Tuesday 11 December 2012

Krugman, Robots and Exchange Value.

Again I will break a promise. Instead of finishing my 2-part comment on Berlin's 775th anniversary, I will write about something that caught my attention and that of my friend Ramanan.

Paul Krugman recently posted a couple of pieces on technology and inequality. Those two pieces are valuable in themselves and I recommend them to my readers. (See here and here)

Unbeknown to me, both pieces were part of a larger debate, of which I became aware thanks to Ramanan. At one hand, Izabella Kaminska (FTAlphaville) and a group support a "technological unemployment" thesis, with clear Marxian overtones (apparently, perceived by Krugman). At the other hand, Tim Worstall (Forbes) and others oppose this. Kaminska herself offers a brief background here.

Frankly, I haven't got a clear picture on the whole debate. Therefore, I will abstain from general comments. I will, however, comment specifically on today's Worstall post, "That Robot Economy and the Rentier Class". (See here)

Worstall summarizes Kaminska's party's view thus:
"The argument is that as robots become capable of doing everything then there will be two very stark classes. The ones who own all the robots and thus get all the money and then the rest of us who live on whatever scraps anyone bothers to tax out of the robot owners".
I am in no position to comment on the fairness of Worstall's summary; so, I'll take his word for it.

After that summary, Worstall goes on to argue that machine-produced goods would become cheaper faster than workers would lose income. So, for instance, if cars are entirely made by machines, their prices would fall faster than car-buyers' income. In other words, real wages, measured in cars, rise.

The first point to make is that that is an assumption which Worstall did not argue. Here, I'll make mine his words: "I'm afraid I just don't see it".

The second point is that Worstall inadvertently stumbled upon a proof, by reductio ad absurdum, of the proposition "labour is the source of all value" (proof sketched by Ernest Mandel in 1967 in his "An Introduction to Marxist Economic Theory").

Let's assume the negation of the proposition (i.e. "is not the case that labour is the source of all value"). Knowing of this, capitalists everywhere invest in machines and sack all their staff.

Now, machines make everything: cars, washing machines, iPods, clothes, medicines, sausages; they write news stories (see here) and even books, diagnose and treat diseases, work the land; run movie-theatres, shoot the movies, compose their sound track, write their script and act; they make and pilot commercial planes; run radio and TV stations; cook, clean and look after babies at home, and have sex for money, in the streets; teach at schools and universities, keep law and order and break them, as required; sweep the streets and help in shops, and design other machines.

In a society where no one works, because machines do all the work (as said literally by Worstall himself in his summary), no one earns wages. Leaving aside "whatever scraps anyone bothers to tax out of the robot owners", if such scraps were forthcoming, no one has incomes.

Now, if no one has incomes, then no one can spend, no matter how cheap the cars, the washing machines, iPods, clothes, medicines, sausages, movie tickets, doctors' fees.

If these things cannot be purchased in the market, because would-be buyers have no income, they have no market price. Their market price is not defined.

Is not that those things are not useful. There is no a priori reason to doubt machines can make quality, useful stuff: stuff with "value in use" (as Adam Smith used to say); or with "use value", as Marxists say.

Is not that those things lack "utility", as I suppose Worstall would say.

Is that would-be purchasers lack the wherewithal to pay for them. Their market price is undefined, and so is their "value in exchange" or their "exchange value". Whatever utility they might have, it's irrelevant, without the income to pay for the goodies.

So, we find a paradox here: in their effort to reduce costs, capitalists (intent on profiting from their "entrepreneurship") end up unable to sell their much cheaper and more abundant produce, let alone to make a profit or even recoup their investment.

We must conclude that the proposition "is not the case that labour is the source of all value" is false. That is, whether readers like it or not, the proposition "labour is the source of all value" must be true.

To put things differently, it is the labour exerted to produce things, for which workers are paid, that allows workers to earn an income, income they use to buy those things they produced. This is what confers "value in exchange"/"exchange value" to the commodities.

And although Worstall stumbles quite badly with other things (among them, his highly idiosincratic definition of communism), I'll let them pass. With this is enough:
"For here’s a little known point that I like to make. We don’t actually care about jobs: don’t care if no one at all has one. We also don’t care about incomes: it’s not a problem if everyone has very low incomes. What we actually care about is that everyone has the opportunity to consume".

32 comments:

  1. hmmm. I don't think that the lack of consumer demand would be the biggest issue in a fully-automated economy. Capitalists have their own investment demand for productive inputs and this investment demand makes up a great deal of the economy. (and this is why a crisis can't be the result of underconsumption...)

    A fully automated society would still have plenty of demand for productive inputs. It is conceivable that all production could be production for inputs in other industries, thereby by-passing personal consumption altogether.

    The larger question this raises is: upon what basis would the exchange value of these commodities be based? Without the domination of labor by capital what would be the social basis of commodity values? Obviously the required inputs and exchanges between machines could be worked out in a technical fashion based solely on the needed inputs and outputs to keep the system of production going. But there is no reason why we would need exchange value for any of this. Prices and money would be redundant. It could all be planned by one big computer.

    (Eerily this looks kind of like the theoretical starting point for many aspects of neoclassical econ: parties meet in the market endowed with unique natural endowments of commodities which they then exchange based solely on subjective valuations. Production is wiped from the picture, value doesn't exist in any objective form, and money is redundant. In Walrasian general equilibrium the market is replaced by an auctioneer who more or less looks like a planner.... Thus neoclassical econ abstracts away the very thing it is trying to describe: the system of production based on the domination of labor by capital.)

    ReplyDelete
    Replies
    1. Hi K101

      Thanks for the thoughtful comment.

      After thinking carefully about your reply, I suspect you are missing an important point. Further, this point -I believe- could explain your first paragraph.

      The clue is this: "It could all be planned by one big computer".

      No, I don't think it could. We have exactly the same number of individual capitalists as we had before, when human workers had a job. Each and everyone of these capitalists aiming to profit.

