Everyone is talking about open source, even presenters at BIO 2004, the international biotechnology conference recently held in San Francisco. As the argument goes, if open-source software can be developed in centralized communities of interested parties less expensively than proprietary software, can a decentralized community of interested parties also develop less expensive drugs?
This sounds like a nice idea, but software and biotechnology are different, and not all open-source software generates communities of interested parties that would lower costs. Before biotech can reap the benefits of open-source development, several differences between the communities should be considered.
How Are Software and Biotech Different?
In a sense, biotechnology is already open. Experimental data needs to be published and the experiments duplicated by independent parties. If it is not replicable by independent parties, it is suspect. Some biotechnology is done in secret. Generally, though, the processes must be disclosed before downstream users of the technology can rely on it.
Take, for example, drug development. Before receiving FDA approval, a new drug’s method of manufacture, composition, interaction with the body and other considerations must be disclosed, even to the point where others could repeat any experiments on which the drug’s support is based.
Software is a little different. It can be distributed and used just fine in object code with the source kept secret. We can rely on the developers’ word — rightly or wrongly — that the object code really does what the package says it does. If a printer driver causes pages to print properly, we assume it is written well enough. Life forms are much more complicated. It is not a good idea to rely only on just the output of a biological process or the promise of its promoter. We need independent and open access to the “source” of the biotechnology development.
To be a truly open-source development, however, requires more than just openness to the source of a development. It requires an open-source license — otherwise a community of interested parties that could freely build upon the development could not exist.
In software development, open-source licenses are needed because of the nature of copyright. One cannot build on another’s software without the author’s consent. Authors have the right to control the creation of derivative works and, while there might be some dispute about the exact boundaries, building on someone else’s program code typically creates a derivative work. As a result, to get the community of developers working on distributed open-source development, they all need licenses to the underlying work, which an open-source license provides.
In biotechnology, essential developments are facts rather than expression, so copyrights are not typically implicated. Even in cases where the development involves data compiled from experiments, courts have held that mere facts are not protectable under copyright, regardless of how much effort it took to compile the facts.
How Are Software and Biotech Similar?
While they have many different characteristics, in some ways biotechnology and software are alike. Large capital investments are required to develop the technology, and that capital investment needs to be recovered, either from end users or elsewhere.
The investment can be recovered from end users by charging more for the product than the cost of production. A CD-ROM with software on it might cost no more than a few dollars per copy, but the product might need to be sold for hundreds of dollars per copy to recover the cost of development.
Drug development works the same way. A drug might only cost a few cents per dose to produce, but is sold for several dollars per dose to recover development costs from patients. Given the nature of the marketplace, absent legal constraints, costs and prices will converge. As a result, to recover capital investments from end users, legal constraints, such as copyrights and patents, are needed.
How Is Development Funded?
If capital recovery from end users is not needed, no legal constraints are necessary and price and cost can converge. For example, research done in government laboratories with public funding might be free for others to build upon. Software developed as part of government programs is often made available for the cost of copying.
Even some privately funded developments might not need recovery from end users if the investor obtains some side benefit. An embedded hardware manufacturer might pay a software company to port an operating system to its embedded hardware so a market for the company’s hardware will exist. Printer companies might pay for development of printer drivers and give away the software to sell printers.
Any time a development can be marketed with capital investment recovery from end users, proprietary vendors will have an advantage over open-source vendors. Even eager volunteers are no match for dedicated employees and a large marketing budget. Thus, one would expect that the user base for any open-source software would always evolve to the proprietary solution.
Preventing the Takeover of Open Development
Richard Stallman recognized this problem early on. To enable community-developed software to compete with proprietary software, it had to be something that could not be co-opted by proprietary vendors. The General Public License allows for unconstrained distribution, modification and use of software, but constrains against downstream constraints. We need the same sort of open-source licensing for biotechnology developments.
Some believe the GPL is insidious, but it is really a clever device. It allows for open-source software and proprietary software to remain on parallel tracks and compete using different models, giving the former the advantage of lower cost at the input and the latter the advantage of higher price at the output. It is not good enough to have open-source licensing without downstream constraints.
What is needed is some sort of downstream constraint to prevent the “proprietization” of biotechnology. Because copyrights are not implicated much in biotech, the downstream constraint should probably be based in patents. While obtaining a patent is much more expensive than obtaining a copyright, the cost of protection in either case would typically be only a small fraction of the cost of making the development in the first place.
More Than Buzz
As with software investors who are only interested in having the first copy, there might be biotechnology capital providers who are only interested in having the development exist and not worried about recovering their investment from end users.
For example, suppose the Lupus Foundation of America could fund the development of a cure for lupus. It really wouldn’t matter if the cure was made freely available to everyone, unless they partially funded the cure only to see a drug company take the development for free, formulate a drug, and sell it for large profits.
Open source might seem rooted in altruism and noble goals, but it doesn’t get very far if it cannot compete with moneyed marketing muscle. Even when the price is zero, it has to compete to attract investment in its development, in hard cash or in community interest.
Before we can take the buzz about open sourcing biotechnology seriously, biotech must recognize that what makes the open-source movement successful is more than just generating buzz.
Phil Albert, a LinuxInsider columnist, is a patent attorney and partner with the San Francisco office of the intellectual property law firm Townsend and Townsend and Crew LLP.