If you have "no place to go," come here!

The Job Guarantee and the MMT Core: Part Twelve, Theory and Fact

letsgetitdone's picture

Cullen Roche continued his extensive and multi-faceted critique of the Job Guarantee policy and the Modern Monetary Theory approach to economics with a piece attempting to distinguish “theory” and “fact.” His piece is based on the common sense idea that there's a distinction between them, and Cullen tries to use it in his argument. There is, but, unfortunately, the common sense notion of the distinction has long been put aside in the philosophy of science, and in most of the sciences a decade or so later, because of its incoherence. So, in using it, Cullen's argument shares this incoherence.

The “Theoretical” and Something Else

“. . . . It is my responsibility to stick to what my message has always been and that is offering readers a clear understanding of the world we live in. . . .

This is an admirable goal, but, in stating it, I hope Cullen realizes that “a clear understanding of the world” doesn't only include understanding the events and sequences of events that have occurred, or are now occurring in the world, but also includes an understanding of the way the world works. Inevitably, understanding how the world works involves understanding how the systems within it work, and understanding the different kinds of causal and non-causal relations and the dynamics that occur within it. Those relations are descriptive in character, and may even be called “factual” if they correspond in some meaningful sense with reality. But they are not “facts” in the sense they are particular events with specific spatio-temporal locations that are described using singular rather than universal statements.

. . . There are parts of MMT that do that and there are parts that are slapped on as nothing more than policy fixes. The latter is a clear theoretical part. And yes, the JG is ENTIRELY theoretical (which could be ENTIRELY right – I don’t know).”

My problem with this characterization is its attempt to distinguish the parts of MMT that offer a clear understanding of the world and other parts that “are slapped on” and are “nothing more than than policy fixes,” on the basis that the first part is not “theoretical,” while the second part is clearly “theoretical.” And the further notion that the JG is “ENTIRELY theoretical.” I say this characterization is “incoherent” because no radical distinction can be made between the “theoretical” and the “non-theoretical” in attempts to offer narratives about the world. I also say that the “theoretical” parts of MMT come down to the parts that Cullen doesn't agree with and the “non-theoretical” “factual” parts come down to the parts of MMT he likes.

That “theories” and “facts” (in the sense of statements of fact) aren't separable, has been widely recognized beginning, at least, with Popper's Logik der Forschung, published in 1934 in German, and published in English as The Logic of Scientific Discovery in 1959. Since then, Popper extended his argument further, in many of his own publications in opposition to logical positivism and logical empiricism, and his views on this were famously adopted by Paul Feyerabend, Thomas Kuhn, and many other philosophers of science in the 1960s, and then became dominant throughout the sciences in the 1970s, and since that time.

Stated briefly Popper's argument is that a singular statement about a particular of fact “uses universal names (or symbols, or ideas).” (See pp. 94-95) These universal names involve theoretical commitments (See p. 95), For example, if you make a statement about a “glass of water” you're talking about “a physical body that exhibits “certain lawlike behavior,” and you can't even understand that statement unless you understand something about how “glasses” filled with “water” behave (See pp. 94-95). In short, “theoretical commitments” and assumptions pervade our descriptions/interpretations of fact, (pp. 423-424) and we can only construct the facts by making such assumptions. MMT descriptions of how the monetary system operates also involve theoretical commitments, interpretations, and assumptions. For these reasons, it is just wrong to say or imply that the parts of MMT describing monetary operations are non-theoretical, while other parts that deal with the JG are “theoretical.”

So, you can't say, I will accept one part of MMT because it's non-theoretical and reject another part because it's “theoretical” either partly or “ENTIRELY.” Instead, you always have to address the specifics of the theory in the various parts of MMT, and try to test and refute theoretical proposition you find questionable. Cullen has tried to do that in connection with the MMT hypotheses that the JG will work to create full employment with price stability. In previous posts, however, I've considered his various arguments, and given reasons why I consider his theories about this issue to be false.. My point here, however, is that both Cullen and the MMT theorists are theorizing when they write. They may agree that certain sets of propositions describe facts, the way the world really is, and so are true. But where they disagree about the descriptive aspects of MMT, both sides are disagreeing about “facts” as they are theoretically interpreted, or about factual things like the way the world works. There is nothing “factual” that is not also “theoretical.”

