Open Source Software and the "Private-Collective" Innovation Model
von Hippel & von Krogh 2003
These are notes I made interspersed with my responses. The headings follow those in the original paper.
History and Characteristics
The paper starts with an overview of the origin of the FSF/GPL, and the reasons why Raymond and Parens created 'open source' as a more business-friendly term. it is pointed out that Raymond sees OSS development happening where the programmer loves the code – ie it's not a normal job. It's about attracting interest. People ("users") who simply use the code a described as "free riders".
Two examples of OSS projects are given: Apache and Fetchmail. Both are infrastructure tools that support techies in their work. I wish there were more examples of end-user oriented products such as OpenOffice.org, CMSs, Personal media like WinAmp. Or even WordPress.
The bare sentence "A significant amount of software developed by commercial firms is being released under open source licenses as well" [p210] is all there is on the growing use of OSS as part of a business model by organisations -public or private sector.
Finally, OSS is seen as of interest to Organisational Science because it challenges conventional models of innovation and the process is open and easily studied. Build the data and the researchers will come…
OSS Projects – Exemplar of a "Private-Collective" Innovation Model
Two major models of innovation are distinguished:
- private investment (PI), with expectation of a return on investment, and incentives to prevent spillover
- collective action (CA) model to create a public good, eg science. Reputational motivations are more important.
Differences from PI model
Firstly, users rather than manufacturers drive the innovation. They do go on to acknowledge that manufactorers such as IBM can still make money out of their investment if it improves sales of their products and services. Users' incentives are seen as improved profits.
Secondly, free revelation of the s/w leads to loss of proprietary rights, and costs of diffusion. Makes sense if there are "low loses from free revealing if they have low rivalry with potential adopters… and/or expect gains from the increased diffusion…" [p214] – ie participation in OSS makes sense at the point where the code is not core to the business model. To me that must be one of the reasons why OSS in the commercial world happened at the level of Apache, Linux and Fetchmail – but the paper does not go on to draw that conclusion.
Since (direct) costs of dissemination are low and many OSS programmers are students (ie with economic motivation to share), it is argued that it does not make sense for organisations that have developed the same code to keep it private.
Differences from CA model
It is argued that OSS projects do not conform to the public good frameworks established in the lituratare, and that still they work because of strong private incentives for the developers. The focus of the discussions is on the individuals' reasons – though I like the idea that participation in a project inherently generates private benefits, through the learning process and other psychological motivators. I guess it's important even for geeks to feel they belong to something.
I would imagine that organisations benefit by participating get to have a say in steering the direction of development – but there is nothing in this paper on the organisational motivators for participation in OSS projects.
In their discussion I think that vH&vG have missed a key point about OSS communities, namely, it might be free for a person to in on a mailing list, but any well run OSS project will be very careful about who they let at the source code. In most projects, you have to prove yourself through commitment to the community (answering user questions) and submission of patches for approval. That is, there is a highish barrier to entry, and even if the mailinglist subscribers might change, the core community is not ephemeral at all. For instance, the Drupal community (ie Dries) is very jealous of the committer priviledges. This paper conflates the two communities.
That is why free riding is not an issue in OSS, in my opinion, and why in their words "contributers actually regard free riders as an asset". They are not free riders at all: they are users. Users are good.
There is some discussion on the nature of leadership and management issues – but it's all from the point of view of motivation of individual participants.
The paper concludes by summarising the functionality and features of sourceforge-type OSS communities. That's not the focus of my interest (and I've done it for real), so I'll just note that again the focus of the proposed research is the surface: patterns of conversions and comments in the code. The nature of the application is not considered in the article at all – there is an assumption that there is a limitless pool of potential developers, and enticing them into a particular OSS project is purely a matter of social organisation and incentives.
I need to find something that looks drives the direction of development (what features to add?) and funding.
The discussion of OSS project leadership models compliments the dictator/democracy models in Karl Fogel's Producing Open Source Software: How to Run a Successful Free Software Project (2005) . As he says in the preface:
"Running a free software project is not exactly like running a business (imagine having to constantly negotiate the nature of your product with a group of volunteers, most of whom you've never met!). Nor, for various reasons, is it exactly like running a traditional non-profit organization, nor a government. It has similarities to all these things, but I have slowly come to the conclusion that free software is sui generis."
But that's for a different blog entry