Here is a fun chart showing the percentage of U.S. patents that cite to Wikipedia as prior art. I created this from data downloaded from the USPTO.

by Dennis Crouch

Masimo Corp. v. True Wearables, Inc. (Fed. Cir. 2022) (nonprecedential)

This trade-secrecy decision includes a few interesting holdings:

1. Published and Widely Distributed, but Still a Trade Secret: The fact that information was found in a prior publication “that has been cited over 1,200 times . . . does not necessarily compel a finding that the information cannot maintain its status as a trade secret for a party in an entirely different field.”
2. Serious Questions: A preliminary injunction in trade secrecy cases require only a “fair chance of success on the merits or questions serious enough to require litigation.”  Here, the plaintiff raised “a serious question” and that was enough.
3. Order to Keep the Patent Filings Secret: The defendant has several patent applications pending at the USPTO which apparently include the trade secret algorithm within its specification.  The district court ordered the defendant to “take all necessary steps to prevent publication” of those applications or resulting patents. On appeal, the Federal Circuit affirmed that preliminary injunction as warranted based upon California state trade secrecy law.

The companies here compete in the market for creating specialized finger-LED equipment and accompanying data analysis services.  This case focuses on calculating total hemoglobin (“SpHb”) based upon light absorption readings from the finger-LED.

How does it work: The LED emitter sends out a pulse of light through the fingertip; a sensor on the other side then measures how much light passed through at various wavelengths. Hemoglobin has a particular light adsorption pattern, and so those results can be used to calculate a relative amount of Hemoglobin.  They do this with a simple linear aggregation: “SpHb = Ax + By + Cz . . .,” where x, y, and
z are various absorption measurements from the monitoring device (such as different wavelength readouts) and A, B, and C are coefficients that reflect the relative importance of the different measurements.

Each product line will have a different equation because it will likely have a different LED configuration; have different sensors; and have a different physical form.  So, a product developer needs to run a clinical trial of its product in order to estimate the most accurate coefficients. For each person in the trial, you do a blood test for SpHB level (the gold standard) and then also take measurements with the finger LED.

What is the Trade Secret: The trade secret at issue here is the optimization algorithm for determining all the equation coefficients.  How do you take data from the clinical trials and turn that into an operational equation for measuring SpHB?

Answer, if you have studied linear algebra this is all straightforward and largely developed in the 1960s.  Math and engineering students across the country take classes on linear programming and linear optimization and have done so for decades. These same models are a basis for most of the system we call “AI.”

The defendants pointed to a particular 2008 IEEE article that spells out the particular approach claimed to be a trade secret here — – the parties refer to this algorithm as “TSS.”  We don’t know the title of this article (it is redacted), but we do know that this is a very well known article with 1,200+ citations.  The description below comes from the appellate briefing:

The footnote explains: “[T]he complete citation for this publication is not reproduced in this brief as Masimo has asserted that the identity of the publication is  confidential.”

Under California law, information that is “generally known” cannot be a trade secret.  Here, however, the Federal Circuit affirmed the lower court ruling that the publication did not render the information “generally known” to folks in the particular field in question.

Under some circumstances, the publication of an alleged trade secret will clearly be sufficient to indicate that the information is generally known. However, the fact that the trade secret has been revealed in some publication somewhere [even if highly cited] does not necessarily compel a finding that the information cannot maintain its status as a trade secret for a party in an entirely different field from the one to which the publication was addressed.

Slip Op.  The Federal Circuit previously wrote that disclosure in a patent renders information “generally known” for trade secrecy purposes, but endeavored to distinguish that conclusion here. Ultimax Cement Manufacturing Corp. v. CTS Cement Manufacturing Corp., 587 F.3d 1339, 1355–56 (Fed. Cir. 2009).

I’ll note here that there are some other factors at play – namely, the founder of True Wearables (Dr. Lamego) is a former Masimo employee. He is the one that  mathematically derived implemented the TSS algorithm at Masimo. And, while there, labeled it a “trade secret.”  These actions of the parties tend to make it look like a trade secret.

I have a Serious Question: On appeal, the Federal Circuit did conclude that the trade secret issue was questionable, but still found that the plaintiff had raised a “serious question as to the validity of the trade secret” which is apparently sufficient for a preliminary injunction.  That appears to be a lower standard than the more traditional “likelihood of success.”

