A recent post by LSE based International Relations scholar, Nick Srnicek, asked ‘Has the Left given up on economics?‘, arguing that in spite of living through one the greatest economic depressions in history the left has notably failed to incorporate economic analysis or alternatives into its program for change. What we have, instead, is the usual focus on racism, police brutality, defending the welfare state, and such like issues – a repetition of the kind of lowest common denominator politics of resistance little different to what the left focused on during the boom time. Of course, the reasons for this state of affairs are just as much political as to do with the collective psychological state of the left. A counter post here makes a few important points on the institutional reasons for the weakness; the most important of which I consider to be the lack of support by union structures for developing and disseminating alternative economic analysis. As a supplement to this, I would also add that the current predominance of (neo) anarchist ideas in activist communities – with the valorisation of sponteneism, horizontal (non) organization, and a focus on means rather than ends (so called prefigurative politics) – serves to reinforce the irrelevance of expert, technical analysis of economics. Ultimately, the end game of this kind of extreme democratic politics cannot tolerate analysis that would privilege the scientific knowledge of a minority. But in any case, the above political argument is not the focus of my present commentary, so I will leave that aside for another day.
Here, I want to present what I believe to be one possible reason for the seemingly minimal influence of economics on left wing politics: the fact that the Marxian economic project has stalled on a number of levels. As a Marxist myself, I feel at liberty to place at least some of the blame for this at Marx’s feet, even if, of course, contemporary Marxists should shoulder most of the weight of responsibility for the state of affairs. In brief, there are two levels at which I think Marxian economics has stalled:
- Failing to work through the categories elaborated in Capital to more adequately conceptualise transitionary structures to post-capitalism.
- Not bringing Capital up to date in terms of economic model theory, hence leaving it as a limited ‘fundamental structural model’ of the economy, unable to incorporate enough variables and economic objects into its analysis.
1. Capital and post-capitalism
There is a persistent tendency within Marxism to reduce the economic to the political, as if none of the economic laws under one system carry across even during a period of transition to another system. This is reflected in the somewhat endless debates regarding to what extent Capital represents simply a critique of capitalism or a work of positive economics? One could certainly make arguments both ways. My personal view on this matter is that by the time on reaches Vol. III it becomes indisputable that Marx had, whether intentionally or not, created a work of positive economics with tendential predictions unique to the Marxian economic paradigm. That is, the exploitation labour theory of value leads to a theory of tendencies within the profit rates not found elsewhere in other economic theories, dependent upon historically variant relative composition of capitals, and also leading to a theory of tendencies in employment levels.
Now, none of this fits very neatly with Marx’s political philosophy, which is all about the real, historical political movement. One of the mysteries during my research into Marx as part of my PhD thesis is why, if Marx felt his own analysis mostly useless for thinking a post-capitalist society, did he seemingly invest so much time in it? And why, at the end of 20 years study, did he abandon his economic studies to learn Russian and focus almost exclusively on events there, even to the point of ditching the thesis that communism would preserve the best of capitalism to endorsing the populist idea that peasant communes could act as a direct launch pad for communism?
My feeling is that Marx never really resolved these issues intellectually. You see the same thing in his mathematical writings, where he never really reconciled himself with abstract mathematical entities, trying to shoehorn the whole of calculus into the study of variable magnitudes. So the legacy of this Marxian reduction of abstract logics to the real is the presumption – common across most of the left in fact – that a revolutionary political upheavel would be sufficient to create a new historical dynamic to reconfigure the economic. What follows from this presumption is the idea that left wing economics may be able to analyze capitalism and critique it, but – and this ‘but’ is what I want to contest – the categories it uses to do so are no use to us in thinking what an alternative, transitionary structure might look like. Lenin repeated much the same line in State and Revolution where, despite a chapter nominally devoted to economic transition, the focus is unrelentingly political. For two reasons I believe that today such a position is untenable.
