MMT founders Warren Mosler and Randall Wray like to claim that ‘monetary policy has it backwards’, see these 1,5 min video quotes:
To offer some explanation of why I do not subscribe to W. Mosler‘s claim, I‘ll sketch another, imho more coherent view of how interest rate policy ‚transmits‘ to the price level of consumer goods that‘s pretty ‚Keynesian‘ to me (and not Moslerian at all), but also goes beyond Keynes in one important aspect by identifying a core implicit general assumption Keynes made, and specifying the situations where this assumption actually applies – and those situations where it does NOT apply.
This overall view is based on my reading/interpretation of Wolfgang Stützel‘s work mainly from the 1950s-1980s, which like MMT is also based on an ex post sectoral accounting framework (i.e. an accounting/balance sheet based circuit model of a closed economy). It imports Keynes‘ core GT ideas on employment, interest and money into the sectoral accounting framework, but specifies them and clarifies them by added conceptual precision, imho.
Unfortunately, no english translation of this particular – core – part of Stützel‘s work exists yet. We did very shortly present it at our 2017 seminar at BI Norwegian Business School, which had focused more on our overall legal institutionalist framework and its connection to macro accounting. The seminar talks & presentations are linked at the end of this text.
Here, I‘ll try to be as brief as possible, hoping it will still be accessible.
Reading texts by MMTers, specifically Wray, had helped to prepare me for studying Stützels often difficult work, but as I did, I found that in a number of significant points, he achieved what I consider significant additional clarifications beyond MMT (and other post keynesian monetary circuit models).
I will presuppose the standard MMT ex post sectoral accounting framework where all financial assets (if accounted for at face value!) always net to zero with their corresponding financial liabilities, and where there is a clear awareness of how different transactions lead to different types of quadruple entry balance sheet changes.
So: a raise in central bank interest rates would raise yields/period on financial assets and costs on financial liabilities. But this would also immediately reduce the current price of those financial assets the central bank offers to buy & monetize, leading to direct reductions in net worth of the holders of those assets. If financial institutions‘ net worth in relation to their financial assets is very low, a drastic raise in interest rates can directly lead to insolvencies of financial firms, even to the point of chain insolvencies of banks and bank runs like we saw in 1931. Thus, an important post 2008 reform effort was to raise capital requirements (net worth to total assets ratio) for financial institutions. I‘d recommend Perry Mehrling‘s money & banking course on that topic.
But in this short blog entry, I don‘t want to focus on what interest rate policy means for the financial sector, which Perry Mehrling exclusively focuses on, but what it means for the nonfinancial sector (nonfinancial firms, i.e. firms producing nonfinancial products/assets & services) from one particular perspective, focussing on one particular ‚transmission‘ from the financial to the ‚real‘ sphere and the pricing decisions of nonfinancial firms and thus, the ‚price level‘ for the nonfinancial assets & services they produce. Perry Mehrling does explicitly NOT cover that – imho because with his conceptual apparatus, he fails to conceptualize clearly enough how financial and nonfinancial spheres interact.
Let‘s presuppose firms want to maximize their yields and minimize their costs/period – they want to maximize their profit.
If on average, expected net yields on financial assets are > expected yields on real assets (durable investment goods), firms will plan to increase their stock of net financial assets (financial assets held promise yields, financial debt held incurs costs), i.e. restructure their portfolio by shifting from real to net financial assets (net worth = real assets + net financial assets, therefore ΔNW = ΔRA + ΔNFA).
Net financial assets (sometimes more aptly called ‚net financial position‘ since it is a balancing item that can take on both positive and negative values) is the crucial link between financial and ‚real‘ spheres. An increase in net financial assets can be accomplished only by achieving a current account surplus, which is roughly equal to a trade surplus. To be able to achieve that, subjects need to plan to sell more real assets (in terms of monetary value) than they buy. Sales imply revenues, i.e. increases of net financial assets, purchases imply expenditures or ‚spending‘, i.e. decreases of net financial assets. So to achieve a trade surplus, subjects need to plan to achieve expenditures > revenues, or an increase in net financial assets.
