Intellectual property often creates strange bedfellows on the left and the right sides of the political spectrum. On the left, many socialists oppose private property in all its forms. On the right, some libertarians, such as Tom Bell of Chapman Law School, are deeply suspicious of the use of intellectual property to block the right of other individuals to think and speak as they choose. While they regard private property as acceptable for physical resources that cannot be used by everyone at once, they draw the line at intellectual property, which can be copied at close to zero cost.
All this anti-IP rhetoric begs one question: how do we produce IP in the first place? On that question, the open source movement - which has already generated some real successes in both operating systems and various kinds of servers - offers its own distinctive institutional response. The movement, whose principles have been expertly analysed by James DeLong of the Progress and Freedom Foundation in his paper “The Enigma of Open Source Software (Version 1.0)”, is organised around three tenets. First, access to source code (the master plan that generates the zeros and ones to which computers respond) must be made available to all. Open availability allows other individuals to tinker with the original program with an eye to improvements in its operation or extension in its use. Second, once someone incorporates open source software in his own programs, then any licence that he issues cannot charge others for its use or restrict them from making further modifications of the program. Third, each licensee agrees that all subsequent licensees may use or modify on the same terms as the original licensee.
The linchpin of much, but not all, of the open source movement is the General Public Licence (GPL) prepared by the Free Software Movement, which covers, for example, the Linux operating system. Its key provision reads: “You must cause any work that you distribute or publish, that in whole or in part contains or is derived from the Program or any part thereof, to be licensed as a whole at no charge to all third parties under the terms of this License.”
Does open source software represent a viable alternative to the competing forms of proprietary software, which is licensed only for a fee? In answering this question, first note that open source software relies on the very private property regime that its supporters, most noticeably Richard Stallman, disdain on moral grounds. As DeLong rightly notes, the GPL does not place open source software in the public domain, where any one may use it as he pleases. Rather, like the “copyleft” movement in general, the GPL often supplies, as with Linux, an all-enveloping ownership structure by which a central committee decides whether to incorporate proposed changes into the basic public program. (Anyone can keep whatever version he likes for personal use.) No cash compensation is paid to the self-selected improvers, who either work for the love of the game, or because they are supported by some third party payers, either in universities or industry, who want to keep this alternative platform alive.
There are two serious weaknesses with this governance system. The first is that the critical provision of the GPL set out above has not been tested. Yet it may prove vulnerable on least two grounds. First, as a straight interpretive matter, it only states what the obligation of each programmer is with his own private improvements. It does not in so many words specify the appropriate remedy when some portion of the open source code is incorporated into an otherwise proprietary program. The apparent intention of the provision is to “infect” that new program so that all of its content becomes open source software subject to the GPL. In principle, the entire Microsoft operating system could count as “the work” that becomes open source because a few lines of open source code have been incorporated into it by inadvertence. I doubt very much whether courts will tolerate that extreme remedy, if they enforce the clause at all. Just imagine if Microsoft insisted that it had exclusive rights to any derivative work that incorporated its code! In both cases, it is much more likely that courts would allow the incorporator to remove the offending lines of code, or to pay some damages for the improper inclusion.
Does open source software represent a viable alternative to competing forms of proprietary software? To join our discussion click here
Second, the clause might only bind those people who know that they are using open source code. Suppose, for example, that A uses some open source code in his program, which is in turn used by B. If B has no knowledge of how A cobbled together his program, then the GPL may be read not to apply at all. Indeed, even if B knew of the provision, the GPL might not apply on the ground that it constituted an illegal restraint on alienation that everyone is free to ignore. One way or another, courts are likely to counteract the creeping imperialism of the GPL licence. Once the contract protection lapses, then the open source movement is left only to its copyright remedies, which are likely to prove far weaker.
