2015 U.S. Patent Practitioner Trends

Guest Post by Zachary Kinnaird, Patent Attorney with International IP Law Group

We are currently in the midst of a noticeable downward trend in the number of new patent practitioners each year.  As recently at 2009, nearly 2,000 new patent attorney and agents earned registration numbers, however this has fallen more than 40% in just five years.  Based on registrations from this January, only 1,000 new patent practitioners are projected to register in 2015.

Registration timing data also shows:

  • A weak correlation between law school enrollees and new patent practitioners
  • A third of current patent attorneys were previously patent agents
  • The average time to convert from agent to attorney is slightly less than 3 years
  • Dramatic shifts in registration frequency around changes in the law and USPTO policy


In addition to the current downward trend, other interesting points include the roughly 55% decrease in registration numbers earned from 2003 to 2004.  Also notable is the doubling of new registration numbers earned from 1997 to 1998.  However, as noted in the methodology below, any data prior to 1998 may not be consistent with more recent data due to USPTO surveys and database updates.  Accordingly, fewer conclusions and points of interest can be identified for these earlier years.

January 2015 Below Average, Projecting Only ~1,000 New 2015 Practitioners


In the chart above, the averages for each month from 2005 – 2014 are shown with error bars showing the standard error for the number of new practitioners in the past 9 years.  Based on data pulled from the month of January 2015, a prediction can be made about this year’s total new practitioners.  As no satisfactory correlation is currently found between the number of new practitioners and any other identified factor, these predictions are made only by comparison to averages over similar time periods.

A simple proportion is used as follows:


This is of course a very loose estimation.  Based on the standard error of January months used in the average, it would not be surprising to see up to 1,112 or as low as only 787 new US patent practitioners in 2015.  These lower projections fit the recent downward trend seen year to year since 2009.  If this downward trend continues, I am interested to see its effect on the employment market for patent attorneys, patent firms, and patent educators.

Weak Correlation between Law School Enrollees and New Patent Practitioners


In this graph, the number of newly enrolled 1L law students as reported by The Wall Street Journal is compared to the number of new registration numbers earned each year.  Although at times there appears to be a weak correlation, overall there does not appear to be any correlation between the number of students attending law school and the number of new patent practitioners each year.  Indeed, the correlation coefficient in excel for these two trends was 0.066.

From the lack of a strong correlation presented by these values, one conclusion to draw is that the factors that convince a person to enter law school are different or are weighed differently than the factors that convince a person to pursue a career as a patent practitioner.

1/3 of Patent Attorneys were Agents First, Usually Converted in 3 Years or Less

With date of registration data, it is possible to find the number of patent attorneys who were previously agents.  Of the 43,064 practitioners listed, 13,232 were listed as a patent agent first.  This represents approximately 31% of the listed practitioners.

It is also possible to find the average time between these practitioners’ registration as agents and their later registrations as attorneys.  On average, the conversion time was 1039 days, or roughly 2 years and 10 months.  Since the primary requirement to convert a registration status from agent to attorney is passing a state bar, it is reasonable to conclude that this number is close to 3 years because of the typical three year duration of law school in the United States.

However, more interesting is the fact that this average value is just below 3 years.  One interpretation is that this figure suggests the group of practitioners who change from patent agent to patent attorney, on average, decided to pursue work as a patent practitioner only after entering law school.  Otherwise, this average might be longer than the average duration of law school, not shorter.

 Month To Month Registration Frequency Shifts with Changes in Patent Law

Fig5 Since the start of online testing, registrations are more evenly spread through the year.  However there are still outlier months, and recently these outlier months correspond almost perfectly to follow the timing of the phases of AIA changes were set to be added the patent bar.  The slight delay from implementation month is likely explained by processing times at the USPTO of registration paperwork after passing the patent bar.  Unsurprisingly, the months immediately following the large increases show dramatic drops in the number of freshly registered practitioners.

Further, these outlier months make sense both theoretically and personally.  I myself was part of the May 2013 outlier month, and scheduled my exam towards the end of the month in March 2013 – the final month before the third phase of AIA changes were to be tested.  After passing the exam, my paperwork and processing time at the USPTO resulted in a first registration date of 5/20/2013.  It appears I was not alone in strategically scheduling my exam to avoid the uncertainty of being tested on new law.

