« Does vote swapping increase efficiency? | Main | Someone is Getting Health Spending Under Control »


Feed You can follow this conversation by subscribing to the comment feed for this post.

to suggest the market interest rates don't determine investment is not the same as saying
investment decisions are made without a time discount of some sort
at least mobilizing in the back of a deciders head

the system can be both determined and its decision makers faced with radical uncertainty
if the agents have their causal reasoning behnd every choice the system is determined
if the deciders lack complete info about everyone elses actions till kingdom come
even secondary uncertainty over lays the primary natural uncertainty
oh it spirals off here

but simple mechanicals jazzed up with a silly randomizer
isn't the only way
to get results that look like we see out there in the actual marketplaces of earth



when uncertainty is as radical as it is when viewing the prospects for a new plant built here starting tomorrow
then to suggest a criterion for rational choice gets messy and problematical in the extreme

the cone heads that suggest otherwise
are prolly performing a card trick on you

it delights me to see lerner tossed about here
might i urge his perspective

and why
because now we have the tools of a propr demiurg we can make happen ..more or less...what we want to make happen
its not realying on divine providence or best of all possible worlds spontaneous conjunctions
or the worship of jesus's miracle of the markets
as described in the lost gospel of saint mammon

in the case of norte amigo lands
the state can make ull employment and stable development partners
without pretending there is a single all rational agency like an over soul behind the micro foundations

as lerner shows us

we need a thermostat to control total output motions
and a cap and trade system to control price level motions

and as to micro
all we need to be liberal legit
is a healthy skill at mechanism design
and hawk eyes looking for pareto improving moves
to guide us thru the stiglitz-ian millions upon millions of small inefficiencies failures incompletenesses
etc that post samuelson-arrow post modern micro theory reveals to us

---1970 -1990 ---

hows that for give me a second
to me tell you my plan eh ??

recall the mothers of invention goof ranting :

" Do you think that I'm crazy?
Out of my mind?
Do you think that I creep in the night
And sleep in a phone booth?

Lemme take a minute &
tell you my plan
Lemme take a minute & tell who I am

If it doesn't show,

Think you better know
I'm another person

“Let me put it another way. The orthodox New Keynesian view is that Y equilibrates S and I in the short run, given the r set by the Bank. But in the long run, when Y is set at "full employment" Y cannot adjust to equilibrate S and I, so the r set by the Bank must be adjusted to equilibrate S and I.”

So r follows rn.

But why is rn not a (partial) function of the path of r?

I.e. why does r not (partially) determine rn?

That would mean rn becomes obsolete as quickly as r (t) changes value to r (t+), and that r is always the dominant force.

“If MMTers did believe what my crude model says they believe, then the MMT policy prescriptions make sense. In that sense, my reverse engineering gets the right answers.”

You may be confusing sufficient for necessary and sufficient.


Parenteau’s essay is about the MMT sector financial balances model – to which Krugman’s graph is equivalent (according to Parenteau).

Can you explain simply the connection/equivalence between the Keynesian cross and the MMT sector financial balances model?

Andy, You said;

"Isn't the QTM itself an example (perhaps the clearest example) of a theory that fails via the Lucas Critique?"

You're right. I had two other ideas in mind. The QTM can give you a ballpark estimate of the price level--it can explain why Japan's price level is roughly 100 times higher than the US price level. (Nick says Keynesianism can't do that--which I see as a huge flaw.) Second, quasi-monetarism (not the QTM) is the best way to model the monetary system. It has a nominal anchor.

Suppose the US adopts Japanese monetary policy. The QTM predicts the US price level will rise by several orders of magnitude. And the Keynesian theory predicts . . . I'm not quite sure what it predicts. What would happen to US interest rates right now if we adopted Japanese monetary policy? I have no idea.

Monetary policy has certain long run effects on nominal aggregates. The short run affect is largely determined by expectations of that long run effect. Thus the QTM explains the long run effect, and quasi-monetarism explains the short run effect.

Scotty seems never to consider the pricing system might determine the money supply in a discrretionary credit based market economy
The rough relationship is produced from the other direction
Output and prices induce the necessary money supply

Not that Beau summoner will respond to masked commenters of unknown pedigree
But Others might notice something to consider

Scotty seems to me mystified by the difference between a dollar based economy and a penny based economy
Summers famous catsup bottle economics

QTM was good enough for Hume so it's good enough for ........
I want to see a duel between new monetary theory and modern monetary theory

Mars versus Neptune

What happens when the size of the money stock is partly endogenous ?

