I first read Capital when I was about 16, and probably understood little and certainly remembered even less. So recently I decided to take another run at it, this time aided by Harvey’s Companion to Marx’s Capital, mostly because I recently read Piketty’s Capital in the 21st Century and wanted to see if there is any overlap. As I was reading I began to think that nobody would read Capital for pleasure, until I remembered that some people enjoy having leather straps applied to their backsides.

Please note that I’m writing this as I finish each major section, and it may be a while before I finish. So don’t expect a unified critique or review of the book.

Anyway, buckle up your latex hood, and let’s get down to business.

Part 1

The first three chapters, a hundred pages or so, introduce fundamental concepts and form the basis for what follows. Amazingly, those 100 pages can be condensed as follows:

  1. The basic unit of (capitalist) production is the commodity.
  2. There are three types of value associated with a commodity:
    1. Its use-value: its utility to some person somewhere. This is a relationship between the commodity and the person.
    2. Its exchange value: a relationship between one commodity and another, when those commodities are exchanged.
    3. Its value: this is the amount of socially necessary labor power employed in its manufacture.
  3. In a barter economy, exchange value can be represented as a set of equations eijci = cj where it is understood that eij = 1 / eji for all i, j, and the eij is the exchange rate between commodity i and commodity j.
  4. As a first step towards a money economy, a “universal commodity” c0 can arise or be selected by convention, so that the above equations reduce to e0jc0 = cj, and the terms e0j can be thought of as the “price” of commodity j, expressed in units of the “universal commodity”.
  5. The value of the “universal commodity” is as defined above: the amount of socially necessary labor power employed in its production.
  6. The “universal commodity” has two necessary attributes:
    1. It must serve as a medium of exchange.
    2. It must serve as a store of value.
  7. The amount of the “universal commodity” in circulation at any instant must be at least as great as the amount of all exchange transactions at that instant.
  8. The “universal commodity” is typically gold or silver, owing to their durability and measurability.
  9. The transition to actual “money” occurs when the state sanctions the use of the “universal commodity” and asserts a monopoly on its control (via coinage, for example).
  10. In what follows, I will replace “universal commodity” with “gold”.
  11. Paper money is a symbolic representation of the underlying gold, and there must not be more paper money than there is gold, or the paper money will be discredited.
  12. The amount of money in circulation can be increased by the use of “credit money”. This is basically an accounting of payments owed, and in many cases such debts will be cancelled, because of mutual debt. But in the end, net debts must be settled in actual money.

That’s the gist of it, though I’ve probably missed some important points and gotten others wrong. So how, you might wonder, did Marx manage to expand this into three chapters and 100 pages? Mostly through absurd amounts of repetition and tedious detail. We are treated to several pages of commentary on the exchange of a certain amount of linen for a coat, and treated to several more pages when that same transaction is mediated by money. I can tell you that by the end I was thoroughly sick of that linen seller, and didn’t think much of the coat seller and the bible selling brandy drinking character that makes an occasional appearance. I will admit that there may be more going on here than I am aware of, and maybe Marx was justified in such a large expansion of such simple ideas. Maybe.

Although Marx introduces three kinds of value, the great bulk of the first three chapters deals with exchange-value. Use-value is mentioned only in connection with exchange transactions: namely, the commodity does not have use-value for the seller, but does for the buyer. In a sense, the commodity acquires use-value as the last part of the transaction. Nowhere does Marx attempt any quantitative analysis of use-value, and in fact suggests quite strongly that use-value is a purely qualitative attribute. If that was what he believed then I think he missed some opportunities. Apple corporation and every design studio in the world would argue that there is a quantitative side to use-value, and that it can be measured, and is largely responsible for brand value. Anyway, for Marx it seems that use-value is a sort of binary property: a commodity has use-value for a person or it does not.

The concept of value as being the “congealed” product of labor, and defined as the socially necessary labor power needed for its production, is one that becomes central in the following chapters. It is, at best, a very odd sort of definition. Let’s start by examining the concept of “socially necessary” labor. Marx introduces this term because he wants to make it clear that merely because a commodity is produced in an inefficient way, it does not thereby have greater value than one that is produced efficiently. So he introduces “socially necessary” labor power as the average amount of labor power needed to produce that commodity (average labor productivity, average labor intensity, average degree of automation). Why average? Wouldn’t the “socially necessary” amount of labor be the labor applied with maximum efficiency? Since we introduce “socially necessary” as a way to discard inefficient processes, why reintroduce those inefficient processes as part of this average?

And just what is labor power? If a person driving a Caterpiller scraper can clear a section of brush in 1 hour, and 100 men with shovels and machetes can clear that same section of brush in 1 hour, what is the socially necessary labor power? Is it 1 person-hour? Marx suggests that is so with an example of the introduction of the power loom as having reduced the value of cloth by one half. But what of the actual production of complex commodities, where the production process requires the application of many types of labor ranging from unskilled to highly skilled? Marx acknowledges that there are such differences, and that highly skilled labor is worth more (contributes more value, I suppose), but states without any explanation that all such labor can be reduced to a common denominator, and that the socially necessary labor should be expressed in terms of “simple labor”. This is unsatisfactory, in my opinion.

