patent laws need to be looked at more closely, and not just for
drugs. There is sometimes need to provide incentives to some
innovators, through the rewards of a temporary statutory monopoly, to
offset risky and costly research. There is sometimes a need to
encourage innovators to disclose their inventions so they become
available at the end of the term of the monopoly. Sometimes this
monopoly is not only not needed to stimulate innovation or
disclosure, but can also get in the way of innovation.
Brian Masse, NDP critic for Industry,
Science and Technology, wrote in "Patent
Protection - Who should it serve?"
The drug patent
laws have attempted, through policy and regulations, to find a
balance between the needs to reward innovators who experience lengthy
and costly research and development periods and on the other side, to
allow for strong generic competition after a 20 year patent period to
ease the price burden on Canadians and their health care system.
The problem I see is that current patent
policy treats all innovation as similar when in fact the market
conditions for different subject matter currently granted patents by
CIPO (Canadian Intellectual Property Office) are extremely different.
The market dynamic is very different in pharmaceuticals (composition
of matter), the manufacture of consumer electronics (manufacturing
process), or computer software (information process).
Last year I was hired by the ICT branch
of Industry Canada to do a A
Review of Software Patent Issues.
The two major themes of this report were:
We should return to first
principles and question any assumptions. Patent policy should be
evaluated as economic public policy, not something that should be
left to legal analysis (within government or within business)
Each type of subject matter should
have independent economic analysis recognizing differences. Broadly,
software (information processes) is very different from hardware
The problem is getting quite critical in
software. Software and business model patents (broadly known as
information process patents) are of extremely poor quality, with most
software developers assuming that if they author software of any
substantial size that they will infringe patents that they are
unaware of. Researchers who have done independent analysis have
suggested that between 60% to 95% of information process patents
granted by the United States Patent and Trademark Office (USPTO)
would be found invalid if adequate tests of useful, novel and
unobvious were done.
Patent offices seem to go under the
assumption that patents filed, if they are not obviously invalid,
should be granted. Poor quality patents would then be sorted out in
the courts when an infringement case is brought before them. This
method may work well in pharmaceuticals where both the
innovator/brand and imitator/generics are large pharmaceutical
companies, but this does not at all work in information processes. In
information processes the players can range from teenagers using
their parents computer to implement an innovative concept in
software, or the largest transnational software companies. Small
innovators, arguably where most of the innovation has come from in
the software field, are locked out of this system.
While my main goal at this time is to
encourage the government to do subject matter independent economic
analysis, I believe that when the analysis is done that governments
will agree that information processes should be excluded from patent
policy. This will not be suggesting that some innovations in this
subject matter may not happen without the added incentive of the
temporary monopoly, but that from a public policy standpoint that we
would be better off without those few innovations that require the
monopoly than the larger amount of innovation that is stifled by the
monopoly. Many studies suggest that in software that it is
competition that is the major driver of innovation, not patent policy
which deliberately creates temporary monopolies to stop competition.
My major area of policy interest is
and Open Source Software, what
is called "commons-based
peer production" when
extended outside of the production of software. Patent policy being
extended recently to include information processes is entirely
incompatible with this new mode of economic production. This mode of
production relies on liberal copyright licensing to allow the creator
to add their work to the commons, and to authorized the creation and
publication of derivative works by peers. Patents on information
processes in the commons effectively removes all possible expressions
of these processes from the commons, forming a harsh anti-commons
that is chilling participation in this extremely innovative
Open source advocate Bruce Perens, in an
by BBC technology correspondent Clark Boyd,
said that "we're looking at a future where only the very
largest companies will be able to implement software, and it will
technically be illegal for other people to do so."
production is not only incompatible with the temporary monopolies
produced by patents, but is a demonstration against the assumed
necessity of patent policy in information processes. Peer production
demonstrates a model of production where improvements are made in
smaller incremental innovative steps, largely removing the necessity
for large up-front development costs. This advanced methodology also
relies on knowledge being publicly published into the commons, and
provides its own non-patent incentive to publish innovations.
The software-side of the Internet is
co-dependent with commons-based peer production methodologies. Peer
production methodologies need good worldwide communications
facilities which the Internet offers. Most of the software that forms
the infrastructure of the Internet is either FLOSS or a FLOSS
derivatives, and much of the most productive aspects of the Internet
turn out to be forms of peer production. The two leading leading
browsers are Internet Explorer which is a derivative of the FLOSS
Mosaic project, and Mozilla which is a FLOSS project. The vast
majority of Internet server software is FLOSS with the Apache
web-server having a 67% market share. The BIND project (domain name
server) and Sendmail (email server) have similar numbers.
It would be hard for anyone to argue
that the Internet and methodologies that are co-dependent with it are
not innovative. If we are to receive the benefits of these new
production methodologies we need to re-think some of the existing
regulations and assumptions.
I am available to any parliamentarian,
policy makers or citizen who wishes to discuss this area of policy
and work towards the required changes.
- Russell McOrmond, private sector
co-coordinator of the GOSLING
Community (Getting Open Source
Logic INto Governments)
Copyright 2004, Russell McOrmond
This work is licensed under a Creative
Editor's note: This article was first published at flora.ca. The author may update it from time to time, so please look at the original for the "definitive" version, especially if you plan to quote it elsewhere.