The difficulties with the open source movement , moreover, go deeper than the problems with a single provision of the GPL. The open source movement shares many features with a workers’ commune, and is likely to fail for the same reason: it cannot scale up to meet its own successes. To see the long-term difficulty, imagine a commune entirely owned by its original workers who share pro rata in its increases in value. The system might work well in the early days when the workforce remains fixed. But what happens when a given worker wants to quit? Does that worker receive in cash or kind his share of the gain in value during the period of his employment? If not, then the run-up in value during his period of employment will be gobbled up by his successor - a recipe for immense resentment. Yet that danger can be ducked only by creating a capital structure that gives present employees separable interests in either debt or equity in exchange for their contributions to the company. But once that is done, then the worker commune is converted into a traditional company whose shareholders and creditors contain a large fraction of its present and former employers.
“The open source movement shares many features with a workers’ commune, and is likely to fail for the same reason: it cannot scale up to meet its own successes”
The bottom line is that idealistic communes cannot last for the long haul. The open source movement may avoid these difficulties for outside contributors who work for credit and glory. But how do the insiders, such as Linus Torvalds, cash out of the business that they built? And in the interim, how do they attract capital and personnel needed to expand the business? Traditional companies have evolved their capital structures for good reason.
But suppose this analysis is wrong. One clear policy implication remains: this novel form of business association should succeed or fail on its own merits. The do-or-die question is whether open source offers a low cost solution to particular problems. Ordinary companies will make just those calculations, but government agencies may be swayed to take a different tack, as has been suggested by a number of EU studies. That temptation should be avoided. Governments are bad at forcing technology by playing favourites. If open source is less effective than proprietary software, that gap should not be ignored by positing some positive network externalities that come from giving it a larger base. Proprietary systems also show positive network effects from increased users, as software designers are always attracted by a larger installed base. It’s a tough world out there, in which no one should be exempted from the general competitive pressures of the marketplace. The fiduciary duties of government to all citizens demand no less.
The writer is the James Parker Hall Distinguished Service professor of law at the University of Chicago and Peter and Kirsten Bedford Senior Fellow at the Hoover Institution
James Boyle: Give me liberty and give me death?
The price of liberty is death, at least so far as free software is concerned. Or so goes the argument in Richard Epstein’s column, “Why open source is unsustainable”. It is a characteristically provocative title, but I am unconvinced.
The article gives two main reasons for open source’s doom. The first is an attack on the vagueness and, somewhat paradoxically, the imperialism Professor Epstein sees in the General Public Licence, the licence common to most “free software” and much “open source software.” Prof Epstein claims that the licence is silent on certain key issues, vague on others and likely not to be enforced by the courts in certain cases. I disagree with most of his arguments, and think the fears are exaggerated. An extended analysis would require a law review article, not an Op-ed. How is a reader to judge whether there are deep flaws in the licence? Two handy guidelines suggest themselves. Listen to the market, and assume judicial common sense.
Listen to the market
Global businesses such as IBM have very good lawyers. They are not known for investing billions of dollars into businesses built on licences that are simultaneously vague and imperialistic. (I imagine an absent-minded Genghis Khan.) Unenforceable licences are also unpopular. In his scholarship, Prof Epstein has pointed out eloquently that the market is the best information processing system we have: we should assume that it is incorporating all available information. If we apply his principle here, it indicates that the market has weighed his fears and found them wanting. In my view, the market is discounting Microsoft’s stocks moderately because of fears about the competitive challenge posed by open source, and discounting open source-reliant stocks mildly, because of fears about legal challenges to the GPL or software produced under it. (The much-hyped SCO litigation, interestingly, is not a challenge to the GPL itself.) That does not mean that the free software movement will inevitably triumph. Nor does it imply that the GPL is seamless - no licence is. But every business has an element of legal risk, or contract-uncertainty; the GPL seems to me - and, so far as I can tell, to the market - less uncertain than most. As for the implication that the intent of the GPL’s authors or users is to “infect” proprietary code, I think the evidence runs exactly to the contrary. So far as one can tell from their words and deeds, the authors of the GPL want passionately to avoid the entanglement Prof Epstein describes. They certainly take pains to specify the ways that one can avoid problems, including distribution of separate programs on the same disk, appropriate ways to use proprietary “software libraries” and plug-ins, and so on.