The Preliminary Injunction: Lamego has several patent applications pending, and the preliminary injunction orders Lamego to make sure that those applications do not become public.  Here, the court found the remedy appropriate since “Once the information is in the public domain and the element of secrecy is gone, the trade secret is extinguished.” (Quoting Ultimax).  In this portion of the opinion the court did not contend with its statement a few paragraphs before that publication does not necessarily extinguish a trade secret.

I held a contest and Apple wins–for having the patent with the largest number of cited prior art references (patents issued 2005-2021).  Apple’s U.S. Patent No. 11,037,565 is directed to a version of Siri that works across devices – an “Intelligent digital assistant in a multi-tasking environment.”   The total, 9,441 cited references — 3 of which were cited by the examiner. Apple.Patent.

Here is claim 1:

1. A non-transitory computer-readable storage medium storing one or more programs, the one or more programs comprising instructions, which when executed by one or more processors of a first electronic device, cause the first electronic device to:

provide a media output;

while providing the media output: receive a speech input requesting to continue providing the media output at a second electronic device different from the first electronic device;

determine, based on the speech input, whether to continue providing the media output at the second electronic device; and

in accordance with a determination to continue providing the media output at the second electronic device: cause the second electronic device to continue providing the media output by resuming the media output based on where the media output was previously stopped at the first electronic device.

The application claims priority back to a 2016 provisional application and has three additional family member patent already issued.

In re Netflix (Fed. Cir. 2022)

This was an easy mandamus transfer.  The plaintiffs, CA Inc. and Avago Technologies Int’l,  are both subsidiaries of Broadcom Corp. The plaintiffs are both headquartered in San Jose and Broadcom is down in Southern California.  Netflix is also Silicon Valley. Still, the plaintiffs sued Netflix for patent infringement in the Eastern District of Texas.  Although Netflix itself does not have a place of business in EDTex, the district court found that its intimate relationship with ISPs to provide local content delivery was sufficient.  The vast majority of Netflix content is delivered directly from local residential Internet Service Providers (ISPs).

The district court also refused to transfer for inconvenient forum. Although N.D. California was certainly the locus-of-relevant-events, the court noted court congestion and access to evidence weighed against transfer.  Although the source code is located in N.D. California, the court saw no evidence showing that it would be difficult to get that data to Texas. Netflix does have more internet traffic than any other company, right!

On mandamus, the Federal Circuit only addressed the inconvenient forum question — ordering the case transferred.   The appellate panel particularly focused on the district court’s apparent bias against transferring venue.  In particular, the district court discounted Netflix arguments of evidence availability because Netflix did not “articulate the precise way that the evidence supports its claim or defense.”  At the same time, the court did not place the same precision requirement on CA regarding its evidence.  This bias was improper:

[W]e see no basis for reasonably demanding more from Netflix than it provided at this stage of the litigation, when a transfer motion must be filed and ruled on. . . . Nor have we been shown any reasons here to doubt that the documents and sources would be relevant and material to the issues here.

Id.

Venue – we have a very broad federal venue statute, except that it is super-narrow for patent lawsuits. Should we expand patent venue?

— Dennis Crouch (@patentlyo) January 19, 2022

by Dennis Crouch

Back in 1969, long before Bubba Gump, Louisiana Judge Alvin Rubin explained the virtues of shrimp: “Shrimp, whether boiled, broiled, barbecued or fried, are a gustatory delight.”  Laitram Corp. v. Deepsouth Packing Co., 301 F. Supp. 1037 (E.D. La. 1969).  The Deepsouth case was about patented equipment for deveining shrimp in order to render them “more pleasing to the fastidious as well as more palatable.” Id.  Deepsouth litigated its case to the Supreme Court, and the Court eventually allowed the company to escape some portion of its adjudged liability based upon the territorial limits of U.S. patent law.

We know that U.S. patents are territorially limited.   Although Deepsouth was barred from using Laitram’s patented inventions throughout the United States, Deepsouth began selling its deveining machine to folks outside of the US in a partially constructed form.  The company explained the setup to one of its customers in Brazil:

We are handicapped by a decision against us in the United States. This was a very technical decision and we can manufacture the entire machine without any complication in the United States, with the exception that there are two parts that must not be assembled in the United States, but assembled after the machine arrives in Brazil.

The Fifth Circuit had sided with the patentee — holding that “Such a dependence on technicality would require us to countenance obvious schemes, perhaps as simple as omitting an important screw, designed to evade the mandate of § 271(a).”  However on certiorari, the Supreme Court felt itself bound by the statute and precedent. The court particularly rejected the notion that “substantial manufacture” can constitute direct infringement ”

[W]e have so often held that a combination patent protects only against the operable assembly of the whole and not the manufacture of its parts.

