A Challenging Matchup

By Thomas K. Grose

With a hint of nostalgia in his voice, Joe O’Brien recalls an era that ended some 20 years ago. He remembers when corporate-sponsored research contracts with university labs were casually reached over a cup of coffee with the faculty member who would lead the investigation. Back then, “we could have a collegial dialogue with faculty,” and deals were quickly agreed upon, recalls O’Brien, university relations program manager at Hewlett-Packard, an information technology giant that can trace its lineage back to campus labs at Stanford University.

The reality today is that research negotiations often get hung up on issues involving intellectual property (IP) rights: Who owns them? Will they be licensed nonexclusively or exclusively? And at what cost? Difficulties in resolving questions like these can lead to negotiations lasting a year, or even two. For industry, that’s much too long and costly a wait. Notes R. Stanley Williams, director of HP Quantum Science Research: “All too often, the company spends more on attorneys’ fees than the value of the contract being negotiated.” Companies complain that too many university technology transfer administrators have an unrealistic notion that they can make money off of all research.

Moreover, companies say, although they paid for the research, if any useful IP results from it, schools claim it’s university property and sponsors are entitled only to first crack at negotiating (and paying for) a license to use it—which smacks of being asked to pay for it twice. Research schools, however, say companies fail to take into account sunk costs needed to build a lab’s infrastructure and body of expertise. The problem, says Gerald Barnett, director of the Office for Management of Intellectual Property, University of California, Santa Cruz, is that industry considers IP contracts another form of procurement; but to schools, they’re licensing deals.

The upshot is that Corporate America is increasingly moving academic research programs to schools overseas, particularly to the developing world, where results are outstanding, costs are low and arguments over IP are nonexistent. That certainly is a cause for worry among America’s engineering school deans, many of whom are sympathetic to industry concerns. They fret that the negotiations impasse means that too many of their faculty and student researchers are losing opportunities to work on useful and lucrative corporate research.

To be sure, there is ongoing dialogue to bridge the differences and speed the pace of IP deals. Indeed, the American Society for Engineering Education (ASEE) is sponsoring an all-day workshop on intellectual property negotiations in February in Washington, D.C., that will bring together engineering deans, industry officials and tech transfer administrators to talk about problems and possible solutions. But the issues are complex and, in some circumstances, based on varying interpretations of federal law. Another hurdle is a lack of unanimity within industry on what the problems are and what fixes are needed. Notes a white paper, commissioned by ASEE to coincide with its workshop and also the basis of this article: “… each sector tends to approach IP rights negotiations with different—and sometimes conflicting—goals. What causes problems in one sector can be the complete opposite of what’s a stumbling block in another.”

Certainly, the trend in federal spending for research and development in engineering and the physical sciences can make one sympathetic to the plight of research schools and their need to find new sources of revenue. According to the National Science Foundation, federal spending has been essentially flat for about 30 years, fluctuating between $5 billion and $7 billion. What critics say is a mistake, however, is when schools try to compensate for lost federal funding with revenues from IP royalties—particularly in the area of engineering. Only very few schools gross large sums from royalties. And most of the winners hit the jackpot with discoveries in the life sciences, particularly pharmaceuticals. Individual engineering patents rarely have much value—particularly in the IT world. As University of California, Berkeley, business professor David Mowery recently told Fortune magazine, a single piece of hardware often comprises about a hundred different patents, which diminishes the worth of just one. But critics charge that too many technology transfer offices fail to distinguish between engineering and the life sciences and too often apply to engineering the “big hit” royalty model of pharmaceutical patents.

According to the Association of University Technology Managers (AUTM), many schools earn only enough from IP royalties to cover the costs of running a technology transfer office, and a significant number don’t even manage to do that: They’re in the red. “There is a lot of mythology out there” concerning royalties, says Don Giddens, dean of the Georgia Institute of Technology’s engineering school. And even if a tech transfer office’s overhead is only just covered by royalty revenues, “What are the benefits of that?” asks Nino A. Masnari, dean of the College of Engineering at North Carolina State University. In 2003, AUTM figures show that U.S. research schools completed 4,516 licensing deals and total licensing income reached an impressive $1.3 billion. But just over half of that cash went to just 10 schools. And even big grossers don’t have much left after expenses.

