Forum Shopping Within the United States Patent and Trademark Office.

• Author: Mehrle, Hannah

CONTENTS INTRODUCTION I. ART-GROUP SHOPPING WITHIN THE USPTO II. FORUM SHOPPING: OVERVIEW AND HISTORY A. How Plaintiffs and Defendants Engage in Forum Shopping B. Domestic Forum Shopping 1. Interstate Forum Shopping 2. State-Federal Forum Shopping C. Forum Shopping for Nationwide Injunctions III. FORUM SHOPPING AT THE APPELLATE LEVEL IN PATENT CASES AND THE CREATION OF THE FEDERAL CIRCUIT A. Forum Shopping Before and After the Creation of the Federal Circuit B. Compulsory Patent Counterclaims and Forum Shopping at the Federal Circuit C. Federal Circuit Choice of Law IV. FORUM SHOPPING AT THE DISTRICT COURT LEVEL V. FORUM SHOPPING WITHIN THE UNITED STATES PATENT AND TRADEMARK OFFICE A. A More Efficient Classification System B. More Concise and Exact Guidelines for the Examiners to Follow C. Internal Controls within the USPTO D. Easier Transfer of Cases by Examiners and Patent Applicants E. Decreasing the Number of Art Units CONCLUSION INTRODUCTION

The American legal system has historically frowned on efforts to gain advantages through forum shopping. (1) Lawyers, judges, legal scholars, and lawmakers have taken steps to address and dissuade litigants from forum shopping through articles, court orders, and legislation. (2) But every day patent drafters gain unfair advantages through a process similar to forum shopping by drafting patent applications to avoid unfavorable classifications within the United States Patent and Trademark Office ("USPTO").

After a patent drafter files an application with the USPTO, the patent gets classified into a broad class, and an art unit within the class. This classification, based on subject matter, determines which group of examiners will examine the application. (3) Drafting a patent to obtain classification into a specific art unit is similar to forum shopping because the classification of a patent application into one art unit instead of another can have a huge impact on whether the USPTO eventually issues the patent. For example, in 2016, a patent application classified into art unit 3628 was five times more likely to get issued than an application classified into art unit 3689. (4) But these issuance-rate discrepancies were not because the two art units covered different subject matter. In fact, art units 3628 and 3689 both encompass similar subject matter: "business processing, cost/price units and reservation." (5)

Recent increases in technology, especially in the area of big-data analysis, allow patent drafters to more easily predict the art unit into which the USPTO will classify an application the before the patent drafter actually files her application. (6) Thus, the patent drafter can change the wording of the application before filing to ensure a more favorable art-unit classification. In short, careful patent drafting can produce a much higher chance of patent issuance. (7) Albeit a narrow issue in a specialized area of law, creative patent drafting is a problem because patent drafters can game the USPTO easily when historically lawyers and scholars have actively opposed similar forms of forum shopping in other settings. (8)

This Note discusses forum shopping within the USPTO by placing the problem in the broader context of forum shopping in other areas of law. Part I addresses the issue of forum shopping within the USPTO, and how patent drafters are using creative drafting to gain more a favorable patent classification. Part II summarizes the legal treatment of forum shopping in general. Part III addresses patent forum shopping at the appellate level, including the creation and the impact of the United States Court of Appeals for the Federal Circuit. Part IV examines patent forum shopping at the district-court level. Part V suggests ways for the USPTO to address this rampant forum shopping, including recommending that the USPTO give examiners more specific guidance, make transferring an application to a different art unit easier, switch to in-person classification, implement more internal checks to manage examiners, or consolidate the number of art units. These changes will promote fairness and efficiency by minimizing the discrepancies in the issuance rates of similarly drafted patent applications.

