JKH on the Recent MMR/MMT Debates

By JKH (This document can also be read in PDF format here)


A recent round of blogosphere debates between some of those associated with MMR (Modern Monetary Realism) and some MMT (Modern Monetary Theory) enthusiasts examined the question of the nature of saving in economics and finance. As part of that discussion, an equation appeared suddenly, S = I + (S – I), whose origin I have some personal responsibility for. It met with mostly disapproval and even scorn from the MMT side, particularly given its innocuous tautological form. Several people did plumb its intended meaning. The MMR folks grasped it. In my view, Tom Hickey at Mike Norman Economics was one person from the MMT side of things who by paying attention and following and documenting the development gained some understanding of how and why this equation came to be and its context. He made reference to how it originated several times in various blog conversations, attempting to be helpful for the benefit of fellow MMT commenters. But I don’t recall seeing much evidence that his effort prodded corresponding investigation by those who were scratching their heads. Whatever Tom may have thought of the thing, he did understand that there was a more complicated back story to how this item bubbled to the surface of conversation on the subject of saving. The other MMT affiliated person I’d mention is Neil Wilson, who while sceptical of the notion that such an equation might be meaningful, did take the time to identify the correct technical specifications of its underlying assumptions and variables including the importance of the sector context. Several commenters not directly affiliated with either MMT or MMR grasped the key points. FDO15 and Vimothy come to mind, but I may be overlooking others. Special mention also goes to Ramanan, who is a persistent seeker of accuracy when it comes to the subject more generally.

Briefly, the equation mostly came about in recent form as my own version of a well known blogger’s cogent observation on MMT’s stylistic use of the language of “saving”. In that sense of origination, I am the messenger, not the message. But I happen also to agree with the message, in verbal or math form, and had thought of it myself before seeing it articulated so effectively by someone else. The blogger in question was Steve Waldman at Interfluidity. I want to emphasize here that Steve Waldman seems to be a fairly strong supporter of the best that MMT has to offer, and has documented that in several posts. So his concern as it relates to the subject here is focused and exceptional in that context.

The full sequence of the original verbal portrayal and my math translation of it are included in section 2. But I don’t want this to be just about that equation. I haven’t been on any great promotional campaign in its favor. But I do think it’s a concise illumination on a particular MMT style of exposition. More general observations in this category have been recorded by Marc Lavoie and Brett Fiebiger in their recent papers regarding the MMT form of monetary system description.

But when it reaches the point where an MMT leader tries to trample the thing to death in a flurry of venomous analytical naiveté, it deserves renewed briefing, at least for the potential benefit of those not yet fully biased against it. However, it’s not been intended to be such a center of attention. That only distracts from “Modern Monetary Realism’s” (MMR) more sensible reference to it as a symbolic representation of a balanced approach to describing the modern monetary system. MMR picked up on the expression as confirmation in balancing “vertical” and “horizontal” views of the monetary system, when providing explanations of its operation and especially in dealing with the description of “saving” in fully accurate fashion.

Some of the blogs involved in discussions around this general subject included, but were not restricted to, Interfluidity, Asymptosis, Mike Norman Economics, MMR, winterspeak, 3spoken, and John Carney’s CNBC blog. In a way, Interfluidity was the origin of the critique in full, starting with several posts in 2011. Much of the subsequent discussion took place in February 2012, which coincided with the early days of MMR blogging. John Carney of CNBC reproduced a particular comment of mine from MMR, relating to the theme of the meaning of saving. That comment referenced MMT’s preference to compress the idea of saving into a particular model that emphasizes the measure of consolidated private sector net financial saving. The comment reproduced at the CNBC blog made no direct reference to the suspect equation, but covered in general terms the originating theme that spawned it.

As far as the suspect equation is concerned, the general reaction to it was mixed, including those who were negative. A good deal of this was just puzzlement as to the intended purpose of such an innocuous form. Some of it went further and immediately assigned it zero value as contributing to the understanding of monetary systems. Some of it was even contemptuous, in particular the rather rancid, dismissive assessment by Professor Randall Wray. Some of it, including Wray’s analysis, condemned the surface tautological appearance of the equation as an indictment of vapidity and meaninglessness. In contrast, Warren Mosler in a separate blog discussion exhibited more measured scepticism, while making insightful and useful points about the topic of saving in economics, and posing a reasonable question indirectly for me in the process. Perhaps he felt he didn’t have quite enough information before arriving at the entirely dismissive point of his academic colleague. I appreciate that he didn’t render the death sentence for the crime of tautology. His marshalling of the conversation flow at that particular blog is worth a look. Accordingly, I’ve excerpted some of the dialogue there in section 3, and added some of my own interpretation in the form of accompanying remarks.

My response to L. R. Wray is in section 4.

The interesting thing to me is how spontaneously the general theme was picked up by MMR, because I don’t recall doing any heavy lobbying to that effect. Steve Waldman also expressed a favorable opinion regarding the development of the equation. How is it that both Waldman (an MMT supporter in some ways) and MMR came to a quick and similar view on this? What would be the common factor there? A few others joined positively in as well.

Waldman, Lavoie, and Fiebiger all record legitimate observations in my view. These observations are inextricably linked to specific stylistic tendencies of MMT. The recent collection of conversations has not only confirmed this, in my view, but exacerbated it with inconsistencies of argument that continue to betray internal contradiction.

There are 4 sections – a discussion of the concept of saving, a description of the development of the equation S = I + (S – I), and reviews of each of the noted Mosler and Wray discussions/responses.

1.  Saving

Saving From Income

John Carney of CNBC published a blog post on March 1, 2012, titled “More on Saving”:


In that post, he reproduced a comment I had left at the Modern Monetary Realism post “Paul Krugman does S = I + (S – I)”:


That comment included the following:

“The correct economic definition of saving is disposable income not spent on consumer goods.”

That’s not perfect, but it is close enough. As preamble to the full discussion on S = I + (S – I), its worth adding some meat to those bones. As part of this, I want to make reference to Andy Harless’ post from several years ago “Investment Makes Saving Possible”:


Here is Harless’ definition from that post:

“Savings equal income minus consumption.”

Pretty close. He uses the plural form savings. To be more precise, saving refers to a subset of income flow, and savings usually refers to the cumulative balance sheet result of such a flow, i.e. a stock. But the plural form is a minor point in the debate about saving.

Apart from that, I agree with just about everything in the Harless post. So my objective here is not to reproduce his post, but use it as a credible and consistent reference. Most of what follows here is an extension from that base.

(Saving in the case of corporations is net undistributed profit, which follows after consumption in the form of capital depreciation cost.)

Harless is very specific regarding constraints he imposes on sector structure. He does this to streamline his treatment of causality in the relationship between saving and investment. His model eliminates government and foreign sector imbalances by assumption. This means that private sector S = I. Because of this, his particular model, although wise on issues of causality, is not directly relevant for purposes of examining sector specific issues in the MMT “sector financial balances” (SFB) model in its most common 3 sector form. It cannot provide insight into the implications of sector multiplicity. I do agree with his overview of causality in the relationship between saving and investment. But that is unrelated to issues that arise with sector complexity. As will be explained in section 2 below, the equation S = I + (S – I) was created and written as specific to the 3 sector model. The S used in the equation is the same private sector saving S used in a 3 sector model, not the consolidated single sector saving that Harless uses.

Harless confirms that the primary causal direction at the macro level is that investment “allows” saving. This is a well known Keynesian analytic theme. There’s no need to elaborate on it here.

Choosing the Number of Sectors

Saving equals investment (in magnitude, not in substance) in a one sector economic model.

A one sector model can be simulated by constructing a global consolidation in which international current accounts sum to zero, and government accounts are consolidated with private sector accounts.

A more common approach in constructing a single sector model is to posit a national economy where both government and international current accounts are assumed to be in balance. A national system that assumes balanced government and foreign accounts can assume an S = I relationship, where S is private sector saving, albeit in the context of the constraints noted. This is an artificial construct.

The only real world working economy that can be represented as a single sector model is the global economy. Current accounts net to zero on consolidation, and remaining government and private sector accounts can be consolidated globally as if one. The resulting system is a globally consolidated fusion of government and private sectors. Still, for S = I in such a global construct, S must be defined as the combination of government and private sector saving. That is very different than the private sector S savings used by convention in a 3 sector national model.

From either single sector basis, one can move on to 2, 3, 4 and higher sector models.

The 2 sector model, used by MMT for its highest order demonstration of such concepts as “net financial assets”, consists of government and non government.

The 3 sector model, consisting of government, private, and foreign sectors, is derived from the standard national income accounting equation:

C + I + G + (X – M) = C + S + T

In this model, S is defined as private sector saving.

The MMT 3-SFB model is simply a rearrangement of terms in the national income model. For example:

(S – I) = (G – T) + (X – M)

That says that private sector “net saving” equals the sum of the budget deficit and the international capital account deficit. This interprets private sector “net saving” as the amount of financial surplus that spills over as funding for some combination of the government deficit and the capital account deficit.

(One can think of a capital account deficit as the automatic net debit to a pool of funds created by a corresponding current account surplus (X – M).

Another rearranged expression is:

(S – I) + (T – G) + (M – X) = 0

This says that the sum of the net financial surpluses of all three sectors must be zero. That must be the case because net financial assets in total equate to mirror image reversing liabilities in total. MMT often presents the 3-SFB model in graphic form in this way, with the 3 sector balances represented above or below the zero line, and adding up to zero.

Finally, yet another expression:

S = I + (G – T) + (X – M)

That says that private sector saving is the sum of the amount of saving required to fund investment I, plus the government deficit (G – T) plus the capital account deficit (X – M).

THIS is the form that links directly into the interpretation of S = I + (S – I) in a 3 sector model, which is the subject of section 2.

Recall that the Harless post describes a system in which S = I because of specified 3-sector balance constraints – i.e. financial balance in government and foreign accounts. But no actual open economy situation can be represented this way, because government and current accounts are not identically zero.

Stocks and Flows

The expressions used so far all pertain to flow accounts. Economic units that save from income produce a flow of saving. That flow accumulates over multiple time periods to a corresponding stock account. That stock account is referred to as savings, net worth, or equity. These are all conceptually interchangeable expressions.

Economic units that spend less than their disposable income on consumption save by definition. As a result, they increase their stock of savings, net worth, or equity.

Economic units that spend more than their disposable income on consumption dissave by definition. As a result, they decrease their stock of savings, net worth, or equity.

A unit that saves can acquire physical assets (investment) or financial assets, or repay liabilities.

However, the act of saving from income is separate from subsequent assets of asset-liability management. This is a critical characteristic in the correct definition of saving.

For example, a unit can increase liabilities to acquire assets, or liquidate assets to reduce liabilities. The use of funds in such cases is the same as it might be when saving or savings are the source of funds, but such asset-liability transactions require no saving.

Financial Accounting and Economics

From a financial accounting perspective, the following is true:

Saving is an income statement “event” in the sense that saving must be a subset of income. (Harless has the same view of it.) The accumulation of saving is recorded as a balance sheet item usually called net worth or equity (and classified as savings).

The deployment of saving or savings as a source of funds into increases in investment or financial assets (or reductions in liabilities) as a use of funds is classified as an accounting flow of funds. (Flow of funds is sometimes known as sources and uses of funds in corporate financial reporting.)

As noted above, economic units can alternatively acquire investment or financial assets or reduce liabilities with the proceeds of borrowing. But both lending and borrowing are separate and distinct from saving. Borrowing and the deployment of funds borrowed are also recorded through accounting flow of funds. And the same holds for the reversal of such transactions.

In summary, the three types of financial statements are income, balance sheet, and flow of funds. The income statement connects income and expense flows to balance sheets via the equity or net worth account. The flow of funds statement connects flows between different balance sheets in time and space. Although such statements apply formally to corporate units, the accounting concepts carry over to household finance in a comparable fashion. Household saving is comparable to corporate saving in the form of undistributed profit.

This is all straightforward accounting that is essential to understanding the monetary system and arguably economics more generally. It is necessary but not sufficient in that sense.

While the described accounting framework is more familiar in the corporate environment (and the household sector by extension), it is certainly applicable in concept to the government sector. Moreover, the described set of “micro” accounting statements has “macro” correspondents in the form of such accounting statements as NIPA, Fed Flow of Funds, SNA, etc.

There are 1, 2, 3, and 4 sector models. But in theoretical concept, one could have a 7 billion sector model, with a balance sheet corresponding to each individual. These balance sheets could include connecting accounts for the internal flows of funds within households, for example. And each balance sheet has for example a “net financial asset” measure.

One aspect of the recent debate touched on the question of whether expenditure on investment should correspond to underlying saving. The conventional definition of course includes investment as a use of funds sourced from saving (at the macro level). But if saving were defined instead such that gross investment were to be treated similarly to consumption, then saving would be identically zero for a one sector model and for the global economy. Saving would not be equal to investment.

Such an idea results in some crazy arithmetic relationships. In the US, the household sector would hold approximately $ 40 trillion in net financial claims corresponding to the value of its cumulative saving. At the same time, the US as a whole would have dissaved in stock form by approximately $ 2 trillion, and the global economy would have no cumulative saving at all. This would be a chaotic mix of saving relationships indeed. My own view is that this approach to the definition of saving would be highly dysfunctional and wrong. MMT thinkers seem to be highly flexible on this aspect, though. Warren Mosler alluded to it in recent comments excerpted in section 3. The correct generic definition of saving in my view is still that expressed by Harless. It’s pretty much the language I used in my comment that John Carney used in his CNBC post.

Net Financial Assets (NFA)

MMT uses the term “net financial assets” or NFA. But there’s lots of confusion around this idea.

First, MMT didn’t invent the terminology “net financial assets”. It is a generic accounting categorization. For example, the United States has an NFA position with the rest of the world equal to its net international investment position, which is currently a liability, making it negative NFA or NFL. Any individual household whose financial assets exceed its liabilities has an NFA position. The US household sector as a whole holds a very substantial NFA position due to its financial claims on the business sector. Any corporation that holds physical assets in any amount has an NFL position.

NFA is an accounting stock term rather than a flow term. It is most correct to refer to “change in NFA” when incorporating such changes relative to income and saving.

The NFA characteristics of a particular economic representation depend on the chosen sector decomposition of that representation.

In the MMT 2 sector SFB model, with government and non-government sectors, NFA positions of the two interfacing sectors are the symmetric inverse of each other. This is the generic version of the basic MMT point regarding the government as a supplier of NFA to the non government sector.

The frequently used MMT 3 sector SFB model admits 3 sectors, each with their own NFA position (positive or negative).

3 and higher sector models are more complicated from an NFA analytical perspective. One can’t conclude automatically that any particular non-government sub-sector has an NFA position that is the direct result of government issuing NFA. That has to be assessed by examining the portfolio structure of NFA across all sectors. For example, the fact that much US treasury debt is held outside the US means a more nuanced view of government NFA provisioning is required.

Thus, the impact of government supplied NFA must be qualified as one explores increasingly higher dimensional sector decompositions. This becomes critical when decomposing the private sector into household and business sectors, where household sector NFA dramatically exceeds private sector NFA. The consolidated private sector view obscures important underlying dynamics relative to the all important household NFA and broader savings position. That is the subject of following sections.

Other Considerations

Recent debate has included some mention of the use of the terms “real” and “nominal”. Both terms are fraught with ambiguity. Perhaps “physical” is an option when distinguishing the physical nature of investment from the monetary nature of financial assets. “Nominal” seems to be a term whose destiny is assured as eternally ambiguous.

Warren Mosler has said that saving is the record of investment (section 3). I think such terminology reflects an inherent MMT bias toward loosely defined consolidation concepts, something that both Marc Lavoie and Brett Fiebiger have criticized MMT for in recent papers. (I’ve referred to those papers further below.) In my view, the technique of conceptual consolidation is overused and pointedly insufficient as a central analytical technique for understanding the modern monetary system. And I’d say saving is not the record of investment at all. The record of saving is found in the lower lines of an income statement. The record of saving accumulation, or savings, lies in a section of the balance sheet that is usually referred to as net worth or equity. The record of outstanding stock investment is found on the asset side of the balance sheet, whether micro or macro. And the relationship of saving to investment is recorded in a flow of funds accounting statement.

In summary, saving is a subset of income. It is a flow, not a stock. It is the residual of after-tax spending on consumption. (In the case of corporations, it is undistributed profit after the payment of all expenses including taxes and depreciation.) Saving is not the actual deployment of funds into asset acquisitions or liability reductions. Those events are defined subsequent to the fact of saving. Harless recognizes this distinction in his piece. Saving flows into net worth, savings, or equity account as the subsequent stock form accumulation. Saving or savings is not a placement of funds “from itself” into other stock categories, a logical error made by many in an attempt to describe the essence of saving.

A system can be closed by construction or closed by assumption. The global system is closed by construction, assuming accounts are appropriately consolidated for closure purposes by intentional definition. A national system in which government and current accounts are balanced by assumption artificially constructs the private sector as a closed system in which S = I. When that system is opened for real world analysis, S becomes defined as private sector saving, according to the normal national accounting identity. Private sector saving can be decomposed into the amount of saving required to fund investment I and the amount deployed in net financial assets (S – I). That is the focus of the equation S = I + (S – I). The detail of the deconstruction is explained in the next section.

As background, there is more discussion of this general theme in the following Asymptosis post and comments:


2. S = I + (S – I)

A local blogosphere dust storm kicked up in early 2012 around a particular equation that I have some responsibility for:

S = I + (S – I)

In particular, the MMT movement got quite agitated about it, especially since the MMR blog became active in promoting the idea. I thought it might be useful to lay out the facts as to how this particular expression came about.

As background, the equation S = I + (S – I) began recent life, at least with any momentum, as a response I made to a comment by made by Steve Waldman at Steve Roth’s Asymptosis blog on February 1, 2012.

Here is part of Waldman’s comment:

“There is no “to the penny” sort of accounting relationship between household-sector financial saving and Government Issue of NFA. You can tell stories of how there might be a positive relationship between the two, but those are a function of policy choices and behavioral models, not logically necessary as a matter of accounting. And it is household-sector saving that conventional morality so strongly proclaims as a virtue. No bourgeois moralist ever complains when the value of a firm’s assets rise, implying an increase in business-sector “indebtedness” to shareholders. (Firms owe their full value to financial claimants, as their value rises, so does what they owe.) It is a bad rhetorical trick that MMTers sometimes pull, to confuse an increase in “private sector net financial assets” with an increase in household-sector savings in order to recruit bourgeois support for the latter in the cause of promoting net issuance of government securities. There are a lot of perfectly good reasons to support net-issuance of government securities during times like now, and I’m certainly allied with MMTers in their promotion of wise fiscal policy. But claiming “saving” is impossible as a matter of accounting without a government deficit is a bait and switch, a game played with definitions by rhetorical confusing household sector financial surplus with an aggregate private sector surplus.”


(This quote is included as part of a more complete record, at the end of this appendix.)

Waldman’s Interfluidity blog is a popular spot for discussion about economics and finance. Notwithstanding the quote above, he seems to be a relatively strong supporter of MMT. He does have stated qualifications with this aspect noted above being one of them. Overall, my impression is that he is scrupulously honest in stating his full position on MMT. Importantly, he’s done a lot of research on MMT, including both blogosphere and academic writing.  The footnotes to his posts demonstrate that. So in my opinion he has some bed rock credibility as an observer of MMT who has actually scrutinized it quite closely. That doesn’t mean he’s always right. And it doesn’t mean MMT has considered his criticism seriously. Anyway, here we have an opinion about MMT coming from somebody who in many ways is a supporter of MMT, but in this case describing one characteristic of its style that he perceives as unattractive. And this is a recent comment.

I responded to Steve Waldman’s comment at the Asymptosis blog, by constructing the equation S = I + (S – I) as my own algebraic translation of his thought process on this issue. This was not spontaneous, in that I’d made such a connection previously. And Waldman himself had articulated the same point previously, in a comprehensive blog post in April, 2011, that critiqued MMT more generally. But his Asymptosis comment was the raw material for the equation’s most recent incarnation, from my perspective.

I explained the rationale in several more comments at the same Asymptosis blog post, which I’ll reproduce here as well:

This is what I wrote on February 1, 2012, in response to Steve:


“Steve Waldman,

For reference only (you know this cold):

S = I + (S – I)

Private sector saving = investment, plus the change in private sector net financial assets.

You’re saying that you can increase S as a result of increasing I rather than (S – I), which is true, and very important, as you explain.

MMT is not entirely innocent of the “rhetorical overreach” you suggested, as I think you clarified in your second comment.

MMT frequently adopts short hand mode, in which it conflates terminology, describing what is actually (S – I), as saving.

This is unfortunate, in that the distinction is very important, not only in positive terms, but in normative terms.

To the degree that MMT associates saving with (S – I), it leverages the message of its normative view, that the private sector is driven by the motivation to satiate its desire for net financial asset accumulation. This general view has great associated consequences for its normative view on deficit financing, etc. etc.

This can be problematic from a logical perspective, and therefore from the perspective of normative balance.

It is not clear at all that the desire for net financial assets is in fact the driving force, in my view.

It is more the case I think that the true force in this Keynesian context is the desire for saving per se – i.e. for S.

And S can be mapped into two components, as above, “I” and (S – I).

To the degree that the “I” component is submerged in a lazy conflation of the meaning of “saving”, the (S – I) outlet for saving desires is the one that becomes accentuated in the MMT story.

The more accurate story, in my view, is that the (S – I) outlet is a critical point of adjustment, as per MMT, but that it is VERY importantly not the only critical point of adjustment.

So I think for the most part I may be agreeing with and reformulating your comment(s) in my own words. And you’re right – the merging of business and household sectors into the private sector is a catalyst for lazy and perhaps overly convenient interpretation. After all, should we not hesitate just for a moment before blithely ticking off corporate debt and equity as “negative financial assets” within this private sector consolidation?

(To give MMT its due, they have published sub-sector analysis in this area, although it’s not so prominent in the inherently fast marketing world of the blogosphere.)

“There is no “to the penny” sort of accounting relationship between household-sector financial saving and government issue of NFA.”

Right, and again, this statement is probably implicit in MMT’s recognition of the difference between what it classifies as “horizontal” versus “vertical” financial assets. The horizontal nexus includes the financial asset interface between business and households, and there is real investment off to the left of that interface. But your point is a fundamental one that is submerged in the frequently applied MMT shorthand version of the world. MMT should be given credit for understanding this, but not for advertising it.

Perhaps we should start to look at the problem as one of portfolio mix, and portfolio balance, where the two major categories for the application of saving are investment and net financial assets. I see MMT driving a theme of persistent marginal adjustment through net financial asset accumulation; not one of deliberate portfolio balance between investment and net financial assets. And in doing so, there is a tendency toward open ended flow arguments that facilitate persistent budget deficit rationales, rather than targeted stock balance arguments, as a matter of overarching investment and saving strategies.”



Now, you can agree or disagree with Waldman’s comment or my response, or both, but that is the record of the recent development of S = I + (S – I).

Let me make a few further comments regarding my response noted above:

a)      The equation in question is the mathematical translation of somebody’s observation about MMT. It was specifically for the purpose of illustrating, in algebra terms, the point that Steve Waldman was making verbally. He hasn’t been the only one to make such an observation. And I did not think that math type connection up on the spot. But it was an opportunity to record an instance of such a connection, using the thought process of somebody else as the input and reference.

Waldman had made an earlier comment in the same thread as lead in to the one above. Moreover, he did a complete post on MMT on April 11, 2011, in which he raised the very same point. For completeness of the record, I have included those two additional excerpts below, with corresponding links.

Again, the reader can be the judge of how my initial comment reflects what Steve was saying. You may not agree with either one, and/or you may not agree that the equation represents a good translation of his point, but you can be the judge of that. And I don’t disagree with his observation about MMT at all. As I said to him:

“So I think for the most part I may be agreeing with and reformulating your comment(s) in my own words.”

b)     Given the general discussion on saving that has taken place since then, it is important to point to the assumptions that were documented as underlying S = I + (S – I). As described above, this was the standard 3 sector “Sector Financial Balances” (SFB) model of MMT. The household and business sectors are consolidated into a single private sector in this 3 sector model. It is private sector S that is being used in S = I + (S – I). Thus, it is essential to note that the context for the framing of the variable S in this equation was very specifically a 3 sector model of the type most frequently used in MMT’s sector financial balances model and in the national accounts equation from which that 3 sector model is derived. The parent framework specifically is a 3 sector model, not 1 or 2 sectors. And while MMT has occasionally made reference to a 4 sector model in its writing, such a 4 sector partition is not its standard form of presentation and it’s certainly not the context for MMT’s typical reference to private sector saving. In this equation, as in the 3 sector SFB model, S is the very specific variable that is the designation for private sector saving.

