R' Polk Wagner

1
IS THE FEDERAL CIRCUIT SUCCEEDING? An Empirical
Assessment of Judicial Performance

• R. Polk Wagner
• University of Pennsylvania

2
The Patent Law as an Institutional Design Problem

• The increasingly central role of the Federal
  Circuit
• an institutional mandate for jurisprudential
  management
• a robust (aggressive?) implementation of this
  mandate since 1990
• Recent events call for such an inquiry
• widespread dissatisfaction among the patent
  community
• and yet ... a growing awareness of the importance
  of decisions
• Critical analytic need the performance of the
  Federal Circuit against its mandate (clarity,
  stability, predictability)
• a look at the court in action

3
Unit of Measurement The Federal Circuits Caselaw

• Using court opinions as data (jurisprudence as
  the dataset)
• basic strategy gather large numbers of opinions
  on a particular issue, code those opinions for
  relevant characteristics, and mine the resulting
  data for patterns
• Advantages
• allows systematic, large-scale, long-term
  analysis reveals patterns, trends, differences
  that many not be otherwise apparent
• allows results to be expressed quantatively
• Disadvantages
• works best for big issues
• highly dependent on the coding strategy and
  implementation

4
First Up Claim Construction

• The Federal Circuits mandate is especially
  powerful here the basis of Markman II is
  jurisprudential management.
• The academic literature and commentary suggests
  opportunities
• high reversal rates (30-50) in literature (is
  this high?)
• practitioner and (district) court commentary
• universal agreement about importance
  (dispositive)
• Claim Construction offers an excellent data
  profile
• a pure Federal Circuit question question of
  law, no deference
• a large dataset claim construction is raised in
  virtually all cases, typically dispositive

5
Overview of the Study

• Dataset all Federal Circuit claim construction
  analyses since Markman II (April 1996)
• rolling updates (typically within 2 to 3 months
  of current)
• as of September 10, 2005 665 opinions (610 for
  the court, 55 alternative opinions) all data
  released to public immediately
• Case-Coding (converting court opinions into
  useful data)
• each opinion is coded for 14 categories usual
  categories methodological approach
• all opinions independently coded by multiple
  coders (early dataset 2, later dataset 5 or
  more) any discrepancies among coders resolved,
  tracked, and tabulated.

6
A Methodological Split Procedural versus Holistic
A Taxonomy of Methodological Approach
7
Why Methodology MattersWhich Approach is Used
Affects the Results

• Universal agreement that claim construction is
  most critical factor in patent disputes.
• The Procedural/Holistic dichotomy is drawn from
  the jurisprudence itself this is what people
  argue about (typically described as two lines of
  cases).
• (Some) empirical testing of this point
• Claim construction disputes among Federal Circuit
  judges tracked the detailed methodological
  categories 95 of the time.
• Federal Circuit reversals of D.Cts claim
  construction analysis tracked binomial categories
  82 of the time.

8
BASIC RESULTS
9
Frequency of Methodological Approach
10
A Move to ProceduralismClaim Construction
Methodology Over Time
11
JUDICIAL METHODOLOGY( PROCEDURALISTS, HOLISTICS
SWINGS )
12
The Methodology of Federal Circuit
Judges Proportion of Opinions Coded Holistic
13
The Methodology of Federal Circuit Judges The
Factions of the Federal Circuit
14
Factions on the Federal CircuitJudges grouped by
M-index (lower M-index more procedural)
15
Judicial (In)consistencyFederal Circuit Judges
Ranked by Consistency in Claim Construction
Methodology
16
PANEL-DEPENDENCY
17
Judges Impact on Methodological Approach
significance _at_ .10 level significance
_at_ .05 level
binary logistic regression, dependent variable
binomial category
18
Factional Impact on Methodological Approach
19
Panel Impact on Methodological Approach
20
predictor.claimconstruction.com
The data allows some reasonable predictability
concerning the claim construction results of a
given panel. This tool provides an interface.
21
search.claimconstruction.com
Search and download the entire dataset. Build
your own analysis.
22
PREDICTING PHILLIPS
23
The Building Storm Over Claim ConstructionThe
Rate of Alternative Opinions
24
Predicting PhillipsClaim Construction
Methodology at the Federal Circuit, 2004-2005
order
argument
opinion
25
Troubles Ahead?The Rate of Alternative Opinions
on Claim Construction, 2004-05
order
argument
opinion
26
IMPLICATIONS CONCLUSIONS
27
Recap The Basic Findings

• A distinct, yet persistent split in
  methodological approach.
• 65 procedural, 35 holistic.
• An overall trend towards proceduralism
  (statistically significant), at least until
  Phillips.
• Increasing polarization, at least until Phillips
  ?.
• The emergence of distinct factions
  (proceduralists, holistics).
• Panel composition absolutely affects methodology.
• For some panels, results can be predicted with
  90 reliability.

28
The Post-Phillips Environment

• The real effects of Phillips wont be clear for a
  year or more
• A single line of cases, but multiple
  methodologies.
• No reason to think that the distinctions will be
  eliminated.
• Phillips is a decision about the Federal
  Circuits role
• In the no rules environment, they make the
  decisions
• Hard to see how this is anything but less
  predictable (look at 96-99)
• Bottom line were far from the end of the story
  on claim construction this new chapter puts us
  back to 1996 (will history repeat itself).

29
Is the Federal Circuit Succeeding?

• Maybe The performance of the Federal Circuit on
  claim construction has not been an unqualified
  success a persistent split increasing
  polarization a division among Judges.
• note pre-1982, we had variable results, based on
  circuit now we have variable results, based on
  panel assignments!
• But some of the trends might suggest a
  turnaround
• a growing stabilization on the procedural
  approach
• polarization working it out
• Phillips remains an open question it might have
  been better to avoid en banc treatment.
• Future issue informal deference to formal
  deference
• The results highlight the importance of personnel
  changes.

30
IS THE FEDERAL CIRCUIT SUCCEEDING? An Empirical
Assessment of Judicial Performance

• R. Polk Wagner
• University of Pennsylvania Law School
• polk_at_law.upenn.edu
• 215.898.4356