Data Dashboards, Higher Education Finance, and Student Outcomes

In fall 2016 the University of Tulsa administration abruptly announced that it would “temporarily suspend contributions to retirement for all employees” (see appendix A). Almost three years later, another TU administration announced its plan to cut 40 percent of the university’s academic programs and reorganize most surviving academic departments into multidisciplinary divisions, despite all the data that shows teaching quality is the most important factor in improving student outcomes.

TU has a billion-dollar endowment and only four thousand students, so the announcement of severe educational budget cuts caught my attention. I wondered, Why can’t TU operate like a “normal” university? At that time, I believed most institutions prioritized instruction and the educational mission in their budget decisions. This question inspired a major shift in my research. My earlier work had focused on pre-K through secondary education policy, but I have recently turned my focus to higher education finance. What I have found in this line of research is disturbing. I have seen tremendous waste, misplaced budget priorities, misleading and false statements from university administrators, and institutions appearing to abandon their educational missions in favor of maximizing profits.

Unfortunately, I can provide many examples beyond TU. The president of Mills College, which had a sizable endowment and valuable land holdings, falsely claimed it did not have the ability to continue as a degree-granting institution. The Saint Xavier University administration pushed to cut instructional spending while the campus generated outsized profits. And I have seen many other colleges and universities prioritize cuts to instructional spending over cuts to all other functional spending categories.

This problem of misplaced budget priorities raises the biggest red flag. Since at least the late 1960s, education researchers have known that the most important factor influencing student outcomes is teacher quality. Increasing administrative and other noninstructional spending, improving the quality of buildings, and even reducing class size do not have a significant influence on student achievement in comparison with even modest improvements in overall teacher quality.

The implication of this research is clear: to improve student outcomes, administrators need to promote policies and practices that allow the institution to attract, hire, and retain the best instructors. To support this objective, administrators need to pay faculty and staff more than the competition, offer better working conditions, and foster high morale on their campuses. It is simple labor economics. Budgets should maximize instructional spending and minimize other noninstructional spending, particularly spending on administrative functions.

Yet many upper administrators and trustees in higher education ignore data about what interventions make a difference for student outcomes and instead do the opposite. They push practices and initiatives that bloat administrative spending while shrinking budgets for faculty and instruction, leaving faculty members feeling demoralized and undersupported. In such environments, student learning, enrollment, and retention inevitably suffer.

Ironically, administrators often promote the idea of data-driven decision-making but fail to make data-driven decisions themselves. In my current research, I have seen several instances of administrators overemphasizing enrollment or revenue losses to justify their decisions to cut instructional spending, frequently ignoring or even increasing excessive administrative spending.

To address this contradiction and put learning back at the top of budget priorities, all members of a campus community (including administrators and board members) must have better information and greater transparency. I have been working to provide both by creating and publishing data dashboards for colleges and universities. Dashboards contain detailed data about institutions’ revenue, expenses, enrollment, debt, staffing, and more. They allow anyone to track and compare trends in their institution’s key financial and performance indicators over time and relative to peer institutions.

Dashboards offer an ongoing opportunity to undermine false narratives and improve student outcomes by providing clear information about revenue streams and budget priorities. Armed with this information, campus activists can expose misaligned budget allocations, reshape budget priorities, and challenge austerity measures.

I want to give readers the confidence to use accessible dashboards like the ones that appear below as powerful tools for shared governance. First, I describe the information contained in the dashboards and highlight key financial indicators that enable monitoring over time and relative to peer institutions. Second, I present three case studies of institutions where dashboards are helping community members resist bad policies and practices. Finally, I urge you to use data and similar dashboards to inform colleagues at your institution, counter false financial narratives, and promote positive changes.

Information in the Benchmark Dashboards

To date, I have created three public dashboards using data from the US Department of Education through the Integrated Postsecondary Educational Data System (IPEDS), IRS Form 990 tax statements, and audited financial statements:

1.     Mills College Benchmark Dashboard

2.     Saint Xavier University Benchmark Dashboard

3.     University of Tulsa Benchmark Dashboard

One key insight users gain from the dashboards is an objective and rigorous identification of peer institutions. The dashboards identify peer institutions using a “K nearest neighbor” (KNN) algorithm. This KNN algorithm identifies peers based on similarity across many variables, including operating revenue, revenue per student, enrollments by student type, Carnegie Classification, on-campus housing, Carnegie instructional classification, faculty-to-student ratio, admissions selectivity, student characteristics (gender, percentage qualifying for Pell Grants or student loans), and location.

