Coming Soon:

Now Available: Volumes I, II, III, and IV of the Collected Published and Unpublished Papers.



Total Pageviews

Thursday, February 3, 2011


Mike's comment on the last part of this tutorial indicates that I have failed to make clear a very elementary point about the classical school of political economy. Since this is really important, I think I had better interrupt my exposition to explain this point again, in hopes that I will make it clear enough so that no one will be confused or misled.

Mike notes that in a situation in which people are exchanging goods that they did not produce, the intersection of supply and demand is sufficient to explain the relative prices of the foods -- which is to say the proportions in which they exchange with one another. He is quite correct. Indeed, since in this situation the goods are not produced [at least not by the people exchanging them], they might as well be treated, as he says, as manna from heaven. In his POW camp example, the goods are presumably introduced into the system by the Red Cross. One might also reference the World War II phenomenon of "cargo cults." [Google it.]

The classical political economists were quite aware of this fact. Ricardo, bless him, refers to wines made from grapes grown on a particular side of a hill [I guess he led a rather comfortable life.] He also references old masters, which is to say works of art which, though once produced by someone's labor, are now non-reproducible [save by forgers]. Such goods exchange in ratios that have nothing much to do with the amount of labor required to produce them, and everything to do with supply and demand. [Just imagine what it would do to the price of van Goghs if someone discovered a cache of three thousand additional authentic van Goghs.]

Why the difference? Well, if goods are reproducible [corn, iron, cloth, and so forth -- the standard commodities of the modern industrial economy, whose quantity is in effect unlimited], then a rise in demand will result in a temporary rise in price, which in turn will result in a momentarily higher profit for the entrepreneurs lucky enough to enjoy that blip in demand. Other entrepreneurs, noticing that higher profit rate, will shift their capital into that industry, with the result that there will be an increase in supply. When everything has settled down, Ricardo argues, the ratio in which goods exchange will be determined by the labor embodied in them. The one exception is of course a situation of some necessarily scarce factor of production, which is to say land. That is why he thinks he needs to show that rent is not a determinant of price, but a diversion of profits from entrepreneurs to landowners.

Note that when Leon Walras introduced the modern notion of price determination by the intersection of supply and demand, a decade after Marx published CAPITAL, he posited a pure market situation in which possessors and desirers of commodities shouted out offers until through a process of "tatonnement," an equilibrium had been reached in which no further mutually advantageous exchanges were available to them. He was not talking about production at all. Ricardo [and Marx] would not have disagreed with this theoretical result. They simply would have pointed out that it does not speak to the actual situation of the production of commodities by means of commodities [to echo the title of Sraffa's book.]

We have a long way to go before I am finished with the entire story of Marx's revision and critique of Ricardo's theory, but if we are not clear about this elementary point, the rest will just be utterly mysterious.


JP said...

A request for clarification: I thought that Marx, etc. would agree that anything, including land rents, wages, etc. is a function of supply and demand. So that the Walrassian mechanism would explain wages as well. But then they just add that this is mere surface, and underneath that these prices reflect labour-power expenditure (meaning that in the long run, prices determined by supply and demand will track the amount of labour expressed in the commodity).

Is this not correct? So they would not claim Walras is irrelevant, but say it reflects mere surface.

Robert Paul Wolff said...

It would be more accurate to say that Walras explains market prices, but not natural prices [going back to Smith's distinction]. Everything that is bought and sold in the market is subject to the interplay of the forces of supply and demand [as Smith himself says]. But the natural, or long-run equilibrium, prices cannot be explained in that fashion. [I am not going to attempt here a fullscale critique of modern economic theory. That would take me too far afield, and be a trifle above my pay grade. So please, no comments about the exquisite beauties of marginal utility and general equilirbium theory. I have other fish to fry.]

Mike said...

This is somewhat helpful, but I'm still puzzled. You say that long run equilibrium prices can't be explained entirely by supply and demand. But why not? And even if they can't, what makes you think labor input fills the gap?