Deepsouth Packing Co. v. Laitram Corp., 406 U.S. 518, 528 (1972).  The decision was 5-4, with Justice Blackman writing in dissent and arguing that “The Court has opened the way to deny the holder of the United States combination patent the benefits of his invention with respect to sales to foreign purchasers.”  Bills eliminating this loophole were quickly introduced into Congress, and eventually became law in 1984, creating 35 U.S.C. § 271(f).  Patent Law Amendments Act of 1984, Pub. L. No. 98-622, 98 Stat. 3383.  The new law creates two avenues for liability:

1. Supplying a “substantial portion” of the uncombined components of the  patented invention from the US in a way that induces combination abroad. 271(f)(1).
2. Knowingly supplying a part from the US that is especially designed as a component of a patented invention and not a staple article or commodity with substantial non-infringing use. 271(f)(2).

Deepsouth has important implications beyond the component-export arena.  The case is an important example of the Supreme Court refusing to fill a gap in statutory language. However, the court majority was also clear that the refusal to gap-fill was one-sided — the majority did not want to expand patent rights without Congressional input:

[W]e should not expand patent rights by overruling or modifying our prior cases construing the patent statutes, unless the argument for expansion of privilege is based on more than mere inference from ambiguous statutory language. We would require a clear and certain signal from Congress before approving the position of a litigant who, as respondent here, argues that the beachhead of privilege is wider, and the area of public use narrower, than courts had previously thought. No such signal legitimizes respondent’s position in this litigation.”

Deepsouth at 531.  The dissent suggested that the precedent was not so clear and so the demand to overrule or modify was fairly slight.

The anti-patent-right-expansion policy of Deepsouth was repeated in Parker v. Flook, 437 U.S. 584, 596 (1978) (“we must proceed cautiously when we are asked to extend patent rights into areas wholly unforeseen by Congress”) and again in Justice Brennan’s dissent in Diamond v. Chakrabarty, 447 U.S. 303, 319 (1980) where he wrote:

[W]e must be careful to extend patent protection no further than Congress has provided. In particular, were there an absence of legislative direction, the courts should leave to Congress the decisions whether and how far to extend the patent privilege into areas where the common understanding has been that patents are not available.

Chakrabarty.  See also, Asgrow Seed Co. v. Winterboer, 513 U.S. 179, 195 (1995) (“A statutory restraint on this basic freedom [of Farmers to have right to collect and replant seeds] should be expressed clearly and unambiguously.” Stevens, J. in dissent).

Deepsouth was decided before the PCT became available and well before TRIPs and  the global expansion of patent rights. We might question whether there is still a need for the US to extend rights to overlap those available abroad, especially as it tends to injury U.S. manufacturing capability.

We have snow on the ground here in Missouri. That makes me think of Gary Reinert’s new patent entitled “method and apparatus for rapid stopping of a motor vehicle particularly on snow or ice.” U.S. Pat. No. 11,091,154.  The image below from the patent tells the story.  Conventional brakes are awful. Antilock Brakes are much better, but the Reinert Snow Brakes take the cake.  Rather than just stopping the wheels, the Reinert Snow Brakes actually shift the car into reverse and spin the tires backward to slow-down the car.  I feel like I saw this in Cannonball Run II, but I’m probably misremembering.

Here is claim 1:

A method for emergency stopping of a motor vehicle comprising the steps of:

activating the emergency stopping of the motor vehicle;

stopping the forward rotation of each wheel immediately following the activation of the emergency stopping of the motor vehicle;

activating a transmission of the vehicle, which shifts an engine of the motor vehicle between driving the wheels in the forward rotation direction, a neutral park position and a reverse rotation direction, immediately after the forward rotation motion of each wheel has ceased such that each wheel maybe driven in the reverse rotation direction;

rotating each driven wheel in the reverse rotation direction until the vehicle is stopped;

activating the transmission of the vehicle once the vehicle is stopped to place the transmission into the original forward rotation position or into the park position,

wherein the step of activating the transmission includes using an on board electronic control unit whereby the method for emergency stopping of a motor vehicle is controlled by an on board electronic control unit, and

wherein the step of activating the transmission includes using sensors on each ground engaging wheel.

