Few people in higher education know their costs at the course level, insists Maria A. Anguiano, vice chancellor, office of planning and budget, University of California, Riverside. She’s hoping to change that at UCR with the use of software that allows the integration of various data systems, including human resource, core, financial, and asset management.
“By putting all our data sources in one software platform and creating different assumptions around how we want to allocate costs, we’ve been able to basically understand both our revenue and costs at the course level,” she says. “It produces more granular information than could be achieved with using just financial data.”
Anguiano estimates the data will be rolled out to academic leaders in the winter and spring quarters. Only the central office currently has access to the stockpile of information.
Tools for Academic Leaders
UCR’s activity-based costing project, which has been in the works for more than a year and a half, is designed to be a decision-making tool for academic leaders. “Increasing productivity or reducing costs is not what this is about,” Anguiano emphasizes. “This is about giving academic departments enhanced information to allow them to plan better for future growth and to help departments and deans think about how they put together their portfolio of courses.”
A good example of this type of work is the Looking Under the Hood Institutional Aid Metrics project, conducted by NACUBO and the Association of Governing Boards of Universities and Colleges. The joint study, which gathered data on recent aid allocation trends, by institution type and other characteristics, resulted in a Web-based tool that allows colleges and universities to compare their aid allocation and distribution practices with national averages and peer group results. The goal, similar to that of UCR’s initiative: to provide tools that lead to more effective decision making.
One of the five priorities of NACUBO’s new strategic plan (see sidebar, “Data Analytics Is a NACUBO Strategic Priority”) is to lead higher education’s integration of analytics to achieve institutions’ strategic goals. That includes facilitating enhanced quality, reliability, and consistency of data to improve comparability across higher education, which actually starts at the individual institution level with work such as that being done at UCR.
Pushbacks and Positives
“Some people have voiced concerns about a reduction in quality, but this is like any other data,” says Anguiano. “I believe the more data we have, the better decisions we can make. Ultimately, it’s still up to the academic leaders to make the decisions about quality and how we’re spending our resources to produce quality programs. That’s an important piece that is sometimes forgotten when you look at costing data.”
Right now, having insufficient data could be resulting in suboptimal choices, she says, citing the commonly held belief that institutions can’t afford to enroll more students as biology majors—or other STEM majors—because their education costs so much more than other majors. However, when looking at the overall cost of these majors, Anguiano indicates that the assumption is not necessarily true.
“Yes, the biology department is expensive, but, because we have this cost-of-course per student, we were able to do an analysis that added up all the classes a biology major takes over a four-year period. When you add all the components, a biology major is not [that] much more expensive than any other major.”
A major that does carry a high price tag, she says, is English. “We think it’s important for upper-division English classes to be small; therefore, it’s an expensive major,” she explains. “If the dean wants more English majors, she might need to figure out how to balance her portfolio so that it doesn’t affect the rest of the college.”
According to Anguiano, cost is just one factor in understanding programming. After analyzing the full cost—and quality—of classes, some deans and departments may decide that they need to invest more in a particular class.
“Part of this is being able to give data to academic leaders so that they can look at the quality and say, “OK, is it worth the cost we are spending?” or “Are we spending too little?” If it’s a cheap course and 50 percent are failing, we might want to invest more so that more students succeed. With this information, we have a way to do a cost/benefit analysis.”
Many institutions charge one tuition amount to all students, yet offer a multitude of programs. Thus, by definition, some courses are subsidizing others. This is fine, she says, as long as there is transparency.
“If we’re making decisions about which programs to grow or how to make our current programs affordable, we’ve got to understand the components to make sure that the whole balances,” Anguiano says. “We have hundreds of programs. They all have different costs, yet everyone pays the same tuition. The costs have been hidden—not on purpose, but because they haven’t been known. This type of information allows us to understand what those cross-subsidies are and have real conversations about the choices we have made. Most importantly, it provides transparency to the entire institution and to the public.”
Where We Need to Be
Another institution that is exploring activity-based costing, thanks to a $300,000 grant from the Bill & Melinda Gates Foundation, is Johnson County Community College in Overland Park, Kan. “We have begun the journey of using an activity-based costing tool,” says Barbara A. Larson, executive vice president, finance and administrative services.
“We have run one year of data through this tool, which enables us to take information from disparate data systems, such as Banner, as well as facilities and human resources data and student data, and to really look at costs in a more comprehensive way,” she says.
With limited licenses to the software, access to data so far is primarily restricted to financial and institutional research staff preparing reports and sharing those with academic and other units. “We haven’t dispersed the data widely yet, but we plan to do that,” Larson says. “We have an academic program review process, and we’re just starting up an administrative review process. Our plan is that in the future, the revenue and cost data that the various programs receive will be coming from this model, as opposed to our general ledger. This will be a new way of looking at revenue and costs.”