Consider the price of oil. If you asked me its natural price, I'd have no idea what you're talking about. Oil has a market price, which fluctuates considerably in accordance with supply and demand. If you asked why oil was cheaper a decade ago, I could, presumably, explain it entirely in terms of supply and demand. I wouldn't need to reference the cost of labor. Similarly, if I knew what the supply of oil would be 10 years from now, and what demand would be, that would be sufficient for predicting a price in constant dollars.

Now let's do a thought experiment. Suppose a technique is invented that halves the labor power required to bring oil to market. But suppose that every oil producer continues to produce the same amount of oil. Thus, the supply of oil hasn't changed, but the amount of labor required to manufacture it has. I would predict no change in the price. Is there evidence to the contrary?

Of course, over the long term, if labor costs drop then producers will be able to bring more oil to market, thus increasing its supply, and thus lowering its price. But supply and demand still appears to be doing explanatory the work. After all, if, for whatever reason, the reduced labor costs did not increase supply, the price wouldn't budge.

In all, there is surely a relationship between labor and price, but it seems we can explain the latter without reference to the former. Is that not so?

Robert Vienneau said...

I think to understand different approaches, one has to enter into the spirit of each approach as it is being presented.

One needn't get too hung up on "the cost of labor" - which differs from labor values - when thinking about "natural prices", also known as "prices of production".

In the post-Sraffian tradition, non-reproducible resources are examined in the corn-guano model. Seriously - you can look it up. (I prefer to think about wine.)

M said...

Mike, you suppose that every oil producer will continue producing the same amount of oil. But why suppose that?

If I run an oil company and obtain a way to produce twice the oil for half as much, then I'm going to produce twice the oil and sell it at the going rate. That way, my profits double until my competitors catch on to this new technique

john c. halasz said...

"The law of supply and demand" is more of a truism than an all-encompassing explanatory principle. (And oil is a bad example, as it is a natural resource, and much of its price consists in the equivalent of land-rents rather than production inputs).

It was Smith, not Marx, who made that invidious distinction between productive and unproductive labor. The intuition is simple: something must first be produced, before it can be exchanged. So the farmer or factory worker is doing productive labor, whereas the lawyer is doing unproductive labor, (which is not the same as unnecessary or non-functional labor), since, when the lawyer draws up the contracts facilitating exchange, he doesn't add vslue, but only realizes already produced value.

Obviously, something that is difficult/expensive to produce will be produced in small quantities and sell for a high price, which will limit its demand, whereas something that is easy/cheap to produce will be produced plentifully and sell at a low price. But in an economy dominated by commodity exchange, nothing will be produced if it sells below its production-cost, (which will come up as an issue in economic crises). But obversely, demand will be constrained not just by what the production system can actually produce, but also by production incomes and their distribution. The oddly named doctrine of "consumer sovereignty" rather obfuscates that basic point.

Marinus said...

Mike, you say the claim is that supply and demand can't fully determine the price of commodities, but that's not it. The position being defended by Ricardo and Marx is consistent with that claim: like Prof Wolff says, that's a different issue. So, it is demonstrable that supply and demand are full determinants of price, and that price is proportional to labour, both at the same time at a position of long-term equilibrium. The reason for this is that labour is an input at demand, and at equilibrium the contribution of the various elements have been worked out and balanced to allow both the supply-demand and the labour theories to be true. So, supply and demand might fully determine the price, but they aren't the only determinants.

All the magic comes in with the assumption of equilibrium (for both theories). Like M observes above (and Prof Wolff has stressed a few times), once there is a change in the supply price, then the market is out of equilibrium because everybody will try to cash in on the bigger profits. There is an upper limit to this rush though: it will only go on as long as the profits in the oil sector (to use your example) are higher than elsewhere. Once the profits are no better than elsewhere, people won't move their capital into the oil sector. If, in fact, people overshot and put too much capital in oil, making the profit rate there fall behind what it could be elsewhere, they'd draw their capital out and go elsewhere. So, eventually, we'd get back to long-term equilibrium.

So, price discrepancies like you describe will (given the common assumptions about market equilibrium, which is a different matter) work themselves out. And, with the other background assumptions in place about equal organic composition of capital (and other factors described in Understanding Marx which we are still getting to here), those equilibrium prices will be proportional to labour.

john c. halasz said...


