Dissecting Dissents for Ex Parte Appeals

Dissent is not the highest form of judgment for judges on the Patent Trial and Appeal Board (PTAB) of the United States Patent and Trademark Office (USPTO).  As discussed in further detail below, our own analysis indicates that dissents for ex parte appeals are found in about .5% of decisions issued by the PTAB.  A PTAB judge deciding an ex parte appeal is more than ten times less likely to dissent than a Federal Circuit (CAFC) judge.

The PTAB decides thousands of ex parte appeals per year.  Each appeal is assigned to a panel of three Administrative Patent Judges (APJs).  While one judge is designated to write the initial opinion, all three judges are supposed to take an active role in adjudication before the final decision is issued.

Relying on internal USPTO policies and former PTAB judges’ personal experiences, a recent spate of commentary has provided different explanations regarding the rarity of dissents for ex parte appeals.  We were still left wondering why some judges go out of their way to write dissents.  In an effort to better understand this issue and what the dissents might reveal about the ex parte appeal process in general, we conducted a statistical analysis of dissents in recent ex parte appeal decisions.

Methodology and Qualifiers

Initially, we identified 96 publicly available final ex parte decisions that contained dissenting opinions over a 20-month period.  The date range was chosen to reflect the current fiscal year (FY 2018), from October 1, 2017 to May 31, 2018, as well as the past full fiscal year (FY 2017), from October 1, 2016 to September 30, 2017.

In order to identify the 96 decisions, we searched for dissents with a date range of [10/01/2016 to 05/31/2018] at USPTO’s e-FOIA page.  Out of the related results, 96 actually contained dissenting opinions for ex parte appeals.  The other results populated due to the presence of the word “dissent” in majority opinions when quoting other dissenting opinions.  There are likely more dissenting opinions that weren’t identified in the search results due to some decisions not being publicly available and due to possible errors in text recognition and/or search functionality.  Similarly, the total numbers of decisions identified using USPTO’s e-FOIA page, discussed below, do not exactly match statistics provided by the USPTO because some of the decisions aren’t publicly available and because of other possible issues with USPTO’s e-FOIA page (e.g., duplicate search results being returned for a decision).

Typical ex parte decisions are for appeals filed when an applicant thinks that an impasse is reached with an examiner during the normal course of prosecution.  The 96 identified, dissenting opinions do include dissents for ex parte decisions that aren’t representative of the typical ex parte decisions, such as decisions for pre-AIA reexaminations and for requests for rehearing.  For some of the analysis below, we use only a subset of the 96 dissents to make comparisons to the publicly available statistics that categorize ex parte decisions in different ways.

After identifying the dissents, we identified the following data for each one of the identified 96 final decisions that contains a dissent:

  • Technology Center
  • Outcome (i.e., affirm, reverse, etc.)
  • Rejection Type (i.e., §101, §102, §103, etc.)
  • Judges
  • Page Numbers

Analysis of Statistics Regarding Dissenting and Concurring Opinions for Ex Parte Appeals

Table 1: Rate of Dissents for Ex Parte Appeals

As shown in Table 1, the 96 dissents represent less than .5% of the total number of decisions that have been issued for ex parte appeals over the 20-month period.  The total number of decisions, in Table 1, were identified using USPTO’s e-FOIA page.  Surprisingly, even though the dissents represent such a small proportion of the total number of decisions, the percentage of dissents has remained incredibly similar between FY 2017 (0.47%) and the beginning of FY 2018 (0.51%).  This suggests that dissents for ex parte appeals is a consistent – even if very small – phenomena at the PTAB.

A recent study found that Federal Circuit judges dissent less than 7% of the time.  Therefore, dissents for ex parte appeals at the PTAB are about 14 times less likely than dissents at the Federal Circuit.

Table 2: Rate of Concurrences for Ex Parte Appeals

As shown in Table 2, we also identified 49 final decisions for ex parte appeals that contained concurring opinions over the same 20-month period.  To identify the concurrences, we searched for concurring opinions with a date range of [10/01/2016 to 05/31/2018] at USPTO’s e-FOIA page.  There are likely more concurring opinions that weren’t identified in the search results due to similar issues discussed above with regards to dissents.  The 49 concurrences represent .25% of the total number of decisions that have been issued for ex parte appeals over the 20-month period.

