Can NSF centers achieve sustainability after graduation? evidence and argument

In my last post, I promised a look at why the term “graduated” makes little sense to me in the context of the National Science Foundation’s centers programs. I can adduce some theoretical arguments for my position, and I’ve also assembled a quick-and-dirty table (after the break) summarizing what I could find via superficial Web searching about the current status of centers that NSF regards as “graduated” from the ERC, I/U CRC, and MRSEC programs.

This empirical exploration revealed a bit more evidence for sustainability than I’d suspected, but on the whole I think it’s still unduly optimistic to believe that once federal funding stops, a university-based center will necessarily retain the essential characteristics of what was originally envisioned. That’s not to say the funding has been a failure, only that expectations for sustainability may be unduly high, or perhaps irrelevant.

Of course, it’s the ambition of every funder with public goals (whether federal or philanthropic) to leave behind an enduring, systemic change — if you like, to create a lasting institution focused on the particular set of issues of concern to the funder. However, when you think about what that would mean in the university context, it’s simply not realistic. Universities endure. Centers, typically, do not.

The long-term fixed asset in university research is the faculty member, whose loyalties are actually owed primarily to his own discipline rather than to the nominal hierarchies of university management ((As was observed years ago by the brilliant organizational theorist Henry Mintzberg.)). Each faculty member operates much like an independent entrepreneur, expanding or contracting his business (i.e., his “lab”) by hiring or attritting labor in the form of graduate research assistants, postdoctoral students, and professional research staff — all according to the amount of external financial sponsorship (“soft” funding) that can be secured. Typically no one in the lab except the faculty member (and sometimes not even all of his salary) is paid from university budgets funded by “hard,” assured sources like tuition, fees or endowment revenue.

That means that when grants expire or disappear, behavior reverts to the organizational norm: what you might call dollar-tropism. The faculty member remains in place, engaged in essentially the same set of research interests, but if she wishes to maintain the size and staffing level of her lab and to provide continuity for research assistants who are actually graduate students on a multiyear dissertation cycle, she will find a way to couch the same underlying topics in forms that appeal to a new or different set of sponsors. As a consequence, whatever color or flavor a new funding source wants, they will get — provided there’s a match with what the faculty member and her lab team can provide by way of knowledge and services. But then the activity takes on the character of the new funding source: if features like interdisciplinary collaboration or industrial knowledge transfer are not important to that new source, then they will inevitably fade from the active research agenda.

What was innovative about the large-dollar NSF centers programs — when 30 years ago they began to compete with “investigator-initiated” grants for faculty attention — was their insistence that more than one faculty lab work together as a “center,” often on topics that transcended disciplinary and/or institutional boundaries, and often also in collaboration with industrial labs and scientists. To incentivize these arrangements, center grants provided enough funding to pay for center infrastructure such as an administrative director, an associate for industrial liaison, and sometimes clerical support staff who could be dedicated to center work rather than shared across an overburdened academic department or chairman’s office. Sometimes the university or state government kicked in matching funding for facilities so that faculty labs previously scattered across departments or schools could be collocated, especially around expensive equipment that needed to be shared in common. But the work unit is still the faculty lab!

When government funding provides a high level of support, faculty members can be induced to plan their labs’ research goals around commonly agreed themes and to write new, complementary grant proposals through the center rather than via home departments (as would be conventional with investigator-initiated grants), thus expanding the scope and scale of the center beyond what the federal grant actually pays for. Faculty may even agree to discuss and transmit their findings in an integrated way with the same set of federal and industry sponsors — but when center funding does finally disappear, the organizational glue is significantly if not fatally weakened.

With money no longer available to pay for common infrastructure, the faculty member may no longer see an incentive to collaborate, and work can easily migrate back to disciplinary silos or subcritical-mass groupings of faculty labs. While funded, a center may display impressive team spirit, depending on the quality of its academic leadership, but without continued core funding, it amounts to a mere cost-accounting construct, and one of indefinite lifetime at that. No senior university administrator ever mistakes this temporary status for a long-term institutional commitment to an academic department or an endowed institute. Defunded centers are not mourned: if not viable, they are simply shut, and if a show of sustainability is nonetheless required, they are kept in what might be called undead status.

Therefore, writing a workable sustainability plan — as the NSF now demands of its center grantees — depends on building structures that can not just replace the federal “glue” funding so that thematic focus can be maintained. I submit that this is much, much harder than it looks.