      But, there is no variable capital (V), because by hypothesis there are no wages. No workers producing cars or bread, and no workers producing the steel and the wheat flour used in making cars and bread.

      The machines pass on their full value to their output; no more, no less. Therefore, there is no surplus value (no S), either. No law would forbid US Steel to sell steel to General Motors, of course, but they won't be making any profits from their workers' efforts.

      (For readers not familiar with Marx: this roughly means there is no value added, which is wages and gross profits; and if there is no added value, GDP is 0. Note that this is a stronger result than the one I argued in the text).

      The only way some capitalists could profit would be by selling their output with an arbitrary margin of profit. Say, US Steel adds a markup of X% on each metric tonne of steel. But General Motors couldn't pass on the extra expense. Why would they keep in business?

      A big enough computer theoretically could do the job of matching technical inputs and outputs, but it could not reconcile the capitalists' attempts at profiting from each other. And I can't see how it could provide the capitalists with any profits.

      Can there be a capitalist without profits? I think of it this way: why would a capitalist run a firm if the best she could do is to make sure she keeps her constant capital (C)? She can do that by depositing the money in a bank account, with no interest.

      Does this look like a capitalist economy? Not to me.

      Incidentally, this explains why I am not convinced about the impossibility of underconsumption. You are welcome to try and argue this point, however.

      Cheers,

      PS,
      We could also have a look at numbers, if you prefer.

      Delete
    2. If there is no profit than clearly there is no reason to invest and it's not capitalism. That's why I say that it just become say a technical matter of adjusting inputs and outputs and that this could just be handled technocratically by a computer rather than through the anarchy of the market.

      If the point of the law of value is to discipline labor then value loses all relevance in a society with no wage labor. Inputs and outputs are measured as use-values, not values.

      In terms of capitalists trying to rip-each other off in exchange, this would, first of all, not allow for any aggregate profit so it is unlikely to be a real source of dynamism to encourage investment. Secondly, without labor time as substance behind value there would be no way of actually determining whether one has been "ripped off" in exchange. This is because use-values would not have any homogenous intrinsic value that could be added, subtracted, etc, compared in any way except in a purely technocratic, physical sense. There would be no money. Just use-values.

      Example:
      I have a robot that produces WidgetX but it requires WidgetY as input. You have a robot that produces WidgetY but you require WidgetX as input. We decide to trade. There is no reason why we would need any money or concept of value to make this trade. All we would need to know would be the technical proportions of inputs and outputs.

      This is why I think the non-capitalist nature of a fully automated society is the result not of the lack of consumer demand but rather the result of a more fundamental reason: the absence of value production.

      Delete
    3. I think we both agree that a fully automated economy would not be capitalist. I am trying to argue that the reason it wouldn't be capitalist is not because of the lack of consumer demand but because the absence of any workers. The function of value is discipline and apportion labor. It has no function in an automated system. Adjusting inputs and outputs would become a purely technocratic matter.

      In regard to capitalist competition in this scenario: First they'd merely be competing to rip each other off in exchange. There would be no aggregate profit. And without an aggregate profit I doubt there would be that much wiggle room for ripping each other off in exchange before people would go out of business. So maybe there would be a massive consolidation of capital to the point of one big robot factory...

      ... but regardless I don't think there would be any basis for exchange without labor anyway, and no need for money. Without labor goods would just be heterogeneous use-values, not reducible to a common substance, and there would be no value in the abstract. The exchanges between robots would not be an issue of value exchanges but purely technical exchanges of inputs and outputs to keep the system operational. This is why I say it could just be all calculated by one big computer.

      The parallel with underconsumption is this: it is not consumption which drives or defines capitalism but value production.

      Delete
    4. (In response to the last of the two comments)

      "I think we both agree that a fully automated economy would not be capitalist."

      Yes, we do agree on that.

      "I am trying to argue that the reason it wouldn't be capitalist is not because of the lack of consumer demand but because the absence of any workers."

      I know.

      ----------

      For brevity's sake, let A represent "the lack of (effective) consumer demand" and B, "the absence of workers".

      You believe there is a dilemma, where (A) precludes (B), and (B) precludes (A), as cause for the demise of capitalism. (In general, I'd go further and I'd say you consider that capitalists crises are caused exclusively by falls in the rate of profit and that underconsumption has little if anything to do with them).

      I am trying to argue that there is no such a dilemma: I believe (A) and (B) are linked; they are the two sides of the same coin, so to speak.

      ----------

      This is how I, so far, have attempted to argue the link between (A) and (B).

      If you go through my previous reply, you'll notice that if machines were to replace workers entirely, V would be zero.

      This implies TWO things:

      (1) final aggregate demand = 0 (the underconsumption crisis); and
      (2) as only machines are used, S = 0 (profitability collapses: the profit rate falls to 0).

      Both things happen simultaneously and not by coincidence. At least in the scenario considered, both are causally linked by the fall in V, their common cause.

      Allow me to quote this passage, which I hoped had been descriptive enough:

      "(...) it is the labour exerted to produce things, for which workers are paid, that allows workers to earn an income, income they use to buy those things they produced". So, no labour? No wages and no demand, either.

      But it is this very same labour that would have provided capitalists their profit: So, no labour? No profits, either.

      ----------

      Assuming an economy with no government and no foreign sector, I can put this all in a much more colourful and succinct way: by sacking their workforce capitalists eliminate their customers. And we better understand this "elimination" literally.

      On top, they also eliminate their own profits and (if workers just lie down and die) create a society without class division.

      Delete
    5. But not all demand is consumer demand. Capitalists have investment demand for Department 1 goods. Department 1 capitalists buy and sell to each other. So your argument that sacking all workers means demand is 0 is just not true. Profits would be 0 but not from a lack of purchasing power from workers, but from a lack of surplus value creation. That is why I am saying that you have to go further than just looking at circulation if you want to show that automation ends capitalism. It has to do with value production, or the lack thereof.

      Delete
    6. Brendan,

      I was able to put my ideas together and argue them. I went into details. And I am no special in any particular way.

      You say I am mistaken. Fair enough. It's very possible.

