Percentage of Patents Where Applicants Filed Disclosure Statements

I looked at a sample of 100,000 patents that issued in 2006 and 2007 to see what proportion of them include at least one applicant filed information disclosure statement (IDS) in the electronic file wrapper found in PAIR. About 83% of these recently issued patents include an IDS. The graph below breaks-up results roughly according to Art Unit at the PTO and sorts results according to the percentage patents that include an IDS in the file wrapper.

Because most of these categories include well over 1000 patents, the difference between individual groups is statistically significant (.99 CI) whenever the actual difference between two groups is at least four percentage points.

Evidence Based Prosecution: More Claims Filed Equates with Longer Time in Prosecution

I am looking generally at how the contents of a patent application may impact the prosecution process. In this study, I looked at a sample of 50,000 original patent applications (excluding continuations) that have issued as patents in the past few years. This first graph shows the average time in prosecution (filing to issue) as a function of the total number of claims in the original patent application. The result is that the claim count positively correlates with time in prosecution. This correlation is strongest when the claim count is less than 30. As an example, patent applications with 30 claims took about 33% longer to issue than those with only 2 claims. The correlation makes sense: more claims would generally mean more work for the PTO to reject each claim, and more work for the applicant to respond in kind.

The second graph shows a similar pattern for independent claims.

Appellate Decisionmaking: Unanimity or Dissent

The graph above is based on a review of almost 400 CAFC patent decisions that include a dissent. To calculate the percent of patent decisions that include a dissent, I pulled each decision associated with each judge and then added up the number of decisions that involved a dissent (regardless of whether the judge in question dissented). I am confident in these results, but a little more work is needed to handle the en banc decisions (which are more likely to involve a dissent). My RA Ryan Starnes helped in compiling the data.

Evidence Based Prosecution: Why are Deadlines Driving Response Timing?

  • I have compiled some data on the timing of initial non-final office action and the ensuing response for 400,000+ patents with patent numbers ranging from 5,800,000 – 7,100,000.  The graphs below show that office action responses are strongly deadline driven. About 60% of responses are filed by the initial three month deadline. However, 2/3 of those early responses are filed within only a couple days of the three month deadline. Similar “rushes” are seen to meet the four, five, and finally the six month date.

PTO Fees are the explanation for the deadline push. Missing the initial three month deadline requires payment of a  $120 fee – not a steep fine – but something that some clients are now refusing to pay. Subsequent monthly surcharges rise more dramatically: first $460 then $1050.  For clients on more of a fixed prosecution budget, attorneys would rather money go toward their fees rather than to the PTO.  Of course, the six month date has the significant legal ramification of abandonment.

The  immediacy of the deadlines hides an important fact in today’s prosecution: Each day of by the practitioner results in one less day that the patent is in force.  The fact that very few responses come quickly (apart from the deadline) indicates a latent demand for delayed prosecution — It seems that most clients do not care whether the patent issues now or six-months from now. Why not eliminate these deadlines or push them back considerably? Applicants with important or easily patentable inventions would continue to respond quickly.  Those who could care less or who need time to build a patenability case could wait. 

Now, there are several reasons to strive for compact prosecution timing. Perhaps most importantly is the notion that both the applicant and examiner will work better and more efficiently if they have seen the case recently.  The tradeoff, however, is that applications sit much longer awaiting initial consideration.

A note about the data below: Responses that appear below to be just after a three, four, five, or six month deadline are typically ‘on time.’ My measurement is from the date the OA is entered in PAIR to the date the response is entered in PAIR (not the mailing dates). This typically adds an extra week or so.


  • Jeff Steck, a patent attorney at MBHB LLP, examined the small ‘ripple’ in the histogram and explains that it can be explained by the fact that (most) OA’s are sent out on weekdays and responses are received on weekdays. This ‘weekend effect’ causes the ripple overlay. My original postulation that the ripple is caused by the PTO’s bi-weekly counts is not sufficient since the ripple clearly has a one-week wavelength.

Trends in Inventing: Design Patents

Most design patents are based on the work of a single inventor. Although the average “inventor count” per design patent is still under two a time series appears to show that the lone-mind story of inventing continues to shift to a collaborative effort. Two additional points of information: design patents associated with an assignee have a higher average inventor count as do design patents that list a law firm as handling the patent. Part of the story may be that patent attorneys are getting better at ferreting out the true inventor. However, I suspect that the trend does reflect a shift in the process of invention and product development.


  • This data is based on an analysis of all 300,000+ design patents issued since 1976.
  • I’ll redo this graph for utility patents. I suspect that the results will be similar – especially the continued trend toward more inventors per patent.
  • In response to a comment, I parsed the number of inventors by country of assignee. Outcome: Amongst the 10 countries with the most design patent holders, US assignees have the highest average number of inventors per patent. (Non-assigned patents are left out of this analysis). (Click on graph to expand)
  • In response to another comment, I looked at the patent with the most inventors. U.S. Patent No. Des. 563,280 has ten inventors listed. The patent covers the “ornamental design for a car” and is owned by the Chinese auto company Shenyang Brilliance Jinbei Automotive Co. The company manufactures BMW’s Chinese 3 and 5 series. (Image below).

What’s Wrong with Software Patents?

by James Bessen and Michael J. Meurer

[This post is the third in a series of posts based on empirical research in a new book, Patent Failure, by James Bessen and Michael Meurer. Read Part I and Part II]

Patents on computer programs, financial processes and business methods have been controversial at least since the 1960s. Surveys regularly find that computer programmers are opposed to patents on software by a wide margin. In what other field is the class of inventors so opposed to patents?

Is there a problem?

Some people contend that there is nothing particularly wrong with patents on software, arguing that "patent thickets" are not preventing innovators from entering software markets. But the latest evidence [LINK] suggests that patent thickets do, in fact, inhibit startups in software. More important, patent thickets might not be the only or even the most important problem with software patents.

Indeed, our evidence suggests an important problem of another sort: software patents are four times more likely to be litigated than are chemical patents; business methods patents are twelve times more likely to be litigated; finance patents are 49 times more likely. Moreover, the evidence also suggests that these patents have lower values than chemical patents, so these patents are not being litigated more because they are more valuable.

