Thread by @besttrousers

Here's an attempt to illustrate the core differences between MMT and mainstream economics in a DAG (directed acyclic graph) /1

"DAG" = Directed acyclic graph en.wikipedia.org/wiki/Directed_…

It's a standard (though, uh, not required cc @causalinf) way of representing causal chains in economics and other fields /2

We can start with the government budget constraint:

G + (1+i)B = T + B(t+1) + [M(t+1)-M(t)]

/3

The government spends money and pays debts. It finances this through a combination of 1.) taxes 2.) New debt 3.) New money. /4

I *think* that MMT and mainstream macro agree on this accounting identity. Sometimes it's unclear from the rhetoric. /5

I think the difference is the behavioral relationship *between* the variables. That's why I think a DAG might be illustrative! /6

Here's how I think the mainstream folks represent the relationship /7

Congress sets G and T. If G>T, they finance the additional spending through selling new bonds. /8

The @federalreserve sees the fiscal stance, then sets an interest rate i (pursuing their goal of price stability and full employment). /9

The interest rate determines both the cost of servicing the debt, and the change in inflation /10

The important thing here is that the monetary authority moves last - they can see what Congress does and react /11

Here's my best guess of the DAG for MMT /12

(Quotes are from Fullwiler "Interest Rates and Fiscal Sustainability" papers.ssrn.com/sol3/papers.cf…. Thanks to @geerussell for the pointer) /13

In MMT, spending, taxes and bond sales are all independent decisions (instead of bond sales being determined by G-T /14

Instead, G-T determines the change in the monetary base /15

The change in the monetary base creates financial assets for the non-government sector (private savings). That determines the interest rate on government bonds /16

In MMT the @federalreserve doesn't make an independant choice - it's constrained by the pre-determined decisions made by Congress and the Treasury. /17

The accounting identity is the same; it's which actors get to make active choices, and which are constrained by the accounting identity itself /18

That's why MMT sugests we should use G and T to deal with the business cycle, /END