Nr's mistake
A post that didn't resort to toy scenarios drew my interest
I promise if he sticks to bake scratchonomics
I'll stay silent

I don't claim to actually know the MMT literature, but I think I understand intuitively its critique of the stuff that's taught in econ grad schools. And I would say that it is precisely the following assumption that is (one of) the problems that MMT seeks to address:

NR: luis: "thanks. this is what confuses me - the thing about how net financial assets are created/destroyed seems to me, when I translate it into terms I understand, as a perfectly unremarkable restatement of text book macro. But (some) MMTers portray it as an insight (we know how the monetary system really works!) so I must be missing something."

Same here. Apples bought must equal apples sold. Financial assets bought must equal financial assets sold.

Because financial assets aren't real goods, every financial asset identically also represents a liability. The issue is that some financial assets (bank deposits, money market funds, repos) are also money. And these forms of money are "backed" in the private sector by the countervailing liabilities (loans, commercial paper, etc.) that may or may not have a "true" value commensurate with the value on the banks' books -- or with the amount of money created in exchange for them.

Thus (i) the financial asset "identity" is dependent on the quality of banks' loan origination procedures, and (ii) the money supply (and its stability) are also dependent on the quality of these procedures and the central bank does not control the money supply, so it's not so clear what it is that an LM curve represents.

Paine, I think we can all agree that a monetary reform making pennies the medium of account will increase prices by precisely 100-fold (the ketchup example.) The implicit assumption is that all nominal contracts will be revised 100-fold. The more interesting question is what happens when the base increases 100-fold, but nominal debt contracts are not re-adjusted.


I am not a card carrying MMTer but let me try my hand at explaining the bond versus cash issue.

1. Insofar as net financial wealth is concerned, bond and cash are similar and unique when compared with other financial assets. All other financial assets are directly or indirectly claims on other physical assets or cash flow from them. Government bonds and currency are not. As such. they are countercyclical--their value relative to other assets goes up during downturns and vice versa.

2. Under boom conditions, lots of financial assets behave as if they are money or close money substitutes, especially if they are short-duration and are nominally fixed claims. For example commercial paper. In downturns, they lose their "moneyness." In booms you don't need the government providing the net financial assets because the private sector is working overtime to produce net financial assets (as asset prices go up in relation to the credit that was created to finance and fuel them). However, this process is inherently destabilizing (Minsky).

3. During busts, only the government can provide net financial assets. Whether this comes in the form of bonds or money is distinctly secondary. Once the private sector demand for net financial assets if satified demand for goods and services will take care of itself.

4. Mainstream monetary policy is almost entirely focused on rearranging the composition of money+government bond held by the public. This makes sense in a gold standard world, where government government bonds are "inside" asset as well and gold (and money backed by it) is the only uncorrelated financial asset. So, an increase in the amount of gold (say a new find) would be expansionary. it has no relevance for the fiat world.

5. In the fiat world, although money is technically an "inside asset" in the sense that its value is ultimately underpinned by the society's productive capacity and the government's power to tax a significant portion of it, in practice it behaves like an outside asset. In places where this does not hold true--you have Zimbabwe (or name our country). And in the fiat world, government bonds are not really different from money.

6. The exchange property of money (vis-a-vis) bonds is vastly exaggerated. Insofar as pure exchange for mere economic activity (as opposed to financial market and credit activity) as long as you have some concept of money, a system of book-keeping and trade credit is generally sufficient.

7. To Scott Sumer: when money supply is endogenous, the idea that money stock determines the price level is meaningless. It is tautologically true (if you can define the correct money stock to use) but vacuous in terms of understanding the causal forces.

8. The IS curve cannot be seen in isolation of the LM curve--they are not independent. Common variables drive them. A rise in perception of uncertainty will drive demand for money from a portfolio allocation perspective and simultaneously drive down the demand for physical investment (IS).

9. It is not that interest rates have no influence. Rather the influence of interest rates is rather nonlinear, context-dependent, and generally quite uncertain. Other than housing, most investment is not sensitive to cost of capital, at least at the margin. Yes, financial asset prices are sensitive to interest rates but that relationship is obviously not invariant and quite volatile. Besides, if asset price appreciation is your objective why not use fiscal policy to increase net financial assets? At least that is more transparent and perhaps less plutocratic.