The examiner found several on-point references, including US5492512A – Brake system with wheel-reversing means for an automobile (Wu). Reinert distinguished those references by requiring a particular shifting through “neutral park position” before heading to reverse, and also requiring sensors on the wheels that are used to activate the emergency stop.

Field v. De Comeau, 116 U.S. 187 (1886)

Field’s patent uses a spring to make it easier to slip on tight-fitting gloves.  The case reached the Supreme Court in 1886, but the court sided with the accused infringer.  The case involved admitted-prior-art.  In particular, the patent specification admitted “that springs had been combined before with the wrists of gloves” in the past. That admission assured the court that the patent cannot cover “the combination of springs in every form with the wrists of gloves to close them.”  The accused infringer was also using a spring-glove combination, but a different form than that claimed. As such, no infringement.

In this case, the court primarily used the admitted-prior-art to limit the scope of the patent in the infringement context, after recognizing that a broader interpretation would render the patent invalid.  That approach is common of the era.  Today, courts rarely rely upon the prior art when interpreting claim scope.  Still, admitted prior art would seemingly be quite relevant as part of the intrinsic analysis of claim scope.

by Dennis Crouch

The chart below shows the number of Supreme Court patent decisions broken-up by decade. The ramp-up in the late 1800s is tied to the “second industrial revolution” in the US.  The drop in cases after 1890 is largely procedural. Congress created the regional circuit courts of appeals as a buffer between the trial courts & Supreme  Court.  Subsequently, the Court limited the number of patent cases it heard via the certiorari process.  Read more about this transition in my Kessler Cat paper (with Homayoon Rafatijo).

We are 2 years into the 2020s and have 5 decisions thus far (counting Google v. Oracle, which discusses substantive patent issues).  If that trend continues, then the  final figures for the 2020s would be similar to that of the 2010s.  Of course, thus far in the 2021-22 term, the Supreme Court has not yet granted certiorari in any patent cases.

To get this data, I used Westlaw’s case search and limited the results to Supreme Court decisions identified as “patent” decisions by Westlaw and that have a patent law related WL KeyCite headnote.  Reasonable minds will differ on what counts as a “patent case.”

by Dennis Crouch

Plastronics Socket Partners v. Hwang (Fed. Cir. 2022) (non-precedential).

There is a lot going in this decision, but the crux of the appeal is a license interpretation question.  Here are the rough facts:

• Hwang licensed its patent rights to Plastronics Socket.
• Plastronics Socket split into two companies: Plastronics Socket and Plastronics H-Pin under Texas “divisive merger” statute.   (It is odd, but under Texas law, a division in this way is legally defined as a merger).
• As part of that split, Plastronics H-Pin took on sole responsibility for the license. Plastronics H-Pin made the licensed products (h-pins) and sold them its sole customer Plastronics Socket at a very low rate — thus greatly reducing the royalties owed.
• Plastronics Socket resold the product, but argued that it was no longer bound by the contract because the contract liabilities had been shifted to Plastronics H-Pin.

In the appeal, the Federal Circuit confirmed that the “divisive merger” cannot be used to escape from contract obligations.  Thus, Plastronics Socket still owes the 3% that it agreed to pay for its own sales.

= = =

Statute of Limitations: Plastronics also had a counterclaim. In particular, Hwang had a right to further license the patent, but only with approval from Plastronics. Hwang admitted that he had licensed the patent, but that the license occurred 9 years before the lawsuit began and thus was outside of the 4-year statute of limitations under Texas law.  The Texas allowed the claim — since it was an ongoing license.  On appeal, however, the Federal Circuit reversed — holding that the “breach of contract … arose from a single, unauthorized license grant … almost ten years before Plaintiffs filed suit.”  As such, the case was beyond the statute of limitations.  The court distinguished other situations regarding periodic payments of royalties, where each missed payment is seen as another breach and thus can restart the statute of limitations. See Hooks v. Samson Lone Star, LP, 457 S.W.3d 52, 68 (Tex. 2015).

= = = =

These little H-pins are tiny, but have been important to the semiconductor industry:

= = = =

by Dennis Crouch

Prior to 1891, appeals in patent cases went directly to the Supreme Court, and the Court decided lots of patent cases.  In 1891, Congress created the regional circuit courts of appeals as a buffer between the trial courts and the Supreme Court and the number of high-court patent cases began to fall.  The court decided a number of big  patent cases during the period of 1891-1952, although many of them have been rejected or are no longer followed.  Many are also primarily anti-trust cases involving the use (or misuse) of patent rights.