In german standard business accounting (which Stützel used), the terms ‚revenues‘ and ‚expenditures‘ are defined as ‚increases of net financial assets‘ and ‚decreases of net financial assets‘):
Ex post, aggregate sales (revenues) and aggregate purchases (expenditures) will of course be equal by accounting identity. On the ex ante level of expectation based future plans, however, there is no such accounting identity. Rather, aggregate planned sales (revenues) and aggregate planned purchases (expenditures) can deviate from each other (and thus, be incongruent) for the simple reason that in a market economy, independent ‚free‘ legal subjects plan independently from one another and do not know everyone else‘s plan. Plan coordination happens ex post on the market when subjects try to realize their plans; and incongruent plans will lead to market tensions.
Aggregate ex ante planned sales of real assets is synonymous with aggregate revenues and aggregate ‚supply‘ of real assets; aggregate ex ante planned purchases of real assets is synonymous with aggregate planned expenditures (‚spending‘) and aggregate ‚demand‘ for real assets.
As far as my MMT understanding goes, in MMT, the concepts of supply and demand for real assets are not yet coherently related back to the sectoral accounting framework in this way. Sometimes one gets the impression that MMTers see ‚spending‘ as such as synomous with ‚demand‘ for real assets, not explicitly distinguishing ex post actual spending from ex ante planned spending. The former can be traced back to balance sheet entries as they relate to transactions that have already happened. The latter cannot be traced to balance sheet entries, as it relates to transactions that have not have happened yet.
This distinction is crucial, however. Aggregate demand cannot be equal to ex post aggregate spending: in that case, aggregate supply would have to be equal to aggregate revenues, and we know that aggregate ex post expenditures are always equal to aggregate ex post revenues by accounting identity. So if agg. ex post expenditures would be equal aggr. aggr. demand could never deviate from aggr. supply – they would always be equal by definition.
So if the CB raises the interest rate above the average rate of expected yields on nonfinancial assets, this will induce firms to plan trade surpluses for the following period. If agg. planned trade surpluses > planned aggregate trade deficits, this amounts to aggregate supply for real assets being > aggregate demand.
The same plan constellation of agg. planned revenues (supply) being greater than agg. planned expenditures (demand) can also be expressed as aggregate planned change in net financial position being > 0, or as the familiar Keynesian
S(desired) > I(desired)
This simply results from S = ΔNet Worth (NW) and I=ΔReal Assets (RA):
since we know that
NW = RA + FA – FL
Net Worth equals Real Assets plus Financial Assets minus Financial Liabilities, and
FA – FL = NFA
Financial Assets minus Financial Liabilities equals Net Financial Assets so that
NW = RA + NFA
and regarding flows,
ΔNW = ΔRA + ΔNFA
S (ΔNW) can deviate from I (ΔRA) if ΔNFA deviates from zero.
For a closed economy as a whole, therefore, ex post S will always equal I by accounting identity. Ex ante, however (in terms of ‘desired’ or ‘planned’ S and I), S and I can deviate from one another.
For any open economy (individual firm or sector), S and I can also deviate from each other ex post. Sex post will deviate from Iex post by ex post net exports (X-M)ex post, as S = I + (X-M). Exports X are sales of real assets to rest of world, imports M are purchases of real assets from rest of world. The ex post accounting identities are Xsector = Mrest of world (any sector’s exports/sakes equal rest of world’s imports/purchases) and vice versa, so that ex post,
(X-M)sector = -(X-M)rest of world
(X-M)sector + (X-M)rest of world = 0
(S-I)sector = -(S-I)rest of world
(S-I)sector + (S-I)rest of world = 0
Which in turn are just 2 more ways to express the standard ex post accounting identity regarding net financial assets (NFA):
ΔNFAsector = -ΔNFArest of world
ΔNFAsector + ΔNFArest of world = 0
Of course, we can subdivide into any other number of sectors; I’m sticking with 2 sectors for simplicity, here. The basis of this simple use of standard ex post accounting terms is well established in MMT, albeit not always as explicitly integrated with income accounting categories as described above, and not always clearly distinguishing ex post from ex ante variables (which is crucial, see below). Compare for example the section on ‘macroeconomic accounting’ in Tymoigne/Wray’s 2013 ‘The Rise of Money Manager Capitalism’, p. 20-21.
But ex ante, in the planning stage, the above aggregate balances will equal zero only in a very unlikely special case. Usually, they will deviate from zero.
Xplanned are planned sales to /revenues from, Mplanned are planned purchases from/expenditures to rest of world.
