DOCUMENT RESUME

ED 449 719 HE 033 719

AUTHOR Stringer, William L.; Cunningham, Alisa F.

TITLE Cost, Price and Public Policy: Peering into the HigherEducation Black Box. New Agenda Series[TM], Volume 1, Number3

INSTITUTION USA Group, Inc., Indianapolis, IN.; Institute for HigherEducation Policy, Washington, DC.

PUB DATE 1999-08-00 ,

NOTE 68p.; "With Jamie P. Merisotis, Jane V. Wellman, and ColleenT. O'Brien."

PUB TYPE Information Analyses (070)

EDRS PRICE MF01/PC03 Plus Postage.

DESCRIPTORS Educational Administration; Educational Economics;*Educational Finance; *Higher Education; Paying for College;*Public Policy; *Student Costs

ABSTRACTThis report contains a conceptual framework for analyzing

costs and prices by evaluating the higher education production function andthe determinants of both prices and costs. The framework can be used tostrengthen understanding of costs and prices within individual institutionsand to inform macro level investments at state and national levels. Despitethe breadth of choices in higher education, average costs differ in somewhatpredictable ways, by level of instruction, discipline or field, andinstitutional type and control. Any analysis of price structures mustincorporate subsidies and strategies for differential pricing. There aredifferentials in sticker prices based on residency status, level ofinstruction, and institutional type and control. Price-to-cost ratiosdeveloped in this report point to broad trends across institutions. In spiteof differences in missions and resources across institutions, categories ofinstitutions exhibit somewhat consistent pricing and cost patterns. Theanalyses in this report suggest that greater use of available data willenhance understanding of cost and price structures within higher education-tohelp institutions ensure that their financing decisions are compatible withinstitutional goals. (Contains 17 figures, 58 endnotes, and 51 references.)

(SLD)

Reproductions supplied by EDRS are the best that can be madefrom the original document.

1

PERMISSION TO REPRODUCE ANDDISSEMINATE THIS MATERIAL HAS

BEEN GRANTED BY

K cTO THE EDUCATIONAL RESOURCES

INFORMATION CENTER (ERIC)

usAGroup Foundation

New genda Serier

01.

Cost, Price and Public Policy:Peering into the Higher Education Black Box

William L. Stringer, Alisa F. Cunningham, with

Jamie P. Merisotis, Jane V. Wellman, and Colleen T. O'Brien

Mo The Institute for Higher Education Policy

U.S. DEPARTMENT OF EDUCATIONOffice of Educational Research and Improvement

EDUCATIONAL RESOURCES INFORMATIONCENTER (ERIC)

This document has been reproduced asreceived from the person or organizationoriginating it

Minor changes have been made toimprove reproduction quality.

Points of view or opinions stated in thisdocument do not necessarily representofficial OERI position or policy.

VOLUME 1 NUMBER 3 AUGUST 1999

2 BEST COPY AVAILABLE

Acknowledgments

This report was prepared by William Stringer, Senior Associate, and

Alisa Cunningham, Research Analyst, at The Institute for Higher

Education Policy, with guidance and analytic support from Jamie

Merisotis, President, Jane V. Wellman, Senior Associate, Colleen

O'Brien, Managing Director, and other members of The Institute's staff.

The Institute would like to acknowledge several individuals andorganizations who offered technical expertise, advice, and guidance for

the report. In particular, we would like to thank: Ronald Phipps, Senior

Associate at The Institute; Michael Middaugh, University of Delaware;

Dennis Jones, National Center for Higher Education Management

Systems; Gil Kline, Stategic Communications; and the staff and

associates affiliated with the USA Group Foundation, includingMartha Lamkin, Bob Dickeson, Jean Rose, and Natasha Swingley.

