A few weeks ago, administrators at Penn State University did something they believed had never been attempted in American academia: The school put about 70 engineering patents up for auction and tried to sell them to the highest bidder. They weren’t so successful—not many patents sold—but the project has disturbing implications. What if all this intellectual property, based on research done at a public institution, were to end up in the hands of someone less interested in innovation than in hauling companies to court? Penn State tried to discourage bids from sleazy agents, and Lisa Powers, chief spokesperson for the university, says in an email that "not one patent aggregator made a bid during the auction."* But what if it had auctioned its inventions to a greedy patent troll by accident?
It wouldn’t be the first time that an institute of higher learning had partnered up with patent trolls, or mimicked their behavior. Universities and patent trolls have some major traits in common: Both make money off of legal rights; both let other businesses implement ideas and then pinch a portion of the revenue; both purport to bring that money back to those innovators who most deserve it. Looked at from a distance, and with squinted eyes, a school might seem to be a patent troll itself—and that resemblance is growing stronger.
Trolls come in many forms, of course, and most engage in practices that a university never would. Some function like protection rackets: They file suits against small companies, then settle for a modest fee—$50,000 or $100,000, something less than the cost of litigation. Others wield poorly worded patents that cover broad swaths of industry—all podcasts, for example—then spring their rights on lots of companies at once. “Aggregators” buy up vast portfolios of patents and shop them around in package deals.
Universities, for their part, try to make their research useful. They often license their technologies to startups and hope the businesses succeed. (Stanford University wisely gave the Google algorithms to the students who created them, in exchange for 1.8 million shares of the company.) But the nature of the knowledge business makes it so that many university patent deals are exclusive: A professor invents something; a school files for the rights and sells them to a single, private company. (The professor gets a share of the profits.) That private company may have no interest in the university’s mission or the broader public good. It might even work to stifle rivals’ innovation.
Indeed, this happens with some regularity. When researchers based at the University of Utah cloned a pair of genes that predispose women to getting breast cancer, the school granted the discovery to Myriad Genetics. Myriad later used its patents to prevent doctors from screening women for the dangerous genes at a reasonable cost. In 1993, a biologist named Michael Doyle built a Web-based, interactive image-viewer for the University of California–San Francisco so doctors could examine 3-D reconstructions of human embryos. The school patented his invention and licensed it to his startup, Eolas Technologies, which went on to sue purported infringers. At one point, Eolas settled with Microsoft for an estimated $120 million, with $30.4 million going to the University of California. (The patents that Eolas used to make these claims have since been ruled invalid.)
Of course a private business needs protection if it’s to make investments in a new technology. But we all supply the seed capital for university inventions, through grants provided by the National Institutes of Health, the Defense Advanced Research Projects Agency, the National Science Foundation, and other tax-funded incubators. That doesn’t mean inventions should be free to one and all, but it makes you wonder whether the patent game should be played so aggressively with money from the government.
The university’s patent libraries—and the problems that they inadvertently create—are only growing. Until 1980, and the passage of a law that made it easier for them to profit from inventions, universities were issued about 250 patents per year. According to a report from the Association of University Technology Managers, in 2012 schools filed more than 14,000 new patents, were granted about 5,000 patents, and made about 5,100 new licensing agreements. Their total income from such agreements amounted to $2.6 billion. (Some argue that the law didn’t matter much and that the patent boom would have happened either way.)
The schools don’t engage in much patent trolling on their own. Despite the huge expansion of their patent warehouses, they rarely go to court. (There are some notable exceptions.) James E. Bessen, of Boston University School of Law and Harvard’s Berkman Center for Internet and Society, estimates that lawsuits by patent trolls cost the country $29 billion every year, but that universities participate in just 1 or 2 percent of cases.* However, Bessen says this statistic might conceal a bigger problem: The firms that are doing business with the schools engage in patent mischief of their own. “The universities aren’t acting as trolls themselves, but they’re licensing their technology to trolls,” says Bessen.
That is to say, your alma mater may end up working with a patent troll, and making money from a patent troll, even if it would be disinclined to go to court itself. To pick only the most prominent example, Intellectual Ventures has now acquired patent rights from at least 60 American universities, according to Paul Basken’s reporting in the Chronicle of Higher Education. (The California Institute of Technology sold exclusive rights to 51 of its patents to the notorious aggregator in 2008.)
In the past, university administrators might have tried to avoid these arrangements. In 2007, officers at a group of 11 major research institutions—including Caltech, the University of Wisconsin, Stanford, and the University of California—signed a document called “In the Public Interest: Nine Points to Consider in Licensing University Technology.” Point 8 suggested that schools “be mindful of the implications of working with patent aggregators.” They “would better serve the public interest,” it went on, “by requiring their licensees to operate under a business model that … does not rely primarily on threats of infringement litigation.”
But the dynamics of the patent business have made these noble goals seem distant. For all the work they do to sell off good ideas, most universities barely make a nickel from their intellectual property. A recent report from the Brookings Institution found that 84 percent of research universities lost money on their patent deals in 2012. In other words, they spent more on filing patents with the government, and on paying staff who try to turn those patents into profits, than they earned from actual licensing arrangements. In fact, it’s often said the system functions like a lottery: Among universities, the top 10 percent of earners take in 70 percent of all the license income. A few might break the bank with a whopper of a deal—or a whopper of a lawsuit—but most end up sitting on a stack of worthless paper.
One administrator told Paul Basken that about 95 percent of university patents go unused and that schools are “looking for newer and better ways to try and license them out.” That could mean they’re seeking deals with well-known patent trolls, or merely selling to the highest bidder. Either way, academia seems on the verge of falling into some increasingly unsavory behavior.
The president of the Association of University Technology Managers says these concerns are not supported by the evidence. Yet in the midst of the ongoing debate in Washington over patent reform, universities have been lobbying in ways that would seem to align them with the patent trolls. They’ve fought troll-busting legislation, like the expansion of an earlier law that made it easier to challenge overly broad “covered business method” patents. And they’ve already forced some legal changes in their favor, says Mark Lemley, Stanford law professor and author of a 2008 paper on universities and patent trolls. Under pressure from lobbyists in 2011, Congress lowered fees for patent applications that happen to originate at universities. Institutes of higher learning now count as “micro entities,” a designation normally reserved for individual inventors.
The nation’s institutes of higher learning may not intend to act like trolls, nor to benefit from others’ patent shakedowns. But they’re working in a system that sets its own rewards. Faced with these incentives, what a university wants and what it does could end up very different.
*Update, May 8, 2014: This post was updated to reflect that in addition to being at Harvard's Berkman Center, James Bessen is a lecturer in law at Boston University.
*Update, May 12, 2014: This post was also updated to clarify that before its patent auction, Penn State put in place guidelines to prevent bids from patent trolls, and to include a quote from a Penn State spokesperson.