You wouldn’t have thought that this was a difficult or controversial question. But actually it’s both. It’s controversial because it’s more or less the central battleground for Modern Monetary Theory, and therefore an absolute magnet for bad tempered online debate. And it’s difficult for the same reason that a lot of things are difficult: the question looks like a reasonable one that should admit of a short answer, but that’s because all of the complexity and ambiguity is packed up into the fact that words are used in an ordinary language sense but the statements made with them need to be precise.
There are two possible courses of action in a situation like this. The first is to say either “very few people in this debate are confused about the actual facts, so this is really a dispute about semantics and since I am not in an iron lung, I have better things to do” or, depending on your circumstances, “despite being in this iron lung, I just got a new adult colouring book”, and then get on with your day. This is quite attractive, and I wouldn’t blame anyone for taking it.
But it’s not wholly satisfactory, because although few people actively involved in this controversy are confused, there are a lot of people in general who are confused about the relationship between taxes and spending and who think “taxes fund spending” is an uncontroversial and obviously true statement. And the process of un-confusing them is made a lot more difficult by the unclear semantics. That’s the point (when there is one) of semantic debates : if you have the semantics squared away then when people say dumb or contradictory things then they sound dumb and contradictory, but while the language is confused, they might sound profound or practical.
And so we come to the second course of action, which is “ENDLESS SQUABBLING”. Hurray!
I’ll start by just noting a few principles which I myself regard as uncontroversial, stupidly. First we have what might be called the “Very Weak-Form Laffer Curve” or less annoyingly the “Blood/Stone Theorem”. This is just the statement that at any given time (italics there to indicate some foreshadowing of a point to come in three paragraphs’ time), there is a maximum possible output of the economy, and if the economy can’t produce more than this, you can’t take more tax than this either. Furthermore, you can’t get round this constraint sneakily : even if the government finances some of its spending by means other than taxation, the total amount of government expenditure plus the total amount of private sector expenditure have to be lower than or equal to the productive capacity of the economy. Most importantly, if you try to print money to exceed this level, it will just result in inflation.
Caveats! What’s an arid semantic debate without caveats? I’m ignoring the overseas sector here; you can’t, in some important sense, get round this constraint by importing the difference. Or at least not sustainably : you could in some shorter periods, and the choice of period to consider will matter a lot. I will help myself here to an airy and misplaced mathematician’s “without loss of generality”.
The second important basic principle is the Keynes Possibility Principle : “Anything we can actually do, we can afford”. This is somewhere between a truism and a definition of the concept of “monetary sovereignty”, a concept that’s very important to MMT people. If you’re an imperial province, or if you’re otherwise subject to requirements to make payments overseas, then it’s not true; there are a bunch of things which you could do, but can’t afford because of your obligations. I am, however, for the purposes of this post, going to assert the principle, because I think it’s important to make clear that the semantics of “taxes fund / do not fund spending” have nothing to do with debt sustainability or borrowing constraints. That really is a separate and largely empirical issue which unfailingly confuses things when it’s brought up.
So do taxes fund spending? What does that even mean? The physical image is obviously of some coins going into a bag, and then being taken out again to spend on things. Abstracting several levels from that, we’re really talking about things like “is the level of taxation a constraint on the level of government spending?”, “does a change in spending need to be matched by a change in tax receipts?” and “can spending increase without raising taxes, sustainably?”. It’s clear when you look at these questions why people quickly start talking about debt and borrowing constraints, because the argument quickly switches to a debate about fiscal policy. But really it’s more fundamental than that, because “do taxes fund spending?” is a question about the underlying workings of the system, not about any particular state which the system might be in.
Semantic debates like this one are often driven by the fact that the words used describe a phenomenon which is itself different depending on how you look at it. In the case of the relationship between taxes and spending, and as foreshadowed above, the time period under consideration is very important. You can see this by taking two corner cases:
If your accounting period is “the whole of history, past present and future”, then of course taxes fund spending, because over that period, “taxes” and “government spending” are the same thing. They are two different names for the amount of economic output that is directed by the government sector rather than the private sector.
At the other corner, as your accounting period becomes infinitesimal, it gets more and more accurate to say that there is no relationship at all between tax receipts and spending : nobody at all thinks that receipts have to match up to spending in real time. On any given day, the government will spend some money, take in some tax receipts, and either issue or redeem some kind of IOU, either money or securities; you can construct an accounting system such that the three categories add up to zero, but you need to be clear that this is something you’re imposing on the transactions, not an insight into underlying economic reality.
So what happens at a more sensible, middling length accounting period? Let’s think about the actual system, in terms which I stupidly believe to be uncontroversial and universally accepted among people who know what they’re talking about, and then we can decide whether the reality we’ve described is accurately described by the phrase “taxes fund spending” or not.
Under a gold standard, of course, the only kind of IOUs that the government can issue are debt securities which have to be paid back. So, there’s an economically meaningful accounting period over which the taxes collected have to equal the money spent, adjusted for timing differences. In this case, the feeling is overwhelming that “taxes fund spending” is a good description of that relationship.
But, of course, this is the whole point of “Modern Monetary Theory” : as I regularly and apparently irritatingly point out, the first word of that phrase is an adverb modifying the adjective, not an adjective modifying the noun. It’s a theory of modern monetary systems, not a modern theory of monetary systems. That’s why people shouldn’t be surprised to find that there’s not much to it that wasn’t in Keynes. The word “modern” here means “not gold standard” and it is meant to describe a system in which the government sector is able to issue IOUs which don’t need to be paid back; money. That changes things a lot.