Institutional Cost ConsiderationsThe Determinants of CostSpecial Characteristics of Higher Education

-Average-Cost PatternsCost Measurement and AllocationEducation and General (E&G) Expenditure CategoriesInstitutional Type and Carnegie Classifications

Alternative Pricing StructuresThe Determinants of PriceSources of Revenue

Data on Pricing Patterns

The Relationship Between Price and CostThe Expected RelationshipRelating the Data on Average Costs and Prices

ImplicationsInternal Cost EvaluationPricing Policy

---State-Policy---Future Concerns

AppendicesThe University of Delaware National Study of

Instructional Costs and ProductivityThe Integrated Postsecondary Education Data System (IPEDS)Comparing Costs and PricesAdditional Data Needs

BEST COPY AVAIEABLE 4

usAGroup Foundation New Agenda Series'VS'

by:

William L. Stringer,

Senior Associate

Alisa F. Cunningham,

Research Analyst

Jamie P. Merisotis,

President

Jane V. Wellman,

Senior Associate

Colleen T. O'Brien,

Managing Director

The Institute for

Higher Education Policy

usaGroup Foundation'

Cost, Price and Public Policy:Peering into the Higher Education Black Box

Executive Summary

Escalating tuitions and public accountability demandshave brought considerable attention to the cost and price

structures of American colleges and universities. This report is

intended to serve as a primer on costs (the overall expenditurepatterns, or what institutions spend in support of theirmissions) and prices (the charges to students and other consumers).

The report includes a conceptual frame-work for analyzing costs and prices byevaluating the higher education productionfunction, and the determinants of bothprices and costs. This framework can beused to strengthen understanding of costsand prices within individual institutions, aswell as informing macro level investmentsat the state and national levels.

Insight into differential cost structuresflows from the recognition that the highereducation production process varies frominstitution to institution. Colleges anduniversities have broad latitude in choosingthe quantity and quality of faculty, supportstaff; physical capital, and other inputs inorder to produce educated students(instruction), knowledge (research), andpublic service. The nature of these prod-

ucts depends fundamentally upon themission and goals of an individual institu-tion. The ability to meet the mission andgoals depends upon the productionprocess employed by the institution.Factors such as the use of distance learn-ing, the choice of course sizes, and theemployment of adjunct faculty are part ofthe higher education production process.Despite the breadth of choices, averagecosts differ in somewhat predictable ways:

By level of instruction (lower division,upper division, and graduate): Eachlevel has successively higher costs.

By discipline or field: Such disciplines

as engineering, health professions, andmulti/interdisciplinary studies have

EST COPY AVAILABLE 5

relatively higher costs; English and

literature, psychology, and protectiveservices have relatively lower costs.

By institutional type and control:Research universities have relatively

higher costs than comprehensiveuniversities; private institutions haverelatively higher costs than publicinstitutions.

Because higher education institutionshave sources of revenue in addition totuition and fees, an understanding of pricestructures must incorporate subsidies andstrategies for differential pricing. For eitherincentive or revenue-maximizing purposes,some groups of students are chargedtuition and fees that diverge from theinstitution's average level. Although thesechoices are complicated by the overlappingpatterns of tuition discounts, nationalaggregate data indicate some specifictrends in average sticker price differentials:

By residency status: Out-of-statestudents generally pay higher rates oftuition than do in-state students atpublic institutions.

! By- level- of-instruction:- Graduate -andfirst-professional students pay slightlyhigher tuition per credit hour than doundergraduates.

By institutional type and control:

Private, non-profit institutions chargehigher prices than do public institu-tions, and research universities chargehigher prices than do communitycolleges.

These price differentials do not reflectcost differences exactly, even when costsand prices vary in the same direction.Decisions to differentiate tuition rates restto some extent on conscious choices bystates and institutions regarding which

groups of students should (and would bewilling to) pay more of their educationalcosts. Because tuition levels do not varyexactly with instructional costs even with-in the same institution, different studentspay different proportions of those costs.

What proportion of the costs of highereducation is met by stated levels of tuition,on average, for various subgroups ofstudents? At an aggregate level, price-to-

cost ratios presented in this report point tobroad trends across institutions:

Students at research universities andcommunity colleges tend to pay alower proportion of their instructionalcosts than do their counterparts atother types of institutions.

At public institutions, out-of-statestudents pay a higher proportion oftheir instructional costs than do in-state students, at both undergraduateand graduate levels.

Students in low-cost disciplinesgenerally pay a higher proportion oftheir instructional costs than dostudents in higher-cost disciplines.

Students-at-private-institutions tend-topay a higher proportion of theireducational costs, on average, than dostudents at public institutions.

