Remodeling in 3–D
Up, down, and sideways. That’s how the University of Michigan studied its back-office services, applying private-sector principles to transform its administrative structure. The strategy features three critical elements—benchmarking, business case development, and an implementation road map.
By Rowan A. Miranda
At a time when many of its peers face significant financial challenges, the University of Michigan (UM)—widely considered a model public university—has managed to maintain financial stability while enhancing academic reputation. This progress has continued, despite the university's own recent fiscal restraints resulting from a decline in state funding from $416 million in FY02 to $308 million in FY12. Indeed, had the university's appropriation kept pace with the consumer price index for the Detroit region, state funds would have amounted to $498 million in FY12.
Amid these challenges, 95 of UM's academic programs rated in the Top 10 by the U.S. News & World Report College Rankings, while the university placed 18th out of 400 institutions in the Times Higher Education World University Rankings for 2011-12. UM also maintains the highest bond ratings possible from both major credit rating agencies, conducts $1.2 billion annually in research, and has the seventh largest endowment among universities in the United States.
To help offset the funding decreases and continue to invest in new priorities, UM's leadership spent much of the past decade implementing cost-saving measures. The university achieved general fund reductions and reallocations of $135 million between FY03 and FY09. Some of the strategies implemented during this period included improved purchasing practices, elimination of warehouses, energy-efficiency measures, health-benefit strategies, labor-saving enabling technologies, space allocation, and improved staff productivity.
Although challenging in their own right, these measures yielded only incremental annual savings of 1.5 percent to 2 percent. By 2009, UM leaders realized they had already picked the low-hanging fruit, and they would need bolder initiatives going forward.
Assuming that the economic climate would not improve markedly in the short run and state funding might well continue to decline, senior leaders acknowledged that the university needed to make even more significant changes to maintain its financial resiliency while protecting its academic quality. Specifically, achieving UM's reduction/reallocation target of $100 million for 2010-12 and $120 million for 2013-17 would require decidedly more aggressive and innovative measures than previously considered. Such actions would eventually involve a multidimensional look at costs and quality of administrative services—and designing a different model for the University of Michigan's administrative structure.
Begin With a Big Target
As newspapers from coast to coast frequently remind their readers, the cost of higher education has risen rapidly. While a number of factors can account for this continuing cost escalation—such as increases in the real price of inputs, higher output, or improved quality of academic and research programs—the widespread public perception points to unchecked administrative staffing growth as one of the main culprits.
In fact, Administrative Bloat at American Universities: The Real Reason for High Costs in Higher Education (Goldwater Institute, 2010) documents that tuition at leading public and private universities grew 67 percent between 1993 and 2007. During the same period, administrative staff growth at these institutions significantly outpaced increases in positions related to teaching and research. The report's authors argue that, rather than achieving economies of scale in administration, higher education institutions are becoming less efficient as they grow, which leads to tuition-increase spirals.
Universities undertake a broad range of programs and activities, so why focus on back-office functions such as finance, human resources, payroll, procurement, IT, communications, facilities, fundraising and development, student services, and research administration? Administrative services offer a prime target because they typically consume 15 percent to 30 percent of the budget. Simply dispensing with administrative services as budget pressures increase isn't an option; without administration, most organizations would stop operating altogether.
Yet cost escalations in support functions detract from teaching, research, public service, and patient care. While few organizations can garner support for cost cutting for its own sake, most stakeholders like the prospect of reinvesting realized gains from greater efficiencies in the academic mission.
By its nature, an administrative services transformation program goes against the grain of long-held traditions and norms.
With that in mind, UM's CFO, provost, and executive vice president of medical affairs sponsored a campuswide study of the costs and quality of administrative services. Conducted in mid-2009, the study aimed to pinpoint opportunities for efficiencies that might help the university meet its cost-reduction targets in FY10 and beyond. UM's finance department, which served as the project manager of the overall effort, retained the Hackett Group to conduct the benchmarking analysis.
Four areas formed the core focus of the study: information technology, finance, human resources, and procurement. In addition, more limited data were collected on headcount and transactional output for communications, development, research administration, and student services.
Some Surprising Findings
The University of Michigan has an enrollment of 59,000 students on three campuses, supported by 42,000 faculty and staff. Prior to the study—which involved only the main campus in Ann Arbor—it was difficult to determine the cost of specific administrative services throughout UM's 19 schools and colleges because resources were fragmented across central units, academic units, and auxiliaries. The study enabled us to measure, for the very first time, what was actually being spent by functional area and business process, as summarized in Figure 1.
Through specific metrics—such as cost per invoice, errors in transactional processes, HR cost per employee, level of technology utilization, use of strategic suppliers, and IT cost per user—benchmarking also enabled us to compare UM's performance in various administrative services to the performance of some of its peers and of similar-sized organizations across industries. Here are some of the study's findings, many of which demonstrate opportunities for improvement at UM:
- About one third of the university's administrative costs are centrally located, with two thirds occurring at the local or unit level. This finding ran counter to the common perception on campus that the highest portion of administrative costs sits in central units.
