Hindsight Bias: The Human Condition1
by Frank S. DiGiglio
"Hindsight is 20/20." We have all heard and appreciate the comment. "Monday Morning Quarterbacking," a recognition that after
the fact, the prior circumstances are readily known, appear predictable and even inevitable. These are each actually a recognition
of hindsight bias, a human condition which appears to defy escape. A psychological exaggeration of foresight in the time before
the event, once the outcome of the event is known.
In patent law, this phenomenon presents a reality of particular problems at the core of the analysis of invention, measuring what
is patentable and what is not. Consider first, a principal condition of patentability found in 35 U.S.C. §103 which provides that
a patent cannot issue if:
the difference between the subject matter to be patented
and the prior art as a whole would have been obvious
at the time the invention was made to a person of ordinary
skill in the art to which said subject matter pertains
(emphasis supplied).
Accordingly, a critical measure of patentability is that threshold between the prior art and the departure of the new invention, determined,
according to the statute, at the time of the invention. And while the statute requires this initial constraint on the concept of hindsight bias,
it offers little more than the simple mandate. The statute does not suggest how this analysis is to be accomplished.
Accordingly, patent office examiners, judges and juries learn the invention after it has been achieved and subsequently measure
whether it was obvious to the skilled artisan in view of the prior art, at the time the invention was made.
The problem is, once we know the invention, it is cognitively impossible for us to objectively measure the obviousness of the innovation
in the time prior to the invention. The knowledge of the invention and a certain inevitability of its achievement is always brought to the analysis.
It is noteworthy too, that case statistics demonstrate that most cases (in the USPTO and the Courts) turn on the assessment of obviousness,
in the context of patentability and validity, respectively.
Thus, at the center of many legal determinations of who secures a monopoly and who does not, an essential element of any intellectual
property policy, is a determination of the difference between the innovation and past technology. And, at the center of that analysis is a human
condition -- once we are told what the innovation is -- it is impossible for us to objectively assess the innovation prior to its arrival. In other words,
once we know the invention, we cannot assess it in the absence of a bias -- a hindsight bias.
Various studies have demonstrated that hindsight bias is not only very real in patent law, but consistently skews the analysis of what
is and what is not a patentable departure from the prior art, while also affecting the analysis of enablement, claim construction and even
the doctrine of equivalents. Invariably, suggests the data, hindsight bias makes it more difficult to find patentablity under an obviousness
analysis because the invention always appears more readily achievable after it is, in fact, achieved. The bottom line is that fewer inventions
are found patentable as a consequence of the bias of hindsight. With enablement, for example, the invention appears to be more readily
enabled after we have seen that it has been achieved. This is particularly true in technologies that develop very quickly. There is ultimately
a bias to conclude that more is enabled at the time of the invention after we know how the invention was resolved. The after-the-fact knowledge
is always imputed to the earlier process.
The law has, of course, attempted to constrain this manifestation of the human condition. First, in the context of an obviousness analysis,
the statute requires that the analysis of the differences between the inventive subject matter and the prior art be measured "at the time the
invention was made." However, as earlier indicated, the imperative is, at best, elusive. Once we know the invention achieved, what the
inventor found at the time of the invention has been inescapably altered in the analysis, by the achievement of the invention itself.
The case law, and particularly the Supreme Court decision, in Graham v. John Deere Co. of Kansas City, 383 U.S. 1, 86 S.Ct., 1966
offers a principal constraint to the hindsight bias. The decision requires that the prior art must reflect some teaching, suggestion or
motivation (TSM) in order to sustain a finding that the present invention is obvious in light of the prior art. And various cases hold that
a hindsight reconstruction of the prior art is prohibited. Nevertheless, this requirement does not obviate the inherent hindsight bias
when reviewing the completed invention in the context of the prior art. Again, recent studies show that the bias is unabated and that
the Supreme Court TSM test does not mitigate hindsight bias. The net effect is that less subject matter is determined to be inventive
as a consequence.
Then, recently in KSR Intern. Co. v. Teleflex Inc., 550 U.S. 398, 127 S.Ct. 1727 U.S., 2007, the Supreme Court has diluted the
TSM test in view of the fundamental criticism that just because the prior art does not provide a specific teaching or suggestion,
it does not mean that the proposed invention is not obvious. While there is merit in the analysis, removing the teeth of the TSM
test does nothing to constrain the application of the hindsight bias in the equation. In fact, the decision in KSR fundamentally
diminishes the bias constraint and makes it easier to find new inventions obvious. In terms of tempering the hindsight bias,
the decision in KSR appears to exacerbate the problem. Even less innovation will be determined to be inventive as unobvious.
Various efforts and suggestions have been made to ameliorate hindsight bias in the assessment of patentability but with little or
no success. Starting with an assessment of secondary considerations as part of the obviousness determination in the law, further
education of the prior art and offers of incentives to the decision makers, nothing appears to mitigate the hindsight bias. It appears
to be an inescapable part of the human condition, once the knowledge of the invention is provided to an individual, that person can
no longer assess the subject invention objectively in the absence of the bias.
The only effective means to achieve an objective view of the invention at the time prior to its realization, in the context of an obviousness
determination, is to avoid showing the investigator the completed invention in issue. Such analysis must be conducted with an
appreciation of the ordinary level of skill in the subject art, the relevant prior art and an articulation of the problem to be solved.
An analysis independent of the resolution of the problem must be conducted in order to avoid the hindsight bias. The question
to be answered is therefore whether a solution to the unsolved problem is obvious from the prior art to one ordinarily skilled in
the art. There does not appear to be any other way to satisfactorily address the hindsight problem.
Of course, such a conclusion presents as difficult a practical solution as the problem it seeks to resolve. We have all probably
experienced some success with such an approach with experts and expert witnesses providing those individuals with the problem
to be solved and the prior art in the context of an inquiry to determine an objective opinion of obviousness. However, presenting the
issue to a judge, a jury or a patent examiner in this manner is likely to create as many difficulties as it seeks to resolve. Nevertheless,
it may be the only way to truly achieve an analysis of an invention in the context of a determination of obviousness. Since methodologies
to provide the investigator the inquiry without the knowledge of the invention are practically elusive, hindsight bias will remain part
of the human condition in the fundamental assessment of patentability.
1Based on the work and study of Professor Gregory N. Mandel.
Gregory N. Mandel, Patentably Non-Obvious II: Experimental Study on the Hindsight Issue before the
Supreme Court in KSR v. Teleflex, 9 Yale J.L. & Tech. 1 (2007).
Gregory N. Mandel, Another Missed Opportunity: The Supreme Court's Failure to Define Nonobviousness
or Combat Hindsight Bias in KSR v. Teleflex, 23 Lewis & Clark L. Rev. 323 (2008).
|