There are two qualitatively different states that the economy can be in from this point of view; full employment or less-than-full employment. The qualitative difference is that in the first case, increasing government sector spending must, by the blood/stone theorem, involve reducing private sector spending. Let’s look at this first because it’s a bit easier.
Assume that the economy is in a state of full employment at its potential output level P, and government spending G is (perhaps fortuitously) equal to tax revenue T. Because G=T, there is no need for the government to change the number of IOUs in circulation. Now we want to increase G.
There is a choice of methods here; either raise T by the same increment as G, or print some IOUs. But whichever you do, the blood/stone theorem will still apply. One way or the other, if you can’t produce more than P, then the increase in G is going to have to be matched by a decrease in private sector spending, which is the same thing as an increase in T. In fact, this might be one basis for a choice of accounting period : one economically meaningful time scale for this question would be to ask “over the period of time for which it is reasonable to take P as fixed, do taxes fund spending?”.
Right here you can see why this is a semantic debate. The two statements:
“raised taxes in order to bring in tax receipts to spend, issuing IOUs to cover the gap between the spending and the tax receipts”
and
“issued IOUs to spend, then raised taxes in order to prevent inflation”
are pretty clearly describing the same set of events. But the second one feels considerably less natural as a description; “the tax rise funded the spending” seems to describe the causal connection better.
Now let’s take a look at the more difficult, interesting and arguably relevant case. The actual output of the economy is some level less than P, and this is accounted for by private sector and government sector spending. The choice of methods here looks a lot more obvious, because there is no need to reduce private sector spending by increasing T; you can just print IOUs and spend them to increase G. If you don’t need to raise taxes in order to raise spending, then the taxes-fund-spending description feels pretty wrong.
Except that the two cases need to be made symmetric. Although misbegotten austerity is, annoyingly, a thing in the real world, we shouldn’t build it into the abstract argument that, in the example above, the policymaker is seemingly indifferent to unemployment. Let’s consider, without loss of generality, the case where the desired increment to government spending happens to equal the difference between current total output and P:
“Issue IOUs equal to the desired increment to G and spend the proceeds, thus bringing total output to P”
and
“Raise taxes to bring in tax receipts to cover the desired increment in government spending, then carry out monetary stimulus by issuing IOUs to raise private sector spending until output equals P”
The amount of IOUs that you would have to issue in the second sentence above would, of course, be equal to the tax increase; you’re increasing G by the difference between current output and P, so in steady state you need private sector spending unchanged in both cases. So the two sentences are, once more, describing fairly similar realities. But this time round it’s the taxes-fund-spending framing that seems unnatural; you’ve raised taxes and then rebated the proceeds right back to the taxpayers. (I have ignored the fact that the increase in government spending and output will itself tend to increase the tax take in most taxation systems; I have also ignored distributive effects. Without loss of generality, oh yeah).
I think what we’re establishing here is that “taxes fund spending” isn’t a useful phrase, and nor is its negation. Taxes, spending and monetary stimulus are three parts of a system which is regulated so as to allow public services to be provided in accordance with the Keynes Possibility Principle, while ensuring that the sum of government and private sector spending obeys the blood/stone constraint.
Which means that if you’re more worried about inflation and you’re thinking about the supply side, you’ll say that taxes fund spending, but if you’re worried about austerity and thinking about the demand side, you’ll say that taxes don’t fund spending. In aggregate.
“In aggregate” there is very important, because the most common context for people to argue about taxes funding spending is with respect to “pay-fors” for proposed new spending programs. And I think this is a case where the semantics cause people to say very wrong things indeed.
As we have seen above, it’s most defensible to say that an increment to spending needs to be “funded” (and therefore, that a new program needs to be justified with either a new tax or an equivalent reduction in existing programs) during an accounting period when the economy is at full employment. But this would mean that a “pay-for” promise or proposal could only really be made for a maximum of one accounting period. In the next period, P will be different; either the spending program no longer needs to be funded, or a new “pay-for” is needed whether the program is taken on or not.
Added to which, in many cases, the spending program and its associated taxation will themselves have an effect on the future output capacity of the economy. This is true of a lot more things than those which are classified as “investment” in the conventional system of accounts; education expenditure affects future productivity, prison spending affects future property destruction, even something like an Olympic Games is likely to have some effect on the future.
In other words, the “funding” decision is one that needs to be made separately in each accounting period, and at the level of the fiscal balance in aggregate. Paying for individual programs out of individual taxes is an analytical fallacy, and it’s not a harmless one as it’s deflationary (there’s an implicit assumption that the full-employment case is the relevant one and that the stimulus effect of a new program can be ignored or regarded as actively harmful). After all, when Toyota opens a new car factory, it’s very rare for them to be asked “how will you pay for it?”.
{ 50 comments }
steven t johnson 09.17.21 at 4:30 pm
“Issue IOUs equal to the desired increment to G and spend the proceeds, thus bringing total output to Pâ€
and
“Raise taxes to bring in tax receipts to cover the desired increment in government spending, then carry out monetary stimulus by issuing IOUs to raise private sector spending until output equals Pâ€
Either way, the assumption that spending increases output is what needs justifying.
Output depends upon investment, which is triggered in this system by the prospect of profits. Thus, changes in profit rates are the issue.
