Chapter 22. Patterns of Governance in Open Source

Steven Weber

The hardest problem facing a political community is how to increase the probability that the whole will be greater than, not less than, the sum of its parts. People join together voluntarily to solve problems because they believe that the group can do things that an individual cannot. They also believe (in some abstract sense and often implicitly) that the costs of organizing the group and holding it together will be smaller than the benefits the group gains. If you put aside for the moment the affective and emotional needs that individuals satisfy in groups and focus instead on the part of politics that is about problem solving, the bet that people make when they enter a political community is simply that "none of us is as smart as all of us"maybe not on any particular issue or at any particular moment, but on the vast set of problems that human beings confront and try to manage over time.

It doesn't have to work out that way. Everyone has been part of a community or a company where the whole is less smart than the individuals who comprise it. Political systems often seem to suffocate under their own organizational costsnot just national governments, but smaller systems like city councils and co-op boards. And even if a community does create net benefits for at least some segment of the group, the distribution of those benefits can be so grossly unequal that most of the community members would be better off on their own. Get the balance wrong, and you can easily create situations where no one is as dumb as all of us.

These are very old problems confronting political thinkers. The rise of the Internet adds a small but significant twist by making it much easier to discover potential collaborators and pull together "affinity groups," networks of subcontractors, outsourced component makers for production systems, and the likeall of which are political communities that aim to solve some kind of problem. The core idea is joint production at a distance, the opening up of a universe of collaborative projects in which physically separated individuals contribute to the creation and refinement of a solution. Because the opportunities for creating collaborative communities have been expanded greatly by Internet technology, the boundaries and borders of existing communities are open for redefinition, and the possibility for new communities seems vast.

Which means that the stakes are high for getting it "right," or at least getting it right enough. This, I believe, is where some of the most important lessons of open source collaboration are likely to emerge. This chapter poses the question this way: if patterns of collaboration within open source communities were to become surprisingly pervasive, or pervasive in surprising places, what would this suggest about institutional design for communities of knowledge and practice in politics, outside of the realm of software or even technology per se?

To answer that question takes at least four steps. I first bound the question by limiting the argument to a class of problems most likely susceptible to open source-style principles. I then describe more precisely some of the theoretical issues at stake in group problem solving. The third section of the chapter lays out seven design issues that follow from the experience of what works (and does not work) within open source communities. The final section suggests some actionable implications. If we view the politics of problem solving through this kind of prism, what might or should we do differently?