Proof of X

In casual talks with Douglas MMT Trader questions about what constitutes a “proof of MMT” or a disproof arose.

New Zealand is an MMT Economy

We already know, from the Banking Acts and Public Finance Acts that New Zealand is an MMT monetary economy.

Just thought I’d mention this upfront. Although the legal statutes only prove half the case. We’d also need the accounting rules in the software and public accounts to obey the sectoral balance identities. Since the software is not open-source, we have to infer from the sectoral balances published. It turns out New Zealand somehow magically obeys the accounting rules! Imagine that? What on earth could be going on with our gold miners?

TODO: I had some old Three Sector Balance data for NZ, but need to 
update *live* on a dashboard using the BIS/FRED data, but was missing the 
private credit balance code. I think Douglas had it.

Research Objectives

The DougBot project seeks to use MMT analytics (balance sheets and fiscal flow essentially) together with neural network algorithms to predict forward prices on the more macroeconomic sensitive indices, like the SPX.

The motive does not need to be to make money ahead of the average market returns, but (we think) more importantly to show the monetary system, which underlies the financial sector, is an MMT system, even if the traders do not like it! So an educational project.

To get such “proof of MMT” we’d be also looking to predict (or retrodict, which is just as valid) movements in unemployment and tax flows. For this The Sequence of MMT is getting tested.

Do governments issue currency effectively first, tax it away later, or the other way around? Just looking at balance sheets from a single starting point in time today cannot ever prove this Sequence, because you are not using the appropriate initial conditions. (Who first emitted the first currency records?)

The latter question could be considered “lost in the sands of time” if you think Benjamin Franklin’s or Susan B Anthony’s mug-shots have nothing to do with “government”. (See, we are prepared to give Austrian Kindergartners a lot of rope.)

To test a sequence that has cyclical components that has no discernable $t=0$ stopwatch start, you need a method for determining differential motions within the cycle.

This is extreme sort of “proof” — like saying to some joker, “Hey, suppose you are right that anti-gravity inertia free propulsion exists. Now let’s see if that shadowy thing yonder that looks like a shadow, is truly accelerating instantaneously and is not a shadow…”

While this could be construed as violating Lerner’s Law (never concede the principle) this is ok, because we’re only doing it as an internal part of our project. In public debate we’d never concede the principle. It’s always just a Null Hypothesis: “suppose New Keynesian assumptions are true.”

Or, if it suits your data better: “assume MMT operations are the reality.” There are then some things you can do to test these hypotheses.

Testing the Sequence

Imagine then we are total idiots and have no thought of looking at the legal system, how could we show the MMT Sequence is the reality, rather than the alternative Null Hypothesis of pure endogenous money (government has to borrow or tax)?

One way is to examine the software of the troika of Treasury – Central Bank — Money Market operations when the government spends. Randall Wray and others have already done this, and they show governments do in fact issue or create new currency when they spend. There is no borrowing or taxing involved.

Incidentally, Richard Werner did the analogous thing for private bank credit creation — the software proves banks do not loan deposits nor reserves, they just issue money by keystrokes (in return for a promissory note to repay — every credit matches a debt).

However, this is too smart alecky for us. Suppose we were truly dumb Austrian Kindergartners and refused to regard the Central Bank as a branch of government.

Well, in fact Randall Wray’s balance sheet analysis still shows that the money is issued by the Parliament/Congress, not the Central Bankers. The Central Bankers are just scorekeepers.

Again, too smart aleck by far. We might want to be very knuckleheaded and even pretend we never read Adam Smith’s observations about “the Colonies” emitting “paper money” and then “burning it upon receipt”.

OK, so now I think we would be backed into the toughest corner to “prove MMT”. Finally. The only resort such knuckleheads would have is to look at impulses during the fiscal cycles.

The shorter time interval has to be found from,

• (step-$1$) government spend, to (step-$2$) real output growth,
      and then,
• also the shorter time lag must be from ($1^\prime$) accelerated increasing general tax returns to ($2^\prime$) downward price and employment movements.
      and,
• also the shorter time lag must be from ($1^{\prime\prime}$) increasing employment to increasing rate of FICA tax returns then a lag to ($2^{\prime\prime}$) stalled price and downward employment or stalled wage movements.
• also the shorter time lag must be from ($1^{\prime\prime}$) increasing government spending to ($2^{\prime\prime}$) higher savings rates for households and firms (not the other way around).

