In the January, 2006 issue of Intellectual Property Today, attorneys Thomas J. Van Gilder and Carl A. Kukkonen cited to a document on the webpage of the United States Patent and Trademark Office [USPTO] for the proposition that “patent filings have grown from 353,394 to 406,302, an increase of nearly 15%, from FY 2002 to FY 2005.” In footnote 3 of their article, they carefully provided the page number of the document and the link to the document: See USPTO 2005 Annual Report at 61,
The interesting, and troubling, aspect about this is that one will not find in the cited document at the given link the number “406,302”.

The number 406,302 did not arise out of thin air. Intellectual Property Today had mentioned the number in its December 2005 issue: The agency received 406,302 patent applications, and
323,501 applications for trademark registration as reported in its fiscal year 2005 Performance and Accountability Report released in November.
The magazine eWeek had mentioned the number on January 16, 2006: Last year, the USPTO granted 165,485 patents, up from 99,000 in 1990. According to the patent office, a record 406,302 new applications were filed
last year alone.
On January 25, 2006, Bruce V. Bigelow of the San Diego Union-Tribune used it: In November, the patent office said it issued 165,485 patents in the fiscal year that ended in September — and had received 406,302 patent applications, along with 323,501 applications for trademark registration. Andy Holloway of Canadian Business used the number on January 30, 2006: It’s easy to see why the patent office is swamped. It received 406,302 patent applications
and granted 165,485 of them last year, adding to a database of patents that numbers roughly seven million.
On February 20, 2006, Dan O’Shea had used the number in Telephony: During 2005, the agency received 406,302 patent applications and 323,501 applications for trademark registration, according to the agency’s Web site.

As of April 23, 2006, one will find at page 61 of the above-identified link a table stating that there were 409,532 filings in FY 2005. One will not find an explanation of why the previous number of 406,302 is gone. Data that exist only on an internet website can be changed, without explanation, and thus cease to exist. In this particular instance, we have evidence of the previous information, because so many people cited to it. However, a person looking only to footnote 3 of the article by Van Gilder and Kukkonen, and following the link to page 61 might be led to the conclusion that Van Gilder and Kukkonen “got it wrong.”

In fact, the “error” of Van Gilder and Kukkonen would have been referring to a link to a site that was capable of changing the numbers. Those authors who cite to internet websites for numerical data should be aware that such citation may be perilous in the sense that it is not permanent. The situation presents an interesting issue in the way law review cite checking works. If a law review author utilized the 406,302 number and cited to Van Gilder and Kukkonen, the cite would be approved. However, if a law review author utilized the 406,302 number and cited the USPTO document, the cite should NOT be approved.

In this particular case, there is a secondary sustantive issue as to patent law. The figure of 409,532 (or the previous figure of 406,302) comprises the combination of utility patent applications AND design patent applications. Typically, in such things as the debate on how high the patent grant rate is, one looks only at utility applications. Taking the previous data as presented, one would find a grant rate of 165,485/406,302 = 40.7%, minuscule in comparison to the 97% number once alleged by Quillen and Webster, or the 85% number accepted by Lemley and Moore in the article “Ending Abuse of Patent Continuations.”