The relevance of the blood/stone principle is also highly questionable. The only relevant examples are war economies (or, in different ways, planned socialist economies.) Talk of taxes without talk of rationing and price controls in this context is too incomplete to be useful.
The blood/stone theorem is also suspect because of the difficulty in defining potential output, which is a moving target to say the lest. Further, the assumption that “inflation” is a necessarily bad thing does rather depend on a careful understanding of what “inflation” might be. I suspect that in practice it tends to mean 1)wage rises and b)lower return to creditors, property owners and such and c)lower profits.
It seems to me that the discussion might be clarified by seeing taxes as funding government debt. And seeing government debt in the modern economy as the underpinning of the financial instruments. Adam Smith told us (I thought) that the real wealth of nations was not gold but the more productive division of labor in a functioning economy, where returns to scale, network effects and the provision of physical capital changed “P,” generally in the long run, upwards. Taxes fund government debt because they are tickets to income from the real source of wealth, the economic system at work, not mere stockpiles of commodities.
The upshot is that austerity in taxation/spending, where one must give way to the other, is always a “rational” response, not because of nonsense about “P” in the blood/stone theorem, but in maintaining the financial structure. (In my eyes, for the sake of those who own those instruments.) Abstractly viewed apart from the question of who owns what, austerity may be deemed irrational, I suppose. But the conclusion that a correct policy, taken case by case for any given program, doesn’t follow. The need for a national government to follow policies that support a national currency and a national market are in the end overriding. I don’t think the conflict can be resolved within the system, but that’s me.
Peter Dorman 09.17.21 at 4:53 pm
This is a very clear, confusion-dispelling look at MMT and its verbal warfare with conventional “policy macroâ€. The only quibble I have is that the Blood/Stone principle is not quite so dichotomous in practice. The notion of potential GDP as a single number is a huge oversimplification (both collapsing a probability distribution into a point estimate and a heterogeneous collection of constraints into a single aggregate), and this complicates the accounting problem described in the OP.
But let’s pivot to a different aspect of the MMT catechism, that issuance of fiat money by a sovereign monetary authority is economically equivalent to issuance of loans. No. There is a difference between the economic roles of the existing stock of monetary vs loan assets. If financial markets are resistant to accepting an infusion of new bonds, that raises interest rates and reduces the price of all previously issued bonds in secondary markets. Changes in i may be consequential, but the Treasury can safely ignore devaluation of bonds issued in years past. If, however, new money is issued and financial markets resist, the entire stock of existing money is devalued, for instance via forex markets, and that is consequential. It matters that money rather than bonds operates as a means of exchange.
eg 09.17.21 at 5:09 pm
Thank you so much for this cogent attempt to grapple with this issue in a serious and pretty comprehensive way.
All I can add is to observe that the economy is almost NEVER operating at full capacity, outside of total war. I don’t know which side of the Atlantic you are writing from, Daniel, but if it’s the Western side a casual familiarity with the U6 gives us some idea of the current (and historical) amount of “slack” in employment, which signals a lot of unused capacity. The labor participation rate is also instructive here. I like to remind “there’s no free lunch” economists that there are an awful lot of lunches going unmade …
All of which is to say that “how are you going to pay for it” is not really a sensible question under current capacity utilization, and won’t be until considerably larger deficits and debts are incurred. Bizarrely we have imposed that constraint upon ourselves far, far short of the point where actually available real resources (labor and materials) make the question sensible. It strikes me as a sort of communal hallucination, at best, and a nefarious plot to hoodwink the public at worst.
dilbert dogbert 09.17.21 at 6:09 pm
Please take Niall Ferguson back. I looked at this from the Economist and screamed so loud it could be heard in London.
https://www.economist.com/by-invitation/2021/08/20/niall-ferguson-on-why-the-end-of-americas-empire-wont-be-peaceful
Business Insider tells the story:
https://www.businessinsider.com/niall-ferguson-bad-track-record-on-economics-keynes-comments-2013-5
Niall may want the US to reestablish the British Empire but I don’t
Mike Furlan 09.17.21 at 6:20 pm
Friedman “famously” said, but I can’t find the original source:
“Keep your eye on one thing and one thing only: how much government is spending, because that’s the true tax […] If you’re not paying for it in the form of explicit taxes, you’re paying for it indirectly in the form of inflation or in the form of borrowing. The thing you should keep your eye on is what government spends, and the real problem is to hold down government spending as a fraction of our income, and if you do that, you can stop worrying about the debt. “
Charlie W 09.17.21 at 6:58 pm
Find myself wondering that if there’s talk of accounting periods in here, then perhaps something about tax incidence and generational churn shouldn’t be in here as well. After all, people who kvetch about spending and taxes seem to be worried in the first place about taxes falling on them. Sometimes they pretend to be future minded with talk about ‘our grandchildren’ (but we don’t really take them seriously).