So if on the aggregate level,
Xpl > Mpl
aggregate planned sales revenues are > agg. planned expenditures on purchases, then
agg. supply > agg. demand
This implies that
(X-M)pl > o
the closed economy (world as a whole) plans a a trade surplus (which it by definition cannot achieve as for a closed economy, there is no external trading partner by definition, so its balance of trade can only be zero at all times), and
Spl > Ipl
the world s a whole plans to save more than to invest (the classic Keynesian version of the paradox of thrift), which is also impossible since ex post, S will be equal to I by accounting identity; and
ΔNFApl > 0
the world as a whole plans to increase its net financial assets, which is also impossible since ex post, the aggregate economy’s change in NFA will be zero by definition.
What does that mean for the outcomes of such incongruent, not fully achievable plans?
If aggregate plans to increase net financial assets are > 0, this will mean ex post some plans must fail as ex post, the change in net financial assets of any closed economy always equals zero, by definition (by accounting identity). If aggregate planned sales are > aggregate planned purchases, either some planned sales will not have been realized ex post, or aggregate planned purchases will ex post be greater than was planned ex ante.
That crucial (and fairly trivial) relevance of the ex post macro accounting identities of the sectoral accounting framework is – as far as I‘m aware, and correct me if I m wrong – not yet explicitly described and used this way in MMT discourse. It was, however, a standard insight in any german national accounting textbook of the 1960s, such as Alfred Stobbe‘s ‚Volkswirtschaftliches Rechnungswesen‘ (1966, p. 130ff).
To explore some consequences, let‘s first use the standard implicit Keynesian presupposition of ubiquitous buyer‘s market tensions (where buyers – the ‚demand side‘ – have power over sellers – the ‚supply side‘), as that will be an uncontroversial assumption for people considering themselves Keynesians. New for Keynesians will only be Stützel’s insight that this is an unexplained general assumption generalized mostly from the great depression period, and, although in fact a (very frequent) special case, not a ‚natural fact‘ and ‚general‘ case, and therefore requires an explanation based on a careful collection of historical counterexamples (which I will only cover shortly in this short sketch).
In that particular situation, demand will win over supply: planned purchases (spending) determines ex post realized sales (revenues), so that the result will be that some planned sales will ex post not have happened as was planned. There will be unplanned/unexpected failures to sell products.
Total planned sales are equal to total quantity multiplied by the price of real assets. So either, not all real assets that were planned to be sold ex post could be sold: some firms, at the end of the period, still have some products left in their stock which they originally had planned to sell. Or, the other possibility: all real assets were actually sold as planned – but some of them could only be sold at a price lower than the price that was planned ex ante.
In that latter case, the price level will have dropped as a result of the raised interest rate, not risen.
Now if in the next period, the CB will keep interest rates – expected yields on fin. assets – above the average expected net yields on real assets, and firms continue to maximize their profit, plans will remain at the same constellation. But those who failed to achieve their plans will probably try to achieve their plans next period by reducing their expenditures, as the expenditures/demand side wins on buyer‘s markets. This will lead to a drop in total (ex post) expenditures, with again the same result of some failed plans as above etc. – the deflationary spiral that should be familiar to any Keynesian from the 1930s evidence that led to Keynes‘ 1936 General Theory in the first place, I‘d say, as a result of a raise in interest rates.
If in that situation, governments plan additional expenditures / purchases of nonfin. assets (of bombs or whatever – as we know, in the later 1930s, in increasingly became bombs, tanks etc.), it can change aggregate plans so they will either be congruent (‚equilibrium‘) or even aggregate demand > aggr. supply. In that case, the above process reverses, and as long as firms still expect & experience buyer‘s market tensions, total expenditure/production will rise, not necessarily the price level.
But the plan constellation could also be changed if CB‘s lower interest rates significantly below the average expected net yields on real assets. But if average expected net yields on real assets are very low already, there is little room to manouver with interest rate policy, and fiscal demand stimulus policy – planning additional gov‘t expenditures – must help out (which is what Central Bankers had been calling for since 2019, at least).
