The battle between Microsoft Windows and Linux has attracted increasing attention over the years. Scholars and journalists have repeatedly analyzed the virtues and drawbacks of open-source communities, the different ways they coordinate scientists who produce new knowledge and the two systems' different incentives for innovation. This has led to extensive debates on which of the two operating systems will become dominant. This well-known example encapsulates the most relevant tensions between public and proprietary domains in the production of new knowledge.

What drives people to produce new knowledge? What drives scientists to invent new devices and scholars to develop new theories? What motivates those developing software and databases for academic research? The prospect of an economic pay-off surely constitutes an essential factor, but there is more than that at stake. Curiosity, the quest for fame and reputation, and the pursuit of promotion are equally relevant motivators and economic rewards are often simply a means to collect funds to invest in further research.

When knowledge is produced, two regimes can be observed. The first assigns the producer specific property rights that exclude others from using the newly created knowledge. An obvious example is the patent system, which guarantees a temporary monopoly to the inventor. The second regime, the so-called “Republic of Science”, is the realm of pure scientific research in which knowledge is freely shared with other researchers.

The institutional foundations of open-source science emerged in the late sixteenth and early seventeenth century in Europe when intellectual endeavors began to receive support from the State instead of through patronage from elites. The logic of open-source science assumes that the production of knowledge is a public endeavor characterized by norms that encourage universal standards based on competence or merit. Its practice demands full disclosure of methods and findings in order to allow scientists to verify and replicate each others' work. While the goal of commercial investment in science is to increase the profits that accrue from rights to private knowledge, the goal of open-source science is to add to the stock of public knowledge. Instead of property rights, scientists are granted priority for discoveries made, gaining reputation and promotion. This provides scientists with an incentive to share their discoveries early to avoid duplication of findings and advance the field more rapidly.

One question that arises is how these two regimes can coexist and whether one should be expected to prevail over the other. The solution - according to the model developed in the paper - lies in researchers’ individual preferences. Researchers may seek to secure property rights and restrict diffusion of their findings or they may prefer to put the outcome of their research efforts in the public domain. While in the private domain researchers enjoy a higher monetary compensation, in the public domain the benefits are twofold. First, scientists derive enjoyment from the work itself; second, they enjoy the satisfaction of actively contributing to increase the level of public knowledge. Researchers will decide to operate in the public domain if the benefits of contributing to the public knowledge are greater than the monetary gain they will forego by renouncing the proprietary regime.

The authors’ model shows how the proprietary and the public regimes can actually coexist and how researchers are split between the two regimes. The proportion of researchers who join the proprietary regime is larger when profit opportunities are stronger, and is smaller when the lure of the research itself is stronger.

That the two models can coexist is consistent with what can be observed in the real world. Over the last few years, the United States has witnessed a significant increase in the number of patent applications filed by universities and public research institutes. Conversely, the industrial giant IBM has recently announced that it would make 500 of its patents freely available, and that all of its future patent contributions to the “standards group for electronic commerce on the web” would be free.

The model also shows that the only way to get the two regimes to coexist is to coordinate the individuals who operate under the open-source sharing rules. If this is not done, researchers have a strong incentive to leave the open community and join the proprietary regime. The authors show that the Generalized Public License (GPL) arrangement constitutes an effective mechanism to achieve the required coordination. With GPL, each producer of an open-source program requires that all the subsequent modifications and public contributions to the program are subject to the same rules of openness. By committing to GPL, the originator of an open-source project may attract a large number of both contributors and users.

Coordination of members is difficult, but this not the only drawback of open-source communities. Frequently, newly created knowledge requires further investments to create economically useful technologies or to commercialize products. In the case of software, additional investments to enhance its usability and to produce documentation and support are necessary. In order to sustain these downstream investments, some profits are needed. Open communities lack this incentive and may therefore be unable to carry out the required investments.

One possible solution to this problem is to confer property rights on the innovations of downstream producers while keeping upstream knowledge under an open regime. In this way, the two regimes can not only coexist, but can also complement each other in different steps of value creation.