Therefore, there are almost twice as many dissents as concurrences.  In other words, PTAB judges are even less likely to spend time writing a concurring opinion than writing a dissenting opinion.  The dissenting opinion can serve as a useful tool when a judge refuses to sign on to a decision because she thinks the outcome is wrong and the other judges on the panel continue to disagree.  For a judge to write a concurring opinion, he would have to agree with the outcome of the decision but disagree with how the decision is written.  Unsurprisingly, PTAB judges are less likely to go out of their way to write an additional opinion when they just disagree with the reasoning in the decision than the actual outcome of the appeal.  This also understandably suggests that PTAB judges don’t view the reasoning presented in their decisions as nearly as important as the actual outcomes of those decisions.

Table 3: Dissents by Technology Center (TC)

As shown in Table 3, we also identified the USPTO technology centers (TCs) that handled the applications related to the 96 identified dissenting opinions.  Over 20% of the 96 dissents were written for applications that were handled by TC 1700, the Chemical and Materials Engineering TC.

Table 4: Comparing Results by TC with Respective Fiscal Year

Table 4 illustrates representation of TCs in final decisions involving dissents for FY 2017 and FY 2018.  Since FY 2018 is still not over, the data set for FY 2018 is smaller (33 samples versus 63) and does not represent an entire fiscal year.  The quantities of dissents are based on the 96 dissents identified using USPTO’s e-FOIA page, while the total numbers of ex parte appeal decisions for those fiscal years are based on the statistics provided by the USPTO.  We analyzed Table 4 by comparing the percentages.

For example, TC 1700 is overrepresented in both FY 2017 and FY 2018, although by a smaller margin in FY 2018 than in FY 2017.  So while the proportion of total ex parte appeal decisions for TC 1700 remained relatively constant – at 13% in FY 2017 and 12% in FY 2018, the proportion of dissents has decreased from 24% in FY 2017 to 15% in FY 2018.  Since the same group of PTAB judges decide ex parte appeals for particular TCs, this suggests that there is no consistent behavior within these groups of PTAB judges that leads to the particular number of dissents.  Overall, the proportion of dissents for applications from a particular TC is relatively comparable to the proportion of total number of decisions for that particular TC.

Table 5: Dissents by Decision Outcome and Comparison to All Decisions by Fiscal Year

Next, we analyzed decision outcomes (i.e., affirmances, affirmances-in-part, reversals).  Table 5 compares the proportions of decision outcomes in decisions with dissents to proportions for those same types outcomes for all decisions.  Table 5 does not include the six dissents of the 96 dissents that were for “Request from Rehearing Denied” since that outcome was categorized differently in the statistics provided by the USPTO.

The smallest number of dissents is for “affirm-in-part” decisions, which is not surprising since “affirm-in-part” decisions occur much less frequently than affirmances and reversals.  For FY 2017, 12.7% of final decisions containing dissents were for “affirm-in-part” decisions and 12.3% of all final decisions were for “affirm-in-part” decisions.  We were surprised to see the percentages being relatively similar since we expected a more significant proportion of the “affirm-in-part” decisions to be reached as a compromise between PTAB judges instead of one of the judges writing a dissent.

However, judges are much more likely to dissent from reversals than affirmances.  For example, while only 29.5% of all final decisions were reversals in FY 2017, 32.3% of final decisions containing dissents were reversals.  And, while only 56.4% of all final decisions were affirmances in FY 2017, only 48.5% of final decisions containing dissents were affirmances.  This indicates that reversals are more controversial amongst APJs.  As a result, judges are generally less willing to agree to decisions that are reversals than affirmances, and judges are more likely to go out of their way to write dissents when reversals do take place.  In turn, this suggests why the rate of reversals is so low in the first place: a judge who writes an opinion wants to avoid a situation where the other judges on the panel refuse to sign the opinion because it’s a reversal.

Table 6: Types of Rejections Discussed in Dissenting Opinions

Table 6 identifies what types of rejections were discussed in the 96 dissenting opinions.  Some of the 96 dissents discussed multiple different rejections.  However, many more of the final decisions discussed multiple different rejections while the dissents usually focused on only one of those different rejections.  Section 103 rejections were by far the most commonly discussed in the dissents, which is not surprising since § 103 rejections are the most common type of rejection made by examiners.

Table 7: Types of Rejections Discussed in Dissenting Opinions and All Decisions

Next, as shown in Table 7, we looked at how rejections represented in dissents compared against rejections represented in all final decisions.  To identify the decisions related to the different types of rejections, we used search terms related to those types of rejections at USPTO’s e-FOIA page.  We then took the percentage of dissents based on a particular rejection type out of all decisions of a particular rejection type to compare to the percentage of all dissents out of all decisions (.48%).

