IPcentral Academic Advisory Council member, Professor Ronald Mann, has written a groundbreaking paper on the implications of open source software for the landscape of innovation. Commercializing Open Source Software: Do Property Rights Still Matter? (March 2006). ExpressO Preprint Series, Working Paper 1042. Mann analyzes how open source business models are affected by, and influence, proprietary business models. He makes several arguments that seem non-intuitive until their simplicity shows that this paper contains fundemental insights; that open source will continue to rely on investments by "IP firms," push the technology industries towards more concentration and raise entry barriers for new firms.

Citing the research of Petra Moser (patents will affect where innovation occurs) and Ashish Arora (that stronger IP systems often facilitate industries marked by small and specialized firms), Mann argues that patents may have helped the fragmentation of innovation in the last decade of the 20th century, making possible opportunities for small firms to profit from solving complex and highly specialized technological problems. With the deconcentration of industry, and increasing range of entities undertaking innovating activty, Mann posits that patents helped small firms interact with larger firms, and remain competitive. The result is that today, the technology industries are less concentrated, and with healthier levels of entry than other industries.

Establishing that intellectual property rights can affect the structure of industry and innovation, Mann then integrates open source into the picture.

Mann argues that the growth of open source business will continue to rely on integration into traditional commercial value chains of large corporations that appropriate returns from other parts of the chain through proprietary means of return. Open source can be viewed primarily as a means of firms commodotizing industrial markets where they want to reduce the price of inputs, or markets in which they are not competitive. Big money corporate investments in open source makes sense in this context even if the open source revenue stream does not suggest a net gain for the corporate adopter. Consequently, open source has been “unduly romanticized.” It does not change traditional business economics nor signify the displacement of profit motive with altruism on the part of firms like IBM, Intel and Sun.

…it is not clear that the models of the [OSDL] corporate participants would work without the internalization of R&D rights that patents facilitate. If much of the participation of proprietary firms in open source development is motivated by “value-chain” returns – i.e., firms accepting commoditization at one point of a value chain at the same time they attempt to stake out a profitable point of competency elsewhere on the set of products and services their customers use – then patents presumably will be just as important in the remaining core areas...

This is a stirring position. Most open source supporters are not fond of big corporations, yet the irony that open source businesses and open source customers are often large commercial entities with profit interests rarely occurs to them. Open source proponents even see themselves a revolutionary individualists at times, yet scaleability is simply a drawback of open source, making it less friendly towards individual consumers than pin-striped professionals in large enterprises

Mann makes another key point. Open source will likely result in increasing concentration in the software industry despite its low entry costs, distributed development process and licensing models that ensure free availability of source code. Mann argues that small firms will be unable to leverage open source business models for gain because of their limited ability to adequately control different parts of their value chain; those best posed to succeed with open source are comparatively large service like IBM and Red Hat. Smaller firms have limited ability to ensure continuing service support and assure customers against legal uncertainty.

Further, firms relying more on service rather than product based revenue streams are often at a disadvantage in attracting venture capital; suggesting that entry for open source firms will be less than the overall industry, that stand-alone open source business will tend to concentrate to a few large entities, and that large service firms with concurrent proprietary revenue streams and the ability to achieve economies of scale will be the primary backers of open source commercial activity.

Mann’s critique on the effects of open source on innovation extends to how industry entry levels and the viability of new product firms will be affected by the growth of open source businesses.

… a property rights system favors new entrants because large firms can use other tools related to their market power to continue to grow. Small firms have nowhere to turn except property rights… it is much easier for a small startup to pursue an idea to the point of having...patent(s)... sufficient to protect the idea from competitors than it is to develop the kind of brand identification and market power that would make it a strong competitor against the large incumbent firms in the industry…

… what open source does…is to remove from the market firms that are developing discrete products from which they wish to get revenues... a fuller vision of open source reveals a great interconnection, if not outright dependence, on proprietary property rights...

A comment on competition is important here. If open source makes some proprietary product firms irrelevant, then isn’t open source simply beating rivals with different development models, and shouldn’t it be patted on the back? In my book, yes. However, the real issue which Mann may or not imply, but which I believe is of utmost significance, is that if the spread of open source may have negative effects on industry entry levels, concentration and viability of product firms, then there is even less reason to weaken product firms by eliminating software patents under the belief that doing so will increase innovation. Mann writes: "it is open source, and not patents, that pose the biggest threat to the “polyarchic” decision structure under which the software industry has flourished for the last decade."

An interesting aspect of this paper is that it effectively casts open source as just another chapter in economics and business books, that developed a lot of rhetoric in a path dependent mission to spout misconceptions of the proprietary industry. Yet open source falls on its own swords by potentially causing the very harms to innovation it purports to resolve. The notions that open source is all about decentralization, non-existent barriers to entry, alternative incentives, or even the common prattle that open source is the future of the software industry- all arise primarily from the ideological rather than realistic bases for open source. In the end, the hype around open source simply proves its one of the most effective marketing campaign the technology industries have witnessed.