The University of California system grossed $74 million in licensing income in 2001, but expenses and other costs sucked up $69 million of that. John Preston, former head of the Massachusetts Institute of Technology’s technology transfer office and now a senior lecturer at the school’s Entrepreneurship Center, bluntly says, “Royalty income is such a horrible means of measuring success.” Schools should instead focus on wealth and job creation, economic development and corporate goodwill. Those results can help universities attract top talent, both faculty and student. Mark Crowell, associate vice chancellor for economic development and technology transfer at the University of North Carolina at Chapel Hill and current AUTM president, agrees but argues that, unlike a decade ago, few universities today expect IP to make them rich.

The Bayh-Dole Patent and Trademark Act of 1980 is credited with encouraging schools to commercialize potentially valuable and useful technologies. And that was the law’s intent: The government would permit schools to assume ownership of IP resulting from federally funded research and license it to businesses. But industry critics—as well as many engineering school deans—say schools now widely apply an overly broad interpretation of Bayh-Dole by arguing that they own all IP coming from their labs, including discoveries from industry-sponsored research. But as Crowell explains it, if a single dollar of federal money pays for equipment used, or a salary paid, Bayh-Dole does apply. But, he adds, other laws also give schools ownership of IP. Labs in campus buildings funded by tax-exempt bonds are, for all intents and purposes, barred by U.S. tax laws from assigning IP rights to for-profit entities. Pioneering tech-transfer schools like MIT had policies in place that assumed ownership of IP long before Bayh- Dole’s enactment.

If universities need to focus more on making technology available, industry should stop being hung up on ownership, Preston says. “ ‘We funded it, we should own it’ is B.S.” Santa Cruz’s Barnett says sponsored research builds upon a school’s knowledge base, a foundation that took many years and many billions of dollars to construct. “It is not just the professor you’re hiring, it is the entire infrastructure,” Barnett insists. Adds Preston: “Obviously a company is going to a university lab because the lab is uniquely capable of solving a problem the company has not solved.” And here’s another complicating factor: Not all companies care who owns the IP. Within the IT sector, ownership matters rarely, and most companies just want a nonexclusive license. But in other sectors, including aerospace and bioengineering, companies prefer to own the IP they funded; absent that, they almost always want exclusive licenses.

Having ownership of all IP coming from their labs is no reason for universities become greedy, Preston says, noting that no law, including Bayh-Dole, sets a minimum price for a license. “A university could own the IP and license it for a dollar,” or some other token amount, Preston explains. He recalls that when he ran MIT’s transfer office, he would figure out the middle pricing range, then go 10 percent toward the industry side when setting his starting price. His philosophy was that it was better to savor small nibbles from many big deals than choke on big bites from a few small ones. Nevertheless, it rankles industry negotiators to pay anything beyond the cost of the research. “We don’t feel that we should pay for a … license for IP we’ve already paid for,” O’Brien argues. When fees are paid, there’s a preference within industry to pay a lump-sum amount, not ongoing royalties. “The cost of tracking royalties through product cycles and derivatives costs more than the value of the invention,” O’Brien explains. Deborah Kilpatrick, director of new ventures at Guidant Corp., a California bioengineering firm, agrees. “Downstream royalties give us serious concern in early-stage research and technology development.” It is, she explains, very difficult to commit to them so far upstream of any commercial product.

Real Time

Beyond conflicts about ownership and payments, there are many other issues that can cause sponsored-research contract talks to drag on ad infinitum. And then the length of those negotiations becomes a problem, too. “Speed is a bigger issue than cost,” says Paul Peercy, dean of the University of Wisconsin’s engineering school. “GM says it is easier to merge one of its units with a company from Japan than to do IP negotiations with an American university,” says Peercy, who works regularly with the automaker. (General Motors funds an engine research center at Wisconsin.)

