by Philip Pilkington
PP: Let’s move on to another interesting point about neoclassical economic theory. It’s my understanding that the whole system is assumed to be, at its very essence, one of ‘truck and barter’. In that I mean that they tend to ignore the fact that capitalist economies are, in reality, monetary economies and have nothing to do with barter-based economies. Yet, am I correct in saying that the neoclassicals abstract away from this in order to get away from all sorts of things that make their type of analysis more difficult? I’m thinking in particular of where profits come from and the role of credit in the capitalist system of production. Could you talk about this a little?
SK: That’s actually a strange point of the theory. There was no need to assume the non-existence of money in building a model of capitalism – fundamentally, it makes as much sense as assuming the non-existence of wings when building a model of flight. But this was an established aspect of pre-neoclassical economics long before Walras came on the scene.
David Graeber notes that right from Smith, economists have had a ‘creation myth’ that exchange began with barter and money developed subsequently. This also turns up in Jean Baptiste Say, who, rather than Smith, is the direct ancestor of neoclassical economics:
Ironically, though this was done to abstract from money so that capitalism appeared as a social system aimed at fulfilling the consumption needs of the producers, rather than the class system Marx described it as, it has made what neoclassical economists do – attempt to model capitalism – far harder than it need be.
There is certainly a political side to it – with money abstracted from, so are profits and accumulation – but I really think that this notion that capitalism is a barter system also became ingrained because they thought it would be easier to model capitalism without money than with.
In fact, it’s as easy to model capitalism without money as it is to model birds without wings. But they won’t consider abandoning this impossible assumption and simply modelling the economy in nominal, monetary terms.
PP: One thing that you raise time and again (excuse the pun) in the book is temporality. Economists seem to be enamoured of static models that don’t take the passage of time into account. Perhaps you could comment on this. Is there some historical reason for this or what?
SK: One word: ‘EQUILIBRIUM’. Three main causes: Adam Smith’s ‘invisible hand’ metaphor; ideology; and the great economic historian Philip Mirowski’s point that neoclassical economics adopted the ‘energetics’ approach to physics that pre-dated the development of thermodynamics. Without the ‘arrow of time’ plus the knowledge that no system can be ‘perfect’ that thermodynamics gives, economists continue believing in things that physicists long ago proved were impossible: perfect efficiency in a system, reversibility of processes and independence of a system’s final state from its initial state (though this depends more on complexity theory than the 2nd law).
They’ve spent a couple of centuries trying to prove ‘the invisible hand’, with proving it meaning that the economy reaches an equilibrium in which everyone is as well off as they can be (the ideology bit). Ignorant of the 2nd law of thermodynamics (“You can’t win; you can’t break even; you can’t leave the game”), they hypothesise economic systems that are perfectly efficient. And with a methodology based still on comparative statics (DSGE models are not dynamic, they’re just another version of comparative statics where two time-separated equilibria are compared in the one model) rather than dynamics, they have remained insulated from the techniques mathematicians, engineers and computer programmers have developed in the last century to analyse non-equilibrium systems.
I also blame the development of ‘mathematics for economists’—economists teaching other economists mathematical methods. That’s all most economists do in terms of mathematical training (along with econometrics, which is economists teaching other economists statistical methods). So they don’t learn what is routinely learned by engineers and physicists who are taught mathematics by mathematicians (even though the teaching of mathematics is slanted to the needs of the dominant service group at the time). So, most of them have no knowledge of differential equations; the insolubility of many mathematical problems etc. They have instead developed their own clumsy modelling methods, which presume you can only model in equilibrium, and dynamic processes involve moving from one equilibrium to another where the time path can be ignored.
In fact most dynamic processes of any interest involve movement from one disequilibrium position through many others to a time-path-dependent disequilibrium position. The equilibria of such systems can be attractors or repellers, but they’re neither the starting point nor the ending point.
The trouble is too that this ignorant way of thinking about dynamics dominates not just the neoclassicals, but most of the rival schools of thought as well – including much of Post Keynesian economics (though there are of course exceptions there!). When I first returned to economics as a mature-age student in my mid-30s, my Post Keynesian colleagues (and then instructors) were making much noise about ‘the traverse’. I couldn’t fathom what they were on about – until I realised that they were saying that neoclassical economics was wrong to ignore dynamics since that was ‘the traverse’ from one equilibrium point to another.