I notice there has been occasional discussion in the recent debate that the term “private sector” in general might refer to an MMT 2 sector model. I must say that I mostly recall the usual relevant term in such a 2 sector model being “non-government”, complementing government, and that I don’t recall much use by MMT of “private sector” in a 2 sector SFB model context. And I would think that the rationale would be to avoid ambiguity with respect to analysis referencing either 2 or 3 sector models.

c)     In any event, the (domestic) private sector is definitely the construct that corresponds to Waldman’s description regarding the decomposition of same. It clearly accords with Waldman’s intended meaning, and Steve himself didn’t seem to disagree with this when he made the following remark in a separate post of his own, written shortly after, on February 12, 2012:

“JKH — wildly off-topic, but I returned too late to the discussion to congratulate you on how great your S = I + (S – I) discussion was. (Michael Sankowski did a service by highlighting it.)”


The two links he included there are:



Once again, whatever you may think of the equation or the underlying view, these are the facts as to the recent development of the equation in question.

d)     Finally, I repeat a point I made in my comments to Waldman, above:

“To give MMT its due, they have published sub-sector analysis in this area, although it’s not so prominent in the inherently fast marketing world of the blogosphere … MMT should be given credit for understanding this, but not for advertising it.”

Thus, I’m quite familiar with the scope of MMT’s inquiry into the subject matter as it concerns saving. This becomes relevant to my response to Randall Wray’s post on March 4, 2012, which is the subject of section 4.

I made some additional related comments at the same Asymptosis post, which I wanted to extract here also. The next one was highlighted by Waldman himself, noted in c) above. I’m gratified he picked this, because it goes to the underlying logic behind the equation as a tautological expression. It might be interpreted as meaningless by those who gave little thought to it or who didn’t probe its origins. A number of commenters have apparently reacted that way. Indeed, it has been depicted that way by Randall Wray. Based on the quoted comment that follows here, I hope some readers may consider the reason for the expression in a different light, even if they disagree with the underlying rationale for the expression. Indeed, at the time, the commenter (Vimothy) to whom I responded here inquired whether or not I had made a mistake in presenting an expression that that seemed so oddly redundant. This is what I said:


“S = I + (S – I) is not a mistake.

And rather than being a small error, it’s a critical analytical decomposition of private sector saving, most particularly in the context of interpreting the (sometimes ambiguous) MMT intended meaning of saving. Not understanding it can be the source of much confusion around MMT. The fact that you thought it was an error might hint at some. Moreover, it’s the saving algebra that encapsulates what Steve W. is driving at in his comment. That’s partly why I started my opening comment with it.

It’s an obvious identity/truth from the algebra alone, just as a self-referencing rearrangement of terms.

But it’s much more than that in the context of deconstructing MMT’s approach to the interpretation of saving.

Before getting to that interpretation, here is the national income accounting/algebra that gives background context:

C + I + G + (X – M) = C + S + T

(S – I) = (G – T) + (X – M)

That’s one permutation of “sector financial balances”, as a derivation of national income accounting.

That particular permutation says:

Private sector net financial asset accumulation = the government deficit plus the current account surplus.

This is the SFB formulation that defines NFA (actually the flow change in NFA) as the left hand side – it’s the excess of private sector saving over the amount required to fund investment.

Private sector NFA is MORE than just the government deficit in an open economy. Only in a closed economy does it collapse to (G – T) as you suggested. The always correct definition starts with (S – I), not (G – T).

Now, in parallel, but at the core of the relationship between private sector saving, investment, and “net financial asset” accumulation:

S = I + (S – I)

Just looking at the algebra, this is an obvious identity (or truth, really) just as a result of being a mere rearrangement of entirely equal terms. But it has considerable meaning in the context of how the pieces of the saving puzzle come together.

Using the earlier definition of net financial asset accumulation from the sector balances equation above:

Private sector saving = investment plus private sector net financial asset accumulation

That equation is the foundation algebra for illustrating the point Steve W. was making, in my view, as per my explanation above. I agree with SW’s point, as per my comment. And I also covered this same aspect (among many others) in recent comments under previous posts here at Steve R.’s site.”



So that again is the explicit mapping of this equation to the 3 sector MMT SFB model, and its parent national accounting framework. And this also becomes relevant in my response to Warren Mosler’s comments (further below) when he participated in an extended blog discussion on the same day that Professor Wray put out his own response.

Back to the Asymptosis comment stream, there is one more comment of mine that I wanted to document here:


“You may be thinking of the NFA position of “non-government”, which combines private and foreign sectors. THAT position is identically equal to the government deficit (or debt in stock terms). That’s both a convenient and useful consolidation, that MMT defines as a high level conceptual reference point for the communication of its core message, about government as a marginal supplier of desired saving to non government, and that is a very valid point. But it can obscure what lies beneath.

EACH sector has its own NFA position. That’s important, and again goes to the crux of the potential confusion around this issue.

Moreover, the household sector has its own NFA position, and it is VERY substantial, because it frees up the household savings analysis from the consolidated treatment of corporate liabilities and equities as negative financial assets. This is the crux of Steve Waldman’s point. And the household sector position is VERY important in the interpretation of the Fed flow of funds analysis, which in itself is an essential element in the overall coherence of the subject matter.”



So there a distinction is drawn between the intended 3 sector mapping for this equation and the alternative MMT 2 sector presentation. Again, S = I + (S – I) is the same private sector S that appears in the MMT 3 sector financial balances model, which itself is derived from national income accounting. Moreover, I’m confident that I’ve specified this precise meaning of S on every occasion where I’ve engaged in extended discussion about it.

The exclusive application to the 3 sector model was also confirmed most recently on February 28 in the following MMR post (prior to the response of both MMT leaders), where I mapped the intended meaning of S = I + (S – I) into a 3 sector graph that Paul Krugman used to make a point about saving and investment in the United States:


It couldn’t be clearer. Paul Krugman’s graphical illustration of S = I + (S – I) is congruent with 3 sector model – the same one used by MMT in deriving the 3 sector financial balances model. Hence the title of the MMR post, “Paul Krugman does S = I + (S – I)”.

So those are the facts to the background of the recent blogosphere appearance of S = I + (S – I).

Some who have been active in criticizing the form of the equation apparently were passive in their research into it. Those who have taken a strongly negative position haven’t necessarily been paying attention to the subject they’ve been predisposed to dismiss. I have no interest in immediate persuasion on conversion on the matter. A number of commenters have noted the same tendency for the language and the meaning of the word “saving” to have been confused in the context of MMT’s use of the SFB model (e.g. Waldman, JKH, Roche, Ramanan, Carney, and others). This is a shared observation. It is a noteworthy stylistic characteristic of MMT. The equation S = I + (S – I) highlights that incongruity. The equation itself is not going away. It is what it is.

For my own part, I’ve also said little about any type of “desired” relationship between I and (S – I) in explaining the meaning of the equation. Evidence for a trend relationship is a matter of recorded history. That aspect, while fundamental, is separate and beyond that of S = I + (S – I) as a more basic clarification of language and logic. But there’s no direct normative inference just from the equation S = I + (S – I), in terms of desirable private sector financial behavior, consolidated or deconsolidated. As criticism, that allegation is mostly a straw man. However, a normative argument, using the form of the equation as an analytical point of measurement reference, becomes eminently arguable, a point which MMR is already picking up on. It is a second step.

Additional References

Here are Steve Waldman’s fuller comments (excerpted) from the discussions at Steve Roth’s Asymptosis blog, combined with an excerpt from Waldman’s own Interfluidity post from April 7, 2011, originally critiquing MMT:



“It is perfectly possible to hold the international balance constant, have the government reduce debt, and have “people” save more. “People’s” financial savings consists of claims on firms and claims on government. If I perform some work for a firm that (however infinitesimally) increases the firm’s real economic value, and I accept as payment a share of that firm’s stock, I have performed the economic act of saving, and increased the net saving of “people” — of the household sector. Net private sector financial assets have not increased: my “savings” is the firms’ obligation; the household sector’s surplus is offset by the business sector’s deficit. But much of what we call saving is exchanging real resources for claims on the private business sector. And as long as the private business sector doesn’t entirely squander those real resources, that act contributes to macroeconomic S. If the private business sector does squander the resources, then while I still perceive my contribution as “saving”, the value of macroeconomic S = I does not increase, and my claim amounts to a transfer from other shareholders of the firm. But even in today’s terrifyingly %$&*-ed up world, people make a lot of productive contributions to private business in exchange for financial claims. The act of doing so increases S = I, but has no effect on “net private-sector financial assets” and requires no government accommodation.”



“But much of what we call saving is exchanging real resources for claims on the private business sector. And as long as the private business sector doesn’t entirely squander those real resources, that act contributes to macroeconomic S. If the private business sector does squander the resources, then while I still perceive my contribution as “saving”, the value of macroeconomic S = I does not increase, and my claim amounts to a transfer from other shareholders of the firm. But even in today’s terrifyingly %$&*-ed up world, people make a lot of productive contributions to private business in exchange for financial claims. The act of doing so increases S = I, but has no effect on “net private-sector financial assets” and requires no government accommodation…

There is no “to the penny” sort of accounting relationship between household-sector financial saving and Government Issue of NFA … but claiming “saving” is impossible as a matter of accounting without a government deficit is a bait and switch, a game played with definitions by rhetorical confusing household sector financial surplus with an aggregate private sector surplus. – [this section quoted in full, above]

It is perfectly fair, and quite possibly correct to argue that in the present economic environment, due to absence of demand and business-sector risk-aversion and broken household balance sheets, that household saving will not be accommodated by increased claims on expanding value in the business sector, and so the best way to repair those balance sheets is for the government sector to run a deficit. I very much support the use of government balance sheets to help the US household sector save. But that is because I think the other path will not work very well right now, not because as a matter of logic and accounting no other path is possible.”



“MMTers sometimes blur the distinction between “private sector net savings”, which is necessarily backed by public sector deficits or an external surplus, and household savings, which need not be. In doing so, MMTers rhetorically attach the positive normative valence associated with “saving” to deficit spending by government. This is dirty pool, and counterproductive. The vast majority of household savings is and ought to be backed by claims on real investment, mediated by the liabilities (debt and equity) of firms. There is no need whatsoever for governments to run deficits to support household saving. When household savings increases, an offsetting negative financial position among firms represents increase in the amount or value of invested assets, and is usually a good thing. Household savings is mostly a proxy for real investment, while “private sector net financial assets” refers to a mutual insurance program arranged by the state. It is a category error to confuse the two. Yet in online debates, the confusion is frequent. Saving backed by new investment requires no accommodation by the state. It discredits MMT when enthusiasts claim otherwise, sometimes quite aggressively and inevitably punctuated by the phrase “to the penny”.”


Here are several examples of other commenters engaging along the same general theme:






Игры рынка:




3. Blog Discussion

W. Mosler – March 4, 2012

Two MMT leaders – Warren Mosler and Randall Wray – participated in the debate. Mr. Mosler offered his thoughts in an extended discussion at the winterspeak blog:


Much of the discussion concerned itself with the meaning of saving in the context of the MMT analytical framework. Mosler approached the question of the S = I + (S – I) equation with scepticism. He noted that without further information, not much was to be said about it. He did have a question regarding the purpose of it.

As review, the equation S = I + (S – I) is a decomposition of private sector saving S, as defined in the 3 sector SFB model and its parent national income accounting model. From one of my quoted comments at the Asymptosis blog:

“It’s an obvious identity/truth from the algebra alone, just as a self-referencing rearrangement of terms.

But it’s much more than that in the context of deconstructing MMT’s approach to the interpretation of saving.”

To answer Warren Mosler’s question directly, the purpose of S = I + (S – I) is clarity of terminology, expression, and understanding. MMT may insist it has been clear on these issues all along. But even if that were the case, there is clearly a gap between that belief and the comprehension of those who are actively interested in the subject. Here there is evidence of considerable fog within the full discussion in which Mosler participated. His question is answered partly by the elephant in the room, which is the confused nature of the discussion itself. There was much speculation as to meaning of the S that was the subject of the discussion. Yet the correct S specification falls directly out of the standard 3 sector SFB model that MMT embraces and uses ubiquitously. It is for that reason that I’ve extracted a fairly lengthy sequence of comments from that discussion, including my own remarks along side.

Mosler asked about the public purpose of the S = I + (S – I) expression. It has to do with clarity of exposition as a catalyst for clarity in discussion and debate. There needn’t be a perpetual process of blogosphere ground hog day discovery in the meaning of such terms as saving. These terms have been around for a long time. We all know that the sector approach is important to MMT. It should be important to everybody. But is it necessary to flirt with confusion of terms by adopting non-conventional twists in the meaning of saving?

Mr. Mosler was apparently unaware that S = S + (S – I) had been specified as a 3 sector equation. This was documented (as summarized in section 2) and frequently referenced in lead up discussions over the month of February, 2012. At one point, informed by another discussant that it was almost certainly an intended 3 sector application, Mosler maintained it had nevertheless been written as a 1 sector model. That is incorrect. I know that because I wrote it. Moreover, nobody is in a position to say otherwise about such an equation, without specific information supporting such a contention. There is nothing that precludes such an equation being specified as a 3 sector model, when it has been specifically designated to be so. The S in that equation is the 3 sector S, not the one sector S. As Neil Wilson noted in the discussion, it is meaningless even to consider such an equation in application to a 1 sector model. (The (S – I) factor becomes identically zero in such a model.)

On a related issue, some have debated and did so on this occasion whether or not expenditure on investment constitutes spending absent corresponding saving at the macro level. If not, then S = 0 for the global economy. But if that were to be the case, then in the US for example the household sector would hold approximately $ 40 trillion in net financial claims, constituting its cumulative saving in stock form. At the same time, the global economy would have no saving, the US as a whole would have dissaved approximately $ 2 trillion. This is a truly chaotic porridge of saving mathematics. It is all wrong in my view. But there are indications that this is the intuitive preference of MMTers. Mosler alluded to this definition of saving as his preference. In my view, the correct definition of saving in terms of its relationship to investment is the one contained in the Harless piece and in the language used in my comment that John Carney used in his CNBC post. Moreover, the abandonment of the meaning of the term saving from any related expenditure on investment departs from the standard meaning adopted in economics.

And it’s been noted that the conflation of the terms S and (S – I) “works” if I = 0. It also “works” if the definition of saving in general is adjusted to completely exclude the notion that saving required to fund investment is not saving at all, but just another form of income spent as if on consumption, which means global saving must be zero. Mosler and some other MMTers seem to indicate a predisposition to this sort of interpretation. There seems to be a relaxed informality about flexibility of basic definitions and language use. This was Marc Lavoie’s theme more broadly (see below). MMT’s intuitive preference includes this sort of modification to the conventional definition of saving. And if that is the case, then it is no wonder that the net saving as the intuition for saving per se bubbles to the surface in the form of proactive language transformation and confusion. And when that is the case, it explains the issue that Waldman and others have observed.

Thus, the potential conflation of the meaning of saving and net saving may be connected to an accompanying maverick divorce of saving from investment. It is hard to know for sure. But all of this serves up dysfunctional language malleability. The distinction between the meaning of S and the meaning of (S – I) in a multi sector model becomes desperately conflated with the S in a single sector model.

It is not possible to develop a coherent world measure of saving and investment by stipulating that global saving is identically zero. That is intuitively repulsive. It begs the question of additional motivation for the MMT SFB stylistic language emphasis. Is it any surprise that there is ambiguity around what people perceive in the mixed language model of saving and net saving around that model, when what lies beneath is such an intuitive preference model? Intuitive preferences, if held, mold interpretation and language and logic and explanation, despite any protests to the contrary that “you see we really know what we’re talking about”. “I suspect they know what they’re talking about. But does anybody else understand or want to understand the same thing in such inconsistent terms?” Again, there are two components involved in the confusion. One component is the difference between saving and net saving as in the difference between S and (S – I), and the other component is the actual definition of S itself. This is literally a conflation of two confusions.

Mosler raises the issue of “purpose” in his comments. The purpose is clarity of exposition. As far as the observed MMT conflation of the idea and language of saving, what’s not clear is what public purpose is served by such a thing. One might argue that  effort toward precision and consistency would be beneficial to the rest of the community that is attempting to understand this subject on common ground, as a basis for shared comprehension and moving forward collectively in policy discussions.

This language theme can be taken much wider as it applies to MMT than just the use of the word “saving”. There is the larger realm of the way in which the monetary system is described by MMT. This broader scope has been the object of critical scrutiny in several recently published academic reports, including those of Marc Lavoie and Brett Fiebiger (see further below).

But apart from all this, Mr. Mosler offered some interesting remarks on the subject as a whole. The following sub-section includes some of those comments, along with others that preceded or followed, and my accompanying remarks and observations.

Again, the reference is here:


Selected comments are in quotation marks. Warren Mosler’s comments are preceded by his name. Otherwise, names have been omitted:

“Once again, it needs to be emphasized that one cannot derive substantive quantitative and causal conclusions about which transactions are or are not the backbone of some larger pool of transactions, or about which activities and operations do or do not “drive” the economy, simply by examining identities and accounting logic.”

The term “backbone” was used to emphasize the physicality and relative size of the outstanding investment stock, compared to private sector net financial assets, as noted earlier. Relative magnitudes in the US have been estimated and included in various comments posted in connection with this subject matter. Causal arguments exist, but are secondary to the intended purpose in using that particular term.

“In a state capitalist economy with a very large government, G might very well dwarf I and X and be driving the creation of real wealth.”

I don’t know what “real wealth” means, but yes the inversion of the proportions might be possible in an MMT oriented counterfactual, perhaps. But that’s not the case in the real world we have now.

“S = I + G – T

Suppose the actual quantities instantiating the above equation are:

$7 Trillion = $2 Trillion + $6 Trillion – $1 Trillion.

In this economy, JKH’s equation is still true, because it is a logical tautology:

S = I + (S-I)

But is I the backbone of S in this economy? Clearly not.”

Quite right, in terms of a counterfactual for relative magnitude; same point at above.

“Our own economy is not like the imaginary economy in the example I just gave, so private sector investment might be the “driver” of economic growth.”


“Y = C + I + G + X – M.

Does this now prove C is the backbone of Y? No. But it does point to a big elephant in the room that has been left out of the discussion.”

I wouldn’t analogize C as the “backbone” of anything; perhaps “arms and legs” if one insists on pushing the analogy further, but I wouldn’t.

“Contrary to Monsieur Say, people and firms don’t typically produce things first, with no desire to consume the product themselves and no reliable information about whether someone else is eager to consume it.”

No dispute there, but we’re exploring the nature of saving. And a key point here I think is that what MMT characterizes as a demand for net financial assets isn’t really that at all. It’s a demand for saving that at the margin can be satisfied at times by the government provision of net financial assets. The margin isn’t the total, and the fact that there appears to be a demand met by net saving ex post shouldn’t obscure the fact that there is an underlying demand for saving more generally, including the type of saving that matches to underlying investment in equal amount. The much larger element of total saving is the latter – at least on average and as measured through saving accruals that accumulate to stock form over time.

“MMTers could interpret this as you saying that the household sector is ultimately – for a lack of a better word – the owner of the means of production.”

That is factually correct. Net corporate wealth translates to household net worth. The household sector includes all wealth as it accrues to individuals (there may be some lower level statistical classification issues or data irregularities).

“The whole S=I+(S-I) thing isn’t some “gotcha” moment. It’s a clarification that I think some MMTers are getting way too bent out of shape over”

Right; clear thinking and articulation are continuously desirable, not a one off “gotcha”.

“If during some period of time some group of people is devoting an increasing share of their income to the holding of money and other financial assets, and less to the purchase of consumption goods and capital goods, the question is what motives are causing them to do this, what effects on other aspects of economic activity do those motives have, and what policy changes or other systemic changes would or would not influence those motives.”

That’s a departure point for analytical investigation. But at the level of households and individuals, it has to do with the mix of I and (S – I), not just a focus on (S – I) alone.

“Isn’t MMT the side that has insisted in focusing on “net saving” and S-I? It seems to me that this is a very common-sense use of the term “saving”.

Common sense suggests S be defined the same way it is in the SFB model that includes both S and (S – I), using the same definition of S as in the parent national income model. S is private sector saving. S and (S – I) can’t both be referred to as saving in any sensible use of language. Such is a fundamental contradiction in language and logic.

“SRW original criticism, made quite a while ago, was that MMTers sometimes confuse or conflate net saving with saving.”

Yes. While SRW was not the first to make this criticism, he did so most effectively.

“Now, it’s fine if you wish to go with a new definition of saving – although, it would be nice if MMTers were more upfront about this – but what it implies from the point of view of macroeconomic relationships is totally incoherent. Do you think that aggregate saving and investment is possible? If the answer is “yes”, then you need the proper definitions of the terms.”

Yes. My point, better said.

“You will of course have verified the particular definition given that your extensive education will surely have taught you that ‘saving’ is a humpty dumpty word in economics.”

The saving S in question, as it relates to S = I + (S – I) is a very specific S, the same one used by MMT in its 3 sector SFB model.

“Therefore, S = I. Crystalline in its clarity.”

This Harless 1 sector model specification has absolutely nothing to do with the issue of how S is specified in S = I + (S – I) when S refers explicitly to the S of the 3 sector SFB model ubiquitously invoked by MMT.

“I don’t think the current national accounting convention that includes all capital expenditures under the category of “saving” is a basis for serious economic analysis at all. It’s a strange terminological choice that, in my estimation, deliberately obscures economic reality by combining two very diverse economic activities under the same card in a three-card monte game. So when MMT highlights net national saving, that is S-I, and tends to use “saving” to refer to that rather than to gross national saving S, then I think MMT is on the side of both common sense and a clearer picture of economic reality.”

That’s the view that snuffs out the very idea of global saving, as noted above. This seems to be an undercurrent of MMT intuition. But it is completely inconsistent with the 3 sector SFB specifications and definitions now commonly used by MMT, as derived from national income accounting.

“To side with Ramanan et al, the accounting identities are the accounting identities. We should stick with them and know what they are, but we just need to remember what they are not as well.”

They are what they are. If used at all, they should be used consistently, not as an option to be abandoned or distorted when convenient.

Enter Warren Mosler into the discussion:


“Nominal savings are financial assets – like dollar balances, which are, functionally, tax credits. Real savings is deferred real consumption – like seed corn. You can have either one without the other. Also, I’m new to the s=i + (s-i) thing. Doesn’t s-i= 0? So what does adding ‘0’ illuminate?”

I’m not a fan of that use of the term “nominal”. It’s fraught with ambiguity of general meaning.

(S – I) is zero identically only in a 1 sector model. This is not a 1 sector model.

From this point, it becomes clear in the discussion that there is obvious confusion within the ranks of MMT devotees on this general issue of saving. The first dozen comments or so indicate this fog starting here:


It should be evident to those who paid attention to the original discussion of S = I + (S – I), as it appeared prominently via the MMR blog and earlier, that we’ve been talking about S as it is defined in the MMT 3-SFB model and in national income accounting from which it is derived. (The Mike Norman blog, to its credit, did accurately track the development of S = I + (S – I), under the scrutiny of Tom Hickey and colleagues.)

But only one MMT commenter seems to have considered it in this discussion:

“Warren, if (S–I) = (G–T) + (X–M), then why would S – I always be zero?”



“If you keep it all nominal, G – T = net financial assets of the non govt sector. That’s the 2 sector model. But it begins with the one sector model, where S = I, and S – I=0. In the two sector model, S – I= domestic ‘savings”/accumulation of nfa, and G – T= Govt. ‘savings’/ accumulation of nfa. And total savings of NFA = 0. Going to a 3 sector model, X – M is just foreign savings, defined as net financial assets. So in that 3 sector model where S – I is domestic holdings of financial assets only, and X – M is non resident holdings only, the total domestic plus non resident = (G – T). With sector analysis of ‘inside money’ the sectors add to 0. Expanding the number of sectors doesn’t alter the ‘fundamental maths’ of the 1 sector model.”

Certainly, S = I in the 1 sector model. That is trivial.

But such an identity carries over to a two sector model only if you precisely define S as that exact equivalence. For example, if you define government saving to be (T – G) and non government saving to be (Y – T – C), and define the sum of the two to be S, then S = I in the 2 sector model.

But such a forced definition of S in a 2 sector model is a complete departure from the use of the same symbol as the standard definition S of private sector saving in the 3 sector model commonly used by MMT.

(Ironically, S = I in such a 2 sector model only if it is recognized that government can save. Yet there is considerable MMT language even rejecting such a concept (e.g. Mitchell), and the same holds for the three sector view. It always struck me as bizarre that the MMT SFB methodology turned against the parent national accounts terminology, by denying legitimacy (in the form of MMT approval) to the idea that the government could save or dissave. This particular language policing seems to stem from the desire to emphasize the currency issuing power of the government and its associated capacity to generate unlimited quantities of “net financial assets”.)