Despite this labor theory of value being central to all that follows, Marx really does not define this theory of value very well. He seems to suggest that value is the labor component of the marginal cost of production, and his example of the power loom seems to reinforce this suggestion. But that surely can’t be right. When automation is introduced there is generally a large cost that must be amortized over the life of the automation machinery. So the fact that the machinery reduces the amount of direct labor needed per unit commodity surely must not reduce the “value” of that commodity correspondingly. So perhaps Marx means that value consists of the sum of the transitive labor power over all the input materials and machinery, along with the direct labor. That would make a lot more sense (at least from a simple accounting perspective), but nowehere does Marx make this clear - and, in fact, I do not believe that this was his view.

More fundamentally, why should we accept this definition of value? At best it seems to provide a sort of lower bound on the price that a commodity can be sold while not representing a loss to the producer. Though even then, the “value” of the commodity would have to be augmented by the labor expended to transport, warehouse, and actually sell the commodity, so it would therefore acquire additional value at each step on the road to its eventual sale. But why should this lower-bound price be taken as (fundamental) value?

Value, seen as simply a loose lower bound on a commodity’s exchange-value, then becomes merely the endpoint of an ordinary supply curve. And this suggests that use-value, far from being the binary quantity that Marx seems to suggest, is in fact continuous, and represents the demand curve (as a function of price). This, in turn brings up a point: I do not recall Marx anywhere suggesting in the first three chapters that his three types of value are commensurable with one another. In fact, he goes to lengths to show that they are of a different nature, and represent three distinct kinds of relations: between labor and the commodity, between people and the commodity, and between pairs of commodities. These relationships are also implicit in a supply/demand graph, and perhaps Marx believed it was important to disentangle these relationships. But in so doing I believe he discarded important information, including the meta-relationship embodied in a supply/demand graph.

I found Marx’s discussion of money discouragingly wrong. He was wedded to the gold standard, and went off the tracks as a result. In his discussion of the “universal commodity” he said that it had to have two functions: as a medium of exchange and as a store of value. But nowhere does he question why there must be a single “commodity” that has both those properties. In fact, as we know now, it is possible to separate these properties, so that the medium of exchange does not serve (very well) as a store of value. And this separation has really good properties. It means that the medium of exchange (dollars, euros, yen) can expand or contract to meet the needs of the economy, and that this manipulation of the supply of money can happen literally overnight. This in turn means that it is possible, at least in principle, to use money supply as a counter-cyclical tool to soften economic downturns and pull back the reins on economic booms. The downside of this is that money as a medium of exchange is subject to inflation - though the experience of the past 30 years suggests that central banks have become pretty good at managing inflation to low levels.

Because Marx believed that currency must also serve as a store of value, he therefore treats it as a commodity. As a commodity, its value is the socially necessary labor power needed for its production. So, in this view, the value of gold is the amount of labor needed for its extraction and refinement, and the total value of gold must be at least as great as the total value of commodities being simultaneously exchanged (in a money economy). This, in turn, would put an upper bound on that total simultaneous exchange-value. This fact alone should have warned Marx (and all other economists of the time) that there is a fundamental flaw in tying currency to a gold standard (or to any other commodity standard). His way out of this is to (correctly) introduce “credit money” which is simply the balance of accounts between financial entities - your bank account and the loan that you’ve taken from the bank, for example. The commodity (gold) value of all such accounts is generally far less than the nominal value of such accounts, because of mutual or circular chains of debt. But in the end, the net debts must be settled in currency, here named the “means of payment”, so again there is the constraint imposed by the total amount of currency.

So, sadly, because Marx believes that one and the same commodity must be the medium of exchange, the means of payment, and a store of value, he makes a fundamental error and treats money as a commodity, which it most certainly is not. Whether this error propagates into his later discussion of surplus value and capital accumulation is a question.

Part 2

Things begin to pick up here. In part 1 Marx talked at length about commodity transactions: C-M-C, where a seller has a commodity, sells it for money, and buys some other commodity. He mentions only briefly the cycle M-C-M, where someone has money, buys a commodity, and sells it again (for more money?).

Here in part 2 he examines M-C-M and points out that to be useful it must really be M-C-M’ where M’ = M + ΔM. He then asks, as did many economists before him, where the ΔM comes from. He rightly ridicules the answers given to that point: that it comes about by selling the same commodity for more than was paid for it; that there is a class outside the commodity system that somehow injects money into it (landlards, for example); that it can only come from outside the system (as a colonial possession, for example).

His own answer (not expressed quite the way I present it here) is that the actual cycle is M-C1/C2-M’ where C1 and C2 are different commodities. Specifically, C1 is labor power, which is capable of producing more value as commodity C2 than it itself is worth.

In this discussion Marx finally gets around to quantifying use-value (something he had, as mentioned above, left largely undescribed in part 1), by describing the use-value received by the worker (who supplies the labor power) as the goods or equivalent money needed to sustain himself so that he can continue to sell his labor power the next day, week, or month. And the use-value of the labor power, for the purchaser of the labor power, is the value (I suppose, the exchange-value) of the goods produced.

This was, as far as I know, a new insight; the first to explain the origin of ΔM.

This formulation seems slightly problematic, though, in that it once again reverts to treating labor (and pay) as fungible, and does not address the question of differences in skill levels being reflected in differences in pay. It seems natural to suppose that this is a supply/demand function, but Marx nowhere in part 2 says anything about this. And by tying the pay to use-value, where use-value for the worker is sustenance, Marx seems to scrub away the real and actual differences in pay for different types of work. He goes further, in fact, and says that the level of pay is dependent in part on the general level of civilization of the surrounding society (owing, I guess, to minimum wage laws and the like). But I guess this is part of the simplification Marx stated in part 1, where he acknowledges degrees of labor complexity but says that all labor can be treated as simple labor for purpose of the exposition.