I tried to tell them that this was just as much a fallacy as neoclassical methodology itself: if dynamics were simply the ‘traverse’ from one equilibrium situation to another, then the neoclassical argument that one could ignore ‘the traverse’ and focus simply on comparative statics of the beginning and end equilibria was a legitimate argument. But instead, there’s no ‘traverse’, there’s simply continuous disequilibrium. In that case, comparative statics is not merely too simple because it ignores ‘the traverse’, its wrong – because the economy will never be in any equilibrium situation.
The bizarre thing is that all they have to do is accept some of the results from their own research – like the SMD conditions that show that market demand with heterogeneous agents generates a market demand curve that doesn’t satisfy the necessary condition for equilibrium in a market of a downward-sloping demand curve – and instantly they have to do disequilibrium dynamics to be true to their own research findings. But here ideology – and bad pedagogy – intervenes, and they either distort or obscure results that undermine their own methodology, and persist with a barren, equilibrium approach.
Here I can’t resist taking a swipe at Gregory Mankiw, whose textbooks are among the most dominant in economics today, and whose students recently staged a walkout from his class at Harvard.
One of his students sprung to his defence with this reply.
In it he quoted Mankiw’s ten ‘pearls of wisdom’ that he puts across very early in his book as the insights that economics provides. Three of them are:
PP: Your point about ‘the traverse’ is interesting and I think a good way to lead into a discussion of your own ‘positive’ work, rather than your ‘negative’ critique.
My first encounter with the bones of economic theory was Philip Mirowski’s historical work in his book More Heat than Light (which, not coincidentally given the broad point you’re making, you just cited now). I was instantly struck by the way neoclassical economics saw things in terms of stasis, lack of movement and, above all, equilibrium. I had read some of Norbert Weiner’s work before – particularly The Human Use of Human Beings – and was pretty well-versed in the philosophical arguments of Nietzsche, Hegel and the so-called post-structuralists. I knew that this equilibrium stuff was a complete fantasy. All systems, large and small, are always moving toward a state of entropic decay and dissolution (just as, in the same manner, we’re all moving toward death – depressing as that sometimes may be). They’re not moving toward some fantasised harmony of some sort.
Without getting too deep and philosophical: life, as Freud says somewhere, is one great struggle against death. We’re just marking time. Ditto when we do scientific analysis of any sort. We’re always just catching at the coattails of something always slipping away. ‘Equilibrium’ is just a fantasy we cook up, a harmonised ideal that we project into reality. Physicists and engineers know this. Cosmologists know this. Only theologians deny it. For them there is a heaven somewhere.
Now, for me that always meant that what economic analysis could ever achieve would always and necessarily be very, very approximate and limited. This is because we’re dealing with highly complex systems – and given that we only have a few variables to work with, we’re dealing with these highly complex systems largely in the dark. Yet you’ve spent a great deal of your career trying to model some of these systems without recourse to the priest’s heaven of ‘equilibrium’. Perhaps you could say something about this work – and its nature.
SK: It’s actually profoundly simple once you stop using the “simplifying assumptions” so beloved of neoclassical economists – simplifying assumptions like assuming that the economy is always in equilibrium, that money doesn’t matter, that you can ignore time… In fact, they’re about as “simplifying” as making it “easier” to explain how birds fly by assuming that they don’t have wings.
If you look at the history of economic thought, you’ll regularly find the leading neoclassicals saying that they should for example consider time, but that that would make analysis so much more difficult, so let’s make the “simplifying assumption” that … and on they go from there – into an impasse.
Here are some of my favorite examples of that:
Now of course the tools exist to model complex dynamic systems, from the idea of systems of differential equations and systems engineering software at one end of the spectrum to cellular automata and multi-agent modelling at the other. I’ve stuck with the former “tops down” approach, and started from the perspective that we have to treat capitalism as a strictly monetary, inherently cyclical dynamic system.
The irony is that this decision not to use equilibrium and barter as “simplifying assumptions” has made my work easier, not harder. Centuries ago, mathematicians developed a method for modelling dynamic systems called “differential equations”, and with them you necessarily begin with a system that is not in equilibrium: if you start from equilibrium, then nothing happens! In the last 50 years, engineers have developed a methodology for building such systems graphically using flowcharts, and in the last 20 years, computer programmers have turned that engineering method into live “graphical user interface” programs that simulate these flowcharts on screen.
