Thursday, August 07, 2014

IPSC part 1

Opening notes: usual disclaimers apply. Given the size and scope of the conference, I couldn’t attend half of what interested me. (I wish more people would send in their papers so I could read them!)  My notes aren’t guaranteed to be accurate and reflect my opinions.

Plenary Session

Who Should Pay for Progress?
Jeremy Sheff

IP law embodies a conception of distributive justice; this is legitimate but contestable—competing conceptions of justice, not a failure of justice or justice v. efficiency.  Implementing IP’s conception of distributive justice is difficult practically, which is responsible for common doctrinal tensions.

Access: who should enjoy the benefits of creative activity.  Alternative regimes may have components of compulsory licensing or open access.  But access is only one side of distribution; the other side is burdens.  Who pays?  Assume we’re certain that our regime will produce any unit of knowledge we want. How would the regime distribute the burdens of producing that knowledge?  One feature of alternative regimes is that they often distribute burdens assuming that the gov’t has a role to play and that there’s no specific taxation, though there’s progressive taxation/ability to pay. IP imposes pecuniary and nonpecuniary burdens—the burden of legal exclusion. This burden is only felt by people who feel the absence of a benefit that would be theirs under another regime (access to a drug, for example).

The burdens of producing a unit of new knowledge ought to be correlated to the benefits at the level of the individual. That’s a principle of distributive justice we can infer from the IP regime.

This is not an inevitable principle of distributive justice, but it is a legitimate one. Issues: relevance of luck, desert w/r/t inherited endowments; responsibility of individuals for historical injustices from which they benefit but which they did not cause; moral responsibility for future generations; duties of one moral agent to another; whether and how aggregation of individual needs, values, and preferences ought to be carried out.

When we argue over desirability of IP v. alternative frameworks, these deeper normative/philosophical questions are what we’re really arguing about even if we claim to be arguing about welfare maximization.

Regardless of whether we think that the “correlation at the individual level” principle is just, we have to concede it isn’t easy.  Substantial similarity in copyright; nonobviousness in patent.  Debates over moral duty of filesharers to pay for music they enjoy.

Wendy Gordon: In patent, you don’t have to receive a benefit from someone to be liable for infringing their patent.

Sheff: True, works better in copyright.

Mark Lemley: your theory would be consistent with a user fee theory of gov’t, but most of the things we care about in society we don’t take that view—people who benefit from welfare aren’t the ones we think should pay for welfare (careful about that ‘we’)—why do we think that in IP?

Sheff: not trying to be prescriptive, but descriptive.  Should do the work of addressing the normative commitments in a distributive exercise—what moral individuals are responsible to one another for.

Q: how do economic theories fit in?

Sheff: buried distributive concerns: user pays means we have to figure out who is a user and what it means to pay.

Is the time Allocated to Review Patent Applications Inducing Examiners to Grant Invalid Patents? Evidence from Micro-level Application Data
Melissa Wasserman (presenter; coauthored with Michael Frakes)

Is there empirical evidence that PTO overgrants patents?  The time they’re allocated to review an application is one piece of evidence.  Average: 19 hours to review each application, search prior art, determine patentability; novelty and nonobviousness take the most time.  Obviousness is the most time intensive, b/c you have to cobble together prior art.  Applications are presumed valid.

Lots of anecdotal evidence that examiners lack time, leading to invalid grants. Hypothesis: as given less time, examiners will cite less prior art, make fewer time intensive objections like obviousness, grant more patents.  Take advantage of fact that promotions decrease time allocated to review applications.  GS-7 gets about twice as much time as GS-14 (10.2 hours).  There is a change in level of scrutiny when you get promoted to GS-14—below that all decisions are reviewed by an examiner with full signatory authority. 

Tracked grant rate, rejections, number of prior art references cited over time.  All 1.4 million utility patent applications from 2001-2012 filed on or after March 2001 and published or disposed of by July 2012.  FOIA’ed PTO for annual roster of GS level and experience.

Results: evolution of grant rate over course of examiner’s career.  As you get promoted, you grant more patents.  Is that just an experience effect?  Decouple b/c promotions don’t always happen on a particular schedule.  Turns out it’s true even after controlling for experience.  Grant rate seems to increase on promotion, but when you spend more time at a particular level, grant rate goes down.  It’s a learning story—examiners get better at forming bases of rejection, only to get interrupted on promotion and new time crunch.

Obviousness rejections show the same pattern—they decrease with time crunch, then creep back up after the examiner gains experience with their allocated time period.  Higher GS level = fewer examiner (not applicant) citations of prior art (at least for issued patents). 

PTO may be treating similar applicants in dissimilar ways.  Scaling of time allotments may be too aggressive.

No comments:

Post a Comment