      I can see no reason why you should not be able to put together your ideas, in a level of detail similar to mine and argue them as I did.

      I'd suggest you do something like my previous comment. Not a book-length treatise, by any means.

      I'd even suggest questions you could answer. Explain why the capitalist investment demand will stop final demand from falling to zero. No need to go into too much detail.

      Explain too why capitalists will invest if final demand is low (I suppose at least you will concede that final demand would be low).

      I think it's only fair, after all.

      Delete
    7. I am more than happy to expound upon the point.

      In Vol 2 of Kapital Marx sets up his model of simple reproduction to counter Adam Smith's theory that the entire social product resolves itself to wages paid to workers. "The phrase that the value of the entire annual product must ultimately be paid by the consumer would be correct only if consumer were taken to comprise two vastly different kinds: individual consumers and productive consumers. However that one portion of the product must be consumed productively means nothing but that it must function as capital and not be consumed as revenue." (Vol. 2 p. 440 Progress Publishers Edition, chapter 10 section 'a retrospect on Adam Smith etc)

      Marx sets up a two department scheme where department one produces means of production and department 2 produces means of consumption. D1 produces means of production for both departments. D2 produces mens of consumption for the workers of both departments.

      The commodities of both departments are comprised of the following value components- constant capital (c), variable capital (v) and surplus value (s). Marx then asks what exchanges of value between departments would allow for balanced simple reproduction. His answer is this:

      D1v+s=D2c

      In other words D2 produces all consumption goods for its own workers. It spends the rest of its revenue on constant capital from D1. D1 buys consumption goods from D2 for its workers and sells constant capital to D2.

      All of D2's product ends up as consumption goods. But this is not true of D1. D1 produces means of production for itself. This portion of the aggregate value of society always remains in D1 and never transforms itself into revenue for workers. It never becomes consumer demand. It always remains as investment demand. Imagine an iron mine which sells iron to a steel factory which sells steel to a mining equipment maker which sells mining equipment to an iron mine, etc.

      Later in his model of Expanded Reproduction Marx asks how we can take this model of Simple Reproduction and figure out the exchanges between departments in order to make growth possible. To do this he creates a intermediary production period before expansion can happen. In this period money is diverted away from consumption goods and used to expand D1. This allows for the expansion of means of production which then make growth possible. It follows that in order for capital to grow there must always be a prior growth in investment demand.

      Andrew Kliman provides an empirical illustration of this trend through the 20th century in his book "The Failure of Capitalist Production" stating that "in 2008 investment demand was 72.7 times as large as in 1933, while GDP was only 18.5 as large and personal consumption demand was only 15.4 times as large. So private investment demand grew almost 4 times as rapidly as GDP and almost 5 times as rapidly as personal consumption demand." (p 175)

      Thus, counter to the underconsumptionist instinct, it is not consumer demand that drives capitalism but rather investment demand. This is why capitalism is a system of profit for profits sake not profit for consumer's sake.

      And this is, again, why I think that lack of consumer demand is not a sufficient condition for negating capitalism since capitalism is not a system built around meeting the needs of consumers. It is a system of value production for its own sake and thus it is the absence of any value production that would be the sufficient condition for the end of capitalism under full automation.

      Delete
    8. I am enjoying the discussion between Magpie and kapitalism101. This comment will really just be an aside on an argument made by kapitalism101. It does not bear directly on the main topic of discussion, and will not indicate an opinion on that. I will refer to an economic system such as we currently have, not the purely mechanized economy considered in the original post. For this reason, I have placed my comment down here so as not to interrupt the flow of the discussion.

      kapitalism101: I know Andrew Kliman often makes the argument that capitalists could just focus on investment-goods production without direct or indirect connection to consumption-goods production, but after reflecting on this for quite a few months, I don’t think I buy it. I'll try to explain.

      It is certainly true that there are production circuits that entail producing investment good 1 to use as an input in the production of investment good 2, part of which will then be used as an input in the production of investment good 1. But would closed circuits such as these exist if they were not connected to wider circuits out of which one or more consumption items are produced? I suspect these smaller circuits are ultimately connected to production aimed at final consumers, whether these consumers are of the working or capitalist class.

      If, instead, production of investment goods could occur completely without any connection to the production of consumption items, capitalists could resolve the current crisis simply by nihilistically producing investment goods to exchange with each other in a closed circuit. They could ensure success through contracts specifying the production schedules and exchanges to take place.

      I think the reason they do not do this is implied in Marx's work. For something to have value, there must be a use value (it is a necessary condition, if labor is to be socially necessary) and I'm not sure investment goods producing each other without at least indirectly assisting the production of consumption items really cuts the grade as a use value. Admittedly, I can't prove it, because use value is to a significant extent subjective. But, again, if investment-goods production for its own sake did have use value, there would be nothing stopping capitalists from forming circuits of such production and guaranteeing themselves surplus value without risk or exposure to the vicissitudes of markets.

      If this is correct – i.e. that all production is directly or indirectly connected to the production of consumption items, or, in Keynes' terms, demand for investment goods is a derived demand – then it is open to debate whether capitalism is investment driven, consumption driven, or some combination. Investment depends on the prospect of realizing profit, and realized profit is a function of aggregate demand, including the extent to which consumption is induced out of wage income, since any aggregate saving out of wages subtracts from realized profit. In other words, I think it is possible to conceive of investment as induced by the level of demand or as responding to profitability. In my view, they are really two sides of the same coin.

      Personally, I don’t think this undermines Marx's theory of value or the temporal single-system interpretation of it. I am inclined to agree with both. But I do think it means that demand has to be taken seriously, as does profitability, when considering accumulation and growth.

      I have argued this latter point in greater depth elsewhere:

      http://heteconomist.com/taking-demand-seriously/

      Delete
    9. Peterc,

      I think you are just flat out wrong to say that all production requires a human consumer at some point down the production chain. It is a commonly made statement (for instance Foster and Robert McChesnsey make this argument in their new book Endless Crisis) but it runs against the grain of Marx's analysis of capitalist reproduction in Vol. 2 as well as against reality.