Cullen's attempt to separate “theory” and statements of “fact,” reminds me of the creationists' attempts to emphasize that “evolution” is just a theory, and not a fact. Evolution, is a theory, and, in fact, it may not be true. But, people have been trying to refute evolution since Darwin advanced the theory in the 19th century. All those attempts have failed, and, in addition, there is much evidence that corroborates the theory and refutes the narrative about creation in the Bible, and many other versions of creationist theory. So, as far as we know, given our best present knowledge, while evolution is a “theory” it is also a “fact,” showing again that the kind of distinction Cullen is trying to make between the “theoretical” and something else, is invalid.

Selling MMT?

Cullen objects to the views of the MMT economists by saying:

“If the MMT economists want to sell MMT and “modern money” as part of a massive government labor program then that’s fine. It’s my opinion that no one in the world needs to understand large government labor programs in order to understand modern money. Instead, I think we should focus on giving the world a better understanding of modern money by focusing on the proven factual pieces of MMT (monetary operations, monopoly supplier function, etc). . . . .”

As in many of his previous posts criticizing MMT's emphasis on the JG as part of its core, Cullen misconstrues and mis-constructs what MMTers are saying and also shows political bias in his formulation of the issues. First, When MMT writers advocate for the Job Guarantee program they are proposing part of an MMT program for recovery and for strengthening the safety net of automatic stabilizers of the economy that they think will work. Selling that prescription is a separate issue. The primary issue first, is whether their prescription will help to achieve full employment and price stability. That has to determined first -- then you worry about “selling it” or “messaging it”.

Second, MMT writers are not proposing to have “a massive Government Labor Program,” in the sense of a massive expansion of the Federal Bureaucracy, they are proposing an addition to the social safety net in the form of millions jobs that will be funded by the Federal Government but that will be performed for community non-profits, States, and local governments. Also, the size of that jobs program will vary radically with the business cycle, whose swings it will contribute to moderating and stabilizing. In calling the JG a massive Government Labor Program, Cullen is just formulating the issue as the opponents of the JG would formulate it for political purposes. He is just giving them the “framing.”

We will not do that, we will deny the frame and we will frame the unemployment buffer stock as unnecessary and inhuman, as an instrument of 1% and Wall Street control and oppression, and as just the same old trickle down that has been making the rich richer and the poor poorer for more than 40 years. I think we'll win the battle between those political frames, as we won a similar battle in the 30s and the 40s of the last century.

Third, no one is saying that everyone has to understand all the details of the JG in gaining a basic understanding of Modern Money. All the MMT writers are saying is that people need to understand is that price stability requires a buffer stock and that our current choices are mainly between an unemployed and an employed buffer stock being paid a living wage with good fringe benefits. Guess which choice the public ends up favoring if they want price stability? Guess which choice they'll favor if they want full employment? Guess which choice they'll favor is they want a return to prosperity?

Fourth, Cullen talks about the “proven facts” of MMT and teaching people about those instead of the JG. But what actually is “proven” in the sense that it is 100% certain? Is there anything about MMT that is 100% proven? Some may say that the sectoral balance model is an accounting identity, so it's “proven.” However, the sectoral balance model isn't important because it's an accounting identity, and that status doesn't mean that it's been “proven” as a model explaining and predicting “facts”.

It's important because its factual interpretation in real world terms shows that the data we gather to test it (Scott Fullwiler, please note) never refute the factual interpretation of this identity. In other words, it's not the sectoral balance accounting identity that counts. Instead it's the sectoral balance financial model, which, along with its empirical interpretations, constitutes an empirical theory about the macro-economy that counts because it's falsifiable. And that MMT theory isn't “proven” in that sense that it is certainly true, though as far as we know from the data, it isn't false, and may well be true.