Matt 09.18.21 at 12:36 am
This is interesting and helpful, so thanks for it. A question, in hopes of increasing my understanding a bit more. I’ve been reading parts of the Cambridge Companion to Keynes in small bits of spare time recently. In the (very interesting!) chapter on “Keynes and Keynesianism”, by Bradley Bateman, he argues that Keynes’s own view was that it was necessary to make a distinction between the “ordinary” and the “capital” budget, and that it was highly desirable to balance the “ordinary” budget, but that it was not, in general, necessary to balance the “capital” budget (for reasons very much like that suggested by the example of a Toyota factory at the end of the post), and that one of the big mistakes being made by the UK Treasury at the time when he was working was in treating the “ordinary” and the “capital” budget as one thing, or the same thing. (There is also some discussion about the desirability of paying for things out of the “sinking fund”, but that seems to me to perhaps be more of a discussion of the particular UK institutions at the time, though I’m not 100% sure.) As I understand the article, Keynes thought that, in periods of less than full employment, government should increase investment by engaging in public works, and that this should not be thought of as a “deficit” any more than when a company engages in investment. And, the idea seemed to be that over time these activities would pay for themselves, just as Toyota hopes that, when it builds a factory, the output of the factory will (more than) pay for the cost of building and running the factory. For that reason, in a way that is similar to how it would be foolish for a company like Toyota to only fund investment out of cash in hand, it is foolish for a government, if there is less than full employment, to only engage in investment that it can fully fund from current tax receipts. (Bateman also thinks that later economists, both “Keynesians” and their opponents, misunderstood Keynes’s own view here. That’s not without some interest, but less interesting to me.)
What I’m interested in is how far this account is saying more or less the same thing that you are, and is connected to what’s right in MMT. If they are not closely related, that’s also useful to know (especially if you’re able to say, briefly, why.)
Seekonk 09.18.21 at 12:49 am
‘people shouldn’t be surprised to find that there’s not much to [MMT] that wasn’t in Keynes’
That’s my understanding as well. It’s also my understanding that even an autonomous government with a sound currency can’t untether spending from taxes**.
I recommend Brooklyn-based economist Doug Henwood’s piece on MMT. He wants to be supportive, but he’s skeptical. He thinks that the danger of inflation is real, and that big projects like a Green New Deal must be funded largely by taxation. Kinda long but clearly written: https://www.jacobinmag.com/2019/02/modern-monetary-theory-isnt-helping
(** which is a reason for governments to implement their programs by direct mobilization of resources rather than via taxation.)
Rapier 09.18.21 at 1:52 am
You lost me but let’s look at how the government pays for things. The amount of money the Treasury has to spend is contained with an account it has with the Federal Reserve. That is here, in Section 5, U.S. Treasury, General Account. As pf 9/15 that was $344bn.
https://www.federalreserve.gov/releases/h41/current/
The account is funded by payments made to the Treasury; taxes, fees, fines and money borrowed from the sale of bonds notes and bills. There is no theory about how this works. This is how it works. Now one does not need a theory to envision other ways of funding that account. All one needs is a method and probably laws and regulations about how to do this. One needs mechanisms, not theories.
So what are these methods? F theory, pardon my French
J-D 09.18.21 at 4:12 am
My personal inclination would be to begin the endless squabbling by asking ‘What difference does it make whether taxes fund spending?’ or, equivalently, ‘How does it matter?’ or, equivalently, ‘Why do you care?’
If the answer is, as it well might be, ‘People say that increasing spending without increasing taxes will have negative consequences’, then I might respond, ‘Well, let’s try to figure out whether that’s true; if we find out we can agree about the answer to that question, maybe it won’t matter whether we decide to say that taxes fund spending’.
If the answer it, as it also well might be, ‘I just want to understand how things work’, then I might respond, ‘Well, let’s try to figure out what the consequences are when spending is increased without increasing taxes; if we find out we can agree about the answer to that question, maybe it won’t matter whether we decide to say that taxes fund spending’.
I feel confident that if you’re looking for squabbling, you can produce plenty of it by this kind of approach.
Rapier 09.18.21 at 3:33 pm
The squabbling could be mostly eliminated if the MMT people would simply state how money is created now and how they propose to create it in a new “modern” monetary system.
So first things first. Would some MMT person tell us how money is created, and destroyed now?
Even here we are presented with a semantic minefield because no description of money creation is a theory. Well not a theory in any scientific or logical sense. Just as declaring Trump the winner isn’t a theory. There are only ‘theories’ about how it was done. The declaration that MMT is a theory muddles the whole thing into a mess. To wit, this article. Can’t we scrap the theory. Let’s here how money is created now and how you propose its creation in the future.
Glen Tomkins 09.18.21 at 5:30 pm
The Blood/Stone Theorem is only true even as a first approximation at a stage of history in which private enterprises are of such marginal productivity that they struggle, and still fall short, to accumulate enough capital to take advantage of all possible investment opportunities that increase the output of the economy. Aren’t we way past that stage of history? If there were enough investment opportunities out there that would actually increase the output of the economy if they found capital, wouldn’t we see them funded in preference to the nth order derivatives from the economy of goods and services, the bubbles and Ponzi schemes we see private individuals and enterprises direct so much of their spare cash towards?
The only stones here are the massive stones of the malefactors of great wealth who whine that govt spending stifles their investment in the economy, either by taxation or by increasing interest rates. The private sector isn’t sitting on a stone, it’s sitting on a blood bank.
MisterMr 09.19.21 at 7:39 am
@Rapier 5
According to MMT money is created when the government prints it and spends it (thus deficit spending), it is destroyed when the government taxes it away but doesn’t spend it (thus government surplus).
The government going into debt (emitting bonds) is counted usually as the same as printing money (a 100$ bond is treated as if it was a 100$ bill),though I believe the MMTers actually think that printing money is better than printing bonds.
The Fed adding to the government account is also counted as printing money.
I think they skip private creation of money, also known as banks making loans, and I think they are wrong at skipping it.