      When you say this is implied in the concept of use-value I believe you are conflating human use-value with use-value in general. There are lots of things that have use-value for capital and not for humans at all. The usefulness is not defined by human subjectivity but by the practical usefulness in fulfilling some aspect of the production process.

      There is a portion of the aggregate capital of society that is always in the possession of Department 1. It does not end up on the shelves of Wallmart. And, as in the Kliman argument I quoted above, this part of the total social capital must grow at the expense of human demand if the economy is to grow. This, at least, is my understanding of the issue.

      When you say "if, instead, production of investment goods could occur completely without any connection to the production of consumption items, capitalists could resolve the current crisis simply by nihilistically producing investment goods to exchange with each other in a closed circuit. " you are forgetting that the reason capitalists can't do this is because this is a crisis in profitability not a crisis of demand. It's fine to talk about underconsumption problems but only if it is in the context of the rate of profit. Without a rate of profit consumption questions don't make any sense. Capitalist investment demand is driven by profitability.

      At heart is the question of whether capitalism is a system of value production for its own sake or of a system for meeting the needs of humans. This is what differentiates a Marxist perspective from a Keynesian one which seeks a way out through redistributing wealth and priming demand.

      Delete
    10. K101, thanks for your reply. We may have to agree to disagree on this particular point. I'll take one more stab at defending my position. If it seems like I am just repeating myself and you got it the first time, I apologize in advance.

      When Marx writes (end of section 1, ch. 1) that "nothing can have value, without being an object of utility. If the thing is useless, so is the labour contained in it; the labour does not count as labour, and therefore creates no value", I think it is a reasonable reading to think that production of investment goods for no other reason (even indirectly) than to maintain investment-goods production for its own sake qualifies as useless. Capitalists have no use for investment goods unless they are enabling surplus value creation, and if they could increase surplus value simply by producing investment goods by means of investment goods with no ultimate connection to consumption even indirectly, they would do so. I agree completely that some production remains entirely within department 1. No question about that. I just doubt that these circuits would exist if it weren't for an indirect connection to the production of consumption items.

      You wrote "you are forgetting that the reason capitalists can't do this [produce more investment goods without connection to consumption as sure profit] is because this is a crisis in profitability not a crisis of demand. It's fine to talk about underconsumption problems but only if it is in the context of the rate of profit."

      I am not making an underconsumptionist argument. Demand (not consumption demand per se) and profit are inextricably connected. A substantial part of demand can be for investment goods. No question about that. But the demand for investment goods is still derived, ultimately, from the demand for consumption goods, in my opinion.

      Realized profit = capitalist consumption + capitalist investment + budget deficit + net exports - worker saving. In other words, realized profit and demand are two sides of the same coin.

      But, also, the amount of surplus value created in production depends on autonomous demand. A higher level of autonomous demand means a higher level of output and surplus value creation unless a decrease in the rate of exploitation more than offsets it.

      Delete
    11. @K101,

      I've taken time to think about your argument and assess it properly, as fairly and dispassionately as I am capable of being. I went back to Marx, I've checked Rosa Luxemburg, I even checked John Fox's popular account.

      While I concede your point on purely formal grounds (namely, there is an investment demand by capitalists, theoretically independent of consumption demand by consumers), I am not persuaded of it and, indeed, paradoxically, became even more convinced it is fundamentally flawed.

      I don't believe the problem lies in the simple reproduction scheme, but in the way you conceive capitalists.

      I am sure you know what I am going to say much better than I, but I wonder if you have realized the implications. So I'll ask you to bear with me.

      Under capitalism, workers do not own the means of production, they are compelled (note the word) to sell their labour-power to the capitalists. In a sense, workers are little more than automata, with very little freedom. This, for instance, manifests itself in the usual assumption that workers consume whatever they earn.

      Capitalists, at the other hand, own the means of production and can dispose of them as they see fit. In this sense, capitalists are much freer than workers. They are not forced to buy labour-power, raw materials, machinery or anything really. This is for them an option. The simple reproduction scheme highlights, among other things, the consequences of this freedom: basic instability. If capitalists take more than V out of circulation, full reproduction is not achieved.

      Once Department I collapses, unlike you seem to assume, there is no reason for capitalists within Department II to trade with each other. In fact, there is good reason to believe this unlikely: as the surviving capitalists lost their own source of income (IIs), they will be tempted to use IIc for consumption purposes. They are free to do whatever they decide.

      The "non-capitalist" reproduction seems, if anything, less likely than the simple reproduction.

      This makes your belief in the powers of the investment demand extremely shaky. The very idea is, to me, unimaginable:

      "Imagine an iron mine which sells iron to a steel factory which sells steel to a mining equipment maker which sells mining equipment to an iron mine, etc."

      For space reasons, I'm forced to split this comment in two. The second part is below.

      Delete
    12. Second Part

      Ultimately, I agree with PeterC in that surplus value needs to be realized in money terms. And this is something that the simple reproduction scheme takes for granted.

      Surplus value is created in production, but it is realized in circulation.

      Much more importantly, Marx's own conclusions look much different, once he considered financial questions:

      "Let us suppose that the whole of society is composed only of industrial capitalists and wage-workers. Let us furthermore disregard price fluctuations, which prevent large portions of the total capital from replacing themselves in their average proportions and which, owing to the general interrelations of the entire reproduction process as developed in particular by credit, must always call forth general stoppages of a transient nature. Let us also disregard the sham transactions and speculations, which the credit system favours. Then, a crisis could only be explained as the result of a disproportion of production in various branches of the economy, and as a result of a disproportion between the consumption of the capitalists and their accumulation. But as matters stand, the replacement of the capital invested in production depends largely upon the consuming power of the non-producing classes; while the consuming power of the workers is limited partly by the laws of wages, partly by the fact that they are used only as long as they can be profitably employed by the capitalist class. The ultimate reason for all real crises always remains the poverty and restricted consumption of the masses as opposed to the drive of capitalist production to develop the productive forces as though only the absolute consuming power of society constituted their limit." (Capital Vol. 3, part v, chapter 30, at MIA)

      Delete
    13. ERRATA:

      Where it appears

      "Once Department I collapses, unlike you seem to assume, there is no reason for capitalists within Department II to trade with each other. In fact, there is good reason to believe this unlikely: as the surviving capitalists lost their own source of income (IIs), they will be tempted to use IIc for consumption purposes. They are free to do whatever they decide."

