Monday, March 7, 2011

Patent Fees and Counts and the Standard Application Metric - PTO Reform Ideas 03

Both the fee structure and the examiner count system should be based, not on the unit of the filed application, but instead on a metric that compares the actual application to a standard application.   The comparison would be based on such parameters as numbers of:  specification columns in standard format, figures and reference characters in the figures, independent and total claims, and IDS references and the sum of the individual references' pages (or word counts, figures, etc.). 

This way fees and counts would be appropriately tailored to the complexity of the task of examination.  Today's one size fits all system prevents such tailoring.  (Excess claim and page fees do not address complexity adequately and may not affect the count system at all.)

Patent prosecution professionals know that the more complex the case the more time is needed to do a proper job, and their fees, whether by the hour or by prearrangement, reflect this.  R&D expenditures are also related to the complexity of the technology, as are the pay scales in different fields.  Likewise, revenues from commercialization are often proportional to complexity because the price per unit, the margins and the volume of sales are often higher for more complicated and sophisticated technology.

Why do PTO fees and examiner counts essentially ignore complexity?  Answer:  "Because they always have."  That answer is not good enough any more.

Employing a standard-application metric (SAM) for the count system and fee structure acknowledges the reality that all applications are NOT created equal.  Examiners can be given twice the time to examine an application that is twice as complex.  This means that they will have enough time to understand complex technology, read a lengthy specification, consider voluminous information disclosures, and examine numerous, highly nuanced claims.  The system will also encourage applicants interested in cheaper and faster examination to be more succinct and direct.

Who will devise this metric?  Answer:  The PTO.  Director Kapos, and management, and the examining corps, all know what the measurables of patent applications are and how those measurables affect examination time.   

The public, too, can consider the definition of the standard application.  Publicly available databases and the PTO's $200 DVD of patent statistics are helpful.  For example, the CLAIMS.TXT data on the DVD permits calculation of the median, mid 50%ile and average for three aspects of complexity of issued patents:  numbers of claims, figures and drawing sheets.   Unfortunately, the number of specification columns and IDS references -- useful information for our SAM -- are not as easily found, nor are data for applications as opposed to issued patents.  That information can be compiled without too much difficulty, however.

To begin working toward the SAM solution, I have looked at some numbers.  For simplicity and speed,* I have segregated out two groups of ~1000 patents from the DVD, representing past and current complexity levels:
     For the 1000 patents numbered 3,999,000 to 3,999,999, issued in 1976:
          Claims: median 8, mid50 4 to 12, average 9.2,
          Figures: median 5, mid50 3 to 9, average 6.9, and
          Drawing Sheets: median 2, mid50 1 to 3, average 2.8.
     For the 990 patents numbered 6,999,000 to 6,999,999 (ten patents were withdrawn prior to issuance), issued about thirty years later in 2006:
          Claims:  median 15, mid50 9 to 21, average 19.6,
          Figures:  median 8, mid50 5 to 13, average 13.3, and
          Drawing Sheets:  median 6, mid50 4 to 9, average 9.5.

It is undeniable that serious thought and hard work will be needed to determine
     (1) the parameters to use,
     (2) the standard application values,
     (3) the weights to apply to those parameters, and
     (4) the structure -- linear?  step function with how many steps? -- by which to calculate the fees and counts for actual applications compared to the standard.
Consideration should also be given to how often the definition of the standard application should be adjusted.  It might be when the median actual application goes above 1.5 SAMs, for example.

The PTO can do this.  And should.  And the time to do it -- while Microsoft v. i4i is pending before the Supreme Court, patent reform legislation is before Congress, and public attention is being given to the quality of examination and in particular to the problem of inadequate examination time and resources -- is now.

* I used batches of 1000 patent numbers rather than a larger or more random sample.  Consecutive patents, all issued on the same date or a week apart, may all have been granted by a small subset of examiners or art units and thus may not necessarily represent the whole.