On the other hand, take the JG proposal. As Cullen rightly says, the part of MMT that says that combining the JG with payroll tax cuts and State Revenue Sharing isn't proven. But it's also not been tested, and so, it too, may be true. However, facts bearing on the Job Guarantee proposal, like the direct job creation efforts here in the 30s, the Jefes program in Argentina, the current experience in Hungary, don't refute that part of MMT theory. And since the JG has never been tested in the larger context of an MMT economic program like Warren Mosler's or Randy Wray's, the MMT theory about its likely effectiveness has also never been refuted. So, the JG idea, as far as we know, is promising, especially compared to the unemployed buffer stock, whose side effects we know very well are extremely harmful to much of the population.

So, why not teach people about the JG? Why are we doing MMT anyway? To teach people only “the facts” everyone who says they are part of MMT agree on?

What will people do with such knowledge? They will only use it anyway to go beyond that knowledge to use what they think are its policy implications in their own work and lives. Will their theories about the effects of MMT policies be better than ours? Will they be better than theories about the JG whose implications have been explored for many years by MMT economists?

Well, that's certainly a possibility. But the likelihood that these theories won't be, is very high. And the likelihood that an MMT recovery program based on payroll tax cuts, state revenue sharing, and other traditional stimulus measures, along with spending targeting productivity, will be more effective in creating full employment and price stability than Warren's, or Randy's, or Stephanie Kelton's, or Pavlina Tcherneva's MMT-based policies including the JG is also low, because such a program won't add anything to the automatic stabilizers. Of course, it may produce better results on the productivity dimension of economic change. But if that's true and we want to directly target productivity increases by including spending for that in an MMT-based recovery program, then I don't see anything in MMT that bars its inclusion.

Operational Aspects of MMT?

Cullen likes to talk about the “operational aspects of MMT” and the notion that it's policy proposals are “secondary”:

”. . . . If we offer policy proposals then that’s fine, but that’s secondary. I would expect the MMT economists to do that and it should be encouraged that they use their expertise in doing so! But the JG is a very clear case of embedding a policy proposal on top of operational aspects of MM(t) . . . .

Hmmm, “. . . embedding a policy proposal on top of operational aspects. . . .” What does that mean exactly? Well MMT provides an account of the nature of our fiat money system, how the Government relates to the banking system and actually creates and destroys money and net financial assets in the private sector, how the banks create money, but no net financial assets, how the Consolidated Government can and does control interest rates, a theory about the origin of money, a theory about the causes of business cycles, an account of Governments that are sovereign in their own currency differ from Governments that lack or have given up that currency, how all this relates to international trade, theories about how international bond markets and ratings agencies relate to national governments, both those who are currency sovereigns, and those who are just currency users. In addition, MMT economists have drawn policy implications from these theories, about how Governments with currency sovereignty can recover from the crash of 2008 and provide for full employment and price stability in the future.

Now my question is, what is the method by which I and others can tell, which parts of this mosaic are “operational” and which are not? Are all policy proposals “not operational”? If so, how about policies that are currently in place, like not allowing the Treasury to run a negative balance at the Federal Reserve? That's a current policy. Is it also not “operational”? If not, then what do we mean by “operational?” If so, then why is a policy currently in effect about “operations”, but a policy which is an alternative to that not about “operations”?

Moving along, perhaps by the “operational” aspects of MMT Cullen means to include only its narrative about monetary operations, the Government's role as a monopoly supplier of currency, and also any monetary policy that flows from this, but mean to exclude any current or future fiscal policies from the primary focus of MMT, and simply view them as secondary. While I recognize that this may be a clear distinction that could be made, I don't think he'll find broad agreement among people in the MMT movement on this sort of a distinction. Especially since most MMTers seem to think that fiscal policies are much more important than monetary policy in determining economic well-being.

