In 1986, more than 900 independent colleges derived at least 75 percent of their revenue from students, according to a 1989 study by Minter and Associates. This figure has probably changed little in the past eight years. Too often, enrollment-dependent independent colleges find that actual enrollments are dramatically different from budget projections. Because financial reserves at these institutions are so meager that small discrepancies between forecasts and fiscal performance can be disastrous, fi­nancial officers must find a way to monitor budget status.

The longer it takes to discover that tuition targets do not match the budget forecast, the more difficult it is to solve the problems. Options disappear during the academic year as discretionary expenditures are allocated. Conversely, the pleasant possibility of extra funds can be dashed if planners have made false assumptions about the size of the an­ticipated surplus. Plans for expected excess tuition revenue must take into account as­sociated costs such as faculty salaries and instructional supplies.

Clearly, an ongoing system of short-term budget controls that test actual performance against budget forecasts is essential for enroll­ment-dependent independent colleges. By definition, these institutions often lack the resources to supplement enrollment revenue with endowment or other funds.

CASE STUDY: WILMINGTON COLLEGE

The experience of Wilmington College, a small, fast-growing independent college in Delaware, demonstrates the difference a well-tailored budget control model can make. The college, which has developed and applied its budget control system over the past eight

years, offers degrees at the bachelor’s, master’s, and doctoral levels. More than 70 percent of Wilmington’s students attend part-time; their average age is 27. Enrollment has grown at a compound rate of 12 percent over the past five years, reaching nearly 3,900 in 1994. The enrollment dependency rate (tu­ition and fees plus bookstore sales, divided by total revenue) was 93 percent in 1994.

Instructional programs are conducted at five sites throughout the state, each of which sets a tuition rate based on its fixed costs. Therefore, the college publishes five rates of tuition for each program. A corps of veteran adjunct faculty members (most of whom work at one of the state’s Fortune 500 corpo­rations) teaches 82 percent of the courses.

The combination of a multiple price struc­ture and a large number of adjunct faculty results in a complex budget. Financial con­trols are imperative. Prior to the introduc­tion of the budget performance tracking model in the mid-1980s, annual deficits were commonplace. Deficits generally occurred because of an unanticipated decline in en­rollments or because direct expenses out-paced enrollments. Even when enrollments exceeded the budget forecast, the college often could not avoid a deficit. The extra revenue only fueled the demand to spend, which is not unusual for an institution that depends on surplus revenues to fund capital improvements and equipment. The extent of the deficit was never discovered until the end of the fiscal year, when it was too late to take any deficit-reduction actions.

The problem Wilmington College faced was how to stop this chain of unforeseen

Michael K. Townsley is senior vice president at Wilmington College.

deficits. The fall semester of 1985 found the college with a smaller student body than expected and a governing board that had become exasperated with the college’s in­ability to control its finances once an enroll­ment trend developed. The finance office was obliged to devise a system for closely monitoring its tuition revenue; the rudi­ments of the control model began to take shape that year.

During the fall, the finance office started to track tuition revenue closely as it was posted. Financial officers soon found that regular budget reports covering only tuition were inadequate, and decided that figures on adjunct faculty expenses also needed to be included. In other words, the original pre­sumption had been that if enrollment fell, the number of classes offered would decline proportionately. But that proved to be false. The number of classes is driven by instruc­tional program demands, as well as enroll­ment. Even with small class enrollments, certain courses must be offered periodically for students to be able to complete their program requirements.

In the spring, the costs of adjunct faculty expenses were added to the tracking pro­gram. Any possible deficit could be estimated by comparing tuition revenue and the cost of additional adjunct sections to the original budget. For the first time, this gave the col­lege sufficient forewarning to initiate a cost-cutting program to eliminate a deficit.

During the summer of 1986, this simple monitoring system was reviewed to deter­mine how it could be revised to improve efficiency. Four flaws were immediately ap­parent:

• Tuition and direct expenses were not accrued.

• Gross income (tuition minus adjunct faculty) was not computed.

• Payroll tax variances were not estimated.

• The budget was not subdivided accord­ing to the academic schedule.

As these defects were remedied, the exist­ing control model took final shape. The first change was to begin accruing tuition rev­enues and adjunct payroll by academic pe­riod. (Over the years, accrual accounting has been expanded to the bookstore and to all revenues and expenditures that had a sign ifi­cant potential for varying with the budget.)

