Classical economic theory, from Smith and Ricardo, holds that the amount of labor used to produce a commodity is the determinant of its price; neoclassical economists reject this, arguing rather that supply and demand — as constructed by preferences and productivity — are what determine prices. In the computer age, digital reproducibility in particular foregrounds the role preferences play in determining price. However, I think the phenomenon of open-source and freely distributed software — while certainly digitally reproducible — gives all economic models (classical, neo-classical, or Marxian) fits. In fact, I wonder whether economists would even call such software a part of the economy, or whether they would name it as an externality, or a product of a non-class process.
But that’s just the start, here. Let’s add another complicating factor, and talk about the place of depreciation on a balance sheet. We understand that the depreciation of a business asset — its loss in value over time — can be written off, in the U.S., as a tax loss. (This tax writeoff is what makes certain REIT mutual funds potentially solid investments.) When Darla the Wal-Mart greeter decides to start her own small Web design business on the side, the $1000 computer she buys will lose value (and not just due to Moore’s Law) over time, and she can deduct that loss from the taxes that her business pays. So one question I have would be, if Darla gets a big client and needs to invest in a $699 copy of LIFT NN/g Pro, will that software depreciate in the same way (for tax purposes) that her computer does? (Anybody with business experience have insights to offer here?)
I wrote recently about possible ways in which university administrators might view computers as capital, generating return on investment; a better way to put it for that model might be to call them “business assets.” However, Charlie reminded me yesterday that the state university, unlike a for-profit corporation, doesn’t pay taxes, and so computers can’t be written off as depreciation. So I need to keep in mind that, as much as American culture’s understanding of economics seems dominated by the neoclassical ideology, I need to be careful when trying to understand the university through such a lens: capital may do strange things in not-for-profit environments.
But see, I’ve also been looking at how writing circulates as a product in the classroom, and whether education itself can be understood as a product. These questions also become more difficult when coupled to the contexts of the university and open-source software.
Wolff and Resnick, citing Capital, equate “the production and distribution of surplus labor” to “the class structure” (128), and point out (my apologies; I’m putting on my Mr. Obvious hat again here) that “Marxian theory is a class theory. The originality of this theory lies not in its claim that classes exist, but in its proposition that they have a particular structure — exploitation — and that this structure shapes what we see, think, and do” (Wolff and Resnick 125). But freely-distributed open-source software seems to be, almost by definition, the very definition of the product of surplus labor rather than necessary labor, and so the wrench it throws into neoclassical economists’ understandings of valuation seems to almost justify the cries of “Communists!” often hurled at the open-source community. It obviates exploitation, and is infinitely reproducible at practically zero cost. The products of for-profit software vendors, on the other hand, leave an audit trail of receipts, warranties, tech support numbers, copyright protection, and budget expenses, all of which serve useful purposes in capitalist economies. (After all, if you don’t spend money on software, your superiors may assume you can do something with nothing, and cut your budget for next year: in some ways, if you can’t put a price on it, it’s useless.)
So we’ve done a little work with the neoclassical perspective. Let’s see what the Marxists say. Here’s Mr. Obvious: “class is actually two economic processes: in one, people perform surplus labor; in the other, the fruits of that surplus labor are distributed. . . Every society of human beings is assumed to require that at least some of its members interact with nature and one another to produce goods and services. This interaction is called ‘the labor process’: the expenditure of human muscles, nerves, and brain power to transform objects in nature into goods and services satisfying human needs and wants. Those members of society who do this labor are called ‘direct laborers.’ . . . direct laborers always perform more labor than the necessary labor. They participate in the labor process for a longer period of time than that which is needed to supply their own needs and wants. This extra time of labor is what Marx called surplus labor” (Wolff and Resnick 144).
Sometimes, Wolff and Resnick explain, laborers (the direct producers of surplus labor) collectively appropriate their own surplus labor, as in an agricultural collective. Sometimes laborers individually appropriate their own surplus labor, as with an artisan who produces and sells enough pieces to support herself but also produces and sells additional pieces, the profits from which she uses as she likes. When one person appropriates another’s surplus labor, though, it’s exploitation: “Wages or salaries are given by the capitalist to direct laborers in exchange for their necessary labor. These laborers give their surplus product to the capitalist — who thereby obtains profits — without obtaining any product in exchange” (147). This production of surplus labor, Wolff and Resnick tell us, is what Marx calls the “fundamental” class process, as opposed to the “subsumed” class process, which details how the appropriated surplus labor gets distributed: “It [the subsumed class process] is motivated by the appropriators’ aim to continue the fundamental class process and their positions in it”; it “is the way appropriators pay for the performance of certain nonclass processes without which the fundamental class process could not exist” (150). Resnick and Wolff list policing and education as two examples of subsumed class processes. The questions for class analysis, they suggest, then become questions of identifying the class and nonclass processes, and their subvarieties, and how they overdetermine one another.