I’ve simply ported these advances into economics, while the vast majority (though certainly not all) economists don’t even know they exist. The minority who are aware of these methods and programs generally call themselves complexity theorists or something like that – people like Carl Chiarella, Thomas Lux, Peter Flaschel, Willy Semmler – and they’ve done some brilliant work. Technically, I don’t see my work as better than theirs – I don’t think I can hold a candle to Carl Chiarella for example, whose knowledge of mathematics is far stronger than mine. My advantage has come from having a richer knowledge of the history of economic thought, and therefore applying this method to turn the insights of the truly great thinkers – Marx, Schumpeter, Keynes, Fisher, Sraffa and Minsky – into dynamic mathematical models.
My first forays into this built on Richard Goodwin’s “growth cycle” model to introduce debt-financed investment into Goodwin’s simple “predator-prey” model of a class struggle between workers and capitalists (in which, to confuse both Marxists and right-wingers, technically speaking the workers are the predators and the capitalists are the prey!). With this I could model the fact that investment is debt-financed, and the model captured Minsky’s basic idea that capitalists borrow to invest during booms, but have to repay debt during a slump, with the result that debt can “ratchet up” over time, leading to a final crisis in which the debt taken on simply overwhelms the economy and it falls into a Depression.
That model generated an apparent “Great Moderation” followed by a “Great Depression”, which led me to finish my first paper on it with what I then regarded as a rhetorical flourish:
But I was dissatisfied with my own model because it wasn’t explicitly monetary. I finally worked out how to model money creation dynamically after becoming aware of Graziani’s work on “the monetary theory of production”, which made the point that in a monetary economy, all exchanges involve one commodity and three agents: a buyer, a seller, and a bank that effects the sale by transferring funds from the buyer’s account to the sellers. That ultimately led to the breakthrough that I could model money creation directly from double-entry bookkeeping, and now I can develop models of monetary dynamics of any desired level of complexity that way.
I’ve now developed a method of building models of financial dynamics and physical production that is embodied in simulation software. A prototype called QED (which stands for “Quesnay Economic Dynamics” in honor of the person whom I believe should be seen as the father of economics) is already available on my blog, and with the help of an INET grant I’m developing a far more comprehensive modelling tool called “Minsky”.
My vision for Minsky is that it will make it possible for undergraduates to build meaningful models of monetary macroeconomics within minutes, and also scale to the level of enabling researchers to build large scale monetary models incorporating both government and private sector money creation, multiple sectors, multiple countries, and trade and financial flows. I hope it will be the “Yellow Brick Road” to lead new students away from neoclassical economics, because one of the great weaknesses of the heterodox approach is that it hasn’t been able to match the apparent sophistication of the neoclassical paradigm and model: in place of systems of equations describing “general equilibrium”, heterodox economists have only been able to wave objections to ignoring uncertainty, etc. – they haven’t been able to offer a compelling alternative methodology (Wynne Godley’s “Social Accounting Matrix” approach was a step in the right direction, but not enough). This will I believe be a compelling alternative that will expose neoclassical methodology for what it really is – passe.
PP: And what do you make of the critique that economic models can only really give us a small insight into the actual working economy – i.e. that they are just ‘toy models’ that ‘train the brain’? The heterodox highlighting of fundamental uncertainty you highlighted above would be a good example of such a critique. Another, more recent example, would be what Yanis Varoufakis et al have called the ‘Inherent Error’ in all economic models in their new book Modern Political Economy – that’s a critique showing that the lack of any possible coherent theory of value that corresponds with any theory of growth means that models will never reflect the real world. Do you buy these critiques and if so where does it leave those, like yourself, that seek to build models of the economy?
SK: Generally I agree that economic models have to be toy models. We can build a model of the weather that lets us predict it with moderate accuracy for a week or so – a significant achievement given how chaotic fluid dynamics are. But predicting the future course of the economy over its relevant time frequency – which is years rather than weeks – is probably a forlorn ambition.
However, building a mathematical model which incorporates the factors that are qualitatively important in capitalism is not a forlorn ambition. For example, neoclassicals build models in which private debt has no important macroeconomic effect. I build models in which it does. As to which model is qualitatively more accurate, I rest my case.
There’s also a role for mathematical modelling in accurately expressing verbal logic. For example, Graziani and Circuit theorists developed a brilliant verbal model of money, but reached the bizarre conclusion that capitalists couldn’t borrow money, make a profit and repay the debt. They’d actually made a simple stock-flow error, which I was able to correct with a simple dynamic model of banking. Without that maths, Circuit Theorists had spent 20 years floundering on that issue.