In any case, my purpose in the above isn't to put words in Cullen's mouth, It is to point out that the distinction between the “operational aspects” of MMT and its other aspects is not an obvious one or easy to make. And if one can't make it coherently, then his advice about considering the “operational aspects” as primary, and other aspects as “secondary” will be impossible to talk about, much less to follow.

Next, Cullen continues his attempted distinction between the “operational” and “theoretical” or “secondary” aspects of MMT and his attack on the JG policy with this:

In this regard, it really is entirely theoretical and I appreciate and understand that the economists have built it this way. But there’s a difference between reality and theory and educators have a responsibility to very specifically differentiate between the two so as to avoid misleading those they are teaching.”

It's too bad that Cullen seems unable to take his own advice. As I argue above all of MMT, and every other economic narrative/ideology/approach/paradigm/theory, as well is “theoretical,” in the sense that it is permeated with theoretical assumptions and notions. We cannot say that any of these is ever reality. They are never reality. They are linguistic constructions of human beings.

Now, if you are a “realist” in your epistemology, you may want to claim that your linguistic construct corresponds to reality in some sense of that term. And you can certainly take parts of MMT and say some parts of the body of knowledge claims that is MMT are true, or perhaps closer to the truth, or perhaps more likely than other parts of MMT, and everyone teaching MMT does have a responsibility to do that.

The problem here is that there will be varying degrees of agreement even among MMTers about the parts that are closer to the truth and the parts that are further away from it. So, I think the best way to teach MMT is to teach all the parts of it as simply as we can. Express our individual views and the views of others about the strengths and weaknesses of various parts of it and leave it there. That's unbiased teaching. It is not biased teaching to leave out the parts of MMT that one doubts and call what remains, MMT. That is partisan, and keeps one's readers in ignorance.

One final note, if you're not an epistemological realist, then you'll make a slightly different argument from the one I just made, but there is no currently viable epistemology that will support Cullen's argument distinguishing fact from theory. The only kind that will do that is a foundationalist empiricism that few in philosophy or most of the sciences believes in anymore after the withering criticism it received over the period from the 1930s to the 1980s, and continuing to the present.

Descriptive and Prescriptive Components of MMT

In addition to the distinctions he makes between the “theoretical” and “non-theoretical” or “factual” parts of MMT, and also between the “operational” and the “secondary, ”theoretical” parts of MMT, Cullen also has something to say about the descriptive and prescriptive aspects of MMT:

“I had long been under the impression that there were two components to MMT – the descriptive and the prescriptive. Readers had always been perplexed by the name “Modern Monetary Theory” because what I taught them (by removing the JG) did not appear so theoretical at all. MMT has always been described to me as having a descriptive and the prescriptive component, but it is clear that this is not the case. There is only the “theoretical” under the MMT umbrella. And that’s great. I think it’s a fantastic theory (even though I disagree with parts of it). But there’s a big big difference between the theory and the truth. Parts of MMT are entirely factual (the operational aspects, monopoly supplier, etc) and then there are parts that are entirely theoretical (like the JG). I guess the name alone makes that clear enough (though it never occurred to this idiot)….“

In light of the comments I've made above, it should be clear that what the MMT writers are saying and also what Cullen is saying are both “theoretical.” Nor can Cullen “prove” that what he advocates is less “theoretical” than what they advocate. But there's something else about the above quote that glosses over an important distinction.

Cullen starts off by mentioning the descriptive/prescriptive distinction and then denies that MMT has these two components because, he claims: ”there is only the “theoretical” under the MMT umbrella”, as if to say that economic paradigms or approaches can't have both “descriptive” and “prescriptive” aspects because they have a “theoretical” aspect. But what can this mean?

Doesn't Political Theory have a descriptive and a prescriptive aspect? Doesn't policy analysis have both of these two aspects as well as a theoretical aspect? Doesn't Cullen's work have all three aspects. I think that it's very clear from the various posts in this series that his writing does reflect all three, and that he is no different from other MMT writers in that respect.

