Governments don’t need taxes to finance public expenditure. They just really want them.

MMTers set great store by the idea that the government ushers money into existence when it spends, because if you understand how government finances actually work, you will know that the government can electronically credit accounts at the stroke of a keyboard.

Now I may be wrong, but this looks to me like another case of putting too much importance on the order in which things happen (this is the other case: Banks don’t need deposits to lend. They just really want them).

MMTers then say that because governments create money in this fashion when they spend, taxes are not needed to finance spending. Taxes are there to hoover money out of the economy, thereby preventing inflation that would otherwise occur if there was too much money injection and not enough money extraction.

I think this is nuts. Let me explain why through the medium of parable!

A society uses clams for money. People buy bread and pay school fees with clams. Then somebody invents the government which says: “we think it’s better if we take some clams from each of you, pay the teachers and provide free schooling.” Taxes are being used to finance government expenditure. Then the economy grows and they run short of clams. The government grants itself the monopoly to produce artificial clams. Each year they produce a number of new artificial clams judged to keep aggregate demand consistent with price stability. The government gets to spend more than it taxes, by those clams. Expenditure is now financed by taxation and seigniorage.

Then somebody invents the internet and smartphones, physical clams are abandoned and everything is done with virtual clams. People receive their wages in virtual clams, transfer virtual clams to bakers to buy bread and to the government to pay taxes. The government creates some new virtual claims each year, and makes transfers to teachers equal to the sum of taxes and seigniorage (it’s a very efficient government).

But then some bright spark in IT looks at the details of how the computerised virtual clam transfer system actually works, and notices that technically the government creates a new virtual clam whenever it transfers one to a teacher, and destroys a virtual clams when it receives a transfer from a taxpayer. Inspired by this observation, this person goes on to form a briefly successful political party which promises painless fiscal expansion because it turns out taxes are not actually needed to finance expenditure.

Financing spending

[I have updated this section, because some MMT academics have helpfully published a paper that tells us “actually the government creates clams, sorry money, whenever it spends.” Here is a blog about it.]

I get paid a salary, and it finances my spending. When I try to spend, the system checks that I have the funds on hand (including the existence of things like authorised overdraft arrangements) before I am allowed to spend. My income finances my expenditure, although I can borrow too.

The government has a rather more generous overdraft arrangement — it owns the Bank of England! It can raise income, it can borrow, but it can print money too. Should we now say that its income no longer finances its expenditure?

But hold on, there’s more! Actually the government “creates money when it spends”, and its income from taxes and borrowing is netted off afterwards. Does this detail about tell us anything about what’s possible, that we did not already know? We certainly do not learn that because the government “creates money when it spends” it does not need the taxes or loans. The system we have is one in which the government does tax and borrow. There is no reason to think the system would keep working if we kept the creating money but stopped the collecting money. We cannot infer from the technicalities of the government’s banking arrangements that we could sustain anything like the current level of government expenditure without taxing and borrowing. What matters is whether taxes and loans are necessary for the government to do what it wants. If they are, I don’t care if you want to say they are financing spending (I think they are) or doing something else.

So what?

Economists have always known governments can print money. That’s why (some of them) warn about the Weimar Republic or Zimbabwe. The government can finance expenditure in three ways: taxes, loans and printing money. My spending is constrained by two things (my income and my ability to borrow); government spending is constrained by three things.

If I wanted the government to spend more, I was always able to suggest cranking up the printing presses (and if anyone cares, after 2007 I often suggested the UK should indulge in some money-financed expenditure). As far as I know, mainstream economists only advocate a ban out outright monetary financing because they don’t trust politicians with the temptation.

The quantity of base money grows as the central bank’s balance sheet expands. When the CB creates reserves to buy government bonds that will be rolled-over in perpetuity, that’s what I am calling seigniorage or money printing [3]. Government expenditure is what, roughly 40% of GDP, and whilst I have not checked I am sure that the average annual net increase in the CB balance sheet has been a lot less than that. For the sake of argument, let’s say the typical split is 30% tax, 8% net borrowing and 2% money creation.

Taking as given a view on how far we are away from potential output, what changes if you think taxes don’t finance spending? Is there anything in the idea to suggest that holding spending constant we could get away with more money creation and less net borrowing and/or taxation?

MMT says the only constraint is real resource, or potential output. It claims that money and bonds are perfect substitutes, so that financing a 10% of GDP deficit by printing money is no more inflationary than doing so by borrowing.

I am not sure that is true, but let’s stick with taxes which MMT says don’t finance spending. What happens if you start to reduce taxes and increase money printing? If raising taxes removes money to tamp down inflation, doesn’t cutting taxes (holding spending constant) increase inflation? If yes, then the constraint on government spending spending is not just about real resources, where inflation occurs if you attempt to spend more than you can produce, but it’s also about the financing mix. Which is the same conclusion I’d reach if I thought taxes did finance spending (along with loans and seigniorage). If no … well I struggle to believe MMT really claims we could hold spending at 40% of GDP and set taxes and borrowing to zero, with no inflation problem.

The bottom line

I want to see big spending on climate change, more spending that benefits the least well-off, and fiscal policy used more aggressively in recessions. If MMT is how we will get that, I should be cheering it on [4]. Prominent MMTers are writing op-ed suggesting we can have whatever we want, without raising taxes. That may or may not be true, depending on how far we are beneath potential output (and how the financing mix affects inflation) [5]. But when asked how we can afford to spend more without taxing, MMTers usually invoke this supposed insight that taxes do not actually finance spending, and I am buggered if I can see what difference that makes.

Coda

My impression is that the debate around MMT is moving towards the more fundamental questions of whether the economic outcomes and policy decisions (interest rates policy, job guarantees etc.) are really possible, and less about this ‘we understand how modern monetary systems work’ stuff. Which is good, because to my taste at least, that stuff leads nowhere. I’d also like to see more thinking about unintended consequences of low interest rates, such as potentially increasing the market power of dominant firms.

[1] yes yes I am abstracting away from detail. Add the detail back if you like, it won’t change anything

[2] I started writing this post years ago but should have moved faster. I have been intellectually gazumped by Cullen Roche.

[3] There is another form of seigniorage which is to do with interest income earned by the CB. I think?

[4] It could go horribly wrong, if politicians try a large money-financed increase in expenditure and it turns out the economy does not have the slack then people will get hurt. But as MMTers will point out, people are getting hurt now, and so is the planet. There are risks on both sides.

[5] I can’t help feeling it’s not true, but maybe I’m like those mainstream economists who keep thinking we need to raise rates only to see employment expand but not inflation. I don’t know the extent to which demand will create its own supply, if there’s a big splurge on replacing existing kit with renewable tech.