“So S= I + (S – I) is a 1 sector model identity, as there either is no X or M or G or T or it assumes that they equal 0 and therefore those sectors are of no consequence for this point of analysis? This all reminds me of the long history of thought including smith, ricardo, marshall, keynes, hayak, von mises, keynes, sraffa, lerner, friedman, lucas, modigliani, and all the rest who raised points of endless debate that get resolved via MMT.”

This includes the error already noted. This is not a one sector model; the variables S and I have clearly been specified as 3 sector variables, and this was documented repeatedly in lead up discussions. That documentation commenced formally at the Asymptosis blog, noted above, and was subsequently channelled via the MMR blog (and noted at the Mike Norman blog) and presumably monitored by MMT followers who gathered there. The S being used in S = I + (S – I) is the same S that is used by MMT in its 3 sector financial balances model.



“It may not be what they are ‘talking about’ but what’s written is a one sector model. Adding sectors is adding degrees of freedom to the model, and this model doesn’t go there.”

It’s not a one sector model. It would be useless in such a specific context, as (S – I) is identically zero in that case. It was developed with specific reference to the 3 sector financial balances model used by MMT, and with S defined very specifically to have the same meaning as it does in that model.

The explicit connection between the equation and 3-SFB has already been explained numerous times, so there should be no need to repeat it. But here again is a summary of the direct relationship:


(S – I) = (T – G) + (X – M)


S = I + (T – G) + (X – M)

S = I + (S – I)

Where (S – I) is substituted from the first equation above

This is a tautological decomposition of private sector saving, as it relates directly to the SFB model and the SFB parent national accounts framework. The equation in its crude form captures the two components of private sector saving – a component that offsets investment, and a residual net financial balance component. The 3 sector SFB MMT model highlights an (S – I) component relating the private sector financial balance to those of the government and foreign sectors.

“Consolidation” of a 3 sector equation in more compact form doesn’t mean the 3 sectors have been eliminated. And it’s an equation, not a model. The equation can be used to make any argument that one wants to make of it, but the equation itself has no inherent model or ideological content. It is open source, and free to use, from that perspective, just as is SFB.

“Warren, specifically on the issue of saving, my take is that MMR agrees that (S-I) is a measure of net financial assets accumulation in the private sector, but they claim that it is not savings per se. Savings is “S”, not (S-I).”

That’s right

“This is why MMR is particularly fond of this type of graphical representation, where I and S is treated separately: http://monetaryrealism.com/paul-krugman-does-s-i-s-i/




“Definitions are always ‘for further purpose’ “

OK, but so are most things, I think.

But it’s a reasonable question, and I’m responding to it throughout this section.


“Also, the s=i+(s – I) is a one sector model. Yes, it applies to multi sector models, but so do all one sector models, much like the one bank model also applies to multi bank universes, etc.”

It’s not a one sector model, as explained. It would be irrelevant as a one sector model, since (S – I) is identically zero in such a model.


“Nor can I be anything other than equal to S as a matter of accounting, whether the analysis is in nominal or in real terms.”

This is incorrect, when you acknowledge the specified private sector S as it’s used in MMT’s 3 sector SFB model, as documented above.


“In nominal terms it just means there is an asset and a liability, one side accounted for as savings and the other as investment.”

Actually, that’s not the correct accounting. Household net worth is not a financial claim and not a liability for the household unit.


“The best way to think of it is to think of savings as the accounting record of investment.”

I disagree it’s the best way to think of it. The accounting record of savings (stock) is the net worth or equity account. The accounting record of investment (stock) is the investment account. These accounts are on opposite sides of the balance sheet. See my discussion in section 1.


“So subdividing sectors can be a tool of discovery.”

“Private sector consolidation within SFB is not an indicator of saving per se” (me)

“Why do you care?”

I care pretty much for the reason you stated yourself, but in a more general way – subdividing saving can be a tool of discovery.


“For purposes of my analysis I focused on net financial assets, which are commonly also called ‘savings’ though ‘savings’ is clearly used in other ways at other times by others.”

This is representative of the flexible MMT approach to the meaning of the term “saving”.

It is apparent from this overall discussion that MMT devotees themselves seek clarity on this issue, so it’s hard to understand why MMT types would continue to supply language that confuses. On this issue, there is obvious confusion within MMT follower ranks. And at this point we might add that to ridicule S = I + (S – I) because of its tautological appearance makes about as much sense as ridiculing SFB as a rearrangement of national income accounting, which is all it is.

Finally, in this discussion, “winterspeak” objected to my earlier definition of saving as income not spent on consumer goods. (Perhaps I should have specified that services are included as spending, although that wasn’t the objection.) The objection had something to do with the timing of consumption of apples or something like that. This is a lower level point. We know that economic and financial accounting is delineated in time periods, so the issue of crossing multiple time periods is well understood in the casting of definitions. Of course consumption as a flow refers to what is being consumed in the current accounting period, because flow accounting is period specific. Stock accounting is point in time specific. And of course there are various options for delineating things like consumer durables as consumption or investment. And there are further issues of data and time classification. But this detail is not the stuff of the concept itself. We’re interested here in categorizing the forest, not the trees. Andy Harless for example had the good sense to ignore this lower detail in his account of saving and investment.

I’ve noted the Andy Harless post on saving and investment several times. He is specific on constraining the sector structure for that model. It eliminates both government and foreign sector imbalances, which allows S = I. He avoids the sector complications inherent in the derivative MMT SFB model. And while much understanding flows from the Harless exposition, it has absolutely no explanatory power relative to the S specificity of the 3 sector model. And therefore it is insufficient in the same way relative to the equation S = I + (S – I), which was written as it relates to the 3 sector model. The S used is the private sector saving S in a 3 sector model, not the consolidated single sector S that Harless uses.

The same blogger winterspeak more recently has developed some sort of trademark notation for HPM. This appears like some kind of functional notation that seems to conflate characteristics of asset, liability, and equity all at once. Look, this whole area is a matter of simple, straight forward double entry book keeping. When the government deficit spends, it creates incremental income and saving for the non government sector in the national accounts and micro accounts, ex post facto. That is an income statement phenomenon. The settling of funding accounts in HPM or Treasury bills and bonds is an entirely separate matter. It is represented correctly in a properly constituted flow of funds statement, with corresponding beginning and end of period balance sheets. One of the entries must be to a net worth or equity account for some agent. It is impossible for it to be otherwise. But HPM itself is an asset to whoever holds it, and does not in itself end up in this type of net worth or equity account.

This is an example of a more general methodological tendency for MMT to grasp at sections of double entry accounting and mold them into MMT specific accounting clusters with potentially ambiguous results. There’s been a recent tendency for example to treat “NFAs” as if they were government distributed fruit baskets or something. “Net financial asset” is a generic accounting consolidation term, which existed long before MMT got hold of it. Its use as an accounting measure extends well beyond MMT’s application of it in the government context. If one actually believes that NFA is only an MMT designated government item, it is little wonder that the natural household sector position in NFA has been submerged in common discourse on MMT blogs, with resulting confusion. See section 1.

4. NEP Blog

L. R. Wray – March 4, 2012

On March 1, 2012, John Carney of CNBC published a blog post that included an MMR blog comment of mine quoted in full. Carney’s post is reproduced in its entirety below:



More on Saving (Note: This Gets Wonkish)

(Carney): “Over in the comments section of Monetary Realism, the mysterious and ubiquitous commenter JKH spells out how we should think about the concept of “saving.”

Since this is central to the accounting identity dispute I was writing about yesterday, I thought it deserved to be reprinted in full here.”

(JKH): “The correct economic definition of saving is disposable income not spent on consumer goods.

Individuals often save and deploy their saved income into financial assets such as stocks, bonds, and pension funds. (They may also invest saving in newly constructed residential real estate, which is separate from financial asset acquisition, of course.)

Such financial assets represent direct and indirect claims on corporations and governments.

The recent blogosphere kerfuffle about saving arose in part because MMT embraces the sector financial balances model (SFB), which features the consolidation of household and corporate sectors as a unified private sector. The model treats financial claims on corporations as negative financial assets for corporations, so the consolidated result is that household saving deployed in such financial assets makes a zero net contribution to private sector saving after counterparty corporate netting. At the margin, such deployment of funds becomes a net financial asset for the household, a net financial liability for the corporation, and a net financial asset wash for the private sector as a whole.

The private sector as a whole adds net financial assets when either the household or corporate sub-sectors deploy funds from saving into the acquisition of financial claims on either the government or foreign sectors. This can occur for example with the purchase of government bonds or foreign financial assets. Saving thus used can be identified as a particular subset of saving, but by no means does it account for saving per se.

Private sector consolidation within SFB is not an indicator of saving per se. Consolidation obscures the core underlying saving dynamic of the private sector.

All private sector saving can be condensed, in effect, to a measure of household saving alone — by projecting the cumulative value of corporate saving onto the household balance sheet. This occurs when household financial claims on the corporate sector are valued by the marketplace to reflect those underlying corporate saving changes. E.g., the value of stocks tends to increase when corporations save and invest in real assets. The issue there is one of valuation translation, rather than the conceptual correctness of corporate saving being reflected at the separate level of household balance sheets as well.

MMT alludes on occasion to a definitional change for saving that it deems desirable for purposes of delineating saving as portrayed in the sector financial balances model. It sometimes describes the act of deploying the proceeds of normally defined saving into physical asset investment as spending, without associated saving. This revised interpretation of saving treats spending on consumer and investment goods similarly, with zero associated saving under such a revised definition.

But on that basis, it would only be consistent to extend the implied revised definition of “non-saving” to include the use of normally defined saving proceeds to acquire financial assets. In either case, funds that have been saved according to the normal definition of saving have been “spent”, which would allow for consistent abandonment of the idea that saving has been the source of such spending. (Indeed, MMT has occasionally referred to central bank acquisition of financial assets and associated creation of reserves as “spending.”)

The problem here is that the correct definition of saving precisely specifies the passive act of not spending on consumer goods. It does not specify how the proceeds of such saving should be used, whether to acquire real or financial assets. Saving is described in proper accounting terms as funds sourced from income by virtue of being saved from income. The eventual deployment of that source of funds is described properly as a use of funds — whether such deployment and use occurs in the form of a bank deposit, a bond, a stock, newly produced residential real estate, or newly produced plant and equipment. The deployment or use of funds is separate from the act of saving itself.

In summary, the consolidated private sector account obscures, not only the view of saving as it materializes within a given accounting period in bifurcated fashion across household and corporate sectors separately, but also the view of total private sector saving as it is projected fully into the household balance sheet, when captured as a cumulative measure over a sequence of such accounting periods. As a result, the consolidated private sector presentation within the sector financial balances model obscures the measurement of the core component of saving.”

(Carney resumes): “A couple more thoughts to add.

If saving does not include households purchasing financial assets of the corporate sector, why should purchasing financial assets of the public sector count as saving? The attempt to exclude “investment” from saving just makes saving vanish altogether, if taken to its logical conclusion. If all you mean by “saving” is one sector accumulating claims on a different sector, then you aren’t talking in ordinary language any more. You are probably confusing most of your listeners or readers, and if you do this over and over again, its fair to wonder if you are intentionally befuddling people.

The problem with constantly thinking in terms of a consolidated “private sector” is that it encourages you to skip over the most important economic dynamics, most of which take place within the private sector. Businesses get started, venture capitalists commit funds, people take out loans, and customers buy stuff with cash and credit — within the private sector. If all this just “nets out” to you and is therefore uninteresting, you aren’t really doing economics at all. You are doing something else that may be interesting but it isn’t economics. It isn’t a new economic perspective or heterodox economics or any kind of economics at all.”


The comment quoted by Carney is here:


Professor Randall Wray published an NEP blog post on March 4, 2012 in response to Carney:


Wray’s response included remarks specific to three different things – my comment, Carney’s remarks, and the underlying S = I + (S – I) equation. Carney had used my MMR comment as support for a similar argument he was making in an earlier post. Wray responded to Carney’s post, and separately as well to the entire S = S + (S – I) debate, which he must have observed independently of the CNBC posts.

(The original post mistook my quote as Carney’s words. That was corrected at one point. The original error has since reappeared.)

There is nothing in the quoted part of my comment that Wray questioned that is factually incorrect. The professor was apparently so agitated by the underlying S = I + (S – I) equation that he attacked regardless of the neutrality of the verbal description.

Here are the supposedly offending passages:


“The recent blogosphere kerfuffle about saving arose in part because MMT embraces the sector financial balances model (SFB), which features the consolidation of household and corporate sectors as a unified private sector. The model treats financial claims on corporations as negative financial assets for corporations, so the consolidated result is that household saving deployed in such financial assets makes a zero net contribution to private sector saving after counterparty corporate netting….”

That’s entirely accurate. The standard 3 sector MMT SFM model consolidates the private sector as one. Wray resurrects an old blog that references a description of the internal deconsolidation of the private sector. But I have acknowledged the existence of such background work (see my comments quoted earlier from the Interfluidity discussion.) The point is that the standard MMT presentation is a 3 sector model, not a 4 sector one. And private sector deconsolidation is obscured in the 3 sector model, by construction. I doubt any MMTer will claim the 3 sector model is not the standard presentation, because the empirical evidence to the contrary is overwhelming.


“The problem here is that the correct definition of saving precisely specifies the passive act of not spending on consumer goods. It does not specify how the proceeds of such saving should be used, whether to acquire real or financial assets. Saving is described in proper accounting terms as funds sourced from income by virtue of being saved from income. The eventual deployment of that source of funds is described properly as a use of funds — whether such deployment and use occurs in the form of a bank deposit, a bond, a stock, newly produced residential real estate, or newly produced plant and equipment. The deployment or use of funds is separate from the act of saving itself. In summary, the consolidated private sector account obscures, not only the view of saving as it materializes within a given accounting period in bifurcated fashion across household and corporate sectors separately, but also the view of total private sector saving as it is projected fully into the household balance sheet, when captured as a cumulative measure over a sequence of such accounting periods. As a result, the consolidated private sector presentation within the sector financial balances model obscures the measurement of the core component of saving.”

Again, this is entirely accurate and factual. The definition of saving is consistent with that used in the earlier Harless reference, and both are consistent with standard economic meaning of the term saving. And while I don’t know if MMT ever acknowledges income statement, balance sheet, and flow of funds as necessary constituents of fully coherent accounting, as indeed they are, the rest of what I wrote is perfectly accurate as well.

Wray quotes from MMP “blog # 4”:

“Of course, much of the debt issued within a sector will be held by others in the same sector. For example, if we look at the finances of the private domestic sector we will find that most business debt is held by domestic firms and households.”

That’s interesting, but it says nothing about how MMT handles the definition of saving for the private sector as a whole.

And he quotes from his 1998 book:

“To simplify we can assume that it is the household sector which saves and the business sector that invests; the net (inside) indebtedness of the business sector is exactly offset by the net (inside) financial wealth of the household sector. It is frequently the case that the household sector wishes to save more than the business sector wishes to invest.”

That’s interesting, but again says nothing about how MMT handles the definition of saving for the private sector as a whole. It only references the intra-sector decomposition of net saving – not the full composition of saving.

Wray resurrects yet another old blog as defense, although for what purpose I’m not sure.

Then, this is telling:

“To briefly summarize, at NEP we prefer to use the Godley sectoral balance approach, where he defined private sector saving as “net accumulation of financial assets” (NAFA), using the flow of funds data. Typically economists use the GDP equals national income equation where saving is defined as a residual: the net income received but not consumed (I’ll use it below in discussing the MMR approach). In theory these would lead to approximately the same result; in practice they do not because the NIPA accounts include imputed values. Godley preferred the flow of funds data but even they had to be carefully adjusted to ensure that every spending flow is actually financed and actually “goes somewhere” (ensuring “stock-flow consistency”). That is all quite wonky. My bigger point is that we can come up with alternative definitions of saving that would include unrealized capital gains as real and financial assets appreciate in value.”

This seals the case regarding MMT’s confused interpretation of the term “saving”. I don’t know if that correctly reflects what Godley does or not, but if private sector saving is defined that way it has nothing to do with the private sector saving S that’s cast in MMT’s own 3 sector financial balances model. S cannot identically equal (S – I) unless I is identically zero, which is ridiculous.

For example, in a closed economy with a balanced budget, if the (alleged) Godley definition of private sector saving is combined with the explicit definition of private sector saving S in the 3 sector SFB equation, then saving must be identically zero, whatever the level of investment. In cannot be otherwise for the Godley and MMT SFB specifications of S to be consistent. And in the real world economy, any such consistency would force global S = 0. And this is the point of it all – such inconsistency in the use of the term “saving” is fundamentally misleading.

Remarkably, Wray just demonstrated the same conflation of saving and net saving in MMT language and logic that is at the heart of the issue in question, which I find flabbergasting in the circumstances. Furthermore, he suggests this is consistent with the NIPA definition of saving. The definition of saving (allegedly) attributed to Godley above is not the same as NIPA defined saving at all. And the issue of imputed values or capital gains has nothing to do with the primary question. That is a secondary consideration having to do with the reconciliation of stocks and flows, not the outright confusion of flow definitions. It is a trees rather than a forest issue.

If this is not the five star version of the point that Steve Waldman makes, and the point that motivated the equation S = I + (S – I), then I don’t know what is.

Then, laughably, Wray writes the following:

“Three Sectors: I + Def + NX = S”

That’s the same as:

S = I + (S – I), where:

(S – I) = (G – T) + (X – M) = Def + NX

I’ve already developed that relationship at length.

Wray has just written the S = I + (S – I) equation as it was developed, and as it has been explained in previous sections. He has defended the very equation that he spends the rest of his time ridiculing.

Preposterously, having just written out the actual equation used, Wray then says:

“Carney has adopted the amusing (nay, hysterical) and confused exposition over at MMR which rewrites the one sector model as … S = I + (S – I)”

It is not a one sector model. Wray makes the same error as Mosler. See the previous section.

Wray says:

“Here is what the MMR folks claim about it:

Hence, our focus on S=I+(S-I) with the emphasis on the idea that “the backbone of private sector equity is I, not Net Financial Assets.” The idea is not novel, but simply clarifies the understanding of the private sector component.”

He should have noticed the phrase “private sector component”. What does he think MMR means? Does he understand the meaning of the word component? Does he actually think this refers to a closed economy without a government? How could anyone pay attention to this and believe it refers to a one sector model?

I can tell you what “the MMR folks” are thinking. It’s exactly in line with the documented form of the equation. You see, the MMR guys actually examined what was said about the equation when it was developed.

In Wray’s next step, it is impossible to tell what he is trying to demonstrate, as it seems to be a confusion of made up symbols in combinations that nobody else would consider, combined with a reversion to a language of net saving that totally contradicts what he just claimed was NEP’s definition of saving. The entire muddle is actually hard to believe.

He then degenerates into silly examples that include such dexterities as adding and subtracting infinity, etc., as permutations of the equation construction. If it is an earnest attempt at “exposing” something, the whole tirade only exposes a deficit in mathematical intuition. The equation is a decomposition of S into two parts, not a construction of S from random selections as portrayed by Wray.

All in all, the professor trots out numerous personal blog references, none of which addresses the issue, and all of which do nothing to repair the contradictory language and logic that is the actual issue.

Let’s pause now for some elementary but needed repair of his logic.

Decomposition does not equate to meaninglessness. Those who dismissed the equation merely because it was an algebraic tautology might consider the following simple analogy:

The United States = California + (The United States minus California)

That reduces to:

The United States = The United States

Does that mean the first equation is meaningless?

The point of the first equation is not that it reduces to the second, but that it decomposes the second.

There may well be information that one is interested in for the United States, for California, and for the rest of the United States apart from California. That’s seems like a reasonable investigation.

Apparently, according to the professor, such decomposition becomes meaningless, just because more than one such decomposition exists.

For example:

The United States = New York + (The United States minus New York)

According to the professor’s logic, the fact that New York can be separated out for analysis as easily as California must imply that neither case of decomposition can be of interest or value to anybody.

In the case of the S = I + (S – I) equation, is net saving of the private sector. MMT regularly uses the term “saving” to refer to that element. But at the same time, private sector “saving” is unambiguously defined as S, in the context of the national income model from which the MMT 3-SFB model is directly derived. So, the MMT stylistic habit of referring to (S – I) as S, is obviously contradictory. It could only not be so if private sector S were identically equal to I, which is clearly not the case. To suggest that would deny the purpose of 3 sector decomposition in the first place. The puzzling question is why would MMT persist with this sort of conflicting language and meaning?

Decomposition is not trivial or meaningless simply because it becomes X = X on consolidation. To adopt such a view is to adopt the view that analysis as a generic endeavor is useless and worthless, because after all this is the type of thing that analysis does. And to conclude that X = Z + (X – Z) is useless because it always consolidates to X = X is mathematically naive in the extreme. The idea that sets decompose into elements is at the foundation of mathematical analysis. I’m not surprised that some people would find the tautological expression of such decomposition curious, but to extend this to Wray’s type of response must indicate a total lack of interest in analysis or some agenda otherwise.

Moreover, the idea that an equation must be useless because people have considered it before is equally absurd. Do people simply not ever conceive of California as being a separate state and at the same time part of the United States? Do they ban further thought on such a perspective? Is it unthinkable to decompose a map of the United States ever again in such a fashion? Who actually thinks this way?

I find Warren Mosler’s reaction more interesting, because he acknowledged at least the potential for explanation, as opposed to shutting down the thought process with a defensively naive X = X demonstration. The immediate purpose of S = I + (S – I), to reconsider Mr. Mosler’s question here, was to add clarity to the discussion of saving, clarity that is obscured by terminological confusion. MMT, in focusing on private sector net saving, frequently transitions directly into the language of generic saving, to describe something that is only one part of private sector saving according to the definition of S in the 3 sector model. The strange case of the professor’s response is evidence of this in itself. Through SFB, MMT’s prime focus is net financial saving. Given this focus, there is no reason for MMT to draw particular attention to the amount of saving that funds the “I” component in S = I + (S – I) as a result. This is the essence of the Waldman observation.

Somewhere along the line, I used the term “backbone of S” to describe investment I. I’ve seen that interpreted in various places as if it is meant to imply some sort of causal requirement, in the sense that “horizontal” activity is pre-eminent to the point that “vertical” activity may not even be required. Nowhere has that been suggested by me or anybody from MMR. When I coined that phrase, the first thing that came to mind was literally the physical presence and stamina of a backbone. Investment I does represent physical investment for the most part, after all. Extending that metaphor, one could easily suggest that (S – I) is something like a brain or a central nervous system. That’s not intended, but the intuition is optional for those who wish to exaggerate the strength of the Keynesian mindset, perhaps. And I wasn’t even referring to the generally accepted causality of the relationship between I and S in a single sector economy – i.e. the Harless type causality. The equation as specified doesn’t apply to a single sector economy. Then, it really would be trivial, because (S – I) = 0 in a single sector economy. Conversely, if one recognizes and understands that the equation applies specifically to a 3 sector economy, as was documented from the start, then one should begin to suspect that there may be a non-trivial aspect to the meaning it is intended to depict. And that meaning is simply the exposition available from a particular decomposition of S, as the sum of private sector net saving and the amount of saving required to fund I.

Wray says:

“But MMRers ignore that as they jump to the conclusion that it is perfectly fine if the business sector’s deficit exceeds the household sector’s surplus—so the private sector taken as a whole is running a deficit.”

I’ll let others from MMR speak for themselves, but this claim is ridiculous and false. And again it frames the issue in the context of (S – I), without reference to I.

More false allegations are summarized by Cullen Roche here:


In my own writing on this subject, I’ve actually not referred to causal or normative relationships that arise directly from this equation. The only thing I’ve done is refer to what common sense would observe as a typical stock relationship between outstanding investment and cumulative saving, as in the United States for example. My comments elsewhere include summary data from the US Fed Flow of Funds reports regarding such quantitative relationships. In particular, the “backbone I” description refers to the typical quantitative mix of I and (S – I) in stock or flow terms; i.e. (S – I) is dominated by I in trend terms. Volatility around trend exists over time, obviously.

The Harless article “Investment makes saving possible” explores causality between those two actions in a 1 sector context. That is a compatible but separate topic from the 3-sector specific equation itself. I’ve made no comment that draws any systematic inference from the equation S = I + (S – I) to some consequence for desirable private sector financial behavior, consolidated or deconsolidated. This is mostly a straw man with respect to the reasoning behind the equation.

Wray makes other comments regarding John Carney and the MMR founders. I’m not going to respond to those.

The specific connection between MMT exposition and S = I + (S – I) extends to more general fare.  Other observers such as Marc Lavoie and Brett Fiebiger have drawn a wider net in capturing the peculiar nature of MMT’s use of language in general. The use of “saving” is only one part of it.

Next up from Wray:

“OK, but JKH is also called “brilliant” so I’m not sure if that is meant as a compliment or a diss.”