      It should read

      "Once Department II [consumption] collapses, unlike you seem to assume, there is no reason for capitalists within Department I (production goods] to trade with each other. In fact, there is good reason to believe this unlikely: as the surviving capitalists lost their own source of income (Is), they will be tempted to use Ic for consumption purposes. They are free to do whatever they decide."

      Apologies.

      Delete
    14. Peterc,

      On the use-value point. Use-values are not homogenous. They cannot be transferred down the production chain like values are. Some of the constant capital value of D1 enters D2 and some of it stays in D1. But if a machine in D1 stays on the shop floor in D1 its use value cannot be transferred down the production chain to D2 and to a final consumer. Its use value is purely a use-value for capital in D1. But the same is true for machines in D2! They are only use-values to capitalists, not consumers. They might produce a use-value for consumers but they themselves are not use-values for consumers. Use-value can refer to the utility a person gets from consuming a commodity. It can also refer to any other useful purpose an item is put to. If there is demand from D1 capitalists for a machine then it has a use-value.

      On what basis do you say that investment demand is ultimately derived from demand for consumption goods? It is just an instinct or do you have an argument? I honestly have never heard a defense of this position. I've only heard it asserted. I'd love to hear someone actually put forth an argument. Stating that consumer demand is part of the total demand in society is not the same as saying that all demand resolves itself into consumer demand, or that all investment demand is for the purpose of personal consumption.

      Delete
    15. Magpie,

      I'm not arguing that it is actually possible, under full automation, for have D1 capitalists exchanging goods. I'm arguing the opposite: that without labor there is no value and no basis for exchanging goods, nothing making goods homogeneous, and no surplus value to compel investment. The point of value is to compel labor. Without labor there is no point in money or value. I think the parallel with your argument is that labor is compelled to work for capital because it must purchase means of consumption from capital. Without the consumption of workers we would have no value production.

      My point is not that consumption is irrelevant, but that we have to be clear what the relevance is for specific questions. For instance: I would not claim to have crisis theory all figured out, but one thing that I'm fairly convinced on is that without a theory of the profit rate we can't really say anything definitive about consumption. If demand is being propped up by credit, if there is a rise in financialization rather than accumulation, etc. then we have to ask why there wasn't a robust expansion of D1 to step into the void and make accumulation more stable. Perhaps because there wasn't adequate profitability! Without some theory of profitability we can't make sense of consumption phenomenon.

      In terms of the question of automation, this means that I think the primary point to make is not that without workers there would be no consumption. This can't be true, in a formal sense, since there already is production for investment demand within D1. Therefore the better argument to make is that there would be no value production and hence no capitalism.

      I'm not sure if I understand the relevance of your remarks about surplus value needing to be realized in production to simple reproduction. I don't think referring to Marx's model of simple reproduction implies some version of Say's Law where demand and supply always meet. It is merely a model for explaining the relation of different sectors of production to each other. It is not a claim that capitalism actually achieves balanced growth. But the simple and expanded reproduction models were often taken to mean this by 20th century marxists. But the actual problem that Marx is trying to solve through that whole brutal chapter on Simple Reproduction, is not how to model capitalism in ideal balance, but rather to respond to Adam Smith's assertion that all production resolves itself to wages paid to workers. To do this he discusses the different component of the social product and the exchanges between all these components. He concludes that there is autonomous investment demand within D1 that does not resolve itself to consumer goods.

      Delete
    16. kapitalism101: You wrote: "I'd love to hear someone actually put forth an argument … that all demand resolves itself into consumer demand." I'll try. Equally, I will be happy to be convinced otherwise.

      I agree with all of your first paragraph including this part: "Use-value can refer to the utility a person gets from consuming a commodity. It can also refer to any other useful purpose an item is put to. If there is demand from D1 capitalists for a machine then it has a use-value."

      My claim is that the demand from D1 capitalists for a machine (and its use value) is contingent on it being indirectly connected to production circuits that culminate in the production of machines for D2 that will be used to produce consumption goods. That is, I agree that "D1 produces means of production for itself", as you noted earlier in the discussion, but claim that this production is only socially necessary (and productive of value) to the extent that it indirectly helps to maintain the capacity to produce means of production for D2.

      Actually, for an open economy with government, the demand for a D1 machine could come from foreign capitalists, which would be sufficient from the perspective of domestic D1 capitalists, or from government investment expenditure, which again would be sufficient for domestic D1 capitalists (or, for that matter, global D1 capitalists). So, I am not saying that all production ultimately has to be for consumption within the domestic private sector, although I do think it would ultimately have to be connected to private or public consumption globally.

      I will try to make the following argument: If there can be no way ever to realize profit from a production circuit, whether directly or indirectly, the production could not be socially necessary, and so could not be productive of value or surplus value. (… continued below)

      Delete
    17. (continued …) Take your example of iron: "Imagine an iron mine which sells iron to a steel factory which sells steel to a mining equipment maker which sells mining equipment to an iron mine, etc."

      Initially, let's suppose this is the only productive activity in the economy. Iron is mined using a combination of labor and equipment. Steel is made using a combination of labor and iron. And equipment is made using a combination of labor and steel.

      There will be no consumption goods produced, so for simplicity we may as well assume workers are living on air, receiving no wages. Any value they produce will be surplus value.

      Is it possible for the capitalists to gain as a class from this activity in price terms? I would say no. Whatever one capitalist pays to another is a gain for one and a loss for the other. As a class, they gain nothing. Of course, they could be producing more and more physical iron, steel and equipment but there would be no way to realize profit as a class from this activity. To me, this suggests there is no use value for the capitalists in this example, so no value, and no surplus value. Producing iron, steel and equipment for their own sake, and their own sake alone, cannot profit the capitalists.