Beyond the benefit of peer-institution comparison, the dashboards contain many key performance indicators (KPIs) organized on eleven web pages. I use links to the Mills College dashboard to provide readers with an example for each:

1.     Peer Institutions—results of the KNN machine learning algorithm

2.     Key Financials and Ratios—net income, net income margins, net assets, spending ratios

3.     Endowment Benchmarking—endowment value, returns, payout for operations

4.     Debt Benchmarking—total debt, interest expense, debt to revenue, debt to net assets

5.     Expenditure Benchmarking—functional expense allocations

6.     Revenue Benchmarking—operating revenues: tuition and fees, auxiliary enterprises, other

7.     Staff Benchmarking—noninstructional staff sizes and salaries by job category

8.     Faculty Benchmarking—faculty sizes and salaries by rank 

9.     Contractor Benchmarking—number of contractors, estimated expenses, food-service contracts

10.  Net Price Benchmarking—sticker price, net price after institutional aid by family income

11.  Enrollment Benchmarking—applications, admissions, students enrolled, retention rates

Dashboards versus Static Documents

Dashboards have crucial advantages over static graphs or tables, including their easy availability. Whereas PDFs and Excel files may be difficult to locate through web links or saved files, the dashboards I create are quickly accessible on any device in the middle of a discussion about policies and practices.

A second, and perhaps more important, advantage is that dashboards enable user interactivity, which helps generate acceptance among community members through their more transparent analysis. Users can modify the comparison group, see all data points in a variety of views, and switch between aggregated or disaggregated comparisons. Interactivity also allows users to perform validity checks on the fly, providing timely information.

Dashboard Applications

A dashboard provides at least three tangible benefits: it informs discussions of institutional policy and practice, it supports data-driven decision-making, and it promotes transparency.

As François Furstenberg and Naveeda Khan note in their fall 2022 Academe article on budget activism at Johns Hopkins University, dramatic changes in KPIs over time or differences relative to peer institutions naturally generate focused questions among campus community members, who are likely to ask, Why are we different? What happened to cause this change? Using dashboards sharpens the conversation by encouraging people to ask the right questions and seek answers based on interactive and wide-ranging data rather than ideological beliefs.

By encouraging data usage, dashboards discourage reliance on rhetoric, intuition, or anecdotes, which are often unsubstantiated. Dashboards allow quick access to information widely available to everyone with a stake in the conversation, forestalling the dangerous practice of relying on narrative abstractions when making budget decisions.

The transparency of a public dashboard (or at least one that is available to campus community members) makes potential problems clear for everyone. Kelly Grotke asserts in her fall 2022 Academe article on coalition-building at Oberlin College the necessity of transparency in maintaining shared governance and promoting positive morale among faculty and staff. Expanded transparency also makes it more difficult for administrators to rationalize decisions with narratives and jargon that could erode the institution’s ability to build what Grotke calls a “culture of trust and cooperation.”

Three cases provide models for how to use the dashboards to undermine false narratives, to counter ill-advised policies, and to expose suboptimal budget allocations.

Case Study 1—The University of Tulsa

In September 2016, TU faculty and staff received an email message from the president outlining several cost-cutting measures. The cuts included a temporary suspension of retirement contributions, a cancellation of university-paid long-term disability insurance coverage, the elimination of forty-three staff positions, and a hiring freeze (see appendix A).

The president justified these cuts by indicating that net tuition revenue had recently declined and that the administration anticipated future enrollment declines. As Jill Penn explains in her fall 2022 Academe article on austerity cuts in the University System of Georgia, limited or inaccurate demographic information has become a key part of such narratives of scarcity. The letter we received from the president stated, “We are not immune to the challenges of today’s economy. We continue to compete for a shrinking number of highly qualified students. In addition, lingering effects of the 2009 economic downturn and a slumping energy sector are putting greater pressure on U.S. family income. As a result, we are seeing increased financial need in a much greater percentage of student applicants.”

