By Kevin E. Noonan —
On April 13th, the Department of Commerce released a White Paper entitled "Patent
Reform: Unleashing Innovation, Promoting Economic Growth & Producing High-Paying Jobs." All admirable goals, but unfortunately
the paper is more hortatory than informatory. Disappointingly, department economists, including Mark Doms,
the chief department economist, and Stuart Graham, Chief Economist of the U.S. Patent
and Trademark Office, offer precious little evidence or even data supporting
their assertions that the proposals they recommend will lead to these goals. (The first author of the report, Arti Rai, is the Administrator of the PTO's
Office of External Affairs and thus responsible for advocating for these
proposals, but realistically she can only work with what her economists provide
her.)
The Executive Summary sets forth some economic
statistics related to the goals contained in the Paper's title. According to this portion of the report,
75% of post World War II economic growth in America is "linked" to
technological innovation. This
statistic is extrapolated from the proportion of the average annual growth rate
(3.4%) due to capital investment and increased efficiency (2.5%). In addition, average compensation has increased from
1999-2007 at a rate 2.5-fold in "innovation-intensive" sectors (not
expressly defined) than in the economy as a whole. Not surprisingly, the Paper also states that 76% of venture
capital investors "consider" a company's patents when deciding to
fund "[h]ighly innovative firms." The consequences of delayed patent grant (purportedly due to
the backlog of 750,000 applications) has "substantial costs," which
the Paper estimates "could ultimately cost the U.S. economy billions of
dollars annually in 'foregone innovation'." The Executive Summary ends with the two proposed
solutions: giving the USPTO fee-setting authority, which the Paper states would "contribute
significantly" to a 40% reduction in patent pendency; and establishing an "enhanced
post-grant review" regime. Claims for the latter are even more dramatic: according to the Summary, "[t]he cost of such
proceedings is expected to be 50-100 times less expensive than litigation and
could yield $8 to $15 in consumer benefit for every $1 invested."
The Summary provokes the expectation that the rest
of the White Paper will disclose the economic and statistical underpinnings to
these claims (after all, in economics of all the sciences it seems reasonable
to expect to "see the work" underlying the conclusions). Unfortunately, that data is not
found in the Paper. The Discussion
following the Summary sets forth policy presumptions ("[s]timulating
economic growth and creating high-paying jobs are key priorities for the Obama
Administration"), a history of recent patent reform efforts, and the need
for the latter to achieve the former: "[p]atent reform legislation will
accelerate [a virtuous cycle of innovation, growth and additional innovation]
and speed the pace of growth and of job creation."
The Paper then cites academic sources relating
innovation/technological change and job growth, something akin to bringing
coals to Newcastle for an intended technology-savvy audience. (It is unlikely even any troglodytes in
Congress have failed to hear the message linking technology to Progress.) The 75% (2.5%/3.4%) statistic recited
in the Summary is supported by a 2007 paper by Jorgenson et al. ("Industry Origins of the American
Productivity Resurgence," 19(3) Economic Systems Research 229) and a 2000
National Bureau of Economic Research study (Boskin & Lau, "Generalized Solow-Neutral Technical Progress
and Postwar Economic Growth," Working Paper 8023). Innovation also produces high-paying jobs, the Paper
notes, citing Basu & Fernald (2009, "What
Do We Know (and Not Know) about Potential Output," 91(4) Federal Reserve Bank of St. Louis
Review 187), with 75% of the differences in industrial output between countries
being attributed to "innovation-driven productivity differentials,"
citing Hall & Jones, 1999, "Why Do
Some Countries Produce So Much More Output Than Others?" 114(1) Quarterly
Journal of Economics 83. Technology-related differences can be discerned between industries: while the Paper reports that "the
average rate of real compensation per employee" in the U.S. private sector
increased by 20.2% from 1997-2007 (supported by data in accompanying Table I),
in "the most innovative industries" (at least computers, electronics
and chemicals) the increase was more than 50% (a greater than 2.5-fold
difference).
New technologies are "disproportionally"
generated from venture capital-backed startup companies, and VCs rely on
patents when making investment decisions (assertions supported by Kortum &
Lerner, 1998, "Does Venture Capital Spur
Innovation?" National Bureau of Economic Research, Working Paper 6846; Mann &
Sager, 2007, "Patents, Venture Capital and
Software Startups," 36 Research Policy 193; and Graham et al., 2010, "High
Technology Entrepreneurs and the Patent System: Results of the 2008 Berkeley Patent Survey,"
forthcoming). (As has been noted by others, it is a
pity Dr. Graham's work is not yet available; it should be posted on the PTO
website.) The Paper also
cites certain "anecdotal" evidence supporting these "large-scale
empirical findings," ranging from medical device incubators to companies
acquired by big pharma companies. The Paper picks out the pharmaceutical industry as one dependent in
particular on "high-quality" patents, this being one of the reasons
why patents in the pharmaceutical industry do not support Bessen and Meurer's
conclusions that patents are inimical to innovation (2008, Patent Failure: How Judges, Bureaucrats and Lawyers Put Innovators At Risk). This produces a "patent premium" for biotechnology,
pharmaceutical, and medical device companies, according to the Paper (appropriately citing Arora et al., 2008, "R&D
and the Patent Premium," 26(5) International Journal of Industrial Organization 1153) for this
proposition. It should be heartening
for members of that sector to hear the Office understands the importance of
patents for these industries.