Dramatic Shift in Registration Timing after USPTO Shift to Year Round Testing


The above chart shows the number of new registration numbers earned every month from January 2002 to December 2007.  This range was chosen to highlight the effect of the USPTO switching from administering the patent bar twice a year to the year round method used today.


The USPTO provides both a database for practitioner information as well as a zip file of this data in spreadsheet form that is updated daily.  However, the spreadsheet provided by the USPTO does not include the dates of registration as an attorney or agent.  Accordingly, the process of retrieving the registration date information from the online database was automated to yield the data that is analyzed in this post.  As many practitioners have been registered as both patent agents and as patent attorneys, only the date an individual first received a registration number was used for this analysis.

Unfortunately, it is unclear how accurately the USPTO database reflects active U.S. Patent practitioners.  Dennis Crouch wrote about this issue in 2012 and also covered one of the Patent Office’s attempts to refresh its database to reflect a more accurate count of current practitioners.  In fact, as recently as October 2014, the Office of Enrollment and Discipline (OED) has conducted another survey for registration numbers 35,000-39,999 to update the information in its database.  For the curious, these most recently identified registration numbers correspond to practitioners who first registered between August 1991 and February 1996.  Due to these surveys, and the lack of more complete data, the following graphs and charts only represent the USPTO attorney and agent database as of January 31, 2015.

Future Analysis

In the future, I will take a look at the ratio of agents to attorneys on from year to year to see if the relative percentage of patent agents is increasing, decreasing, or does not follow a trend.  I also plan to flesh out the historical parts of this data by calculating yearly values of registration numbers from earlier years to compensate for the data removed by the USPTO through OED surveys.

54 thoughts on “2015 U.S. Patent Practitioner Trends

  1. 14

    Don’t have time to read through all these comments (busy with 3l year which I am not finding to be the relaxing time common wisdom says it is) so sorry if this has been stated before. With the downward trend in hiring of lawyers even in the patent law field (or the perceived one at least) the employment counselors at my school and I expect others recommend taking the patent bar either during your 1l summer or definitely your 2l summer. So I expect people who enter law school knowing they want to do patent law are not really practicing as patent agents (a summer law job can you really call a patent agent job?) but taking the bar early to look better on their resume for finding a job after graduation. That was certainly what I did. My school has also certainly seen an uptick in the number of law students interested in patent law – given I think scientists jobs are drying up as well. So though for my own selfish reasons as a 3l I hope this trend continues I am not sure it will.

  2. 13

    For what its worth, a lot of firms will have anyone who is qualified to take the patent bar take it, even if they are not patent prosecutors, as they think it makes them appear more qualified when doing the beauty contest for patent litigation. So we had licensing attorneys get reg numbers who were never going to practice with them.

    As patent litigation has waned, the number of firms doing this may wane as well. This could explain some of the drop in numbers: people who have no intention of doing prosecution are not getting reg numbers any more.

  3. 12

    I’m not sure how much of a factor it might be these days, but back in the day when I took the registration exam (1996), the degree of difficulty sometimes varied significantly from one test to the next. I recall a passage rate of about 35% for one test and over 60% for the next one. (I attribute swings of that magnitude at least in significant part to the test, not the test takers.) That could help explain some of the big swings, at least from that era.

  4. 11

    One reason there are so far fewer older registered PTO practitioners on the chart is the special “Geezer Removal”* program the OED conducted a few years ago to get old practitioners to either remove their names voluntarily, or to get involuntarily removed if they were dead, retired, or out of it by not responding to their mailed demand to old practitioners.

    [Actually not a bad idea.]

    *[that’s the name I gave it]

      1. 11.1.1

        Given the absolute fixed nature of the data set, lots of fun can be had with stats here.

        For example, running with the Paul note, and looking at the rough buckets, by assigning a designation of “Not Active” to each serial number missing from the active roster, we can see an interesting phenomenon.

        The “2.5” bucket is half empty.
        The “3.0” bucket is 80% full.
        The “3.5” bucket is 90% full.
        The “4.0” bucket is 95% full.

