In the past, LaGuardia Community College, Long Island City, New York—like many of the colleges and universities for whom I have worked and consulted—had not set a high priority on obtaining program or project cost information, or many other kinds of information for that matter. One excellent exception was a system to collect cost information from several different expense systems on several different computers, including the CUNY Research Foundation system. The goal: to create an expense analysis by functional area within our large (more than 70,000 educational experiences each year) and diverse Division of Adult and Continuing Education. In the past 10 years, the college has taken very seriously the goal of basing decisions on information and has expanded its institutional research and assessment department accordingly.
As the demand for better cost information increased, LaGuardia’s business office obtained supplemental reporting software to "tack onto" the legacy (and inflexible) accounting software maintained by the City University of New York (CUNY) system. The business office personnel got very good at recoding and programming to create new reports requested by users. Then the university began a major project to replace all the old systems.
While the primary goal of institutional research and assessment has been to improve outcomes reporting and analysis—especially the assessment of our retention efforts—we have begun working more and more with the business office on projects that involve cost analyses. When senior management told us in May 2007 we would be getting a new integrated information system as soon as all modules of the new enterprise resource planning (ERP) system were implemented across several years, our five-person research and assessment team had dreams of being able to produce all kinds of informative cost reports, especially with a former business officer heading the team.
Our first goal was to calculate the cost of a degree versus the cost of a dropout. We then hoped to be able to produce reports on the cost per student retained in each of our many retention projects, but we never seemed to have enough time to dig through the expense data kept in several separate accounting systems.
Beginning in fall 2012, the integrated finance system and student modules were both up and running; and we began to review the kinds of reports and analyses we wanted to produce. We soon found that we still couldn’t create reports that helped us evaluate costs and retention efforts. Although we had integrated the correct type of systems, we learned that the transition hadn’t positioned us for our analytical needs.
Now, we want to start over. Others may learn something from our experience.
Think of the Possibilities
These days we find we need better cost information to inform evidence-based decisions. However, the thought of enhanced cost analysis scares many. For example, faculty fear that cost data will cause decision makers to ignore information on the value of programs that can’t easily be quantified. These trepidations can keep us from trying. The irony is that the real goal of analysis is to push out the least effective efforts, not necessarily the most costly. Consequently, even if it’s politically difficult to discover what doesn’t work, institutions that don’t try will never have an opportunity to cut losses and change business models.
Business officers and accreditation associations agree that we must be able to gather and present effective cost information so that data-informed decisions can be made. Recent foundation grants that have funded projects at LaGuardia have demanded to know how we will "transition" our grants to the regular budget. Our response should be that we have analyzed the costs and benefits of the projects and will have no trouble justifying the replacement of several less-effective programs—but we just don’t have the data.
Inadequate cost information is not the only stumbling block to making such decisions. We have great difficulty even counting dropouts, for example, and have failed at keeping track of the costs invested in students who transfer among institutions. The dream we had, and that is increasingly shared by other managers at LaGuardia—especially those involved in our entrepreneurial, continuing education area and those working on funded projects—was to have regular access to better cost reports that would provide data on which to base many of the decisions and leadership challenges we face in higher education.
If we turned this into something of a wish list, we’d want to create a cost reporting system that could support decision making in the following areas:
Strategic planning. We’ve realized, through several years of relatively hard times, that strategic planning should be all about managing change in our colleges and universities. Too often, such efforts devolve into annual rituals of discussion and a circuitous barrage of words with little focus on anything remotely strategic. However, with accurate cost information at hand, leaders can concretely evaluate data that reveal the truth about our change efforts and help focus the campus community on further improving the outcomes of such activities.
Ongoing review of priorities. Although we tend to cost out strategic initiatives only when an accreditation report is due, this type of information can give us a more continuous focus on priorities. The difference between a strategic slogan and a real institution-changing initiative is often just money. Do we know how much has been spent on strategic initiatives such as "improving undergraduate education" or "internationalizing the campus"?
A cost figure for change efforts. For actions aimed at changing outcomes, accurate data allow us to establish the payoff of a particular effort. Here are a few examples:
- Student attrition, retention, and degree completion. How many institutions have been able to put cost figures on their efforts to reduce the number of students who drop out? What has been the payoff of these efforts? Is the institution willing to continue spending money on retention efforts once the full cost and benefit are understood?
Most of us have not measured very well the costs of such programs against the gains in outcomes. And, although we may have many fine retention programs, we rarely hear the people managing these programs reveal how many dropouts they’ve prevented or the cost of the program per dropout prevented. Often, that is because we do not have systems that can give us that information.
Our dream was to use the cost per degree and cost per dropout to focus attention on the financial need to improve retention. Certainly the humanitarian need to improve retention exists, but there is nothing like real dollars to focus the mind.