Next, the college began computing gross income, which is important because it is the predominant source for the funds that cover fixed costs. The only refinement to this com­putation has been to add the gross income from each site to give a total gross income.

Another line was added after the gross income, estimating social security taxes based on any variance in the adjunct payroll. This too was summed to give a total payroll tax variance for all sites. Finally, the budget was divided into academic periods so that it could be compared to the accrued balances in the ledger. (An initial breakout by month pro­vided more detail than proved necessary.) The last major revision to the control model, carried out in 1990, translated it into a Lotus program, which quickly compiles the data and produces a summary report.

The budget control model concentrates only on those variable revenues or expenses where unexpected differences from the bud­get may have a major impact on the status of the current unrestricted fund. The reporting procedures used are based on commonly accepted practices for variance reporting. This model accounts for all expenses that vary, whether related to instructional, ad­ministrative or support services.

The model uses five equations, each of which depicts a benchmark in the flow of tuition revenue through direct expenses to net income. (See Exhibit 1.) The equations are computed in sequence at the end of each drop/add period on a year-to-date basis. Each equations’ output is compared to a budget benchmark to ascertain variances.

The equations depend upon a budget that forecasts the variables and benchmarks for each academic period during the budget year. The ledger system also must provide data that clearly depict the financial transactions for the academic period.

Benchmark equation 1 identifies the amount remaining after the adjunct and full-time faculty overload costs are deducted from tuition. The initial step in computing this equation is to translate enrollment figures into tuition revenue, based on data from the registrar’s office. Enrollment is subdivided by site and by any deviation from the posted tuition not due to a scholarship award. Tu­ition revenue is then calculated by simply multiplying the appropriate tuition rate times the enrollment.

The second step in computing the equa­tion is to add the adjunct faculty overload contracts for each site. (The contract should describe the amount paid and any adjust­ments made, such as a prorated payment based on class size.) The resulting figure,

$B 1, is the contribution margin—the expenses, as defined by R.N. Anthony and difference between variable revenues and

D.W. Young in their book Management Control in Nonpro fit Organizations. The final step is to compare the resulting figure to the budget forecast to see if a variance exists.

Benchmark equation 2 deducts direct ex­penses related to tuition or to instructional expenses. Adjustments to tuition revenue include payment defaults, withdrawal re­funds, and scholarships in excess of the bud­get. Adjustments to instructional expenses are instructional support and social security taxes. A payment default is included only if it exceeds the amount set aside in the budget for a student’s outstanding bill. Similarly, scholarships are recognized only when they surpass the budgeted limit. A withdrawal refund denotes a refund made after the drop! add period ends. Prior to that, any adjust­ment to tuition revenue due to withdrawals should have been divulged through the en­rollment report.

Instructional support is usually revised only when enrollments exceed the forecasts, as extra funds may well be needed for in­structor materials, classroom space, computer software, and equipment. Excess social secu­rity taxes are computed by multiplying the variance for adjunct and full-time faculty overloads by the rate for social security taxes. (Some institutions may have to consider other payroll items such as unemployment taxes, workers’ compensation, or institutional ben­efits.)

Benchmark equation 3 takes into account any variances in nontuition revenues, such as money from student fees or from the net income projected for the bookstore. (These items are not intended to be exhaustive. Other institutions may have other revisions to revenue that are related to enrollments, such as dining hail or housing income. For some colleges, nontuition revenue plays a significant role in the financial structure of the institution. If this revenue or its atten­dant direct expenses vary substantially from the budget, the variance needs to be reorga­nized in benchmark equation 4.)

Unanticipated changes to the expense budget for other instructional expenditures are recognized in benchmark equation 4. The original budget often needs to be fine-tuned because it was based on incomplete information. For instance, medical insur­ance rates may be higher that originally estimated, or the rate of inflation may have risen faster than expected, causing utility expenditures to skyrocket. Furthermore, un­expected emergencies, such as equipment breakdowns, accidents, or bad weather, may have befallen the campus. As a result, other budget items must be modified other than those directly related to tuition revenue.

Variances from the first three equations are summed in benchmark equation 4. The total variance resulting from this equation, $B4, equals the estimated net income pro­jected for the budget year. A positive net income signifies that benchmark 5 is to be computed. However, if $B4 reveals a possible deficit, the college devises an expenditure reduction plan. (The nature of the spending cuts will depend on the institutional climate and governance structure.)