Other people admit that there are problems with software patents, but they suggest that this is only temporary: once judges and patent examiners understand this technology better, once they have become familiar with the prior art, etc., then the uncertainty about these patents will abate and litigation rates will go down. But the evidence shows that after a decade of issuing software patents in large numbers (over 200,000), the probability that a newly-issued software patent will be litigated is continuing to rise.

So it does seem that patents on software and related technologies at least have a particular problem with litigiousness. And this problem is central to the poor performance of the patent system generally. In the previous post we highlighted how litigation costs have substantially outgrown the profits that public firms receive from patents outside of the pharmaceutical and chemical industries. In 1999, 38% of the cost of litigation among public firms arose from lawsuits involving software patents; preliminary data suggest that this share has increased since then. Litigation over software patents is clearly a major factor in the poor performance of the patent system. So in a very real way, the overall performance of the patent system cannot be fixed unless the particular problems of software patents are also fixed.

Why are there so many lawsuits over software patents?

A variety of evidence leads us to conclude that software patents are involved in relatively more litigation because they are more likely to have "fuzzy boundaries." Statistical evidence shows, for example, that software and business method patents are much more likely than other patents to have their claim construction appealed to the Federal Circuit. Part of this tendency arises because of the nature of the technology and part arises because of the way the courts have treated this technology.

Our reading of the case law convinces us that patent law tolerates too many software claims untethered to any real invention or structure; in such a world clear boundaries are unattainable. When patent claims relate to actual devices or chemical structures, then their meanings can be interpreted by reference to those physical or chemical entities. However, when the words refer to abstract ideas, they are often subject to multiple interpretations and are therefore more ambiguous. For example, many people thought that "point of sale location" (in the famous E-Data patent) was computer industry jargon for that place in a retail store where transactions take place, formerly occupied by a cash register. When the Federal Circuit interpreted this claim, they decided that it referred to any location where an e-commerce transaction might take place, although it is highly unlikely that this is what the inventor had contemplated 17 years earlier. Thus a broadly worded invention for a kiosk in a retail store was read to cover a broad swath of e-commerce. Not surprisingly, this patent generated quite a few lawsuits.

Patent doctrines that might serve to prevent such fuzzy claims have been undermined. For example, the enablement doctrine has historically been used to keep patents from claiming much more than what was actually invented. Unfortunately, as a result of Federal Circuit decisions on software patents during the 1990s, these patents no longer need to provide computer code, a flowchart, nor any detailed description of specific operation in order to be enabled.

Fixing the problem

A lot of people have very strong opinions about how patent policy should or should not change regarding software patents. We wish we had such clarity, but we do not. We are convinced that the current treatment of software patents creates significant problems and that these are getting worse. But the problem is complex and fixing it will likely involve multiple changes in law and institutions.

Certainly, KSR should help and so would a stronger indefiniteness requirement. Additionally, it might help to restore a substantial enablement requirement for software patents so that these patents are restricted to claiming more or less what was actually invented and disclosed.

Possibly, a subject matter test might help. We confess we do not have a rule that cleanly distinguishes inventions using software that should be patentable from abstract processes that should not be patentable. Some people argue that any attempt to proscribe subject matter will only increase uncertainty and encourage avoidance through clever claim drafting. But the evidence suggests that the subject matter tests used following Benson and Flook did not, in fact, encourage excessive litigation during the 1980s, even though there was some evasive claim drafting. Litigation rates for software patents then were about the same as those for all patents. On the other hand, we doubt that a subject matter test by itself would be sufficient to fix the problems of software patents.

Ex Parte Reexamination Statistics II

Patent.Law088To get some sense of the PTO’s current ex parte reexamination practice, I looked at the 2000+ reexamination certificates that have been published in the Official Gazette since January 2000.  There are three potential outcomes for any particular reexamined claim: validity affirmed; validity affirmed with an amendment; and cancelled. In addition, claims may be added. 

Outcomes: From my sample, 71% of the completed reexaminations resulted in claims that were either amended or cancelled while only the remaining 29% resulted in all claims being confirmed. A much smaller number – 12% – of the reexaminations resulted in all claims being cancelled. 

The middle ground: 59% had at least one independent claim confirmed while 50% included at least one amended claim.  For the 34 completed design patent reexaminations, the single claim was confirmed as valid 16 times and cancelled 18 times.

Timing: The reexamination certificates in my sample had an average pendency of 33.5 months (median 28.5 months).* The PTO’s reported average and median are 24.0 months and 18.6 months respectively. The difference apparently comes from their inclusion of reexaminations from the 1980’s and 1990’s which had much lower pendency than more recent numbers. [PTO Data 1981-2007]. 

Outcome & Timing: Reexaminations that simply resulted in a confirmation that the the initial claims are valid were resolved more quickly (averaging 24 months) while those resulting in cancellation of all clams were much slower (averaging 45 months). Cases at the longer end of the spectrum often involved either BPAI appeals or lost files.


  • Thanks to several Patently-O readers for showing me how to find reexamination certificates. I was able to use the Westlaw database for this search. (Remember law school and the free academic access). The data can be compiled from the PTO OG, but that would be a larger job. The PTO does maintain a list of all the issued reexamination certificates here.
  • * I roughly calculated pendency by measuring the elapsed time from publication of a reexamination request to publication of the reexamination certificate.
  • Reexamination Statistics I

Certificate of Correction

According to PTO figures, the Office has issued around 25,000 certificates of correction on issued patents each year since 2001. On a per patent basis, about 15% of issued patents are associated with a certificate of correction.*

Generally, a certificate of correction can be used to correct a minor mistake such as a “clerical or typographical” error or other minor error. In Superior Fireplace, the court held that a certificate of correction that broadens the claim is only permissible when it is clear from the specification and prosecution history how to correct the error. Similarly, in Central Admixture Pharmacy the CAFC found that the correction from osmolarity to osmolality was void because the error “was not clearly evident to one of skill in the art and the result was to broaden the claims.” The relief for an improper certificate of correction is simply to cancel the certificate. 