The Paper then identifies the problems it intends
to address: untimeliness in the
patent procurement process and inconsistent quality of granted patents. "Delay, uncertainty and poor
quality" are the principal characteristics of the current U.S. patent
system, according to the Paper, which "makes private investments in
innovation less likely," undermining economic growth and job
creation. The consequences of this
situation are not illustrated by academic studies but by anecdotal evidence,
using stories of companies that had not been able to obtain high-quality
patents in a timely fashion. The one large study cited, by London Economics (2010, "Economic Study on Patent Backlogs and a
System of Mutual Recognition"), was used as the basis for the
assertion that problems with patenting could account for "foregone
innovation" costing "billions of dollars annually" (albeit this
study was not confined to the U.S. but included the Japanese and European
Patent Offices, which are usually not thought of as suffering from the problems
of the U.S. Patent and Trademark Office addressed in the Paper).
For poor
patent quality (defined as "patents that are obvious, overly broad or
unclear in the inventive territory they cover"), Bessen and Meurer were
again cited, for the proposition that even low-quality patents can be "profitably
asserted against genuine innovators in litigation." (It would be a benefit to this debate
if examples were provided showing this morality play of "false"
versus "true" innovators;
while this may happen, it seems to be an overreaction to overhaul the
U.S. patent system without evidence that such behavior is widespread.) Such "bad patents" are
also asserted to constitute a "tax" on innovation, purportedly
because they are more likely to be licensed than challenged, both because of
the inherent costs of patent infringement litigation as well as the economic
behaviorialist position that a challenger is unwilling to bear the cost and
risk for an outcome in which all potential infringers will share (i.e.,
invalidating a bad patent). While
these scenarios are plausible on their face, without evidence they are just
that — mere scenarios.
Having discussed the problems the Paper provides
the solutions. The fee-setting
provisions of the Senate Manager's Amendment would provide as follows:
The Director shall have authority to
set or adjust by rule any fee established or charged by the Office under
sections 41 and 376 of title 35, United States Code, or under section of the
Trademark Act of 1946 (15 U.S.C. 1113), or any other fee established or charged
by the Office under any other provision of law, notwithstanding the fee amounts
established or charged thereunder, for the filing or processing of any
submission to, and for all other services performed by or materials furnished
by, the Office, provided that patent and trademark fee amounts are in the
aggregate set to recover the estimated cost to the Office for processing, activities,
services and materials relating to patents and trademarks, respectively,
including proportionate shares of the administrative costs of the Office.
The benefits of enactment of these provisions,
according to the White Paper, would be a decrease in the 750,000 application
backlog and a reduction in the average 34-month pendency for applications (the
time between original filing date and "a final disposition"). (Unexplained in these statistics is how
the 750,000 figure for the backlog was calculated, and whether a "final
disposition" is limited to allowance, abandonment, or appeal, or if
intermediate stages, such as a Request for Continued Examination, are
counted.) The Paper does correctly
point out that the pendency period is even longer than average for arts such as
information and communications technology, and that these delays can have
significant impacts in arts that are characterized by "rapid technological
turnover and short product life-cycles." Here, the Paper asserts, untimeliness in procuring patents
can lead to "worthless and obsolete" patents. While undoubtedly true, in such fields
it is reasonable to ask whether patent protection, procured at significant
investments of time and money, is an appropriate avenue for protecting
intellectual property that is ephemeral at best; such questions are not
addressed in the Paper.
Reducing pendency takes money, of course, and some
of the uses for the cash the Office anticipates it will collect if granted
fee-setting authority (and in the context of the economics set forth in the
Paper, this is really fee-increasing authority)
include "expenditure on IT infrastructure upgrades and additional hiring of
examiners." Fee-increasing
authority is needed, according to the Paper, because "the fee schedule in
the current patent statute fails to provide the USPTO with the flexibility it
needs to assure that its future revenues are commensurate with the costs it
will incur to modernize its operations." Further, "[t]he current fee structure is inflexible and
poorly aligned with actual costs, making it exceedingly difficult to fund
long-needed modernizations" according to the Paper. Some of the ways the fee structure is
thus "poorly aligned," the Paper asserts, is that applicants do not
pay fees sufficient to support the actual cost of the services the Office
provides them. The estimate set
forth in the Paper is that applicants pay fees equivalent to about one-third
the actual cost of examination, making "back-end" fees (such as issue
fees, publication fees, and patent maintenance fees) necessary to make up the
shortfall. And these fees are
sufficiently indeterminate that they do not provide the needed reliable source
of necessary funds.