Firstly, because after the experience of 20th century communism, post or anti capitalism has been fully associated with the command economy. It may be the case that workers’ self-management is held out as a potential model that never received a proper test, but the fact that this never materialized seems to suggest that it would not, alone, hold out promise for transcending capitalism. Secondly, and surely as a result of the first point, today the left seems to lack any positive economic vision; and this lack of ideas is now part and parcel of the problem the left has drawing people in under a program for radical change. Most of the unpoliticised general population are not willing to put their faith in political upheavel necessarily leading to a better economic state of affairs, and they associate, in the absence of any countervailing evidence, the leftwing economy with simply a return to the state run, command economy. So ideas for post-capitalist changes are needed. Even simple, single policies that would begin a transitionary process would be welcome.
2. Capital out of date?
Marx’s Capital has recieved a rough ride ever since its publication. Even in the late 19th century it must have been percieved as antiquated, relying upon an unfashionable Hegelian mode of exposition and a radicalised Ricardian labour theory of value that would soon be superseded by the marginalist neoclassical thinkers. Things got no better once the critiques of inconsistency rolled in; critiques that only recently seem to have been laid to rest with the TSSI interpretation pioneered by Andrew Kliman, Alan Freeman, and others. So only today, almost over 150 years after Capital Vol. I was published, do we have a workable Marxian paradigm shared by a community of scholars, and used to conduct econometric work – generally focusing on the long term tendency for a declining rate of profit in explaining crisis.
Why do I believe TSSI Marxism is not enough? The most glaring problem seems to be because there is no clear epistemological reasoning as to why this consistent economic set of laws represents reality most accurately. This allows critics to accuse it of being out of date, and solely reflecting a mode of accumulation associated with Victorian era industrial production, of no relevance to IT, finance, the service sector, and immaterial labour in general. In this context reliance on Hegelian reflection theory as a grounding epistemology is obviously completely inadequate.
Moreover, its dialectical construction seems to leave just too many holes, or residual, undelineated categories. The entire financial edifice, for one thing, is simply lumped into an undifferentiated sphere of exchange. Marxian economics does not seem to extend much further than looking at underlying profit tendencies and capital compositions within a limited, fixed number of categories. The problem thus appears to be how to expand it from simply a posited set of fundamental structural laws to become a complete model able to incorporate more variables and levels of analysis. The impedimenta to this progress appears to be Capital‘s generally dialectical structure (and I write this with a few caveats – for example, I do not believe it is an irrevisably dialectical theory) which makes it very hard to add more variables and levels of analysis. For Capital to become a more serviceable model of the economy it needs to rest on a foundation that would allow it to add more levels of analysis that would ultimately be able to be fed into an econometric model – not, that is, circumscribed to only intuiting very long run tendencies simmering beneath the surface of economic phenomena.
In order to achieve this, I believe a model of Capital needs to focus on its strongest aspects. This would ditch some of the philosophical baggage around commodity fetishism and any unworkable categories, and focus on where it does best – namely on the hypothesis of long term profit tendencies being responsible for more short run economic phenomena such as crisis and financial movements. The model would need to be explicitly a temporal model of how various structures in the economy interact. We also want this model to be revisable to the introduction of new variables and structures. It has to be an open model able to incorporate all relevant structures and phenomena. The axioms of geometry provide a good example here. Euclid’s axioms are consistent, but when placed in an ‘inner model’ on the surface of a sphere they break down, leading to proofs of the independence of some of the axioms and the need for revision. We need to be able to do the same thing for economic theory – we need to be able to test our fundamental structural model to make sure it does not rely on dogmatic assertions.
What I have sketched above is no small task. Indeed, from my initial research into economic models there appears to be very little out there to begin with – there is no off the peg model structure into which Capital‘s categories could simply be dropped. At present it thus feels like something of a tabula rasa effort, and to be successful it will no doubt need to be a collective project conducted amongst colleagues, some of which will have to be more technically and mathematically proficient than I.
The ultimate point of this endeavour is based on the wager that there is a structural truth to the economic crisis to be discovered (in the realist sense), and that a correct diagnosis will help the project to conceive a determinate economic project on the left, and give us tangible ideas to fight for.