        The gen er al time frame may be seen to account for this.

        But this is where it gets interesting:

        The “4.5” bucket is 80% full.

        All of the remaining buckets are 95% full.

        It appears that there is a substantial anomoly centered around the specific “4.5” bucket, quite distinct from BOTH those registered before and those registered after.

        I wonder if the “purge” focus was centered on this particular group. Did the purge effort lapse subsequently? What – if anything – singles out those with reg numbers between 45000 and 50000 for this anomolous distinction?

      2. 11.1.2

        Yay! Look at Me! I’m anon! I’ll claim credit for an idea even though Morgan’s point about the Geezer Removal Program no where near “exactly the point I was making to Ned below at 1.1.1.”.

        Draw your attention toward me, everyone! I’m special!



          Lol – that’s some serious jealousy you have there my friend.

          Hmmm, the office seeking to remove older people from the rolls and the notion that age-related drop-off have nothing in common….

          Um sure, that’s the “extent” of your “logic” and you want to draw more attention to yourself on that point…

          Too funny / but not in the way that you think…


            “Hmmm, the office seeking to remove older people from the rolls and the notion that age-related drop-off have nothing in common….”

            Whoa there, goalpost boy!

            Cf. “have nothing in common” with “that’s exactly the point I was making to Ned below”.

            Exactly the point. LoL.

            Yay! Look at me!


              You say “goal post” and forget(?) the operative word of “move.”

              Or perhaps you realize that your l@me attempt at a reply would fail as there was no movement of goalposts.

              You really are not good at this. Maybe you should try looking up funny GIFs…?


                Your sauce is pitifully weak.

                Tell us again, tb, how is it that “have nothing in common” relates to “that’s exactly the point I was making”?

                We’re waiting.

                Weak, weak sauce.

  5. 10

    I am an outlier, for sure. Registered 15 years ago and never been to law school. What do people like me do to your 3 year average?

    At this point I doubt if law school would get me much in the way of extra remuneration, and certainly not as much as it cost … unless I did a cheap online law degree, but I’m not sure if that would increase my employability or reduce it, LOL!

    As for the stress levels, I do wonder if I should have taken this up atall. We have one guy here who often says he wishes he’d gone into plumbing instead! At least this is cleaner … I think.

    I await the next survey with baited breath.

    1. 10.2

      (But there is a big assumption that an agent will turn into an attorney)

      A very rough look points out that this may in fact not be so. I put the data into rough buckets by registered serial number and fairly consistently since the 5 series began roughly 30% of agents may not in fact choose to become attorneys.

      Again – this is a rough estimation based on a single snapshot in time, relying on listed status (which admittedly may not have been properly changed by agents who have become attorneys), but the shear number of agents with mid-level reg numbers does indicate a fair number do not proceed on the assumed path. If the three year turn mark is only for those who have turned, then I am uncertain how that mark takes into account those attorneys who were never agents at all (or even the respective sizes of those two groups).

    2. 10.3

      Alun, if you don’t mind, I would prefer your breath to be ‘bated rather than baited, and anon’s number to be sheer rather than shear (unless, that is, he’s alluding to the numbers caught in the shears of the Geezer Cropping program).

      1. 10.3.1

        A wise man once said:

        Although if you read this blog regularly, you’ll know I often avoid the normal rules of grammar, spelling, and punctuation.

        Guess who.

      2. 10.3.2

        I’m never sure what anon is alluding to.

        However, I’m confident that Alun meant he would be snacking on sardines.

  6. 8

    Ned, regarding the numbers of registrants, you may find this 2014 paper of some interest. Abstract (emphases my own):

    This article pursues two distinct, but related hypotheses. First, as total LSAT takers decline, we expect to see a decline in the number of new attorneys admitted to the patent bar. Second, as the number of new patent attorneys shrinks and the number of women pursuing engineering degrees increases, we expect that the patent bar will become more female.

    In order to test these hypotheses, we gathered and collated data from the Law School Admission Counsel (LSAC) regarding students taking the Law School Admissions Test (LSAT), the United States Patent and Trademark Office (USPTO), the Society of Women Engineers (SWE), and the American Bar Association (ABA). The data establishes that the first hypothesis is true, but that the second one is false.