Contract talks can commence with each side having different time expectations. Says Barnett: “University sponsored-research offices usually work with a proposal and future deadline while companies work with a ‘need it now’ approach.” But MIT’s Preston says universities must learn to appreciate industry’s limited time horizons; many industries are pushed to ever shortening windows of opportunity by stock markets that want continuous performance. “Too many universities are too used to taking their time,” Preston says.

However true that may be, Crowell says there are many instances when it’s industry foot-dragging that forces talks into extra innings. “Companies can match us with delays in ways that are just phenomenal.” Crowell explains how in a recent negotiation with a big pharmaceutical company that once the agreement was finally drafted, the corporate team had to take it to an in-house management board for approval—not a quick process. “But I had signing rights. I was authorized to do a deal.” That sort of thing happens regularly, he says. “In a lot of cases, our (academia’s) processes are more streamlined.” However, Dale Parker, director of intellectual asset management at Northrop Grumman Corp., fires back a similar complaint about schools. Northrup negotiators, he says, often find themselves dealing with people who “have no power to negotiate the terms of the IP clause and don’t understand the policy behind them.”

Lawyers are a root cause for many delays. That’s something schools and industry agree on. “Both sides sometimes lose control of their legal counsel, which can lead to gridlock,” Barnett admits. “Legal counsel may have a different mandate (and line of reporting) from that of the technical or administrative points of a contract. Some negotiators simply ‘want to win’ rather than create an exchange of value.” Crowell agrees: “Either side can become so obsessed with winning, it loses sight of the objective.” Lawyers, Crowell adds, should be kept on the sidelines and “they should speak when spoken to.” But only then. There is no shortage of other issues that can bog down negotiations.

Schools and companies at times wrangle over control of the patent-filing process. Crowell says if a company opts to license a discovery, it typically wants its lawyers to write the patent. “We can’t let that happen,” he says. That would be like letting corporate counsel represent the school. But HP’s Williams is critical of the way universities sometimes handle patent applications, saying that they’re often too provisional and don’t contain carefully crafted claims sections. And they “are often too weak to defend,” Williams adds. Schools tend to wrongly apply the lessons of drug patents to engineering, he adds, and regularly overvalue the worth of a patent.

There’s not even consensus as to where most ideas for corporate-sponsored research come from. Santa Cruz’s Barnett says this is how many universities define sponsored research: “financial support for a university-supplied proposal.” But Williams, in testimony before a U.S. Senate committee in 2002, claimed that “in many cases, the root idea originated with the sponsoring company in the first place, not the university.” AUTM’s Crowell takes a diplomatic stance, saying both sides develop ideas. “It’s all over the map.” He adds, however, that industry is certainly interested on a regular basis in research funded by Washington or other sources.

Background rights are often a source of contention. That’s when a company wants to license any discoveries coming from research it sponsored and also rights to other relevant IP belonging to the university. “So if one research lab uses inventive stuff from another lab, the company wants to license that work—even though it has not supported the other lab—for the price of the research materials,” Barnett says. O’Brien, of Hewlett-Packard, says his company expects all background IP to be disclosed upfront and “to be included in the agreement, or we will enter into a separate negotiation prior to completing this one.”

Other times companies ask for, ahead of any research to be conducted, a grant of rights for all further applications and fields of use. These are called upfront licenses. For researchers who collaborate on the sponsored research but are also working on other projects, that’s problematic, Barnett says. “While an upfront license is convenient for the company, there is virtually no way to track what is and what isn’t included in the license.”

Northrop’s Parker says, however, that restrictions are typically by field of use—military, commercial and automotive, for example—and often are also limited to a geographical area. “The smaller the field of use or general use limitation, the more licenses—and revenue—a university may squeeze out of IP. This is true of companies, as well. I don’t see this as a unique issue.”