Zamfir 09.19.21 at 11:54 am
Rapier, I am not an MMR person, but I don’t think the answer to your question is a mystery? Governments can pay their bills by issuing new money, and declaring that the new money has equal legal standing to the money already in circulation. The central bank could simply declare that the treasury account is now 345 billion instead of 344 billion, and the government would have an extra billion to spend.
They follow a more complicated procedure in practice . The government issues a bond to the market and deposits the proceeds in the treasury account, then the central bank buys bonds from the market and pays with new money. The treasury even pays interest on those bonds, which eventually become profit of the central bank and are paid back to the treasury, etc.
This extra steps have their functions, but the underlying principle is no different from declaring the treasury account to be a billion higher.
Gorgonzola Petrovna 09.19.21 at 4:58 pm
It’s a global economy. Taxes, employment, government spending are attributes of a nation. What nations do is they compete for investment. There isn’t much else going on. MMT is missing the point; the game’s in who you tax, where you spend, and how orderly and carefully you weaken your currency.
Mark Pontin 09.19.21 at 10:45 pm
“That’s why people shouldn’t be surprised to find that there’s not much to it that wasn’t in Keynes.”
I’m not sure there’s much to MMT that wasn’t in chartalism in 1924.
https://en.wikipedia.org/wiki/Chartalism
Elites have long known that this is how fiat money systems truly work, of course; thus, more than $28 trillion to the world’s financial industry since the GFC in 2008. Still, one does admire the barefaced audacity of the continued telling of the Goebbels-level Big Lie:
‘Let us never forget this fundamental truth: the State has no source of money other than money which people earn themselves. If the State wishes to spend more it can do so only by borrowing your savings or by taxing you more. It is no good thinking that someone else will pay – that ‘someone else’ is you. There is no such thing as public money; there is only taxpayers’ money.’
Falsehoods from start to finish! What a piece of work that woman was.
RichardM 09.20.21 at 10:25 am
Thought experiment time. Space aliens show up, and offer the governments of Earth 100 trillion Rigelian credits in return for their vote in the space UN. As it turns out, Rigelian credits are the hardest of hard currencies, having maintained their current value within 1% for the last 5 million years. The aliens visibly and publicly convert those credits into local currencies before handing them over to governments. They don’t do anything else, like handover advanced technologies.
With their funding met by the aliens, do the governments of Earth still need to levy taxes? MMR would I think say yes; you still need the taxes to avoid inflation. The credits are useless as credits. Only when used to actually buy galactic consumer or capital goods do they do anything.
Conventional economics would I think say no, you can use the galactic foreign aid to drop the tax rate and nothing bad will happen.
rsm 09.20.21 at 6:48 pm
》The account is funded by payments made to the Treasury; taxes, fees, fines and money borrowed from the sale of bonds notes and bills.
Doesn’t Treasury chiefly sell bonds high and buy them back low? Especially when the Fed buys, because interest comes right back?
If inflation is noise as Fischer Black proposed in “Noise”, how can you use it to measure the mythical output gap?
John Quiggin 09.20.21 at 8:32 pm
RichardM MMT is typically discussed in terms of a closed economy. Once the Rigelians turn up, things change. The simplest way to get back to the original MMT story, is to treat Rigel+Earth as a unit, with the Rigelians as both central government and central bank. Earth is like a poor region getting federal assistance The Rigelians need to levy taxes to prevent inflation, but the regional government of Earth can just spend and not worry.
John Quiggin 09.20.21 at 8:34 pm
I’m assuming the Rigelian money can be used to buy consumer goods from the Rigelians. Otherwise, all they”ve done is print some Earth currency
John Quiggin 09.21.21 at 5:24 am
The other big point for lots of MMTers is the specific claim that there is no need to tax “the rich” (an undefined group). The underlying model is of someone like Scrooge McDuck with a swimming pool full of money. If Scrooge gets a tax bill, he just hands over some of the money, but doesn’t change his consumption or investment. So, there is no change in real activity.
RichardM 09.21.21 at 9:41 am
Well I think that’s very much the point; MMT and conventional Keynesian economics give the same results in the special case of a closed economy. It’s only in the real world, that contains things like foreign aid, reserve currencies and wars that the differences become apparent.
The obvious analogy is relativity, which is the same as Newtonian mechanics at slow speeds. The difference being that the MMT people don’t yet have, at least as I understand it, any really worked out and experimentally backed theory as to what actually does happen in that general case.
For example, if you talk about taxing the rich, you could base that discussion on the premise that there are no other currencies in which to store wealth, outside the jurisdiction of the relevant tax authorities. But that is not talking about the world as it is.
I don’t believe that anyone really knows what would happen if the US used some legal trick to simply fund something like medicare for all without either raising taxes or borrowing. But it would be an interesting experiment.
John Quiggin 09.21.21 at 10:09 am
We have plenty of experience of governments undertaking wartime expenditure beyond what can be funded with taxation and borrowing. The result, more or less invariably, is inflation.
In the standard presentation of MMT, it’s the need to control inflation, and not the need to balance the budget, that generates the requirement for taxation.
aubergine 09.21.21 at 10:12 am
I tracked down the Milton Friedman quote that Mike Furlan mentions above @5:
…to 13:44 in this video.
Now, note the words Friedman chooses: “that’s the true tax”; “you’re paying for it”; “the real problem is to hold down government spending”; “our income”. And, of course, “debt”.