A predictive model based on Capital would be somewhat of a holy grail. Being neither a philosopher nor an economist I can’t really help here, though for comparison the mathematical models used (or, indeed, proposed by Soviet economists and rejected) in the 1960s may, or may not, be a productive avenue for enquiry. http://www.amazon.co.uk/Red-Plenty-Francis-Spufford/dp/0571225233 – the slightly fictionalised account where I encountered them, which does include full references.
That looks really interesting, thanks, and a fun read too, which is a relief after trawling through dense material about economic models.
I found this a provocative piece but needs working out; why not try to put something together for publication in one of the growing number of journals that carry articles dealing with Marxian economics?
It contains two points (developing the categories to deal with transition, and incorporating economic model theory) that are thinly connected. Do you think there is a relation between the two, or are is this a wishlist?
I agree with your first point, but I think you could phrase the second more carefully, because it has let to misunderstandings.
I would separate these two issues – you’ll get a different audience for each one. I recommend you approach Andrew Kliman to discuss the first point. I know he and Anne have given serious attention to the whole range of issues that revolve around ‘does Marx have a theory of postcapitalist society’.
So I will focus on the second point. The basic problem is the word ‘model’ and I think that this is the source of the animosity that your post has generated. Andrew and I both argue that among economists (both neoclassical and most Marxists) the word ‘model’ is used in a highly tendentious way. What economists call ‘models’ are in my view not models at all in the sense that you or I, as mathematicians, understand them (eg semantic model, models and ultraproducts, etc). Nor are they models in the sense that engineers or physicists understand them (that is, a set of necessary relationships between variables on the basis of which one may write down predictive equations or families of equations, like say Maxwell’s wave equation or equations governing motion under gravity). What economists really mean by a ‘model’ is a concealed ontology. This is most transparently obvious when simultaneous equations and the simultaneous, or equilibrium method, come into play. The Bortkiewicz/Surplus School/post-Sraffian equations make no real pretence of predicting, or even describing, how an economy behaves or might behave. In fact most serious simultaneists begin by recognising that their systems have no predictive merit at all and that the real world could not possibly behave as their systems describe (profit rate everywhere equal, instant adjustment to change, input prices = output prices, etc). To the contrary, their claim is that the magnitudes one may calculate with their system constitute a ‘good approximation’. One only has to recognise what nonsense would be made of the wave equations, or equations of conservation, to describe them as ‘approximations’ or ‘not governing reality’ to see how far apart are the two usages.
Their systems in fact constitute *disguised definitions*. Value and price are actually defined by the equation systems, and that is the purpose of the equation systems. All the alleged inconsistencies flow from this.
Thus I am rather sympathetic to those who argue that, in order to develop Marx’s ideas, one most first free the Marxists from the prejudice that such ‘models’ constitute a contribution to thought.
As a mathematician, I suspect that you are making a rather different case. I myself developed, in 1992, a simulation ‘model’ of Marx’s schemas of reproduction (which ‘worked’ surprisingly well in that it generated cycles whose periodicity was that of the turnover time of fixed capital) but I did not conceive of it as a predictive instrument. Its purpose was to show that a capitalist system could actually reproduce materially, and in value terms, as Marx suggested in Volume II, if exchange took place at market prices as defined in Volume III. It was to test whether the theory ‘worked’ and was fully consistent. I think this is a valid use of models.
Then the question is whether there is merit to ‘predictive’ modelling of any kind. In terms of a simulation model, this could be done by adding additional, behavioural equations. For example, it turns out one has to define how prices react to excess demand or supply, in order to fully specify the behaviour of the simulation, or how capitalists react to falls in the profit rate. Frankly, I have no real idea how modern capitalism reacts to excess demand or supply. It strikes me as an utterly empirical question.
I do think we can attain a reasonably clear picture of how they respond to declining profit rates (financialisation, speculation, and withdrawal from investment in production). But the problem here is not that Marxists have paid no attention to models, but that they have imposed a ‘model’ in which the profit rate can’t even fall. So they can’t even ask the question, let alone come up with a half plausible answer. This in my opinion is the fundamental reason that Marxism has ‘stalled’; it does not provide a framework in which to explain or understand the most elementary phenomena of the crisis.
One could conceivably test various hypotheses about capitalism by experimenting with different additional specifications.