Previously, academic departments could see only actual costs that came straight out of the general ledger and not the overhead, space, and the cost of administrative areas that support them. “Activity-based costing will allow us to allocate resources to those activities that are going to have the greatest benefit for our students’ success,” Larson says.
By providing financial resources to an academic program with a very low enrollment, an institution may be unable to meet the demands of 30 or 40 students for a different program. “That’s an opportunity missed,” she says. “We want to ensure that we are directing dollars to those activities that will drive our students’ success.”
The data from an institution’s enterprise resource planning (ERP) system provide hard numbers on how much, for instance, an institution pays for faculty salaries, utilities, or custodial services, she says, but it doesn’t generally combine those statistics in a meaningful way.
“What this tool is doing is ascribing overhead costs,” Larson says. “For example, having brick-and-mortar facilities is a necessary overhead cost of teaching and learning. When you then take scheduling data, you have a real sense of, ‘Well, this is what it costs to have this room scheduled for this period of time to meet this group of students’ needs.’”
She hopes that the data will eventually encourage in-depth conversations about the facilities required to teach various classes. “We need to really look at facilities in a different way in terms of how classes are scheduled and to be as accountable as possible with the dollars we are given by taxpayers and students,” Larson says.
“We are an innovative community college, and one of our largest areas of growth is online delivery,” she continues. “We want to understand the cost avoidance of not scheduling a classroom, as well as to be cognizant of the cost of supporting an online experience. Those are discussions we will have more and more. We need to be thoughtful about building a building, for instance, to make sure we don’t overbuild—because education is changing.”
Larson predicts that one outcome of this activity-based costing model will be using the institution’s square footage more effectively. “We have never looked at our operation through this lens before, because we have never had this data presented in this way before.”
The use of data analytics in all parts of higher education, not just for class management, is inevitable, she says. “I’ve seen here and in my career at other institutions that there is more comfort with evidence-based decision making,” Larson says. “We all recognize that this is where we need to be, both to be accountable to ourselves, our students, and our external constituencies. The data can be very powerful. Now, culturally we need to start using this data and helping people understand it.”
Improving the Commute
Improving the student experience through the use of analytics is also a priority at Rutgers University, The State University of New Jersey. “One of our president’s goals is to improve the student experience by having students spend less time on a bus schlepping around our campuses,” says Michael Gower, executive vice president for finance and administration, Rutgers, New Brunswick, N.J.
To that end, the institution has purchased an analytic product that examines how classes are being scheduled and assigned. “That product ties in with a classroom management tool that we are building [ourselves] because we could not find something acceptable on the market,” he explains. “The idea there is factoring in not just demand, but things like where students are coming from.”
According to Gower, with five different campus areas that surround the city of New Brunswick, Rutgers operates one of the largest bus systems of any higher education system in the country. “We have a very distributed campus,” he says. “That’s not necessarily a good thing, because we are making students go from point A to a point B that is way across town. The idea is to find ways to schedule classes closer to where students are coming from or going to, especially at the beginning or end of the day.”
For example, the institution recently analyzed where students start their days and where they go to classes, creating a scatter map of the entire campus, with all of the lines based on the students’ class schedules. “It was extraordinary to see and to be able to answer questions, such as ‘How often did they have to get on a bus per day? Were they going from one campus to another? What was the path?’ All this comes back to the need for analytics and the utilization of space,” Gower says.
With the emphasis on improving the student experience, the data will be used to set up classes and assign them to rooms. If Gower has to choose between convenience for faculty and staff or convenience for students, the latter wins. “All of this ultimately ties into dollars and cents as we look at what is the appropriate size, location, and number of sections for individual courses.”
Early changes implemented in the fall have already resulted in 3 percent less travel for students, Gower says. “We expect to get that number up to double digits through better utilization of data as we go along.”
Piggyback on Success
Whether using data analytics to recruit students, benchmark productivity, optimize space, improve transportation, or conquer other higher education challenges, Gower thinks it’s a mistake to start from scratch. “We’re best when we learn from those who have already made mistakes.”
Gower encourages business officers to share their missteps and successes so that others may learn from them. “If the University of Arizona has done something really well, I want to copy it,” he says. “And if I can develop a best practice in a new area, I want to share it so, perhaps, it can become a standard.”
Other advice from institution leaders includes:
Build alignments between information technology and institutional research or other analytical units, says Joel Hartman, vice president for information, technologies, and resources, and chief information officer, University of Central Florida, Orlando.