1. ART-GROUP SHOPPING WITHIN THE USPTO

   When a patent drafter files an application with the USPTO, the application is classified first into a class and then into an art unit within that class, based on the application's subject matter. (9) The art unit determines which group of examiners--and thus which specific examiner--is assigned to examine the application throughout the prosecution of the application. (10) The USPTO publishes a list of the different art units into which the applications can get classified as well as the subject matter of each art unit; but beyond that list, the USPTO provides little public information about how its classification system works. (11)

   On its face, this classification process seems harmless, but a closer examination reveals that it is not. Several recent Supreme Court decisions regarding patentability have affected the examination of applications in some Art Units much more than others. In Alice Corporation v. CLS Bank International, (12) the Court held that an automated method of mitigating settlement risk in financial transactions was an abstract idea, and thus invalidated the patent under 35 U.S.C. [section] 101. (13) In Alice, the Supreme Court further held that the patent's claims did not contain an inventive concept that transformed the claims from an abstract idea into something patentable. (14) Additionally, in Mayo Collaborative Services v. Prometheus Laboratories, (15) the Supreme Court held that patent claims covering a drug-testing process were laws of nature and therefore unpatentable under 35 U.S.C. [section] 101. (16) Following the Supreme Court's decisions in Alice and Mayo, examiners in some art units, such as the business-methods and biotechnology art units, are much more likely to reject applications because the application consists of an abstract idea or a law of nature than before these decisions. (17)

   Of the twenty hardest art units, from which to obtain a patent, eight come from art unit 3600 (a business-methods art unit), which is hardly surprising considering the Supreme Court's decision in Alice. (18) Other than business-methods units, another five of the hardest art units are in the 1600s, which deal with biotechnology and organic chemistry. These groups were most affected by the Supreme Court's decision in Mayo. (19) At first glance, this may not seem like an issue, because the USPTO's examiners are just following the Supreme Court's guidance. But upon further inquiry, regardless of the subject matter of the patent, a patent classified outside these affected art units has a significantly lower chance of getting rejected for containing a law of nature or an abstract idea, inversely increasing, drastically, that application's chances of getting issued. (20) This is "extremely difficult to reconcile" when, despite covering more or less the same subject matter, examiners from one art unit are five times more likely to issue a patent than examiners from another art unit. (21) In one example, a patent with very similar subject matter to a business-method patent was classified into art group 3661 (Data Processing: Vehicles, Navigation, and Relative Location), (22) which, "according to PatentCore, has an issuance rate of 84.4%," and the alleged business-method patent was issued shortly after filing. (23) But had the same patent been classified into a class 705 art unit, the class that typically covers business-method art units, the issuance-rate percentage would have dropped to, at best, 43%. (24) This phenomenon shows that there are discrepancies in whether a patent is eventually issued based solely on the art unit in which the application is initially classified. (25)

   The differences in issuance rates between the different art units is hard to reconcile. Art unit 3689, which covers e-commerce, transportation, and national security, had the lowest issuance rate in 2015. (26) But, that same year, art unit 3659, which covers the same subject matter as 3689, had the highest issuance rate. (27) While eight of the hardest art units come from the 3600s, four of the easiest art units come from that same technology. (28) For example, art unit 3681 relates to transportation, construction, electronic commerce, agriculture, national security, and license and review; patent applications in that art unit have a 29.8% issuance rate. (29) But art unit 3688, which the USPTO has described as covering the same subject matter, has an issuance rate of about 81%. (30)

   The numbers show that the Supreme Court's decisions in Alice and Mayo are contributing factors as to why very few applications are issued out of certain art units, while other, similar art units remain largely unaffected. (31) For instance, work group 3690 examines applications in finance, banking, and insurance. Prior to Alice, the work group issued 2,100 patents a year, which corresponded to 25% of that office's total actions (i.e., nonfinal rejections, final rejections and issuances). (32) But after the Supreme Court's decision in Alice, issuances dropped to only around 2.6% of the office's total actions. (33) This drop seems to stem from examiners in finance art groups simply equating, without much more analysis, almost all applications to the patent application in Alice, and therefore issuing rejections because the patents are abstract. Additionally, the median number of patents issued each year by individual examiners in art unit 3690 is one. But this number is only a median; while some examiners are issuing new patents, others are not issuing any. (34) The chances of the USPTO issuing a patent should not depend so significantly on the art unit in which the application is initially classified.

   With all of the unpredictability and lack of uniformity...