Over the past several years, a number of automated and out-sourcing services have emerged to check against PTO printing errors and other typographical errors. What are the best services and best practices regarding corrections?


  • * To account for the time delay in obtaining corrections, I looked at only patents numbered 6,000,000 – 7,000,000 for the per patent estimate.

Ex Parte Reexamination Statistics I

PatentLawPic332Through its attorneys Microsoft requested ex parte reexamination almost thirty Avistar patents. In recent days, the PTO denied Microsoft’s request in 14 of those cases.

Standing alone, this denial does not appear to be big news — except for the fact that over 95% of requests for ex parte reexamination are granted. In FY 2003–07, for instance, the PTO reviewed 2389 requests for ex parte reexamination and only denied the request in 105 cases. (4.4%). (At this point, there is some possibility that the denials are procedural  as the documents are not yet available on PAIR).

To be granted, a reexamination request must present a “substantial new question of patentability” that affects at least one claim of a patent.  The request must be based on a prior art patent or publication.  Prior public use or sale are not admissible as evidence in an ex parte reexamination proceeding (although affidavits explaining the prior art publications is admissible).

In general, there has been a rise in requests for ex parte reexamination. As the chart below demonstrates, increase is entirely due to an increase in requests for reexamination that are filed by third parties.


A First Look at Who Files Provisional Patent Applications

Patent.Law083As part of the 1995 patent law overhaul, the USPTO began allowing patent applicants to file provisional patent applications.  Over a decade later, these lower-cost provisional filings have taken hold.  According to PTO annual reports, over 132,000 provisional patent applications were filed in fiscal year 2007. In perspective, that number is over 30% of the number of the 439,000 non-provisional utility patents filed during the same period. This proportion has been slowly rising since 2002 when the provisional applications filing rate was about 27% of the non-provisional rate.[1] That year (2002), the PTO recorded just under 90,000 provisional applications and 332,000 non-provisional patent applications.

Interestingly, in my study of recently issued patents, only 21% reference a provisional application as a parent. (In the study, I look at approximately 15,000 utility patents issued in April and May 2008.) [2] 

Patent.Law085National Tool: Over half of the recently issued patents that listed an assignee indicated that the assignee was a foreign (non-U.S.) corporation or agency.[4]  Although foreign entities are not prohibited from filing provisional applications, the provisional tool was designed to benefit U.S. entities. Thus, it is not a surprise that only 5% of the patents assigned to international applicants were associated with a provisional application while 30% of the patents assigned to a U.S. applicant were associated with a provisional application.  Two countries – Israel and Canada – stood out as filing the highest proportion of provisional parent claims. Both of these countries are known for having patent attorneys with a high level of familiarity with U.S. laws.  Only 2% of the Japanese & Korean patents included provisional parent claims. [Updated June 03 with Corrected Figure]

The provisional application provides a potential extra year of patent eligibility at the end of the term.  Thus, it is also not surprising that new drug inventions – where a potential year at the end of the term is most valuable – have the highest rate of association with a provisional application.[5]  Likewise, patents on electrical and electronic applications had the lowest rate of provisional filing even after excluding the international applications.

The provisional filings appeared to have almost no impact on the pendency time of a patent application as measured by the number of days from filing the nonprovisional to issuance.  This makes sense as no examination takes place until the nonprovisional application is filed.

The following table also provides some interesting comparisons of patent strategy.[6]

Type of Technology< ?xml:namespace prefix ="" o />

Number of Utility Patents in the Sample

Number of Patents Referencing Provisional Parent











Drugs & Medical












Boston Scientific Scimed





Computers & Communications

























Bristol-Myers Squibb




Procter Gamble




3M Innovative Properties




EI du Pont de Nemours




General Electric[7]






[1] This information comes from PTO annual reports for FY 2006 and 2007.

[2] These patents were downloaded on May 30. I excluded the few patents filed prior to the 1995 introduction of provisional patent applications.

[3] In a follow-on study, I will look at published patent applications claiming priority to provisional applications to get some sense of how often folks abandon provisional patent applications.

[4] 6823 were assigned to foreign entity while 6457 were assigned to a U.S. entity. Another 1922 had no listed assignee.

[5] My study included 677 patents having a primary U.S. Classification in a “drug” field and also assigned to a U.S. entity. Of those, 48% were associated with a provisional application.

[6] Note, my assignee ‘scrubber’ is not yet perfect. Thus, it is likely that some patents associated with listed assignees are not included in the table.

[7] Some companies, such as GE hold patents in several different categories – this table looks only at those patents in the particular identified category.

Theory of Dependent Claims: Survey Results

Patent claims define the scope of an invention much like metes and bounds define real property boundaries or a statute delineates the difference between criminal and non-criminal activity.  Often, patent applicants submit a series of claims all directed toward the same invention. Various independent claims allow an applicant to focus on various aspects of the invention. These broad aspects may be directed to various nodes of a larger system or regions of a compound, processes for creating or using various elements, etc.  Dependent claims by definition add additional limitations. Thus, for instance disputed claim 1 and 2 from the recent Lovanex dispute reads as follows:

1. A heterogeneous intimate admixture of sulfated heparinic polysaccharides … consisting essentially of

from 9% to 20% of polysaccharide chains having a molecular weight less than 2,000 daltons
from 5% to 20% of polysaccharide chains having a molecular weight greater than 8,000 daltons, and
from 60-86% of polysaccharide chains having a molecular weight of between 2,000 and 8,000 daltons ….

2. The heterogeneous polysaccharide admixture as defined by claim 1, comprising less than 2% of dermatan sulfate.

Claim 2 is interpreted to include all the limitations of claim 1 with the additional limitation of “less than 2% of dermatan sulfate.” See 37 CFR 1.71(c).

Earlier this week, I polled Patently-O readers on why they file dependent claims and received just over 1000 responses. (Listed below).