The following are the most cited Supreme Court patent cases during this period:

1. U.S. v. U.S. Gypsum Co., 333 U.S. 364 (1948).  U.S. Gypsum is an antitrust minimum-price-fixing case. The U.S. Gov’t wanted to show that the patent covering the products was invalid, and therefore that the price fixing was improper. The district court prohibited the Gov’t argument, but the Supreme Court reversed.  The case is primarily cited for its explanation of the “clearly erroneous” standard for reviewing factual findings by a judge. A finding can be clearly erroneous even if based on some evidence if the reviewing court “is left with definite and firm conviction that mistake has been committed.”
2. Graver Tank & Mfg. Co. v. Linde Air Products Co., 339 U.S. 605 (1950).  Graver Tank is generally cited for its statement of the function-way-result test for the doctrine of equivalents. “A patentee may invoke [DOE] against … a device if it performs substantially the same function in substantially the same way to obtain same result as patentee’s device.”
3. Precision Instrument Mfg. Co. v. Automotive Maintenance Machinery Co., 324 U.S. 806 (1945).  Precision describes inequitable conduct as form of unclean hands — holding that the patentee had committed fraud and thus lost access to the court of equity — “closing doors of equity court to one tainted with inequitableness or bad faith relative to matter in which he seeks relief.”
4. Hazel-Atlas Glass Co. v. Hartford-Empire Co., 322 U.S. 238 (1944). Hazel-Atlas is another fraud case. The lower court found that the defendant had not done enough to fully uncover the fraud. However, the Supreme Court found that patents are special because they implicate a great public interest. “The public welfare demands that the agencies of public justice be not so impotent that they must always be mute and helpless victims of deception and fraud.”
5. Great Atlantic & Pac. Tea Co. v. Supermarket Equipment Corp., 340 U.S. 147 (1950). The A&P case is a precursor to KSR v. Teleflex (2007). The Supreme Court explained that a collection of known elements cannot be patented unless “the whole in some way exceeds the sum of its parts” — something that is usually not the case in mechanics.
6. The Fair v. Kohler Die & Specialty Co., 228 U.S. 22 (1913).  This short decision by Justice Holmes was cited hundreds of times for the principle that a patent infringement lawsuit has proper jurisdiction in Federal Court despite a lack of diversity between the parties, and that the party making the claim decides what law is being asserted.
7. American Well Works Co. v. Layne & Bowler Co., 241 U.S. 257 (1916). The defendant had been telling people that plaintiff was an infringer; plaintiff sued for defamation.  The Supreme Court confirmed that the patent laws did not create Federal Jurisdiction for the case.
8. Keystone Driller Co. v. General Excavator Co., 290 U.S. 240 (1933). In this case, the court explained that the inequitable conduct (unclean hands) only applies when the unconscionable act has “immediate and necessary relation to the equity that he seeks in respect of the matter in litigation.”  In the case, the court found that bad acts with regard to two patents in-suit did not tarnish three other asserted patents (despite being from the same “family”).
9. Washburn & Moen Mfg. Co. v. Beat ‘Em All Barbed-Wire Co., 143 U.S. 275 (1892). This famous case is most cited for the Supreme Court’s rejection of oral evidence as sufficient to invalidate an issued patent.  Rather, proof must be “clear, satisfactory, and beyond a reasonable doubt.”
10. Eibel Process Co. v. Minnesota & Ontario Paper Co., 261 U.S. 45 (1923). In Eibel, the court found that a continuous-sheet paper machine invention did not rise to the level of a “pioneer patent, creating a new art.”  However, the court determined that the improvement was of such merit that it was still “entitled to liberal treatment.”  The court also held that the invention’s “immediate and successful use” across the industry was “convincing proof” that the specification was sufficiently detailed and precise.

Note here, I ranked these according to the number of citations by courts.  We would get a different sorting if ranked by law review citations — a case such as Funk Bros. Seed Co. v. Kalo Inoculant Co., 333 U.S. 127 (1948), would rise up toward the top of the list and the Barbed Wire case would fall off.