This one deserves brevity of response:

It is churlish and ignorant.

Others have observed an MMT gene that could be related:

Professor Marc Lavoie:

“While criticisms and counter-criticisms are healthy in a scientific setting, neo-Chartalists occasionally seem to over-react to criticisms, blasting away even people that are essentially on their side.”


There should be in principle a similar interest shared by MMR and MMT in advancing the broader public comprehension about modern monetary operations. Cullen Roche has been focused on the importance of such clear exposition, as far as MMR is concerned. Lavoie covers a larger frame of reference insofar as MMT is concerned, that of a confused language and logic in the description of modern monetary operations.

Brett Fiebiger directs a more technical laser at MMT in his paper:


There is much to be said later about the analysis contained in each of the Lavoie and Fiebiger papers. But in a broad sense, both papers are quite similar in that they seriously question the clarity of MMT exposition on monetary operations. Both papers allude to potentially misleading presentations in this regard, to which MMT retorts that this couldn’t be the case, that there is nothing to fix. And yet both papers cite a common complaint regarding accessibility in terms of clarity of the material. Importantly the point is made in both cases specifically and painfully with respect to student comprehension. This has to do with public purpose. Perhaps it isn’t quite sufficient for MMT to respond to those authors by saying “You don’t understand what we’re saying”.

It is ironic that a model – the MMT 3-SFB model – positioned as the centrepiece of a heterodox approach, feeds off input variables from conventional national income accounting, when those generally accepted definitions of measures of saving are then converted to different ones of MMT preference. That language conversion and the confusion it creates are at the heart of the issue here, at this specific level, and more generally. The question is why, not only because the communication dysfunction it creates, but because of the strangeness of whatever motivation might drive such inconsistency. The responses cited here show inconsistency even while attempting to defend against the perception of inconsistency. How they can be active users of the 3 sector SFB model and still confuse the language of S for the language of (S –I) is extremely puzzling. It can’t be in the public interest to be the provocateur of such language confusion. I concur with John Carney when he says “This is not Economics”.

MMT does not have a monopoly on the understanding of monetary operations. It carves out its place under the pretence that it does. It does so at its own risk. The MMT map is not the territory. There are alternatives. Criticisms of a similar nature are too widespread and coincidental not to have a reliable grain of truth. There are too many comparable instances of blogging and academic objection, including those noted here. Surely it would be in the public purpose for MMT to discipline itself away from such a noted presentation style.



  1. Oh. My. This could potentially explain JKH’s absence over the past several weeks.

    • Cullen Roche says:

      Yes. I’d say tour de force is a vast understatement. He should be applauded for his efforts here. It’s quite an accomplishment and a thorough explanation of the thinking behind our (his) position. Bravo JKH.

      • Yes, I very much appreciate the effort put forth here, independent of where I stand, which I frankly don’t know yet. Once I find the month’s worth of time to digest this :) , hopefully I will better understand where all parties are coming from.

  2. Deus.Ex.Machina says:

    JKH is in the house and he’s loaded for bear. MMTers take cover.

  3. This. Nice work JKH.

    I think the big problem here is that MMT views (S-I) as a proper representation of the private sector when it doesn’t really tell us much at all. In fact, from their understanding you’re inclined to believe if I>S then the private sector is experiencing a “net loss” as some of them have said. That’s crazy of course since I>S often represents a booming economy.

    The idea behind this is to give the reader the impression that the private sector is in dire straits if the government sector does not always spend. This is obviously wrong as I’ve previously mentioned since household net worth increased 120% between 1997 and 2008 when the private balance was negative in 38 of 42 quarters.

    The focus should be to acknowledge that government is a facilitator of growth as the MMRists show and not the driver of economic growth. It’s nice to see the language starting to get cleaned up, but don’t expect MMT to change it’s political (read, government based) approach to macroeconomics.

  4. JKH,


  5. Wray on Godley:

    “To briefly summarize, at NEP we prefer to use the Godley sectoral balance approach, where he defined private sector saving as “net accumulation of financial assets” (NAFA), using the flow of funds data.”

    This is flat out wrong and poor scholarship by Wray.

    Godley worked his whole life seeking accuracy with loving detail. Not only would he write “Net Saving”, the next phrase in brackets would be “Saving net of Investment”.

    It is highly embarrassing for me to have overlooked this even while I was reading things with a lens! But it also shows how easy it is for MMT to get away by making inaccurate statements.

    Wray says:

    “And if Carney and MMR were correct about the three balances, then that means Wynne Godley has to be wrong.”

    Wynne Godley is right.

    MMR is right.

    MMT is wrong here as Wray proves.

    Great JKH you were careful in using the word “alleged” when commenting on Wray’s quote.

    • Good stuff, Ramanan.

      BTW, my 2012 Lavoie/Godley is on order.

      • Oh that’s nice!

        Do let me know your thoughts. First impression and then maybe after a year or so. Initially when I came across it, I really liked it but was unsure as to why so much pains have been taken in many places. Gradually the book grew on me and I slowly started to appreciate the inner workings of the minds of two great people – actually a third because Godley wrote one before in 1983 with another partner. Marvellous stuff even though the authors admit there are omissions – but that’s just their modesty.

  6. JKH,

    Good point on monopoly and pretense. Outside banking and central banking even among the nonMMT community, Alfred Eichner who was a PKEist but died early wrote on this in 1984 – about the federal reserve’s operating procedures.

  7. JKH,

    Thanks for writing this. Much of the previous debate and the resulting monster thread passed over my head. I couldn’t get what you were trying to say. In bringing it all together in one place you have done everyone a great service.

    I am an advocate of MMT but don’t comment often as I don’t want any mistakes in understanding I make to be construed as representing the MMT viewpoint. I didn’t like Wray’s response to the Carney post and said so at the time, and this post has shifted my thinking slightly.

    Just so I can get this clear in my head, could someone confirm whether the following is true or not?

    When someone like Wray says something like “the private sector cannot save without a government deficit”, to be accurate he should say “the private sector cannot save (net of investment) without a goverment deficit”. For completeness he could say that “without a government deficit, either the household sector or the business sector could be saving (net of investment), but not both at the same time. Is this about right? I’m ignoring the external sector here.

    It’s been suggested that MMTers don’t make the distinction for nefarious reasons, but my guess would be that they just feel they have adequately made the distinction previously, so don’t need to be explicit every time they talk about sectoral balances. I think they are probably wrong if that is indeed their judgement, so I would like to see one of the academic MMTers respond to this post. As someone who has followed MMT for a while, when I read at MMTer write that the private sector can’t save without a government deficit, I know they mean saving net of investment, but I agree that not being explicit about that creates confusion. If this discussion leads to more clarity in the use of terminology, that will be a great thing.

    • For a given conversation (and better yet for a given subject, in general) it helps if people can agree to a coherent set of definitions. Here, that includes definitions for investment, saving, net saving, assumed number of sectors, etc. etc.

      I think the evidence of dysfunction is pretty substantial in this regard. The various section 2 links, the various commenters at the section 3 Mosler discussions, and even the Wray response itself are the exhibits, as I described them.

      So reinterpret and rephrase as one may, it persists.

      For a short cut tour, just go to the first link for Ramanan at the end of section 2. Bill Mitchell actually denies any distinction between net saving and saving. It’s astonishing. Then read the Steve Waldman quotes. Then read about the more general level of the problem via Lavoie.

      There’s no magic bullet. The accounting and the explanation of it is pretty simple when you get down to it. But people should be able to explain it in terms of straight forward accounting, not an (MMT) “magic bullet” of fuzzy consolidation.

      Fact, assuming a coherent set of definitions:

      The private sector cannot “net save” unless everybody else taken together dissaves.

      That, plus some coin of the currency issuer, will get you a cup of coffee.

      (But your interpretation of “people like Wray” looks OK. I’d be careful about assuming away the external sector though. It’s convenient, but convenience gets habitual. I think that’s what happens when mainstream economists start discussing S = I. But real world sector composition is an essential part of the discussion.)

      • Q1 again here 18 March 2012:


      • I wish Wray and Mitchell would respond more positively to Ramanan’s questions on their use of the word ‘saving’. ISTM they need only respond with one of the following:

        “when we use the term saving in the context of SFB, we mean saving net of investment. We are happy with using the term saving in this way and will continue to do so because…”


        “when we use the term saving in the context of SFB, we mean saving net of investment. On reflection it seems like many people find this confusing so we will stop using the term saving in this way in future.”

        • Ben,

          Yes they can possibly say that but when one uses

          S – I = …

          one commits oneself to using saving as per national accountants. Hence the other definition of saving is not possible – i.e., usage two definitions makes the whole thing incoherent.

          • “commit”

            very appropriate word

            The critical commitment in principle should flow from the fact that the 3-SFB model derives directly from national income accounting that relies on the same sector composition.

            • I agree, commit is spot on. Once you use those terms, you can’t just go and change the meaning.

              I am a bit surprised Scott doesn’t show up here instead of Mike Normans place. It would be useful for everyone.

  8. Joseph Laliberté says:

    Nice to see you back. I have been thinking some more about your piece published on CNBC. Below is a friendly criticism (my guess is that you may be already aware of this point).

    Corporate savings (corporate S) has often been described by yourself and other MMR commentators as addition to retained earnings (and therefore as being reflected in the book value of the corporate sector). The problem with conflating Corporate S with corporate earnings is that Corporate S is earnings before depreciation. In a sense, S is more a measure of cash flow than earnings (although it would also be wrong to assimilate S with operating cash flow as build up of inventories is a negative event for operating cash flow, but not for Corporate S). I would contend that this distinction is quite important. For a business that is highly capital intensive, the output price necessary to produce a positive S is extremely low. In a nutshell, these businesses could produce positive S almost no matter how bad the market for their product is.

    Therefore, I would say that any mention of corporate S being tantamount to “addition to retained earnings” is somewhat misleading. It is at least theoretically possible that the corporate sector has a positive level of S while depleting its equity. So I would disagree with the notion that corporate S could be consolidated -via holding of stock- on the balance sheet of the household sector. Corporate S does not mean an increase in retained earnings, and therefore, does not mean an increase in the book value of the corporate sector.

    • You’re right. It’s a detail, but an important one.

      I did reconcile that somewhere else in comments. It might have been with you?

      S corresponds to the I of GDP. Both are gross measures.

      When you net depreciation out from the outstanding stock of I, and also as an expense on the income statement, you get to the level where cumulative I is the correct measure of the outstanding stock of I (plus market value adjustments etc.), and cumulative retained earnings is the corporate book value contribution to total savings or net worth or equity. That all gets translated subsequently through market valuation as that value is transmitted to household balance sheets.

      So you’re right in the sense that you need to add back depreciation to retained corporate profit in order to get the flow measure of saving that corresponds to the I in GDP. But when you accumulate to stock measures, it gets taken out again. And given that retained earnings is a stock measure, it sort of works out that way.

      Suppose you calculate GDP for a time period epsilon, and depreciation is the only thing that “happens” (OK, there are interest accruals and other continuous stuff, but just making an other things equal point here).

      Then gross investment is zero, GDP is zero, gross saving is zero, and retained profit is a loss equal to depreciation.

      Add back depreciation to profit/loss, and you get gross saving of zero which is equal to gross investment of zero.

      But net investment and net saving (both net of depreciation) are both equal to the negative charge for depreciation in the period.

      At the end of the period, the stock of outstanding investment is diminished by depreciation and the stock of cumulative saving (savings or net worth or equity) is diminished by the same amount, all adjusted for market values etc.

      So it is the retained earnings portion that’s accumulating as the contribution to total savings, net worth, equity piece.

      That’s fairly quick. It might not be precisely correct, but I think its close. Does it sound OK? Hope I didn’t miss something. I think we’re agreeing overall; but it’s a very good point in the detail of flow/stock reconciliation.

      P.S. on something related, the Harless piece has a very insightful take on how household saving in particular occurs at the instant of income generation when you’re looking at it from a continuous time perspective.

      • Yep,

        Even with net saving defined as saving net of consumption of fixed capital (depreciation), it is possible for net saving to be positive with the budget in surplus and the current account of balance of payments in deficit.

        So the whole discussion is independent of depreciation.

        (this net saving is different from saving net of investment!).

        In fact one can say Gross Saving defined as per FoF adds to while depreciation subtracts. i.e., if one wants to include one can still use the same language of adding to etc.

        • Joseph Laliberté says:

          Agree with your comment. My comment was linked to the following extract from JKH’s CNBC piece:
          “All private sector saving can be condensed, in effect, to a measure of household saving alone — by projecting the cumulative value of corporate saving onto the household balance sheet.”

          My points were twofold.
          First, there might be a terminology issue here. The above statement is correct to the extent that “Savings” is defined as “net savings”. Perhaps it should have been specified (understood that JKH may not have wanted to add more fog to the savings debate by specifying that he is really talking about “net savings”).

          Second, there is the issue of stock/flow reconciliation. When talking about flow (i.e. GDP), we usually look at gross investments. When we look at savings and investments from a stock standpoint (balance sheet), we usually look at it from a net savings and net investments standpoint. Therefore, we need to bear this in mind when doing stock-flow reconciliation.

          • Joseph,

            I think JKH did mention that “Saving in the case of corporations is net undistributed profit, which follows after consumption in the form of capital depreciation cost.)”

            And also he talked only of undistributed profits not profits itself.

            The usage of the terminology net saving would have made the original post rather difficult to read because there the net saving is saving net of investment.

            • Joseph Laliberté says:

              A bit tricky here, I think JKH has talked at lenght how savings are deployed. Gross savings are “deployed”, not net savings. Households/corporations decide what to do with their gross savings, not with their net savings. Gross savings may well be fully invested in machinery. Only once you consolidate this flow to a balance sheet will depreciation be netted out, and the investment stock / equity adjusted accordingly.

      • Joseph Laliberté says:

        Browsing through the previous thread, you did address the issue of depreciation with Steve Roth.

        To be franked, I am not totally sure at this point how important this detail is. We both agree that in a closed economy, with the government budget in balance, S and I (in flow form) will not generate an equivalent amount of investment in stock form because of depreciation. The correct measure of the impact on the stock of investment is actually the flow of net savings (i.e. which means gross savings minus consumption of fixed capital, or to simplify, gross saving minus depreciation). Said another way, GDP calculation (a flow) is estimated based on gross investment. GDP measurement is not affected by depreciation of fixed capital, but the productive capacity of the economy is (the productive capacity of the economy is a stock). As MMR appears to decidedly emphasize the productive capacity of the economy as “the variable to optimize”, I would offer that this distinction between gross savings and net savings has important macro ramifications.

        Please note that Statistics Canada use “savings” as a synonym for “net savings” (http://www.statcan.gc.ca/nea-cen/gloss/iea-crd-fra.htm#epnt )
        Net saving: The current income of a sector, less its current expenditure. Includes current transfers but excludes capital consumption allowances and capital transfers. Synonym of saving.

  9. Dunce Cap Aficionado says:

    “There are 1, 2, 3, and 4 sector models. But in theoretical concept, one could have a 7
    billion sector model, with a balance sheet corresponding to each individual”

    I hope Google finds a way to put together a 7 billion sector model someday. Imagine if we could study flow of funds that way…..

  10. JKH,

    That was a particularly clear-sighted and level-headed appraisal of the situation. Thanks for that.

  11. This is epic. I’m afraid to start reading it.

    • Yes. Puts some of Bill Mitchells verbosities to shame.

      • Just when you’re worth of sitting in the same figurative room with these folks and contributing to the conversation (instead of frustratingly pontificating about what “saving” and “savings” actually mean)… then they have to go and toss this in front of us.

        Damn you, JKH, for making me feel dumb again ;).

        • JKH,

          To clarify, I am loving this, and I’m smarter for reading it… I didn’t mean for this to sound like your post was incredibly too wonkish or confusing.

          This is incredibly helpful. Thank you for taking what must have been a tremendous amount of time to lay this out.

        • Pierce Inverarity says:


          Who is the JKH? Where can I take classes from him?

  12. JKH,

    Btw minor point. IMO, savingS is used as a plural generally. Outside MMT and in literature on national accounting (BEA authors etc) use is like savings of American households even though they are referring to the flow. In your readings (outside MMT), have you come across places where savings is used as a stock?

    But for the case at hand, savings as a stock looks okay. Certainly MMT – when not mixing up – uses savings as a stock so it’s okay to keep it as a stock.

    • certainly from MMT (Scott mostly, I think)

      I had thought somewhere else before that, but maybe not

      I tend to use net worth/equity more often; I actually prefer either one to savingS

      I like to use equity to get some fluidity of the concept as it should apply across sectors

      It always starts as a book value measure, and progresses to market value through valuation of associated financial claims (although there are none in the case of the household sector)

      (I even like to use negative equity for the government sector (treasury at the institutional level), but that’s pretty radical. Has to be emphasized that’s entirely a book value concept in that particular use. I like it for its congruence with a view to other sectors. And the “currency issuer” gets to run negative equity as a by product of its liquidity manufacturing capability, etc. There are other interpretations of government debt as equity, that I don’t particularly like, and at times I’ve discussed those with Steve Waldman.)

      thanks for pointing that out – I’ll be more circumspect about using it now

      • That’s a good way to think about it, negative equity. It does then mirror what we see in the private sector balance sheets>

        “And the “currency issuer” gets to run negative equity as a by product of its liquidity manufacturing capability, etc.”

        But you raise a good point, it does then seem to slide into the fiscal theory of the price level writings. I do find those useful, but the FToTPL use of equity seems such a convoluted way to get to the point of “the government can’t go broke but can debase” and then it really breaks down in the valuation of the equity stage. And then Buitler and Kocherlakota just go off the deep end with their analysis of it.

    • Ramanan,

      I would like to defend the distinction JKH draws between saving and savings. It is one that I tend to make myself.

      Saving” is a verb, a doing word, a thing which is done. “To save” is therefore a flow.

      Savings” is a noun, and not a verb. Savings are things which have been saved, a stock.

      To use saving to refer to the flow and savings to refer to the stock is therefore the correct way to do things from the point of view of the English language, as far as I can see.

      Admittedly, what’s correct from the point of view of the English language is probably not the first thing on the minds of economists or even national accountants (although a quick and random scan of nearby textbooks does not suggest disagreement with this usage), but I think that following standard rules of grammar and syntax is generally the way to go in these things, and that consistent use of these terms in this fashion is generally preferable to the alternatives, even if it is not completely standard.

      • Vimothy good points. Nothing inherently right or wrong about SavingS. Since MMTers tend to use SavingS as a stock, it’s fine here. Also in everyday terminology, we (including I) tend to use “you need to have higher savings” etc. My point was just about standard terminologies as per national accountants who sometimes use savings as a simple plural of saving.

        Also banks use “savings accounts”.

        JKH can use it either way because he is JKH!

  13. I’d like to offer my own reflections on the general phenomenon of The Great Savings Throwdown. Readers should probably bear in mind, and discount accordingly, that I’ve no particular affiliation but am merely an interested outsider to both MMT and MMR, and heterodox schools more widely.

    One of the things that I like and admire about MMT and MMR is the general principle of trying to make very important economic facts, theories, events, processes and so on, to the general public and the enthusiasm this generates amongst advocates.

    To me, the goal of increasing the degree of economic illumination and accountability is what is important and not the particular vehicle that might further that goal in any given instance. If I’m critical of any particular vehicle it’s not meant to be a slight to those affiliated with that vehicle, but rather an attempt to further the ultimate goals which I assume (and strongly suspect) that we all share.

    If proponents of mainstream theory, for example, are wrong about some specific thing, then it is in their—and everyone else’s—benefit to have that thing pointed out to them, and so to advance the terms of the debate and bring it more in line with the truth of the matter at hand.

    Where I’m coming from in this particular instance is a (mediocre, I can assure you) student of economics who has felt for some time that MMT often confuses or even outright abuses the proper economic understanding of “saving”, and not merely the proper accounting definition, although with hindsight it seems clear that it does that as well, at times.

    For that reason, when SRW and JKH made their arguments in the thread at Steve Roth’s excellent blog Asymptosis, I was keen to have my own two-penneth worth, in my own relatively much more inept and incoherent manner.

    Initially, I thought that JKH’s saving flow identity was a typo, because I read the repetition of terms as analytically redundant, and was slow to connect it to the wider points made on the thread by other commenters. JKH was kind enough to explain it all to me at length, and doubtless I would have remained ignorant otherwise. So I can sympathise with MMTers for whom the meaning is not obvious (but not those for whom politeness falls out of ideological affiliation rather than simple human decency—but I digress).

    Although the identity is symptomatic rather than causal, and dwelling further on it risks obscuring the more serious problems at work here, in the debates that followed I found myself coming back to it often as a useful heuristic tool for the purposes of exposition.

    I would characterise the events that followed that initial conversation at Asymptosis in the following manner: Perhaps because of concomitant MMR controversy in the MMT blogosphere, the criticism of MMT presentation of saving as a macro concept was picked up by many MMT followers and enthusiasts, and generated a lot of debate. Rather than address the criticism outright, most commenters instead sought to argue that this presentation of saving was not in fact misleading, because it actually represented the correct understanding of saving. In other words, the conflation of saving and net saving was a feature and not a bug.

    Since in my view MMT really does present saving in a confusing and confused fashion, this was not surprising, and certainly not to the discredit of those commenters who are after all only trying to make sense of the same crazy mixed-up world as the rest of us.

    More surprising was the confusion emanating from the professional branch of MMT. It seemed that professional economists were not merely guilty of a confusing presentation of the objective facts of the matter, which is understandable and excusable, given that we are all fallible and human, but rather that they were somehow unfamiliar with those facts—facts which are basic and fundamental to a coherent picture of the macroeconomy, let alone to a theory that might explain the macroeconomy.

    It should be obvious from a moment’s reflection on the economic significance of saving as a concept what the truth of the matter is, and so it is hard to understand how there can be dispute over the facts themselves rather than impressions of presentation of those facts. And that’s without even reflecting on the implications for the values of the relevant flows given the usual (and oft used by MMT) identities.

    Where that leaves us, it seems to me, is that it is not really arguable that MMT does not present such things in a confusing manner. The whole controversy is testament to that. And it also seems that there is confusion and disagreement amongst the top tier thinkers in the school what saving is and what it means. Since saving is so fundamental a concept in generl and to MMT’s analysis of the economy in particular, in my view (discounted per my first para) this suggests the need for serious consideration from the MMT community.

    • Thanks for repeatedly pointing out confusions such as the government cannot save and all that. Which is wrong (them, not you) of course because not only can the government run a surplus, it has expenditure on investment and can have positive saving even with a deficit.

      On top of it MMTers tend to define the budget deficit as the “national saving” a completely unnecessary step. It’s wrong because the budget can be in deficit and the current account in a bigger deficit which means the private sector is in deficit! An internal contradiction and self-inconsistency. Because with the domestic private sector in deficit in the scenario, the budget deficit in the scenario can hardly be called “national saving” – even if “saving” was restricted to accumulation of financial assets.

      National accountants simply add the saving of all resident sectors including the government to get the national saving – which is self-consistent.

  14. Thanks for pulling all this together, JKH. Delighted with all the refs to Harless’s post; I found I incredibly useful in understanding, and hadn’t gotten around to posting on it.

    Some rambling thoughts:

    Saving is an income statement “event” in the sense that saving must be a subset of income. (Harless has the same view of it.) The accumulation of saving is recorded as a balance sheet item usually called net worth or equity (and classified as savings).

    I have had trouble with this because saving seems to be a *non*-event (compared to income or spending, for instance). But I realize it’s another instance of consolidation — of income and expenditure — that both reveals and obscures aspects of the underlying flows. In that consolidated view, it’s an “event.”

    Saving or savings is not a placement of funds “from itself” into other stock categories, a logical error made by many in an attempt to describe the essence of saving.

    This is great. Saving is about sources of funds (ultimately, de-consolidated, from Income). It says nothing about uses.

    Private sector saving can be decomposed into the amount of saving required to fund investment I and the amount deployed in net financial assets (S – I).

    I find a whiff of ex-post/ex-ante confusion here: “saving *required*” and “amount *deployed.*” But that may just be my confusion.

    I’m confident that I’ve specified this precise meaning of S on every occasion where I’ve engaged in extended discussion about it.

    Right. I think the ultimate takeaway here is that we all need to be equally rigorous and clear. That’s all.

    The distinction between the meaning of S and the meaning of (S – I) in a multi sector model becomes desperately conflated with the S in a single sector model.

    Right some more. I have no doubt that the leading practitioners understand this distinction; it’s not really that complicated. But many like me have been quite desperately confused by it. That suggests not so much a failure of understanding by those leading practitioners, but a failure of explanation. I will leave motive imputation to others.