This seemed like a reasonable explanation. It was not until three years later that I discovered that the justification for the cuts was probably not true. Net tuition at TU had not declined leading up to the September 2016 announcement. In fact, just prior to the email, TU had completed a fiscal year in which it had earned a near record $89 million in net tuition and fee revenue. During the 2017 fiscal year of the announced cuts, which ended nine months after the email, TU nearly matched that same net revenue at $88 million (see figure 1, panel B).

The story for net tuition revenue per student is nearly the same. In fact, nine months after the announced cut, TU matched its all-time high net tuition revenue per student (see figure 1, panel A).

So why did the administration blame declining tuition revenue for the cuts? Perhaps budget forecasts had predicted the $9 million drop in tuition revenue that eventually did occur during fiscal year (FY) 2018—nearly two years after the announced cuts. That is possible, but it begs the question of why, if the administration had predicted the revenue decline, did it dramatically increase operating expenditures during FY 2015 and FY 2016 (see figure 2, panel A).

During these two fiscal years prior to the announced cuts, the university’s annual expenses increased by $19.3 million (9.4 percent). Most likely, the actual reason TU needed to make the announced cuts was related to this large spending increase—not because a loss in net tuition revenue, which did not happen until FY 2018.

In 2019, a new administration rolled out the True Commitment plan to stabilize TU’s finances through cuts to instructional spending, including the elimination of 40 percent of academic programs and the reorganization of most surviving academic departments into multidisciplinary divisions. This plan was based on the false assumption that TU needed to cut spending on academic programs when in reality instructional spending was in line with that of the university’s peers.  Its administrative spending was not.

In 2019, TU was spending over $60 million on administrative functions—that is, institutional support or upper administration and academic support or middle administration such as deans’ offices, information technology (IT), and advising—while the average peer school spent $37.5 million. Although TU’s annual administrative spending was $22.5 million higher than that of its peers, the administration rolled out a plan to instead cut instructional spending (see figure 3, panel A).

The goal of the dashboards is obviously not to trigger abrupt layoffs, like those done in the name of austerity, but to realign budget allocation around instruction, learning, and student outcomes. In 2019, TU employed 861 full-time noninstructional staff—including those for business, financial, and IT operations—compared with an average of 561 at same-size peers (see figure 3, panel B), resulting in about $20 million in excess annual payroll costs. TU could have closed its structural budget deficit of $12.5 million by eliminating about half of this excess non-instructional staffing cost.    

Instead, the new administration intended to focus its cuts on classroom instruction—the functional expense category most closely aligned with the university’s mission, retention, and student outcomes. Predictably, True Commitment generated a bitter struggle between faculty and the administration that resulted in faculty no-confidence votes for both the president and provost. If faculty, staff, and upper administration at TU had been able to rely on information from a dashboard or similar comprehensive analytics tool, perhaps the institution might have taken a different and healthier path.

Case Study 2—Saint Xavier University

Saint Xavier University is an interesting example of an institution that had spending problems related to nonacademic staffing and excessive contractor use prior to 2017 (see figure 4). SXU’s problem now is that it can’t seem to stop cutting its spending even though it has generated excessive net operating-income surpluses (profits) for the last five consecutive years (see figure 5).

During FY 2021, SXU generated $10.5 million in net operating income (that is, profit) on $73.9 million in operating revenue—a net-income margin (profit margin) of over 14 percent. Prior to 2021, SXU generated net operating-income margins of 5 percent in 2017, 2018, 2019, and 8 percent in 2020 (see figure 5, panel A). SXU’s peer schools target an operating margin of about 3 percent (2.8 percent is the six-year peer average) (see figure 5, panel A). Despite achieving operating-income margins (profit margins) that are more than double its peers, SXU’s administration has continued to push a narrative of financial crisis and continues to operate accordingly.

In May 2020, SXU president Laurie Joyner and board chair Patricia Morris notified SXU faculty that they would no longer recognize the faculty union as a collective bargaining unit (see appendix B). In that letter, they indicated that SXU “must prepare for a 10%–20% decrease in associated revenue, which translates to an anticipated budget gap of $4 million–$9 million for the next fiscal year.” The SXU president and board chair also emphasized the need to create a “nimble operation” to meet a coming decline in the number of college-aged students that they expect to have a negative impact on SXU’s enrollment (see appendix B).

