International Patent Application Outcomes

1
International Patent Application Outcomes
Paul H. Jensen Melbourne Institute and
IPRIA University of Melbourne Australian
Conference of Economists University of
Melbourne 26-28th September, 2005
2
STATE OF PLAY

• Firms apply for patents in each jurisdiction they
  want legal protection disharmony?
• Likelihood of disharmony attenuated by
• Patent offices apply the same patent threshold
• Triadic patent offices (USPTO, JPO, EPO) promote
  consensus on examination protocols
• However, recent evidence suggests there are
  differences in patent outcomes
• Use of aggregate statistics means existing
  studies cant control for quality of patent
• Controlling for quality, we ask
• How much disharmony really exists?
• Does it vary by patent characteristics?

3
PATENT APPLICATION OUTCOMES

• We define a patent application outcome as either
  granted, rejected, pending, or withdrawn
• 4 factors affect patent application outcomes
• Legislative differences first-to-invent (US) vs
  first-to-file (rest of the world)
• Institutional factors different resource
  allocation decisions and incentives in the patent
  offices. Strategic trade factors?
• Applicant behaviour is the applicant persistent?
• Patent characteristics
• area of technology
• priority country
• size of inventive step

4
DATA

• We constructed a dataset consisting of
• 70,477 single priority non-PCT (i.e.
  directly-filed) patents granted by the USPTO and
  applications in the EPO, JPO
• Matched sample of patent applications (unit
  records) with priority years 1990-1995
• Of the 190,583 triadic patent family
  applications
• 18,488 PCT patent families
• 172,095 non-PCT patent families
• 70,477 single priority families
• 101,618 multiple priority families
• Potential problems with the dataset
• Selection bias PCT more valuable?
• Truncation increase in patents pending?

5
DESCRIPTIVES

• Interesting observations (as at end 2004)
• High withdrawal and pending haze of
  indecision
• Only 37.6 of patents granted in the US also
  granted by both the JPO and EPO
• 0.6 were rejected by both the JPO and EPO
• JPO rejected 10 of USPTO and EPO patents
• Type I or Type II errors?

6
MODELING APPLICATION OUTCOMES

• Multinomial logit model used to estimate the
  effects of technology, inventiveness and priority
  country on the probability of each outcome in
  each patent office
• Inventiveness proxied by backward/forward
  citation ratio
• Results (marginal effects) presented for 2
  models
• JPO Model based on 50,958 observations (72.4)
  where USPTO and EPO granted, but JPO did not.
• EPO Model based on 31,314 observations (44.5)
  where USPTO and JPO granted, but EPO did not.

7
RESULTS AT THE JPO
8
RESULTS AT THE EPO
9
COMPARATIVE RESULTS

• Technology area
• Marginal grant effect in all tech areas (except
  biotech) at the JPO were negative, but not at the
  EPO where there was
• strong positive effect in auto industry
• weak positive effect in drug/chemical sectors
• strong pending effect in software
• Measures of inventiveness
• Positive effect at the JPO, but small at the EPO
• Priority country
• Strong negative effect at JPO for all foreign
  applications
• Weaker priority country effects at EPO

10
CONCLUSIONS

• Despite recent efforts at patent harmonization,
  substantial disharmony exists across trilateral
  offices
• Distribution of disharmony varies across tech
  area, priority country and inventiveness of
  application
• Economic effects of disharmony are unknown
• Two important caveats
• We cannot observe grounds for rejecting
  applications
• We cannot observe individual claims within the
  patents
• Question Are the USPTO/EPO granting too many
  patents or is the JPO rejecting too many patents?