I don’t want to appear overly negative, but I am not really sure what the point of doing that would be. Perhaps you could explain that a bit more precisely? The purpose of my article on positivist Marxism was to suggest that attempts of this nature are ‘missing the point’, because it assumes that determinism reigns supreme – but, actually, the behaviour of capitalist society is contingent on action. So one cannot actually produce a genuinely ‘predictive’ model since contingent human action can always override it. Rather, one should seek to establish relations between such action and the expected effect. This would indeed be useful. But is it modelling?
Regards
Alan
Alan,
Thank you very much for your thoughtful response, which far from coming across as ‘overly negative’ ask some very important questions and helpfully points to holes in the argument.
On the first point of the ambiguous connection of these two items on the ‘wish list’, it is my intuition that they are somehow connected, but at this stage it is difficult to connect them very precisely except on the level of an intuition that a more sophisticated and multi-faceted model of Capital needs to be developed.
I acknowledge that ‘model’ as it has come to be used in economics comes with a host of wholly negative associations from neoclassical economics and certain neo-Ricardian schools. So there is admittedly a risk in talking about ‘models’ in economics, when one means something very different to how the mainstream has colonised the term. I am extremely interested in the model you developed back in 1992, because this kind of consistency modelling is much closer to the modelling I am interested in, even if in some way I would like the theoretical model to ultimately become transitive with an econometric model.
Again, I am totally with you, and the rest of the TSSI theorists, that the models imposed by previous generations have been generally detrimental to the field. So it is a bit of a gamble I suppose that modelling techniques imported from outside the economics profession could improve matters.
I will try to flesh all this out in more detail as soon as the ideas crystallise in my own mind better, hopefully through discursive engagements such as the present one with more experienced practitioners in the field. My background is philosophy (even if I have an undergraduate degree in economics) so this is rather late entry.
Thanks again for your comment. Any further critique or discussion will be most appreciated.
Best Wishes
One more point, much of the inspiration for this project comes directly from Andrew Kliman’s discussions on ‘the day after’ – even though I’m not convinced the quantity-quality transformation is adequate for thinking the sudden change from the law of value to a new, complex economic transitionary structure.
I think that if you want to pursue the idea of models in Marxism, it would be best to start seriously attempting to distinguish between the different types of ‘model’ in use. To be precise, it would be excellent if you could undertake a critical analysis of the different meanings that are assigned to the word ‘model’ by different users of it.
The basic problem is this: let’s take four uses of the word ‘model’. An simultaneous equation ‘model’ of the structure of an economy, a simulation model of the same economy, the Bohr model of the atom, and a semantic model of a set of statements in the predicate calculus. In each case we see the word ‘model’,
The logical model is certainly not the same as the Bohr model, I am sure you would agree. The first exhibits a set of objects whose relationship can be precisely described by the desired syntax; the second is a conceptual tool for thinking about how an atom ‘might behave’. There is a fundamental difference. The objects in the logical model really do behave as described by the syntax. The correspondence is precise. No deviation is conceivable, or it is not in fact, a model. In contrast, the atom definitely does not behave like the Bohr model; in fact all wave-like aspects of its behaviour directly contradict the Bohr model.
One use of the word ‘model’ refers to an exact logical correspondence. The other is a rhetorical or imaginative crutch for thought. These have no more in common than the use of the word ‘heaven’ to describe the abode of God, and to describe the skies above us.
Similarly, the simultaneous equation ‘model’ directly prohibits to a dynamic simulation model. It is a third use of ‘model’ in that it imposes a priori a set of relations that cannot be violated – for example that profit rates cannot differ, and that prices cannot vary from beginning to end of a period of production. These suppositions are integral and intrinsic and cannot be varied. Without them, one cannot even write down the equations. This is actually an ontology. It *defines* price and profit.
On the basis of this so-called ‘model’ one cannot even write the simulation because, by definition, the quantities one wishes to work with are not allowed. It is a perfect Orwellian tool; the ideas that one needs in order to express dynamic movement, simply cannot be stated in the language permitted by the simultaneous equation ‘model’.
