Fee Shifting as Patent Policy Lever: How to Ensure Sufficient Torque

Guest Post by Professor Shubha Ghosh, University of Wisconsin Law School

Congratualtions to Hannah Jiam for her great work on attorney’s fee shifting after the Supreme Court’s 2014 Octane Fitness and Highmark decisions! The forthcoming article provides a deep portrait of how attorneys have responded to the new regime under Section 285 and how courts have ruled.  She also makes an insightful observation about the role of judicial discretion and the need for legislative reform of Section 285 to provide more detailed guidance for judges in order to direct and focus their discretion. Many scholars speak about attorney’s fees as a policy lever to limit frivolous litigation and the questionable practices of NPE’s. What I learned from Hannah’s article is that we also need to consider the torque of any particular policy lever (to take the nerdy techno-metaphor to the next level).

I have been pursuing research on Section 285 during my leave year from the University of Wisconsin as an AAAS Law, Science, and Policy Fellow at the Federal Judicial Center in Washington, DC. A completed paper summarizing my research will be available in a few months. Here, I would like to present some preliminary findings in order to complement Ms. Jiam’s work.

Where my methodology differs is to focus on the regimes pre-Octane Fitness as well as decisions after 2014. I am examining published judicial opinions ruling on 285 motions over three periods: (1) from the founding of the Federal Circuit in 1982 to the Brooks Furniture decision in the first week of 2005; (2) from 2005 to the Octane Fitness decision in 2014; and (3) from 2014 to the present. While there are methodological problems in looking at published opinions (which I address at the end of this post), looking at what courts have done and said with regards to 285 motions offers useful information on when courts have found a case to be exceptional and when not.

My principal finding is that judicial discretion matters regardless of the stated standard for the award of attorney’s fees.  In other words, raising the standard, as occurred with Brooks Furniture, or lowering the standard, as occurred with Octane Fitness, may not matter over all as to how courts exercise their discretion. The tentative implication is that the torque of attorney’s fees as policy levers depends on an articulation of what factors courts are required to use in awarding fees and perhaps ultimately on whether or not the fees are automatic.

Some preliminary findings from various data sources illustrate these points.  Table 1 reports findings from Lex Machina, which reports on various patent litigation events starting in 2000.  The database captures awards of attorney’s fees across jurisdictions.  Unfortunately, the database, when examined in March, 2015, did not report denials.  Nonetheless, consideration of the three periods for my study, we can see an above average of attorney’s fees awards per year during the Brooks Furniture period (from 2005 to 2014).  The relative size of this number, as compared to the pre-Brooks Furniture period, may be surprising since Brooks Furniture presumably made it more difficult to obtain attorney’s fees. The magnitude is most likely the result of increased patent litigation during this period, as reported by other scholars.  More careful study of the types of cases arsing during the Brooks Furniture period would explain the larger annual award rate by the district courts.

Terminated cases with award Yrs Cases/yr
Period 1 (2000-2005) 113 5 22.6
Period 2 (2005-2014) 292 9.33 31.29
Period 3 (2014-2015) 24 .9 26.67
TOTAL 429 15.23 28.17

TABLE 1: Lex Machina data on awarded attorney’s fees from 2000 to the present

Looking more closely at the 24 identified awards of attorney’s fees post-Octane, 17 of these cases involved awards to patent owners while 7 were awards to the alleged infringer (who of course prevailed at trial). Of the 17 cases in which the patent owner obtained attorney’s fees, 4 were default judgment cases and a different 4 also involved the award of enhanced damages. The last finding suggests that fees are awarded in cases that would not meet the standard for willful infringement.

One way to obtain a sense of types of cases is to look at the time towards final judgment from initial filing, what is reported in Lex Machina as “time to terminate.”  Of the 7 cases in which the alleged infringer obtained attorneys’s fees, 5 were cases with above average time to terminate, meaning cases that were initiated sometime before the Octane Fitness decision. Of the 17 cases in which the patent owner received fees, 8 involved above average time to terminate. If cases with longer times to terminate are viewed as more contentious cases, it seems that cases in which alleged infringers prevail on fee shifting are more contentious than the cases in which the patent owner prevails on fees.

The main limitations with respect to the Lex Machina data are the selection effects that arise with respect to examining reported cases and the exclusion of fee denials.  Nonetheless, even looking solely at instances of attorney fee awards, there does not seem to be a pro-defendant orientation in judicial decisions. But to fully understand whether Octane Fitness serves as an effective policy lever to control frivolous patent litigation, we need to compare the post-Octane world with the regimes that existed before.