I addressed a bit of this in my lost comment.

I think you're missing the very basic problem that is being addressed: why is it that two qualitatively utterly distinct things with incommensurable use-values come to be exchanged in certain proportions as equivalents, so that, say, 2 bottles of whiskey "=" 1 Bible? That can't be explained as barter, since barter would entail no constant proportions, (which, incidently, is what is wrong with claims for the "neutrality of money", which suppose an "as if" barter economy). It's that "third" thing, the commodity form stamped on different things, that is to account for this peculiarity. You could just say that actually the "third" thing is money, as a medium of exchange. But then what is money and what qualifies it as a universal medium of exchange and gives it perceived value? (Whatever is adopted as the medium of money, as a general representative of "value" rather than "value" itself, will tend to be something that itself has little or no use-value, is outside, absent from, the system of production: despite the enthusiasm of hard money fetishists, money itself is something of a cipher).

The answer to be elaborated is that a quite extensive set of social conditions involved in a commodity-production economy, that is, an economy in which, with ever increasing intensity, all use-values are produced for purposes of exchange rather than for direct use, is required for such exchange of "equivalents" to be continuously sustained. And not the least of those conditions is a plentiful supply of "free" commodified labor-power.

AFAICT the basic question you've been trying to formulate is just why Marx feels a need for an objective account of "value" and why does he then use labor-value as his unit of account. Isn't that just a thoroughly retrograde piece of 19th century metaphysics, whereas we've now advanced into a much more sober-minded positivism? That is, indeed, a good part of the explanatory burden here. But any claim that Marx is positing some "deep", noumenal reality beyond the realm of appearances, would have to account for all those points of satirical raillery where Marx makes out the commodity-form itself as a piece of metaphysical mystification and styles economics itself as a kind of theology!

Besides which an auction-like account of nominal price formation is not actually a very good empirical one. It offers little by why of explaining the sources of supply and of demand, nor an analysis of the components of costs and value-added.

Mike said...

M: I note that in my initial comment's penultimate paragraph.

John: I don't know why you describe the law of supply and demand as a truism. Surely it's not an analytic truth, and surely we can imagine worlds in which supply and demand have no bearing on price. As for oil being a bad example, just substitute refined oil, which takes lots of labor to produce. The price of that commodity seems to depend entirely on supply and demand and not on the amount of labor it takes to produce it, whatever that means.

Marinus: Lots to say here. I'm curious what you'd say is the long term equilibrium price of a barrel of oil. I'm no economist, but my sense is that there is no such thing. There are short term fluctuations in the price of oil and there are long term fluctuations, both explained by supply and demand (and by expectations of future supply and future demand, of course). Now, I grant that expectations of future labor costs affect expectations of future supply. But it seems to me that, ultimately, it's supply and demand that does the real explanatory work.

One more thing. Reading your comment, I started to suspect that there's a crucial ambiguity in the notion of labor. You say that, according to Marx, equilibrium prices will be proportional to labor. Now, do you mean to the COST of labor, or to labor itself? If you mean the latter, then I don't know what you're saying. If you mean the former, then you're surely right but I fail to see how you're saying anything interesting. Of course the cost of labor affects supply, which in turn affects price. Surely that isn't Smith's or Marx's great discovery. After all, any feudal lord could have told you that. So I suspect Marx's key idea -- that "natural" price is proportional to "labor" -- is either false or trivial. Of course, I'm happy to be educated on the matter.

Robert Vienneau said...


What is being discussed is not the"cost of labor". It is the quantity of direct and indirect labor. So you are correct to say that you "don't know what [others] are saying". I've tried to explain things myself here and here as well as with various posts on my blog (e.g., here.

You will see to me, it's a matter of mathematical economics. Prof. Wolff is much more systematic, and his book Moneybags..., by the way, is focused on something like John H.'s point about why commodities are commensurable.

Marinus said...