This is semantics, I suppose, but it’s also metaphor, and metaphor can be powerful. Friedman is carefully tuning his language to encourage his audience to think of the government’s budget as a large-scale version of their own household budgets – “you’re paying for it”, “our income”, “the debt”. Nobody wants their household to owe too much debt, right? Of course you want to reduce spending so you don’t have to load yourself down with debt; that way, you’ll have more to spend in the future.
This framing naturally serves the purposes of right-wingers trying to cut government spending that benefits people they don’t like, so as to preserve the wealth and spending power of the people they do like.
(Keynesianism, unfortunately, buys into this framing to some extent with its talk of “deficits”, which sound bad, and “surpluses”, which sound like nice things to have.)
One of the best things about MMT is the way it unshackles economics from these kinds of metaphors and shifts the focus to actual flows of money. In particular, it makes it clear that policies which require spending to be linked to taxation aren’t dictated by economics, but rather reflect political decisions about the distribution of spending power.
MisterMr 09.21.21 at 11:49 am
I’ll just note that we have a lot of experience about what happens to governments that fund at least part of their expenses on continuous deficits, since basically no government balances its deficits since 1945.
This is happening in the real world, not in some theory.
Ashley Nanton 09.21.21 at 12:28 pm
That was mostly very clarifying, thanks. However, I am struggling to reconcile the idea that taxes fund spending over the whole of history, with the later point about capacity expanding investments. How can that be if the long run is a stringing together of short runs? I know you wanted to abstract from debt issue, but this would also seem to touch on whether debt is left to decline in nominal terms, versus taxed to be paid off (and I understand it’s mostly the former).
Rapier 09.21.21 at 1:08 pm
All money is created by bank loans. All of it. Loans create money. Banks create all money. An account is created and the money declared to exist in said account. Now enters double entry bookkeeping. The loan becomes an asset for the bank, and the deposit is a liability. The books balance. This scheme was first proposed in 1494, pretty much coincident with the rise of The Age of Reason.
Now MMT has it that some debts don’t exist. The asset doesn’t exist. Bye bye double entry bookkeeping. Kellyanne Conway was right. There are alternative facts. The Age or Reason is over. Okee Dokee.
On the street level no government in the world “issues” money. That is, pays bills with currency it prints. Such a thing could exist, but doesn’t. (Here this new money could not say Federal Reserve Note because it would be independent of the banking system) The maddening, well actually the crazy thing about MMT, is the insistence that governments now do “issue” currency. No they don’t. Which is why I harp on the need for MMT to explain the mechanisms by which they could. Which will never happen of course. Just like no MAGA fan will ever admit Biden won the election. Same dif really. They’ve got their theory and will always stick to it.
Hey, the Age of Reason wasn’t really all that great, looking back on it.
steven t johnson 09.21.21 at 3:07 pm
Again, it seems to me to be useless to talk about war economies without considering price controls/rationing.
On the general subject of the rare economy that is approaching maximal output, I’m pretty sure that any such economy develops serious bottlenecks. Physical capital is not “leets.” Physical capital isn’t produced/invested in increments guaranteed to produce a smoothly linear supply curve, even if “leets” may be, either. It is customary to shriek that socialist planning produces bottlenecks, of course. That’s like how bad harvests never have any relationship to famines in newly socialist countries.
The role of taxes in funding government debt is key to the value of money, in practice, if not economics departments. That’s why the Fed pumping money into banks and the stock market for a decade hasn’t caused general inflation, merely (!) asset inflation in FIRE. The whole economy backs up the currency but it can only be accessed in a guaranteed stream of income, which is to say, taxes. That’s why losing wars, military or economic, is the only sure road to hyperinflation. (Yes, there are special definitions that manage to discover the same phenomenon elsewhere, but conflating those cases with the others is I think confusing, if not confusionist/obscurantist.)
If it isn’t obvious, the above is based on putting more weight on economic history than economics. That may be judged not even wrong, but the generalizations from real world?
eg 09.21.21 at 7:14 pm
@John Quiggin #23
I would point out that the MMT money story begins with the imposition of a tax obligation denominated in the tokens over which the sovereign exercises a jealous monopoly of issue. So taxation is required from the outset, not just to control inflation; the latter is a function of getting the amount of taxation correct.
J-D 09.22.21 at 12:16 am
This brings me back to my earlier question. If there’s agreement about a requirement for taxation, what difference does it make what is (said to be) generating that requirement; or, equivalently, how does it matter; or, equivalently, why does anybody care?
Seekonk 09.22.21 at 3:44 am
Supporting public projects by borrowing and creating money has its limits because it does tend to be inflationary; but the good news is that virtually everyone recognizes this, so runaway inflation almost never happens.
Pro-austerity deficit hawks lack credibility because they invoke their balanced-budget mantra selectively — always against popular projects such as Medicare-for-all or a Green New Deal; never against right-wing schemes such as tax cuts for the rich, military spending, or Trump spending his weekends at Mar-a-Lago rather than at Camp David.
Zamfir 09.22.21 at 11:51 am
@ J-D, central bank independence is the difference, I guess. Most countries nowadays at least try to separate monetary policy and fiscal policy, and inflation is supposed to be part of the monetary side of that separation.
Afaict, MMT is mostly a policy advise to reverse that separation. Give the fisc more direct control over the money supply, in addition to taxation and bonds, and in return the fisc becomes responsible for inflation management. The central bank is then relegated to a lower executive role.
Dividend Power 09.22.21 at 1:27 pm
Interesting question. But spending rises whether taxes are low or are high at least at them federal government level. Debt makes up the difference. People argue that spending is inflationary but debt has grown multifold in the past 20 years and inflation has been low. Inflation has ticked up recently but that is due to the negative impact of COVID-19 on the supply chain and input costs.