We expected for § 101 rejections to be most overrepresented in dissents since there is by far the least clarity and the most disagreement regarding § 101 compared to the other sections.  Instead, there were only 11 dissents – out of the 96 dissents – that discussed § 101 rejections and they made up only .41% of all decisions that discussed § 101 rejections.  The .41% is less than the .48% cumulative rate of dissent for all types of rejections.  Therefore, surprisingly, the lack of clarity regarding § 101 is not leading judges to write dissents more frequently for § 101 rejections than for other types of rejections.  One possibility is that judges on the panel are more willing to give judges writing the opinion more leeway to reach the decisions with regards to the § 101 rejections.  Another possibility is that judges who would normally write a dissenting opinion are less willing to do so for § 101 rejections because of the complexities involved with applying § 101 and the reluctance of doing so on record in a dissenting opinion.

Table 8: APJ Statistics

As Table 8 shows, we also examined the data to see if there is a small number of judges who are outliers that are writing almost all of the dissenting opinions.  We were surprised to find that there were 55 different judges who wrote at least one of the 96 dissenting opinions.

According to the USPTO, 226 judges decided appeals in FY 2017 and 224 judges decided appeals in FY 2018 (“RE: Number of PTAB Judges,” Email from USPTO Statistician to author Eli Mazour, August 7, 2018).  Therefore, the 55 different judges who wrote dissents represent about 25% of all judges who write appeals.  It is unlikely that this many different judges would have written dissents if there was an active policy of discouraging judges from doing so at the PTAB.  This indicates that dissents are a tool widely used by PTAB judges for ex parte appeals in exceptional situations when they do not feel comfortable signing on to the majority opinions.

However, we did identify one outlier judge who is responsible for writing 16 of the 96 dissenting opinions.  Most of the 16 dissenting opinions were written for applications examined in TC 1700.  In doing so, the judge clearly demonstrated an incomparable work ethic and approach for deciding to dissent.  The judge dissented from all types of decisions, including decisions to affirm, affirm-in-part, and reverse.  Therefore, the judge does not seem to be driven by some ideological agenda to consistently side with applicants or examiners.

Figure 1: Length of Opinions

Finally, we wanted to see if the amount of work involved in writing dissenting opinions might explain how few of them are written.  To do so, as shown in Fig. 9, we identified the number of pages in the majority opinions of the 96 decisions and the number of pages in the dissenting opinions of the 96 decisions.  We removed dissents for requests for rehearing since those dissents typically only contained a couple sentences indicating that they dissented for the same reason as the previous hearing.

We found that dissenting opinions are typically significantly shorter than majority opinions.  This seems mostly due to the fact that majority opinions often tackle multiple grounds of rejections, while dissents usually include arguments for only one of those rejections.  Nonetheless, if PTAB judges were able to get credit for writing a dissenting opinion, the amount of work involved would be less than for writing a majority opinion.


The statistics discussed above show that while dissents for ex parte appeals are extremely rare overall, about one out of every four APJs who handles appeals has written at least one dissent in the 20-month time period.  The statistics do not provide any nefarious explanations for the dissenting practices (or, more precisely, the lack of dissenting practices) of PTAB judges.

Most charitably, the lack of dissenting opinions is a sign of a culture at the USPTO where a panel of PTAB judges is encouraged to work together to form consensus regarding the best possible decision.  However, anecdotal and other statistical evidence indicates that a decision for an ex parte appeal is driven by the judge who is designated to write the opinion.  In rare situations, one or both of the other panelists disagree with at least parts of the initial written opinion.  The quickest way for the writer of the opinion to get credit is to either convince the other panelists or to modify the opinion until all of the panelists agree to sign the opinion.

For exceptional situations when neither one of those options can be easily achieved, dissents serve as a useful escape valve.  The designated judge who wrote the initial opinion can make sure that the issuance of the opinion is not held up by asking the dissenting judge to just write a relatively short opinion that records his disagreement.

The biggest issue is that, like the use of expanded PTAB panels for post-grant proceedings and the great USPTO outage of 2018, the extremely low number of dissents for ex parte appeals contributes to the perception problems of the USPTO.  The USPTO must come up with a public explanation regarding why panels are used for ex parte appeals if dissents are so rare and what purpose they expect dissents to serve.  The USPTO then should explain if these goals are being met or implement an appropriate plan.

From a practical standpoint, a practitioner should not expect PTAB panel members to spend time coming up with individual positions for an appeal.  Instead, an appeal brief should make it easy for the designated judge to write an opinion reversing the Examiner’s rejections and to get the other judges on the panel to quickly agree.  A very nuanced argument that takes a considerable amount of time to digest is unlikely to be successful.


Top Image Source: Deposit Photos


Warning & Disclaimer: The pages, articles and comments on IPWatchdog.com do not constitute legal advice, nor do they create any attorney-client relationship. The articles published express the personal opinion and views of the author as of the time of publication and should not be attributed to the author’s employer, clients or the sponsors of IPWatchdog.com.