- Approximately three fourths (75 percent) of the work done at the unit level is transactional in nature rather than being strategic, advisory, or analytical. Again, this finding did not support the common perception that units do specialized work that can't be done by a central service provider. As one dean commented, “Until I saw the data, I always believed our people did strategic, high-value work. Not only are we not doing that, we are a very expensive provider of transactional work.”
- Rather than engaging in strategic, analytical, or advisory work—such as assembling a budget or improving business processes—UM managers and professionals spend more time (compared to benchmarks) on essentially clerical tasks, such as processing invoices.
- Opportunities for process improvement abound. Many units, for example, maintain a primarily paper-based workflow for HR transactions rather than an all-electronic one, and they underutilize self-service technology.
- Many units maintain shadow systems that replicate what central systems do better. As an example, UM made an investment of more than $100 million in an ERP system 15 years ago, yet some colleges and schools have continued to develop redundant or “shadow” systems.
- UM makes minimal use of outsourcing, even for functions such as commodity IT services, where stable and thriving competitive market providers operate.
- Hyper-decentralization of operations often means layer upon layer of administrative organizations. A large college may have a central accounting function and department-level accounting, as well as accounting within programs and institutes.
Closing the Gaps
After review—and considerable discussion—of the results and data comparisons, the study's executive sponsors identified two major initiatives to address UM's performance gaps: IT rationalization, and administrative services transformation. These opportunities have a combined savings estimate of approximately $60 million to $70 million per year. The IT rationalization project includes initiatives such as consolidating e-mail systems and desktop support, eliminating redundant storage systems and reporting applications, developing a high-performing computing service for researchers, and network consolidation.
The other initiative—what UM decided to call administrative services transformation (AST)—represents an organizationwide effort to deliver sustainable reductions to the cost structure while at the same time enhancing the quality of service. AST tailors best practices from the private sector to the unique operating environment of higher education, implementing new service delivery models built upon business concepts such as enabling technology, shared services, strategic sourcing, process reengineering, and business process outsourcing.
The University of Michigan approached AST much like an enterprise systems implementation, developing a comprehensive, four-phase strategic effort (see Figure 2). Phase I builds the overall strategy for the program and includes three critical elements—benchmarking, business case development, and the implementation road map. Phase II focuses on the design of various components, such as new business processes, organization structures, service management, enabling technologies, and workforce impacts. Phase III involves implementation of the designs and includes activities such as building new service delivery organizations, configuring new applications, delivering training, or building out new facilities. Phase IV involves deploying each initiative in a sequence established during Phase I.
This timeline requires 36 to 48 months, underscoring that AST is not something undertaken lightly or as the answer to an immediate budget crunch. Transformation involves radical or non-incremental changes to organizations, allowing them to migrate from “low efficiency” and “low-to-moderate effectiveness” structures to higher performance on both dimensions. While the term “transformation” is often used liberally by consultants and administrators alike, the scale and ambition of AST justifiably merits its use.
Seven Steps to Strategy
In addition to explicitly defining goals and the steps for achieving them, any strategy must also take into account environmental opportunities and threats, competitive positioning of services, skill and resource levels, stakeholder support, and governance (decision structure and process). Because Phase I of AST provides the foundation for each successive phase, the resulting strategy needs to be both fact-based and inclusive of senior executives and key stakeholders across the university.
Following are the seven steps that form Phase I of UM's administrative services transformation.
1. Establish strategic intent. This step involves assembling key stakeholder groups—such as central service providers, school- or unit-level business officers, and deans—to define and confirm expectations. For example: Does the university seek to reduce costs, improve service quality, or both? Are there certain units or functions outside the scope? How will major decisions be made? How and when should information be communicated to campus? Who funds the investments, and who retains the savings?
UM leadership facilitated several workshops so participants could become comfortable with goals, approaches, and expected outcomes. These forums also established a set of guiding principles for our overall effort.
2. Benchmark performance. You can't apply a solution until you fully understand the problem—which requires measuring an organization's performance and comparing it to its peers. You might, for example, find out that an accounts payable transaction in your organization costs $30, compared to $15 per transaction in peer organizations and to $5 per transaction in the highest-performing private sector companies. Assuming you could at least match the performance of similar organizations, how much savings could your institution achieve?
In our case, the Hackett Group compared UM's performance to other industries in its database and to organizations demonstrating world-class performance in finance, HR, IT, and procurement. Overall, across all functions, organizations performing at a world-class level report costs that are 24 percent to 41 percent lower than the average across all industries in the Hackett Group database in 2010.
3. Assess the current state of operations. With the benchmark process complete, we conducted interviews with university professionals and collected additional data on a unit-by-unit basis. This enabled unit staff to assist in evaluating features of service delivery such as staffing, process standardization, customer service, organizational complexity, utilization of enterprise systems, reliance on shadow systems, and capacity for change.