MisterMr 09.22.21 at 8:24 pm
@Rapier 27
Pedant note: Apparently the first uses of double entry bookeeping that we have trace of is from 1299:
https://en.m.wikipedia.org/wiki/Double-entry_bookkeeping
J-D 09.23.21 at 12:56 am
If that’s correct, then it seems to me that there is a practical question, namely, ‘Does the level of interest rates have an effect on the rate of inflation?’* If I’m right about that, then I would ask, again: ‘If there is agreement about the answer to that question, what difference does it make how we choose to describe the mechanism through which interest rates might affect inflation, or, equivalently, how does it matter, or, equivalently, why should we care?’
If it’s agreed that interest rates do affect inflation, that doesn’t by itself settle the question of whether interest rates should be determined directly by government or by a central bank independent of government, which would still have to be settled at least partly on other grounds.
Richard Melvin 09.23.21 at 1:22 pm
@35 As a simple mathematical model they are the same, at least for closed economies; at that level the ‘debate’ is mostly just a matter of what you call the different variables involved.
When you take that model and apply control theory to it, that means acknowledging that the numbers involved have both known and unknown sources of noise and error. Hence basing the decision of what the tax rate needs to be on some actually measured metric for inflation is different from making that decision in a different way. Even if, by the theory, they ‘should’ be the same.
Same principle as turning right when you see the junction, versus turning right after proceeding for 2 hours at 30 km/h from a known starting point.
Zamfir 09.23.21 at 7:51 pm
I am on uncertain ground here, but I think MMT people prefer direct government investment as business cycle management, instead of indirect manipulation of private investment through interest rate manipulation.
Rapier 09.23.21 at 8:02 pm
Meanwhile; in fact MMT is here now. Here is how real world MMT is working at this very moment.
The Fed, the BOJ, the BOE and the ECB are monetizing the equivalent the total of their nations deficits. I say equivalent because the Fed purchases as many mortgage securities as it does Treasury securities. Then all the returned principal from when those securities are redeemed, they are immediately used to buy more. It is perpetual debt. Perpetual money in other words. You do get that don’t you?
This is modern in the sense that the very first rule of the Bank of England, the original Central Bank, was not to fund a significant portion of the governments debts. Well now that’s out the window , but please don’t mention it. It’s embarrassing. Actually Jay Powell nor any of them needs to be embarrassed. Everyone, everyone in the whole world wants more money. Money is popular. It’s going to solve all our problems is the general consensus. Still Jay is a little embarrassed by the whole thing, throwing out a cardinal rule of central banking. Inside he has to be ecstatic because banks now shape the entire world. Well except for those 21 million who support violence to reinstall Trump as president.
Well anyway more money will solve all our problems is what bankers and their friends say too, and everyone believes them. That they, the banks and only the banks, create money is besides the point I suppose. Perish the thought that people and institutions which create money would use this power for their own benefit. What a silly idea. Excuse me for broaching it.
I’m pretty sure, baring an unforeseen circumstance, they will print enough money to send Elonnauts s to Mars. Hooray.
J-D 09.23.21 at 11:27 pm
The question I asked in comment 35 was ‘Does the level of interest rates have an effect on the rate of inflation?’
The question with which this discussion began was ‘Do taxes fund spending?’
‘As a simple mathematical model they are the same’ is not a wrong answer to either of those questions; it is not any kind of answer at all to either of those questions. It could be an answer (perhaps right, perhaps wrong) to some other question, but I can’t tell what that question might be, because I don’t know what ‘they’ is referring to.
It seems as if somehow Richard Melvin has read what I have written and understood it as meaning something completely different. I guess I could have expressed myself more clearly and would appreciate any suggestions about how I could do so.
From the reverse perspective, I know that I have not misunderstood Richard Melvin’s meaning because I don’t understand Richard Melvin’s comment at all. It doesn’t seem to be about the same topic as my comment, but I can’t be sure because I’m not sure what it’s about at all. Maybe this is just because I am obtuse, although I have to hope not.
eg 09.23.21 at 11:30 pm
@J-D #35
The effect of interest rates on inflation is contested — not that there is an effect, but rather what that effect might be. I have heard Mosler claim that the interest rate is (or eventually over time becomes) the inflation rate.
John Quiggin 09.24.21 at 2:04 am
MisterMr @25 The value of existing debt, expressed as a percentage of GDP, falls as (the money value of) GDP grows. If the annual deficit is too small to offset this effect, debt/GDP will decline, as happened from 1945 to the late 1960s.
Peter T 09.24.21 at 2:13 am
The history does not support the proposition that interest rates are connected in any direct way to inflation. The history also does not support the proposition that government spending (money supply) in itself leads to inflation. For every cherry-picked example supporting either there is another in contradiction.
Pedant note for Rapier: the very first act of the Bank of England was to raise a huge loan on behalf of the government.
J-D 09.24.21 at 3:25 am
That’s a needlessly patronising way of expressing yourself. Do you not get that?
J-D 09.24.21 at 3:37 am
I did not know this; thank you for the education.
I observed in my previous comment that whether interest rates have an effect on the rate of inflation is a question with practical implications; perhaps I should have added explicitly that I get that this is also true of questions about what that effect might be and how it might vary. However, I still don’t get the relationship (if any) between these questions and the question of whether taxes fund spending.