    That is, the number of new entrants to the patent bar will drop precipitously. By 2018, new entrants will number one half of what they were in 2008. However, the number of female patent attorneys compared to the number of male patent attorneys will not change in this same time period. That is, even though the patent bar will shrink, the patent bar will not become more female.”

    Ken Port, Lucas Hjelle, and Molly Littman

    link to papers.ssrn.com

    1. 8.1

      I think DC is saying there is no correlation between law school LSAT takers and the number of new patent attorneys whereas this article is saying there is.

  7. 7

    Dennis –

    Your thoughts and analysis attention seem to be skewed a bit toward law and law school. I wonder if the reason for the trend might be an the economy as a whole. For example, if more engineers and scientists are finding gainful employment in industry, perhaps there is less motivation to look at careers that are tangential to their training, such as patent law.

    1. 7.1

      From people that I have run across this is true. Many of laid off engineers think maybe patents.

                1. That’s going to be a problem, since ALL technology (and all engineering) is nothing but “applied math,” the logic being used eviscerates all intentional innovation.


  8. 6

    “1/3 of Patent Attorneys were Agents First, then”

    Similar to my other comment, this number should be looked at with some skepticism given how many people take the registration exam as attorneys and become registered, but who had to spend some time getting the necessary certifications required by OED to prove (to OED satisfaction) that they were attorneys, despite their own jurisdiction having been long ago satisfied of their status as attorneys.

    1. 6.1

      Not only what you said, but people that get jobs at big law firms after graduation from law school often spend the summer taking the registration exam and the bar exam. The bar exam results take longer, so they are agents first and then become attorneys.

      In fact, this is very common.

      1. 6.1.1

        Also, my guess is that a lot of numbers have to do with the rise of the large IP boutique that pays for the registration preparation class and gives associates a couple of weeks off to study for the test.

      2. 6.1.2

        Sorry for all the posts, but also remember that for many, many people it is take registration exam and bar exam in the summer and then the law firms didn’t care if you upgraded to attorney from agent, so most people just let it stay at agent.

  9. 5

    The “interesting” observation between 2003 and 2004 was the direct result of the shift from a biannual, paper exam to an on-demand, electronic exam. If I remember correctly, after the 2nd regular administration of 2003, the test was not made available until late July 2004 (after many delays and false starts), and for those of us that passed early in the process, we didn’t receive registration numbers until November or so, despite the exam results being instantly available to the PTO. I suspect they were waiting for statistical validation or other correlation data to green-light the registrations. I guess those were interesting times. Problem was, I had taken the Kayton course way back in January, and was getting pretty antsy by August.

  10. 4

    “The average time to convert from agent to attorney is slightly less than 3 years”

    Assuming Mr. Kinnaird is trying to get this information from the current rolls made available by OED, he should be made aware that a substantial percentage of patent attorneys are quite tardy in getting the updates to OED.

    Some just don’t bother until they get around to it. Many others were delayed by certain OED directors who demanded their certificates in good standing be issued from “the highest court in the jurisdiction”, but who were barred in states that would not permit attorneys to swear in before the highest court without X years practice experience. My own state did not require any X years (2-4) but required an applicant for admission to the supreme court bar have recommendation letters from 3 current members of the bar and then get an appointment to be sworn in, which could take a year or more to accomplish.

    What this all means is that there are a substantial number of people in your dataset who were attorneys on the day they became PTO eligible, but were registered initially as “agents” due to external/artificial factors.

    1. 4.1

      Sorry, the “years (2-4)” above was intended to go as an example with the first iteration of “X years”, not the second where I inserted it above.

  11. 3

    I’m on-trend. I passed the patent bar after my 1L year, and converted to patent attorney almost exactly two years later.

    I always found that excellent patent practitioners popped up about 2-3 years after the engineering industry had a downturn. So the bump in 2011 was probably from the recession.

    The economy is picking up, and engineers and biotech folks are finding tech jobs, so fewer of them go the patent route. And that’s a good thing for those of us in the patent business. More creative folks create more work for us.

