July Regents Meeting: Presidential Resolution Likely to Decrease rather than Increase Efficiency

The Finance Committee of the Regents will consider an item at their July 14th meeting called "Adoption of Resolution Regarding Administrative Efficiencies" (F2).  At first glance, it is guaranteed to increase the president's executive authority, but is unlikely to increase UC efficiency.

The resolution has the following features:

• It is the first item associated with the UC Commission on the Future (UCOF)  to be considered by the Regents.
• It was not a UCOF working group recommendation, but comes from one of the Expanded Recommendations inserted by UCOP into the documents for UCOF's fifth meeting in June (page 68, Recommendation 9).
• It starts with harmless "whereas" clauses about the value of administrative efficiency. Then, in Whereas 5, asserts that "the Regents consider administrative commonality and consolidation a requirement for reaching the efficiency objective."
  
• On the basis of this equation of efficiency with commonality and consolidation, it grants the President the power to approve and disapprove "all new or substantially revised campus administrative systems to ensure commonality and best practices across all locations" (Resolved clause 2).  Resolved clause 3 grants him veto power over any internal variation in the system. 
• The burden of proof is on the campuses to show that local campus needs are best served by customized systems.  The judge and jury would be the president.
• Having granted executive authority to the president, the 4th Resolved clause calls on the president to "employ the shared governance model."

This resolution empowers the president to centralize administrative systems, to centralize them around the presidency, and to select and reject directly proposed changes in administrative systems on the campuses. 

The resolution is a step backward.  It rests on a key management fallacy. The fallacy is that "commonality and consolidation" in themselves increase efficiency.  In reality,  centralization in most cases reduces efficiency in administrative systems. Centralization increases the number of layers of management, increases the authority of the supervisory layers over the conduct of the local activity, and reduces the control of front-line employees over the work they themselves perform.  Hierarchy, mediation and complexity reduce what we can call epistemological or communicative efficiency - information has to travel through more layers, becomes separated from practice, and gets distorted in the transmission process.  Executives become information poor even as they remain authority-rich, and the operational outcomes have been regularity critiqued in the management literature both before and after management classics like In Search of Excellence and Built to Last showed that long-term, sustainable, organizational success requires the engineering of administrative systems to reflect the deep intelligence of actual employee practice, liberated by non-authoritiarian managers to improve continuously. (Ironically, in the 1980s and 1990s, such classics were trying to make corporations into "learning organizations" that resembled universities.)

UC had its own upwelling of "Built to Last" analysis in 2007-08, and the current resolution is also at odds with these previous evaluations of UCOP practice. These include Regent Richard Blum's critical memo of August 2007 ("We Must be Strategically Dynamic"), and various consultancy reports, which, according to a UCPB overview that I helped draft, encouraged UCOP to be less secretive, more interactive, more supportive and more coordinative of campus initiatives rather than closed, top-down, and unilaterally directive. Administrative inefficiency was the problem, and network coordination was the solution - not centralization and standard practice.

One concrete UC example of the value of decentralization is the Office of Technology Transfer.  The UCOP office of that name used to handle everything related to tech transfer on the campuses from its central Oakland location -- all contract and grants compliance as well as patent licensing, industry research sponsorships, and so on.  A faculty inventor at UCSB would know her contracts and grants officer from the grant application process, and would have a network of colleagues in her field and specfic contacts in particular companies.  If Professor Rodriguez, let's call her, had an invention to disclose, the local C&G person would send it up to Oakland for processing.  Oakland decided whether the disclosure was worthy of a patent application, not Santa Barbara, and how the patent would be prosecuted.  If Prof. Rodriguez received a patent and knew a company with an interest in licensing this invention, it was again Oakland that would proceed or not, negotiate the terms, and decide the university's position, in effect controlling the outcome of the negotiation.  Both Prof. Rodriguez and the UCSB C&G officer were bystanders in a centralized process.

Though OTT had and still has a valuable compliance role, as well as experience and insight to contribute, the campuses argued that its distance from the inventor, local industry, professional networks, and so on caused it to make many mistakes. Just as importantly, OTT was more expensive to run than were the local offices.  One by one, the campuses achieved relative independence for their OTT offices. The decentralization of technology transfer and industry alliances away from UCOP has increased effectiveness and lowered costs.  It has not eliminated OTT input. It has made that input more reciprocal and interactive -- more intelligent, in other words, in part because it is now less unilaterially authoritative.

This doesn't prove that what worked for TT will work for all administrative systems.  That is in fact my point. You need to find out whether the problem is insufficient centralization for a specific function before you "solve" it with more centralization.

There certainly are exceptions to the value of decentralization is when a very high percentage of a system's functions really are in common. But one needs to have empirical data to decide how many of UC's admin functions really are common, and if so, which ones.  Are 85% of UC's functions common, or 15%? I think it is closer to 15% --  I don't even think payroll can be standardized, given what I learned about the extraordinary divergence of campus pay practices when I sat on the Salary Scales Working Group a few years ago.  But this needs to be determined, and the specific systems that would really be improved by centralization need to be specified.

The president's resolution would be plausible to me if it resolved to:

• use the resources of the president's office to identify those systems that really are common to UC, and would be imporoved through standardization and uniformity.
• authorize the campuses to engage each other directly to find their preferred solutions, and to offer a UCOP endorsement of any obtained consensus
• facilitate all campus-specific customized solutions that optimize local efficiency and leverage local knowledge as long as they do not thereby reduce the efficiency of the overall system.

Is it too late for the campus administrations to get the F2 resolution modified into something that will actually help the campuses work better?