Now, imho this is all pretty much compatible not only with standard Keynesian theory, but congruent with how Central Banks actually have learned to conduct monetary policy in practice and are doing monetary policy today– and I don‘t see how to square that with Mosler‘s imho confused claim + added ad hoc ‚theorizing‘ that
‚Monetary policy has it backwards – raising interest rates is actually expansionary, lowering them is actually contractionary‘.
Based on the view sketched above (and further below), it seems to me that quite the contrary could be the case: it seems to be Mosler who ‚has it backwards‘ even by Keynesian standards (let alone by standards of actual central banking practice).
I find this a little surprising for someone claiming to be an expert on ‘money’ and monetary policy, and someone claiming his views to be compatible with Keynes’ thinking.
To see that more clearly, let’s take Stützel’s view one step further: once buyer‘s market tensions are dismanteled and seller’s market tensions develop, which will inevitably happen if average expected net yields on real assets consistently remains above expected average yields on net financial assets and therefore, nonfin. firms consistently plan to achieve their profits by trade deficits (by decreasing net financial assets): then, power will move over to the seller‘s side (seller‘s markets) and the choice to raise prices will be preferred by sellers, which eventually will lead to stagflation.
Buyer‘s markets also put pressure on producers regarding product quality, customer service, efficient organization of work and efficient use of resources. Sellers have to try to persuade and ‚bribe‘ buyers into buying their stuff (we all know the kinds of presents sellers sometimes offer in that regard). Seller‘s markets remove that pressure from sellers put it on buyers – customers. They now have to run after sellers, bribe them to get the products they want, and stand in line in the shops (a well known phenomenon in the 1970s & 80s in real socialism, analyzed by Janosz Kornai in his 1980 ‘Economics of Shortage’, and even re-emergent today in some countries, notably in Venezuela in an extreme form). Product quality declines, delivery times get longer. In the 70s, mild forms of that could be witnessed during the stagflation era in capitalist economies as well (I very well remember the bad customer service at the time, compared to today).
As Stützel noticed in 1953, this is not clearly conceptualized by Keynesians yet because to them, buyer‘s market tensions are something ‚natural‘ and ‚ubiquitous‘, whereas seller‘s market tensions exist as little for them as for their Walrasian foes.
So, in the following 4-way matrix, the two Keynesian ‚buyer‘s market tension‘ cases can be found in the top row, the two ‚anti-keynesian‘ ‚seller‘s market tension‘ ones in the bottom row. The historical illustrations were not given by Stützel. I have added them in an attempt to integrate & clarify Stephan Schulmeister’s Keynesian model of ‘long business cycles’ (summarized here) by viewing it through the lens of Stützel’s scheme.
As Stützel put it,
‘Pure General Equilibrium Theory cannot give a solution for any of these 4 cases of plan divergences. Keynesian Theory only knows the two buyer’s market cases’. (W. Stützel 1953: Paradoxa der Geld- und Konkurrenzwirtschaft, Aalen 1979, p. 188)
‘If economic subjects in a closed economy are planning to decrease aggregate net financial assets, this does not self-evidently – and as claimed by most theories of business cycles – lead to an increase in ex post total expenditures/revenues. This only happens under the special conditions of prevailing buyer’s market tensions. If seller’s market tensions prevail in the closed economy, the same plan divergence (planned expenditures/sales/purchases or agg. demand > planned revenues/purchases or aggr. supply which implies planned increases of aggregate net financial assets) leads to the opposite result: ex post, total expenditures/revenues will decrease. ‘ (R.D. Grass/W. Stützel 1988: Volkswirtschaftslehre – eine Einführung, München: Vahlen, p. 341).
One actually fairly trivial, but nevertheless clarifying takeaway insight from this is, in my view: Keynesian policy can only work after buyer‘s market tensions were created, first; and this can happen only by aggregate supply (planned revenues) being > aggregate demand (planned expenditures), which will be a result of expected yields on financial assets > expected net yields on real assets. Constant application of demand stimulus policy, of trying to keep aggregate demand > aggregate supply by way of fiscal or monetary policy, will remove the preconditions for that very policy to lead to the desired results, and eventually produce the opposite result: instead of growing output and employment, declining output and employment stagflation, with rising prices.
Which leads to the quite trivial insight that what’s needed is anti-cyclical monetary and fiscal policy that aims at regulating market tensions, i.e. the level of pressure/challenge or ‘discipline’ sellers are faced with in the economy. Which monetary policy already practices, with fiscal policy currently experiencing a full swing return to the (pre-Stützel) Keynesian perspectives of the 1950s/60s.