Perhaps at some point SXU will experience an enrollment shock or reach a “demographic cliff,” but as of June 2021 it had not. Yet in addition to unilaterally dissolving the faculty union in 2020, SXU imposed severe cuts to instructional spending and seriously gutted its faculty ranks (see figure 6). Since 2015, SXU has slashed instructional spending by about $4.6 million despite experiencing an increase in student enrollment (see figure 6, panel A). SXU’s instructional spending now lags behind its same-size peers by about $4 million.

SXU cut instructional spending by significantly altering its faculty composition. Since 2015, it has lost sixty-four full, associate, and assistant professors (see figure 6, panel B). That is a 42 percent loss of stable, full-time faculty in just five years—but student enrollment has increased over this period. To replace the lost faculty, SXU has significantly increased its reliance on contract faculty and now employs significantly more contract faculty and significantly fewer full-time ranked faculty than its same-size peers (see figure 6, panels B and C).

The SXU president and board chair declared in 2020 the need for the university to become nimble, but it appears that it had already been acting quite nimbly over the past several years. In FY 2017, SXU cut its operating expenses by $8.5 million (11.6 percent). Between 2016 and 2020, it cut $13.5 million overall out of its operating expenses—over 18 percent—in anticipation of future revenue losses that have yet to materialize.

Why, one might ask the SXU administration, does the university not conduct better demographic research or develop a plan to make drastic spending cuts only if and when enrollment losses actually occur? The administration’s current practices, particularly the excessive cuts to instructional spending and reliance on contract faculty, risk unnecessarily causing an enrollment decline. In that sense, the current SXU policies might be a self-fulfilling prophecy.

In summer 2021, I created the SXU Benchmark Dashboard as a gift to the SXU faculty after learning of their no-confidence vote against President Joyner. They are using the dashboard to push back against continued budget austerity in spite of consistently posting outsized annual net operating incomes (profits).

Case Study 3—Mills College

In March 2021, Mills College president Beth Hillman announced that the college would cease operating as a degree-granting institution (see appendix C). The college has since merged with Northeastern University. She justified the move by stating, “Today, because of the economic burdens of the COVID-19 pandemic, structural changes across higher education, and Mills’ declining enrollment and budget deficits, Mills must begin to shift away from being a degree-granting college and toward becoming a Mills Institute that can sustain Mills’ mission.”  Not long after this announcement, Mills College faculty voted no confidence in the Hillman administration.

Mills had been struggling financially for several years, but the financial situation had nothing to do with COVID-19 or structural changes across higher education. It was a symptom of two problems: Mills suffered from administrative overspending (see figure 8), and it faced a sudden 50 percent decline in applications in 2014–15 that the college did not correct (see figure 7, panel A).

The application decline resulted in smaller incoming classes from 2015 through 2020, with full-time equivalent undergraduate enrollment falling as each successive graduating class was replaced by a smaller incoming class. Mills lost 26 percent of its undergraduate enrollment by 2020 (see figure 7, panel B), having prevented a full 50 percent loss by adjusting its admission criteria and experiencing a higher yield rate for admitted students.

On top of this enrollment problem, Mills had to address an administrative spending problem. In 2021, the year President Hillman announced the plan to close, Mills’s administrative spending (for institutional and academic support) was about $13.5 million higher than its same-size peer institutions (see figure 8, panel B).

Where was the spending? Like TU and SXU, Mills had more noninstructional staff and, like SXU, excessive independent contractor expenses. In 2020, Mills employed twenty-six independent contractors, each paid at least $100,000, versus an average of fourteen among same-size peers (see figure 9, panel A). This accounted for about $6 million in excess annual spending.

Another $5 million in excess spending came from excess noninstructional staffing. In 2020, Mills employed 222 full-time noninstructional staff, compared with just 162 employed at the average same-size peer institution (see figure 9, panel B).

Rightsizing these expenses would have more than corrected Mills’s operating income deficit, which was around $8 million at its worst (see figure 8, panel A). If the Mills administration, board, faculty, and academic staff had had the Mills Benchmark Dashboard or similar analytics in 2016, they might have addressed their enrollment and spending concerns then, and perhaps Mills College would still exist today as an independent women’s college.

Conclusion

These cases share common themes, including a collective lack of understanding of the problems the institution faces. There is also a tendency among administrators in times of crisis (or perceived crisis) to cut instructional spending first, ignore excess administrative spending, or even increase it—as was the case at Mills. Our knowledge of the relationship between student outcomes and institutional inputs indicates that this is the opposite of the optimal response.