We invest a great deal in every student, often over several years. When a student leaves, either as a graduate, transfer student, or dropout, how much on average have we invested in each "unit of output"? In industry, we would monitor and measure how much of our "production" did not end up as a finished product, and work hard to find the causes for the inefficiencies and eliminate such production flaws. With better information, we might be able to calculate the real costs of our outputs in higher education.
- Establishing tuition amounts. Many institutions could be using the cost of a course to assist in tuition setting. Strong cost analysis and savvy graphical displays might help to shift arguments from simple "for and against" tuition increases toward an understanding of the basis of cost pressures. This type of presentation is necessary for raising the dialogue to higher-level arguments over the proper resolution of the trade-offs among raising tuition and improving salaries, increasing faculty numbers and offsetting dwindling state aid.
- Investment opportunities. Another item on our wish list was the ability to calculate departmental revenues and costs to find opportunities for investment. For example, back in the 1990s, when I worked with Teachers College, Columbia University, we took the tuition revenue and allocated half of it to academic departments according to teaching hours and the other half according to majors, dissecting detailed finance data manually and building spreadsheets.
We started by allocating departmentally restricted revenue to departments, but made no other allocations of nondepartmental revenues or expenses. Then, using a ratio of departmental revenues to expenses, we divided the departments into thirds. The top third generated the highest departmental revenues compared to expenses. We considered these departments to be under some stress in terms of resources and looked somewhat more favorably at their requests for more faculty resources. The bottom third was looked at less favorably and those departments had to build their requests on some basis other than resource constraints. Thus, our growing psychology department’s requests were warmly welcomed, while those from departments whose focus had wandered away from enrollments and research revenue were less positively received, at least by the chief business officer.
Similarly, as a consultant, I once helped a university dispel the myths that its business school was highly profitable (it had only a moderate revenue-to-expense ratio) and that its evening division lost money (even though evening courses were discounted, the program was highly profitable). Although pulling the supporting data out of the information system was difficult, having the numbers available kept the university from making what would have been a bad financial decision to close the evening division.
These examples show that we can come up with data to answer particular questions on an ad hoc one-time basis (perhaps hiring a consultant to help), or we can regularly produce the necessary reports.
Challenges and Errors Derail the Dream System
In retrospect, we see four roadblocks to creating an information system that can deliver accurate cost information.
Confusing responsibility with purpose. When coding costs, is instruction considered a responsibility or a purpose/effort? The confusion between the two categories started some time ago with the dual nature of instruction. Some accounting systems allow us to run two different dimensions, but we don’t always think that way. We should be able to imagine two separate hierarchies with very different definitions. The responsibility hierarchy should derive from the standard organizational chart, focusing on responsibility for controlling expense and revenue centers. Budget resources go to people with expense-center budget control responsibilities. These responsibilities roll up in a fairly fixed hierarchy of authority and responsibility.
The purpose/effort hierarchy, on the other hand, can be much more flexible than the responsibility hierarchy and should reveal the efforts that we make to get things done. While the activities that go on in an English classroom might be under the control of the expense center, English department, the effort is instruction. Nevertheless, that effort might also be "improving undergraduate instruction," a strategic plan initiative.
The way we code for the first dimension determines who has responsibility for controlling the expense. The way we code for the second dimension determines which effort gets credit for the expense.
For example, when we purchase an item, we go online and fill out a requisition, which requires a responsibility code. If we have budget responsibility over several areas, we must choose the budget area from which we wish to take the funds. We then must assign a purpose, with each responsibility area possibly allowing a different set of purposes. The chosen field may automatically populate—for example, all English department charges might default to the instruction purpose/effort code—or we may need to choose from several options, especially when there are several strategic purposes that the responsibility area must serve. We can choose "instruction," or we may choose "improving undergraduate education."
Additionally, if we are responsible for several grants, we may also have to designate a source for the funds. Finally, we will need to designate a categorical description of what it is that we are buying, like office supplies.
Differing codes for like items across systems. I once needed 94 queries to derive departmental costs and revenues for one university client. Most of these queries were applications of "crosswalk" tables to convert codes for one thing into departmental codes, relevant for the institution’s reporting needs.
Take a department of English, for example. How is "English" coded for payroll costs, departmental expenses, English majors, English courses, tuition from English majors, restricted gifts to the English department, grants to English faculty, endowment payout restricted to the English department, restricted scholarships to English majors, English teaching labs, and English office space? Very few of these "English" classifications would have a common code. Some might not even have coding at the department level.
In all probability, the payroll people and the facilities people, to use two examples, each designed their own coding structures without regard to the needs of overall cost reporting. You cannot have a truly integrated information system if your codes for the same target are all over the map.
Inflexible hierarchies. Did you ever find that to do one report, say for the auditors, you needed to roll the accounts up one way; but for the trustees, you needed to roll up the reports another way? We have three choices for reporting accurately to different constituents: (1) include a lot of different dimensions for each expense code (like the auditor’s departmental code, the NACUBO Functional Expense Code, and the trustees’ department code); (2) write special reports that can identify and recode all our expense information to fit a new set of categories; or (3) establish alternate roll-up hierarchies that produce a number of higher-level reports with different sets of aggregations. The third alternative is similar to the second, but it would be built into the reporting software, making the roll-ups significantly easier.