Theory of Dependent Claims            
Why do you file dependent claims?             
  Percent of Responses  
Answer Options Strongly Agree Mostly Agree Neither Agree nor Disagree Mostly Disagree Strongly Disagree Rating Average
Dependent claims are backup in case the independent claim is rejected in prosecution. 59% 34% 4% 2% 0%        1.51
Dependent claims are backup in case the independent claim is found invalid in litigation. 59% 31% 6% 3% 1%        1.56
Claim Differentiation: The dependent claims expand or alter scope of other claims. 47% 37% 8% 6% 3%        1.82
Dependent claims focus on particular commercial embodiments to make infringement easier to explain to a jury. 29% 44% 21% 4% 2%        2.07
I primarily include dependent claims with nonobvious additional limitations.  19% 45% 25% 9% 1%        2.29
I was taught to draft dependent claims. 26% 34% 26% 8% 7%        2.35
USPTO fees: Patent office does not charge extra for first twenty-claims. 26% 34% 24% 9% 7%        2.36
Dependent structure better supports foreign filing. 7% 19% 58% 11% 3%        2.84
Dependent claims are easier to draft and manage than independent claims. 6% 23% 37% 23% 11%        3.08
Client demands a certain number of claims. 6% 16% 29% 25% 23%        3.42
Dependent claims help avoid restriction requirements. 2% 13% 39% 31% 15%        3.44
A longer application is more valuable. 2% 13% 29% 34% 22%        3.60
          answered question 1029
          skipped question 0

Responses were also allowed to add additional reasons in the comment section. I have compiled a few of those individual responses below.

  • Drafting Process: A few responses commented on the iterative process of drafting a patent. For some, drafting the dependent claims helps refocus and redefine potentially problematic terms from the independent claims. For many, the dependent claims also serve as a guide for fully drafting the specification.
  • Prosecution Process: Some responses proposed that the dependent claims are helpful during prosecution for several reasons. First, there is a perception amongst patent practitioners that examiners read and consider the claim-set before looking at the actual specification. For that reason, the claims should be considered as a way to teach the examiner about the features and purpose of the invention.  For some, obvious limitations in a dependent claim are useful in that regard to show the examiner an intended use (practical embodiment) of the invention.  For those who see patent prosecution as a negotiation, a series of narrowing dependent claims speed up the the process — the applicant and examiner need only agree upon how narrow the claims need be.  Some propose at least some dependent claims being “super narrow” to serve as a “foot in the door” to increase the likelihood that there is at least some patentable subject matter.  Once you have agreed that a patent will issue, it may be easier to debate the scope of the eventual claims. On a practical note, the existence of dependent claims means that an applicant can get to a post-final rejection allowance without filing a request for continued examination. These practical concerns are important considering that the vast majority of patent applicants amend their claims during prosecution.
  • Claim Differentiation: Most practitioners agreed that the claim differentiation benefit of dependent claims is important. The doctrine encourages courts to consider that each claimed element covers something different than the others. Pushing that line, a typical reasoning is that two or more dependent claims providing alternative implementations of a term from an independent claim will help ensure that the independent claim element is not limited to a single embodiment. Some folks quibbled with my statements regarding dependent claims “expanding scope” of other claims and instead used a more PC version of “aiding construction.”  It is important to realize that the originally filed dependent claims also form part of the specification and can thus help overcome enablement and written description hurdles.
  • Appeasing Clients: Many inventors see their invention narrowly — as their particular embodiment. Dependent claims provide an easy way to satisfy the inventor’s desire to claim their invention without limiting legal rights. On this “politics” side — dependent claims also give reason to add joint inventors, which may be good for office politics. “An invention may only warrant one independent claim embodiment, but many clients like to see many claims. They believe the more claims they have the stronger the patent and that they are getting value for there money.So you add them, e.g., a blue widget, a red widget. It is like chicken soup for a cold, it can’t hurt.”
  • Ease of Analysis: “Filing dependent claims rather than only independent claims makes it much easier for several parties (attorneys prosecuting the application, examiners and the BPAI examining the patent, attorneys analyzing the issued patent, courts analyzing the issued patent) to more readily evaluate the differences and similarities of claims.”  Others noted that it is also useful as a way for potential licensees to know that the proposed product is literally and explicitly covered. For some patent owners, this “ease of analysis” is actually a negative. (Try to read NTP’s $600,000,000 claims). A long claim-set makes it more likely that a potential defendant will easily cave rather than pursue an expensive invalidity / noninfringement study.
  • Capture “minor inventions”: Dependent claims allow an applicant “to capture improvements to the invention that are not important enough to justify a separate application, but that one would want to avoid having a competitor step in and patent something similar.”
  • Tradition: There is nothing wrong with tradition and custom. It helps ensure that we’re all on the same page.  For many the 3/17 rule of independent/dependent claims serves as a default rule that can be adapted for unique circumstances.
  • International: Internationally, patents use fewer independent claims than in the US. One European attorney made the following comment: “Multiple independent claims in the US style cause major headaches and costs in Europe, so one good independent, with well gradated dependents, is the best solution for me.”
  • I will attach the full comments once reformatted.

Extending the Patent Term: Most Patents Are Extended Due to PTO Delay

ScreenShot04435 U.S.C. 154(b) provides for an adjustment of the standard patent term in cases where a patent’s issuance is delayed because of delay at the U.S. patent and Trademark Office.

I looked at all 6155 original utility patents issued between March 1 and March 13, 2008 to see how often patent terms are adjusted. In this sample, the vast majority of issued patents have an extended patent term due to Patent Office delay. Specifically, 72% of the issued patents include at least some 154(b) extended patent term. Among those with an extended term, the average extension is 392 days. A few patents from the group have very long extensions that approach five years. These are generally explained by (1) secrecy orders and (2) BPAI reversals. (See Patent Nos. 7,342,534 and 7,339,604). 

The first chart below is a histogram of the the patent extensions. As can be deduced from above, 28% of the issued patents have no extension. Slightly over 10% of the issued patents were granted an extension of three months (0.25 years) or less. A cumulative frequency chart of this same data would show that just over 50% of all the issued patents have a term extension lasting more than six months.


Of course, the extension term is correlated with the length of time in prosecution. The following scatter plot shows this relationship quite well. The demarcated boundary on the plot is an indication of the potential extension when the entire delay is attributable to the PTO.