Assume judicial common sense
Another guideline is more a matter of legal culture. To the extent that there are problems with the GPL, they are unlikely to produce either the vagueness or the imperialism Prof Epstein describes because, as he concedes, courts strive to interpret licences so as not to undermine legitimate expectations. Legitimate expectations here would include “multi-billion dollar enterprises that people have erected on the premise that this licence actually works.” Courts are also unlikely to doom multi-billion dollar proprietary software businesses just because someone inadvertently included a line of GPL code. Despite occasional examples to the contrary, courts are fairly commonsensical institutions, and their decisions are unlikely to bring about the legal apocalypse for either side of the proprietary hedge. The market valuations that I mentioned before probably reflect that point.
“In this kind of commune, each of us could take all the property the community had created with us when we left, and the commune still be none the poorer. Copying software isn’t like fighting over who owns the candles or the VW bus”
But legal uncertainty is only part of the reason that Prof Epstein thinks that open source is unsustainable. His key criticism is that “idealistic communes cannot last for the long haul.” Well, the Catholic Church is also a relatively idealistic institution, based on canonical texts that are subject to conflicting interpretations. It is doing pretty well so far. Presumably the key word here is “commune.” But is open source a “commune”, holding tangible property in common and excluding the rest of us, worrying about how to split up the proceeds if someone leaves because of bad karma? Or is it a community, creating and offering to the entire world the ability to use, for free, non-rival goods that all of us can have, use and re-interpret as we wish? In that kind of commune, each of us could take all the property the community had created with us when we left, and the commune still be none the poorer. Copying software isn’t like fighting over who owns the candles or the VW bus.
How about idealism? Prof Epstein himself is careful to point out that it is by no means clear that the production of open source software is based solely on the idealism of its creators. There are lots of reasons that people write open code. They want to solve a particular problem and don’t mind others getting the fruit of their efforts, because they themselves benefited from the earlier work of other programmers. They believe in free software. They hope to get a better job. They are good at coding, and like to display their virtuosity. They are paid to do it. The last category is an increasingly large percentage of the whole. Amazingly, IBM now earns more from what it calls “Linux-related revenues” than it does from traditional patent licensing, and IBM is the largest patent holder in the world. This does not seem like a community that is declining.
People used to say that collaborative creation could never produce a quality product. That has been shown to be false. So now they say that collaborative creation cannot be sustained because the governance mechanisms will not survive the success of the project. Prof Epstein conjures up a “central committee” from which insiders will be unable to cash out - a nice mixture of communist and capitalist metaphors. All governance systems - including democracies and corporate boards - have problems. But so far as we can tell, those who are influential in the free software and open source governance communities (there is, alas, no “central committee”) feel that they are doing very well indeed. In the last resort, when they disagree with decisions that are taken, there is always the possibility of “forking the code”, introducing a change to the software that not everyone agrees with, and then letting free choice and market selection converge on the preferred iteration. So far, “forks” have been comparatively rare, but are not unheard of; the tradition of “rough consensus and running code” seems to be proving itself empirically as a robust governance system.
Prof Epstein is careful to note that he might be wrong about the future of open source, but he concludes with an admonition nonetheless. Even if he is wrong, “this novel form of business association should succeed or fail on its own merits.” It should not be aided by government agencies “playing favourites.” “If open source is less effective than proprietary software, that gap should not be ignored by positing some positive network externalities that come from giving it a larger base.” Given that initial “if”, I think this is a reasonable point. If open source software is less effective, government should not be investing in it. (Some people assume it will always be superior: I do not.) The point, of course, is that most of the government recommendations to invest in open source are based on assessments that, for a particular task, open source is actually superior and that adopting open source has important benefits because of its design - including ease of modification, and ability credibly to pressure proprietary providers to lower their prices.
The UK, for example, concluded last year that open source software “will be considered alongside proprietary software and contracts will be awarded on a value-for-money basis.” In fact, in a serendipitously timely report about British pilot study trials issued a few days after Prof Epstein’s column, the Office of Government Commerce had this to say. Open source software is “a viable desktop alternative for the majority of government users” and “can generate significant savings... These trials have proved that open source software is now a real contender alongside proprietary solutions. If commercial companies and other governments are taking it seriously, then so must we.” Sweden found open source software to be in many cases “equivalent to - or better than - commercial products” and concluded that software procurement “shall evaluate open software as well as commercial solutions, to provide better competition in the market.” Sounds sensible to me.