The supply-demand theory of price is a theory about price at equilibrium, and nothing else (this isn't controversial, this is just what the theory is). You say fluctuations are explained by supply and demand, but that can't be right. Supply and demand doesn't explain bubble economies, for instance, when people pay high dollar for commodities when the actual demand for them doesn't warrant it. Hence all the empty houses, whole suburbs worth, we see after the mortgage-driven finance crisis. This isn't a problem for the supply-demand theory, because it's only a theory of prices at equilibrium. But it's a problem for the application of that theory (just like the issues Prof Wolff have been discussing are problems for the application of the labour theory of value), since it makes it harder to find situations where the theory actually works.

You could try to explain bubbles or recession economies (which are out of step with actual supply and demand) by stating that whatever determines how attractive a commodity is to buy counts as 'demand', and whatever affects its ease to sell is 'demand'. In that case the collective mania of a bubble economy drives up demand (despite there not being anybody for the buyers to resell the commodities to). But this renders the theory trivial: you have given a definition of supply and demand, not used them in an explanatory manner. This is because you no longer have a specification of supply and demand which is independent of price the commodity actually trades at, so there's no interesting comparisons to be made. This is the point John C. Halasz and others have made here.

The reason I keep talking about the supply-demand theory rather than, say, the subjectivist theory, or the marginal utility theory, is because I'm trying to pick out just the supply-demand part of the equation, which plays a part of many different theories of value.

Mike said...

Robert: Thanks for the links. I'll check them out.

Marinus: Again, lots to say! First, while I'm hardly an expert on these matters, I suspect supply and demand can explain bubble economies, possibly in terms of expectations of future supply and demand. But even if it can't, I don't see how the labor theory of value does any better. Can that theory explain why my home is worth half as much as it was three years ago? After all, the amount of labor that went into building it hasn't changed. In these matters the labor theory of value seems useless (the supply and demand theory less so).

Now, you might insist that the labor theory can explain my home's natural price. But if you said that I'd have no idea what you mean. As far as I can see, there's no price that my home is gravitating towards.

Finally, of course I don't want to define supply and demand in such a way that the theory is unfalsifiable. But is that really my only option? I'm not an economist, so I don't know the technical definition of these terms or how they are measured, but I find it hard to believe that blatant circularity is my only option!

Marinus said...

Blatant circularity is a very common part of all kinds of theories: Newton's Second Law of Motion is a good example (F=mA is a definition, not a theorem), and all of Euclidian geometry is either trivially true, or false (because it misdescribes geometry on curved surfaces). You can avoid circularity in your definitions of supply and demand, as most of the actual theories do, by ante-ing up and actually saying how you measure supply and demand. The thing is, though, that whenever you do this the claim that prices track supply and demand becomes contingent, and frequently false. The point of all the theories of price we've looked at (for both market price and natural price), including the neo-classical theory, is that they are demonstrably true at a condition of equilibrium (plus some other conditions), but not elsewhere. That's an important and interesting result. But it means we need to be careful to be sure to only apply the theory where its preconditions hold (there being the very serious question of how often we are actually justified in treating our actual economies as systems in equilibrium).

If you can prove that market prices follow supply and demand in a non-trivial way, making use of an independently verifiable form of future supply and demand (whatever that might be!), well, let's just say that there's a lot of economists that would be very interested in seeing that result. Good luck!

There's a reason that the assumption of equilibrium plays such a large part in the theories of Smith, Ricardo, Marx, Walras and most everybody else: it's because it's a special case where a hell of a lot of complicating factors drop away. For instance, in our actual non-equilibrium economies, market prices track the price at equilibrium, but does so in fits and starts. With bubbles the actual market price overshoots the equilibrium price (which means that people who buy at the inflated price eventually lose, and those who bet that prices will drop eventually win), and in depressed economies the market price is below the equilibrium price (rewarding people who invest at the deflated prices with eventual pay-offs). Very often, when the prices go back towards equilibrium, they overshoot. Many critics of capitalism claims that the system can't actually stabilise to equilibrium, that it swings from boom to bust and back again, and thus that the equilibrium theories apply only to cloud cuckoo land. In any case, the relationship between actual market prices and equilibrium prices is stupendously complicated, so much so that I don't think there's even a candidate for a theory to give a full account of it. But, if you assume equilibrium, that problem simply drops away. If you want your theory to hold outside of equilibrium, like you seem to want to, then you'll have to wrestle with that. Once again, best of luck!