The third sequence recognizes FICA tax is regressive, it penalizes work, and so is a major component of driving the small cycle of wages against profits (the predator-prey mini cycles in macroeconomics). The price tends to never go downwards too much, instead the wage bill will tend to be what suffers if tax return accelerates.

I mean this is really tying our hands behind our backs and nailing our shoelaces to the floor boards level of stupidity. But we are prepared to do it for all our Austrian and Neoclassical Nursery School friends.

Note that the requirement for the “shorter lag time” means we cannot go beyond a fiscal cycle, otherwise the order of monetary operations (and hence their effects) can get thoroughly confounded, and the data can be manipulated to appear in any arbitrary sequence within the cycle.

If the sequence is, $$ A \rightarrow B \rightarrow C \rightarrow D $$ then we seek shorter time lags $\Delta t(A,B)$ than $\Delta t(A,C)$ (otherwise reject the Null Hypothesis). If the actual dominant lags are longer than a cycle, so, for instance, rate of change in tax return $2$ years ago has a stronger impact on unemployment than rate of change in tax receipts $1$ year ago, then the cycle is confounded and the proper sequence is indeterminable (without a lot more sophisticated analysis).

Fooling the System without Fooling the System

You can have a system that runs by certain rules, but which when looked at purely empirically seems to not run by said coherent rules.

Need an example? Consider Rugby: get all the equipment and trained teams et cetera, but then they all decide they “do not believe in Rugby” (too tankie-commie) and start playing football. Interesting game, right. No tries get scored. Always a 0–0 draw. Even though plenty of “goals” are scored past the goalkeeper I suppose. (We are presuming the scorekeeper is only noting true Rugby scores. I guess the odd drop-kick might score 3 points.)

In a converse sort of example, imagine the rules are Newtonian gravity or general relativity, the system is our Solar System, but the rules you find are Ptolemaic epicycles. “Ha! We made the earth go around the Sun, suck on that Kepler!”

What about playing cricket with baseball equipment. It can be done. Pretty hard to get clean bowled, so some odd scores again. You think you’ll get 1 point for a “home run” but the cricket scorekeeper logs a 6. Babe Ruth: “What’s going on here?” Don Bradman: “Lol!”

If you think a bunch of New Keynesians running the IMF and central banks is any different to these analogies,… maybe think again!

It can be a way to “understand” some of the craziness and chaos of financial markets.

How about anti-chaotic markets analogies? What if everyone suddenly became altruistic? What about if all chess games were suddenly played with legal moves but players refused to capture each other’s pieces? You could only deduce half the rules from empirical observations.

The more stupid of these analogies are the ones that are more alike to having neoliberals run the apparati of government fiscal policy.

Playing an Austerity Game with MMT Rules

So this is what all neoliberals do, whether left-wing or right-wing or up-wing or down-wing, red-wing, blue-wing or green-wing, dirty-wing or clean-wing. I feel a Dr Seuss cartoon coming on.

Would a numbskull who had no concept of rule of law be able to then look at just the financial data and conclude it was not an MMT operations system?

Douglas said ‘No!"

That’s the stance that a screwed up game of chess is still bound by the moves of the pieces.

I reckon, “Yes!”

That’s the stance that there is a poverty of input, and it becomes possible to imagine something non-MMT about the monetary system, even though underneath it is all MMT operations, down to the last tax liability and bank credit.

Which is it?

I suspect we’d find neither is entirely true nor false. There is in fact too much poverty of input without an understanding of the legal system and parliamentary law. Which means, just examining balance sheet data, one would not be able to get a statistically significant score to separate some of the theories. MMT would just become one of many.

Suppose, for instance, US Congress decided once and for all not to raise the fictional “debt ceiling”?

There is a conditional: anyone failing to see this is a free vote of Congress would be perhaps unable to tell the system is fiat currency. They might be led to conclude bank credit constrains the government net spending. This would be despite the fact the monetary operations and software systems are still all in accord with MMT.

It would be weird though. You’d not be able to compute the determinable amount of bank credit which constrains government. The banks certainly would not know. But mere weirdness is not a proof.

This is at least something to think about if the DougBot project “proof of MMT” is to be rhetorically successful, even if technically unimpeachable.

But… the Triple Sector Balances!

The Three Sector Balance always shows government deficit equals private surplus, to the penny.

Is this really a “poof of MMT” though? It’s just an accounting identity. In my explanation of MMT Basics this is only one of three principle ingredients. I use (1) real resource limits, and (2) the legal framework as the other two principles. So knuckleheads will not regard this as “proof of MMT”. (We are assuming bone-headed Austrians are incapable of looking up The Law).