Benchmark equation 4 indicates only the direction and the potential scale of the net income at that period in the budget year. It does not signify what will subsequently hap­pen. Data from the control model need to be extrapolated to forecast the balance of the budget year.

The purpose of benchmark equation 5 is the allocation of any excess revenues identi­fied. Of course, any surplus indicated by $B4 may have to cover revenue shortfalls or un­expected expenditures later in the budget year. Thus, it would be prudent to set aside a portion of the surplus until the end of the drop/add period for the last academic period of the budget year. The allocation of the surplus from benchmark equation 4, like the deficit reduction plan, should come from plans made by the appropriate decision mak­

ers in the governing structure. In most in­stances, those plans should be approved by the president.

The results of the benchmark equations can be assembled in a report, to be sent to the president and other key officials. Those re­ceiving the report will depend on the polity of the institution. The reports should be prepared and distributed at the conclusion of each drop/add period in the academic sched­ule. These reports provide data on the finan­cial status at a critical point in the flow of new tuition revenue, when the decisions on direct expenses are still fresh in mind.

SAMPLE CALCULATIONS

The results from each of the benchmark equations can be assembled in a standard budget report, as shown in Exhibit 2. The sections correspond to the five benchmark equations; the rows conform to the variables in the benchmark equation; and the col­umns show amounts for the budget, perfor­mance, and variances, using year-to-date fig­ures. At the end of each benchmark section, rows and columns are summed. Each section after the first carries forward variances from the previous section. A grand variance is given at the end of the report.

ACCOUNTABILITY CRUCIAL TO SUCCESS

The success of this model depends on number crunching as well as combining the model with an administrative mechanism that builds in accountability by defining the performance levels expected for each bench­mark, and by ensuring that an administrator is responsible for the performance at each level of the benchmark. Identifying a re­sponsible administrator means that the in­stitution can turn to someone for a report on what is happening with each benchmark variable. This manager can also suggest a course of action when problems arise.

The budget should define the performance levels expected for each benchmark. Centers should be set up to monitor, control, and manage the major revenue and direct expen­diture streams for an institution. Reports should be directed to center administrators or to decision makers in the governance structure so that plans can be developed in response to positive or negative variances.

Financial controls should focus on those components of the budget transactions where variances may occur and may have a sub­stantial impact on the financial condition of the institution. Specifically, the model should track variable revenues, variable expenses, the contribution margin, and the mix of services. This model concentrates on van­

able revenues, represented by tuition rev­enue. It works with variable expenses in terms of adjunct contracts, instructional sup­plies, payment defaults, withdrawals, schol­arships, taxes, student fees, and bookstore net income. The contribution margin is con­trolled by taking the difference between tu­ition revenue and variable expenses. Finally, the model employs centers as the mecha­nism to oversee the mix of services.

THE BOTTOM LINE: NO DEFICIT IN EIGHT YEARS

The biggest challenge to Wilmington Col­lege since the 1985-86 budget year has been to keep pace with the college’s rapid growth. In 1992 alone, enrollment grew by more than 21 percent, mainly due to the addition of three new instructional programs. Deci­sions during such growth spurts have to be made quickly: classrooms must be con­structed and staff hired. The control model has proved a boon by giving the administra­tion a quick, precise picture of its net income. In most of these years of rapid growth, the surplus income was allocated toward capital expenditures and the hiring of new staff well before the end of the fiscal year. Neverthe­less, the college has not had a deficit in more than eight years, despite these demands on its surplus revenue.

This happy state of affairs has resulted partly because financial risk is lessened by having better financial data disseminated quickly to those who need to know. The model helps to militate against the concen­tration of power in the business office be­cause it depicts the dynamics determining the bottom line. As indicated, the reports produced by the application model are given the widest dissemination within the governance structure.

Finally, as Philip J. Bossert, in his manual for NACUBO, Management Reporting and Accounting for Colleges and Universities, and L.H. Gitman, M.D.Joehnk, and G.E. Pinches, writing on financial management in Managerial Finance, note, variance analysis should be used as a reference for improving the budget and the control model. The model lends itself to this usage in two ways. First, it keeps in the forefront the dynamic condi­tions that drive an enrollment-dependent college. In particular, it is a tool to continu­ously inform management of any changes that are taking place in the all-important tuition revenue stream. Furthermore, working with the model suggests further refinements that will improve its precision. The current model is not the final form; revisions will continue to manifest themselves as the model is put to further use. —