The Politics of Open Source

As a vehicle for passionate participation and, transparent management, free and open source software (FOSS) has become an icon of our time: a synonym for a happier, more productive and democratic way of producing things. But sometimes the progressive image of open source may skate past some of the messier realities. Developing software code is, after all, a different challenge than journalism or music. And hackers are a different kind of social cohort than writers and musicians, let alone the general public.

The University of Massachusetts Amherst just completed a two-day conference, The Politics of Open Source, which assessed the impact of open source software in a variety of domains. The dominant perspective was political science, with accents of tech talk and activism.

One of the most interesting was David Karpf’s What Can Wikipedia Tell Us About Open Source Politics? Karpf, a research associate at Brown University’s Taubman Center for Public Policy, examined the potential and limitations of Wikipedia as an archetype of "open source politics." This is the realm of self-organized, self-governed networked platforms by which a community manages a shared resource.

Karpf’s first observation was simple, but worth re-stating: the problem with public goods on the Internet is not the "tragedy of the commons" problem — i.e., "free riders" — but rather, how to coordinate a large mass of people to work constructively together. Even today many social scientists continue to see the chief problem with public goods as a problem of how to stop free riders.

But as Karpf pointed out, this problem is essentially moot in an environment of "low transaction costs" (virtually free communication and exchange of content) and abundant information. "There are plenty of criticisms of Wikipedia," said Karpf, "but no one has ever said that the problem is that 'there just isn’t enough of it.’"

A more interesting point raised by Karpf is the structure and scope of online communities. He said that the boundaries of participation are typically governed by what is often called a "power law topology." In layman’s language, this essentially means that as an online community grows in size, and more people decide to participate in it, the website becomes a "power law hub" dominated by a relatively small core of people who do the bulk of the work.

This pattern is remarkably constant in the networked environment. Karpf noted that even though Wikipedia is the sixth-most visited site on the entire Web, there are only 75,000 wikipedians who are "active" — meaning, they contribute five or more edits per month — and only 7,500 "very active" wikipedians.

The elitism of a power-law hub differs from the elitism of "old media" in the sense that a digital community is more open to new participants and more accountable to them. The real challenge for such Web communities is how to manage themselves over time and whether they can surmount distinct challenges at various stages of their development.

Another great presentation was M.I.T. professor Eric von Hippel’s presentation about how users are becoming more powerful and reliable innovators than commercial producers. Von Hippel’s landmark book, Democratizing Innovation, is an empirical challenge to the conventional idea that corporate manufacturers are the source of most innovation. After studying dozens of product categories and doing first-hand research, von Hippel documented that user-centered innovation is in fact the norm.

What happens is that an individual discovers a need and jerryrigs his own solution. For example, a farmer who wanted to water his fields more efficiently rigged together some old farm equipment to make a "center-pivot irrigation system," which draws water from a well and waters the adjacent field through a pipe that moves in a circle.

There are many other familiar products that were in fact developed by users. They include Gatorade, the sports drink; the mountain bike; the sports bra; chocolate milk; electronic mail; protein-based shampoo; and Internet-access in hotels. As these user-made innovations became popular, manufacturers jumped in to commercialize them and improve them further — and to make money.

The prevalence of user innovation is usually overlooked because users are creating improvised innovations, not marketable products. The former are simply meant to serve users needs. The latter are meant to look attractive and make money for the seller. Once the commercial product becomes familiar, the history of its origins as a user innovation is usually forgotten.

Von Hippel believes that user innovation is displacing a great deal of manufacturer innovation these days. In one study, for example, it was estimated that 80% of the U.K. population had modified or developed a product over the past three years — a whole realm of innovation that companies and business schools don’t even study.

Companies would do well to recognize and leverage the power of user innovations, von Hippel advised. Just look at Lego’s Mindstorms product — a robotic toy that users can "hack" on to make their own creative variations. Since releasing Mindstorms, Lego has positioned itself as the Web host of user design innovations, inviting people to share their own daffy and brilliant ideas. Lego makes kits of the most popular designs, relieving itself of having to develop new products itself.

Von Hippel concludes: "Given modularity, heterogeneous users innovating independently and freely revealing [their work] can produce more and better design work that is collectively available than can individual producers who each protect their private innovations."

There were other presentations suggesting the value of open-source methods for improving an given area of life.

Two Indian authors — Shishir Kumar Jha and Amrutaunshu N. Nerurkar — examined how open source techniques are improving biomedical research. One of the key failures of the pharmaceutical industry is its failure to undertake research for new drugs for which the market is small or poor. In Africa, for example, there is a huge need for drugs to treat sleeping sickness, dengue fever and leishmaniasis. But under the patent-based regime for developing new drugs, a company will take on the R&D for a drug only if the resulting product will cover the costs.

A key innovation these days is to use a decentralized Web community to do collaborative research and share their findings with others. The Tropical Disease Initiative (TDI) is attempting to do this. It presents potential volunteers with an array of possible projects that they can participate in, and then lets them share software programs, databases and other tools to try to develop new drug regimens.

There is also an Open Source Drug Discovery project seeking to aggregate the biological and genetic information that scientists need to research new treatments for neglected tropical diseases. Large problems are broken down into simpler, smaller sets of activities with defined goals.

There were other terrific papers presented at the Politics of Open Source gathering — on the use of open source tools and the "open source" ideology in the Howard Dean presidential campaign, and on the failed attempt by the State of Massachusetts to require its government offices to use Open Document Format standards for all government documents, rather than Microsoft’s proprietary standards.

There was also a fascinating paper on the "insurgent experts" in Brazil who embraced free and open source software as a way to combine leftist politics, an anti-neoliberalism agenda, and efforts to support domestic development.

For the text of these papers and video of author presentations, go to the conference website.