      If, on the other hand, as is no doubt the case, the iron-steel-equipment circuit is not only a circuit unto itself (which it is in part) but also necessary for the maintenance of other production circuits producing machines (or materials) with iron or steel, it would be a socially necessary production circuit, even though some of it is entirely closed to itself, and would therefore participate in the division of realized profit among competing capitalists. It would be socially necessary, in other words, for some of the iron, steel and equipment to be produced only for its own circuit because the maintenance of iron and steel production is necessary for other production processes ultimately connected to public or private consumption. Profit, in this case, would be realized for capitalists as a class once machines produced from iron or steel were sold either to foreign capitalists (in the case of an open economy) or government, or when these machines were used by D2 to enable the production of consumption items purchased either by foreigners, the government, or private domestic consumers. The profit realized would form part of the aggregate realized profit competed over by all capitalists who have invested in socially necessary production.

      I am not suggesting that surplus value is created in exchange. Not at all. It is created in production. I am just saying that if a production process could never result in realized profit, either directly or indirectly, its output could not have use value. The production would not be socially necessary, nor productive of value or surplus value.

      Delete
    18. Aside:

      To be clear, I think all three of us (Magpie, kapitalism101, peterc) are equally convinced of the need to end capitalism. My reason for taking the position I am – i.e. that profitability and demand are really two sides of the same coin – is not to argue for the preservation of capitalism. I see nothing to recommend a system in which public policy needs to be geared toward the maintenance of profit for a particular class and in which all social and economic activity is to be evaluated in terms of its suitability to this class.

      A system of slavery can be made to work on its own terms, too. That doesn't make it a system that is desirable to anybody other than slave owners.

      There does seem to be a difference, though, in terms of our relative emphases on production and realization of surplus value. In my view, what really matters to capitalists is whether they will be able to realize profit. If they think they can, based primarily on expected demand and the rate of surplus value (which together influence profitability), they will invest in a production process that, through the exploitation of workers, is productive of the surplus value that they expect to be able to realize in exchange.

      Delete
    19. Peter,

      Why, in your above example, where workers produce surplus value in D1, do you say that "Whatever one capitalist pays to another is a gain for one and a loss for the other. As a class, they gain nothing."? I don't follow this logic. If surplus value is being produced than it can be realized in exchange.

      Conversely, if surplus value is not being produced, than it can't be realized in exchange. This, I believe, is ultimately what links production and exchange. That's why I do not believe I am making some sort of one-sided argument that ignores demand. But ultimately, for there to be demand value must be produced first!

      I would agree that the existence of D2 is a necessary condition for D1, but not because all of D1's demand comes from D2. (I think we've agreed that we agree on this point.) But because the consumption of the worker's means of consumption in the market is a prerequisite for the domination of labor by capital. This does not mean that capital exists to produce for worker consumption. In this sense I agree that all D1 production is INDIRECTLY related to D2 production... not in the sense that all of its demand resolves itself to consumer goods, or that all of its value ends up on the shelves of Wallmart, but because you can't have value production without a working class and you can't have a working class unless they are buying their food from the capitalist class.

      Delete
    20. kapitalism101: Yes, that quote in your first paragraph is poorly expressed/wrong as expressed. Sorry about that. I later considered (to myself) that because it is a production circle, price would need to step up from iron to steel to equipment for each to be producing surplus value, but then if price stepped up again from equipment to iron, the iron producer would lose back all that was gained by the others. So at the beginning of a production period, the input (equipment) for iron would be more expensive than the output, which would not be viable for the iron maker. If, instead, iron, steel and equipment were equally priced, there would be no surplus value, but also no prospective loss, in any of the three production processes.

      You wrote: "If surplus value is being produced than it can be realized in exchange."

      Agreed. But is value produced? That's what is being questioned.

      "Conversely, if surplus value is not being produced, than it can't be realized in exchange."

      Agreed. This is what I am trying to argue.

      "But ultimately, for there to be demand value must be produced first!"

      I would say there won't be any production unless demand for the output to be produced is expected. Otherwise there will be no outlay of v and c. And, again, is value produced? And hence, will there be that outlay?

      "I would agree that the existence of D2 is a necessary condition for D1, but not because all of D1's demand comes from D2. (I think we've agreed that we agree on this point.)"

      Yes, we agree here. Thanks for clarifying that.

      "This does not mean that capital exists to produce for worker consumption. In this sense I agree that all D1 production is INDIRECTLY related to D2 production..."

      We agree here, too, but it seems not for the same reason. I agree with the first sentence above because consumption demand is not just from workers. Capitalists consume consumption items as well.

      The fact we agree "that all D1 production is INDIRECTLY related to D2 production" makes we think we should be agreeing that all production is ultimately, though in many cases indirectly, related to the production of consumption items.

      I think that the D1 production that occurs unto itself (which we agree occurs) can only be socially necessary to the extent it is not wasteful in relation to its reason for existence, which is its indirect connection to D2 production via other D1 production processes -- that is, anything beyond what is necessary to maintain iron and steel for other D1 production directly or indirectly connected to D2 would be superfluous and not socially necessary.

      Delete
    21. Peter,

      I was about to concede your point about the production loop until I realized that this is obviously a case for a temporal model. Let's see if I can figure out how to type it into a blog.

      We have 3 industries: Iron mining (I) sells to Steel making (S) which sells to Equipment making (E) which sells back to Iron (I). At the end of period 1 S buys from I, E buys from S and E buys from I. New value is added in the next production period, and then they buy from each other again. Since we are assuming a bizarre, sci-fi scenario where there is no money spent on wages, but surplus is still created (and yes we both agree that this is impossible), there is no organic composition of capital and thus no need to equalize profit rates. Each industry spends all of their revenue on purchasing their inputs for the next cycle. Thus:

      C= constant capital
      S= surplus value, also profit in this case
      W= total output price. This is also the revenue with which the industry buys inputs of the next period. The seller's W from one period must equal the buyer's C from the next.