From the opposite perspective, levels ofsubsidy (the proportion of costs paid bynon-tuition revenue, including taxpayermoney) also differ among groups ofstudents. The available data reveal theimportance of subsidy at every type ofinstitution and support the previousfindings:

Subsidy makes up a higher proportionof educational spending for studentsat public institutions (77 percent, onaverage) than at private institutions

For either

incentive or

revenue-.

maximizing

purposes, some

groups of students

are charged tuition

and fees that

diverge from

the institution's

usAG rou p Foundation

AVAILABVE

Students at

community

colleges have the

greatest

subsidies

as a proportion

of educational

costs, despite

comparatively

lower subsidies in

absolute irmis.

usAGroup Foundation"

(31 percent). This should not besurprising, given that private institu-tions tend to rely more on tuition andfee revenue and less on non-tuitionrevenue (such as government appro-priations) than do public institutions.

Subsidy also differs by institutionalmission. In particular, private researchuniversities appear to have a largersubsidy as a proportion of educationalcosts (47 percent) than do otherprivate institutions, such as compre-hensive colleges (19 percent). At thesame time, subsidy proportions atpublic institutions are relatively greater,

regardless of mission; students at

community colleges have the greatestsubsidies as a proportion of educa-tional costs, despite comparativelylower subsidies in absolute terms.

Both the level of subsidy and the patternof differential pricing reflect to some extentthe mission and goals of an individualinstitution, as well as policy decisions madeat the state or the institutional level. Afterexamining costs and prices, we offer several

observations relevant to such policydecisions:

Despite differences in missions andresources across institutions, categories

of institutions exhibit somewhatconsistent pricing and cost patterns.

Within individual institutions,differential tuitions among differenttypes of students can be used to alterthe composition of the student body.

Alternative sources of revenue subsi-dize virtually all students to someextent.

Because institutions have enormousflexibility in selecting their productioninputs, each institution has a unique

set of costs and highly differentiatedoutputs.

Differences in program costs withinany single institution mean that somestudents are subsidized more thanothers.

Measurement of costs and relatedpricing data can be adapted by aninstitution and used to compare itselfwith other institutions in order toprovide insight into performance andcost effectiveness, while simulta-neously recognizing differences inmissions, resources, and other relevantconditions.

The observations above raise somefundamental questions about the fitbetween institutional purposes andfunding decisions. For an institution orgroup of institutions, are the mission andgoals that are incorporated into coststructures the same as those incorporatedinto pricing patterns, or are they, in fact,operating at cross purposes? Is the relation-ship between costs and prices reflected inprice-to-cost ratios and subsidy patternsthat are consistent with an institution'smission or a state's-spending-priorities? T -he

answers to such questions have implica-tions for institutional decisions on pricing,the evaluation of cost efficiency, and publicpolicy.

Traditionally, however, the highereducation community has asserted thatdetailed data on costs and prices cannot becollected due to methodological problems,and resistance to cost accounting withininstitutions continues to exist. Manybelieve that such data should not becollected because of the difficulty ofcomparing data among institutions withvastly different missions and resources.Nonetheless, if institutional and otherleaders do not critically examine patternsof costs and prices, actual cost/price

BEST COPY AVAHABLE

relationships may drift out of alignmentwith the stated missions and goals ofinstitutions and states.

The analysis in this primer suggests thatgreater use of available data will enhanceunderstanding of cost and price structureswithin higher education. The analysis canhelp individual institutions ensure thattheir financing decisions are compatiblewith institutional goals. In addition, it caninform decisions about public financestrategies, such as aligning subsidies withother resources to achieve the greatest good.

Introduction

American colleges anduniversities expend a

considerable amount ofpolitical and analyticalenergy devising andmodifying their cost andprice structures.

One reason for this focus on cost and pricestructures is that the interaction among thecost of educating students, the revenuereceived by the institutions, and thepattern of prices charged to students has abroad impact upon educational opportuni-ties. This impact can be seen through thepattern of financial aid that can be offeredto students, the pattern of enrollmentamong income groups, the persistence ofany given set of students, and the availabil-ity and choice of courses of study byvarious students.' Studying the effects ofthis interaction is complicated, however,because such an analysis requires an

understanding of the factors that deter-mine each component and their

interrelationship at the institutional level.Over the years, the need to understand

the relationship between college costs,: revenue, and prices' has driven institutions

to develop measurement tools that capturethe unique and complex characteristics ofhigher education. One essential tool is cost

. accounting (also called cost allocation, costdistribution, or cost apportionment), thetechnique employed to trace the flow of

: dollars from the source of income to theexpenditure of that income within aninstitution. Movement of income betweenrevenue centers and cost centers can, in

: turn, help reveal areas of profitability andloss. Whether areas of profitability areexpanded (and areas of loss eliminated)

. depends in part upon the objectives of theinstitution.