Below, I provide some preliminary summary statistics about the denial and grant of attorney’s fees over three periods from 1982 to the present.  These three periods are divided into the pre-Brooks Furniture regime (1982-2004); the Brooks Furniture regime (2005-2014); and the Octane Fitness regime (2014 to the present). The following results are based on reported decisions in Westlaw on 285 motions.  The decisions include both grants and denials. I conducted the search of these cases in early January. I am currently working on updating the searches and expanding the coding of the cases for various factors pertinent to understanding what constitutes an exceptional case. I should emphasize that what I am presenting is still preliminary. But I believe there is enough here for discussion especially in light of Ms. Jiam’s study.

My Westlaw search identified 24 reported decisions on attorney’s fees post-Octane, 205 reported cases from the Brooks Furniture regime, and 147 from the period between the formation of the Federal Circuit and the Brooks Furniture ruling.  I examined a random sample of 20% of the cases from the Brooks Furniture regime and 16% from the pre-Brooks regime.  These two random samples provide the basis for the statistics presented below.

Table 2  reports on the 24 reported cases post-Octane. Of these 24 cases, 12 resulted in grants and 12 in denials. Of the 12 grants, 9 went to the non-patentee as prevailing party. This may suggest that non-patentees are more successful post-Octane in obtaining attorney’s fees awards. But what must also be taken into consideration is that the 12 denials of awards all were in cases where the non-patentee’s motion was denied by the court. Just to be clear, these denials involved cases in which the non-patentee was the prevailing party but the court determined that the case was not exceptional. Unlike the Lex Machina data, the Westlaw reported cases capture both grants and denials. The denials are important to take into consideration in examining the effectiveness of section 285 as a policy lever. To highlight this point, consider the following. Of all the 24 reported cases involving fee shifting, 21 involved motions from the non-patentee. The disproportionate number of non-patentee motions would suggest that the Octane Fitness regime might be more favorable to the non-patentee.  But of these 21 motions, only 9 resulted in a grant, suggesting that the motions were successful in less than half the cases, as reported in Westlaw.  On the other hand, patent owners had a 100 % success rates based on the 3 reported cases involving a motion brought by the owner.

n Patentee Nonpatentee Infringement Declaratory  Judgment
Grant 12 3 9 11 1
Deny 12 0 12 10 2
 TOTAL  24 3 21 21 3

Table 2: Attorney fee awards reported decisions post-Octane

How the litigation environment has changed under the Octane Fitness decision requires comparison with the world before hand.  Tables 3 and 4 report similar statistics for the pre-Brooks and Brooks regimes, although based on random samples from the respective population of cases.

 n Patentee Nonpatentee Infringement DJ
Grant 9 2 7 7 2
Deny 15 6 9 13 2
 TOTAL 24 8 16 20 4

Table 3: pre-Brooks reported cases. Total=147. N=24 (16% random sample) 

Patentee Nonpatentee Infringement DJ
Grant 12 4 8 12 0
Deny 28 10 18 24 4
 TOTAL  40 14 26 36 4

Table 4: Brooks Furniture reported cases. Total=205. N=40 (20% random sample)

These two tables are based on random samples from the population of cases and hence are very preliminary.  Nonetheless they raise some hypotheses to examine in considering the entire population of cases.

The first point to emphasize once again is that denials need to be taken into consideration in gauging the effectiveness of a particular attorney’s fees regime. Under the pre-Brooks regime, there were about 62 % denials among the reported cases ruling on attorney’s fees. By comparison, there were 70 % denials among the reported cases ruling on attorney’s fees during the Brooks Furniture regime.  Since the Brooks decision raised the standard for the award of attorney’s fees, it is not surprising that the denial rate is higher. But perhaps surprisingly, the rate may not be significantly higher (I have not done this test yet).

Furthermore, even though reported cases deal with motions from nonpatentees more frequently than motions from patent owners, nonpatentees are more likely to be denied fees under both the the pre-Brooks and Brooks regimes. Nonpatentees had a 44%  grant rate under the pre-Brooks regime and a 31 % grant rate under Brooks. By comparison, patentees had a 25% grant rate pre-Brooks and a 28% grant rate under Brooks.

Once again, the various statistical tests of significance have not been applied yet to provide a full interpretation of these data. But two points are important at this stage of the research.  In order to understand the effectiveness of Octane Fitness (its torque), we need to understand what the world of attorney’s fees looked like before the 2014 decision.  Furthermore, we need to examine how grants and denials are awarded between patentees and nonpatentees. What the data suggest so far is that the various standards, however characterized as easy or hard, do not seem to favor one side over the other in patent litigation. To me, this suggest that judicial discretion may be working around the articulated standards to make decisions on a case by case basis. Part of the research is to determine what courts articulate as the relevant standard.