  12. 2

    Can someone ID the dude who was registered in ’37? We need to give him a parade or something.

    Second, when I first started, early ’80s, we counted the number of new registration numbers per year as listed in the gazette. It averaged just above 500, and this number was consistent, year in and year out as far back as we could go.

    It looks like something changed, radically, in ’93, where the number goes over 1000 for the very first time. What?

    1. 2.1

      Yes, I did do so, although now I’ve decided it may not be fair to publish the gentleman’s name online here, so I will not.

      The 3 next most senior people still on the roster were registered in 1940 and 1947, 1947, by the way.

  13. 1

    What is surprising is not that the number of new patent practitioners is (apparently) falling, but that it has been so high for so long. The field has been oversubscribed for about 15 years, and once the USPTO and Fed. Cir. finally internalize the Supreme Court’s recent jurisprudence and get that it is not just going to go away, we’ll have need of still fewer patent practitioners. I expect the field to revert to its pre-Fed. Cir. status as a relatively quiet backwater of the law in about 5 – 7 years.

    1. 1.1

      Egon, I am not so sure that the rise in agents per year that apparently took off in ’93 has anything to do with patent law per se. I think there must have be some change in the testing or admission rate that took place at that time.

      1. 1.1.1

        Took off…?

        Ned, you do realize that this data is from the current rolls and that there is an obvious – and fully natural – age related drop off for the pre-93 time-frame (20 plus years in law – plus a likely couple more for the engineering background, plus since this is a pretty lucrative gig, you are likely to have more early retirements…


            “Patent prosecution is lucrative?”

            Well, it was that for most, once upon a time. And it still is for some. It’s just that “some” is now a declining fraction of “most”.


          I think patent prosecution is incredibly stressful. I know one of our partners retired because of that (and because he had enough money to do so).

          I’d like to see statistics comparing rate of early death for patent prosecutors versus the normal population and versus attorneys in general. I bet you’d find a much higher early death rate for patent prosecutors (in law firms; may not be true in-house). The stress level is always high: you’re always behind; if you catch up even for one day, the next day, you begin to get behind; the amount of minutia in this job is shocking and incredibly difficult to get right (you looked at the name/serial number but didn’t realize two characters/numbers were reversed — that’s going to cost you); you have to work extra days to take off a week; to take off two weeks is very difficult, and any longer is impossible; more than two weeks vacation? It’ll never happen; the client always finds something wrong, which you then have to spend hours determining what to do, which of course isn’t billable. Need I go on?

          I’ve seen so many people drop dead early in this job.


            If you’re behind that much you need to either hire additional help or become more efficient.

            If you can’t hit your flat fees then you need to lower your billing rate, earn less money due to the higher write-off, or become more efficient.

            Or, you just have cr*p clients. I rarely take sh*t from clients unless they are high volume (or potentially high volume) or represent 6 figures of business.

            It’s really not difficult.



              I agree it is difficult if you are the associate and have no clients.

              On the other hand, if you are a partner it is not difficult and a pretty stress free job if you can get the rare high paying clients (e.g., in-house counsel is your old college buddy, a relative or you inherited the client at your firm who has a long standing relationship with the client) and you can get some dogs like Bob to do the work for you. Alternatively, a partner who earned the client based upon merit and cutting rates has stress and it is a difficult stressful job.


                Then move to a smaller firm with lower overhead and lower billing rates. It’s easy. I’ve never, ever heard of a client needing a large GP firm’s or a large boutique firm’s “global platform” w/r/t patent prosecution.


            If you have more work than you can handle, there are several options:
            1. hire someone to handle some of the load
            2. become more efficient
            3. raise the amount you charge. This is what happens in the free market — if there is a scarcity (i.e. you don’t have time to do it all) then the price goes up.

            Frankly, I think no. 3 is your best choice. You raise the billing rate by 10% and 10% of your clients say no thanks as a result. So you make just as much money as you do now for 90% of the work and go home to watch your grandkids soccer matches.

            Your numbers may vary and your mileage may vary, but that is the way most of the world works; why not attorneys?

Comments are closed.