an inefficient distortion, namely a subsidy to banks and a penalty on non-banks. This seems to be the free market position, no?

*It could be argued that one type of debt is a currency, and can be transferred from you to me, while the other isn't. But I don't buy that. Both types of debt are liquid. They can be bought and sold on exchanges. Or they can be transferred bilaterally. With bonds, a bilateral transfer can be conducted by conveying an old style physical bearer bonds, or by transferring a bond to a recipient using Treasury Direct.

**The government does issue banknotes, which are sort of like perpetual floating-rate debt, where the decision has been made to keep the rate at 0%. And it does issue reserves to the banking sector. But the quantity of banknotes and reserves is quite small relative to overall government borrowing.


  1. The British government already issues CBDC, and has done so for decades, via the British government run savings bank, “National Savings and Investments”. I assume some other countries do likewise, but I’m not sure which ones.

    NSI does not offer “absolutely instant access” accounts, nor check books or debit or credit cards, but money in their so called instant access accounts can be got out in about 24 hours and transferred to a normal commercial bank, from where the money obviously can be transferred to whoever the relevant account holder wishes.

    Next, I don’t buy the argument that “The argument could be made that the banking industry is far more important than other industries because it does a lot of lending, and if lending slows then everyone loses.”

    Clearly if lending slows ALL ELSE EQUAL, then “everyone loses” (i.e. GDP falls). But of course there is no need for all else to be equal: that is, any loss in demand and GDP stemming from less lending can be made good by standard stimulatory measures. The net result would be less debt funded economic activity and more non debt funded activity. Given the VAST EXPANSION in banks over the last 50 years or so relative to GDP (a ten fold expansion in the UK) and the very dubious nature of some bank activities, it is very likely that a CONTRACTION in debt based activity and a rise in non debt based activity would be beneficial.

    Finally I suggest there is another bank subsidy, closely related to the content of the above article, which the above article missed. It’s this. If I lend $X to a bank by depositing $X at a bank, that liability of the bank is backed by taxpayers, i.e. by deposit insurance, not to mention trillion dollar loans to commercial banks at near zero rate of interest when banks are in trouble – nice being backed by an insurer with an infinitely deep pocket, isn't it? In contrast, if I lend $X to a non bank corporation, e.g. by buying some of its bonds, there is no taxpayer backing for me! There is no excuse whatever for that preferential treatment for money lenders (aka banks).

    Full reserve banking disposes of the latter anomaly. That is, under FR, those who want total safety can have it: their money is lodged at the central bank, with commercial banks acting as agents for the central bank. But little or no interest is earned on that money.

    In contrast, where depositors want their money loaned on under FR, their money goes into a special investment account (effectively a mutual fund), where depositor / savers carry the ENTIRE risk involved. I.e. there is very explicitly no taxpayer backing for that money. The latter “two account” system has effectively already been adopted by the Money Market Mutual Fund industry in the US.

    1. Good points, Ralph. I had forgotten about the NS&I. There's also postal banks, which are an old type of CBDC.

      You won't get any arguments from me about deposit insurance. Mind you, full reserve banking isn't the only way to solve the problem. Fractional reserve banking without deposit insurance also works. Over here in Canada our banking system has spent the majority of its history without deposit insurance, and it did fine.

  2. I'm genuinely confused here: do you believe that taxes can finance government spending, or do you not? If you treat currency as a debt instrument, then I don't see how you can logically argue also that the government can 'spend taxes': if the currency is a debt, then when that debt is returned to the government, it is redeemed. The government can't 'hold' its own debt any more than a bank can hold bank deposits at itself.

    So when you say things like "that's a win for taxpayers," what do you mean by that? Do you mean that the government will be spending less of their money on interest? I think that's entirely inconsistent with the premise of this article. On the other hand, if you mean that taxpayers must have their consumption reduced in order to free up resources for bondholders to consume out of interest without inflation, then this should only be true at full employment, not true in general.

    1. "Do you mean that the government will be spending less of their money on interest?"

      Yes, that's what I mean. Perhaps the government can issue a 5-year bond at 5%, and CBDC at 2%. Add in an extra 2% in other costs for managing the CBDC network. So it can issue a little less 5% debt and a little more 4% debt and reduce its overall financing costs.

    2. But my question is about whether or not you're saying that the government "spends their money"; as in, is the money the government spends the same money that it previously collected from taxpayers? Does the government "spend taxes"? It sounds like you're saying that yes it does, but I think that is logically inconsistent with treating fiat money as a debt instrument.

      And if it doesn't, then why should we care about financing costs when the economy is at less than full employment? The only reason I can think of is inequality.

    3. "Does the government "spend taxes"?"

      I'm not sure why his question is relevant, as I haven't said anything about taxes in this post. (Yes, I have used the term taxpayers, which is another way of saying citizens. If you prefer, you can reread the entire post and substitute in citizens in the place of taxpayers).

      All I'm saying is that if the government wants to fund some sort of project, and it typically uses 30-year bonds to do so, it may be cheaper to issue digital deposits instead. And this is good for citizens.

    4. It's not especially relevant to your point in this post, I'm just trying to tease out your view on the matter. It is quite important in general, for reasons, eg. covered here: Also relevant is the work of Camille Walsh, documenting how "taxpayer solidarity" has played a role in dramatically exacerbating school inequality.

      It seems philosophical (and it is), but it ultimately makes a great deal of difference to people's lives :)

    5. Ok, fair enough.

      "But my question is about whether or not you're saying that the government "spends their money"; as in, is the money the government spends the same money that it previously collected from taxpayers? Does the government "spend taxes"?"

      If an entity (ie the government) issues a debt, you are right that it extinguishes that debt when it takes it back in taxes.

      But if one branch of the government, the central bank, issues a debt, the other branches, the executive or the judiciary, can't cancel it. They use it in the same way that Walmart uses it.

      So I guess it depends on how you are defining the term "government".