Other studies show that primary examiners are “quicker” — i.e., that patents examined by a primary examiner issue more quickly and with fewer amendments than those examined by assistant examiners.   Here, I also found a difference between the type of examiner.  As may be guessed, issued patents associated with an “assistant examiner” are much more likely to have an extension. (78% of patents with assistant examiner have extension compared with 69% of patents examined only by a primary examiner). It appears that at least some of the quickness is due to the examiner actually acting more quickly (rather than merely compromising more easily).  Similarly, continuations are much less likely to receive extensions than patents without any ancestral history.  (78% of patents with no ancestry have extension compared with only 61% of patents that reference one or more parent).  That result is explained by the PTO’s procedure of reviewing continuation applications on a priority basis.  Combining these numbers, I find a recipe for delay: In this study a patent with no ancestry that is examined by an assistant examiner has an 83% chance of received extensions due to PTO delay.

As with every measure of patents, the results vary dramatically according to the area of technology. Here, we see that Computer & Communication related patents (according to classification) are much more likely to receive a term extension due to PTO delay and, when an extension is proper, those same patents have a much higher expected extension duration.

Category Percent Receiving Extension Count of Patents Average Extension (Days)
Chemical 73% 606 343
Computers & Communications 82% 1690 544
Drugs & Medical 74% 536 387
Electrical & Electronic 64% 1369 270
Mechanical 67% 900 268
Others 65% 713 297

There are several potential practical implications of these findings: (1) a prosecuting attorney should be careful to avoid delay because applicant delay will naturally push-back the issue date and additionally reduce the eventual patent term; (2) In many cases, quicker issuance is preferable to an equivalent extension. However, it is important to consider whether delay is preferable for your case. (Of course taking affirmative steps to increase delay could be interpreted as a form of laches.

Do Patents Stimulate R&D Investment and Promote Growth?

James Bessen and Michael J. Meurer have authored a new book presenting a careful empirical analysis of whether patent rights encourage innovation. Their conclusion: for the most part, today’s patent system does not achieve its stated goal. I asked them to provide Patently-O readers with a cut from their analysis. The following post is the first in a series of four by the authors. The book is titled: Patent Failure and published by Princeton [Web Link]. The analysis done in this book put Bessen and Meurer at the forefront of leaders in economic analysis of patent laws. I suspect that their results will become the talking points of the next round of patent reform discussions. A live symposium will be held at UGA Law School on March 29 focusing on the book and its results. [LINK]


by James Bessen and Michael J. Meurer

As background material, this post reviews the sizable body of empirical research analyzing the impact of patents on R&D investment and economic growth. Three future posts will present new empirical research featured in our book Patent Failure. The theme across all four posts is that patents often fail to perform effectively as property rights. [DDC: To work well as property, the right should be predictably valid; have discernible boundaries; and not have an overly fragmented ownership scheme.]

Economists cherish property rights that provide strong incentives for investment and trade, and that thereby contribute to economic growth. Potentially, patent rights could accomplish these three goals, and surely they sometimes do. Apparently though, it is hard to set up and maintain a patent system that works as property.

The rise of new market economies and strengthening of property rights around the globe in the last two decades provides economists with “natural experiments” that help us evaluate whether and how much property rights contribute to investment and growth. The empirical results are impressive. Countries that expanded the role of markets and strengthened property rights have prospered from these choices. Economic historians find the same results hold going back to the Industrial Revolution.

Comparable studies of patent systems are discouraging. The evidence certainly is consistent with the notion that patents encourage American pharmaceutical R&D. But otherwise, it is hard to find evidence suggesting patents are a major factor spurring R&D investment, that patents contribute to economic growth, or even that the patent system is a source of great wealth to important inventors and innovators (outside of a few industries like pharmaceuticals).


Claim Construction Reversal Rates II – District Court Judge Experience

    By David Schwartz

How do district court judges with varying levels of experience perform on claim construction? Are judges more likely to have their decisions affirmed when they have previous claim construction experience? Yesterday I provided some background on the large database I compiled and some overall results. [LINK] Today’s post provides an analysis of the data based upon the number of previous appeals of claim construction. A draft of the paper, Practice Makes Perfect? An Empirical Study of Claim Construction Reversal Rates in Patent Cases, can be downloaded here.

If district court judges improve after appellate review of claim construction, one would expect that the reversal rate would decrease as the number of appeals increases. More specifically, a district court judge with more prior Federal Circuit feedback should have a lower reversal rate than a district court judge with less experience. Assuming that the judge decided the second case after learning of the decision of the Federal Circuit in the prior case, the second time a particular judge is before the Federal Circuit, he or she should be more likely to have his or her claim construction affirmed. Figure A below illustrates the reversal rates of district court judges broken down by the number of prior claim construction appeals.


The left-most set of bars represents the results from every judge’s first appeal, with the blue bar indicating the percentage of cases that had to be reversed or vacated due to an incorrect claim construction, and the red bar indicating the percentage of cases with any erroneous claim construction. Moving to the right, the pairs of bars represent the results from each judge’s subsequent appeals.

These results suggest that district court judges do not improve as the Federal Circuit reviews their cases. Contrary to conventional wisdom, the reversal rate does not appear to decrease when a district court judge has multiple decisions reviewed by the Federal Circuit on claim construction. In fact, the highest reversal rate is for judges with four prior claim construction appeals. Other than judges who have been appealed exactly four times, the range is very narrow, between 26.8% and 30.5% for reversals and 31.0% and 40.4% for errors. Thus, there does not appear to be significant expertise gained by district court judges via direct Federal Circuit review that causes the claim construction reversal rate to decrease. (As discussed in more detail in the paper, the study is subject to several limitations inherent in studying appellate cases, including for example, a potential selection bias and a potential distortion if the cases are not distributed evenly across the district court judges. Ideally, for research purposes, cases would be randomly assigned to a judicial district (i.e., no forum shopping), and a random subset of those cases would be appealed.)

Up next: Reversal rates based performance by district court judges after a first reversal by the Federal Circuit.

I welcome comments from the readers of Patently-O.