      Period 1:

      I: 10C + 10S = 20W
      S: 10C+ 10S = 20W
      E 10C+ 10S = 20W

      Period 2:

      I: 20C + 10S = 30W
      S: 20C + 10S = 30W
      E: 20C + 10S = 30W

      Period 3:

      I: 30C + 10S = 40W
      2: 30C + 10S = 40W
      3. 30C + 10S = 40W

      Period 4: the original iron from period one comes back to the iron industry in the form of new machines

      I: 40C etc
      S: 40C etc
      E: 40C etc

      TaDa! It looks simple now, but it took me an hour to work that out. Which is why I'll have to wait till tomorrow to reply to your other points.

      Delete
    22. kapitalism101: Very nice! I'll await the rest of your reply tomorrow. My brain needs a rest. :)

      Delete
    23. @K101

      In reply to K101 (December 17, 2012 7:46 AM)

      As far as I am concerned, this was our discussion:

      (A) I offered a description of a process where the collapse of profits and the underconsumption crisis are generated:

      By hypetheis, V = 0, therefore "This implies TWO things:

      "(1) final aggregate demand = 0 (the underconsumption crisis); and
      "(2) as only machines are used, S = 0 (profitability collapses: the profit rate falls to 0)." (cfr December 14, 2012 12:32 PM)

      Notice that (A.1) I am speaking of final demand (not of intermediate demand or demand in general) and (A.2) the crisis are simultaneous, not one preceding the other.

      (B) You rejected this on these grounds:

      "But not all demand is consumer demand. Capitalists have investment demand for Department 1 goods. Department 1 capitalists buy and sell to each other. So your argument that sacking all workers means demand is 0 is just not true." (Cfr. December 14, 2012 6:00 PM)

      You argue that demand in general doesn't need to fall to zero (i.e. there would be an intermediate demand, the demand by capitalists for production goods produced by other capitalists).

      (C) I requested a more detailed exposition and even asked the following question (which you didn't answer):

      "Explain too why capitalists will invest if final demand is low (I suppose at least you will concede that final demand would be low)." (cfr. December 15, 2012 8:29 AM)

      (D) You replied with:

      "All of D2's product ends up as consumption goods. But this is not true of D1. D1 produces means of production for itself. This portion of the aggregate value of society always remains in D1 and never transforms itself into revenue for workers. It never becomes consumer demand. It always remains as investment demand. Imagine an iron mine which sells iron to a steel factory which sells steel to a mining equipment maker which sells mining equipment to an iron mine, etc."

      So, this proves that, in theory, a part of intermediate demand does not need to disappear after final demand disappears.

      In simple terms, D1 may be no longer selling fabric to D2, so that D2 makes shirts for sale, but perhaps the coal mine will keep producing coal to exchange it for shovels made by the blacksmith. Likewise, the blacksmith may keep producing shovels, to exchange them for the coal produced by the coal mining capitalist.

      (E) I conceded this point on purely formal grounds, and pointed out that such a situation would be unstable.

      In any case, while the collapse in profits would not precede the collapse of final demand, it would precede the collapse of demand in general, so this means that the two things are not simultaneous, as I conjectured (see A.2). Thus, my conjecture is wrong.

      The occurrence and sequence of events is a crucial part of my conjecture.

      -----------

      Now I am utterly confused by this:

      "I'm not arguing that it is actually possible, under full automation, to have D1 capitalists exchanging goods."

      Either all demand, including investment demand for production goods, ceases, or it doesn't.

      But it cannot be both ways.

      Delete
    24. Magpie,

      My intent has been to give some friendly advice toward the logic of your original argument, not to actually claim that you can have capitalism without labor. I took the original argument of your post to be that without workers consuming things that value would be undefined, and that capitalist would have no one to buy back their product, and therefore there would be no profit.

      I agree that value would be undefined and that there would be not profit but I don't think the correct argument for this is through the route of consumption but rather through production.

      If you say: "workers won't have wages therefore there will be no demand therefore no profit therefore no capital" then I could say "but not all demand comes from wages paid to workers."

      But If you say "workers won't produce any value and thus there is no way for products to be heterogenous, or any reason for exchange, hence no capital" then I don't think there is an argument.

      M original comment argued that we wouldn't need value at all under automation and that we could just do it all on a computer in a purely technocratic manner. My argument about D1 investment demand is not an argument that I think there would actually be commodity exchange in D1 between capitalists. It was that the existence of production circuits in D1 puts a lie to the claim that workers' consumption is what is propping of capitalist production.

      The argument is sort of hard to have since we are dealing with impossible hypotheticals. I, for instance, just posted a reproduction schema for a society where workers get paid nothing at all. This is of course impossible. (It's actually the same as theorizing full automation now that I think about it...) We've already agreed that there is no capitalist production in such a society. So it's a hypothetical model to based on a abstraction, to make a point about a closed production circuit.

      I'm arguing that investment demand ceases not because it is a byproduct of consumer demand but because there is no value production. I took your argument to be the opposite that consumer demand ceases thus investment demand ceases therefore no value production.

      The difference would perhaps appear to be mere semantics, or just nerdy knit-picking, but, as in a lot of things in Marx, sometimes those tiny difference open up really interesting points. Here we can see the possibility for two different takes on capitalism: one where capitalism is a system of production for use, and one in which it is a system of production for production's sake. We also get to some interesting questions about the role of automation and consumption to crisis theory.

      But I think I'm missing something from your last comment. Can you explain what you mean by "final demand" "intermediate demand" and so on?

      Delete
    25. kapitalism101: On further reflection, I think even without addressing my other points, your reproduction schema satisfies me that value and surplus value can be created even in isolated circuits of D1 production (once we have a D2), and that at least some production never need resolve itself into consumption items. You have helped clear up a conundrum (for me) that's been nagging away for some time. Thanks.