I've been looking at pre-1952 patent cases (1891-1951). The following are the top-ten most often relied upon in law review articles:

1. Graver Tank & Mfg. Co. v. Linde Air Products Co., 339 U.S. 605 (1950) (function-way-result test for the doctrine of equivalents); pic.twitter.com/DvGjkI1LXL

— Dennis Crouch (@patentlyo) January 12, 2022

By Jason Rantanen

Happy 2022! As I’ve done for the last few years, below I provide some statistics on what the Federal Circuit has been doing over the past year. These charts draw on the Federal Circuit Dataset Project, which contains information on all Federal Circuit decisions posted to the court’s website and and the court’s dockets.

Figure 1

Figure 1 shows the number of Federal Circuit opinions and Rule 36 summary affirmances by origin since 2010. Once again, the highest number of merits terminations arose from the PTO. The most notable change for 2021 was a drop in the number of decisions arising from the district courts: from 197 in 2020 to 142 in 2021. These represent individual documents (i.e.: a single opinion or Rule 36), not docket numbers, but there’s also a 25% drop when looking at the data on a per-docket number basis.  Overall, the Federal Circuit issued 653 merits decisions in regular appeals in 2021 as compared with 690 in 2020.

Figure 2

Figure 2  shows the number of opinions versus Rule 36 summary affirmances arising from the District Courts and PTO. Relative to recent years, the portion of appeals that were summarily affirmed was substantially lower. Thirty-five percent of appeals arising from the PTO were summarily affirmed—much less than 2019, when almost half of the appeals arising from the PTO were Rule 36’d.

You’ll also notice that the number of written opinions that the Federal Circuit issued in appeals arising from the district courts was the lowest it’s been at least since 2010, while the number of written opinions arising from the PTO was the highest. The type of opinion paints a different picture, however: as Figure 3 shows, while around half of the opinions arising from the district courts were designated as “precedential,” only 25% of the opinions arising from the PTO were. To me, Figure 3 suggests that the Federal Circuit perceives that it has addressed many of the novel questions created by the AIA and inter partes review, and sees its jurisprudence as more settled than it was a few years ago.

Figure 3

Interestingly, this does not seem to mean that affirmance rates in appeals arising from the PTO are increasing. For one thing, they’re already quite high: over the past decade, the Federal Circuit has affirmed the PTO about 80% of the time.  Nor did they increase in 2021. Last year, the Federal Circuit affirmed the PTO in 78% of its decisions. Another observation is that even though the court’s use of Rule 36’s in appeals arising from the PTO has dropped over the past two years, its affirmance rate has remained about the same.

Figure 4

The Federal Circuit did, however, affirm district courts more frequently last year than any previous year.  The average rate at which the court affirmed-in-full district courts from 2010-2021 was 69% (indicated by the purple line); last year its affirmance rate in these appeals was 79%.

I didn’t create a separate graph for dissents and concurrences, but overall Federal Circuit judges wrote about the same number of dissents in 2021 as in 2020 (35 in 2021 versus 37 in 2022), but wrote a lot more concurring opinions (23 in 2021 versus 10 in 2020).

Figure 5

A new addition this year is a dataset of all Federal Circuit dockets since 2000 based on data from PACER. Figure 5 shows the number of appeals docketed at the Federal Circuit by origin. For the past few years, the court has received well over a thousand appeals per year. That number fell a lot in 2021—especially appeals arising from the PTO and the MSPB. (The MSPB has not had a quorum since 2017, and has not had any sitting members since 2019.)  The downstream consequence of this is that we can expect to see fewer Federal Circuit decisions in appeals arising from the PTO in 2022.

What about petitions for writs of mandamus? Jonas Anderson, Paul Gugliuzza and I are finishing the draft of our paper on those, but it’s no spoiler to say that orders on petitions for writs of mandamus are a big part of the Federal Circuit story for 2021.  Last year the court issued 52 merits orders on petitions for writs of mandamus, as compared with 35 in 2020 and 19 in 2019. For more on those, here’s a link to our series.

You can access the full document dataset here and generate a variety of charts using the widget, or download the data and plug it into your favorite analysis program.  In addition, both the docket and document dataset are archived on the Harvard dataverse, along with the Stata code for generating the tables for the above graphs.  Methodological details are provided in this working paper. Or, if there are specific aspects of the data that you’re curious about, email me or DM me on Twitter and if I’m able to, I’ll take a look.

Thanks to my research assistants for help reading and coding all of the court’s decisions especially Matt Fuentes, Daniel Kieffer, Brenna Kingyon, Lindsay Kriz,  Sara Leibee, Ryan Meger, Madison Murhammer Colon, Charles Neff, Riley Stanton, Connor Williams & Allison Williamson.