While it is already widely believed that “over-declaration” of standard essentiality is due to large and excessive numbers of patents being filed in patent offices and declared to Standard Setting Organizations (SSOs), my new quantitative research suggests that over-declaration is also being pursued with claims that individual patents read on multiple Technical Specifications.
| Some declare patents essential to multiple Technical Specifications |
More and more patents
However, with
the increasing use of patent counts as a measure of companies’ respective patent
strengths, for example in determining royalties, it is widely believed that
some technology developers puff up their positions with numerous declarations in
excess of what is reasonably required to protect their IPR, shield them from assertions
of anticompetitive behaviour such as patent ambush and provide the commitments required
by IPR policies. Over-declaration is thus commonly understood to be the filing and
declaring of large and increasing numbers of low quality patents that would
never be found essential in litigation. Accordingly, there has been an
exponential increase in patent declarations. Rapidly approaching 80,000 patent
families have been declared to ETSI including various communications
standards.
With
over-declaration, raw patent counts and checked-essential patent counts exaggerate
patent strength. There is no essentiality checking in standard setting, such as
by 3GPP or ETSI. While essentiality checking is undertaken by some specialist
firms, my
previous research shows that this does a poor job in correcting the inflated
relative positions of companies that over-declare. Systemic bias prevails because
essentiality checking is far from perfectly accurate. False positive determinations
(i.e. where a patent is found essential when it is not truly essential) tend to
exceed false negative determinations. And, the lower the true essentiality rate
(i.e. the percentage of declared patents that are truly essential), the more
bias there will be.
Throw everything at the wall and see what sticks
In addition
to inflating patent counts by flooding IPR databases with increasing numbers of
declared patents, my new research paper—based on patent
declaration and standards data collected and processed by Dolcera—indicates
that some companies are also declaring individual patents to multiple different
Technical Specifications. While most major declarers declare their patents to an
average of no more than 2.5 Technical Specifications, some companies declare essentiality
to more than twice as many, and with individual patents declared to as many as 12
or even 18 different Technical Specifications. However, essentiality is based
on whether a patent reads on any Technical Specification, not on how many of the
latter are referenced.
As human
and automated checks have to assess each declared patent’s essentiality against
every Technical Specification referenced, the more of those that are referenced
the higher the probability of false positive determinations while the
probability of false negative determinations cannot increase even to partially offset
the above. Assessing any additional Technical Specification can, therefore,
only increase the possibility that a patent is found essential. This means that
the systemic bias inflating essentiality rates found in checking will be higher
than if declarations for each patent were more diligently focused on only one
or two Technical Specifications. Costs also increase with the expanded workload
in checking more Technical Specifications.
My full new research paper analysing patent essentiality declarations to multiple Technical Specifications can be downloaded here and from SSRN.
