Project outline

Final report

Press coverage

related research




Open Source Software Production - The Magic Cauldron?
Margit Osterloh, Professor of Business Administration and Organization Theory, Institute for Research in Business Administration, University of Zürich, Switzerland

Open source software production is a very successful innovation model. It is characterized by a dispersed knowledge production in a virtual network without intellectual property rights. Nobody can be excluded from the use of the source code. Thus the source code represents a public good in the classical sense.

The question arises why programmers contribute voluntarily and mostly without monetary compensation to the public good. The aim of the paper is to understand under which conditions this „magic cauldron“- model works.

In the second section of the paper we give a short picture of the characteristics of open source software. In the third section we distinct different groups of contributors according to their motivation to join the software production. We discuss whether these types are complementary to each other or not. In the fourth section we question under what conditions the open source model can be successful in general. We isolate situational, individual and institutional factors and compare them to the design principles Ostrom (1990) found empirically for successfully governing the commons. We identify factors Ostrom (1990) does not consider. We show how these factors make communities of innovation work without central authorities and strong property rights, even when there exist no clear group and resource boundaries. The conclusion is drawn that the success of the „magic cauldon“- model is due to an interplay of

  • institutional factors (nondistribution constraints, high degree of self governance and participation in rule providing)
  • motivational factors (balance between intrinsically and extrinsically motivated programmers)
  • situational factors (sequential and complementary innovation, low costs of monitoring and sanctioning rule breakers).

Drawing on these results we ask whether these conditions might reveal some important lessons for the distributed knowledge production in other new organizational forms.