Companies also fear that new information will emerge in an area of research it sponsored that will give rise to new and possibly useful inventions or ways to work around discoveries they’ve licensed. So it’s another possible sticking point when a company asks for not only the results of research it sponsored but also information about any improvements or related inventions that may arise, and that can cause schools to balk, Barnett says.

Industry is aware that academic researchers must publish the results of lab work, but this is one area that seldom causes problems. Companies may ask for short publication delays, typically three to six months, and that’s usually ok with most universities. As Dean Giddens of Georgia Tech says, if a company hasn’t used an invention within six months, “it probably won’t use it.”

Efforts to solve these issues and speed the pace of sponsored contract negotiations are hampered because various industrial sectors don’t agree on what fixes are necessary. That elminates any notion of a one-size-fits-all solution.

The IT industry, for instance, shares an awful lot of technology. So it seldom cares who owns the IP, and it rarely wants an exclusive license. In fact, its preferred solution would be to put all discoveries in the public domain, which frees any company that sees a commercial potential for an invention to run with it. Williams, of HP, says since royalty income is often miniscule anyway, schools and researchers would be more than compensated by the resulting increased flow of direct research support to schools from industry. When it comes to licenses, IT companies usually prefer nonexclusive ones. Says Williams: “Many universities attempt to sell only exclusive licenses to companies, again following the false analogy of drug companies.”

But ownership of useful discoveries is preferable in many other sectors, including bioengineering and aerospace. Says Guidant’s Kilpatrick: “Ownership (of IP) is really, really important to us in early-stage biomedical research” because its ultimate value cannot always be immediately ascertained. Typically, her modus operandi in dealing with academic labs is if we invent it, we own it; if you invent it, you own it. But if you invent it while working on a research project or collaboration exclusively funded by us, give us first rights to an exclusive license. Northrop Grumman’s Parker also says: “We prefer exclusive use of the data and IP to develop discriminating technology” over rivals. If he were to go to his board of directors saying he wanted $3 million to pay a university to develop a new technology, but that the IP would then be made immediately available to Lockheed Martin, Raytheon and Boeing and Northrop also had to then pay for the license, “that would be a tough sell to the board.” In those sectors, solutions like public domain IP or nonexclusive contracts won’t work.

Kilpatrick agrees that slow-moving contract negotiations are an exasperating problem. But another timing issue for bioengineering companies is when schools push for quick licensing decisions. “Demanding an answer in a few weeks is often moving too fast for us,” Kilpatrick says, and what the time period will be tends to become a serious point of negotiation. Bioengineering companies usually seek early-stage discoveries from campus labs, which can be 10 to 20 years away from any of them getting to the healthcare market.

Farming It Elsewhere

Clearly, Corporate America has found one solution: taking its research proposals to foreign schools. That’s a growing trend among many Fortune 500 companies, Kilpatrick says, and for good reason. “They have really super-talented people, and you can have what they invent.” Williams told the Senate committee that not only are these schools eager to work with U.S. companies, but the negotiating process is quick and simple, sometimes lasting no more than a few minutes over the phone. “American universities will either have to modify their behavior or lose their industrial customers,” Williams told a senate panel. And as Wisconsin’s Peercy notes, the quality of foreign research schools is “top-notch,” with many of their faculty members having been educated and trained in the United States.

And American companies are not only taking research overseas, they’re recruiting the young graduate students they find there. That could affect efforts to find good jobs for U.S. grad students. Moreover, as the profile of foreign universities rises, and as their links to U.S. industry increase, their appeal to prospective students grows. That could make it harder for American universities to recruit foreign students. “There is a concern that the U.S. is losing ground in innovation,” Georgia Tech’s Giddens says, “and that is central to the issue of recruitment. It might be one way to get traction on this issue from the federal government.” Certainly something needs to be done make it easier for American research universities and industry to work together. Otherwise the migration of research work overseas is likely to continue to increase. That could not only hurt U.S. universities but also harm American economic development, job creation and innovation. The loss of industrial research work to foreign schools, Preston says, “should be a wake-up call to U.S. universities.”

Thomas K. Grose is a freelance writer based in Great Britain.