Study Confirms That Revealing Secrets, Rather Than Hoarding Info, Is Good For Inventors
from the shocking-not-shocking dept
sciamiko points us to an interesting study done by Stuart Graham, who was the first chief economist of the US Patent Office (we were initially excited about his hiring, though the only other time we reported on his work, it was to wonder why a paper hid his connection to the USPTO), looking at whether or not inventors choose to reveal the “secrets” of their invention prior to actually getting the patent. Graham and his co-author, Deepak Hegde, examined what happened after the American Inventors Protection Act (AIPA) went into effect in 2000. Part of the AIPA was that the USPTO would publish patent applications after 18 months (usually well before the patents were approved or rejected) — rather than only making them public after they were approved. The usual suspects (patent hoarders and self-described “small inventors”) screamed like crazy about how this would completely destroy American innovation, because they’d have to reveal their secrets too early. To try to appease these concerns, the bill included a loophole: patent applications could be kept secret if they didn’t file for foreign patent protection — which was the case for about half of all US patents.
That gave Graham and Hegde a nice dataset to look at, to see who chose to keep their patent applications secret until approval, and who let those applications be revealed. If those who freaked out about the publication requirement were right, it should be clear in the data that, when given the chance, businesses kept their patents secret, and the “secret” patent applications should be worth more than the non-secret ones. The reality? The exact opposite was true. Inventors chose to reveal their patent applications readily, even when they had the option of withholding them, and the more valuable patents tended to be the ones that were revealed:
They examined 1.8 million granted patents filed at the USPTO from 1995 to 2005 and analyzed the disclosure preferences of the inventors. Their analysis found that, among those not seeking foreign protection, about 85 percent of inventors filing a patent since 2000 chose to disclose information about their patents prior to their approval.
“Overwhelmingly, those inventors patenting only in the U.S. are choosing 18 month disclosure,” co-author Hegde said.
And the study appears to show that the complaints and worries of those small inventors about this increased transparency was, in fact, complete bunk:
When the AIPA was passed, one of the biggest complaints was that the publication requirement would hurt U.S. small inventors, but the researchers found that individuals and small companies still opted for disclosure during the study period.
“Small U.S. inventors are not choosing the secrecy route,” Graham said. “When they patent only in the U.S., they are choosing secrecy in only about 15 percent of the cases, not statistically different than the rate among all other types of inventors.”
Another major complaint of the AIPA was that disclosing patent secrets would stop the engine of innovation in the United States and that society would get less meaningful inventions. Contrary to these arguments, the researchers found that patents born out of secrecy were overall less valuable than those that opted for disclosure.
“When we examine indicators of patent value, we find consistent evidence that the least-valuable and least-impactful patents are those that opted for pre-grant secrecy,” Hegde said.
This isn’t particularly surprising to us — as, contrary to the way some think, we’ve pointed out for years that hoarding information tends to limit innovation, while sharing it is likely to lead to greater innovation. Of course, it’s the same “small inventors” who insisted the sky was falling over the AIPA who are now protesting the latest attempt at patent reform, making similar claims about how it will drive down quality and drive inventors out of business. They were totally wrong last time, and it’s likely they’re totally wrong again.