I think Prof Epstein’s neutrality principle is a little narrow. There are many benefits to society as a whole that governments could rationally factor into their decision in picking open software - including creating a social good that other citizens can share, and producing specific competition (lower software prices for my department) and general competition (lower prices for the society as a whole.) But let us say that we adopted his principle. Would it change state purchasing policies? I don’t think so, for the reasons given above. What would it change? There, I think the answer is clear. The key implication of a principle of neutrality would be this; it would change our intellectual property policy. If we were truly neutral, we would be as concerned about the impact of software patents on open source software development as about the impact of illicit copying on closed source software development. We would spend as much time thinking about how to encourage distributed creativity as we do about encouraging proprietary “top-down” creativity. That principle of neutrality would be worth adopting. Where do I sign?
The writer is William Neal Reynolds Professor of Law at Duke Law School, a board member of Creative Commons and the co-founder of the Center for the Study of the Public Domain
Richard A. Epstein: A reprise on open source software
The recent response of Jamie Boyle encapsulates many but by no means all of the barrage of criticisms directed my way for my recent column, on the sustainability of open source software. The defenders of the system are able and articulate, and their conviction that this co-operative mode of social organisation is preferable to the more proprietary systems may well supply the strongest evidence that the system will survive and perhaps flourish. But there are still many issues that need to be sorted out. Here is my reaction to some of them.
The General Public Licence
In my earlier column, my sole concern was with the GPL, and not with other open source licences such as the Berkeley Software Development Licence, which allows for more liberal use of code that permits developers to keep their own contributions to the basic code confidential, so long as proper attribution is supplied. It should be evident that open source and the GPL are not synonymous. In speaking of the GPL, however, I quoted from section 2(b), which is more restrictive: “You must cause any work that you distribute or publish, that in whole or in part contains or is derived from the Program or any part thereof, to be licensed as a whole at no charge to all third parties under the terms of this License.” The provision itself reads as if any new work under that section that is made with open source code in whole or in part becomes open source itself. But the exact import of this provision is unclear because its substantive command is not paired with any remedial section that indicates what should be done in the event that the GPL licensee violates this provision.
In email correspondence, just about everyone has insisted that the only remedy they sought was the removal the offending code from the new distribution for it to remain in circulation. They did not claim that the new work became irrevocably open source. My own sense is that, standing alone, this remedy could be regarded as inadequate in that interim damages for the period of publication might be warranted, even if they would be difficult to calculate. I have little doubt that permanent damages in lieu of an injunction would not be sufficient because that would in effect allow forced buyouts that fly in the face of the licence’s basic provision. The permanent damages would preclude the use of the injunction. One German decision to which I have been referred allows the injunctive relief, but does not address all the issues raised by a license which is most notable for the lack of litigation it has spawned. No one thinks that this case resolves all the ambiguities associated with the licence.
“Many corporations both use open source in their middleware and contribute to the basic code. How will they fare in competition with companies that simply incorporate open source code in their own businesses without bearing the additional costs of keeping up the centre?”
Yet even if this approach is followed by most, the GPL has another odd feature, at least to these outside eyes. It does not specify who counts as the “licensor” in cases of sublicenses. To be sure, the agreement does stipulate in clause 4, “Any attempt otherwise to copy, modify, sublicense or distribute the Program is void, and will automatically terminate your rights under this License. However, parties who have received copies, or rights, from you under this License will not have their licenses terminated so long as such parties remain in full compliance.” The first half of this provision makes sense in principle, but again the want of any remedial method to enforce that termination is characteristic of the licence as a whole. In addition, there is a real question of whether the original licensor, or chain of licensors, can sue on the GPL to stop the use by a further sublicensee in the absence of any direct contractual relationship between them. That problem becomes only more acute if the person who receives the code does not have any licencing arrangement with the original violator. Copyright remedies may well be available under this situation, but the licence itself may not be self-sustaining.
