Weekly utility patent grants in the United States, 2006-2008

1. The chart

2. Background

This project started with a simple question from an audience member in a recent talk I delivered - "Is the lower allowance rate resulting in fewer patents being issued?"

While it seemed easy enough on its face, I had to admit that I didn't know off hand. The official calculation behind the 'allowance rate' is less than clear, as is its impact on the number of patents being issued.

The question became a bit more intriguing to me when I realized it might be connected to a post I wrote last year. Shortly after the Supreme Court announced its decision in KSR v. Teleflex, I posted my thoughts on the Top Ten Policy Implications of the case. Specifically, the question got me thinking about number 8 on the list - "Patent maintenance fees increase dramatically as grant and maintenance rates drop just as dramatically (or “Hey buddy, can you spare a couple grand?”)." My original thinking behind that "policy implication" was that, if the decision results in a dramatic drop in the number of patents being granted, Office receipts for issue and maintenance fees will drop in corresponding fashion. In order to maintain the status quo on gross receipts, the Office will have to increase maintenance (or other) fees to make up the difference. This would be true, of course, no matter the reason for the reduction in grants (whether inspired by the KSR decision or some other factor).

With that in mind, and considering the cloudy nature of the 'allowance rate' calculation, I rephrased the question as something more straightforward - "Is the Patent Office granting fewer patents than it has in previous years?"

3. Methodology

Data for this chart were collected directly from the USPTO advanced search page for issued patents. Searches were limited to individual issue weeks by setting the issue date tag to the appropriate date (e.g., isd/11/11/2008). Design, plant, reissues and other non-utility types were excluded by setting the application type tag ('APT/1').

The following is an example search string: "isd/11/11/2008 and apt/1"

For each search, the total number of patents meeting the criteria was taken from the top of the USPTO search results page (e.g., "Hits 1 through 50 out of 3395")

Data for all searches were tabulated in a MySQL database on the Promote the Progress server. A custom php script queries the database each time this page is loaded and pulls the up-to-date data from the database. The script assembles a url from the data and various formatting instructions and passes it to the Google Chart API, which then draws the chart and embeds it on the page.

4. Analysis

The chart above graphically represents the number of utility patents granted by the Patent and Trademark Office on a weekly basis for the years 2006, 2007, and 2008. It clearly shows that, generally, the Patent Office is indeed granting fewer patents this year than it did in 2006. The same can be said for 2007 v. 2006. I haven't done any statistics on the numbers, but the difference at some individual weeks is greater than 10%. Indeed, the overall grants for 2008 fell by more than 21,000 patents as compared to 2006 (21,653, a difference of about 12%):

(total patents granted during each year)
2008 155,600
2007 159,475
2006 177,253

Such a difference will have both an immediate and a delayed impact on Office income - Twenty thousand fewer issue fees were paid in 2008 as compared to 2006. If all of those 'lost patents' would have been large entity patents (which, of course, is not the case), the lost issue fees would be more that $28 million dollars ($14M if all would have been granted to small entities). That's real money, even in Washington.

The impact will be felt again in 3 to 4 years when the first maintenance fees are due on this year's crop (assuming similar maintenance rates). Maximum lost revenue here is greater than $18M (if all of the lost patents would have been maintained by large entities). This will ripple through the maintenance fee schedule as time marches on (fees will be due again in 7-8 years, and again in 11-12 years). If maintenance rates remain the same, proportionate reductions in income can be expected.

Whether and how the Patent Office responds to these lost receipts remains to be seen, but the differences seem to be of sufficient degree to require an adjustment somewhere.

It would be interesting to see whether the Office or the contractor has made any adjustments in response to the reduced issuance workload.

Aside from the difference in grants, the chart revealed a trend that I wasn't expecting. Over the three years examined, the number of patents granted dropped precipitously at at least six periods on the calendar (I've highlighted these periods with the blue vertical bands). A couple of these are likely due to holidays and their impact on the issuance workflow (e.g., the band at the end of the year corresponds to the Christmas week and the band at the beginning of the year is roughly four weeks after Christmas - a ripple effect working its way through the workflow?). The reason for the other bands is not immediately clear, though. Perhaps other holidays, summer vacations, etc.

5. Resources