Cullen goes on to repeat his notions that there's a difference between a theory and the truth, and that parts of MMT (“the operational aspects)” are “entirely factual” while other parts “(like the JG)” are entirely theoretical. Cullen's right that there's a big, big difference between theory and the truth, because many, maybe most of our statements are not true. But he's got the meaning of this difference wrong. It's not that “factual” statements are true, while “theories” are untrue. It's that some descriptive statements are true, and when they are they're factual, and when they're false they're not, even if they purport to be.

Also, as I said earlier, every statement about the world has “theoretical” aspects. So, in some sense every statement about the world is “a theory,” whether or not it's a universal or singular statement. So, parts of MMT may be true and descriptive, and so, factual, but this has nothing to with whether they are “operational” whatever that may mean, and it also has nothing to do with whether Cullen considers some of them (like the JG) as entirely theoretical, whatever that may mean.

“In my writing I like to stick to the facts even though I am guilty of veering from that goal at times. What you want to do with those facts is entirely up to you. It is not my job to force you to use that understanding in a certain way even though I might, at times, interject my opinion on policy. But what we need to be very clear about in the future is that MMT is in fact theoretical in its current form. I won’t spend my time shooting down MMT because I respect those involved in it far too much. But I won’t be spending my time selling it to readers as though it’s fact. It is in fact theoretical and I hope to use the parts of it that are factual to further the public’s education.”

Translation: I'm going to tell my readers about the parts of MMT that I think are true and I'll use those “to further the public's education But I won't spend my time selling those parts I think are false or don't like. Instead, I'll call them “theoretical,” which I think MMT is in its current form.

Well, Cullen that's just fine. I'm sure no one in the MMT group wants you to advocate for policies that you don't think will work or those parts of MMT you don't believe in. Just please don't label your version of MMT as “MMT,” or tell the rest of us that the JG isn't currently part of the core of MMT, because we know that it's currently essential to MMT, and its progressive prescriptive orientation, which following John Kenneth Galbraith might be called Economics for the Public Purpose, and for which a Job Guarantee with a living wage and good fringe benefits is the best policy we know of right now for enabling full employment with price stability.

No votes yet


Submitted by Hugh on

I think that Modern Monetary Theory practitioners opened themselves up to these kinds of debates because they have used both a narrower version of MMT which emphasized sectoral balances and fiat money and a broader one which contained this narrow version of MMT plus a social purpose for how it was to be used.

A common criticism I have made of MMT is that its practitioners tend to explain it poorly. I think that the social purpose of MMT is a case in point. You see it is not that MMT has a social purpose and other monetary theories do not. It is that all monetary theories have social purposes, explicit or implied. This is because they are all about money. Money is not some fundamental constituent of the universe, like mass and energy. It is simply often treated that way. But what it really is is a social contract, and, in this context, it is entirely legitimate to discuss what the social purpose of money is.

Treating money as a "context-free" purposeless entity does really nothing of the sort. It merely means that the theoretical framework in which it is being used remains hidden and unacknowledged. This is particularly useful in modern kleptocracy where the social misuse of money is covered over with an "objective" gloss. The truth though is that money can not be divorced from its social purpose, and that purpose very much revolves around the kind of society we wish to have.

letsgetitdone's picture
Submitted by letsgetitdone on

For the record, I've never viewed MMT from a narrow perspective because I see all theories, scientific or otherwise as involving value commitments. I don't believe in value-free science or value-free narratives. I believe all our formulations are fused fact-value theories, and have thought this since I worked it out in detail about 40 years or so ago, and I agree that all narratives, theories of money included have social purposes.

Here's my argument against the fact-value dichotomy, pretty much the same since the early 70s.

Anyway in my next post or posts, I'm going to talk about my view of the scope of MMT and what MMT needs to do to make its prescriptive side broader and more explicit.