Monday, April 3, 2017

Funding, Patents, and Monetization

There is a new MIT study which alleges:

To conduct the study, the scholars examined 365,380 NIH grants funded between 1980 and 2007 — nearly every NIH grant awarded for decades. Exactly 30,829 were the direct basis for patents; 17,093 of those were so-called “Bayh-Dole” patents issued to universities and hospitals, something federal legislation made possible starting in 1980. Of the NIH grants, 112,408 were additionally cited in a total of 81,462 private-sector patents. And as the authors put it in the new paper, even these NIH-backed research projects that are indirectly cited in later patents “demonstrate the additional reach that publicly funded science can have by building a foundation for private-sector R&D.” Azoulay, an economist who studies the production and dissemination of scientific knowledge, says the bottom-line figures in the study — the 31 percent and 8 percent of NIH grants that contribute to and more directly generate patents — strike him as being significantly large because of the broad scope of research the NIH supports.

Now let us examine this a bit.

First, under the Dole Law the beneficial ownership of patents resulting from Government work  go to the sponsoring institution. Some interest accrues to the Principal Investigator and Patent filer.

Second, patents are not always filed and not always issued. CRISPR is a real counter example. Universities file for patents if the IP Office believes it is worth the effort. That determination varies widely from institution to institution.

Third, patents are not businesses. Patents are often over turned or contested and thus are no assurance of anything. Patents are also an invitation for litigation, lots of expensive litigation.

Fourth, patent portfolios really do not generate that much money, on average. Sometimes one patent can be great but most of the time they are not money generators.

Fifth, dealing with a university IP Office can be near impossible because they are trying to maximize their return with no understanding of the risk in monetizing the patent.

Sixth, many patents are used solely to block others. Namely a patent may be licensed just to stop a competitor. That means that the patent is shelved and is dormant.

The Scientist notes:

They also found that basic research was nearly as likely to contribute to a patent as applied projects. The authors separated grants awarded through a request for applications (RFA)—indicating a focus on addressing a particular disease—from those that were unsolicited—likely representing more basic science—and discovered a similar citation rate in associated patents.

Given what was noted above, one must wonder to what use the patent was licensed for. The study, supposedly to be released, may or may not contain any details. One is left wondering.

EurekaAlert notes:

A 2015 working paper released by Azoulay, Li, Sampat, and Joshua Graff Zivin, a professor at the University of California at San Diego, estimated that every $1 of public NIH funding yielded between $1.5 and $2 in private-sector pharmaceutical drug sales. That study is still undergoing peer review.

 This above connection seems quite difficult to ascertain. First the result is drug sales. That is the gross output for the pharma industry. The input alleged is NIH funding. Just consider the following:

1. A single patent is but one part of a new drug.

2. A new drug has about a 1:100 chance of ever getting FDA approval.

3. A successful FDA drug may also be problematic in total market acceptance.

Thus the claim as noted above seems highly unlikely. It may also lead to negative criticism of NIH. That often is the problem with academics.