Dominance of source. If we have one strategic initiative, for example, "improving undergraduate education," supported by five grants and some unrestricted money, can we have a single revenue and expense report for that effort? The dominance of fund accounting has meant that the head of the project may have to deal with six separate reports. In fact, at my current institution, we have separate corporations and separate computer systems running the restricted and unrestricted funding systems. Preparing a single expense report for a single effort requires a major research project.
Recommendations for Starting Over
Based on what we’ve learned during our information system upgrade, if you say, "Our system can’t do this," I will reply, "Be sure your next one does." And here are some ways to achieve that.
Make certain that your system can handle four dimensions of hierarchical codes: source, responsibility, purpose, and description. Keep revenue and expense authority within the responsibility hierarchy, but allow budget directors to designate portions of their budgets to a number of different purposes. (For example, strategic plan initiative directors would need to request support from responsibility center budget directors.)
The key point in this recommendation is that business officers must develop the purpose hierarchy (or hierarchies) independently of the responsibility hierarchy. It is possible that in 70 percent of the cases, the two areas will coincide: English department expenses are almost always coded as instruction, for example. Nevertheless, by allowing budget officers flexibility in designating purpose/effort codes, we can gain much better information on strategic initiatives and other special projects.
Develop a usable purpose hierarchy, but allow alternate roll-ups for different reporting needs. We need to be able to code once, but report many different ways. We don’t ever want a system where we have to change transaction codes to fit a different report structure. We want simply to roll things up differently to higher levels. All reports are transaction aggregations of some sort. Does "improving undergraduate education" belong under strategic initiatives or instruction? The answer will depend on the purpose of the report. For some purposes we may want to classify it one way and for others another. We should be able to change the structure of our reports simply by changing how detail categories are rolled up to higher categories. We should be able to set different roll-up structures without having to change expense codes.
Code identical things identically, across systems. We could develop special reports much more easily if we created all coding structures under the same coding philosophy. (If the purpose of accounting systems was to prevent us from determining the revenues and costs of an academic department, we’ve succeeded!) Even if there are several different kinds of English majors, for example, the related costs all should roll up to the "master" English department code. Why should the tuition paid by English majors be so difficult to tie back to the expenses charged to the department budget? In achieving an integrated information system, a single database is less important than an overarching coding system.
Allow easy aggregation of financial information across sources as well as responsibility centers. A single responsibility center may be funded from many different sources; and a single purpose/effort may be funded from many different sources, by many different responsibility center budget managers. That being true, I, nevertheless, still want to be able to send out a report on the revenues and expenses of the English department or the physics department, regardless of source and regardless of purpose/effort. I want to be able to show it at a detail level, indicating sources for parallel lines of expenditures and at a higher level, aggregating all sources into a single line. That is, at the detail level I want to show that the administrative assistant’s salary comes from four sources and that one has been completely expended; and, at the general level, I want to be able to show that the same administrative assistant has been paid 75 percent of his or her salary to date.
The same thing is true for purpose/efforts. I want to be able to show where "improving undergraduate education" stands. At a detail level, I need to show each line by funding source and by responsibility center source. If the English department is controlling a grant for the improvement of teaching and the physics department has responsibility for another grant with the same function—and both have allocated funds to this strategic initiative—I want to see both lines of support for each item they have put into their respective grants’ budgets. At a higher level, I will want to combine those lines.
Allow free invention at the bottom level of coding, especially description, but maintain higher levels of roll-up coding. The only reason an institution might have one code for office supplies and one for teaching supplies is to help people think about how much they need to budget in the future. A finance officer rarely needs to know how much we spend on office as opposed to teaching supplies. Sometimes people just don’t know the difference. It’s really hard to tell whether those magic markers are going to be used in meetings or classrooms.
I know also that the auditors ask how much we spend on advertising, federal rules forbid using overhead for certain things, and external stakeholders will always seek to compare functional or programmatic expenses across institutions. Clearly, those are categories that we must keep. The other categories seem to have grown simply to help people think about how they are spending their money and to plan better in the future.
That’s why I suggest that we have as few high-level categories as possible, and then let the budget managers set up whatever subcategories they want or need. As long as we keep the roll-ups clean, we should be able to manage anything.
Some of these recommendations may be beyond the capabilities of current finance systems. Many systems have the flexibility to adapt to these recommendations, but also have the drawback of being too complex to allow easy redesign. However, with careful planning and a close look at lessons learned, these dreams can become part of the design specifications for the next generation of systems.
NATHAN DICKMEYER is director of institutional research and assessment, LaGuardia Community College, Long Island City, New York.