Lecture Series: How Open Source Software is Revolutionizing Global Markets…And Communities

About the Speaker: Steven Weber is associate professor of political science at the University of California, Berkeley; and an associate with the Berkeley Roundtable on the International Economy (BRIE). His areas of special interest include international political economy, political and social change in the new economy, and the political economy of globalization and European integration. His publications include Cooperation and Discord in U.S.-Soviet Arms Control, the edited book Globalization and The European Political Economy, and numerous articles and chapters in the areas of U.S. foreign policy, political economy of the post-Cold War world, and European integration. His current research focuses on the political economy of knowledge-based industries and open-source software models (The Success of Open Source will be published in 2002). He is a consultant with Global Business Network in Emeryville, California where he works with global firms and governments on the political-economy consequences of information and communications technologies.

Rapporteur's Report

On May 24, 2001, the Project on the Information Revolution and World Politics presented a lecture by Steven Weber, entitled "How Open Source Software is Revolutionizing Global Markets… And Communities." In contrast to the proprietary software produced by most commercial manufacturers, open source software is written and perfected by volunteers, who freely share the programming code that would otherwise be kept secret. Under the terms of the most popular open source licenses, anyone can redistribute the software without paying fees to its author, and anyone can modify it if they distribute the new version under the original terms—open source and non-proprietary. In recent years, open source products such as the operating system Linux have emerged as significant competitors to proprietary software products like Microsoft’s Windows.

In his welcoming remarks, project director William Drake explained that the rise of open source software and the challenge it poses to proprietary software producers is just the latest example of a long-standing tension between open and closed systems in the configuration of the global information infrastructure. Since the configuration of the information infrastructure has a strong impact on both the global economy and world politics, the rise of open source software is something that is best seen in a broader light, giving consideration to implications beyond the software industry. Weber concurred that the open source software movement has broad implications: "I think that this phenomenon of open source software actually has quite a lot to teach us about some very general things…about the political economy of the Internet era, and actually, just as important, about the social process by which we are coming to understand it and the kind of policy process that is going to grow up around it."

Weber first addressed open source software as a market phenomenon, stating some basic facts and seeking to clarify some misconceptions that have emerged in recent treatment of the issue. First, he noted that the open source software movement, per se, has barely begun. The method of collaboration involved in producing open source software dates back to the 1950s and 1960s, but its manifestation in the market today, and the legal and policy issues that accompany it, are both new. Second, recent discussion of open source has combined elements of truth with a lot of hype and misinformation. Open source does not represent a "revolution in capitalism" or anything of the sort, and with proclamations like these it is easy to lose sight of what really is unique about the movement. To make matters worse, Weber argued, open source is sometimes actively misrepresented by stakeholders who feel threatened by its rise. One widespread misconception that has resulted is that open source software is inherently "free" software—a commodity that must be given away without charge and cannot be leveraged into a successful business model. Open source software is free in the sense of freedom to view and modify its source code—not in the sense of zero price. For-profit companies such as VA Linux and Red Hat have found ways to sell open source software, e.g. by adding value in the form of customer service and support.

While proprietary software still dominates the market for personal computers (PCs), open source products are used widely on the servers that power the Internet—a development with profound implications for the software industry and Internet economy. As of July 2000, 38% of Web servers were running Linux as an operating system, while only 21% ran Microsoft products. Market share for the open-source Apache server software is substantially higher than for any competing product, and Apache’s share is the only one that has grown over the past three years. "This is a big deal," argued Weber. "The fact that Linux [may] not run on your PC is not important… The desktop is like the steering wheel to your car, not the engine. The engine is the Internet, and increasingly it is built around open source software."

Beyond its importance for the software industry and Internet economy, open source software is an interesting social phenomenon with broader economic implications. Two questions are particular puzzling for social scientists. The first concerns the micromotivations of individuals in the open source software community—why do programmers contribute when they obtain no direct remuneration? Open source software presents a collective action problem, since there are no clear individual incentives for participation. Part of the answer is cultural, argued Weber—sharing is a cultural norm in the software community, and software engineers see themselves as artists who are proud to show others the creative genius of their programming work.

But Weber also maintained that there are real economic incentives for participating in open source software development. A reputation as a good programmer has great value for employees of commercial software firms, who may seek to change jobs in the future—yet the product of their paid work is only visible to a few other individuals at their current employer, and they receive no public credit for their efforts. In contrast, the credits file of a piece of open source software lists who contributed what to the final product, and potential employers can examine the source code to judge the contribution for themselves. "Writing open source code is probably the most efficient way to establish a reputation as a great programmer," argued Weber. "And a reputation as a great programmer is something that you can make money on in other settings." In fact, this reputation-enhancing incentive lends an element of self-selection to the open source process, resulting in a better end product. The best programmers actively seek to showcase their work, while mediocre programmers would not want to bear their sub-par code to public scrutiny.

A second interesting question about the development of open source software concerns the macrofoundations of the movement—how does the group manage to coordinate the efforts of as many as 7000 contributors? First, argued Weber, the development of open source software enjoys positive network externalities—it benefits from the participation of greater and greater numbers of individuals, even they do not directly contribute code to include in future versions. "Every open source ‘free rider’ now becomes at minimum a beta tester, who can find bugs, identify a new feature, and otherwise contribute to the collective good," said Weber. The more copied and widely used a piece of open source software, the more valuable it becomes. Second, open source programmers divide a piece of software into a number of small, self-contained modules, each of which can be developed and perfected without knowing precisely how other modules function. Many contributors can thus work on different aspects of the project in parallel, minimizing the number of colleagues with whom each programmer must coordinate. Finally, the open source movement may have the outside appearance of an anarchic community, but in reality it includes formal decision-making structures and a set of shared norms to coordinate efforts and govern interaction.