    I think is that what MMT characterizes as a demand for net financial assets isn’t really that at all. It’s a demand for saving that at the margin can be satisfied at times by the government provision of net financial assets. The margin isn’t the total, and the fact that there appears to be a demand met by net saving ex post shouldn’t obscure the fact that there is an underlying demand for saving more generally, including the type of saving that matches to underlying investment in equal amount.

    Yes: I understand this as private savings being “stored” in both fixed assets (residences, and firms’ fixed assets) and to a lesser degree net financial assets. Lately I’ve been wondering if those additional (net) financial assets constitute imminent, potential claims on soon-to-be-monetized real assets, living at the monetization frontier. Still pondering how non-fixed, uncounted real assets (probably a far larger quantity than the counted fixed assets) figure into this picture.


    • Steve,

      Yes, the “event” thing about saving is admittedly abstract.

      My rough take:

      Suppose in a single sector model that S = Y – C.

      Y is an event, in the sense that it’s “key strokes” to your bank account.

      (The government isn’t the only one that can do that.)

      It’s a monetary event.

      Monetary events are essentially calculations.

      C is an event in the sense that it requires your action to spend, but it also involves monetary key strokes.

      Of course, Y also required the action of your employer, etc.

      So in that sense, key strokes are representative of actions and events.

      The question is whether S is an event.

      Well, S also gets represented as the result of key strokes to your bank account.

      It’s a calculation.

      Calculations are events, in the abstract.

      So, measuring S as a flow, it proceeds as a chained key stroke event of sorts, and one that coordinates with a parallel chained key stroke (or currency exchange) event of spending, within a given time period.

      (Harless essentially argues that, to the degree that Y is a prerequisite for C, S = Y at the moment you receive Y. I agree with that. S then gets whittled down over the course of the accounting period by C events within the same period.)

      Sadly, I have the feeling this will not do the trick for you. But it is roughly how I think about it in the abstract, by not limiting the idea of an event to a particular physical threshold according to definition.

  15. Dunce Cap Aficionado says:

    Speaking as a self-recognized lay person who has just finished my first read of this epic piece, I can say that I need to read it many more times to truly absorb it.

    But in my first read even my lay mind was brought up a few notches both in general and by several specific points.

    Bravo JKH.

  16. Also thanks to all for your comments so far – Cullen, wh10, Deus.Ex.M, FDO15, Michael, DCA, Vimothy, Dan, Greg.

  17. Hi JKH. Thanks for this piece of work. It’s monumental.

    I have been a long-time follower of MMT and an admirer of their work. In general I think it’s very good. But I too have noticed this error and it’s no minor one. It first hit me in this story by Mosler where he said the USA was losing output because of pension funds. I think his book discusses this as one of the 7 frauds as well. He says:


    “dollars in pension funds ‘come from’ unspent income.”

    This is very wrong. Dollars don’t go “in pension funds”. They go through pension funds. When you invest money in a secondary market you are buying a claim on that money at a future date (the value of which could be higher or lower). But you exchange that money in return for securities. This is all pension funds do. They collect money from retirees and buy claims on future money. But when they buy the securities from someone that person is selling the security in exchange for cash. This is why they call stock “exchanges” exchanges. You are exchanging cash for securities.

    It’s a slightly different point than the one you’re making here, but the same general error. And a big one in my opinion since it’s one of Mosler’s 7 frauds.

    • thanks, Lars

    • Lars,

      In a sense he is right because it so happens that a lot of consumers tend to spend everything they get “in hand” and pension contributions acts to increase the propensity to save which via Keynesian principles acts to lower demand and output. Probably he doesn’t explain it rightly.

      However it is a bad strategy to ask consumers to spend more because they’ll have less when they retire. Surely the foundations of American growth needs to be changed.

  18. MarkoS,

    Someone is sure to do a better job than I, but I’ll offer my 2 cents. When investment occurs from a bank loan, both savings AND investment has occured. Net saving, however, has not. That’s where the foreign sector, or, to keep it simple, gov’t NFA’s come in.

    Regarding 2008, I think it’s probably worth pointing out that while both MMT and MMR’ers seem to agree that the lack of NFA’s was troublesome in 2008, MMR’ers would probably assert that NFA’s wouldn’t have prevented the housing bubble, nor would the have prevented the popping of it, but simply made the adjustment easier after the fact, without massive unemployment and systemic economic risk.

    MMT seems to try to assert that the lack of NFA’s was a huge contributor to the housing bubble and/or the subsequent crash in housing. This view is incredibly flawed, IMO, for reasons we’ve discussed but I won’t get into here.

    The difference in thinking highlights the difference in perceived-influence of NFA’s on investment and inflation over long swaths of time, and therefore create a fundamentally different picture between MMR and MMT as to what truly drives the economy.

    All that said, when the demand for (S-I) went up considerably in 2008, we should have obliged faster than we did… in fact we probably should have been deficit spending quite a bit higher along the way to 2008… though I’d hardly say that would’ve prevented the housing bubble any more than more life boats would have prevented the titanic from hitting the ice berg.

    • It’s not just semantics.

      It increases gross financial assets.

      Banking system financial assets increase (loan) and non-bank financial assets increase (deposit).

      Any measure of net financial assets must specify the sector that is the focus of netting.

      And there are numerous ways to define sectors. I mentioned a 7 billion sector concept in the post.

      In this case, there is no change in the NFA position of the banking sector, because the deposit offsets the loan.

      And there is no change in the NFA position of the private sector, because the banking position is a wash and because the position of the depositor offsets the position of the borrower.

      But there is a change in the NFA positions of each of the buyer and the seller.

      The buyer’s NFA position drops by 250K and the seller’s NFA position increases by 250K.

      • Both Gross and Net matter… but the nature of the two, and how they interact with an economy in different stages is, to me, one of the big discussion points of this blog.

        NFA’s matter, but it can take a long time for them to matter A LOT, especially if banks are properly allocating resources and we don’t have an asset-price shock like 2006-2008.

      • Sort of – remember that the primary purpose of all of this is to try and sort out a framework of understanding for what’s going on in the economy and the financial system. So a fair bit of that is the accounting for how all the pieces fit together that we’re trying to understand.

        Yes, the gross financial asset expansion of the banking system is important over time – it expands credit and money supply – which is helpful for real economic expansion. And yes, the overall NFA effect in your example is nil and not important to the result.

        In your house example, let’s assume it’s an existing house (not a newly constructed one). Then the activity you describe is just a rearrangement of assets and liabilities. The seller now has money and the buyer now owes money. But there’s no change in net wealth involved and importantly this transaction is happening independently of anything else that’s going on involving income, saving, government deficits, whatever. What’s really happening is that the buyer and seller are using the banking system to make the seller “more liquid” (money instead of a house) and the buyer “less liquid” (a house with a debt). The banking system is providing the liquidity services for that transaction. You’re right; the NFA stuff isn’t a big feature of this, but you can use the NFA concept if you like to describe changes in each of the buyer’s and seller’s financial positions, as I did above.

        • Cullen Roche says:

          I’d disagree with you slightly here. I think the last few years have proven that you have to understand both the horizontal and the vertical extremely well (which MMTers generally do though they probably lean a little heavy on vertical, but whatever). In a normal economic environment the horizontal does the heavy lifting. But in a balance sheet recession like the one we’ve just been through the vertical can come in and pick up the slack when the horizontal freezes. That’s a powerful tool to have and a very powerful understanding. Understanding the net financial assets and the way the economy works off both horizontal and vertical money creation helps one understand the system as a whole. It’s about balance. Waldman has called us “diagonlists” which is not a flattering name, but probably accurate in terms of how we’re approaching things….

        • Yes, and it seems to me breaking down the S = I + (S-I) helps to see how/when/where the horizontal/vertical money creation process might be brittle or clogged.

          • Cullen Roche says:

            A reader just sent this to me via email:

            “Private sector saving can be decomposed into the amount of saving created by investment “I” and the amount of net financial assets transferred from other sectors (S – I). That is the focus of the equation S = I + (S – I).”

      • Dan Kervick says:

        When I take out the loan the gross financial assets in the system increase.

        No I don’t think so; it’s matched by an offsetting liability. And that bank liability isn’t just a meaningless accounting figure. When you write a check on that $250,000 deposit and give it to someone, they probably deposit it in another bank. And when the payment is settled and cleared, $250,000 is drained form your bank’s reserve account at the Fed and re-injected into the other bank’s reserve account. Those reserves were an asset of your bank, so your bank just lost $250,000 in assets. The bank of the person you gave the check to has acquired $250,000 in assets. But that is in turn offset by the deposit balance they created for their customer: the person to whom you gave the check. That customer’s balance sheet is the one to which the assets have ultimately been shifted.

        So if the liability is real and is the representation of a payment soon to be made why is your bank willing to do such a thing? Hand away $250,000 in assets? Because you made a promise to give them a much larger sum in return over a period of several years.

        The liabilities recorded on commercial bank balance sheets are not just accounting fictions. They are representations of real negative value for those banks, which the banks are only willing to acquire because they are exchanged for promises of positive value. But it wouldn’t be wrong, or so I think, to regard the similar liabilities recorded on a central bank balance sheet corresponding to a bank reserve balance as an accounting fiction. A commercial bank’s deposit is a claim on a payment for which the bank is liable, and therefore a claim on the bank’s finite stock of assets. The reserve balance at the Fed is similarly a claim on a payment from the Fed. But since the Fed has no finite stock of assets that is meaningfully diminished or augmented by such payments, the situations are very dissimilar.

  19. Thanks for this. Quick summary (of sorts) will go up on NetNet soon.

  20. jhk – wow, incredible amount of time and effort on your part. There has certainly been a tremendous amount of time and energy devoted to the debate that seems to come down to misinterpretation of semantics and context in regards to decomposition of assets, liabilities and flow of funds between mulit-sector accounting. Your paper will help many understand the foundational concepts that often gets lost in the academic and sometimes twisted (obfuscated) dialog.

    I would love to see the same energy and debate around the concept of
    debt based money (existing system) and debt-free money system.
    The $1T platinum coin idea/concept is an implementation of debt-free money, but why not take it to a systemic approach and implement a “Nationalized Money, Privatize Banking” system. I have come to the conclusion that a debt-free system and tax reform/overhaul (scrap income tax, replace with consumption tax) — but tax reform is a topic for a different forum / debate.

    Joseph Huber & James Robertson at New Economics Foundation ( http://www.neweconomics.org/ ) published a white paper
    that describes the theory and operational aspects of a debt-free system
    (Restoring Seigniorag).

  21. Also realize the usage of nongovernment is a bit silly because by construction this includes governments of other countries!

    In a specific setting one can use it. Closed economy for example. Or nongovernment organizations etc. But dumping everything such as foreign central banks and governments into ONE nongovernment sector is bad terminology. :-)

  22. This has all been a great discussion and I think it’s important to highlight the key takeaway from it all. I hope a leading MMRist will correct me where I am wrong, but I think this is the key conclusion from all of this:

    MMR has depicted the economy as a balance of producers and consumers within a monetary system created by a particular society. The monetary system and the government is created to facilitate this process. I really like the use of the word “facilitate” here because it is specific in its meaning for the government’s role.

    MMT is Keynesian in the sense that they always focus on demand. If there’s an economic problem then it’s a demand problem. MMR seems to be attacking the issue from a more balanced approach by acknowledging that demand is a big part of the equation, but not the only part. In this sense, you guys aren’t really Keynesians, but some sort of hybrid balance.

    I like where it’s going.

    • Though I definitely identify more with the left-leaning economists today (I consider MMR left-leaning in todays environment because of fiscal hawkery and gold buggery from the right… though I’m sure if we had a republican incumbancy we could very well be seeing something very different).

      That said, I like MMR better than MMT, and try to think of situations where “supply-side” is truly necessary.

      Would something like a natural disaster in the gulf, a terrorist attack in a populated city, or something else that would affect productive capacity be something that would call for a decidedly more supply-side approach?

      • Dan Kervick says:

        I’m not sure I am correctly understanding the use of “supply-side” in this context. But disasters of such a kind would presumably call for significant re-purposing of existing real national assets. Presumably the public treasury would purchase a large quantity of goods and services from their current owners, and organize public enterprises and public-private partnerships to rebuild. Those activities would constitute a massive source of market demand in themselves, and would have some additional demand-side spillover effects as the payments flowing into the affected areas would provide their recipients with the financial wherewithal to transform their desires into demand for other products. But there would possibly be supply-side effects as well as the subsidized work would lower the costs of some products or generate new desirable products that would never have been produced otherwise.

    • Cullen Roche says:

      I think that’s a fair assessment. I’d only emphasize that we’re trying to highlight the “backbone” of the economy through these discussions. But that doesn’t mean there aren’t other vital parts to the machine. And I keep using a machine analogy because I think it’s wrong to apply a hierarchy of some sort to the system. That’s what bothers me so much about terms like “monopolist”. It implies a hierarchy of power where it’s not necessarily applicable. To me, the monetary system is a machine with many moving parts that all come together to create a particular result. The backbone of this machine is investment leading to production. But importantly, we must realize that the monetary system exists as a tool that helps us facilitate the optimization of our available resources. I like to say consumption and production are two sides of the same coin. The Keynes vs Hayek debate misses the fact that they’re two sides of the same coin. There is no “this one is better than that one” (though at times the pendulum can certainly swing!). But production can make the coin bigger. It’s up to us how fast we’re going to make that coin grow and we massively underutilize the strengths of our govt in achieving this.

  23. Detroit Dan says:

    This is minor stuff, in my opinion. It’s major as a supplement to MMT, but minor as a criticism of MMT, since there’s no real contradiction of MMT here.

    The elephants in the room, for me, are the monetarists, who spout nonsense which is endorsed by the MMR folks. Please tell me what sense there is in NGDP targeting, the IS-LM curve, etc…

  24. Detroit Dan says:

    Steve Waldmann even had a piece entitled The moral case for NGDP targeting. Steve is the smartest of the smart, and a very good guy. But come on!

    “The last few weeks have seen high-profile endorsements of having the Federal Reserve target a nominal GDP path. (See Paul Krugman, Brad DeLong, Jan Hatzius and colleagues at Goldman Sachs.) This is a huge victory for the “market monetarists”, a group that includes Scott Sumner, Nick Rowe, David Beckworth, Josh Hendrickson, Bill Woolsey, Marcus Nunes, Niklas Blanchard, David Glasner, Kantoos, and Lars Christensen. Sumner in particular deserves congratulations. He has been on a mission from God for several years now, and has worked tirelessly to persuade us all that central banks should target NGDP, and that they have to ability to do so even after interest rates fall to zero.” [Steve Waldmann]

    So do the MMR folks agree with Steve Randy Waldmann on this? Does this make any sense? How foolish can you get? This is complete and utter nonsense, and we are standing with it against MMT?

    I consider myself MMR more than MMT, and S = I + (S-I) is a fine addition to the MMT canon. But let’s not lose sight of the fact that MMR is standing on the shoulders of MMT and extends MMT much more that contradicts it.

    • Cullen Roche says:


      I still can’t figure out what the driver is behind NGDP targeting. Animal spirits? I don’t speak for all MMRists, but I sure as heck can’t figure out what they’re trying to do with NGDP targeting. There’s just not mechanism through which it works unless the Fed gets more involved in markets and starts buying up currency or actually buying houses or pinning rates. I mean, theoretically, the Fed could “target” a certain NGDP through massive QE at the 30 year bond at a specified rate. Would it work? Yeah, it sure as heck would. But it would work through the credit markets and we know where that gets us when the govt encourages reckless lending.

      And I like to call myself a Mosler Monetary Theorist. I will never stop reminding Scott and Warren that they are by far the most brilliant and influential economists I have confronted. Is MMT perfect? Obviously, I don’t think so, but that doesn’t mean it’s not damn good.

      Corrected – MR is not a subset of Mosler Economics. It is more appropriately referred to as a set of operational understandings with an influence from Post-Keynesian Economics. Mosler and Fullwiler are brilliant, however. :-)

      • Detroit Dan says:

        Thank you, Cullen. Plus one for MMR!

      • Hi Dan,

        I am similar to Cullen on this. I don’t know what’s driving the fetish over NGDP levels – it seems like it would doom an economy to low growth.

        I do think NGDP level targeting is superior to inflation targeting for many of the reasons Steve W uses. But that’s comparing bad policy to ok policy, not to actual good policy.

        Why is NGDP level targeting only ok?

        1. Population growth: NGDP level targets of 5% = 2% for inflation + 1.2% for population growth leaves 1.8% for real GDP per capita growth. That’s pathetic. And we don’t hit 2% all that much over the last few decades. We’re staring 1% real per capita growth in the face with NGDP targeting at 5%. At 6% NGDP level targets, we’re at 2.8% real growth if inflation is kept at 2%.

        2. Low inflation target means higher odds of liquidity trap: self evident

        3. Measurement of estimates/proxies for natural rate of interest have confidence intervals 5%+ wide: NGDP level targeting has an implied homage to the natural rate of interest buried in its heart. Unfortunately, we can’t tell this natural rate within a country mile of the actual number even years after the fact. There are many problems with the idea of a natural rate of interest.

        4. NGDP level targets get abandoned when you really need them: Check out UK NGDP level growth, and it’s 5.3% for a few decades. This gets abandoned in 2008, when it mattered the most.

        5. NGDP level targeting relies on monetary policy exclusively as structured by the MMs. Monetary Policy “sucks” as you’ve read by me before, and it chooses winners and losers just as certainly as fiscal policy, but then puts a nice layer of opaqueness to make it extra confusing to the general public.

        This should have been a post I know. Also, I like Nick and David Beckworth and they write tons of smart things, so I didn’t want to lay it all out so black and white.

        I do think NGDP level targeting is far superior to inflation targeting. It at least allows for some real world growth. If it was implemented and followed, it would force the CB to at least attempt to help make peoples lives better. Our current inflation targeting regime is just hoping it all works out for economic growth.

        • lol – I didn’t see your comment over at Mike N’s place until after I wrote this. :)

        • Dan Kervick says:

          I guess my problem with the whole NGDP level targeting agenda has little to do with whether targeting NGDP is better or worse than flexibly targeting inflation, or targeting a monetary aggregate. I just don’t think the Fed can hit these targets. Except for interest rates, the Fed has failed to hit every target they have tried to hit in the monetarist era, and then abandoned the target. Why should the NGDP level targeting be any different?

          • Thats a good point and one I don’t have in my list o”why monetary policy sucks” It’s related to the natural rate of interest problems I am outlining (yes, it’s taking me forever and JKH has done 8x as much in 1/10th the time.)

            The fed simply doesn’t hit any targets it’s supposed to hit very well. Thats more of a monetary policy doesn’t work than it’s a slam against NGDP per capita level targeting specifically. I need to update the list.

            Btw – here is a list of reasons monetary policy isn’t all that great.


            yes, yes, MP is theoretically less distortionary than fiscal policy. But there are other considerations

            And just to be clear, I’ve spent at least an order of magnitude more time on outlining problems with the the other side than I have on problems with MMT. Just go over to the Traders Crucible and take a look.


            • Mike (or anyone interested in thinking about monetary policy),

              In general, do you think that interest rates can have an impact on the economy? How / why?

              • Cullen Roche says:

                Maybe the more important question is – why should we assume that monetary policy is totally inept? Didn’t QE1 prove that wrong? Shouldn’t it hold that the Fed can substantially influence the price of credit which can influence the amount of investment and production in the economy? Is it a coincidence that inverted yield curves lead to recession 90% of the time?

                I’m no fan of monetary policy. I think it’s a weak policy tool in most situations, but aren’t some people a little harsh about its impacts and in some cases in outright denial?

                • Cullen,

                  I think that’s a wise attitude to take.

                  In the for example, in the modern way of thinking, one particularly influential channel for monetary policy is the so-called “cost channel”. If interest rates are raised, this then has implications for the cost of funding for firms in the real economy. It seems hard to imagine that this would not influence economic activity.

                  • Cullen Roche says:

                    Yes. And I think one of the powerful lessons we learn from MMT is that the Fed could really dominate the economy if they wanted to. For instance, we know that monetary policy is about price and not quantity. What if QE2 had been about setting the 30 year bond at 1%? And the Fed just sits there and gobbles up all the 30 year bonds? Mortgage rates tank, the cost of financing long-term loans becomes ultra cheap and we likely see loan demand boom and investment soar. I’m not saying that this would necessarily be a wise move because it would likely just trigger another boom cycle, but the weapons are there for the Fed should they decide to implement them in precise ways….

                    • Could long-term rates really collapse in that way? I’m trying to think through this in terms of a bank’s POV…

                      For ST loans, the bank can either buy treasuries with reserves, and get .1% for 6 months, or loan it to an individual via credit, and get a higher rate but risk losing the money… but only for 6 months… the main issue is the credit of the individual, not interest-rate risk.

                      Now looking at a 30-year loan… if the fed lowered the rate to 1% on a 30-year mortgage to get the economy booming, is the bank really going to loan to me at 30-years for 1.8%, or 2.5%, if they think the fed is just going to raise rates soon anyway? I dont’ think that’d work.

                      I think it basically gets back to the fact that long-rates are built on expectations of short rates going into the future, and non-treasury rates are built on top of treasury rates…. so basically it seems to come back to “How long is the fed going to keep short-rates low?”

                      If you start looking at the implications of the fed aggressively (more than in Operation Twist) attacking long rates, the banks would be more concerned (I’d think) with the duration of time they’d CONTINUE to keep the long rates low moreso than whether the long rates are low today… so really it seems to come back to the fed making promises about short-rates affecting banks decisions, as long-rates are built on that base.

                    • Cullen Roche says:

                      I’m just guessing, but if the Fed offered to buy an infinite amount of 30 years at 1% then the 30 year conventional mortgage rate would crater. It wouldn’t fall to 1%, but it would fall a lot from where it is now. My point is, this is a weapon the Fed has. It’s not necessarily a smart one, but it’s a weapon nonetheless.

                      The Fed could also do fiscal via various tricks. Buying FX is one way. Another is buying munis with the states acknowledging the funding source and doing the opposite of what Europe is doing and spending like crazy. Sucking the Fed teet if you will….These are all within the laws today and an aggressive Fed chief could enact these policies regularly and tell Congress to screw off. I’m not advocating any of this, but the Fed is a big powerful entity and their tools are broader than just “The FOMC announces X change in interest rates”.

                    • I meant “if the fed lowers the rate to 1% on a 30-year TREASURY BOND!!!”… not mortgage. Sorry if that was confusing.

        • Dan Kervick says:

          I am similar to Cullen on this. I don’t know what’s driving the fetish over NGDP levels – it seems like it would doom an economy to low growth.

          FWIW, I think it’s politics. I hate to call out my own tribe on this, but the fact is that Obama dropped the ball on the big fiscal initiative he needed when he had the chance and had the Congress, and then pivoted toward the stupid deficit hysterians and threw the unemployed under a bus for a year. In part as a result we have horrible unemployment.

          When it was clear that things were not going well, the mainstream Dem punditry in the economic blogosphere, the ones who get to sit in on the special conference calls and receive the email distributions with admin talking points, suddenly discovered that there were these generally right-tilting economists claiming that the Fed had a magic bullet to fix the economy and boost unemployment all by themselves. Whether it was true or not, that set up the Fed and Bernanke as the perfect scapegoats for 2012.

          The MMers are a generally very conservative bunch who hate fiscal activism, and have their own ideological agenda for promoting a monetary-only approach to macro policy. But lots of insider Dem’s were willing to glom onto this decidedly anti-progressive approach to macro because it offered great buck-passing potential for a President presiding over a lousy economy.

          In every one of his statements Bernanke signaled the need for more fiscal activism. But no luck. Republicans were now in charge of Congress, and lily-livered Dems were in no mood to take them on aggressively. Plus how would Obam explain the about face form Bowles-Simpson and Peterson Foundation deficit hysteria to revving up the fiscal engines?

          I mean, the Fed must be able to fix unemployment, right? After all they have a phony-baloney “mandate” from Congress that says they can.

  25. Excellent. Very clear and reasonable.

    Two points.

    1. When the central bank buys bonds, doesn’t this also add NFA to the private sector? Bond purchases always increase the liability side of the commercial banks’ aggregate balance sheet over the asset side, as well as adding reserves. Also, withdrawing deposits as banknotes decreases bank liabilities but not bank assets. So I get:

    (S-I) + (G-T) + (X-M) + (QE-Banknote Withdrawals) = 0

    2. Ordinary lay people define their own ‘savings’ not just according to the accounting definition of income not spent on consumption in one time period, but as the total amount of money accumulated in their deposit account over many time periods plus the cash in their wallet.

    So perhaps S needs a more precise name than ‘Savings’. I like JKH’s accounting definition of domestic private income surplus to consumption, but this seems to ask for a name like ‘Private Surplus’.

    This would be analogous to the ‘Trade Surplus’ (X-M), or the ‘Government Deficit’ (G-T).