Mike said...

I'm losing track of the dialectic. I take it we both agree that supply and demand are a large part of the explanation of the price of goods and services. Obviously they don't explain every perturbation, but I think we agree that an attempt to explain price without reference to supply and demand would be sorely lacking.

Now, what about the labor theory? Is labor input a large part of the explanation of relative prices? Could we look at how much labor went into building my house (whatever that means), then look at how much labor went into building your house, and conclude from that how the price of my house will compare with yours? Surely not. The price of a house depends mainly on how many people are interested in it and how many houses there are on the market (i.e. supply and demand). Here labor inputs seem immaterial.

Now, my sense of the matter is that these results generalize. For any good or service, we'll get a much better prediction of its price by analyzing market conditions (i.e. supply and demand) than by trying to figure out labor inputs. Do you dispute this? If so, I'd like to see some empirical evidence to the contrary.

Marinus said...

Well, that's the question that is the subject of the series of blog posts we're commenting on. It's a big issue, and I'll just give the barest rundown of it now.

The claim of the labour theory of value is that, at equilibrium (where the profit rate of each industry is the same), market prices are the same as natural prices, and the price of each commodity is precisely proportional to the amount of labour that has gone into the commodity. You measure the amount of labour by seeing how labour-intensive the industry that produces the commodity is. Thus, where the conditions for this theory holds, the price of a commodity is determined by the method of production for that commodity, with the pertinent factor being how labour-intensive it is. We can only say 'proportional' rather than 'equal' because labour and prices are not measured on the same scale: labour is measured in time (man-hours or man-years, or whatever), and price is not. But since it's a strict proportion (in the example Prof Wolff has been discussing, the proportion is 2.5:1) it's still a very strong claim.

In response to your claim that surely its supply and demand that says how much your house is worth, Ricardo and Marx would respond that the point at which supply and demand equalises is determined by the type of production process used for houses of your type industry-wide if our economy were in equilibrium.

This is something we should believe because we have a mathematical proof of it. But that proof has other preconditions than that the system is in equilibrium. What we're doing here on this sequence of posts is investigating the significance of those preconditions and of the cases where the theory fails.

Marinus said...

I guess even a shorter punchline for the point I'm trying to make is that it turns out, despite how it might appear, that the only workable theories of price available are ones which assess price at equilibrium. This means that if we concentrate on actual prices at any moment in time we are likely to be deceived by how much something is actually worth.

Smith, Ricardo and Marx defended a theory which says that this equilibrium price is determined in a complicated way by how labour is utilised across the whole economic system. Walras and others made this view go out of fashion by instead talking about how supply and demand coalesce to equilibrium points. Later economists, whose work Prof Wolff is explaining and expanding on, have shown that these views are consistent, which means that there are things a labour theory might show to us which a subjective theory of value might obscure. But everybody agrees that there is an underlying structure to the price fluctuations we see all the time, and that actual market prices are far from the final word.

This isn't an absurd conclusion at all. All we need to accept to see that actual market prices obscure more than they reveal is that sometimes the market values commodities wrongly, and we can see that because lots of money gets made by buying commodities below the value they eventually correct to, and lots of money gets lost by people buying or selling at the wrong point of price corrections. If the actual market prices were the real value of commodities, they could never be wrong, but they sometimes are wrong, as people why trade at those prices often find out.

Robert Vienneau said...

My take is a bit different.

Smith and Ricardo meant something different by "supply and demand" than did Marshall and Walras. There are no schedules of either in Smith and Ricardo. (I have no idea what Mike thinks he means by "supply and demand".) I hold that, for many and weighty reasons, the theories building on Marshall and Walras have broken down. I emphasize arguments dealing with capital, especially the Cambridge Capital Controversies. (Because economists (outside of, e.g., Amherst) basically lie when they teach mainstream economics (for reasons Marx knew), many students of economics will be totally ignorant of what I am talking about.)

Anyways, I bet Mike has not heard of the non-substitution theorem. Where is supply and demand in this example? (On can put aside labor values when thinking about this question.)