In my view, there is a critical distinction between the Bohr ‘model’ and the simultaneous equation ‘model’ on the one hand, and a dynamic simulation ‘model’ on the other.The first two are metaphors, *rhetorical* devices; they have no logical status and no logical inference may be made from them. The peculiar sin of economics is that it uses metaphorical models *as if* they constituted logical statements. I develop this idea in ‘An invasive metaphor: the concept of centre of gravity in economics’. The notion of ‘centre of gravity’ is a very tempting one for the simultaneous equation theorists. In admitting that profits and prices do not behave as described by their system, they escape with the apparently logical statement that their system is the ‘centre of gravity’ of actual prices. But, as I show in my article, actually prices cannot possibly behave this way. The concept ‘centre of gravity’ introduces a metaphor that cannot hold, to deduce a false conclusion which does not flow from logic.
It is *this* use of the term ‘model’ which I think critical economists quite rightly object to. I think they are right. In fact, it is more or less my answer to your question ‘why has marxist economics stalled’ – because it has accepted this substitute of metaphor for logic, imported via a capitulation to the dominant mode of reasoning in economics with the intention of making ‘Marxist’ economics respectable. The ‘Marxists without Marx’ as I call them, have made a kind of Faustian pact – they will talk the language of models, without challenging it critically, in order to cloak themselves in academic respectability. The reason that they react so violently and abusively to the criticism offered by TSSI scholars is that, were they to accept the discoveries that these scholars have made, they would have to take this cloak off and would no longer be able to parade in academia as ‘some kind of economist’.
Now, is it possible that there is a ‘good’ use of models? Well, that’s an open question. To me it’s a bit like the question ‘is there a good use of the word ‘god”. My initial answer to both questions is ‘maybe, but why would you want to do that?’ One needs to challenge and overturn nearly a century of obfuscation imported via the economists’ practice of substituting metaphorical reasoning for logic under the guise of ‘modelling’. The first task is to clear all this garbage out of the way by resolutely exposing and criticising this primary defect. When that is done – if it ever can be done – then it may be possible to make a genuinely logical and scientific application of modelling techniques in economics. But this will not work, I fear, unless and except accompanied by a sustained and permanent vigilance against the kind of confusion that now surrounds the word ‘model’.
As regards the 1992 simulation of capitalist reproduction, I started on an update last year and will put it up on the web somewhere when it’s at a suitable stage. If anyone out there with PhP and Jscript skills wants to help develop it, it’s waiting for your input!
You have a great project! I have four comments:
1. You will find the explanation to the gap between Marx political analysis and his Capital project in his own plan for the Capital, which was outlined in Grundrisse. Capital was unfinished. It was supposed to be 7 volumes! Volume 3 was heavily edited by Engels, and in some of the English versions also by the Marx Engels Institute in Moscow. Capital was intended to end with an analysis of the state in Volume 7, where the theroy of value introduced in Volume 1 was to be related to the analysis of politics and the state.
2. When it comes to the bridge between contemporary neo-classical economic mathematical models and Marx I will advice you to read Adolph Löwe. Löwe provides a consistent theory explaining the shortcommings not just of of Keynes, but also of the current hegemonic models of capitalism, inspired py physics. He used these models himself, and he saw their failure. Löwe argued for a synthesis between phenological sociology and economics, and he outlined a theory of growth where you can find the answer to your questions of the analysis of post-capitalist economics.
3. Post capitalist economics is the forms of economic organisation and theory which can overcome the weaknesses in the relations of production which prevents not just growth, but also a rational adaptation between man and nature. These weaknesses are unfolding today, through the interrelated economic and ecological crisis.
4. I think the major reason for the failure of contemporary left-wing politics to come up with an answer is the love affair between the left
and Keynes. The Keynesians promised us the solution in 2008, to bail out the banks. In doing so, they went to bed with the class enemy, financial capitalism. But the current Keynesian symbiosis between the state and the financial class is not able to recreate growth, nor make workable new Kyoto agreements. If the “left” is going to have any future at all, it has to start with a critique of Keynes and the neo-classical economic theory which refers to him. In this respect, Löwe provides a useful point of departure.