“All the players involved with the tools, data, interpretation, flow, security, and reporting need to operate from the same set of principles and assumptions,” he says. “Together, they can support each other. One single office cannot do all of this on its own. It takes multiple functions around the institution.”
At UCF, the institutional research function manages the data warehouse and tools. “We, at IT, partner with them in terms of providing the data, validating the data, and setting up the technical resources and software to support their tools. As an IT person, I’ve frankly been surprised at the number of institutions that talk about the difficulty of institutional research and IT people working effectively together.”
Help everyone understand that analytics is part of their jobs. “There are folks who are hungry, receptive, and naturally curious, and need data to do their jobs, such as admissions,” says Keith W. McIntosh, vice president and chief information officer, University of Richmond, Va. “Others are slower to come to the table. They don’t have that sense of curiosity and may not know about or understand analytics.
“When you bring up analytics, they might say something like, ‘Well, that’s great, but I don’t know when I’ll have time to work on that,’” he continues. “They get their information the way they’ve been doing it for years. They see analytics as something separate and in addition to what they do as a daily job. People need to understand that data analytics should be part of what they do on a daily basis. Having access to more information will make them better at what they do.”
Consider data an asset. “To use your institution’s data effectively, even at the level of day-to-day transactions, let alone for analysis, you need to think about it as a holistic resource and asset,” Hartman emphasizes. “The major sources of data are our business systems, which are set up primarily to be transaction systems.
“From an institutional point of view,” he continues, “all the offices that use those systems and provide input should understand the data and train staff so that the data they create day by day, year by year, is sufficiently accurate and timely to be used beyond its intended primary function for analytic purposes. I’ve talked to institution after institution that has not yet come to grips with the challenge of putting itself in a position to create high-quality data for analytic purposes. That has to come first.”
Make routine reports easy to access. While big data products may contain the secrets of the universe, obtaining those secrets from the morass of information contained in them can be a challenge.
At the University of California, Riverside, Anguiano plans to develop for department chairs draft reports with basic information about direct class costs, different class sizes, and so forth. Then, over a period of several months, she will refine the reports based on the needs and specifications of department chairs. “One of the key ingredients of any implementation is collaboration,” she says. “If the report is not useful for the end-users, they won’t use it. We need to involve them in the creation of the reports. We have data and will work with department chairs to create reports that are useful to them. Only when everyone has the same information can we have good discussions. ”
Through the years, the University of Central Florida has built a series of data marts that provide commonly needed data in a format that integrates with desktop tools to meet the routine reporting, inquiry, and monitoring needs of colleges and departments. Through data marts, staff at colleges and departments insert variables, and run reports on their data, such as HR, finance, and student data.
According to Hartman, the data from the ERP system is accurate up to the second, while information from the data mart is based on the previous night. “The data warehouse is certified census-based data as we use it for periodic reporting and analysis,” he says. “We also bring in data from other sources as needed to do more complex analyses, comparisons, and benchmarks.
Use consistent data definitions. “We not only need to have consistent data definitions across our universities, we need to have them consistent among universities,” Gower emphasizes. “All the various benchmarking initiatives that we and others are involved in get very difficult because there are no standards. We feel that this needs to happen across our industry so that we can talk the same language. The key to analytics is the structure and rigor of the data definitions and data governance.”
Gower is encouraging NACUBO to include in a future strategic plan a data dictionary or mechanisms that would promote commonality among institutions.
Build a solid pyramid. Analytics capability can be divided into three tiers of a pyramid, Hartman explains. “The base of the pyramid is the data that you use to run the institution day by day,” he says. “That has to be solid. The next level is what you do internally to analyze and use that data beyond the transaction level, which involves partnerships. The third tier is the outside analytics vendors or in-house system you build.”
McIntosh agrees, using the same analogy. “You can’t jump from the bottom to the top without going through the pyramid,” he explains. “Depending on where your institution is, you may want a dashboard, but there’s nothing in the data warehouse to pull from. Maybe the data aren’t there yet.” His recommendation in this situation: “Start where you are and build on it, filling in holes with each iteration of data. That’s how you can demonstrate how analytics can help with answering questions. Build on your success.”
Be transparent. “You have to be transparent about the data you are collecting and why you are collecting it, because there can be fear and concern by the constituent group being analyzed,” McIntosh says. He suggests a phrase such as, “We’re just trying to do what’s best for the institution by maximizing our efficiency and delivering outcomes that are better for our students,” is a good place to start.
MARGO VANOVER PORTER, Locust Grove, Va., covers higher education business issues for Business Officer.