So much for argument from first principles. Now, what does empirical examination show? Using commonly available sources for three main centers programs — the Engineering Research Centers (ERCs) ((http://www.erc-assoc.org/erc_affiliates_list.htm visited 7/6/10.)), the Industry/University Cooperative Research Centers (I/U CRCs) ((http://www.nsf.gov/eng/iip/iucrc/directory/index.jsp visited 7/6/10.)), and the former Materials Science and Engineering Research Centers (MRSECs) ((http://www.mrsec.org/taxonomy/term/161 visited 9/15/10. )) — I built the table at the bottom of this page, in which I report my simple observation of whether the graduated center still has a website, and my subjective assessment (after visiting that site briefly) whether there remains in place an organization doing pretty much the same thing.

Results were a little better than I had expected. Of the 36 graduated centers pooled from these three sources and observed admittedly superficially in this way, only seven had no website or one that was clearly abandoned, or one that stated plainly the center had closed (a few more seemed to be under reconstruction). Of those that remained “live,” the picture was a bit murkier.

There were, of course, a few centers that had been renamed (indicating a change in theme, a shift in the winds of what newer sponsors were willing to pay for), or transformed into a larger umbrella institute (not a failure by any means, if one wants to build capacity for industrially relevant interdisciplinary research as a whole, but not direct sustainability in the sense that NSF means it). In some cases, a nominal center structure seems to still exist, including possibly still an industry advisory board or affiliates council, but activity seemed much lower than it would have been during the period of funding, or industry funding had disaggregated into individual rather than pooled projects (as, indeed, faculty teams might also have).

Interestingly, the program where the graduated centers looked most like the current centers was the MRSEC — the least industrially oriented of the centers programs other than the Science & Technology Centers. ((ot treated here, since the NSF frankly acknowledges that at least the 1991 and 1989 vintages are “retired,” not “graduated,” and that their websites are essentially archival.)) That makes sense to me. Faculty can continue to apply as teams under investigator-iniated programs, and may continue the same one-on-one industrial relationships they had when operating as a center, when the essential program is highly basic in nature. On the other hand, when the research is more use-inspired, ((Of course, that’s the late Don Stokes‘s term. )) as in the case of the I/U CRCs and even the ERCs, then the “glue” elements that NSF once funded directly become much more important and more noticed in their absence. It’s far from clear to me that some of these graduated centers are actually more than their websites, that they actually have the capacity to take in new research projects as an organizational unit, rather than simply displaying the results of individual faculty initiatives.

At the level of analysis I’ve conducted, I can’t make a quantitative assessment, but I invite you to follow the links and make your own judgments — if nothing else, it’s a good master’s thesis project for someone in S&T policy!

But what it comes down to for me is that the NSF should be more specific and modest about its sustainability goals. It’s simply not realistic in most cases to expect that the particular structure and thematic focus will endure. Maybe one, maybe the other, but unlikely both. Usually the only source of “glue” funding available to replace NSF funds is the pool of affiliates fees, and these loyalties are fickle at best. Unless the center has been elevated to the status of a major university institute (as does seem to have happened in a couple of cases), no university money will be available for glue. After all, as a wise physicist once observed, universities are galactic funding “sinks,” not funding “sources.”

It’s probably enough to hope that 5 to 10 years of sustained funding has built relationships that will endure; that matching awards have brought equipment that elevates the quality of work that can be done on campus; and that the culture of the institution has changed to embrace industrial partnerships, K-12 outreach, and the other accoutrements insisted on by the NSF. But don’t require grantees to pretend they can deliver what they can’t, because the way universities work, faculty will always rotate toward the next available source of funds. That’s not bad; it’s just the way it is.

More at a later date about industry affiliates programs and other issues (such as inter-campus collaboration) related to the NSF centers.