In the interest of such identification, I’ll ask: is freely distributed open-source software part of the fundamental class process, part of the subsumed class process, or a nonclass process? It’s not a fair question, of course, because I’m using two phrases — “freely distributed” and “open-source” — to describe one thing, when Marxian analysis would suggest that they clearly refer to different “channels” of production and distribution. Part of my reason for doing so is that I recognize how tenuous a connection the phenomenon of such software currently has to student writing in the classrooms where I teach: our students and servers use so much branded and protected software that I feel like I should be inviting Eula to parties here, as often as I see her name. (Hell, she’d show up anyway, but she never drinks and she can only talk about one thing. Actually, now, hey, there’s an idea: Eula’s everywhere, right? Make it so! I think Eula, with her number — the registration number for your high-end software of choice — would make a fine graffito: For A Good Time, Call Eula. All Major Credit Cards Accepted.)
The question also leads me to another important distinction: “the useful fruits of human labor are called ‘products'”; however, “For a product also to be a commodity, it must not only be useful; it must also be exchanged (for money or for another commodity) on some market” (155). It clears up some of the questions I’ve asked about the wired writing classroom: in such a context, I think writing is a product, and education a commodity. But the distinction also raises other questions: in the digital economy, is open-source software a commodity?
I acknowledge that I’m not doing a dissertation about open-source software. Still, the model I have of the way writing circulates in the wired classroom’s digital economy leads me to see certain correspondences; correspondences that might give me further insights regarding Marxian and neoclassical understandings of production and consumption, which might themselves lead to insights about the classed contexts of such activities. It feels like fertile ground, where a lot of different theoretical angles might cross-pollinate. I’d be immensely glad for any feedback folks might have.
Interesting questions and tough ones. I can only offer a few observations:
Just as the original authorship of a piece of open source software is usually due to a single person–think Larry Wall for perl and Richard Stallman for Emacs–so it usually is for commercial software, except you usually don’t get to know the person’s name. Many other people may contribute to subsequent enhancements and bugfixes, but this is not a division of labor in the sense Marx subsequently describes in this passage, one where the series of operations carried out by a single craftsperson are split between several. Software that is big and complex enough to absolutely, positively require a team works in another elementary manner of production: different craftspersons are brought together under the same roof. In the same section, Marx gives the example of a carriage as the product of "wheelwrights, harness-makers, tailors, locksmiths, upholsterers, turners, fringe-makers…" (I can’t type in the whole damn list, sorry) all brought together under one roof; likewise, a big piece of software such as distributed application might have one person on the team who does the database stuff, another person who handles user interface stuff, still another person who handles business logic….
For a taste of just how challenging it can be to coordinate a large team software product, check out Frederick Brooks’ The Mythical Man Month. If Moore’s law reflects the triumphs of electrical engineering, Brooks’ law reflects the intractibility of software development problems: Adding more personnel to a project that is behind schedule will only make it fall further behind schedule. And if I had a dime for every manager I’ve worked for who didn’t know that, I might be able to buy us each a case of a Montrachet Grand Cru.
Here the strategy is different–craftspersons themselves are distributing their production below the price of the capitalists–but essentially, they are still skilled workers opposing capital. If, that is, they distribute their software free of cost. Richard Stallman says that the free-software movement means free access to the source code, not free of charge: “Free software” is a matter of liberty, not price. To understand the concept, you should think of “free” as in “free speech,” not as in “free beer.’ (From The Free Software Definition at GNU.org. More pertinent: Selling Free Software, also at GNU.org)
God, I could use a drink after that. Got any Pabst Blue Ribbon? <wink />
Mike..if Darla buys are $1000 computer, then I believe it’s true that (as a small business) she can deduct the full $1000 as an expense in the year she buys it. If she were a large company, or if the purchase were greater than a certain amount, she would have to depreciate it over several years. From a tax standpoint, it’s always better to expense something in the year you buy it than to depreciate it. From a financial reporting standpoint, the opposite is true.
I believe that software can be expensed if it is below some $ threshold, but must be capitalized and depreciated if it exceeds that threshold. In either event, it gets deducted one way or the other.
Wow. Thanks for the really insightful feedback, Curtiss; you’ve answered a lot of questions I had. I had the feeling I was fudging things a bit when I started talking about open-source software, but your quote from Stallman totally helped to clear up the distinctions I was skating over.
So now I gotta ask: where do you get all this expertise? I never thought software engineering and Marxian economic theory would make for happy intellectual bedfellows, but I was clearly mistaken.
That’s a loooooonnnnggg story. Suffice it to say that if what my employer’s really paying me for is my abstract labor power, then it doesn’t really matter if I’m not always writing code.<StupidShitEatingGrin />