Push and Pulls

TODO: was gonna write about the opposing forces.  

Steve Keen’s “Proof of MMT”

What ProfKeen does is code up a system of ordinary differential equations using macroscopic identities (definitions) and balance sheet stock-flow relations (so accounting rules, obeyed by banking system software) and then he runs these models to “show” that a rising government fiscal deficit is fresh currency injection into the non-government sectors of the economy, fuel for GDP growth (not a detraction via tax drain, so disproving the dead weight tax ideas).

This is not a proof of MMT.

ProfKeen had to assume MMT is the operational reality first. That is all in the code he writes. What ProfKeen’s Minsky models equipped with a government sector are doing is showing if what MMT describes is correct, then government deficits are fuel, not fire.

This is a conditional.

To use such models to “prove MMT” one has to then examine the data to see whether rising government deficits add to GDP without subtracting all the add via the inflation deflator, if true then this would falsify the Null Hypothesis of the New Keynesians and Neoclassicals that welfare incurs a dead weight loss.

One cannot prove or disprove MMT without such empirical data.

No Need for that Proof

I will repeat Warren Mosler’s characterization of “MMT”:

MMT is the recognition that the currency is a public monopoly and the consequences thereof.

This is a legal definition. One must examine the legal statutes to, (a) find out what entity is a “public monopoly” and (b) what their “currency” amounts to, and then read up on the legally permitted monetary operations. You will find in every major nation of the globe that they all operate an MMT system — provided the law is obeyed on average or in the majority of transactions. There really is no other proof of MMT.

Of course, on a small island nation with no legal system to speak of on paper, this all could be a ladies and gentleman’s agreement. Same thing though. Instead of reading legal statutes you’d ask these ladies and gentlemen how they operate their social obligation records.

This means the DougBot project proof is more like Steve Keen’s: we pretend we do not comprehend the law, and dumbly see if a few assumptions turn out to match the reality of currency flows.

Taking Away Our DougBot Food

One other conversation we had this week was over the correlation and lag from FICA tax acceleration to Employment and GDP rates of change (FICA would be payroll tax in NZ). If there is a strong correlation it would be proof that “tax drives the currency”. There was indeed a strong eye-ball correlation. That is one of the consequences of MMT in the specific cases where the public monopoly issues their own fiat currency, and drives a significant demand for it by imposing payroll tax liabilities (rather than say promising gold or pizza delivery).

But then a key policy proposal of MMT is to get rid of FICA taxes, because (a) they are regressive, (b) they penalise what we want to actually incentivise (paid employment), and (c) there are far more progressive types of tax liability that could be imposed without risking tax evasions, or indeed incentivizing tax evasion, and (d) we want to eliminate the horrific waste of human life times spent on compliance costs for such backwards taxes. Give the accountants more holiday time.

It is far better to tax, by a simple fee or fine, egregious consumption (of owner unoccupied land or real estate, electricity, water,…) rather than taxing the work that produces the output we want.

Ah! But this would remove one data time series the DougBot would use to “prove MMT”. So we are asking Mosler to shush his smart alec trap shut, for a while anyway.

The Proof in the Eaten pudding

We have already had abundant evidence over the COVID supply shock crisis that,

• government spending adds to GDP.
• Central bank bond rate hikes add to GDP.
• Government deficit acceleration comes after the inflation impulse, not before (is an effect, not a cause) — to check this one must drag the CPI published inflation back in time to when the actual store prices would have risen.
• MMT also “predicts” (conditionally) that central bank rate hikes are directly inflationary — but this was near to impossible to validate for most of the COVID-19 era because all the initial inflationary impulse came from the effects of supply chain disruptions and mass lost output due to lockdowns, death, ongoing supply chain frictions, and then at the tail end egregious price gouging by OPEC+ and other monopolists (oil being the single biggest input to food production).

All consistent with MMT, and difficult to explain with any other mainstream school of thought.

An important caveat: one might expect the FED raising the interest rate floor (a repeller , so not always an add to every households income!) would sustain inflation. MMT would say “Yes, but…” Why? Because the source of the COVID era inflation was not from too much effective demand (money supply) it was from real good supply and labour shocks. You cannot cure a supply shock with a demand add, not through the interest-income channel at least. You need the demand add to be in the form of real investments to boost the supply of goods.

Read Bill Mitchell’s accounts of the COVID inflation era if you are in doubt.