Innovation Must Get in Line for Academic Funding
In a Q&A session, computer scientist Francine Berman, vice president for research at Rensselaer Polytechnic Institute, explains how funding decisions made in Washington help or hinder innovation at universities.
“I think researchers are really struggling to survive in a world where resources are really scarce and innovation is not always the highest priority,” says Francine Berman, a computer scientist and vice president for research at upstate New York’s Rensselaer Polytechnic Institute. Speaking as part of Miller-McCune.com’s series of interviews on the challenges facing research universities, she explained that the innovation enterprise requires complex scaffolding, access to a pool of adequately paid graduate students, up-to-date equipment and money for things like travel to professional conferences. All of these supports, and many others, face erosion due to limited funding.
Prior to joining the oldest technological university in the U.S., Berman was at the University of California, San Diego, where she directed the San Diego Supercomputer Center and its interdisciplinary staff of more than 250 scientists, engineers, and technologists. She was one of the two founding principal investigators of the National Science Foundation’s TeraGrid project, an effort to combine resources at 11 labs to create the world’s largest distributed computer for scientific research. Berman also directed the National Partnership for Advanced Computational Infrastructure, a consortium of 41 research groups, institutions, and university partners with the goal of building national infrastructure to support research and education in science and engineering.
The following is a condensed version of a conversation Berman had with journalist Ken Stier earlier this year.
Q: Rensselaer has about $90 million a year for research money. How much of that is from the federal government, and how much from what other sources?
A: I would say most of it is from the federal government but not all. It’s typically the [National Science Foundation], the Department of Energy and various agencies at National Institutes of Health. And then a few other agencies. But those are sort of the big three. Many universities, including ours, have little bits of Homeland Security, or little bits of other agencies, but those are the big three for universities. …
The challenge for today’s researcher is that, on average, they’re writing, four, five, six, seven proposals for every successful proposal they get. That is a lot of overhead. If you think about your science per dollar, your science per hour, the fact that the federal money is so difficult to get — that means that the research community is spending a lot of time not doing the research but applying for research funding. And that’s been a part of our landscape for quite some time.
Q: Do you expect that to get worse given budgetary constraints?
A: Well, we’re concerned because there are several things happening. Number one, the stimulus money was also stimulus money for intellectual capital, and that came out as programs for universities for research. And while that lasted, the success ratio was more like 1 in 3 proposals, rather than 1 in 5 or 6. That was a really terrific thing for researchers, but that money is, as of now, gone. And the prospects for another stimulus, or a continuation, or something like that, is not good. More generally, there is the problem of the economy that is just hard for everyone, including researchers. And it’s been hard for researchers for some time. And then, there is, of course, the third problem of the current political environment in Washington, it’s hard to know how education and research will get sorted out by Congress and in the administration. So yes, we are concerned.
Q: How much time does it take to make a proposal?
A: Oh, my gosh. The really big proposals are thousands of hours, arguably tens of thousands of hours. If it’s a huge proposal, you’ve got a number of institutions in New York to put in to it — so like Syracuse and Rensselaer working together. And all together I really wouldn’t be surprised if there was a hundred or more individuals involved spending a huge amount of time.
For a single investigator proposal, I wouldn’t be surprised if researchers invested hundreds of hours.
Q: I think I’ve heard you say you think the review process in Washington is quite lengthy.
A: Well, the review process varies per agency, so like everything else in Washington, you can’t really paint it all with a broad brush. [The National Institutes for Health] does things very differently than NSF. In NIH, you have the science panels, and they’ve been there for a long time, while the panels at the NSF are not permanent; they come together for different programs, and then they dissolve, and program officers have a different kind of a function at each. It’s just a very different approach. One can assume that everyone’s essentially doing their best to try to assess the goodness of a proposal. But, you know, results might vary.
Q: Is it all peer reviewed? And generally do researchers prefer that peer review?
A: I think you’d have to ask a researcher; I’m a researcher myself, and I’ll tell you my preference is that experts review the proposal. I think we all would like people who know the landscape to be reviewing our stuff and for those to be as objective as possible.
In my own particular research area with computer science and information technology, many of [my] peers are in the private sector, and these people are often included on these panels. Research really is taking place not just in academia — which does a lot of it — but there is also a lot of research in the private sector, as there is in government laboratories as well.
Q: Are you, in academia, competing against that as well?
A: Yeah, you bet.
Q: I wonder how you might describe how priorities have changed, say, around the [post-9/11] creation of the Department of Homeland Security. How did that change the programs and the proposals that the government is looking for? How important was that change?
A: Certainly over the last decade there has been a much more intense focus on security. One was in the physical realm with 9/11, with respect to critical physical infrastructure. At Rensselaer, we have a particular interest in that because we have oldest civil engineering department in the nation. [Founded in 1824, the private university is 187 years old.]