I think it’s also possible to provide a quantitative context with a model, when something as extreme as our current debt bubble is considered. Given the astronomical scale of private debt today, I’m quite confident in making an empirical prediction that growth in the Western OECD will be below long term trends for some time, until debt levels are radically reduced.
On the theory of value aspects of the argument Yanis and Joe make, I differ because of my own very different analysis of Marx’s theory of value – which is detailed in my 1993 papers on Marx.
So I think they overstate their case.
PP: Do you believe a coherent theory of value exists? If so, can you lay out the basics?
SK: Boy, you are going into detail here Philip! Do you really have space for this in a Naked Capitalism interview?
Nonetheless, yes I do. It takes a lot longer than an interview to set it out, but the précis is that Marx’s basic theory of value is NOT the labour theory of value, but a dialectic between use-value and exchange value – both of which he defined in objective rather than subjective terms. Whereas neoclassicals argue that there is a relationship between (subjective) utility and (objective) cost of production [the relative price of two commodities being set when the ratio of their marginal utilities equals the ratio of their marginal costs), Marx argued that the two were incommensurable: in a primitive stage of society the utility of a product that one tribe could produce and the other could not might play a role in the price paid for it, but as exchange becomes routine he argued that the relative cost of production becomes the sole determinant of exchange value.
In consumption this simply meant that use-value played no role in setting price. But in production it meant that there could be a gap between quantitative use-value and quantitative exchange-value: the quantitative use-value of an input was its capacity to help produce a surplus, the quantitative exchange-value was its cost of production. The two were different, with the former (normally) exceeding the latter, and this was how Marx explained the source of profit.
He thought he could prove this property was limited to labour alone – so that it was consistent with the labour theory of value – but that was wrong.
I use this as the starting point of my theory of profits, but then use the dialectical method to expand it to consider the setting of wages–where the fact that labour is both a commodity and a non-commodity means that the wage will normally exceed the value of labour-power–and the pricing of capital assets–where subjective expectations of profit set price.
I set this all out in my Masters thesis and an unpublished paper “A Marx for Post Keynesians”, which are both available on my blog.
PP: Finally, what advice would you give to a student thinking of doing economics that has a nagging feeling that the neoclassical strain is complete nonsense, yet at the same time sees that it is the fundamental language in which modern policy is cast?
SK: I’d advise him/her to do an undergraduate degree in systems engineering with a minor in economics. That way you learn the language of economics while simultaneously learning how to think systemically—rather than being lobotimised in the name of education.
Other essential courses to include would be basic maths—calculus, algebra and differential equations—with basic computing, and a course in modern history (since economic history courses have been almost totally abolished). Plus do something to keep the right brain functions alive too —music, poetry, party a lot, maybe lead a student revolt…
PP: Let’s move on to another interesting point about neoclassical economic theory. It’s my understanding that the whole system is assumed to be, at its very essence, one of ‘truck and barter’. In that I mean that they tend to ignore the fact that capitalist economies are, in reality, monetary economies and have nothing to do with barter-based economies. Yet, am I correct in saying that the neoclassicals abstract away from this in order to get away from all sorts of things that make their type of analysis more difficult? I’m thinking in particular of where profits come from and the role of credit in the capitalist system of production. Could you talk about this a little?
SK: That’s actually a strange point of the theory. There was no need to assume the non-existence of money in building a model of capitalism – fundamentally, it makes as much sense as assuming the non-existence of wings when building a model of flight. But this was an established aspect of pre-neoclassical economics long before Walras came on the scene.
David Graeber notes that right from Smith, economists have had a ‘creation myth’ that exchange began with barter and money developed subsequently. This also turns up in Jean Baptiste Say, who, rather than Smith, is the direct ancestor of neoclassical economics:
Every producer asks for money in exchange for his products, only for the purpose of employing that money again immediately in the purchase of another product; for we do not consume money, and it is not sought after in ordinary cases to conceal it: thus, when a producer desires to exchange his product for money, he may be considered as already asking for the merchandise which he proposes to buy with this money. It is thus that the producers, though they have all of them the air of demanding money for their goods, do in reality demand merchandise for their merchandise.Now there are many elements in that which are clearly nonsense, but the proposition that producers want other goods for their goods, and don’t seek to accumulate money instead, became ingrained in neoclassical thinking.