Claim Construction Reversal Rates I – Overall Reversal Rates

By David Schwartz

How do district court judges with varying levels of experience perform on claim construction? To study that question, I compiled a large database that includes every single post Markman Federal Circuit decision involving claim construction. Others have studied this issue from the perspective of the Federal Circuit — i.e., what is the overall reversal rate. However, until now, no one has analyzed how district court judges perform based upon experience. My findings do not reveal any correlation between various measures of experiences and the likelihood of being affirmed. There will be several posts on the study and results. Today, I will provide some background on the dataset and a bit on the reversal rate. A draft of the paper, Practice Makes Perfect? An Empirical Study of Claim Construction Reversal Rates in Patent Cases, can be downloaded here.

The Dataset. First, a few brief words about the dataset. The dataset includes all decisions in which the parties disputed an issue of the district court’s claim construction. It includes all decisions in an eleven year period, from April 23, 1996 (the date of the Supreme Court decision in Markman) until June 30, 2007. Those decisions – 952 of them – are all of the published and unpublished opinions, as well as all of the summary affirmances under Federal Circuit Rule 36.

Overall Reversal Rates. For the entire eleven year period, 38.2% of cases had at least one term wrongly construed. Moreover, 30.0% of the cases had to be reversed or vacated because of an erroneous claim construction. These results are largely consistent with those of an earlier study (by then-professor and now-Federal Circuit judge Kimberly Moore).

Reversal Rates by Judicial District. In later posts, I will present the reversal rates based upon the experience of the district court judges. Today’s post presents the reversal rates aggregated by judicial district. Table II below identifies the number of Federal Circuit appeals from the most active judicial districts in terms of claim construction appellate decisions from April 1996 until June 2007. It also provides the percentage of cases that were vacated or reversed due to an erroneous claim construction. In the last column, I have provided the total number of patent lawsuits filed in those judicial districts, and in parenthesis, the rank of the judicial district if measured by the total number of patent lawsuits filed.

Table II: Most Active Judicial Districts on Appellate Claim Construction Experience: 1996-2007


Judicial District

# of Federal Circuit Claim Construction Appeals (1996-2007)

% of Claim Construction Appeals Reversed or Vacated because of Claim Construction Error

# of Patent Lawsuits Filed (1995-2005)


N.D. Cal.



2613 (1)


C.D. Cal.



2260 (2)


N.D. Ill.



1509 (3)


D. Del.



1112 (5)





1184 (4)


D. Mass.



782 (7)


D. Minn.



743 (8)


E.D. Mich.



669 (9)





952 (6)


E.D. Va.



555 (14)


N.D. Tex.



591 (11)


S.D. Tex.



466 (19)


W.D. Wisc.



232 (36)


W.D. Wash.



475 (17)


D. Col.



407 (23)

As is evident from the Table above, several of the busiest districts have reversal rates above the average rate of thirty percent. For example, the second busiest district in the country, as measured by either number of appeals or number of patent lawsuits handled, the Central District of California, has the highest reversal rate of claim construction. (Some may wonder why the Eastern District of Texas is not included in Table II. The trend of filing cases in the Eastern District of Texas began fairly recently, and as a result, many of the cases from the Eastern District of Texas have not had time to proceed through appeal. Consequently, the Eastern District of Texas is not one of the top fifteen districts in terms of appellate claim construction experience during the period 1996-2007.)

Up next: Reversal rates based on the number of claim construction cases previously appealed (i.e., the first case appealed from a particular district court judge, the second case appealed from a district court judge, etc.)

Comments are welcome!

Counting Design Patents


In 2007, Samsung Electronics received over 550 design patents — the most ever issued to a single company in one year.  Sony holds the most design patents, and is followed closely by Nike. The following table shows the company awarded the most design patents each year:

ScreenShot004Louis Zarfas is the primary examiner associated with the most issued design patents. Mr. Zarfas has allowed over 16,000 design patents since his first (as a primary) in 1978. Over 1,000 of those design patents relate to shoe designs claimed by companies such as Nike, Reebok, AVIA, Asics, LA Gear, Rockport, Sketchers, Wolverine, Keds, Louis Vuitton, Timberland, Berluti, and Kangaroos.

Recently issued design patents (issued 2000–2007) were, on average, pending for 16.2 months. Companies with at least 100 design patents during that period had about one month less pendency than those with fewer patents. Nike has prosecution down to a science — and averages less than 9 months pendency (filing to issuance). Toward the other end of the pendency chart, Apple averages over 21 months.

During the 2000–2007 period, approximately 75% of examinations were handled by primary examiners without any assistance from an assistant examiner. As with utility patents, the pendency for cases without an assistant is significantly less than for those with an assistant. Here, the difference is about two months.

 * These numbers are based on a recently compiled dataset of 300,000+ design patents issued since 1976 on file with DDC.

Does Size Matter? Counting Words in Patent Specifications

The following chart does not tell us whether the size of the patent document makes any difference. It only tells us that the average patent has grown considerably over the past twenty years.


To create the chart, I obtained a sample of 10,000 patents issued between January 1977 and December 2007 and counted the number of words in the description of the claimed invention. This excludes claims, title, abstract, references, and other identifying information.  To amplify the results, I added two trend-lines. The first trend-line runs from 1977 to 1987 and has essentially no slope — indicating that description size remained steady over those years. The second trend-line runs from 1987 – 2007 and has a clearly positive trend-line indicating that the number of words is increasing over time.

Reestablishing the Doctrine of Patent Exhaustion

< ?xml:namespace prefix ="" v ns ="" "urn:schemas-microsoft-com:vml" />By Mark R. Patterson* [PDF Version]< ?xml:namespace prefix ="" o ns ="" "urn:schemas-microsoft-com:office:office" />

Quanta Computer, Inc. v. LG Electronics, Inc.[1], presents facts at the intersection of two legal rules. On the one hand, patentees are free to impose restrictions, such as field-of-use restrictions, when they license others to manufacture their patented products.[2] Such restrictions can be permissible even if they would be antitrust violations outside the patent context.[3] On the other hand, a patentee cannot restrict the use of its patented products once they are sold, whether the sale is by the patentee itself or by a licensee.[4] This is the exhaustion, or first-sale, doctrine. [Background on Quanta v. LGE]