JKH: "Parenteau’s essay is about the MMT sector financial balances model – to which Krugman’s graph is equivalent (according to Parenteau).
Can you explain simply the connection/equivalence between the Keynesian cross and the MMT sector financial balances model?"

Let me try.

Let's start with the Keynesian Cross model (closed economy version):

Desired Aggregate Expenditure (demand for output) is divided into 3 components: desired consumption, Investment, and Government expenditure:


That's just accounting. Now we add some behavioural assumptions:


where Y is income, T is taxes, so Y-T is disposable income. a is some positive number, and b (the marginal propensity to consume) is between 0 and 1. So the consumption function just says that consumption demand depends positively on disposable income.

Here's a simple theory of taxes:


(taxes are proportional to income).

Assume I and G are exogenous (fixed).

Substitue the behavioural functions into the accounting identity and you get:


That's the AE curve. It has a positive intercept, and an upward slope of less than one. AE (demand for output) increases when income increases.

Now we add the equilibrium condition:


That's the "45 degree line". Income is determined by demand for output, and the two are equal in equilibrium.

Where the AE curve crosses the 45 degree line we get the equilibrium level of Y:


Solving we get the equilibrium for Y as:


OK. That's the ECON101 macro model, for the last 50(?) years.

Now, let's take the exact same equilibrium condition:


and play with it. Subtract C from both sides:


Subtract T from both sides:


Define savings S as:


So we get:


Rearrange, we get the MMT equilibrium condition:


Substituting in from the behavioural equations for consumption and taxes we can re-write this as:

-a+(1-b)(1-t)Y-I=G-tY [edited to fix stupid arithmetic mistake]

Which (IIRC) are the two curves of the Krugman diagram, with equilibrium Y determined as the intersection of those two curves.

[Funnily enough, back when I was an undergrad we used to use this as our equilibrium condition:


We called it the "Injections equals Withdrawals approach". But we knew it was the exact same model as the Keynesian Cross, just a different way of looking at the same thing.]

Bottom line: what many people think of as the MMT model is mathematically identical to the model I have either taught or been taught for the last 40 years.

I have read Warren Mosler and Ralph Musgrave's comments on Brad DeLong's post on MMT.


OK guys, confess: who was the blogger that came up with the name "MMT"?

It was maybe a brilliant marketing ploy to attract non-economists. But it has caused no end of confusion for us economists. Especially for those of us who have actually read Keynes' General Theory and Abba Lerner's Functional Finance.

Forget wading through the thickets of words and accounting. This is what really matters: this is what I should really have written about:

"Under what conditions would Abba Lerner be right about the long run government budget constraint and Functional Finance?"

That would have been a useful post that could have moved things forward. Instead, I'm trying to hack my way through thickets.

thanks for that, Nick

especially since I just saw the DeLong post and your comment there


(My impression was Bill Mitchell started using "MMT", but I guess that's wrong)

JKH: Sorry :).

You had the guts to ask the question. There must be dozens of others who didn't know the answer but lacked the guts to say so! So I don't really begrudge answering it for you.

Keynes coined the term "modern money," and Wray named his book after that. But we always have used the term neo-Chartalist to refer to ourselves in academic circles. Bill Mitchell started using "modern monetary theory" on his blog a few years ago, and it eventually stuck. I don't know that the latter has actually been used by any of us to describe ourselves in an academic paper yet, though, and I don't know if we will.

Regarding Lerner, FF, and the IGBC, they are perfectly consistent with each other if by FF one means a fiscal policy strategy that improves the primary budget balance as the economy nears full employment and continues to do so 1-for-1 with increases in private spending when the economy is at full employment.

Scott (Fullwiler) (2 Scotts commenting here):

Aha! So I will blame Bill! I admit it was a brilliant marketing ploy. But I still am slightly pissed at him for using it! "Functional Finance" is much more descriptively useful. "Chartalist" reminds me of Knapp's State Theory of Money -- one of the worst books I have ever (tried to) read. Abba Lerner is streaks ahead of Knapp. Because you can understand what Lerner is saying, and if he's right he's saying something important and substantive.

"Regarding Lerner, FF, and the IGBC, they are perfectly consistent with each other if by FF one means a fiscal policy strategy that improves the primary budget balance as the economy nears full employment and continues to do so 1-for-1 with increases in private spending when the economy is at full employment."

That would require a very large fiscal multiplier, wouldn't it? That would make the standard Keynesian Cross equilibrium unstable? MPC greater than one? Or, instead, some sort of "pump priming"/"Kick starting" multiplier where a *temporary* increase in G gets the ball rolling and causes a *permanent* (or at least long-lived) increase in Y?

