There is a FAQ on the labor theory of value at: csf.colorado.edu/pkt/authors/Vienneau.Robert/LTV-FAQ.html
In addition, this item appeared originally on the PEN-L mailing list and was written by Jim Devine in response to a question about whether the labor theory of value was correct:
It's hard to say that Marx's "law of value" is "not true" if one doesn't understand it, just as it's hard to say that it's "true" if one doesn't understand it. In fact, I think there's a lot of questions about what "it" is. One thing is that the "Labor Theory of Value" does _not_ assert is that "average market prices" equal "labor values." Quite the contrary: deviations of prices from values are just as important as their connection with each other. Marx was quite conscious before he started “Capital” that values and prices deviated from each other. BTW, Duncan Foley's article in the most recent issue of the Review Of Radical Political Economics is quite good on this subject.
Marx's "law of value" is first and foremost NOT a theory of prices. Economists have typically approached “Capital” _assuming_ that it's about prices and pricing, but that seems more a symptom of commodity fetishism than a product of a serious reading. That assumption gets in the way of a serious understanding. (Hey, I've been there. I assumed that “Capital” and the "Labor Theory of Value" was about pricing, too. Then I read the book. The section on comm. fet. and the end of volume III are especially revealing of what Marx's purposes were, as is the first page of volume III.) If Marx had wanted to study pricing, he would have started with supply and demand or a Ricardian general equilibrium system (or a Sraffian input-output parable). Rather, the "Labor Theory of Value" is a theory of social relations between people. (I know that this makes him look like a mere worm in the eyes of economists, but Marx was a sociologist. He saw the economy as part of society.) He sees prices as _obscuring_ these social relations (that's his theory of commodity fetishism in a nutshell), so he uses values instead. If prices actually equaled values, then much of the social relations of capitalism would be more obvious to the casual observer within the system and to the economist. But they don't so, Marx needed the "acid of abstraction" to cut through surface appearances. That's what the law of value is for.
Whereas Neo-Classical economists start their analysis with the isolated individual person coping with scarcity (the Robinson Crusoe story, ignoring all the social relations aspects of the original book, including colonialism, as Steve Hymer pointed out years ago in Monthly Review) and then move on to trying to understand the common-sense but superficial world of markets. Marx, on the other hand, starts with _society_, the society that limits, shapes, and sets the context for the operations of individuals and markets. (I wish he had written like a modern academic, explaining what he was talking about and the progress of his presentation better. The first page of vol. III is a notable exception.)
As many observers have observed, Marx starts with the abstract and moves to the concrete. In volume I, he starts with an abstract commodity-producing society, one without labor-power, capital, or exploitation. Under these weird conditions, on average, prices equal values, because of the zero degree of exploitation. Even then, there are lots of deviations due to the constant fluctuations of supply and demand. This analysis also provides some insights into commodity exchange _in general_ and sort of a moral yardstick (from the capitalists' own point of view) for judging capitalism, i.e., equal exchange under which prices equal value.
It turns out that capitalism fails according to its own moral yardstick, since capitals are able to exploit labor despite equal exchange -- and in the end equal exchange does not prevail. That's the subject of “Capital” volume I after chapter 3: he deals with capitalism, bringing in labor-power, capital, and exploitation (while showing that profits cannot be created simply via buying and selling but must be _produced_ by labor). However, it's a very abstract capitalism, since he abstracts from the differences amongst the various capitals. So we can talk about volume I describing a "representative capital" -- or alternatively, about a "societal factory" in which capital in general faces labor-power in general in an abstract class conflict. (He doesn't deal seriously with labor's side of the conflict, as Mike Lebowitz stresses in his “Beyond Capital”, so that it's mostly a story of capital rampaging over labor. I guess Marx hoped that it would arouse labor to resist and fight for something better.) In this story of abstract capital, one of the key differences between capitals that's abstracted from is differences in the "organic composition of capital." Nor are there any scarcity rents. So, just as in the first three chapters, values = prices. It's much more intelligent that the common NC assumption that an aggregate production function exists (or worse, an aggregate Cobb-Douglas production exists), since Marx is talking about the _shared characteristics_ of diverse capitals, i.e., the exploitation of labor and the accumulation of capital. But it's an aggregate theory, a macrofoundation for the microeconomics of volumes II and III. Here, he's developed the central conservation principle that I think defines the "Labor Theory of Value" more than anything else except the theory of commodity fetishism: the total of all surplus-value produced in the exploitation process of capitalism as a whole equals the total of all property income (profits, interest, rent, some of taxes) in that society. (It's more complicated if we bring in the articulation with other modes of production or the family, but Marx doesn't do so.)
It's only in volume II that Marx gets to microeconomics of the sort that economists talk about. He talks about the role of time -- metamorphoses of capital over time, the circuits of capital, turnover time, introducing the differences amongst capitals. He turns to discussions of the relations between different types of industries (in the famous but often-misinterpreted reproduction schemes) while being very explicit that he is _assuming_ that values = prices. In volume III, he not only brings up the differences among capitals but looks at how they interact with each other. At this point what was obvious all along to Marx comes out: prices _don't_ equal value, while individual profits don't equal the surplus-value that each individual capitalist organized the production of (since in reality, organic compositions aren't equal between industries). In fact, someone can earn revenues and profits without actually contributing to total value or total surplus-value, as with those unproductive folks in the FIRE (Finance, Insurance, Real Estate) sector who simply redistribute surplus-value. They receive revenues and profits because people within the system find that they have little choice but to deal with financiers, insurance companies, and real estate agents. "Supply and demand" redistribute surplus-value to that sector. And a redistribution it is, since the conservation principle referred to above applies. The FIRE sector is able to capture a piece of the aggregate surplus-value pie even though they don't contribute to it. (They do contribute in the sense that Doug Henwood can write interesting books about them, though.)
I could go on (and many would say I've gone on too long), but I've got other things to deal with. This is my last missive for the day. But I'll summarize: Marx's "Labor Theory of Value" is a societal theory (seeing the "economy" as implicitly embedded in society), emphasizing the way in which commodity fetishism -- volume III's illusions created by competition -- obscures the reality of capitalist society, using values as a conceptual tool for prying out that reality, while seeing that society as a unified totality involving the exploitation of labor, so that those who receive profits, interest, or land-rent benefit from exploitation even if they don't exploit labor themselves.