The Federal Circuit has caused the first of these rules to swallow the second, by allowing a patentee to convert any sale into a license by imposing some sort of restriction in the transfer. The restriction makes the transfer a “conditional sale,” and the Federal Circuit has held that the exhaustion doctrine applies only to “unconditional sales.”[5] It has applied this rule, moreover, despite the Supreme Court’s application of the exhaustion doctrine to conditional sales.[6]


The Federal Circuit’s approach allows patentees to transform a wide range of otherwise permissible conduct into patent infringement. All the patentee needs to do is forbid that conduct in its “conditional sale” arrangement. The Federal Circuit itself suggested in Mallinckrodt that this could allow patentees to eliminate the right of repair, and indeed other courts have applied the approach to allow patentees to eliminate not only the right of reuse/repair,[7] but also the right to resell patented products.[8] More generally, as the American Antitrust Institute argues in its amicus brief, the elimination of the exhaustion doctrine leads to considerable uncertainty, as downstream purchasers cannot know whether their use of the product is permissible unless they ensure that upstream sellers were in compliance with any license restrictions.[9]


It seems probable that the Supreme Court will reverse or at least vacate the Federal Circuit’s Quanta decision. It is likely to hold that sales of patented products exhaust the patentee’s patent rights, as the Solicitor General and other amici argue that it should. The patentee could still impose limitations on buyers’ uses of the products, but those limitations would be solely matters of contract. They could not be enforced through patent infringement actions, and they would be subject to antitrust law limitations.


But this raises two questions. First, will the Court define the difference between a sale and a license for this purpose? Quanta and the Solicitor General appear to take the position that the distinction should turn on the transfer of title. A problem with that approach is that it would give patentees considerable opportunity to use formal differences in the transaction to alter what should be substantive rules. (The Solicitor General may be unconcerned about this. It discusses with apparent approval Mitchell v. Hawley, 83 U.S. 544 (1872), where the patentee arguably forbad its manufacturing licensee from selling the patented inventions, granting only the right “to license to others the right to use the said machines.” The Court in Mitchell allowed an infringement suit against the downstream users, though the case is somewhat peculiar because the infringement arose after the original patent term had been extended, when the license terms extended only to the end of the original patent term.)


An alternative sale-license distinction might focus on what is being transferred by the patentee. As the district court argued in Mallinckrodt, one can read the Supreme Court’s decisions to allow restrictions on manufacturing licensees but not on ultimate purchasers. This echoes a distinction drawn in < ?xml:namespace prefix ="" st1 ns ="" "urn:schemas-microsoft-com:office:smarttags" />Europe, where the EC’s technology transfer block exemption applies to exempt from antitrust scrutiny certain “technology transfer agreements entered into between two undertakings permitting the production of contract products.”[10] Sales of the products after they are produced are governed by another block exemption.[11] Under this view, it is in the manufacture of the product that the patentee’s technology is first used, and its rights exhausted.


One effect of the exhaustion doctrine is to make price-discrimination more difficult for the patentee. For example, if the patentee would prefer to sell at different prices to different users, an inability to enforce its patent rights downstream would make it more difficult for the patentee to prevent arbitrage. But the Supreme Court has not hesitated to place other practical limits on price-discrimination by patentees, as in disallowing tying arrangements, which prevents metering through sales of the tied products (though only for patentees with market power). Moreover, even when its patent rights are exhausted, a patentee can enter into contracts forbidding resale for arbitrage, but it can only enforce those contracts through breach-of-contract actions, not through patent infringement suits.


The second question is whether the exhaustion rule applies if, as in Quanta v. LG, the product sold is only a component of the patented invention, in that it does not itself satisfy all the claim elements. The Supreme Court has said yes, at least in some circumstances: “[W]here one has sold an uncompleted article which, because it embodies essential features of his patented invention, is within the protection of his patent, and has destined the article to be finished by the purchaser in conformity to the patent, he has sold his invention so far as it is or may be embodied in that particular article.”[12] This rule recalls, though is perhaps somewhat broader than, the contributory infringement doctrine. That is, if this were the rule, the exhaustion doctrine would apply where there has been an authorized sale of a product that would contributorily infringe if its sale were unauthorized.


But the Univis Court followed the statement quoted above with another: “The reward [the patentee] has demanded and received is for the article and the invention which it embodies and which his vendee is to practice upon it.” Id. at 251. The question in Quanta v. LG can be viewed as turning on whether this second statement from Univis is one of law or of fact. If it is interpreted as one of law, then the Court is saying that the patentee must get its returns in the sale of the component invention. If the statement is one of fact, then the Court may just be relying on a view that the defendant in Univis had in fact gotten the return to which it was entitled in that first sale. In other cases with different facts, the patentee might be able to use patent law to enforce downstream restrictions despite the upstream sale of a component of the patented invention.


For example, one can imagine cases in which the maker of the component at issue, like Intel in Quanta v. LG, would have concerns about contributory infringement claims and therefore would seek a license from the patentee. That seems particularly plausible if the component at issue could be resold, so that even if its maker (Intel) ensured that its customers had licenses from the patentee, others to whom they might resell the component might not. It could make sense for the patentee to grant a license to the component maker in this situation, and make the license apply to downstream purchasers, rather than to enter into licenses with each of those downstream purchasers. But perhaps there could be circumstances in which the patentee would like to price-discriminate by use among the downstream purchasers, so that licenses at each level would be desirable.


It will be interesting to see how, or if, the Supreme Court will resolve this issue. As noted above, it could simply say that the patentee must get its profits from the component maker. This would promote certainty in the downstream product markets and be conceptually consistent with contributory infringement law. On the other hand, it would be somewhat odd to require that patentees deal with contributory infringers rather than direct infringers. Moreover, it would place limits on the ability of patentees to price-discriminate, though, as noted above, the Court might not view that as an obstacle. But one could also take the view that where the downstream applications are significantly different, so as to make separate downstream licenses desirable, there should be separate downstream patents, and that the upstream product then would not be “especially made or especially adapted for use in an infringement of [any one of] such patent[s].”[13]


The Court might simply choose not to address this issue. It could, for example, just correct the Federal Circuit’s approach to the exhaustion doctrine and then direct the Federal Circuit to address it.[14] That approach might look particularly attractive in that the LG-Intel license itself includes a provision that calls for application of the exhaustion doctrine: “Notwithstanding anything to the contrary contained in this Agreement, the parties agree that nothing herein shall in any way limit or alter the effect of patent exhaustion that would otherwise apply when a party hereto sells any of its Licensed Products.”