    • Maybe ‘Open market operations (O)’ is better than QE. And Banknote Withdrawals could be symbolised by ‘W’.

      (S-I) + (G-T) + (X-M) + (O-W) = 0

    • BT, when the CB buy a bond, the bank gets reserves in exchange. It’s an asset swap. No change in assets, everything else constant (there could be a valuation effect from a change in interest rates, but let’s put that aside for simplicity). Also, a withdrawal decreases both the bank’s assets and liabilities. The deposit (liability) decreases as does bank cash (asset) used to settle the payment.

      • I see your point. But I’m treating reserves as being on the central bank balance sheet, not on the commercial bank balance sheet.

        • dbl entry accounting. Reserves are most definitely on the bank balance sheet as a bank asset and also on the Fed balance sheet as a liability.

  26. Detroit Dan says:

    The Fed controls short term, risk free interest rates. The Fed does not control inflation, nor does it control GDP.

    What do these guys that Steve Randy Waldmann thinks are so smart believe the Fed should do?

    • Cullen Roche says:

      There’s a legitimate argument to be made for the Fed credibly changing expectations. For instance, if they bought all the 30 year bonds at 1% right now they could make a home loan really attractive to a lot of people. That would boost residential investment and filter through the economy big time. I’m by no means saying that’s the right thing to do, but if the Fed wanted to be really credible they have the weapons. That’s why I don’t really agree with the MMT approach of always (or usually) shunning the Fed. Better to have and not need than to need and not have….

      But you and I both know where the real firepower is. It’s in fiscal policy. Anyone who understands the monetary system knows that….So the Fed is kind of a diversion from the politics that muddy the fiscal waters….What a great world we live in, huh?

      • Yes it is the fiscal side where the political rascals live and breath. What is needed is discipline in supporting free markets and using the fiscal stimulus when needed to invest in value creation endeavors (could be R&D, infrastructure, public/private innovation sponsorship – not to be confused with Gov’t as a venture cap e.g. Solyndra); national defense (minus the internationalism and nation building); AND a whole lot more investigative and enforcement of appropriate level of regulation.

        Target reduction and elimination of waste, fraud and abuse to focus the resources in the value add/creation investments.

        The FED and the monetary policy is a huge enabler, since the FED controls the interest rate and monetary base to meet an inflation target and constrain/loosen the banking system credit level. The FED actions are watched by everyone, sets the tone and expectations and drives the emotional/psychological behaviors due to the fact that everyone believes the FED is the master of the universe.

        and the game continues to go on and on ….

      • The Fed could also reduce CAD by setting up a dual exchange rate system (in effect, subsidizing exports). Stepping into the exchange rate market is monetary policy that acts like fiscal policy but is really foreign policy; which is why they wouldn’t do it unless Tsy, if not the White House, took the leading oar. In any event, Congress doesn’t have to vote on it.

  27. I’ve been trying to follow this over the last month or two, but I feel like i’m too behind to catch up. With that said, I really hope the broader MMR-MMT community truly breaks into the mainstream, and/or wins the Econ Nobel.

    Is the identity of JKH still unknown?

    Keep up the great work everyone!

  28. Wow! I wonder if this post will actually go down in history as a watershed moment? Perhaps i’m being a bit grandiose…

  29. I wanted to post what SRW wrote on twitter, which I think is very fair:

    “so, i find this whole thing kind of painful, bc it is yet another reason for people who should be… allies to squabble. i have a gripe w/what i consider to be occasional sloppy rhetoric that leads… to a misunderstanding that is convenient 2 the politics of some MMTers, + to my own politics, but… that is economically incorrect under conventional definitions of savings. i think it shortsighted… to leave this misunderstanding uncorrected or to try to capitalize upon it. none of the MMTers… i’ve communicate labor under this misunderstanding, though sometimes even the best slip and… make statements that might be misconstrued. I’ve made this point a number of times w/o great… controversy. The current tempest/teapot, mountain/molehill results from what I consider to be… unfortunate sociology, small fissures magnified by the fact that even the best ppl r human+have egos. ***there is no substantive issue here. there is definitional unclarity, and if any good comes of this.. episode it would be to make us all sensitive to defining and adopting terms in a standardized way… and using them consistently.***”

    • Cullen Roche says:

      We are allies. We just have different approaches to all of this and we come to different conclusions. But we’re working from the same set of understandings for the most part. It’s all a Moslerian understanding of the world for the most part…. :-)

    • The whole twitter feed is longer and gets much more specific. He seems to support the premise, and the idea of corrective action – while trivializing the fact that attention has been drawn to it. I don’t quite understand that flow.

      • It’s frustrating that so much time and energy has been spent merely trying to establish the proper definition of saving–it hasn’t even been possible to engage with MMTers over the original point of debate; namely, how that proper definition of saving is communicated.

        • An analogy I might draw is the feeling of wanting to walk out of the first 5 minutes of a lecture on integral calculus, because the professor is repeatedly identifying derivatives as integrals. You simply don’t want to be taught by a teacher with that sort of method and/or attitude. It is vexing, and for some, intolerable. And you can miss out on a good education because of that. And such a teacher can lose a lot of students – even though the teacher may be quite comfortable in his/her own style. At the same time, there are exceptions, and some teachers are clearly better than others.

          Some people may have been turned off from MMT because of this, and they may be people that MMT would be better to have on its side – or at least not turning away from it for stylistic reasons alone. It could be that the people who are getting turned off have a perfectionist streak in them, or an unusual attraction toward discipline and detail – as Tom Hickey alluded to. I guess what’s always puzzled me is that while MMT is built on a framework of accounting precision on the way in, that precision seems to get muffled on the way out. I don’t understand that. If you are a reader that is sensitive to the value of that sort of precision, and were originally attracted to MMT because of it, you are left wondering why this is the case.

          • To continue with that analogy, I think it such an approach could cause problems that are not merely trivial and and in the purely “nominal” domain, even given a desire amongst the students to transcend their teacher’s idiosyncrasies.

            Since students are young and have not yet developed a deep understanding of the structures that lay beneath the labels applied, there is the opportunity for a great deal of confusion, especially when they meet students of other teachers, or when they sit in other lecturer’s classes.

            So rather than facilitate the clear communication of ideas and the analysis of ideas communicated, this method of teaching risks becoming an impediment to learning and understanding–not at all a desirable pedagogical affect.

            (In fact, such an approach almost seems designed with equivocation in mind).

          • There’s definitely an issue here, and it’s funny how some commenters especially in Tom Hickey’s post dismissed as a matter of pedantry like “oh it’s just saving versus net saving” etc and not realizing there is an issue. Some have accepted that there is a sloppiness but I haven’t seen someone realizing this now appreciating the section on saving as a residual.

            The part of the reason for the issue is the overuse of the word currency or dollar. I read somewhere that the way to describe money is to use a passive language.

            So I see commentators using phrases such “dollar receipts” “dollar payments” “dollar saving/s” even more after this instead of simply using receipts, payments and saving! The Moslerism “dollar saving/s” is a put off because it is misleading to begin with. (no offence to the man). It is true that finally saving is recorded in US dollars for a US sector or a unit, but the phrase dollar saving somehow leads one to think of it as increase in dollar denominated financial assets. (“save in the currency of the issuer”)

            Of course as S = I + (S-I) and the explanation of it shows how the counterpart of saving is not only increases in financial assets but also nonfinancial assets.

            But then we have “real saving/s” as mentioned by Randy Wray.

            Also for a sector these financial assets or liabilities for a sector need not be even dollar denominated in a 3/4-sector model. So to avoid complications MMT uses the 2-sector model government and “non-government”!

  30. [This is a longer post that is more of an “open comment” aimed at MMTers. I’m posting it here because I’m not sure where else it should go. I considered positing it at Tom Hickey’s blog, but it’s hard to know which threads are active and where I could post it so that it would actually be seen. I’d be keen to hear how more experienced MMTers feel about my take on it.]

    It’s unfortunate that there is still so much confusion over this. I could agree in theory with commenters who have described it as mountains made out of molehills, but at times it has seemed hard to penetrate through the noise generated by the enthusiasm of the participants in what should be a simple debate about tactics to the actual signal buried somewhere beneath, and this suggests that there might be something here that is worth addressing in a more serious manner.

    I consider the enthusiasm of MMT proponents to be one of the many strengths of MMT, and it is to its credit that it can motivate such passion for a subject that, despite its obvious importance, is often dry and obscure.

    However, enthusiasm and passion are not always helpful. At times they can interfere with cool reflection, and cause what should otherwise be straightforward exchanges of ideas to become lost in a fog of confused excitement.

    Obviously I don’t speak for anyone else in this debate, and do not consider myself to be affiliated with any particular school or tendency, but my understanding is roughly thus:

    There are two issues. The first issue should really not be an issue at all, since it is a fact and not a matter of opinion. This is the technical question of how saving is defined.

    The second issue—given that the first issue isn’t really debateable— seems like the more important and substantive problem, something about which reasonable people can and will disagree, which is how that technical understanding is communicated and should be transmitted to the public.

    When this whole debate got started, people were discussing issue two and not issue one. It was felt that MMTers sometimes present saving as though it is something which it is not, as though it is something that contradicts or is in disagreement with the proper technical definition. No one, as far as I recall, suggested that the professional MMT contingent didn’t know and understand that proper technical definition.

    In the next stage in the debate, lots of non-professional MMT proponents appeared and, rather than address the issue that was being discussed (i.e., issue two, how the proper understanding of saving is presented and communicated by MMT), instead argued repeatedly and at great length that the proper understanding of saving was incorrect!

    Perhaps the idea that someone could completely and repeatedly confuse saving and net saving is hard to countenance for professional MMT economists, who have studied and researched the subject for years and are unlikely to mistake elementary matters of fact. But nevertheless, professional MMT economists are not the only people who represent MMT in the econ blogosphere or on the internet in general, and it would be wise to give consideration to this fact when formulating informal statements about saving and its function in a conceptual model of the macroeconomy.

    In any case, this generated some huge threads and intense debate—all good from the perspective of the vitality of the school in general, but less helpful from the perspective of a productive exchange of ideas about the actual issue that people were trying to discuss, viz. Issue two, how the proper understanding of saving is communicated.

    Hence, it seems to me, the need for JKH’s recent long post above, to clarify the technical issue, the thing that wasn’t even being debated in the first place, to those of us who do not properly understand it. In fact, despite the thousand comment thread at MMR and very long threads at Winterspeak and doubtless other blogs as well, Scott Fullwiler was the first MMTer that I saw who even agreed with the proper technical definition of saving.

    To me, this suggests that JKH and other commenters might well have a point—because, otherwise, why did it take thousands of comments and intense argument before an MMTer appeared who agreed with and understood the premise of the original conversation?

    It seems as though the non-professional contingent really do have the impression that MMT disagrees with the proper technical definition of saving, which is why they claimed that the proper technical definition is wrong at such length and with such vigour.

    No one, again, as far as I know, has suggested that any of this “proves MMT wrong”. I confess that I struggle to even conceive of what that would mean. Is MMT a simple theorem? No, of course it is not. But the events I’ve described do support the initial criticism raised. For whatever reason, lots of non-professional MMT enthusiasts are quite confused about saving and whether it is possible for the private sector to save in the absence of budget deficits.

    Since they are confused, it would be good, IMO, and to the benefit of everyone if we could move the debate forward by communicating the proper definition to them in a way that they can understand. Then people will be able to continue that debate at a higher level, and advance their understanding of important policy issues. Isn’t this what’s important? You could even call it a Pareto improvement.

    It doesn’t have to be about MMT versus MMR. At least, it seems to me that this is, ought to be—or at least could be—a limited criticism of tactical choices and the way that MMT is presented, and not a criticism of any substantive theoretical aspect, about which there seems to be much agreement.

    Unless there really is disagreement over fundamental definitions and concepts, in which case there might be need for a much wider debate.

    • vimothy,

      Fantastic points!

      Scott Fullwiler was the first MMTer that I saw who even agreed with the proper technical definition of saving

      Harsh but true. BTW Scott’s work is awesome and I’ve told him this many times.

      And yes, why did JKH – JKH! – feel he had to write this paper? If the question was as settled as the MMT side seems to think it had been, JKH would have been one of the few people to appreciate their accuracy and diligence and he would have pounded anyone who doubted as only JKH can.

      Yet, he wrote this paper. He spent weeks writing it.

      “No one, again, as far as I know, has suggested that any of this “proves MMT wrong”. I confess that I struggle to even conceive of what that would mean.” ha! me too. What could that mean?

  31. This is where I feel conflicted. Theoretical explanations of the monetary channels seem less than convincing to me, especially in current conditions. But then I ask myself: was the great moderation independent of OMOs? Did it just happen while the Fed fiddled? That seems hard to argue. Maybe assume a small rudder on a big boat; if the necessary adjustments are small, it’s effective.

    But: Volcker’s moves sure seemed to have the intended (big) effects, and very rapidly. Maybe high interest and inflation rates make the rudder bigger?

    I like to think of fiscal policy giving monetary policy the moxy it needs to do its job. Make sense?

    • Cullen Roche says:

      The 70’s are really interesting. What a lot of people fail to note about the 70’s is that we actually had a decent little credit boom. Most Keynesians are quick to shrug Volcker’s actions off as meaningless and claim that high oil prices took care of themselves. But if you actually look at the data you can see that the rise in rates coincided with the peak in the credit cycle. And the peak in the credit cycle preceded big declines in domestic investment. Saying that this had no impact on the economy is just flat out misleading….Now, you can’t be certain that the Fed caused the credit cycle to roll over, but the case isn’t hard to make….

      • I’d love to see a new post/comment thread here on the 70s and 80s. Stagflation seemed to disprove Keynesianism (or at least Hicksianism — in which neoclassicism distorted and coopted Keynesianism via IS/LM while continuing to use Keynes’ name [something Hicks himself eventually disclaimed and discredited]).

        In an MMT/MMR paradigm, how do we explain the stagflation of the 70s (and secondarily, Volcker’s seeming success in taming it)? Labor, at least, seriously underutilized, demand not banging against supply like *at all*. But high inflation. The oil shock has never seemed to me to be a sufficient explanation.

        I’d love to see the kind of clarity MMT often provides brought to bear on that issue.

        • Cullen Roche says:

          I’ve actually had a post lined up on Pragcap for years on this. Never could validate my conclusions with MMT though. Until the last few months….Maybe I’ll post here soon….

          • “I’ve actually had a post lined up on Pragcap for years on this. ”

            Link please! Searched for stagflation volcker and didn’t get any hits.

            • Cullen Roche says:

              Sorry, it’s in my queue. If you can find it there then please let me know because that means I have a security breach at Pragcap. :-)

        • I too have given much thought to the 1970s stagflation, and what the MMT and/or Post K analysis and response would have been. Clearly, it showed a weakness in the Phillips Curve and IS/LM Hicks’ style of “Keynesianism”, which in the mainstream discourse acts as a stand in for “Keynesiansism”.

      • We had a large credit boom in the 1970’s. We should also note Kocherlakota found some effects relating to low government net spending leading to inflation in his paper on the fiscal theory of the price level (FTOPL is similar to MMR/core MMT).

        Then, long term labor contracts for people were simply far more common in the 70’s even outside of unions. By the mid 1970’s people began to demand (and get) guaranteed wage hikes.

        Bretton Woods unwound – this added to the instability of prices.

        Credit bubble + labor costs + huge energy price shocks + Bretton woods = 70’s inflation

        • Nice explanation, puts across the complexity that I’ve always assumed was at play.

          But still wondering: how great a role did OMOs play in maintaining the great moderation?

  32. Meant to be in reply to: http://monetaryrealism.com/jkh-on-the-recent-mmrmmt-debates-2/#comment-3777 . I’ve copied it up there to maintain threadedness.

  33. Vimothy,

    Coincidentally I sort of remembered this but hadn’t followed it well at the time but now I see you comment in the second link below:

    It is important to understand that people have been asking this but realizing the difference. Look at Bob Murphy’s blog:

    “Of Course You Don’t Need the Government in Order to Save”

    Robert [Murphy]: “In particular, I think it is crazy when people say that if the federal government runs a budget surplus, then by simple accounting the private sector can’t save.”
    [Nick Rowe:] That’s perfectly correct, and standard, once you do the translation. Assume [an economy closed to international trade]. Define “private saving” as “private saving minus Investment” … which is how MMTers normally use the word “saving”, or sometimes “net saving”. Then it’s just standard National Income Accounting. Y=C+I+G, and S=Y-T-C, therefore S-I=G-T.


    And there you have it: When MMTers speak of “net saving,” they don’t mean that people collectively save more than people collectively borrow. No, they mean people collectively save more than people collectively invest.

    I’m not trying to make fun of Nick Rowe; he is a professional economist who has written some very nuanced posts relating MMT to more orthodox mainstream economics. But look at what he was forced to type: “Define ‘private saving’ as ‘private saving minus investment.'” As I noted in my response to Rowe, if we define “private saving” as “private saving,” then my critique of MMT stands. (That’s supposed to be funny, by the way — at least insofar as economics can be funny.)


    Mosler commented here but I believe his answer didn’t do much to address “private saving” versus “net private saving” but instead brought in the nongovernment sector and “savings of net financial assets of that currency”


    • Cullen Roche says:

      Ramanan batting clean-up here….This is going to sound awfully critical of MMTers, but I think this is often done to focus attention on the importance of the state. In a perfect MMT world the state would be a true money monopolist and they would be able to wield nearly infinite power over the price and quantity of all money. This is why I often say that the most logical base case for MMT is a nationalized banking system with one vertical component. The state theory renders horizontal money unnecessary. Of course, we don’t live in a state system. We live in a system with a big ass banking system where bankers basically decide how and when things get done in terms of money issuance. MMT tries to downplay the horizontal in order to bring it all back to the vertical. I think they’ll argue that they don’t do this, but I find that hard to believe given their policy proposals and rhetoric….Maybe I am a bit harsh, but balance is necessary in these conversations.

    • I got hung up on this many times when I first encountered MMT.

    • Ramanan,

      Yes, I remember that article. I thought Murphy made some good arguments.

  34. davegerlitz says:

    Good god, I feel like I’m reading Hamlet.

    MMT vs. MMR: One difference is clear: MMT focuses on and emphasizes a 3 sector model: Government, Private sector and External. MMR focuses on and emphasizes a 4 sector model: Government, Private Business, Private Household, and External. Steve Roth now dipping into a 5 sector model by breaking down G into Treasury and Central Bank. Stand by for the next act.

    • “I feel like I’m reading Hamlet.”

      One thing for sure: the guy with the biggest army wins in the end. Fortinbras has my dying voice.

    • Cullen Roche says:

      Like I keep saying on this point – the equation was not intended to “debunk” MMT (as if that’s possible!), but to add clarity. Which, when one understands the depth of the discussion, it does quite nicely. Taking the multi-sector model adds to a better understanding of the monetary system. Making vague statements like “the private sector experiences a loss” when the govt doesn’t spend is very misleading because it’s a narrow view of a broad topic.

      At the end of the day, MMR hopes to bring more balance to the perspective and understanding of the system. That’s all. JKH’s paper is a superb and important step in that direction….

  35. salsabob says:

    Whether agreed with or not, it seems MMT’s important message is a Keynesian one of the federal govt’s insufficient deficit spending both before (decades of stagnant wage growth, falling household savings rate, excessive household debt, predatory lending) the Minsky Moment and after (contraction, lack of demand, insufficient Stimulus, lack of growth, substandard employment). Within that context as well as MMT’s agnostic, if not occasionally hostile, viewpoint of the effectiveness of monetary policies, one might look at their various presentations of the federal govt’s provisioning of (or, lack of) net financial assets (in the form of US Treasuries) as a surrogate for their desired federal deficit spending (see Second liberty Bond Act, 1917). It would seem to be a stretch of the MMT viewpoint to suggest that it believes NFA provisioning alone “would have prevented the housing bubble.”

    At least here in the comment section, we see explicit statements of the netting out of NFA within the private sector, but, of course, with allusion to the resulting grand economic gains (and perhaps a smidgeon of cautionary tales ala 2008). One might assume that the entire discussion doesn’t foreclose on the MMT’s suggested fiscal assist to the private sector’s goings on – i.e. that the obviously grand savings and grander-still investment collusions within the private sector between households and business might, well, how can it be said in a 2008 kinda way, not quite cut it without some occasional outside help.

    The first question is does this focus on “getting it right” regarding savings amongst the players within the private sector help illuminate the larger MMT concern of insufficient federal deficit spending. Perhaps it could by providing a basis for why savings was diminishing – leading to Minsky’s instability and his eventual “moment.” On the other hand, given the pressing direction of Paul Ryan budgets and scheduled Bush Tax Cuts expiration and debt ceiling agreement spending cuts, and the need, as seen by MMT inherents to inform the public of those dangers, does the “getting it right” focus become a distraction at best.

    Then there is that nagging question of whether this is really some sort of sub-conscious attack on the whole Keynesian message of insufficient federal deficit spending to begin with?

    One wonders.

    • Cullen Roche says:

      I don’t understand the backlash against all of this. We were simply clarifying a specific point. I see some people trying to downplay the fact that we’re trying to get the accounting right. Why? Shouldn’t we be very precise about all of this? The concern, one that has become very clear based on non-economist MMTer’s comments, is that these vague and often general terms or descriptions result in a misleading perspective of the way the monetary system works. It’s clear that many people think the economy is a hopelessly lost puppy if the govt doesn’t come in and spend money all the time. Others have gone so far as to say that production doesn’t even matter or that it’s all about how much we consume. These are very misguided perspectives that are the direct result of this loose and imprecise terminology. All we want to do is clear things up and provide a better understanding of these matters.

    • I really thought advocating for deficit spending 16% of GDP would make me immune to charges of Austerianism. I thought putting it up as the first post on MMR would make a point.

      Oh well…

  36. salsabob says:

    Oh, and just to note, there’s no wonder as to where Carney takes it.

  37. JKH,

    Wow, this is a fantastic post. The explanation and investigation into the concept of saving is superb and I can only guess how long it took you to write it. I very much appreciate the economic context, the macroeconomic context, as well as the context in the discussion (with all the appropriate cross-referencing).

    I really have nothing to add from the saving conceptual and accounting viewpoint, because I think it stands on its own. However, I would like to back you up (not that you need it) regarding this point:

    Decomposition is not trivial or meaningless simply because it becomes X = X on consolidation. To adopt such a view is to adopt the view that analysis as a generic endeavor is useless and worthless, because after all this is the type of thing that analysis does. And to conclude that X = Z + (X – Z) is useless because it always consolidates to X = X is mathematically naive in the extreme. The idea that sets decompose into elements is at the foundation of mathematical analysis. I’m not surprised that some people would find the tautological expression of such decomposition curious, but to extend this to Wray’s type of response must indicate a total lack of interest in analysis or some agenda otherwise.

    Completely agreed.

    I will illustrate another ‘tautology’ that while seemingly simple provides huge mathematical and engineering insight to a wide range of problems:

    A x = lambda x

    I would not be at all surprised if you were already familiar with this, but allow me to expound.

    Ax – lamda x = 0 => (A-lambda) x = 0

    Again just rearranging terms and this just seems like silly algebraic manipulation with no purpose. Except there is. This is the eigenvalue problem.

    Since x is a non-zero vector quantity, then det(A – lambda) must be zero. And what this allows you to do is decompose a system of equations (A) into an uncoupled system, where lamba is the set of eigenvalues (roots) that solve the characteristic equation, and X is the set of associated eigenvectors.

    This has all kinds of applications in mathematical analysis, such as determining if a system of partial differential equations is hyperbolic, parabolic or elliptic. (Useful especially in fluid mechanics). In structural analysis it can break the response of a coupled system (such as a linear elastic structure) as the superposition of a uncoupled modes (natural frequencies lamba and mode shapes X).

    While this is not directly applicable to the discussion at hand, it illustrates that seemingly ‘tautological’ manipulations can uncover the decomposition of a complicated system. In exactly the same way the S = I + (S-I) decomposes the components of Saving into investment + net saving. Even though S-I = S-I, S is not necessarily = I, and not necessarily = to S-I. The decomposition is useful because it adds insight. In exactly the same way that while A x = lamda x looks tautological, it also provides a wealth of decomposed information.

    Keep up the fantastic work! I always look forward to your thoughts!

    • Thanks so much

      I’m really glad that general point has been appreciated – and you’ve brought it to additional light with an elegant but heavy weight example as well

      • Absolutely! And while MMT dispels so many fallacies of composition that exist in mainstream economics today, it needs to not fall victim to its sectoral balances being unable to be properly decomposed. And as such these points being brought up by MMR showing where definitions (such as savings) are unclearly stated in the MMT literature that they lead to poor decompositions and offering ways to view these concepts in a clearer light should be embraced by the MMT community. I certainly find these discussions helpful and adding to the body of thought, not detracting from it. So Thank You!