Also, it’s not obvious how the questions about the effect of interest rates on inflation have implications for who should have control over interest rates.
MisterMr 09.24.21 at 4:59 pm
@John Quiggin 41
I agree, I didn’t mean to imply that the only possible outcome is increased debt/gdp.
For various reasons though I think that happened because of a political choice of high inflation and high marginal taxes (that IMO was better than what we have now). The high marginal taxes are there to reduce the saving propensity, IMHO.
@Rapier 38
Private banks always had the habit to overlend in boom times, this is why the crisis of 1929 (and many other smaller crises) happened.
The government literally printing money is an innovation that came to be exactly to counteract the boom/bust cycle of private lending.
By the way I recently read an article of Wray (one of the authors of MMT) that is very against quantitative easing and says that the MMT idea is that the government should print INSTEAD OF the banks:
https://www.nakedcapitalism.com/2018/10/randy-wray-modern-monetary-theory-came-mmt-include-mmt.html
RichardM 09.25.21 at 10:39 am
@39:
Apologies, the point I was replying to was your 30, not your 35:
I’ll try to explain my reply in a bit more detail, as it seems the first attempt didn’t really hit.
All car manufacturers agree on the basic kinetics describing vehicle motion, how friction works mathematically, and so on in a lot more accuracy, detail and precision that any economic model.
But still, there are different ways of implementing cruise control, including not doing so and leaving it up to the intuition of the driver.
https://auto.howstuffworks.com/cruise-control.htm
And ‘non-adaptive’ cruise control doesn’t do that; not because the relevant engineers disagree with F = ma, but because it takes different measurements at different times, processes them differently, and acts differently based on the results.
I am positing that the statements MMT makes on economics topics like taxes, inflation, spending and interest rates, are, where they differ from conventional economic theory, best understood as statements about economic control theory.
Specifically, you could have an ‘inflation sensor’ coupled to a ‘tax actuator’ by a suitable control law, and entirely independent of that a ‘poverty sensor’ tied to a ‘welfare actuator’. And if you engineered this correctly, you might well have a safe system that guaranteed a smooth and safe ride for everyone onboard.
Or maybe MMT is wrong, and you would have something that oscillates wildly before spontaneously disassembling. And hypothetically the only way to fix it would be to add a ‘deficit sensor’ that tied to a ‘government shutdown actuator’.
So finally to specifically answer the question as asked. We are riders in the car; we do very much care which is the case.
J-D 09.26.21 at 9:15 am
I appreciate your making the attempt at clarification, and I hope it won’t seem ungrateful if I respond that your further attempt also doesn’t hit: I found it almost entirely opaque, and am at as much of a loss as ever about how your remarks relate to the discussion that preceded them. I get no sense of what your answers would be to the questions I am interested in having answered. Perhaps this is because I have failed to make sufficiently clear what those questions are?
Richard M 09.26.21 at 10:23 pm
I’ll try one more time.
Say there was a new make of automobile on the market, Mac’s Mega Trucks. They promised a smoother ride, lower fuel consumption and greater crash protection.
Other people were, reasonably, skeptical about the claims of that company. The way the expressed their skepticism was to ask questions about the laws of physics: ‘hey Mr Mac, if there is greater friction, does your car not slow down?’
The correct, if impolite, reply to such a question is ‘what misconception drives your feeling that that is a relevant question to ask?’
‘do taxes fund spending?’ is a question of the same nature as ‘does pressing the leftmost pedal make the car go faster?’ You can reverse engineer a particular car and find out, but that tells you nothing other than someone once though it made sense.
The better question is ‘should taxes fund spending?’ Because the answer for that can only be found in the domain of ‘what would happen once you have built a system that works that way?’.
Which, as it happens is a well-studied mathematical topic called ‘control theory’, which automotive engineers happen to make heavy use of. When they do a good job, the result is a car that does handle better, uses less fuel and so on than last year’s model.
Control theory has sometime been applied to microeconomics, mostly by Marxists in the 70s. But, as far as I know, no-one has made a serious attempt to see what it would have to say about modern areas of controversy in macroeconomics.
John Quiggin 09.27.21 at 12:19 am
Richard M @48: Control theory is used all the time in mainstream macroeconomics. The central macro controversy is whether fiscal and monetary policy can and should stabilize the macroeconomy, and this controversy is most naturally framed in terms of control theory.
I was drilled in all kinds of feedback loops in my mainstream macro courses back in the 1970s. OTOH, although I’ve spent some time looking at Marxist econ, I never ran across control theory. No one sees the whole picture, it seems.
J-D 09.27.21 at 12:37 am
No, it isn’t.
‘What happens when you vary the level of taxation?’ is a question of the same nature as ‘What happens when you press on the leftmost pedal of the car?’
‘Does lowering the level of taxation increase the rate of inflation?’ is a question of the same nature as ‘Does pressing the leftmost pedal make the car go faster?’
‘Do taxes fund spending?’ is a question of the same nature as … well, I don’t know. That’s what I’m puzzling over, pretty much. I know it’s not the same kind of question as ‘Does pressing the leftmost pedal make the car go faster?’, but I don’t know what kind of question it is.
Something else I have no clear idea about is what the difference would be between a system that works that way and a system that works some other way.
If you were to say to me ‘You would understand what I am saying if you understood mathematical control theory’, then I would respond ‘That doesn’t help me, because I don’t understand mathematical control theory’.
Comments on this entry are closed.