A final issue is that of LG’s method claims. The Federal Circuit held that “the sale of a device does not exhaust a patentee’s rights in its method claims.” As Quanta’s brief describes, this holding is inconsistent with Supreme Court decisions and even with other decisions by the Federal Circuit. Generally speaking, method claims that would be contributorily infringed by an unauthorized sale of a product should be exhausted by an authorized sale of the product. But if the method at issue goes beyond the normal use of the product at issue, so that the product arguably does not contributorily infringe the method patent, as suggested above for downstream product patents, it seems that exhaustion should not occur.


The Federal Circuit’s cases beginning with Mallinckrodt and continuing through Quanta have allowed patentees to use contract, or even simply unilateral notice, to eliminate the application of the exhaustion doctrine. By simply prohibiting certain conduct in their “license” agreements, patentees can under current Federal Circuit law transform what would be permissible conduct into patent infringement. The Supreme Court, however, said in Univis that “sale of [a patented article] exhausts the monopoly in that article and the patentee may not thereafter, by virtue of his patent, control the use or disposition of the article.” The Court’s forthcoming decision will likely reestablish that rule.


* Mark Patterson is a Professor of Law at Fordham University School of Law . His article on field-of-use licensing was recently published in the William & Mary Law Review. Mark R. Patterson, Contractual Expansion of the Scope of Patent Infringement Through Field-of-Use Licensing, 49 Wm. & Mary L. Rev. 157 (2007).


Preferred Citation: Mark R. Patterson, Reestablishing the Doctrine of Patent Exhaustion, 2007 Patently-O Patent L.J. 38,


[1] Supreme Court Docket No. 06-937. Oral arguments are scheduled for Wednesday, January 16, 2008. For background on the particular issues, see, Dennis Crouch, Supreme Court to Decide Patent Exhaustion Case, Patently-O (Sept. 25, 2007) at 

[2] General Talking Pictures Corp. v. Western Electric Co., 304 U.S. 175 (1938).

[3] United States v. General Electric Co., 272 U.S. 476 (1926). 

[4] Adams v. Burke, 84 U.S. 453 (1873); Motion Picture Patents Co. v. Universal Film Manufacturing Co., 243 U.S. 502 (1917), United States v. Univis Lens Co., 316 U.S. 241 (1942).

[5] Mallinckrodt, Inc. v. Medipart, Inc., 976 F.2d 700 (Fed. Cir. 1992). 

[6] Motion Picture Patents 243 U.S. 502; Univis Lens. 316 U.S. 241.

[7] Arizona Cartridge Remanufacturers Association, Inc. v. Lexmark International, Inc., 290 F. Supp. 2d 1034 (N.D. Cal. 2003), aff’d, 421 F.3d 981 (9th Cir. 2005).

[8] Pioneer Hi-Bred International, Inc. v. Ottawa Plant Food, Inc., 283 F. Supp. 2d 1018 (D. Iowa 2003). 

[10] Commission Regulation 772/2004 on the application of Article 81(3) of the Treaty to categories of technology transfer agreements, O.J. L 123/11, art. 2 (2004), 

[11] Commission Regulation 2790/1999 of 22 December 1999 on the application of Article 81(3) of the Treaty to categories of vertical agreements and concerted practices, O.J. L 336/21 (1999),

[12] Univis Lens, 316 U.S. at 250–51.

[13] 35 U.S.C. § 271(c). 

[14] Cf. KSR Int’l Co. v. Teleflex Inc., 127 S. Ct. 1727 (2007); eBay Inc. v. MercExchange, L.L.C., 126 S. Ct. 1837 ( 2006).





James B. Kobak, Jr., Contracting Around Exhaustion: Some Thoughts About the

CAFC’s Mallinckrodt Decision, 75 J. Pat. & Trademark Off. Soc’y 550 (1993).


Mark R. Patterson, Contractual Expansion of the Scope of Patent Infringement Through

Field-of-Use Licensing, 49 Wm. & Mary L. Rev. 157 (2007).


Richard H. Stern, The Unobserved Demise of the Exhaustion Doctrine in U.S. Patent

Law: Mallinckrodt v. Medipart, 15 Eur. Intell. Prop. Rev. 460 (1993).


Consideration of Delayed Injunctive Relief Demanded by CAFC

The recent Verizon v. Vonage majority opens the door to an expanded use of what I would call delayed injunctive relief.  In particular, the two-member majority (JJ Dyk & Gajarsa) explains in Footnote 12 that the district court “should have” considered whether to allow the adjudged infringer Vonage more time to implement a workaround before enforcing an injunction.

“One factor that is relevant to the balance of the hardships required by the Supreme Court’s decision in eBay was not considered by the district court, namely whether the district court should have allowed time for Vonage to implement a workaround that would avoid continued infringement of the ’574 and ’711 patents before issuing its injunction.”

The appellate panel also notes that Verizon’s interest in putting Vonage “out of business” is not a legitimate interest — rather, the court’s obligation is to protect Verizon’s patent interests without disrupting other business activity.

“Verizon had a cognizable interest in obtaining an injunction to put an end to infringement of its patents; it did not have a cognizable interest in putting Vonage out of business.”

Historically, a finding of patent infringement resulted in a relatively quick form of complete injunctive relief requiring that any adjudged infringement cease immediately.  The Supreme Court’s EBay v. MercExchange opinion considered injunctive relief a yes/no decision — either an injunction is granted or denied based on the traditional four-factor analysis.  Here, the CAFC appears to be reformulating debate to focus on a timing question —  i.e., if not now, at what point will the four-factors favor the patent holder?

In this case, the Court’s words are relegated to a footnote because Vonage never requested a workaround period.  And, in addition, sufficient time has already passed since the verdict.

“Vonage made no request for a workaround period to the district court, and Vonage has already had several months since the district court’s judgment to implement a workaround.”