    • Cullen Roche says:

      The math always struck me as a tautology until I got to college and figured out how important it was. http://2.bp.blogspot.com/_DXfbDjZYtMM/TQ-u0EC4I0I/AAAAAAAALfY/Q-Jq-B-BcWY/s640/fsu+tri+delts.jpg

      I’ve since forgotten the importance of the math. I should probably revisit it some time in the future.

  38. It says a lot about MMT that they responded to a good insight in such a negative way. It would have been easy to build a bridge out of this discussion rather than viewing it as an attack. MMT is a young theory and it’s not perfect. I think they would be wise to acknowledge that and be more open to criticism and outside ideas rather than being closed minded and pushing everyone away. They have little or no allies in the field of economics and it’s because of responses like this. It’s not surprising that such an excellent theory has gone nowhere fast.

    • Cullen Roche says:

      It’s been a defensive few months. I think we’ve all gotten a little emotional at times. It will pass and when it does we’ll remember that MMR and MMT are moving in the same general direction….

  39. binve,

    That has all kinds of applications in economics and econometrics as well!

    • Hi vimothy,

      Excellent! I figured that was the case. Eigenvalue analysis is such a useful applied mathematics tool that is used in several disciplines. Much like optimization via linear algebra. You can use it to find optima in economic systems just as easily as finding optima in structural systems (even if the systems are inherently non-linear by taking linear approximations and using move limits and iterating toward a solution via sequential linear programming). Do you have some examples where eigenvalue analysis is used in economics? I would love to brush up on that.

      • binve,

        Yes–Lemme see what I can come up with..

        BTW, the linearisation theorem is also commonly used in econ to study the steady-states of non-linear dynamic systems.

        Cullen, the peasants are demanding LaTeX for the comment boxes!

      • Eigenvalue analysis is commonly use in econometrics, where we’re frequently manipulating large data matrices (e.g., to write a transformation matrix in its canonical form), in applied macro-econometrics (e.g., to test a series for a unit root), in macro (e.g., to solve a huge variety of dynamical systems and qualitatively classify their trajectories), in micro (e.g., to model a continuous time tatonement process that brings a system to Walrasian equilibrium), and so on.

        Here’s a simple macro example that I’ve posted elsewhere in the past, using a dynamic Keynesian multiplier-accelerator model, where s in (0, 1) is the marginal propensity to save, Y(t) and I(t) are output and investment at time t and G-bar is govt expenditure (a constant):

        (1-s) Y-dot = -sY + I + G-bar
        I-dot = aY-dot – I

        The system is simple, with a unique steady-state at I = 0, Y = G-bar/s. Using eigenvalue analysis (which could be too messy to write out in the comment box), it’s easy to show that this cannot be a saddle-path—instead it is a source if a > 1, a sink if a < 1. Under some parameter values, that sink is an attracting spiral. Under others, that sink is a repelling spiral, and the economy goes through a series of ever more extreme booms and busts until corrective action is taken or output goes to zero.

        I could maybe email you some of my math econ or metrics lecture notes or suggest a good textbook if you're interested in taking a more in-depth look at it use in economics.

        • Vimothy,

          This is good stuff, thank you. Yes, in engineering, thinking of eigenvectors as a coordinate transformation to some set of natural coordinates is also a common theme (e.g. taking a general 3-D stress state which has normal and shear components and finding principal directions where along which only normal components exist).

          Yes, please do suggest some good textbooks that discuss these ideas in economics and econometrics. I am sure I could fumble around and find a bunch of texts, but not necessarily ‘good’ texts. Your suggestions would be superior choices. Thanks!

  40. binve,

    A good textbook covering economic theory is,

    Further Mathematics for Economic Analysis by Sydsaeter, Hammond, Seierstad & Strom


    It’s a bit harder to think of something comparable for econometrics, but I’ll try to rummage something up.

  41. Hello, I’m from MMR and I’m here to help.

    Much of this discussion centers around the insistence of the importance of the clarity offered. One might turn that around and suggest that a larger clarity requires a discernment of importance of the offered clarity.

    In a review of the bidding, we do get statements that the intent is not to undermine basic MMT tenants nor its inherents. But, one can almost visualize the sighs of exasperation in the speaker caused by those recalcitrant MMTers, once again, not only incapable of grasping the offered clarity but being just too darn sensitive about it.

    Often overlooked here are the various bids, implicit and explicit, that the lack of MMTers grasping the offered clarity does indeed undermine MMT’s fundamental tenants. My favorite, of course, is the particular train-of-though that would allow one to discredit the existence of nuclear bombs after reading “Einstein’s 23 Biggest Mistakes.”

    Further, we also have a blogger, with an audience size that most economists (let alone those of MMT/MMR world) would die for, suggesting that actually-credentialed economists, regardless of pedigree, should de-certify themselves if they continue in their lack of grasping the offered clarity.

    My, my, that does make the offered clarity seem awfully important. But is it? Let’s see if we can clarify.

    I would suggest that the primary “take away” from MMT exposure is that spending by US federal govt is neither constrained by its tax revenue nor by its “borrowing.” While I stand open to enlightenment, I see nothing in S = I + (S – I) that puts that in doubt. Outside of the rather dubious “Einstein Mistakes = No bomb” train-of-thought, the lack of impact of S = I + (S – I) on MMT’s primary “take away” greatly diminishes, at least in my eyes, its importance to MMT. Don’t get me wrong, the equation certainly remains interesting, but not particularly a basis for de-certifying credential economists with a MMT bent.

    Perhaps the importance of the equation is derived from its potential to either foster or hinder attempted applications of MMT’s primary “take away.” There is little doubt that unconstrained federal deficit spending is greatly appealing to those nefarious Keynesians; could the offered clarity of S = I + (S – I) finally be the key to resolve that debate?! Given the enormous numbers of broken hoists as well as petards on that large field, I can only wish you luck. But, I will note, that your battle, for or against, is with John Maynard and a host of others, and not necessarily with a particular Randall, Warren or Bill. Clearly, the importance of your clarification to MMT through this avenue seems to be fairly limited or at least on the other side of a very treacherous and well-trotted field of battle – again, good luck with that.

    But let’s back up to that MMT notion of federal spending neither constrained by taxes nor debt. In regard to the latter (i.e., federal debt), there are MMT voices that have suggested that it is “not needed” – for it only provides, as a service to a relatively few in the private sector, risk-free (arguably) interest-bearing (currently, arguable as well) financial assets. Sure lots of other discussions in MMT world over federal debt – pretty hard to get away from it in any discussion these days whether purely economic or not. Questions arise as to: what is that debt actually for; is it truly debt; does it impact household and businesses in the private sector particularly in regard to lending, money supply, and, oh yes, Saving and Investment (there it is!). It is fine to attempt to make aspects of these issues clearer, even perhaps, hopefully, more interesting. But to be truly clear, perhaps it is important to clarify contextual importance (to avoid misunderstandings and misuse), that within MMT’s primary “take away,” your issue arises within a framework where one can also credibly entertain the possiblity of “not needed.”

    Once those espousing the importance of the S = I + (S – I) clarity come to terms with its actual rather- constrained relevance to MMT’s primary though (putting aside, of course, the “Einstein Mistakes = No bomb” argument), then perhaps we can also have a more constrained presentation of it and then gain the measured, unemotional response that apparently is highly desired. I think that will come about when the offered clarity doesn’t start with “those MMT guys don’t …yadda, yadda” and the dire implication of some substantial difference between two major schools of economic thought with the very continued existence of the world on the line.

    You know, Einstein got the 7th Proof of e=mc2 wrong, but I think that’s generally been forgotten while da bomb remains a reality. My bet is that S = I + (S – I) will have as interesting a fate as the 7th Proof and MMT’s notion of federal deficit spending unconstrained by taxes and debt will be da bomb.

    • Thanks for your comment.

      First, the “constrained” presentation of S = I + (S – I) already happened – several weeks ago at John Carney’s blog. It was so constrained that it didn’t even mention S = I + (S – I), yet it was the same subject being discussed in verbal form. As I said, I never actually promoted the thing. It stands on its own. It was that constrained presentation that elicited some of the not so constrained responses from those who wished they’d never seen it, including that of Randall Wray.

      It was here:


      What has followed in this post is in part a response to that kind of “critique” by LRW, plus some consideration of a range of more constructive observations and questions about the subject and the equation.

      So let’s put that issue of being “constrained” in that proper opening context.

      Second, as I’ll say in my more general comment that follows this one, the direct issue here is not the content of MMT, but the communication of the content of MMT. There’s a difference, which I point out in the comment following this one.

      Third, the specific issue that is the object of the communication question here is of limited direct scope, given the full range of MMT operational and policy concerns. It is not about MMT policy, and it is not about the general subject of MMT’s description of monetary operations. The equation itself is fundamental for how it represents a particular aspect of MMT analytic technology. But it is quite oblique in the sense of the larger range of all issues about MMT, as you have correctly pointed out. The question is how important is it, given its apparent obliqueness in this regard?

      My first exhibit of its importance is the following “bid”, submitted by Tom Hickey:


      My second exhibit of its importance is the following “bid”, submitted by Scott Fullwiler:


      Both bids are very reasonable, but well qualified, which I again address below. The qualifying point has to do with whether I’ve tied the issue I’m actually addressing here to a more general one of MMT somehow being “wrong”. I don’t think I have, as explained below.

      Fourth, the equation has nothing to do with questioning Keynesian “capacity” at any level. That’s been explained enough already.

      Fifth, the equation has nothing to do with the question of whether or not debt is “needed”. That’s an interesting one but quite separable.

      These last two issues are fundamental ones of policy and operations, and they are good ones. But to suggest that they are necessary gatekeepers of what matters to the effectiveness or success of MMT is to miss the point entirely. You have created a straw man requirement for relevance. Even if the car is in top shape and waiting on the lot to be driven away, and even if the rules of the road are well known, you still need a driver who can operate it and follow those rules. And that probably applies even more so for drivers’ school.

      I shall consider the 7th Einstein proof criterion as a sort of personal challenge, although I’m not too comfortable with the bomb part of the analogy. In a world where any demonstration of monetary economics must follow accounting rigor by necessity (not everybody knows that, but nobody knows it better than Scott Fullwiler), and where there is more than one way to get to the end point of demonstrating that rigor, it remains to be seen what version will come out on top.

      Thanks again for your comment. I thought it was a very good one.

      • And thanks for your very thoughtful comment as well. I encourage further drilling per TH and sloppiness cleaning per STF, but suggesting to your readers that an occasional looking up from the task at hand to admire the surrounding landscape keeps things in perspective. A reference to your fine comment here would be more than sufficient.
        Good luck with developing that 7th proof, but do stay clear of da bomb – the FBI frowns upon that particular application.

  42. There’s quite a discussion about this post here:


    I thought the comments from Scott Fullwiler and Tom Hickey were pretty constructive, and as much as I could have hoped for with this post. I just want to remind folks that a good part of the post consists of documented comments of MMT followers and others, and in two cases MMT leaders. The issue that I focused on was not primarily one of the MMT content in substance, or its framework – rather it was the communication of MMT content considered fairly broadly. And I think Scott and Tom honed in on that pretty well. For example, nowhere did I question the value of the sector financial balances model. In fact, I seem to recall doing the opposite.

    I don’t think that I’ve actually said that the content or the framework for MMT is “wrong”. But I think there’s good evidence that communication in some cases has been confused in a particular area, and that it has been confused in a similar way in those cases. This post was a way of putting a lot of that in one place to show that particular pattern. Nevertheless, the scope of the area examined is quite narrow. I did note, but did not dwell on the fact, that a couple of other academics have broadened the treatment of a similar general issue to a somewhat wider scope.

    I’m not exactly errorless myself. I may have been guilty of rhetorical overreach by association, in saying – “I concur with John Carney when he says “This is not Economics” – maybe. That’s a fairly blunt and charged statement. It may not have been the best way for me to say what meaning I might have intended at that particular point. But if you consider it in the context of communication, rather than content, and tone it down a notch or two, it’s more applicable, if no less painful. Also, John’s latest on the issue appeared to be heavily weighted to the communication issue. As far as Steve Waldman’s comments were concerned, I did say I agree with the observation about style. But I don’t know exactly where the bias he asserts as being the cause of the style lies in a range between perception and reality.

    A number of people have suggested that my observations have little to do with the quality of MMT content. That’s true to a degree. That wasn’t the direct subject. This had nothing directly to do with the JG/ELR policy for example. It also has nothing to do with content in the sense of the broad approach to the interpretation of monetary system design. I said to Scott some time ago that I might favor an approach that would be different from the particular analytical framework he uses. There’s nothing about that in this paper. It’s hard for me to imagine that there is any university professor in the United States that understands monetary operations better than Scott Fullwiler. So I would hope that whatever the differences in approach might be, we would end up with the same pieces of the puzzle in roughly the same arrangement. The issue of what you then conclude about a sovereign currency issuer’s operational potential for deficit spending etc. would no doubt be quite similar. But the approach to putting the pieces of the puzzle together might be different.

    There’s also the issue that the MMT leadership group is a relatively small one. There are differences of style etc. within that group. I think there is little doubt that Scott is the leader in putting the finishing touches on monetary system accounting and operations, etc., and no doubt on some other things as well. So those communication patterns I touched on might differ across that group accordingly.

    Finally, had Randall Wray not responded to John Carney’s post in the way he did, my paper/post might have looked different, although I can’t be certain of that. Tom proved himself to be doubly wise here. I recall Tom saying something earlier to the effect that sharply edged rhetoric can sometimes elicit more deeply thought responses. That may have been an effect here.

    I should add that this particular issue of saving is only the tip of the iceberg in terms of my personal preference in presenting the issues that collectively might be considered “descriptive” from an MMT perspective. But that’s not to claim that MMT is “wrong”. Tom is right that I am not a professional economist. But I have some rough instincts on how economics might interface more effectively with accounting and finance as an explanatory framework. The emphasis on the monetary system that MMT uses going into such an exercise is quite relevant to that view, even though my way of describing it might be different. I think it’s a given that MMT has been the market place leader so far in “grafting” an understanding of accounting and banking right into the guts of economic thinking. And I think it’s also apparent that the area requiring the graft is something that has been very important to the financial crisis and the understanding of it – before, during, and now. Maybe some of the specialists assisting in the operating room don’t need to come from the economics profession, and shouldn’t.

  43. S = I – (S – I)

    OK, basic question here: Are the first ‘S’ and the second ‘S’ meant to represent different quantities? If yes, then why are they represented by the same symbol? If no, then how is the equation at all useful?

    On the subject of economics I’m only an interested layperson, but IMO this post takes a lot of space to make the point that in a 4-sector model (where households and firms are separated), households could net save while G and X are in surplus, as long as F (firms) are in deficit. OK, sounds fine. But I don’t see what the big deal is…?

    I mean, if the point is to emphasize household saving, great! But the focus of MMT seems to me to be on the understanding the role of gov’t policy in a modern economy. I’m not sure the household/form distinction is needed to make the primary arguments in that area.

    I assume the reason that Mosler took your equation to be describing a one-sector model is that, taken by itself, that’s all it depicts.

    • By “G and X are in surplus” I mean more precisely: “(G – T) and (X – M) are both negative”. Sorry!

      • No worries. Experts are confusing this, and JKH felt compelled to write this epic essay. It’s hard to get right and easy to get wrong.

        • Thanks :)
          I’d still love to see a concise explanation of why it’s not appropriate to just simplify “S = I + (S – I)” to “0 = 0″, as any 8th grade algebra student would do. “S = I” means something interesting. “0 = 0″ does not.

          Is the 2nd S/I pair conceptually different than the first? If so, why not write “S = I + (S’ – I’)” or some such?

          The point of all this discussion seems a matter of loose terminology; specifically that S ≠ S – I (ie “gross saving” is not the same as “net saving”). Which seems pretty obvious. So S & I could increase together while the other sectors’ balances remain constant. Am I missing something other point?

          • Госбанк says:

            The point is, as I see it, to tighten up somewhat loose terminology: what is left from consumption(S) can flow either into government paper (S-I) or into private sector firms in the shape of commercial bonds, newly constructed RE, newly issued shares, etc (I).

            Some people sometimes inadvertently or advertently obscure that forked flow. One may interpret in various ways what “saving overall” means, but one hopefully will understand the above without any ambiguity. Precision is the goal of the exercise I imagine.

            • OK, well if clarity is the point, I still am not seeing how S=S+(S-I) is helping to express anything, especially to a layman like myself. All I see is 0=0.

              Would the following formulation be another way to communicate the point?

              S = I + (G – T) + (X – M)
              Moving terms around:
              S = I + (G + X) – (T + M)

              Let S’ = G + X [flow into private sector]
              Let I’ = T + M [flow out of private sector]
              S = I + (S’ – I’)

              The “prime” values then just represent the vertical/external components of “net saving” (aka S-I). This simply reflects that in any N-sector model (where N>1), S and I need not be identical.

              Also, given constant gov’t and external sector balances, the sum of the flows (S + I) can vary based on the total horizontal transactions. But their difference, (S – I), cannot. Right? This is simply gross vs net.

              • TheJ,

                I still am not seeing how S=S+(S-I) is helping to express anything

                For one thing, you have the equation wrong!

                In actual fact, we have,

                S = I + (S – I)

                Or in other words,

                [Saving ] [is equal to] [Investment] [plus] [Net Acquisition of Financial Assets]

                • Yeah, sorry, that was hastily typed (a fact I was hoping would be obvious given my subsequent expressions, which I believe are correct).
                  But to your plain-english reading of the equation: why not use distinct symbols to represent “[Net Acquisition of Financial Assets]”? Such acquisition has to come from outside the private sector right? Wouldn’t the use of distinct symbols for NAFA help in terms of the “clarity”? An equation that, on its face, reduces to “0=0″ IMO will continue to confuse people.

              • Госбанк says:

                S = I + (G + X) – (T + M)

                It’s a legitimate decomposition, as any other, so long as you are clear about specific terms’ meaning and the goal of such a decomposition.

                One could interpret the decomposition as “proving” that exports are a “benefit”, contra some MMT’ers, precisely because they increase Net Financial Assets(NFA) so dear to any chartalist’s heart :)

                In the same fashion, one could interpret X as evil being a drain of NFA’s.

                Clearly, none of those interpretations, as well as the diametrically opposing treatment of imports as “benefits”, make much sense when having been derived solely from the algebraic identities and without taking into account the real world that the identities are trying to, however crudely, represent. They can give one only a hint, possibly misleading, in a manner similar to correlation winking at a possible causation.

                • OK, I’m with you. Though I don’t think I’ve read anything out of MMT that advocates for a particular value of X-M, positive or negative. The external balance is what it is. Yet, whatever the value, any negative effect of (X-M) on the domestic economy can be responded to via public policy.

                  And still, I’m not seeing the point in degenerating the useful statement:
                  S = I + (G – T) + (X – M)

                  into the trivial:
                  S = I + (S – I) (i.e. 0=0)

                  It seems to be trying to obscure the fact that private sector NAFA has to flow in from some combination of the other two sectors. As if the only variables that matter are S and I. Screw G, T, X, and M! So is this all just about ideological framing? JKH says S=I+(S-I) is a 3-sector equation, but that leaves out two of the sectors? Odd.

                  • Госбанк says:

                    exports are real costs and imports real benefits
                    Warren Mosler

                    • Cullen Roche says:

                      In my opinion, probably the most misleading position MMT takes….and largely a result of misunderstanding the identity (or wanting to understand it in a certain way for political purposes)….

                    • I find with a lot of Mosler’s comments I’ve read, he states things so tersely that I don’t really understand exactly what he’s saying. I’m sure it’s all perfectly obvious to him, however :)
                      I do suspect the the word “real” is paramount in the quote you gave. Perhaps his point is that exports move “actual stuff” out of the country, and in that sense they represent a “real cost”? As opposed to a monetary or virtual one.

                    • Cullen Roche says:

                      That is what Warren means. It’s better to get real things than pieces of paper. But I don’t think that’s the full story. For instance, in trade with China we send over pieces of paper in exchange for real goods and services. But who’s really winning that trade? Well, we both think we are, but to claim that China is getting the short end of the stick because they’re just getting pieces of paper is nonsense. No, China is getting investment and real jobs through this trade. America gets cheaper goods that they throw in a landfill in 5 years. I don’t know who’s really getting the better end of the deal. I don’t think we’d do it if we thought it was destroying us. But I do think it’s wrong to imply that we should not be concerned about becoming a nation of consumers who produce nothing. Some MMTers have gone so far as to say that production is overrated or that a country doesn’t need to produce what it can consume from abroad. This might be true in some vague mathematical form, but it’s wrong in real-life….

                    • But I do think it’s wrong to imply that we should not be concerned about becoming a nation of consumers who produce nothing.

                      Right, but that’s not something a proper MMTer would imply, is it? If the MMT focus on maintaining (near) full employment domestically is effective, then we won’t ever be a nation of non-producers, right? Assuming the work being done produces valuable, “real” things.

                • “One could interpret the decomposition as “proving” that exports are a “benefit”, contra some MMT’ers,”

                  Yes, that is right net exports increase the financial assets of the private sector. Also X is like G and hence acts to inject demand.

                  “In the same fashion, one could interpret X as evil being a drain of NFA’s.”

                  You meant M right? Yes imports act to drain demand and increases indebtedness to foreigners.

                  “Clearly, none of those interpretations, as well as the diametrically opposing treatment of imports as “benefits”,”

                  Imports are benefits are at full employment but since imports act to drain demand and create more unemployment, it is misleading to call them benefits.

                  The MMT terminology is fraught with confusions on the external sector. Just like they start talking of “net saving desire” they imitate that for the external sector .. such as foreigners’ “net saving desire” “in the currency of the issuer”. But this is illogical. For the domestic private sector, net saving was saving net of investment. But for rest of the world, what is “net saving desire”?? So one hears of frequent commentary about “foreigners saving in the currency of the importing country” but is that saving?? That’s more appropriately portfolio preference in Tobinesque language.

                  As JKH said, all this is just the tip of the iceberg :-)

                  Anyways, I am going to be all ears to the usage of “dollar savings” (such as “world dollar savings to the penny”) for now even though open economy is what I like most.

  44. I think “MMT is wrong” was written by me in the specific context though people took it the different way. I still have strong views but won’t express it here.

    Bill Mitchell wrote a post today but this still continues!



    “Think about this carefully. Clearly, if households do not consume all their disposable income each period then they are generating a flow of saving. This is quite a different concept to the notion of the private domestic sector (which is the sum of households and firms) saving overall. The latter concept (saving overall) refers to whether the private domestic sector is spending more than it is earning, rather than just the household sector as part of that aggregate.”

    Yeah right!

    • Posted this at Billy Blog:


      “Period 1, the budget is in surplus (T – G = 1) and the private balance is in deficit (S – I = -1). This means that the private domestic sector is spending more (via consumption and investment taken together) than it is earning. So it is dissaving overall. Note that households could still be saving (that is, not spending all of their disposable income). But as a sector, the combination of firms and households would be dissaving.”

      First you mention that “The latter concept (saving overall) refers to whether the private domestic sector is spending more than it is earning, rather than just the household sector as part of that aggregate.”.

      So let us take “saving overall” to mean “saving net of investment” (which needs a stretch of imagination but let’s just do it).

      So first you write it is “dissaving overall” in the example given which is fine assuming the definition of “saving overall” you propose.

      But then you say “But as a sector, the combination of firms and households would be dissaving.” which does not contain the word “overall” and hence incorrect because the combination of firms and households may still be saving in spite of having a negative financial balance. It is certainly possible for BOTH households and firms to be saving simultaneously and both can have a negative financial balance (simultaneously). Both individually or collectively.

      Has this happened in the real world? Yes in years around 2000 in the United States. Households had a positive saving and firms had a positive saving. Households were in deficit and firms as a whole also in deficit.

      So it is certainly untrue that “But as a sector, the combination of firms and households would be dissaving.” because it’s not necessarily the case.

      • Finally did it … got banned it seems :)

        • Can’t believe my eyes.

          And can’t let this go.

          Out today, but back later.

          • Well I nailed it.

            His post looked like a sneaky way to defend – trying to push the argument that whatever he has been saying is right if he defines “saving overall” as Saving minus Investment. That’s smart (maybe) but he still ends up conflating.

            Why is it so difficult to understand that

            Private Saving = S = S(h) + FU(f)

            Private Sector Balance = S – I.

            and that I ≠ – FU(f)

          • I’ve been wrong so many times in my life. It’s the path to fun, usually.

        • A badge of honor! I’ve never commented at BB.

        • lol!

