Mar06
Money & Macro1

What’s Radical about MMT?

An Attempt to Pin Down MMT on Public Finance

People from Scott Sumner to Paul Krugman have complained that MMT is hard to argue against because it’s hard to say what the argument even is. Much of the discussion has been a proxy for policy disagreement. So without any claim to originality, I’m going to attempt to break down MMT into two separable parts: an idiosyncratic but technically valid reinterpretation of orthodox public finance, and a backward causal story stemming from the state theory of money.1 And so far as the subject allows, I’ll try to do it without reference to particular policy debates.

1. MMT as a Reinterpretation of Public Finance

Orthodox public finance theory runs something like this: Imagine a government that wants to spend money but doesn’t have any. It therefore has to raise money. A government can raise money one of three ways: (1) taxation, (2) borrowing, or (3) if it controls the monetary authority, monetization. We can write this as G=T+B+M.

Orthodox public finance then defines “taxation” generically as “the government taking stuff from people”, and uses it as the paradigm for understanding the other two. Direct taxation is literally the government taking stuff from people. Borrowing is kind of like taking money from future generations, who have to bear the burden of repaying with interest. And monetization is kind of like indirectly taxing money balances, since monetization expands the money supply, and is therefore (ceteris paribus) inflationary. So taxation is thought of as the “normal” way to raise revenue, made up by borrowing what one can’t raise in the short term, and monetization if neither taxation or borrowing are feasible.

MMT, on the other hand, starts with a different baseline. Instead of imagining a totally unfunded government, imagine a government whose spending is fully funded by monetization, and then use monetization as the paradigm for understanding the other two. A government with control of the mint has the legal authority to spend money into existence, so government spending is not constrained by funding.

It’s important to note what kind of statement this is. So far it’s a legal claim about the authority of the government to issue money, plus an analytical choice to use G=M as a starting point rather than G=0 (There’s also an ontological claim about the nature of money snuck in here, but I’ll postpone that until the next section). So far no actual claims about the world have been made that any economist should disagree with.

Now, the orthodox economist will say, funding government spending entirely with monetization is foolish and leads immediately to hyperinflation. The MMT economist doesn’t disagree, any more than the orthodox economist would say that G=0 is justified just because he starts there analytically. But taxation takes on a somewhat different significance in light of a monetization paradigm. Rather than being the primary and paradigmatic source of government funding, taxation is a way of sucking money out of the economy to prevent the inflation that would result from a fully monetized budget. A sort of reverse monetization, in other words.

This is a sort of weird, but not invalid way of looking at the process of government finance. So far all we have is a different language. We can translate one-for-one between orthodox tax-baseline public finance and MMT monetization-baseline public finance. So a situation that looks like “too much monetization” from the perspective of an orthodox economist would look like “not enough taxation” from the perspective of an MMTer. Analytically at least, these are exactly equivalent.

That difference in emphasis does obscure a lot of important things, but it also illuminates some other important things. Starting from G=M and working backward to taxation might make certain large spending projects sound more feasible than they would for someone starting from G=0 and working forward, but in principle the two perspectives should find nothing to disagree about in a more rigorous budget analysis. I wouldn’t replace orthodox public finance with this perspective, but it might be a useful change of perspective for certain problems.

2. MMT as Backward Causation: The State Theory of Money

The problem with MMT that makes it seem so slippery is that it uses the tautologically true perspective on government accounting as the motte to its bailey of a much stronger claim about the nature of money, the ontological claim I mentioned earlier.

Statements like “government spending is unconstrained”, or “governments spend money into existence” can be understood as either trivially true accounting statements (as in the previous section for the first statement, and as Scott explains for the latter statement) or false substantive statements. And, usually, challenging the substantive claim results in a vigorous defense of the accounting identities, hence the perceived slipperiness.

In its substantive sense, “governments spend money into existence” means something like: government spending not only creates particular quantities of money under certain circumstances (i.e. debt monetization); it creates Money as an institution. Money is supposed to be in its essence a government debt which gains value through its receivability in taxes. Government spending is the backbone and indispensable precondition of a monetary economy. The money supply represents the cumulative deficit of the government, so the fiscal authority’s decisions on deficit and surplus determine (and do not merely influence on the margin, as in orthodox macro) aggregate demand. For a government to pay off its debt, therefore, is to destroy a monetary economy entirely.

This is the State Theory of Money, on which a great deal of ink has already been spilled. I won’t elaborate further here except to say that I regard it as an unwarranted generalization from some particular historical episodes to the nature of money in general, plus an illegitimate inference from that generalization to the nature of money as such. But it’s important to realize that this, and not the use of monetization as the paradigmatic source of public finance, is the radical component of MMT that puts a backwards causal arrow on every part of their mental macro model.

Conclusion

MMT is amorphous, partly intrinsically, but partly because of the dissimulation of its adherents on where the radical part lies. There are other problems too; problems that I’ll probably take up later, such as the use of a consolidated government budget and the disregard of apparently arbitrary social conventions, which – though more peripheral to the core theory, are perhaps more policy relevant. But at its core, MMT consists of two separable components. Acceptance of the formal validity of the accounting identities does not imply anything about the State Theory of Money one way or the other. Despite protestations to the contrary, orthodox macro does take these accounting identities just as seriously, even if it approaches them from a different (and just as analytically valid) starting point.

Footnotes

  1. There’s also a theory of private banking that I’ll leave to the side for now.

Tags

None

SHARE

Facebook Twitter Reddit StumbleUpon

One Comment

  • 1

    OH- Anarcho-Capitalist

    Mar 11, 2019 at 8:11 | Reply

    Nice post. I was unfamiliar with the motte/bailey argumentative style, so I learned something new.

    As regards MMT, the more I read about it from both its proponents and its critics, it does as you point out rely on both trivial arithmetic identities, much as Keynesianism does, and a radical misinterpretation of monetary history via the State Theory.

    The history of MMT is of course rooted in another long debunked monetary theory – Chartalism – which found acceptance in the more authoritarian types of governments such as state socialism.

    As such, this fringe crank theory should be denounced and dismissed as so much mindless blather…

Leave a Reply

More Content

About »

Hi, I'm C. Harwick, an economist in New York State with an interest in monetary theory, institutional evolution, and folk music.

Care to know more? Read on »

Twitter »

    • Mar2321:53
      Lane & Milesi-Ferrettt (2003): The current account doesn’t account for capital gains, so the CA may not actually be very informative for countries that invest in a lot of foreign equity. t.co/N1VXBYKvHf
    • Mar2021:26
      RT @Scientific_Bird: I looked a bit on James Damore’s (@JamesADamore) “memo” again. A thing that stands out to me is how extraordinarily careful he was. Emphasizing massive overlap, many caveats, constantly careful. Suggesting ways to increase female representation. I think the memo is pretty mild t.co/bYVilOjQjv


Design By Cameron Harwick Powered By Wordpress Hosted By Nearlyfreespeech No Copyrights