Join the Discussion

8 comments so far.

  • [Avatar for Eli Mazour]
    Eli Mazour
    September 13, 2018 09:49 am

    @JNG PTAB Judges do share some attorneys who act like clerks and help them write opinions. However, they do not have multiple personal clerks like Federal Circuit Judges. I’d be surprised if having clerks is why Federal Circuit judges write more dissents but that’s an interesting theory.

    @valuationguy I link to Gene’s article on that topic in my article. One surprising finding is the number of PTAB judges (55 – about 25%) who have written dissents in the 20-month period we looked at.

    @B Thank you!

    @Lost In Norway Thank you for the kind words about the article. If you shoot me an email ([email protected]), I can look into whether I can get you that data.

  • [Avatar for Former M.E.]
    Former M.E.
    September 11, 2018 09:28 am

    Lost in Norway: this is pure speculation, but dissents in 3700 could be technology bias. My experience is that a majority of patent attorneys that have non-mechanical educations believe anyone can prosecute mechanical cases. I suspect that this bias extends to APJs. I would like to see the technical expertise of the judges in the panels where there were dissents in 3700. It would be interesting to see if there is a difference in technical expertise between the majority and the dissent.

    I have received decisions in ex parte appeals for mechanical cases (although maybe not assigned to 3700, I don’t recall) where I thought the technical arguments in the opinion were off base. After a little investigation, I found the judge writing the opinion to have an undergraduate education that was something other than mechanical engineering (my recollection is biology or chemistry).

  • [Avatar for Lost In Norway]
    Lost In Norway
    September 11, 2018 02:44 am

    Thank you for the article, Sora and Eli. It was loaded with information that I had not really seen before. I had expected more 101 dissents, but as you said there are almost always 103 rejections as well.

    It was no surprise that art units related to chemistry (1700) and computer software (2100) and Ecommerce (3600) had a higher dissension rate. Anything chemistry related is always going to be complicated and anything that involves computer software or Ecommerce is always going to be a sore spot in today’s climate. But I was surprised that Unit 3700 – Mechanical Engineering, Manufacturing, and Products had such a high dissension rate.Anyone know why?

    Is there any way get the 101, 102, 103, 112 percentages for each of Art Unit 1700, 2100, 3600, and 3700? Just to satisfy my own curiosity.

  • [Avatar for B]
    September 10, 2018 10:28 pm

    @ JNG “In my experience in Federal Court anyway, Judges write very few opinions, that task is left to clerks – Perhaps a germane question to ask is whether these Judges have individual clerks or not . . . .”

    In my experience I don’t even think most CAFC judges officially authoring an opinion read the opinion. Further, some judges are totally ignorant of a fact pattern of an opinion they signed onto.

    BTW, great article. Very interesting.

  • [Avatar for Pro Say]
    Pro Say
    September 10, 2018 06:49 pm


    “Want patent? Got patent?”

    Here today.

    Gone tomorrow.

  • [Avatar for valuationguy]
    September 10, 2018 05:02 pm

    I would also note that I believe that Gene has previous written here that dissents at the PTAB are actively discouraged by the procedures of the PTAB….specifically, dissents REQUIRE the active permission of the Chief Judge….who (in the person of Ruschke) was well known to “promote” unanimity in decisions…even to the extent of removing ALJs from assigned panels and packing the panels to achieve unanimity….due process be damned….

  • [Avatar for desmond p]
    desmond p
    September 10, 2018 04:18 pm

    “From a practical standpoint,. . .[a] very nuanced argument that takes a considerable amount of time to digest is unlikely to be successful.”

    So, in essence, winning arguments presented by a competent attorney will not be successful unless the issues are so simple as to make it an easy call for the judges.

    I understand nuance being lost on examiners, but judges are supposed to be able to grasp and properly apply nuance in the law.

    I am now sadder than I was before.

    And to JNG, in my limited conversations with judges, I think the judge:clerk ratio is more like 10:1, and the judges with clerks are expected to produce more decisions.

    (most of this is hearsay and by no means the official view of any firm or the PTO)

  • [Avatar for JNG]
    September 10, 2018 01:28 pm

    In my experience in Federal Court anyway, Judges write very few opinions, that task is left to clerks – Perhaps a germane question to ask is whether these Judges have individual clerks or not…. I suspect there is not a 1:1 correspondence as in other domains. If that is the case, then no Judge is going to go out of his/her way to undertake the task of actually writing an opinion on their own. If I’m wrong about this, I’d like to hear from someone who knows.