          Bill might be one of those guys that doesn’t have the accounting expertise as another MMTer has suggested might be the case. I still think what he means to say is what you are suggesting he should be saying (dissaving overall), but he is not saying it because he doesn’t realize the word ‘overall’ makes an accounting difference since he already said ‘combination.’ So do you think this is lack of accounting expertise or that he doesn’t get the difference b/w S and S-I in a broader economic sense?

          • I do not know why the same thing keeps happening.

            My guess is that they are so strongly conditioned to think of saving as increase in financial assets that this happens.

            • MarkoS,

              Didn’t notice this comment earlier.

              Bill constantly says the private sector cannot save if the government doesn’t run a deficit. Now there are various versions of this statement some right, some wrong.

              So it requires some attention to figure this out because he has a tendency to make a comeback as in the last post here:


              The point is he is changing the meaning of “overall”.

              The fact is the private sector can have positive saving even in the absence of government deficit. Both the household sector and firms can simultaneously have a positive saving without a government deficit.

      • He must have banned you because he doesn’t want people to see that he is making a mistake.

        JKH, I hope you’ll elaborate on this point later. It would be really helpful for other readers here.

    • Госбанк says:

      You meant M right?

      Yes, I meant “M” of course.

  45. This will interest many of you. MMTers are still very confused by these developments.


  46. I can’t seem to publish comments over there. I think Winterspeak is blocking comments that don’t agree with him. MMTers seem to think that credit does not add money to the monetary system because there is an assets and liability involved in the creation of credit. That explains why they’re focused on net financial assets and don’t seem to understand the importance of the credit system.

    • oh please, I so often see that comment and not just on this subject all over the place on the net. The software breaks down and everyone is up in arms about censorship, try a test first and if that doesnt work either theres a problem.

    • I should say I comment on a lot of subjects and quite a few seem to use Disquis and ive lost count of the times comments have disappeared and its got to the stage I copy my post before hitting send as ive lost so many.

      • Yes I know this isnt disquis, was making a point about these systems not being perfect before anyone gets on to me about being prescise!! :)

    • Ive also just had a look, there was a post 3 mins after yours and nothing since, even though posts before were going up every few minutes before that. Whats that telling you?

    • If you head back over there Lars you’ll see your comments posted…..although if I were you I might wish for a couple of them to be erased.

      A zero percent bank loan the SAME as govt spending?

  47. Lars
    I think Winterspeak is blocking comments that don’t agree with him.

    Just throwin’ that accusation out there, are ya? Sure, why not! Did he also put the fluoride in your drinking water?

  48. binve
    The decomposition is useful because it adds insight. In exactly the same way that while A x = lamda x looks tautological, it also provides a wealth of decomposed information.

    Uh… but S=I+(S-I) simplifies to 0=0. I guess it feels good to cut the other 2 sectors?
    OK look, this is very interesting (because A is matrix and lamba is a scalar; not tautological!):
    A x = lambda x
    OTOH this is not interesting at all:
    A x = lambda x + (A x – lambda x )

    • Cullen Roche says:

      The insight is in a better understanding of the way the economy works. You’re right that the equation simplifies down to something that appears benign. But then again the sector balance equation doesn’t tell you much about the economy either without digging into each of the components and better understanding them. That’s what this equation does. It takes the heart of our economy and breaks it down.

      This problem arises from confusing comments by some MMTers who have said that the pvt sector experiences a “net loss” when I>S or that the pvt sector can’t save without deficit spending. JKH’s equation expands on these points to show us that S-I alone doesn’t give us any special insights about the pvt sector’s position and can lead to very confusing and inaccurate understandings. One must dig deeper. And when you do you see that investment drives the economy. Investment is the backbone of private sector equity. Private saving is comprised of saving created by investment AND NFA.

      I think MMTers have made much more of this than was necessary, but the confusion clearly still remains….

      • Cullen Roche
        You’re right that the equation simplifies down to something that appears benign.

        The simplification doesn’t “appear benign” it is benign: 0=0. It’s to get more boring than that ;)

        But then again the sector balance equation doesn’t tell you much about the economy either without digging into each of the components and better understanding them.

        The standard FSB equation informs us about the relationships between the sectors. You’re right it’s glosses over the details, and thus leaves a lot out. But that’s its purpose. The glossing over intentionally emphasizes the high-level relationships between sectors. It’s big picture, yes, but at least it doesn’t reduce to zero!
        Granted I’m a total outsider to both MMR/MMT with only a math/CS background, but I can’t help but interpret S=I+(S-I) as anything but sophistry. I mean that in a 100% constructive way, I’m not a hater.

        MMTers…said…that the pvt sector can’t save without deficit spending. JKH’s equation show[s] us that S-I alone doesn’t give us any special insights about the pvt sector’s position and can lead to very confusing and inaccurate understandings. One must dig deeper. And when you do you see that investment drives the economy. Investment is the backbone of private sector equity. Private saving is comprised of saving created by investment AND NFA.

        OK, so you’re saying that the gross values (S or I individually) haven’t been given enough attention, compared to the net (S-I). You’ve piqued my interest to hear more about that idea. But, still, I don’t see how degenerating the SFB equation helps clarify your argument. Why doesn’t MMR just say this:

        S = I + Z
        Z = NFAA = sum of the deficits of all sectors other than the private domestic sector

        Pick a more appropriate letter than ‘Z’ if you like. That formulation at least doesn’t reduce to nothingness. Your point seems to be that if Z is held constant, then S will increase if I increases. I.e. more investment means more saving. Is that it in a nutshell?

        • Cullen Roche says:

          One might be inclined to believe that I>S is a bad thing if you read the FSB as is. That’s a very wrong conclusion. You have to break down each component further. Just as someone else has done with the foreign component. The equation X-M doesn’t tell us anything in the FSB. In fact, one might be inclined to conclude that X>M is a bad thing or that imports are real benefits that allow us to afford lower taxes or that “Demand leakages are a good thing” as Warren has previously said. I would argue again that that’s a wrong position. You have to look at each component on its own and elaborate on it and its impact on the economy. Investment is the backbone of the private economy. Not net financial assets. Not the govt. We’re not trying to downplay the govt, but I do feel that MMT often uses the FSB to play to the strengths of govt in order to satisfy an argument for govt spending. Which is fine. One can use the FSB to do that and I am certainly not against govt spending, but let’s maintain some balance here. Investment is where the primary increase in living standards results and that’s largely a function of private credit expansion with NFA playing an important but facilitating role.

        • My views on this are in section 4 of the paper/post.

  49. Ramanan,

    Here’s a theory:

    I think Bill Mitchell defines “household saving” by qualifying it as “after distributions” or something like that. I haven’t read his stuff in a while, but I recall something like that. I assume “after distributions” refers to a (fictitious) distribution of undistributed profits. In other words, it assumes implicitly that undistributed profits are instead paid to the household sector in dividends, and then reinvested by the household sector as a new financial claim. If that’s right, then he effectively defines household saving to be what normally would be viewed as private sector saving. That would be a neat trick – but I’m not certain about this, so I’d appreciate your feedback.

    I believe that’s how he gets away with defining S as “household saving” rather than private sector saving.

    He’s a little nebulous at first with the phrase “total saving” (although narrows it down somewhat as he goes along):

    “(2) Y = C + S + T, where S is total saving and T is total taxation (the other variables are as previously defined).”

    In fact, in order for coherence, the S in the equation C + I + G + (X –M) = C + S + T must refer to private sector saving, because the only other two sources of saving are government (T – G) and external (M – X). That leaves the private sector to fill the gap – not the household sector. But I believe he flips that S into to his own definition of household saving, which for coherence must arithmetically account for the entire private sector, however he thinks of it or whatever he calls it.

    So the point is that it requires an artificial financial construction in order to define the specific S as household saving in this equation. But that’s what he does for both this equation and the 3-SFB model:

    “The sectoral balances equation says that total household savings (S) minus private investment (I) has to equal the public deficit (spending, G minus taxes, T) plus net exports (exports (X) minus imports (M)), where net exports represent the net savings of non-residents.”

    The next thing he does (I think) is define investment as having the same effect on saving as does consumption – i.e. both are treated as withdrawals from income for the purpose of defining residual saving. Of course, that’s not at all how saving is defined normally – certainly not how Harless defines it, or most other economists. And of course it’s definitely out of synch with corporate financial reporting, which doesn’t help.

    Now, in a closed, balanced budget economy, S = I.

    (That’s the economy that Harless uses for his causality illustration paper.)

    So he defines S there as household saving (artificially reconstructed as above).

    Such a constructed S is indeed a positive number that equals I in a closed, balanced budget economy.

    In this closed, balanced budget economy, he would then define saving for the business sector effectively as (I) – that’s negative I. That’s because all profit is assumed by construction as paid in dividends, which leaves no contribution to business saving from profit. And all investment is treated as spending. Such business spending then becomes dissaving from a reconstructed base of zero retained profit. And so such spending on I therefore generates negative business saving in the amount (I).

    So his defined household saving is S = I, his defined business saving is (I), and therefore his defined private sector saving is (I) + S, or (S – I) = 0.

    That’s how he concludes (S – I) = private sector saving = 0, notwithstanding that S is his defined measure of positive household saving.

    On that basis, when you open up the economy to government imbalance and the external sector, you’ll get (S – I) not equal to zero. So it becomes a measure of “net saving”.

    Therefore, in the closed, government balanced case, he would simultaneously call (S – I) = 0 private sector saving, AND private sector net saving. Both are zero in this case.

    (Harless never suggests that the fact that S = I in his closed, balanced model would mean that saving is zero. He speaks of saving in the normal way in his model – net of consumption only. So there’s definitely a contradiction between the comparable Mitchell and Harless models.)

    And so he gets to the following:

    “It is clear that if we had a balanced budget (G = T) and an external balance (X = M) then (S – I) = 0. Does this mean that there is a zero flow of saving in the economy? Definitely not. Households could still be consuming less than their disposable income which means that S > 0. What it means is that the private domestic sector overall is not saving because it is spending as much as it earns. It also means that when the government is running a balanced budget the non-government sector must be spending exactly what it earns and is not accumulating net financial assets (as a sector). When the external sector is in balance, then that conclusion applies directly to the private domestic sector. Think about this carefully. Clearly, if households do not consume all their disposable income each period then they are generating a flow of saving. This is quite a different concept to the notion of the private domestic sector (which is the sum of households and firms) saving overall. The latter concept (saving overall) refers to whether the private domestic sector is spending more than it is earning, rather than just the household sector as part of that aggregate.”

    In response to all of it, I see no reason to be ambiguous by defining S as household saving. It’s not in any coherent way, relative to the equations in question – not without an artificial conversion of undistributed profits into distributed profits. In reality, undistributed business profits exist that constitute a source of saving separate from household saving – just as there are in the Kalecki equation, for example. The S in the equations here can only be coherently defined as a private sector S, since the only other sources of saving are the government balance and the external balance.

    Second, the idea of treating “I” as spending for purposes of defining saving seems pretty weird and forced for specific effect. When you define saving that way, it results in identically zero saving for a global or closed, balanced budget economies (notwithstanding positive household sub-positions). I mentioned this sort of effect in the post/paper. True, the household sector still saves in the form of net financial assets. I went through these sorts of numbers in my paper/post – e.g. US household savings/ net worth would be $ 40 trillion or so while global saving would be zero, etc.

    So household saving would end up being a big positive NFA number. And household accumulation of NFA would be offset by issuance of cumulative business NFA (assuming fictitious distribution of undistributed profits), and business investment would be considered to be spending or “dissaving” that cancels out the saving effect of households – with the net of all of that amounting to zero saving by the private sector, except for any additional private sector NFA with other sectors.

    And that means the ONLY type of saving that occurs in the MMT world is net financial saving – which is why they refer to net financial saving as saving. Because the methodology zeroes out the normal definition of saving, by treating investment the same way as consumption for the purpose of defining the residual saving calculation.

    I must say, it’s kind of easy to downplay the effect of I on S, when you treat I as if it were C.

    Such a measurement approach would in theory be convenient to the objective of emphasizing government NFA provision as saving, rather than net saving. Of course, we must be careful not to suggest that objective here.

    I think Mosler and Wray embrace, or at least intuit, this same sort of treatment. I mentioned that in my paper as well.

    So if all that’s true, MMT views “global saving” as zero, while allowing for substantial household positive NFA stock and flow positions within that context, such that business I is accounted for through those positions.

    In effect, the reason they refer to net saving as saving appears to be that they deny the existence of saving at the consolidated global level (or closed, balanced budget national level), while acknowledging it at the household level. That’s where the value of business I is captured as saving, via financial claims (including fictitious dividend distributions.)

    • Finally, let’s put all of the above analysis of the Mitchell/MMT paradigm into the context of:

      S = I + (S – I)

      The S used in this equation is the same S that is used in:

      C + I + G + (X – M) = C + S + T

      And, as noted in the previous comment, that S must account for private sector saving, since the only other sources of saving in that equation are government and external financial balances.

      This private sector content for S must hold regardless of how MMT (e.g. Mitchell) wishes to redefine or reconstruct S as “household saving”, because MMT uses this same equation with the same S to derive its 3-SFB sector financial balances model. It must be the same S, with the same logical connection to the rest of the equation.

      Returning to:

      S = I + (S – I)

      Again, this is the same S from the parent equation and the 3-SFB model.

      And again, the purpose of this equation is to decompose private sector saving into two fundamental components:

      a) I = the amount of saving as a subset of S that is required to fund I – i.e. to match it; not to “cause” it

      b) (S – I) = the amount of saving as a subset of S that constitutes net saving over and above the amount required to fund I

      The MMT inclination as noted above seems to be to treat I in the same way as C, for purposes of defining saving. That is, MMT seems to want to treat I as spending, with the same dissaving effect on income as C.

      That would mean in the MMT world that there is no amount of saving as a subset of S required to fund I – i.e. there is no first component a) in the decomposition of S.

      And that would mean that:

      S = (S – I)

      And that is a contradiction unless I = 0.

      And that is nonsense.

      • “That would mean in the MMT world that there is no amount of saving as a subset of S required to fund I – i.e. there is no first component a) in the decomposition of S.”

        I think MMT finally makes inconsistent statements, but I think for corporations at least they think of retained earnings funding I as being spent and hence “goes away” because it has a dissaving effect.

        • Right, but …

          There’s a definitional and measurement difference between:

          a) undistributed profits (as a subset of S) funding I


          b) (I) as dissaving, eliminating that part of income as a subset of S

          • ” because it has a dissaving effect.”

            Maybe I should have worded it as: because it has a dissaving effect in their language? (not mine).

            • BTW, also interesting to consider stock/flow consistency

              i.e. if investment is treated as dissaving in the same way that consumption is, how does one conceptually replace the net worth/equity that currently offsets investment on the balance sheet under the normal treatment of investment and saving?

              i.e. if private sector saving becomes zero as per simple example, then presumably cumulative saving = net worth/equity of the private sector becomes zero (which presumably means net worth/equity of the business sector becomes negative)

              investment from a stock perspective is “offset” by net zero


              (this reminds me a bit of your SNA accounting on steroids; somewhat similar thing taken to an extreme, due to the treatment of investment as dissaving)

              • I have an answer to that :) (although difficult to write down in words)

                Did you notice “real saving/s” ?

                The MMTers are smart and know how to make a comeback, hence it takes a lot of time to argue with them typically but in the end the comebacks add more inconsistencies to the general story.

              • What about SNA accounting?

                • The net worth calculation that is “net” of backing out equity as a liability?

                  i.e. the SNA net worth calculation for the corporate sector is quite different that corporate financial reporting?

                  maybe that’s wrong; I haven’t gone back to it since our earlier discussion

                  • Yes different. Because the SNA treats equity securities as a liability and the treatment is similar to that of bonds issued. The residual – difference between assets and liabilities is the net worth of the corporation. And hence

                    End of period Net Worth = Beginning of period Net Worth + Undistributed Profits.

                    (assuming away capital gains)

                    But is there any connection of this to Bill’s way of treating investment expenditure as dissaving? (i.e., unknowingly he seems to do this)

                    • I was thinking of (my) implied first step of assuming distribution of undistributed profit, that being reinvested, then treated as negative NFA on business books rather than positive saving; then all that compounded by investment as dissaving

                    • Oh connections like that – though that connection is still vague in my head … But I don’t think MMTers would have gone so far! IMO, their confusion is simply due to jumping steps rather than making a connection to some approach.

    • Госбанк says:

      treating investment the same way as consumption for the purpose of defining the residual saving calculation.

      That was my impression too — that they treat “I” as if it were “C” thus leaving NFA as the only “true” saving inflow. I thought I was missing the point, still not sure if “they” truly think so.

      emphasizing government NFA provision
      One “small” problem with such an emphasis is the inconvenient presence of the (X-M) component as a source of additional NFA inflow/drain that gets treated in a contradictory way (NFA, as a whole, is “good”, but the (X-M) is “bad”, according to some MMT’ers).

    • On the “distribution” issue I would say that they assume away undistributed profits to begin with (all are distributed) and generalize this for the case when they are not all distributed.

      So I would rather think of them as starting with this basic model and generalizing rather than saying they do some neat trick.

      Any link other than the usage of “distribution” where they seem to use this neat trick?

      (Somehow there is some connection here with foreign direct investment treatment of retained earning in the BoP context but I can’t make the connection).

      I think the basic mental model they have is one in which households’ investment is zero but firms do the investment. When they do so they dissave because investment is counted as such whether they realize it or not.

      So this as you put it

      “The next thing he does (I think) is define investment as having the same effect on saving as does consumption – i.e. both are treated as withdrawals from income for the purpose of defining residual saving. Of course, that’s not at all how saving is defined normally – certainly not how Harless defines it, or most other economists. And of course it’s definitely out of synch with corporate financial reporting, which doesn’t help.”


      “And all investment is treated as spending. Such business spending then becomes dissaving from a reconstructed base of zero retained profit. And so such spending on I therefore generates negative business saving in the amount (I).”

      So this is where their inconsistency shows up because he uses Kalecki’s equation which has firms’ retained earnings.

      My own interpretation is that they generalize a simple model with no household investment and firms do all the investment and distribute all profits. The investment is counted differently than standard accounting. Then they have saving for households and dissaving for corporates.

      So while incorporating retained earning, they may say it is “spent” rather than having a “neat trick”, though I will try to see from this perspective.

      Essentially your comment makes a lot of sense though I had never looked at their writing as having used a trick.

      • “My own interpretation is that they generalize a simple model with no household investment and firms do all the investment and distribute all profits. The investment is counted differently than standard accounting. Then they have saving for households and dissaving for corporates.”

        Right. That simplifies and summarizes it.

        I didn’t really mean “trick” in the pejorative sense. Let’s say neat simplification. But it’s one that’s pretty fundamental to their whole business/household decomposition approach. Makes it much easier to describe it the way they do, since it compresses the saving part of the private sector into the household sector. It’s a real facilitator in that sense.

  50. Hi Everyone. I thought this comment was pertinent. From my comments at Winterspeak:

    “The banking system is self sustaining assuming a few unreliable trends (increasing asset values and increasing lending). I’ve seen some MMTers say that loans get “destroyed”, but this is not true in the aggregate or long-term sense. It is only true in a micro sense.

    The government plays a supporting role in all of this. I think MMRists like the word “facilitator”. The government supplies reserves IF needed and support to the payment system IF needed. But as I mentioned above, MMT gets the causation backwards here and makes false conclusions about monopolist this and monopolist that (see the WS post). Loans create deposits and the government oversees smooth payment IF needed. Not vice versa. There is no monopolist in this story. There is no “taxes drive money” in this story as Mason rightly mentions.

    I believe these are some of the points MMR is trying to say to MMT, but are meeting resistance against.”

  51. Ramanan and JKH – I hope you’ll entertain my thinking for a moment. Doesn’t a lot of this confusion come down to a misunderstanding of banking? MMT appears to think that a loan requires someone else to dissave. But loans create deposits out of thin air. This means savings increases. Net financial assets don’t increase, but lending allows someone to spend without dissaving. Loans in our system are always increasing. So someone is always able to save and spend without prior saving.

    Thoughts? Could this be the inconsistency here? The use of the term net financial assets appears confused to me.

    • I like the term NFA but I do not think it’s captures much of the important interactions in the private sector. I think this is one “problem” with MMT, as much as there is a problem.

      MMT gets more right than nearly anyone and I am hugely appreciative of the insights of MMT.

    • Lars,

      Won’t put it that way exactly but MMT’s definitions around saving and profits are so messed that they find it difficult to see that it is theoretically possible for a closed economy to exist without the government. Banks can then just settle at the central bank. Not that I advocate it but purely from an accounting perspective, firms can make profits and distribute dividends and retain some earnings and households can save and banks can be profitable too.

      Using the saving argument – cannot save etc, they try to make such an argument for the government but it doesn’t work because the argument is wrong.

      Of course, personal opinion – we need more public sector, and worldwide fiscal expansion

    • Lending and deposit creation is a flow of funds (i.e. an asset/liability flow), directly separate from income and saving flows.

      Lending/deposit creation is potential fuel for spending.

      Spending from (lending created) deposits can be dissaving at origination, if all income for the period has already been spent otherwise.

      So you can spend without prior saving via new loans/deposits.

      Spending from loans/deposits can create subsequent downstream income and saving, depending on multiplier effects etc.

      And horizontal activity with such banking system expansion is a catalyst for increasing aggregate demand, without government provided NFA. But such NFA directly increases aggregate demand beyond that.

    • “MMT appears to think that a loan requires someone else to dissave.”

      Not at all! You seem to have MMT confused with Austrians or someone who believes that banks are “lending the money of savers” Sounds like something Scott Sumner might say or Nick Rowe.

      BTW Mike Cullen and Beowulf, when the hell are we gonna get back to making fools out of monetarists? Mike you, especially, used to devote considerable blog space to pointing out the idiocies of monetarism. Sure wish we could get back to fighting the real enemies, so to speak.

      • lol – I know. I’ve been slacking!

        More seriously, I’ve been reaaallly busy and finding the even smallest amount of time for MMR has been difficult.

        Plus, I do think the end result of MMR is going to recognize how monetarism works. It’s a horrible system – I’ve called it a rube goldberg machine many times – but it does “work” and it does have an impact on the economy.

        (Remarkably enough, it’s comes back to the private sector savings discussion we’ve been having in S = I + (S-I), right? Because monetarism works (primarily) through increasing real estate lending, but the only way this can really stimulate the economy is if the lending somehow monetizes real estate equity. )

        There are times when we would like more private sector bank lending. It’s not something I’d advocate for right now, but there are many times when more lending is a good idea. I also think banishing the private sector out of lending entirely is a bad idea just from a balance perspective. Most private sector activity isn’t fraudulent lending, it is normal, honest commerce.

        Just laying here thinking about it, my next articles on monetarism would be on the idea monetary policy has less distortionary impact on the economy than fiscal spending. It’s preposterous to believe monetary policy is less distortionary than fiscal.

        I am guessing there is some assumption buried in the math I’ll have to shred, and it’s probably a core part of the IS/LM too and it’s probably been done before but I don’t know it.

        • Cullen Roche says:

          You see Neil’s latest at WS:

          “MMR is a neo-classical synthesis theory in the same vein as monetarism. It leaves full employment to the fairies.”


          • He’s close to crazy. :)

            • Cullen Roche says:

              I just find it interesting how they’re so extreme left that anything in the world just slightly to the right is WAY RIGHT. My mom was a lobbyist for a bunch of Democratic congressman back in the 70’s and 80’s. She warned me the other day – the only thing more unreasonable than an extremist republican is an extremist democrat….

              • Yeah, the difference between our positions and theirs is so slim, you’d think they’d be happy to just have someone to talk too who was actually slightly different than them.

                So Greg, I can’t smack down the monetarists – apparently I am one. ;)

                • “MMR is a neo-classical synthesis theory in the same vein as monetarism.”

                  As physicists would say, that’s not even wrong.
                  Monetarists criticized neoclassical synthesis (i.e. Paul Samuelson’s Old Keynesianism)

                • Did you see my comment at Winterspeaks’ Mike?

                  I believe that you guys might be cedeing too much to the monetarists by focusing on private banking in the manner that you seem to have. Bankers and banking is essentially a monetarist run institution in my view. I dont think Ive ever heard a CBer or a CEO of a big bank use anything other than monetarist language. Its one big monetarist club (even though their leading lights in academia admit to not understanding banking well…… ahem Mr Sumner….)
                  I think you guys have quite a task on your hand reforming banking in a non monetarist way. These guys will make every reform satisfactory to monetarists…. since they “are one”. Quite a needle to thread, I hope you can do it.

                  Thus, if the reform is satisfactory to monetarists it becomes a monetarist reform, which as Ive understood MMR/MMT, is 180 degrees almost.

                  Ill be following with great interest

          • I posted this in response:
            I like to think abou this study every now and then.

            “Some literally stepped over the victim on their way to the next building!”