Ironically, though this was done to abstract from money so that capitalism appeared as a social system aimed at fulfilling the consumption needs of the producers, rather than the class system Marx described it as, it has made what neoclassical economists do – attempt to model capitalism – far harder than it need be.
There is certainly a political side to it – with money abstracted from, so are profits and accumulation – but I really think that this notion that capitalism is a barter system also became ingrained because they thought it would be easier to model capitalism without money than with.
In fact, it’s as easy to model capitalism without money as it is to model birds without wings. But they won’t consider abandoning this impossible assumption and simply modelling the economy in nominal, monetary terms.
PP: One thing that you raise time and again (excuse the pun) in the book is temporality. Economists seem to be enamoured of static models that don’t take the passage of time into account. Perhaps you could comment on this. Is there some historical reason for this or what?
SK: One word: ‘EQUILIBRIUM’. Three main causes: Adam Smith’s ‘invisible hand’ metaphor; ideology; and the great economic historian Philip Mirowski’s point that neoclassical economics adopted the ‘energetics’ approach to physics that pre-dated the development of thermodynamics. Without the ‘arrow of time’ plus the knowledge that no system can be ‘perfect’ that thermodynamics gives, economists continue believing in things that physicists long ago proved were impossible: perfect efficiency in a system, reversibility of processes and independence of a system’s final state from its initial state (though this depends more on complexity theory than the 2nd law).
They’ve spent a couple of centuries trying to prove ‘the invisible hand’, with proving it meaning that the economy reaches an equilibrium in which everyone is as well off as they can be (the ideology bit). Ignorant of the 2nd law of thermodynamics (“You can’t win; you can’t break even; you can’t leave the game”), they hypothesise economic systems that are perfectly efficient. And with a methodology based still on comparative statics (DSGE models are not dynamic, they’re just another version of comparative statics where two time-separated equilibria are compared in the one model) rather than dynamics, they have remained insulated from the techniques mathematicians, engineers and computer programmers have developed in the last century to analyse non-equilibrium systems.
I also blame the development of ‘mathematics for economists’—economists teaching other economists mathematical methods. That’s all most economists do in terms of mathematical training (along with econometrics, which is economists teaching other economists statistical methods). So they don’t learn what is routinely learned by engineers and physicists who are taught mathematics by mathematicians (even though the teaching of mathematics is slanted to the needs of the dominant service group at the time). So, most of them have no knowledge of differential equations; the insolubility of many mathematical problems etc. They have instead developed their own clumsy modelling methods, which presume you can only model in equilibrium, and dynamic processes involve moving from one equilibrium to another where the time path can be ignored.
In fact most dynamic processes of any interest involve movement from one disequilibrium position through many others to a time-path-dependent disequilibrium position. The equilibria of such systems can be attractors or repellers, but they’re neither the starting point nor the ending point.
The trouble is too that this ignorant way of thinking about dynamics dominates not just the neoclassicals, but most of the rival schools of thought as well – including much of Post Keynesian economics (though there are of course exceptions there!). When I first returned to economics as a mature-age student in my mid-30s, my Post Keynesian colleagues (and then instructors) were making much noise about ‘the traverse’. I couldn’t fathom what they were on about – until I realised that they were saying that neoclassical economics was wrong to ignore dynamics since that was ‘the traverse’ from one equilibrium point to another.
I tried to tell them that this was just as much a fallacy as neoclassical methodology itself: if dynamics were simply the ‘traverse’ from one equilibrium situation to another, then the neoclassical argument that one could ignore ‘the traverse’ and focus simply on comparative statics of the beginning and end equilibria was a legitimate argument. But instead, there’s no ‘traverse’, there’s simply continuous disequilibrium. In that case, comparative statics is not merely too simple because it ignores ‘the traverse’, its wrong – because the economy will never be in any equilibrium situation.
The bizarre thing is that all they have to do is accept some of the results from their own research – like the SMD conditions that show that market demand with heterogeneous agents generates a market demand curve that doesn’t satisfy the necessary condition for equilibrium in a market of a downward-sloping demand curve – and instantly they have to do disequilibrium dynamics to be true to their own research findings. But here ideology – and bad pedagogy – intervenes, and they either distort or obscure results that undermine their own methodology, and persist with a barren, equilibrium approach.