While the open-source method is now actively revolutionizing the software industry, the mechanics of open-source production have potentially much broader implications for the economy as a whole. The key to seeing beyond the software industry is to think of open source as a production process, where the software is simply a side-effect. Weber offered four observations as to what this process tells us about the broader economy. First, open source is yet another demonstration of how the Internet facilitates geographically widespread collaboration, reducing the costs of communication across great distances. Second, the movement has shown the value of distributed innovation and offers a new way to think about the division of labor in other industries. Third, open source software offers new ideas as to how to combine communities and commerce. Weber noted that the open software community is a relatively unique example of an open, value-driven community where participation is directed towards the creation of a significant product that travels outside the community itself. The model which has succeeded here may have potential in other areas of the economy.

Finally, Weber suggested that "open source software is going to be a really significant factor in the international economy in the 21st century." The production of Linux has involved programmers from around the globe, many of them from developing countries that have security or economic reasons to avoid using proprietary software. "I think you can spin out an interesting story where open source software turns out to be a powerful instrument of development bootstrapping," he said. "Open source software shifts the decision making prerogative into the hands of people in developing countries."

Following Weber’s presentation, several members of the audience posed questions for his consideration. Project director William Drake began by noting that Microsoft’s reaction to the open source software movement bears resemblance to the reaction of national telecommunications carriers when the Internet began to emerge and threaten their entrenched positions. As in this earlier struggle, stakeholders have a lot of influence with policy makers, and they have the potential to leverage this influence into decisions that would be detrimental to the open source software movement. What cards do the opponents of open source hold in this battle, asked Drake, and what can be done about it? In response, Weber acknowledged that Microsoft poses a formidable challenge to the open source software movement. The company’s first move was to deny that open source posed any real threat to its interests, but it has since shifted to a two-pronged strategy: spreading fear, uncertainty, and deception about open source; and embracing and extending the concept to render it less threatening. This latter tactic is epitomized by Microsoft’s shared source initiative, in which it proposes to share its code with other large software companies but not allow them to modify it or distribute it further. Through use of these tactics, argued Weber, it would be easy for Microsoft to sow a lot of confusion about open source and get the policy community on its side. If it succeeded in changing copyright law or using patents to prevent reverse engineering of software products, for instance, this could have dire consequences for the open source movement.

Several other audience members asked questions about the security of open source versus proprietary software. Weber expressed confidence that open source software would be more secure than proprietary software because of the greater number of programmers working to expose and correct security flaws. A security problem in software is like a bug, he argued; it will be more transparent in open source software and therefore fixed more quickly. In response to a suggestion that open source programmers could intentionally build back doors (areas of security weakness that the programmers could exploit) into the modules they were designing, Weber argued that once open source software exists in a corporate setting (e.g. Red Hat or VA Linux) there is greater incentive to control the security of the whole product.

Several questions dealt with the specifics of open source software as a business model. One person noted that open source software is currently popular only among more technically-adept computer users, and he asked how well it would fare among those that are less tech-savvy. Weber responded that this problem is slowly being solved in the marketplace by companies like Red Hat, which build user-friendly interfaces and provide customer service for open source software. Furthermore, he argued, all end users will appreciate the value of software that has fewer bugs and is less likely to crash. Another audience member asked how members of the open source community (who write and perfect the software without compensation) feel about other companies making money off of their product. Weber noted that everything sold by companies like Red Hat is itself open source software. They are thus abiding by the ethic of the community, which stresses access to the source code but does not prohibit making money.

Finally, several audience members asked about the potential for further applicability of the open source production process. One person noted that open source production has not spread to some areas of the computer industry, such as computer games. Weber acknowledged that open source development may be more difficult in this area. As another example, he mentioned graphical user interfaces (GUIs), which make software more user-friendly for those without technical expertise. Open source has not helped to develop these, he noted, because programmers writing the software do not need them. A second audience member asked whether the open source production model could extend beyond information industries. Weber acknowledged that open source development would be more difficult in industries that have to move around physical goods, rather than information, as part of the production process. But he emphasized that open source development does have widespread potential in non-software information industries. In particular, he emphasized, the problem of distributed knowledge management in the software industry is very similar to that which exists in the medical community. As with software engineers, each physician is an expert in some particular area, and the open source method might offer a more efficient way of organizing and sharing knowledge to solve complex health problems.

Finally, one participant asked about the potential of open source software production in the developing world. At some point, she noted, you have to go beyond operating systems and develop applications; for this task people may have to be paid. In the developing world, software export is driven by an army of paid programmers. How well will open source work in this environment? Weber noted that companies in the developing world can build proprietary applications for open source operating systems; this would be an effective model in developing countries that prefer open source operating systems for economic or security reasons. Many developing countries now train students with open source software because it is affordable and it allows students to understand how software works on the inside. Once they understand an open source operating system, argued Weber, these students can easily build proprietary applications to run on top of it.

Report prepared by Taylor Boas