My mind was going in a different direction. It all depends on whether r can be kept below g, for the stable-Ponzi regime to hold. And that depends on the shape of the IS curve and the mix of money vs bond financing (not very MMT). Loose money and tight fiscal (relatively speaking) would do it. My vertical IS assumption would be sufficient but not necessary.

But I haven't worked it out. Just came up with the idea a couple of hours back.


As you're familiar with Lerner, how would you interpret the meaning of "functional" as in "functional finance"?

(I'm not so familiar)


I would go this way:

Start with these from your discussion:

End here for sector balances with "behavior":
Y(1-b)(1-t)-a-I = G-tY

The left hand side is desired net saving given Y as defined by MMT'ers. The right-hand side is the govt's budget stance, again, given Y.

To your point about accounting vs. economics . . .

First, one needs economics to create behavioral relationships, as you did, and would need the latter to solve for Y and get actual net private saving and the govt's balance.

Second, one needs accounting to ensure stock-flow consistency in building the model (though this model has only flows) consistent with how "score" is kept in the real world. It also aids in interpreting the results--as when one solves for Y with one of your very early equations (the basic multiplier equation from Econ101). With the sector balances equation, we understand that (1) any fiscal policy action "works" by adjusting private incomes via government spending or taxation (raising G or lowering T lowers Y), that is, govt deficits raise net private saving; (2) any monetary policy action "works" by adjusting desired spending given Y (changing I or a), that is, monetary policy is stimulative when it induces the pvt sector to reduce its net private saving. These insights are generally absent from the traditional interpretation of these models, as the effects of stimulate policy are seen as equivalent (in as much as they are seen to actually "work," which is a separate matter altogether).

Blending the economics with the accounting, MMT'ers along with folks like Richard Koo understand that the non-govt sector isn't likely to re-leverage on the scale necessary to return to full employment anytime soon. It also makes clear that austerity policies can only avoid worsening the recession if the non-govt sector is ready to releverage, and that if the non-govt sector instead tries to continue deleveraging in the face of austerity policies both the economy and the deficits will worsen. One could get some of the latter these points from a typical IS-LM if one wanted to actually do the math, but it's rare that someone specifies the balance sheet positions required for either outcome to be true using only IS-LM.

I'll stop there as I'm being summoned by my better half. Hope that helps out a bit.

"The more interesting question is what happens when the base increases 100-fold, but nominal debt contracts are not re-adjusted."

But the CB cannot increase the base by 100 times because it doesn't have that many government bonds to purchase. There must be deficit spending first, and only then the CB can increase the base by purchasing some of those bonds.

I think you are envisioning the CB conducting fiscal policy. I also like to conceptually view monetary policy as something that is expenditure/revenue-neutral for the government, whereas fiscal policy is an adjustment in expenditures and revenues (apart from interest rate effects).

But this ceiling on monetary policy has an interesting side effect (to me). Namely, because the total stock of money -- MZM -- held by the public is roughly equal to the federal debt held by the public, when the CB purchases a bond for a deposit, then on the margin, the household can sell the deposit back to the banking system in exchange for another bond (or paper, etc.)

Households do not need to spend unwanted deposits by purchasing goods. They can spend unwanted deposits by purchasing financial sector debt, and this pushes the excess base back onto the banks, where it sits as reserves, even as household deposit levels are unchanged. The only thing that changes is the relative price of bonds and deposits. So if the "money" in the money demand function, is deposits and not the base, then the CB has not succeeded in increasing household money holdings.

This would a "monetary pessimism" scenario, in which velocity changes negate the effects of an increase in the base. It's a lower bound on the effectiveness of monetary policy on the price level, and this is a function of the size of MZM.

An increase in the base does not force MZM -- "money" -- to increase as long as the size of the base is less than the size of MZM. And historically the growth of MZM has not tracked growth of the base, but has tracked the size of federal debt -- MZM moves with fiscal policy, not monetary policy.

But as the CB keeps buying bonds then under our pessimism scenario, once MZM = Base, then an additional increase in the Base must lead to an increase in MZM. At that point, households will have no option but to purchase goods if they want to reduce their deposit holdings. The question is, why would they? As their net-worth has not improved, why would they increase spending? But assume that they do. Then there would be a price effect. But unfortunately, the price effect kicks in only after the critical value, and the critical value is the ceiling on how much