NSF centers programOfficial name while an NSF center (and university hosts)Year of final graduation from NSF centers fundingCurrent status of websiteCan it be judged active with more or less the same mission as while funded? Other comment.
ERCERC for Intelligent Manufacturing Systems re-established as the ERC for Collaborative Manufacturing (Purdue)1999GoneNo — described as a "cluster" area of the Advanced Materials and Manufacturing signature area" of Purdue Engineering.
ERCCenter for Advanced Engineering of Fibers and Films (Clemson, MIT, Clark Atlanta)2008LiveUnclear — "Center research structure is currently being revised." Latest posters from 2008.
ERCERC for Net Shape Manufacturing (Ohio State)1997LiveYes — but apparently low activity.
ERCERC for Emerging Cardiovascular Technologies (Duke, UNC, others)1998GoneNo — referenced without details as an historical component of the Dept. of Biomedical Engineering
ERCInstitute for Systems Research (Maryland, Harvard)1996LiveYes — but has become a major umbrella institute for interdisciplinary research, bigger but arguably less focused than before
ERCERC for Reconfigurable Manufacturing Systems (Michigan)2007LiveYes — but apparently low level of activity.
ERCERC for Particle Science and Technology (Florida)2007LiveYes
I/U CRCCenter for Advanced Computing and Communication (NC State and Duke)UnknownRenamedYes — but renamed as Institute for Next Generation IT Systems and operating at a low level of activity. No administrative contact.
I/U CRCMeasurement and Control Engineering Center (Tennessee and Oklahoma State)UnknownDead linkProbably not — no evidence of center. However, a later center for Reliability and Maintenance has a strong industrial program.
I/U CRCSilicon Wafer Engineering & Defect Science Center (NC State and Berkeley)UnknownLive but at UT DallasProbably yes — evidence of industrial program, but much broader academic consortium.
I/U CRCWireless Electromagnetic Compatibility (Oklahoma)UnknownLivePossibly — low activity.
I/U CRCIndustry/University Center for Biosurfaces (Buffalo)UnknownLiveProbably — but low level of activity, few sponsors.
I/U CRCSurface Engineering and Tribology (Georgia Tech and Northwestern)UnknownGoneNo
I/U CRCSoftware Engineering Research Center (Ball State and Purdue)2004LiveYes — but partly due to renewed NSF funding. Combined with graduated Center for Information Processing at Iowa State and re-funded as a new I/U CRC
I/U CRCParticulate Materials Center (Penn State)UnknownUnmaintained (retrieved September 2010)Probably not — No evidence of the center within the broader Materials Research Institute
I/U CRCCenter for Glass Research (Alfred, Missouri-Rolla, Penn State)UnknownNominal (NSF link out of date)No — clearly subsumed into umbrella Center for Advanced Ceramic Technology.
I/U CRCAdvanced Steel Processing and Products Research Center (Colorado School of Mines)UnknownLiveYes
I/U CRCCenter for Advanced Manufacturing and Packaging of Microwave, Optical and Digital Electronics (Colorado)UnknownGoneNo
I/U CRCCenter for Microcontamination Control (Arizona and Northeastern)UnknownArizona (HQ) says center is closed but Northeastern branch carries onNortheastern only.
I/U CRCCenter for Advanced Studies in Novel Surfactants (Columbia)UnknownLiveYes — but note Columbia also participates in a newly funded related I/U CRC.
I/U CRCCenter for Health Management Research (Washington)UnknownGoneNo
I/U CRCBerkeley Sensor and Actuator Center (Berkeley and Davis)UnknownLiveYes
I/U CRCCenter for Nondestructive Evaluation (Iowa State)UnknownLivePossibly — Refers to a larger umbrella center in which the cooperative research program is embedded, with other components funded by NASA, FAA.
I/U CRCCenter for the Built Environment (Berkeley)UnknownLiveYes
I/U CRCCenter for the Management of Information (Arizona)UnknownLiveYes — but only academically. Industrially probably not. Name refers to a center which no longer carries any reference to an industrial affiliates program.
I/U CRCCooperative Research Center in Coatings (Southern Mississippi and Eastern Michigan)UnknownLive (NSF's link out of date)Possibly not — no obvious recent activity. However, a MRSEC also active on campus.
I/U CRCCenter for Precision Metrology (UNC Charlotte)UnknownLivePossibly — now refers to a much larger entity of which the affiliates program, if active, is now a minor component
I/U CRCCenter for Pharmaceutical Processing Research (Purdue, Puerto Rico, Minnesota, Connecticut)UnknownLiveYes — with minor renaming.
I/U CRCCenter for UMass/Industry Research on PolymersUnknownLiveProbably — name refers to a larger center in which consortial industry research is embedded, but split into clusters, and side by side with "one-on-one" projects
MRSECGarcia Polymer MRSEC (Stony Brook)UnknownUnder reconstruction in September 2010Maybe — unclear.
MRSECMRSEC (Virginia)UnknownRenamedNot exactly — reinvented as nanoSTAR Institute with broader mandate.
MRSECCenter for Response-Driven Polymeric Materials (Southern Mississippi)UnknownLiveYes
MRSECCenter for Materials for Information Technology (Alabama)UnknownLiveYes
MRSECCenter for Nanostructured Systems (Columbia)2008LiveYes
MRSECCenter for Thermal Spray Research (Stony Brook)2007LiveYes — but looks thinly staffed and low level of activity.
MRSECCenter on Polymer Interfaces and Macromolecular Assemblies (Stanford, IBM, UC Davis and UC Berkeley)UnknownLiveYes — but low level of activity.
0 comments

Leave a Comment

This site uses Akismet to reduce spam. Learn how your comment data is processed.