Here I can’t resist taking a swipe at Gregory Mankiw, whose textbooks are among the most dominant in economics today, and whose students recently staged a walkout from his class at Harvard.
One of his students sprung to his defence with this reply.
In it he quoted Mankiw’s ten ‘pearls of wisdom’ that he puts across very early in his book as the insights that economics provides. Three of them are:
• Demand curves slope downward and supply curves slope upward (usually).The first is wrong theoretically via the SMD conditions and empirically via research into the actual cost structures of real firms (see Fred Lee’s work and also Alan Blinder’s Asking About Prices); the next two support comparative statics and, from a mathematical point of view, are also wrong. The conditions under which they could be true are that there’s only one equilibrium (or there are many but equally desirable – their ‘Pareto optimal’ nonsense), and all equilbria are attractors. Neither of those conditions applies in economics, but they practice a method that assumes they always apply.
• Sometimes, things happen that make demand curves and supply curves shift.
• Comparative statics can be a useful way of thinking about how changes in some variables will affect changes in other variables.
PP: Your point about ‘the traverse’ is interesting and I think a good way to lead into a discussion of your own ‘positive’ work, rather than your ‘negative’ critique.
My first encounter with the bones of economic theory was Philip Mirowski’s historical work in his book More Heat than Light (which, not coincidentally given the broad point you’re making, you just cited now). I was instantly struck by the way neoclassical economics saw things in terms of stasis, lack of movement and, above all, equilibrium. I had read some of Norbert Weiner’s work before – particularly The Human Use of Human Beings – and was pretty well-versed in the philosophical arguments of Nietzsche, Hegel and the so-called post-structuralists. I knew that this equilibrium stuff was a complete fantasy. All systems, large and small, are always moving toward a state of entropic decay and dissolution (just as, in the same manner, we’re all moving toward death – depressing as that sometimes may be). They’re not moving toward some fantasised harmony of some sort.
Without getting too deep and philosophical: life, as Freud says somewhere, is one great struggle against death. We’re just marking time. Ditto when we do scientific analysis of any sort. We’re always just catching at the coattails of something always slipping away. ‘Equilibrium’ is just a fantasy we cook up, a harmonised ideal that we project into reality. Physicists and engineers know this. Cosmologists know this. Only theologians deny it. For them there is a heaven somewhere.
Now, for me that always meant that what economic analysis could ever achieve would always and necessarily be very, very approximate and limited. This is because we’re dealing with highly complex systems – and given that we only have a few variables to work with, we’re dealing with these highly complex systems largely in the dark. Yet you’ve spent a great deal of your career trying to model some of these systems without recourse to the priest’s heaven of ‘equilibrium’. Perhaps you could say something about this work – and its nature.
SK: It’s actually profoundly simple once you stop using the “simplifying assumptions” so beloved of neoclassical economists – simplifying assumptions like assuming that the economy is always in equilibrium, that money doesn’t matter, that you can ignore time… In fact, they’re about as “simplifying” as making it “easier” to explain how birds fly by assuming that they don’t have wings.
If you look at the history of economic thought, you’ll regularly find the leading neoclassicals saying that they should for example consider time, but that that would make analysis so much more difficult, so let’s make the “simplifying assumption” that … and on they go from there – into an impasse.
Here are some of my favorite examples of that:
If we wished to have a complete solution … we should have to treat it as a problem of dynamics. But it would surely be absurd to attempt the more difficult question when the more easy one is yet so imperfectly within our power.(Jevons 1871)So what they did was throw the dynamic approach into the “too hard” basket, hoping that their successors would take it out of there and do it later. But neoclassicals never did this.
…dynamics includes statics… But the statical solution… is simpler…; it may afford useful preparation and training for the more difficult dynamical solution; and it may be the first step towards a provisional and partial solution in problems so complex that a complete dynamical solution is beyond our attainment. (Marshall, 1907)
The idea of erratic shocks represents one very essential aspect of the impulse problem in economic cycle analysis, but probably it does not contain the whole explanation … In mathematical language one could perhaps say that one introduces here the idea of an auto-maintained oscillation… It would be possible to put the functioning of this whole instrument into equations under more or less simplified assumptions about the construction and functioning of the valve, etc. I even think this will be a useful task for a further analysis of economic oscillations, but I do not intend to take up this mathematical formulation here.(Frisch 1933, pp. 33-35).