      It seems clear to me now, and I'm not sure what possessed me to question the argument. Even in terms of realized profit -- which I frequently consider -- the demand that is relevant is investment, budget deficit and (for an open economy) net exports, in addition to induced consumption entering indirectly through the negative of worker saving.

      I am still seeing profitability and (all) demand as two sides of the same coin, first in that surplus value is created in production and realized in exchange on the basis of demand, and second in that the level of production and surplus value creation (given the rate of surplus value) will depend on the outlays of v and c to initiate production, which will depend on demand/profit expectations. If you were intending to comment on this aspect, I would be interested to read your thoughts.

      Either way, I have enjoyed and benefited this side discussion. Hopefully it hasn't distracted too much from the main discussion between yourself and Magpie.

      Delete
    26. Apologies if my reply sounded too forceful. I did not mean to be aggressive.

      To your question: intermediate and final demand are categories considered in GDP calculations. GDP is defined as the prices (in current $) times the quantities of goods and services provided at that price for final consumption only (intermediate consumption is excluded from GDP).

      Example: a deli sold you a loaf of bread ($0.25). This is a final purchase: the loaf of bread is consumed and disappears from the production process. Note that each loaf sold adds $0.25 to GDP.

      Consider now how the bread was produced.

      The wheat required to make the loaf was sold by the farmer for $0.05.

      The miller bought that amount of grain, made enough flour for a loaf and sold the flour for $0.12.

      The baker bakes a loaf with that flour and sold it to the deli, for $0.20.

      All these transactions are intermediate. The buyers acquire intermediate goods, not for their personal consumption, but as inputs to produce other goods.

      The deli owner sold the loaf for $0.25.

      If one were to add up all the prices paid, one would find $0.62 (=0.05+0.12+0.20+0.25).

      But that is not the value of the net output produced: the price of the flour, for instance, includes the cost of the grain; the price of the loaf (at the bakery) includes the cost of the flour (which includes that of the grain).

      The difference between revenue and material input cost is called value added. The value added by the farmer is $0.05, as it excludes material input costs, it is formed of wages and profit.

      That by the miller is $0.07 (=0.12-0.05) and so on. (There are some simplifying assumptions, but we'll leave them aside)

      If you sum up all the values added, you get $0.25: the same amount the deli owner sold the loaf. This is a general fact. Therefore, one can calculate GDP by final demand (quantities) times price and by added value. Final transactions can be classified in diverse forms, but we'll leave that aside, too.

      Let's pause and think what happened here. Roughly speaking, the sum of values added (also called Gross Value Added) is similar to the Marxist New Value Created (V+S).

      Indeed, as material inputs are excluded, Gross Value Added is made up entirely of Gross Operating Profit and Compensation of Employees (which is the GDP-way of saying S and V, respectively, without sounding too outrageous).

      Incidentally, the organizers of national accounting in the US, during the 1930s, were largely Russian nationals (at least one of them allegedly was a Menshevik) who emigrated to the US during either Lenin's or early Stalin's time (I'm not sure). Politically, these calculations were subject of great controversy in the US.

      In Mandel's intro he suggests an "analytical proof" of the "The Validity of the Labor Theory of Value" (which is how he calls it). I think that, conceptually, the GDP framework could be useful for this purpose.

      PS, If you are interested, I'd be happy to provide links to useful material, both in Australia and in the US

      Delete
    27. Peter,

      The rest of my reply:

      "I would say there won't be any production unless demand for the output to be produced is expected. Otherwise there will be no outlay of v and c."

      The outlay on v and c creates the demand for consumer goods and producer goods. The only thing in question is which goods will be purchased and whether there will be any savings, or whether credit must make up shortfalls, etc. Capitalist investment is what drives the economy, not the demand.

      I think we mean different things when we say that all D1 production is "indirectly" related to D2. By "indirect" I mean that without workers there is no value production. You mean that all production is ultimately production for human consumption (worker and capitalist as you point out.) Does my 4 period reproduction scheme change your mind on this point?

      As to this:

      "I think that the D1 production that occurs unto itself (which we agree occurs) can only be socially necessary to the extent it is not wasteful in relation to its reason for existence, which is its indirect connection to D2 production via other D1 production processes -- that is, anything beyond what is necessary to maintain iron and steel for other D1 production directly or indirectly connected to D2 would be superfluous and not socially necessary."

      Like I said earlier, this autonomous production circuit, rather than being socially wasteful, is actually really important. For one it generates surplus value for capital, which is the point of capitalist production. Secondly, it must grow in order for D2 to expand, in order for us to have expanded reproduction. Hence, Kliman's empirical argument that investment demand grew faster than consumer demand in the post-war period.

      Delete
  2. Let's assume the negation of the proposition (i.e. "is not the case that labour is the source of all value"). Knowing of this, capitalists everywhere invest in machines and sack all their staff.

    This does not follow logically. If it is not the case that labour is the source of all value, it could simply be that labour is the source of some value and there are also other factors that are the source of value. In that case, capitals may retain (some) staff as they are a source of value as well as investing in the other sources of value.

    Unfortunately this would seem to undermine your proof by contradiction.

    ReplyDelete
    Replies
    1. Thanks, Stubby.

      I knew I was looking for trouble when I asked for your opinion... :-)

      But that's a good kind of trouble. I am afraid I myself have to agree with you.

      I guess, it's back to the drawing board. I'll need to research more deeply what Mandel had in mind.

      Delete
  3. There are significant problems with Krugman's treatment of the effects of technical progress, which he presumes occur in a neoclassical world with tendency to full employment and relative scarcities determining allocation of resources. I posted something on the topic at Naked Keynesianism too.

    ReplyDelete
    Replies
    1. I saw your post. The reference to Hicks came a bit like a bucket of cold water.

      I also agree with your general assessment: "Here is a very illustrative case of the limitations of mainstream marginalist (neoclassical) economics, which leads a reasonable and intelligent economist to all sorts of mistakes."

      I haven't, yet, read Krugman's latest on marginalism or your comments.

      Delete