Also note that buyer‘s market tensions are good for buyers, bad for sellers. But firms and individuals always a seller AND a buyer, if only a seller of labour.
I‘d say that this provides a more balanced, integrative perspective beyond the typical quarrel between Keynesians and Anti-Keynesians. That was Stützel‘s explicit goal – much like Keynes, he didn‘t go for a ‚heterodox alternative‘ but for a ‚General Theory‘, and he pursued that goal methodically and had methodical tools he used for that (today‘s ‚heterodox‘ and ‚orthodox‘ economists seem to lack those tools and stuck in their unproductive fighting an in-grouping).
The (trivial) ex post accounting foundations of this view are the same as those of MMT, but Stützel makes some additional explicit precise distinctions MMT does not (yet?) make: distinctions between
- ‘spot’ prices relating to a point in time, like sales prices, and ‘holding’ prices relating to a period of time, like expected net yields on financial and real assets
- ex post accounting identities and ex ante planning deviations from those identities, resulting in the development of
- buyer’s market tensions and seller’s market tensions
This not only coherently relates the concepts of aggregate demand & supply back to the balance sheet bsed sectoral accounting framework, which MMT does not yet do as far as I can tell. It also integrates Keynes’ thoughts on the role of the interest rate in relation to expected net yields on real assets (investment goods), while Mosler’s claim seems to contradict Keynes here. The model is only very abstract and would need to be specified by including the private household sector, assumptions about their typical portfolio choices, a closer look at the components of net yields on nonfin. investment goods (including wages, for ex.) etc., but the basic patterns of this view have hopefully become clear in this short description.
Stützel also uses a model building strategy of specifying areas of applicability of generalizations, which can demonstrate seemingly incompatible generalizations are actually compatible if their area of applicability is specified; in addition, Stützel looks for implicitly assumed generalizations (like the Keynesian implicit generalization of ubiquitous permanent buyer’s market tensions, thus ‘demand dominance’) to then demonstrate to which special cases they apply and to which special cases they do not apply.
So for example, while Keynesians criticize Say‘s Theorem (‚any supply will always find demand‘) and simply replace it with a reverse Say‘s theorem (‚any demand will always find supply‘), generalizing from the historically specific experience of the great depression mainly, this model specifies the situations where each of the theorems applies (Keynes: buyer’s markets, Say: seller’s markets).
The debate over‚ which of them is true and which is false‘ becomes unneccessary and can be recognized as blocking integrative progress/learning, (like many other ‘true vs. false‘ debates over economists hasty overgeneralizations), because it avoids asking the question, which statement applies to which specific set of situations‘). Within the framework shortly sketched above, both are true AND false – true for the situations where they apply, false for those where they do not apply. By specifying their respective areas of applicability, we see more clearly; by fighting for binary true/false beliefs, we block and avoid seeing more clearly.
I‘ll close here, not very confident in that anyone of the few who may actually have read this up to this point will make any sense of this. I‘ve tried this many times and the reactions were mostly, people at best used what fit their preconceived beliefs and ignored the rest. Only in one case, someone asked, ‘I’d really like to know how these seller’s market cases work’ but then didn’t pursue the question further, but instead of inquiring how these can be produced went back to presupposing they ‘usually don’t exist’ – where Venezuela would have been a current example with an inflation rate way above the interest rate, where people of course plan to purchase more nonfinancial assets than they plan to sell, and everyone plans to reduce their net fin. assets, producing seller’s markets (empty shelves in supermarkets etc.).
I’m not a professional economist, so I don’t have the time and resources to engage in constant debate with professionals, or translate Stützel’s entire work. I would have no chance whatsoever anyway against their skilled rhetorical manouvers. However, I do take the freedom as a citizen to express doubt to economists’ claims if I see good reasons for that, whether they come from orthodox or heterodox economist. I also take the freedom to read economists whom I feel made significant contributions to some progress in understanding.
It seems to me that Stützel‘s work was most often used selectively by Keynesians and Anti-Keynesians, who just picked out what fit their theories anyway but ignored the rest, and chose to perpetuate that quarrel between Keynesians and Antikeynesians, and between Orthodoxy and Heterodoxy Stützel wanted to solve rationally by integrative model building a la Keynes’ ‘General Theory’. He was very frustrated with how economics developed during the 70s and 80s.