The other area is in the cyber-realm. With the rise of our dependence on electronic information, for progress and advancement in all kinds of things, I think you saw a much greater scrutiny and interest in cyber-security. … The whole community has really thought a great deal more about the very, very hard problems involved. …
A really good example of a need for cyber-security is as we get more and more successful in creating and disseminating electronic medical records. If we remain concerned about privacy of the patient — and I don’t see any reason that we would change on that — you could imagine we will have to be very, very careful about transferring that information, using that information, anonymizing that information, etc.
Q: I’m wondering how this significant shift in funding on the part of the federal government may have favored some schools — those with the intellectual capital that could address those issues — than others that are not so strong in these areas.
A: For all of us, we want to do what’s interesting. You’re never going to get any good work unless you feel passionate about it — never. Researchers know that and that there are huge problems based on things they’re interested in. That being said, it’s a lot easier to choose a problem in an area that’s being funded. We have a whole pipeline. Graduate students choose problems with their advisors, and as the huge number of interesting problems come up in cyber-security, and there is more funding for them, they may gravitate to those kinds of problems. So, then you start getting an increase in cohorts. Sometimes people stay in their chosen areas that are not particularly lucrative if they’ve been working in it and they love it, but then they might look at new applications of their work in areas that are being funded.
Q: I’m wondering if you think there may have been maybe too much of a pendulum swing in this area.
A: I think the answer is no. Market forces might guide trends within the research community, but at the end of the day research is about solving problems for which we do not now have the answer. Research in academia, in particular, is about solving problems without the worry that in 18 months we have to produce a product. And that’s really important because the hardest problems we have to deal with take a long track record and many good brains and a lot of innovation to solve. In the private sector, there’s certainly more pressure to produce products out of research.
Q: But do you have a sense that other things were neglected?
A: Yes, but I think there’s not enough money to do all of the things that are important. That will always be true no matter how much money we put into it. I think we are in a precarious place now: If you look at the amount of investment in innovation in this country, it is falling short of what we need now, and more worrisome is that it is falling even more short of what we need for the next generation. I think it is a real concern not to have the foundation of research and innovation that will keep the U.S. competitive over the next decade. That’s a concern because it’s the generation that’s going to college now, or elementary school, or preschool, that are going to be taking over. Giving them experience with unsolvable problems, that’s research; training them in areas that are up to date, that’s education. Giving them the skills they need to navigate a complex world — that should be a big priority for us, but I don’t think it’s as big as it needs to be.
Q: How do you evaluate how well you’re doing with your research resources? What sort of metrics do you use?
A: There are two goals in the office of research at Rensselaer. The first thing is to help our researchers have impact, because no one wants to work and do something incremental. Everybody wants to come to work and be innovative and really make a difference. And the way you measure impact is often through feedback that you get through your community, or winning awards or prizes: Are our fellows in their professional societies? Are they in a national academy? Are they being cited for their publications? Those kinds of things indicate impact. Are they being picked up by companies? Are they creating paths? Another measure of impact is funding success, so we want to see our levels of funding get ever higher every year. There are two kinds of funding: There’s the number of awards and the aggregate amount of money you get from it, and then there is expenditures — how much of that award money is being spent per year? So typically universities want their funding awards and their expenditures per year both going up every year.
Q: How are you doing on these two goals?
A: I’m happy to say that Rensselaer is doing terrific on that. We got an 18 percent rise last year in awards, and the expenditures went up. Another measurement is the overhead that comes back to the university, which helps us run the university, to support the basic infrastructure of the university. All universities do that, and that’s negotiated between universities and the federal government — [and can run from 20-60 percent] — and what you want to see is the overhead return go up every year, and for us it did.
Q: How much have you turned to corporations for help in research? How do they figure into your efforts here?
A: There’s a really positive type of win-win situation for universities to work with corporations because oftentimes corporations don’t have the kind of freedom to work on certain kinds of problems and explore certain ways of thinking, certain kinds of approaches, because they need to make a profit.
For universities, corporations are a great source of real-world problems. They often have real-world data to share. They’re interested in our students. There’s interest in mentoring students for jobs. So, they’re great partnerships.
Q: Any examples that come to mind?
A: There are a number of industrial partners that run their codes at our supercomputer center, [the $100 million Computational Center for Nanotechnology Innovations]. They pay some money and they get some help, and the supercomputers provide them computational cycles on bigger machines than they have in their own facilities. Our machine runs 90 to 100 trillion calculations per second, so it’s a good-sized machine to get things done very quickly. Oftentimes, they need some help writing in their codes, so they can run their particular programs. IBM has worked with us to provide a really terrific deal on the machine, and the state of New York helped, too. And this was a win-win-win for Rensselaer, the state of New York and IBM to get this machine out there and used by an academic, private-sector constituency, and a state constituency — I’d say this is a great partnership.