In each case, a dashboard or similar analysis could have helped campus community members, particularly upper administrators and the board, make better-informed decisions. These dashboards are currently helping faculty, staff, students, and alumni at the three institutions roll back prior poor decisions made by their administrators. There are hundreds of other US colleges and universities that could benefit from developing dashboards for financial analysis and then taking action.

Benchmark dashboards allow quick access to accurate and detailed information. It is critical that administrators and board members—as well as faculty members participating in shared governance—use this information to sharpen their policy discussions, decisions, and implementation. The pervasive reliance on anecdote and notions of business intuition is risky because these are often incorrect and usually do not align with the educational mission.

It has become part of my scholarly agenda to create as many of these dashboards as possible. They are crucial to shared governance, fiscal responsibility, and refocusing higher education on its educational mission.

Matthew D. Hendricks was associate professor and chair in the Department of Economics at the University of Tulsa. He researches education finance and policy. His email address is [email protected]. 

Appendices

APPENDIX A

Letter from University of Tulsa President Steadman Upham to University of Tulsa faculty and staff

September 20, 2016

Colleagues,

During the last several years, The University of Tulsa has maintained a steady course and a secure top-100 national ranking, but we are not immune to the challenges of today’s economy. We continue to compete for a shrinking number of highly qualified students. In addition, lingering effects of the 2009 economic downturn and a slumping energy sector are putting greater pressure on U.S. family income. As a result, we are seeing increased financial need in a much greater percentage of student applicants. This need, in turn, is placing more demands on the university’s financial aid.

These budgetary challenges require modifications to the university’s expenditures this year and going forward. TU is not alone in undertaking such measures. Many universities, public and private, are dealing with financial stringency caused by the very same factors affecting TU. In fact, U.S. higher education itself is changing, and it appears that most of the trends we are seeing now will become a permanent part of our operating environment.

Since mid-summer, the senior administrative staff and the deans have been working closely with members of the Board of Trustees to analyze and evaluate the university’s operating expenditures. Longer term planning has also begun focusing on the university’s organizational structure, policies and procedures, strategic directions, and ways to enhance revenue growth. We are guided in this exercise by a number of principles that will among other things: 1) protect the quality and strength of the academic core; 2) protect the quality and richness of student life; and 3) achieve sustainable long-term efficiencies across the enterprise.

Immediate Expenditure Reductions

With guidance and direction from the Board of Trustees, the university is taking the following immediate actions to reduce expenditures:
1. Workforce reduction—We will trim the size of our workforce by 5%, a reduction that will result in the elimination of 43 non-faculty positions. An additional 19 vacant positions will not be filled. The hiring freeze for all open, non-endowed university positions will remain in effect.
2. Voluntary Salary Reductions—Voluntary salary reductions will be implemented for TU’s highest paid employees. These voluntary salary reductions range from 6% to 20% and involve all senior members of my administration
3. Benefit Change—Effective October 1, 2016, the university will temporarily suspend contributions to retirement for all employees. However, employees may continue to make contributions on their own to the plan. A restructured matching contribution retirement plan will be announced at a later date.
4. Benefit Change—Effective January 1, 2017, TU will no longer pay for long-term disability insurance coverage. Such coverage will still be available to employees, but at their own expense.
5. Operational Efficiencies—We will implement several changes to current practices that will result in greater operational efficiency and cost savings. Areas being examined include centralized purchasing, tighter regulation of environmental controls in buildings, use of alternative energy sources, and delivery of services to off-campus sites.
6. Other Cost Savings—There are a number of additional non-personnel cost saving measures specific to various departments and divisions that will be communicated by vice presidents and deans to department chairs and supervisors.

As we have done in the past, we are holding to a model of lean operations, which means deliberate and conservative use of resources. Research grants and contracts and earnings from the university’s permanent endowment of nearly $1.0 billion are not affected.

Longer term strategic review of the organizational structure

These decisions are never easy but taking action now will put us in a position to examine areas of our enterprise carefully and strategically for investment or further expenditure modifications. Priority will be given to strategies for enrollment growth, increased operational efficiencies, and revenue generation. There will be no shortage of engagement as we solicit strategies and ideas from across campus, a process that will be led by President-Designate Gerry Clancy. More information will be forthcoming on this process in the coming weeks. Thank you for your patience as we move forward together.

—Stead

APPENDIX B