I the end, I think, power claims will probably always trump attempts at integrative understanding overall, at best, only few people will want to sidestep those power games the way Stützel did, even though we should expect that of any academic economist. I don’t see any of Keynes’ or Stützel’s caliber on the horizon right now who could foster integrative progress in economics, so the divide between orthodoxy and heterodoxy both Keynes and Stützel wanted to bridge will probably continue to exist, and citizens and politicians will continue to tend to distrust most economists and their diverse promises, while themselves relying on uninformed models often containing fallacies of composition, but incoherently muddle through somehow in practice, changing ideologies ad hoc as needed.
Keynes famously wrote in Chap. 24 of his General Theory:
“… the ideas of economists and political philosophers, both when they are right and when they are wrong, are more powerful than is commonly understood. Indeed the world is ruled by little else. Practical men, who believe themselves to be quite exempt from any intellectual influences, are usually the slaves of some defunct economist. Madmen in authority, who hear voices in the air, are distilling their frenzy from some academic scribbler of a few years back. I am sure that the power of vested interests is vastly exaggerated compared with the gradual encroachment of ideas. Not, indeed, immediately, but after a certain interval; for in the field of economic and political philosophy there are not many who are influenced by new theories after they are twenty-five or thirty years of age, so that the ideas which civil servants and politicians and even agitators apply to current events are not likely to be the newest. But, soon or late, it is ideas, not vested interests, which are dangerous for good or evil.” (full text here)
I used to believe that too. But these days, after observing both political discourse and post 2008 academic discourse, I tend to conclude that Keynes may have overestimated economist’s influence. Much more than their ideas, power relations influence what happens, and pragmatic elite reactions and changes of ad hoc opinions, legitimated by pretty much randomly chosen ideology, in reaction to crises. Sure, Keynes’ belief may have boosted his self-image as well as that of his colleagues (‘it’s us who make history with our ideas’). I tend to think these days that’s probably a little to self-congratulatory and elitist. Maybe economists and their delusions and ideologies don’t matter that much in the long run, and whatever they say, we shouldn’t take their claims all too seriously, but maybe dare and try to think for ourselves as well. Or not, it doesn’t make that much of a difference. Which reminds me of another Stützel quote. On the first pages of his last book, an introduction to economics he completed about 4 years before his death in an attempt to pass on the gist of his work to his beginning students, he summarized what he had learned about ‘economics’ in the then almost 60 years of his life:
“The only thing I think I really know for sure is this: the job of an economist, the job of improving the world, i.e. the job to prepare suggestions how the human economic life can be organized in a more humane way, is a difficult job. Whoever wants to really take on that job has to keep in mind many different aspects. Otherwise, all too easily the best intentions result in disastrous unintended results.” (W. Stützel/R.D. Grass: Volkswirtschaftslehre – eine Einführung. 2. Aufl. München: Vahlen 1988, p. VI).
Indeed, sometimes I think that the zeal with which someone promotes his (however half-baked) economic ideas are often inversely reciprocal to their empirical and pragmatic quality. And I fear that Mosler’s ‘monetary policy has it backwards’ could be a case of that. I could say a lot more but I probably said much too much already for anyone to listen (those who I might think would ‘should’ listen will not listen anyway, so I shouldn’t be talking). It’s definitely time for me to shut up and do something constructive instead of trying to ‘improve the world’ – at closer look, possibly a rather presumptous, vain and maybe even ridiculous goal.
To sum up and for those who actually feel they can’t stay away from thinking more about this, despite all the warnings, and if it will be anyone at all it will probably be few, I’ll link our video presentations of Stützel’s model described above, including a further elaboration of the monetary aspects of that model, and the only english paper written by Stützel that I am aware of, which may help to get a feel for his thinking.
ANEP Seminar at BI Norwegian Business School, including a summary of the above and a second talk by Johannes Schmidt on further aspects of Stützel’s business cycle theory that I did not cover above
W. Stützel 1974: How to Forecast and Explain the Balance on Current Account of Small Countries. This paper also demonstrates Stützels systematic microfoundation of the macro sectoral accounting model and the micro level portfolio choice approach that is also the basis for what I sketched above.