Now of course the tools exist to model complex dynamic systems, from the idea of systems of differential equations and systems engineering software at one end of the spectrum to cellular automata and multi-agent modelling at the other. I’ve stuck with the former “tops down” approach, and started from the perspective that we have to treat capitalism as a strictly monetary, inherently cyclical dynamic system.
The irony is that this decision not to use equilibrium and barter as “simplifying assumptions” has made my work easier, not harder. Centuries ago, mathematicians developed a method for modelling dynamic systems called “differential equations”, and with them you necessarily begin with a system that is not in equilibrium: if you start from equilibrium, then nothing happens! In the last 50 years, engineers have developed a methodology for building such systems graphically using flowcharts, and in the last 20 years, computer programmers have turned that engineering method into live “graphical user interface” programs that simulate these flowcharts on screen.
I’ve simply ported these advances into economics, while the vast majority (though certainly not all) economists don’t even know they exist. The minority who are aware of these methods and programs generally call themselves complexity theorists or something like that – people like Carl Chiarella, Thomas Lux, Peter Flaschel, Willy Semmler – and they’ve done some brilliant work. Technically, I don’t see my work as better than theirs – I don’t think I can hold a candle to Carl Chiarella for example, whose knowledge of mathematics is far stronger than mine. My advantage has come from having a richer knowledge of the history of economic thought, and therefore applying this method to turn the insights of the truly great thinkers – Marx, Schumpeter, Keynes, Fisher, Sraffa and Minsky – into dynamic mathematical models.
My first forays into this built on Richard Goodwin’s “growth cycle” model to introduce debt-financed investment into Goodwin’s simple “predator-prey” model of a class struggle between workers and capitalists (in which, to confuse both Marxists and right-wingers, technically speaking the workers are the predators and the capitalists are the prey!). With this I could model the fact that investment is debt-financed, and the model captured Minsky’s basic idea that capitalists borrow to invest during booms, but have to repay debt during a slump, with the result that debt can “ratchet up” over time, leading to a final crisis in which the debt taken on simply overwhelms the economy and it falls into a Depression.
That model generated an apparent “Great Moderation” followed by a “Great Depression”, which led me to finish my first paper on it with what I then regarded as a rhetorical flourish:
From the perspective of economic theory and policy, this vision of a capitalist economy with finance requires us to go beyond that habit of mind which Keynes described so well, the excessive reliance on the (stable) recent past as a guide to the future. The chaotic dynamics explored in this paper should warn us against accepting a period of relative tranquility in a capitalist economy as anything other than a lull before the storm. (Keen, 1995, p. 634; emphasis added)Some flourish: in fact that’s what then happened: a period of apparent calm occurred from the 1990s till 2007, and then all hell broke loose. Neoclassicals were lulled into complacency by the apparent tranquillity, and then bang.
But I was dissatisfied with my own model because it wasn’t explicitly monetary. I finally worked out how to model money creation dynamically after becoming aware of Graziani’s work on “the monetary theory of production”, which made the point that in a monetary economy, all exchanges involve one commodity and three agents: a buyer, a seller, and a bank that effects the sale by transferring funds from the buyer’s account to the sellers. That ultimately led to the breakthrough that I could model money creation directly from double-entry bookkeeping, and now I can develop models of monetary dynamics of any desired level of complexity that way.
I’ve now developed a method of building models of financial dynamics and physical production that is embodied in simulation software. A prototype called QED (which stands for “Quesnay Economic Dynamics” in honor of the person whom I believe should be seen as the father of economics) is already available on my blog, and with the help of an INET grant I’m developing a far more comprehensive modelling tool called “Minsky”.
My vision for Minsky is that it will make it possible for undergraduates to build meaningful models of monetary macroeconomics within minutes, and also scale to the level of enabling researchers to build large scale monetary models incorporating both government and private sector money creation, multiple sectors, multiple countries, and trade and financial flows. I hope it will be the “Yellow Brick Road” to lead new students away from neoclassical economics, because one of the great weaknesses of the heterodox approach is that it hasn’t been able to match the apparent sophistication of the neoclassical paradigm and model: in place of systems of equations describing “general equilibrium”, heterodox economists have only been able to wave objections to ignoring uncertainty, etc. – they haven’t been able to offer a compelling alternative methodology (Wynne Godley’s “Social Accounting Matrix” approach was a step in the right direction, but not enough). This will I believe be a compelling alternative that will expose neoclassical methodology for what it really is – passe.