These are only some of the complications that arise under this licence. My sense is that lawyers who have worked with it can find more. Let us stipulate that Jamie Boyle is right that the licence was crafted to the highest professional standards. But its extensive use could raise serious questions, for which future redrafting is not an easy option in light of the huge number of transactions that have already taken place under that licence. In general, any systematic effort to build a coherent whole through a network of contracts is subject to structural weaknesses greater than those in a standard ownership structure. Only time will tell whether this novel system of contracts will prove robust.
My second point in the article went to some of the features of the open source structure, which, as its defenders proudly note, differs from the usual form of proprietary system. In speaking of that I compared it with a workers commune (by which I did not mean to analogise it to some communist institution, but only a worker-owned co-operative established in accordance with the usual rules of property and contract) to point out the difficulties of trying to move a radically decentralised organisation along a common path. Here one point that illustrates the situation is the much-hyped SCO litigation, from which I have read the open “Dear Darl” letter of September 20, 2004 that was signed by Egan Orion on behalf of a diverse group of open source users.
Let me state right at the outset that I venture no position whatsoever on the merits of the suit, but only comment on this dispute insofar as it reveals something of organisational structure of the open source movement. Challenges of this sort, when directed toward a single company such as IBM, present no organisational difficulties. But such challenges can be extended to put the entire open source community on trial and thus require some form of collective response, of which the Orion letter is one. Once a matter of this magnitude gets to litigation, it raises a variety of tactical choices and funding obligations that it is difficult for a loose coalition of like-minded individuals, however committed, to co-ordinate. A proprietary company has a better handle on these issues. So that if this matter comes to litigation, it will provide a test of the durability of the open source business strategy.
This particular example gives rise to the larger question about open source, which is how long systems based on reciprocal altruism can survive in large-scale enterprises. Only time will answer this, but I was struck that most of the defenders of open source quickly moved to discuss matters that fall under the heading of the nature of man. The self-interest model that dominates traditional economic thought is said to be inadequate to explain the new development. Here I am of two minds on the basic question. I think that the standard views of rational choice theory, which postulate narrow and relentless forms of self-interest, are indeed wrong for a range of evolutionary reasons. Without some instinct toward co-operation, or the mutual respect of the rights of others, the entire project of human sociability falls to the ground. But by the same token, among strangers, the level of “confin’d generosity”, to use Hume’s felicitous phrase, may not be strong enough against the powerful tide of self-interest. Many corporations both use the open source in their middleware and contribute to the basic code. How will they fare in competition with companies that simply incorporate open source code in their own businesses (which is legal even under the GPL if there is no further distribution) to sell their own products without bearing the additional costs of keeping up the centre? Again, I do not know the answer. But the issue is one that strikes more traditional students of industrial organisation as somewhat odd.
The discussion also revealed a further rift inside the open source community. Many of its members were decidedly libertarian in orientation, noting how the system was organised in accordance with traditional principles of property and contract, which they think are sustainable because of the non-rivalrous use that multiple individuals can make of software. Some of them were even uneasy about the GPL, thinking that a “fork” that allowed commercial development off the public base could be sustainable as well. After all, that is the model on which much biomedical research takes place: public domain knowledge that leads to private innovation. Individuals on this wing of the movement had little use for grand denunciations of the system of private property. But others are of a more radical cast, and think that the unselfish nature of the open source community exposes the weakness of all propertied institutions, including the patent system generally. I think that the former is the correct view, and would hardly think that any success of the open source system should be treated as a reason to undo the entire patent system, for example.
The subsidy question
On one issue most of the comments that I received were silent. No one who defended the sustainability of the open source movement made any argument in favour of government subsidy or preferences. At this point, the policy issue is resolved. Libertarians like myself wish to have neither subsidies nor penalties for alternative forms of social organisation. This is in part a maxim of prudence: if you don’t know what the future brings, always follow rules that are neutral between alternatives. It is also in part an effort to curb the size of government: the state which has less discretion will attract fewer suitors, which will allow more activity in the marketplace, which is where these decisions should be made.