Across the Internet can be heard the shrill cry of opponents to the “First to File” patent registration system which will take effect, pursuant to the America Invents Act, in approximately 18 months. The “First to File” system, they say, will disfavor the individual inventor working in his garage and award big corporations with the resources to win the race to the Patent & Trademark Office.
This is a very romantic idea which based on certain populist notions regarding the American Dream, but not rooted in historical or legal reality. It is the purpose of this article to describe the “First to Invent” system as it worked in the real world.
Contrary to popular (and populist) belief, the “First to Invent” system was not a system in which patents were awarded to the first person to conceive of an idea. Hacker News provides these helpful examples to show how it worked. (We have taken the liberty of clarifying their examples):
X comes up with an idea at time 0. X starts working to reduce the idea to practice.
At Time 4, Y comes up with a similar idea, and also starts reducing to practice.
At Time 10, Y gets it all working, writes up and files a patent application.
At Time 15, X gets it all working (poorly), writes up and files a patent application.
Under a “First to File” system, Y will be awarded the patent. Under a “First to Invent” system, Y is presumed to be the one entitled to the patent, but X can initiate an “interference” action. (That’s an administrative procedure in the patent office to determine who “invented” first.) To win the interference action, X must show that he conceived of the invention prior to Time 4 and worked continuously and diligently to reduce it to practice until Time 15. Proving this can be difficult — it requires extensive and objectively verifiable record-keeping. It is also expensive. An interference action may be as little as $50,000 but is typically much more and can exceed three quarters of a million dollars. Fortunately, interference actions are rare: less than one percent of all patent applications become involved in an interference action. This means that in over 99% of patent filings, the patent is effectively awarded to the first (and only) to file.
In those cases where there is a dispute by two or more inventors (or teams of inventors), you only need to ask yourself who has the money to mount an interference action and you’ll know that the answer is “not the small inventor.” An empirical study cited on Bill Heinze’s blog, ip-updates confirms this point. In “Competition, Innovation and Racing for Priority at the U.S. Patent and Trademark Office,” Linda R. Cohen and Jun Ishii from the Department of Economics at UC Irvine analyzed a random sample of 662 patent interference actions and concluded as follows:
“we do not find evidence that the system works to the benefit of small inventors or firms. Small firms rarely avail themselves of the interference process. But if we assume that the winner of an interference is the first-to-invent,* the results suggest that small inventors do not suffer from a filing disability. Alternatively, there is some indication that other groups do have such problems – both the U.S. government and U.S. universities, for example, are far more likely to enter our sample as junior parties, yet they win a reasonable share of cases, which suggests that both institutions systematically take longer to file patent applications than do private firms and individuals. However, we need more data to establish the trend statistically.”
Cohen and Ishii found that in 390 out of 662 cases, or 58.91% of the time, the patent went to the senior filer (i.e., the first to file), while 27.64% went to the junior party, 7.85% went to both parties, and 5.59% went to neither party. (Cohen and Ishii presumed that the party awarded the patent was actually the first to invent, but that probably isn’t true in every case.) In most cases, the winners were large corporations.
Let us continue with the examples comparing the two systems.
X conceives of the idea at Time 0 and begins working on reduction to practice.
Y conceives of the idea at Time 4 and begins working on reduction to practice.
At time 6, X takes time off to work on something else — perhaps he took another job or started working on a different invention — and he only resumes working on the invention at Time 8.
Y files his patent application at Time 10. X files at Time 15.
Here, Y wins under both the “first to file” system and “first to invent” system, because X lost his priority date of Time 0 when he interrupted his work. His new priority date is Time 8. Since Y was working diligently from time 4 to his filing date, Y’s priority date is Time 4.
Same as Example 2, except X files at time 10, Y files at time 15.
X, who was the first to conceive of the invention, wins under the “first to file” system, but loses to Y (after an interference action) under “first to invent” system.
In sum, under the “first to invent” system, the patent wasn’t necessarily awarded to the one who first conceived the idea or first reduced it to practice. The situation was far more complicated.
Walter Hunt and Elias Howe
Moving from the abstract to the concrete, here is a real world example of the “first to invent” system in action in one of the most important patents in the history of American industrialization: the sewing machine.
Walter Hunt built the first American sewing machine in 1834, but decided not to seek a patent because he thought his invention would cause unemployment. His machine sewed only straight seams, but used a lockstitch mechanism characterized by a needle with an eye at the point which passed through the cloth, creating a loop on the other side through which a second thread passed via a shuttle. In 1846, Elias Howe obtained the first American patent for what Hunt was the first to invent: “a process that used thread from two different sources.” Howe’s sewing machine, however, failed commercially.
When Isaac Singer, with the help of Walter Hunt, made design improvements, thereby creating the first commercially viable sewing machine, Howe sued Singer for patent infringement. (In Singer’s machine, the needle moved up and down and the mechanism was powered by a foot treadle.) Belatedly, Hunt tried to file for a patent for his earlier invention, but his claim was rejected by the Patent Office. Had Hunt been successful, Howe would never have prevailed against Singer.
What follows are excerpts from the 1854 decision of Judge Charles Mason, Commissioner of Patents, on Hunt’s application. (Thanks to sewalot.com, where we found this text.)
Hunt claims priority upon the ground that he invented the Sewing Machine previous to the invention of Howe. He proves that in 1834 or 1835 he contrived a machine by which he actually effected his purpose of sewing cloth with considerable success. Upon a careful consideration of the testimony, I am disposed to think that he had then carried his invention to the point of patentability. I understand from the evidence that Hunt actually made a working machine in 1834 or 1835. The papers in this case show that Howe obtained a patent for substantially this same invention in 1846.
Notwithstanding this, the Commissioner was forced to refuse Hunt’s belated application, for the reason that an Act of Congress in 1839 had provided that inventors could not pursue their claims to priority in patents unless application was made within two years from the date when the first sale of the invention was made [one year under current law]. Hunt had sold a machine in 1834, and had neglected to make application for his patent till 1853.
Thus it was that one of the grandest opportunities of the century was missed by the man who should rightfully have enjoyed it; the honors and emoluments of the great sewing machine invention passed to a man who neither had invented a single principle of action, nor applied a practical improvement to principles already recognized
* * *
Elias Howe, Jr., acquired the power, by simply patenting another man’s invention, to obstruct every subsequent inventor, and finally to dictate the terms which gave rise to the great Sewing Machine Combination about which the world has heard—and scolded—so much. Howe’s machine was not, even in 1851, of practical utility. From 1846 to 1851 he had the field to himself, but the invention lay dormant in his hands. He held control of the cardinal principles upon which the coming machines must needs be built, and planted himself squarely across the path of improvement—an obstructionist, not an inventor—
and when, in 1851, Isaac M. Singer perfected the improvements necessary to make Hunt’s principles of real utility to the world Howe continued to obstruct and pursue litigation.”
It was no secret that Howe was not the first inventor of the lockstitch sewing machine — according to Hunt, Howe admitted that to him personally. However, Howe was the first to file.
“First to File” under the America Invents Act Doesn’t Mean First to Steal
For those who have been confused into thinking that “first to file” means that someone can simply steal an invention and run to the patent office to file it, the America Invents Act makes it perfectly clear that it is only inventors (or the owners to whom they assign their patents in writing) can file at all. Moreover, where the first party to file “derived” its invention from another, the prior inventor may bring a derivation proceeding. In many respects this will resemble the interference action, but such actions should arise on far fewer occasions.