PP: And what do you make of the critique that economic models can only really give us a small insight into the actual working economy – i.e. that they are just ‘toy models’ that ‘train the brain’? The heterodox highlighting of fundamental uncertainty you highlighted above would be a good example of such a critique. Another, more recent example, would be what Yanis Varoufakis et al have called the ‘Inherent Error’ in all economic models in their new book Modern Political Economy – that’s a critique showing that the lack of any possible coherent theory of value that corresponds with any theory of growth means that models will never reflect the real world. Do you buy these critiques and if so where does it leave those, like yourself, that seek to build models of the economy?
SK: Generally I agree that economic models have to be toy models. We can build a model of the weather that lets us predict it with moderate accuracy for a week or so – a significant achievement given how chaotic fluid dynamics are. But predicting the future course of the economy over its relevant time frequency – which is years rather than weeks – is probably a forlorn ambition.
However, building a mathematical model which incorporates the factors that are qualitatively important in capitalism is not a forlorn ambition. For example, neoclassicals build models in which private debt has no important macroeconomic effect. I build models in which it does. As to which model is qualitatively more accurate, I rest my case.
There’s also a role for mathematical modelling in accurately expressing verbal logic. For example, Graziani and Circuit theorists developed a brilliant verbal model of money, but reached the bizarre conclusion that capitalists couldn’t borrow money, make a profit and repay the debt. They’d actually made a simple stock-flow error, which I was able to correct with a simple dynamic model of banking. Without that maths, Circuit Theorists had spent 20 years floundering on that issue.
I think it’s also possible to provide a quantitative context with a model, when something as extreme as our current debt bubble is considered. Given the astronomical scale of private debt today, I’m quite confident in making an empirical prediction that growth in the Western OECD will be below long term trends for some time, until debt levels are radically reduced.
On the theory of value aspects of the argument Yanis and Joe make, I differ because of my own very different analysis of Marx’s theory of value – which is detailed in my 1993 papers on Marx.
So I think they overstate their case.
PP: Do you believe a coherent theory of value exists? If so, can you lay out the basics?
SK: Boy, you are going into detail here Philip! Do you really have space for this in a Naked Capitalism interview?
Nonetheless, yes I do. It takes a lot longer than an interview to set it out, but the précis is that Marx’s basic theory of value is NOT the labour theory of value, but a dialectic between use-value and exchange value – both of which he defined in objective rather than subjective terms. Whereas neoclassicals argue that there is a relationship between (subjective) utility and (objective) cost of production [the relative price of two commodities being set when the ratio of their marginal utilities equals the ratio of their marginal costs), Marx argued that the two were incommensurable: in a primitive stage of society the utility of a product that one tribe could produce and the other could not might play a role in the price paid for it, but as exchange becomes routine he argued that the relative cost of production becomes the sole determinant of exchange value.
In consumption this simply meant that use-value played no role in setting price. But in production it meant that there could be a gap between quantitative use-value and quantitative exchange-value: the quantitative use-value of an input was its capacity to help produce a surplus, the quantitative exchange-value was its cost of production. The two were different, with the former (normally) exceeding the latter, and this was how Marx explained the source of profit.
He thought he could prove this property was limited to labour alone – so that it was consistent with the labour theory of value – but that was wrong.
I use this as the starting point of my theory of profits, but then use the dialectical method to expand it to consider the setting of wages–where the fact that labour is both a commodity and a non-commodity means that the wage will normally exceed the value of labour-power–and the pricing of capital assets–where subjective expectations of profit set price.
I set this all out in my Masters thesis and an unpublished paper “A Marx for Post Keynesians”, which are both available on my blog.
PP: Finally, what advice would you give to a student thinking of doing economics that has a nagging feeling that the neoclassical strain is complete nonsense, yet at the same time sees that it is the fundamental language in which modern policy is cast?
SK: I’d advise him/her to do an undergraduate degree in systems engineering with a minor in economics. That way you learn the language of economics while simultaneously learning how to think systemically—rather than being lobotimised in the name of education.
Other essential courses to include would be basic maths—calculus, algebra and differential equations—with basic computing, and a course in modern history (since economic history courses have been almost totally abolished). Plus do something to keep the right brain functions alive too —music, poetry, party a lot, maybe lead a student revolt…
No comments:
Post a Comment