4. Develop operating models for the future. Based on benchmarking data and results from the current operations assessment, the project team proposed alternative modes of operating to reduce performance gaps. Examples of operating models include organizational restructuring, ad hoc staff sharing/resource arrangements with other units, creation of shared service centers, and outsourcing
of commodity services.
5. Evaluate the business case. Typically, a business case quantifies the magnitude of factors, such as cost savings and quality improvements, against an “as is” or “do nothing” baseline. Sound business cases don't stop at benefits and costs, however; they also describe risks related to each action, assumptions that underlie the model, and the level of investment required to achieve specified levels of performance. Business cases should be thoroughly discussed with the different advisory committees and workgroups that participate in an AST program.
To help develop the business cases, UM retained two outside management consulting firms with expertise in finance, procurement, and HR service delivery. After achieving a reasonable degree of consensus among the advisory committee and workgroups that represent UM's key stakeholders, the project team shared the results with executive sponsors and a broader group, including budget administrators from the colleges and auxiliaries.
6. Confirm strategic initiatives. A business case might identify a dozen or more economically justifiable initiatives. Implementing all the initiatives at once, however, might prove problematic because of competing initiatives, lack of leadership, skill levels of executives and managers, project management discipline, organizational culture, and available resources. Although consulting firms can help address some of these issues, few universities have a track record of successfully implementing large-scale programs in a short time frame. For this reason, UM prioritized the business cases and selected a few for implementation.
7. Build and source the road map. The strategy phase concludes with a description of how the implementation process will unfold. In building and sourcing the road map for UM, we asked such questions as these:
- Do we implement shared services in IT, HR, and finance all at once?
- Do we instead start with one major function, prove the concept, and develop a subsequent plan?
- What is the appropriate timeline for our university?
- What can we do to minimize business disruptions?
- Do we really have the attention and support of our top executives?
- Have we invested enough in training, communications, and change management?
- What is the realistic level of consulting support that we will need, given that we've never implemented change on this scale?
- How will the project be staffed at the central and unit levels?
To date, the university has completed Phase I (strategy) and much of Phase II (design) for AST, spending about $1.3 million and $1.5 million, respectively. We estimate that Phase III (implementation) will take a one-time investment of $30 million to $40 million to generate recurring cost savings that are estimated at nearly $270 million over 10 years. Many schools and colleges have already begun implementing shared services concepts—such as key performance indicators, service-level agreements, customer orientation, and centralization at the college level—prior to universitywide changes. To lead the way, the university's executive offices have already combined their accounting, purchasing, and HR staffs into the Fleming Administration Building shared services unit that now serves the offices of the president, vice president of communication, vice president of government relations, vice president and secretary to the Board of Regents, and general counsel.
Although we remain focused on strengthening the university's financial position in 2013-17, UM has already realized savings. In the area of strategic sourcing, for example, we have taken steps to secure $10 million in savings per year, bypassing our original target of $7.5 million. In health benefits administration, we have implemented steps to save approximately $3 million per year.
By its nature, an administrative services transformation program goes against the grain of long-held traditions and norms. Initiating one requires you to navigate the treacherous waters of institutional inertia, unearth entrenched business practices, and overcome opposition—which often increases, if only to preserve the feeling of stability associated with the status quo. In all organizational change efforts, but especially those in higher education, don't expect to simply document analysis and facts in a business case and have it drive a decision process. In fact, depending on organizational complexity, approximately 15 percent to 25 percent of overall programmatic effort in an AST program should be dedicated to change management.
The University of Michigan's experience confirms that change management includes patiently working through concerns and assumptions held by managers and staff about goals, roles, and job security. Other change management activities include communications, informal networks, executive support, and training.
As an example, we calculated that UM could save about $1.6 million annually by using remanufactured toner cartridges. Mandating the purchase of remanufactured cartridges, as a corporation or government agency might, could have resulted in a negative reaction on campus. Instead, we educated administrators about the environmental benefits of refilling and reusing the plastic cartridges that would otherwise end up in a landfill. By riding the wave of environmental sustainability, we encouraged behavioral change that has allowed us to meet our savings goal for toner.
AST efforts include the collection and analysis of data for presentation to UM decision makers. This enables senior administrators to show deans, for example, how much their schools could have saved on computers, office supplies, or scientific supplies had they used one of UM's strategic contracts for a higher percentage of their total annual purchases. (Translating those savings into potential faculty positions proves especially persuasive to deans.) With such data in hand, budget conversations become more concrete and solution-oriented.
Most of the AST program is still in the implementation phase. Based on the results so far, we remain confident that adapting best practices from the private sector can help the University of Michigan achieve a large portion of its $120 million cost-reduction target for 2013-17 and enable it to invest those savings in its teaching and research mission. To keep our efforts on track, we recommend a simple philosophy: Think big ... but implement incrementally.
ROWAN A. MIRANDA is associate vice president for finance at the University of Michigan-Ann Arbor. He teaches in the university's Gerald R. Ford School of Public Policy and in the UM Law School.