The biggest strategic deficiency in this proposal
is that no matter what Congress does with regard to fee-setting authority, what
needs to be achieved is a commitment, backed by legislation, that that same
Congress will not expropriate the additional fees for its own, non-patent
related purposes (Indeed, after
several years of eschewing this type of fee-diversion, this Congress mandated
that any fees in excess of $100 million would go into the general coffers for
other expenses.) Arti Rai (at left), lead
author of the White Paper, explained to a gathering of the Biotechnology
Industry Organization (BIO) Intellectual Property Counsels Committee in New
Orleans this week that demanding this type of assurance was unwise because it annoyed
Congressional appropriators. Regardless of the soundness of this realpolitick,
however, without such a pledge
there is an automatic disconnect between the proposed solution and any evidence that it will achieve the
goal of appropriate funding for the Office. Additionally, permitting the Office to have fee-setting
authority creates the possibility that the fees will be used not just to
modernize the IT infrastructure or hire more examiners, but to modify applicant
behavior (as has been proposed in the past, unsuccessfully at least in part because
Congress would not adopt the proposed behavior-modifying fee structure in the
face of the resulting political firestorm of criticism). Having endured five years of the
Dudas regime, it is not irrational for members of the patent community to have
these fears; even with the general
acceptance of the benevolence of the Kappos administration, there is no assurance
that the next Director will have the public interest at least as much in mind
as meeting bureaucratic goals.
The other proposal for patent reform is post-grant
review, needed according to the Paper to ensure only "high-quality"
patents are granted (or at least maintained). The issue is framed in David v. Goliath terms, with patent
challenges being cast as particularly difficult for "small firms with
limited resources." While
certainly true, the Paper ignores the equally-likely converse situation: a resource-rich company using
post-grant review to harass a small firm with high-quality patents but limited
resources. While the Paper's
scenario poses a problem, the converse situation is even more fraught with
innovation-stifling potential, inter alia
because small firms with limited resources are highly dependent on patents for
investment (a reality discussed at the beginning of the Paper with regard to
VC-investment in technology-based companies), and anything that creates
uncertainty about such a small firm's patents will inhibit investment (making
acquisition by resource-rich companies of their undervalued patent assets more
likely). Regardless of these
unexplored possibilities, the Paper discusses threatened enforcement of "poor-quality"
patents by "large firms" that could "forc[e] small competitors
[of such large firms] with breakthrough technologies out of business,"
citing Benjamin & Rai, 2008, "Fixing
Innovation Policy: A Structural Perspective," 77 Geo. Wash. L. Rev. 1. The purported benefits of post-grant
review are supported by academic studies, including Graham & Harhoff, 2009, "Separating Patent Wheat from Chaff: Would
the U.S. Benefit from Adopting a patent Post-Grant Review?" (curiously, an
unpublished paper "on file
with the authors," which would benefit from being made publicly available
on the Patent Office website).
Disclosure of the bases for the assumptions is not set
forth in the Paper, however: in
the post-grant review section, the authors make most of the economic benefit
assertions, including that "the cost of post-grant review is expected to
be 50-100 time lower than the cost of patent litigation," that "between
one-third and one-half of [post-grant review] challenges can be expected to
result in an invalidity decision (which is frankly a dubious assertion unless
the post-grant review is expected to encompass the full range of invalidity
defenses available at trial), and cost-benefit analyses (which almost cry out
for data supporting them) estimating a return of from $8 to $15 for every $1 invested in the post-grant review
system. Instead of a quantitative
review of the assumptions behind these statements, the Paper merely asserts
that "almost every academic economist who has ever examined whether an
enhanced system of post-grant review should be adopted has favored such
adoption," and notes that the Federal Trade Commission, the National
Research Council, and the National Academy of Sciences also support this
reform. While comforting, this is
hardly a substitute for the kind of hard-headed, explicit economic analysis
that might be persuasive for such a proposal. The Paper does note that there is at least one academic
study that differs with the seemingly universal approval of another post-grant
review system (Shane, 2009, "Problems to
be Expected from Expanded Administrative Challenges to U.S. Patents," prepared for the Manufacturing Alliance for Patent Policy), but disputes its assumptions
(which are expressly set forth for examination). Benefits from decreasing the delay in invalidating patents
(which the Paper asserts can take 8-11 years in litigation) are supported by
another unpublished paper by Dr. Graham (2010, "Slow courts and the cost of uncertainty: How patent post-grant reviews
may offer a partial solution") (being another paper that would benefit the
public by being posted in the PTO website).
The Paper is valuable in setting forth goals the
current PTO administration believes to be important and how they propose to
achieve those goals. It is less
valuable in providing the public with the evidence it needs to assess whether
these proposals are related to achieving the goals or whether the reasoning
behind choosing these goals is sound. The patent community cannot and should not be expected to blindly accept
these proposals or their relationship to the purported goals, and it is justified
in expecting and demanding that the Office provide sufficient evidence in favor
of these proposals. This White
Paper does not fulfill those needs and expectations.
Patent Docs 
Leave a reply to Kevin E. Noonan Cancel reply