Table 5 partially addresses the big question of what courts deem to be exceptional cases. The table reports frequently used phrases across the various regimes in the published opinions.

Pre-Brooks Furniture “dilatory tactics”; “inequitable conduct”; “willfulness”; “failure to investigate”; “fraud on USPTO” “no intent to deceive”; “no willfulness”; “no vexatious litigation”; failure to meet burden
Brooks Furniture “vexatious litigation”; “misconduct in USPTO”; “bad faith”; “obj and subj baseless in light of claim construction” failure to meet burden; “no willfulness”; “no misconduct”
Post-Octane litigation misconduct; discovery abuse; relitigating arguments; “lit theory inconsisitent with claim construction”; uncivil tacticsnuisance suit; Fogerty factors No misconduct; Reasonable tactics; No bad faith

Table 5: What courts say and do 

An advantage of examining reproted cases, despite the limitations, is the ability to discern what courts are thinking in granting or denying fees. Let me focus on a few highlights here given my space restrictions.  First, notice the prevalence of identified misconduct by the patentee or nonpatentee and attorneys as a basis for granting awards. Second, notice the prevalence of failure to meet the burden of proof in denying awards under the Brooks Furniture standard. Finally, notice the discussion of willfulness across the various regimes indicating that the determination of awarding fees occurs in the broader context of inequitable conduct and enhanced damages. My main conclusion from reading and assessing these cases is that courts seemingly exercise their discretion in determining what constitutes an exceptional case in similar ways despite the stated standard for granting fees. The policy prescription would be that the torque from using fee shifting as a policy lever may be dampened by judicial discretion. Whether legislative reform should cabin such discretion is at the core of how effective and politically feasible such reform would be.

One last word about my methodology.  There are obvious selection effects from examining reported cases as I have done here. These selection effects are also relevant for Hannah’s article. Reported cases are not the typical case. Although we need to consult reported cases to determine what courts consider “exceptional,” we need to keep in mind that the reported cases are not representative of the population of cases that fee shifting rules are supposed to regulate.  One goal of fee shifting rules is to prevent frivolous litigation and specifically extracted settlements based on frivolous claims. The effect of fee shifting on settlement is a complex one. Assuming that frivolous cases can be precisely identified by a judge, patent owners may be less willing to bring frivolous cases. At the same time, nonpatentees may be less willing to settle all cases, particularly frivolous ones. Looking solely at reported cases does not allow a researcher to determine how effective a fee shifting regime is in controlling frivolous litigation and settlement. Therefore, we should be wary of making global conclusions from looking at reported cases.  Nonetheless, the summary statistics I provide here, in conjunction with Ms. Jiam’s excellent work on post-Octane developments, shed some light on a critical aspect of the patent reform debate.

9 thoughts on “Fee Shifting as Patent Policy Lever: How to Ensure Sufficient Torque

  1. 4

    Your conclusion that individual judicial discretion still plays a big part in whether a party will get sanctioned or not correlates with the good advice old trial lawyers give young ones: don’t play games in litigation, don’t burden judges with filing overly lengthy non-essential papers they don’t want to read*, and don’t do anything else that may irritate the judge or affect your case credibility.

    *besides the time burden on judges who have more pressing matters, like their criminal case docket, this can lead to a suspicion that you are doing so to run your client billing meter.

  2. 3

    Noting the innate sense of “any law suit is bad“…

    Noting that a true norming is not present (the norming factor is not merely the number of law suits, which does help, but the population of possible lawsuits),…

    I would push the meme on torque to an observation that too much torque may be being desired to be applied and that your screw threads might be str ipped.

      1. 2.1.1

        Witness the overblown “Tr011” hype by contrasting the Whitehouse paper (being challenged by Ron Katznelson) and the more objective other government study that came out later that same year, that even with the arguably biased Lex Machina data (Lemley being a principal thereof).

        There is an inordinate unspoken assumption that a “problem” exists particular to patents, an assumption that is clearly fanned out of proportion by a group (or groups) whom are active in pursuing their agenda.

        Need I remind you just who coined the term “Patent Tr011,” and just why they did so? Hint: it had nothing to do with anything except protecting Big Corp profits.

  3. 1

    Prof. Ghosh,

    Thanks for gathering and analyzing all this data. It is very interesting, and worth considering by anyone who wants to know about the changes in fee shifting through time and what changes could be proposed.

    One quick note though: I think your autocorrect may have changed all the “Brooks” in “Brooks Furniture” to “Brooke”.

Comments are closed.