The pressures within the educationalcommunity to provide financial account-ability, promote private collaboration, andfollow the new Financial AccountingStandards Board's accounting principlesrequire some form of cost accounting.However, postsecondary institutions findthat the existence of indirect costs, theinterdependent nature of their products,and the subjective nature of choosing unitsof measurement make traditional costaccounting techniques difficult to adapt.

Because of these difficulties, cost

accounting has not been fully imple-mented within higher education.Nevertheless, colleges and universitiesmust strive for financial accountability andgreater understanding of the relationshipbetween the costs of higher education andthe prices that students are charged. Thatgoal becomes increasingly important aspressures mount to account for each dollarof revenue, as institutions rely more ontuition, and as public anxiety aboutcollege prices escalates.

Policymakers have been forced toexamine both prices and costs in order tomeet policy objectives. This accountabilityis especially true at the state level, where

The pressures

within the

educational

community to

provide

financial

accountability,

promote private

collaboration, and

follow the new

Financial

Acccllll

Standar& goal&

accoun

some

cost accouri

BEST COPY AVAILAVIE

Colleges and

universities must

strive for financial

accountability and

greater

understanding

of the relationship

between the costs

of higher

education and the

prices that students

usAGroup Foundation

taxpayer funds appropriated directly topublic institutions to keep tuition and feeslow must be used in alignment withoverarching state policy goals. On thenational level, discussions from the 1973Carnegie Commission on Higher Educationto the 1998 National Commission on theCost of Higher Education have recom-mended strengthening institutional costcontrol and management, with the goal ofkeeping college affordable to all who wishto obtain a postsecondary education.

This report presents a conceptualbackdrop for understanding the factorsinfluencing the costs of higher educationand the prices students are charged, as wellas the relationship between themfocusing primarily on instructional costs.The report examines empirical evidencefrom existing studies regarding price,revenue, and cost patterns at various typesof institutions. Data are drawn from theU.S. Department of Education's IntegratedPostsecondary Education Data System(IPEDS), which collects comprehensivefinancial information from virtually allU.S. postsecondary institutions annually.In addition, data are derived from morespecific studies such as the University ofDelaware National Study of InstructionalCosts and Productivity, a data- sharing-

consortium of almost 300 colleges anduniversities. Due to the nature of theavailable information, the data and analysisfocus on four-year, degree-granting, publicor private, non-profit institutions, althoughmany of the concepts are applicable toinstitutions with other characteristics.Differences in average instructional costs

by discipline, level of instruction, andinstitutional type are compared with theaverage prices charged. Finally, the reportdiscusses the implications of the findingsto institutions and public decision-makingbodies regarding longer-term institutionalfinancing, student composition, andpricing strategies.

DefinitionsSeveral aspects of higher education

financing are fundamental to understand-ing the approaches and data presented inthis report:

Cost structures are the patterns ofexpenditures incurred bypostsecondary institutions; theydepend upon institutional missionsand choices made to achieve theirmissions. Costs ultimately dependupon the manner in which faculty,facilities, capital, administration, andother "inputs" are combined in orderto produce certain products or"outputs." The transformation ofinputs into outputs is called the"production function."

Revenue structures are the level andpatterns of financial resources receivedby institutions. Ultimately, inputs intothe production of higher educationoutputs must be purchased withrevenues obtained from specificsources, such as tuition and fees, gifts,and government appropriations. Withthe exception of certain income that isrestricted (or "earmarked"), each dollarof revenue usually flows-into-a generalfund from which it is dispersed.

Pricing patterns are the schedule oftuition and fees charged to differentstudents with varying characteristics.'Examples of pricing strategies includeuniform tuition and fees to all stu-dents; tuitions reflecting programcosts; tuition differentials by studentresidence (in-state versus out-of-state)or institutional mission (researchuniversity versus baccalaureate

institution); and differentials in nettuition charged to exemplary orminority students.

EST COPY AVM

To a large extent, revenue structures arethe link between cost and pricing patterns.For example, as increases in state appro-

priations have failed to match rising costs,public institutions have increased theirreliance on revenue from tuition and fees(see The Institute for Higher EducationPolicy, 1999). Thus, the following sectionsof this report primarily focus on costs andprices, but incorporate discussion ofrevenue structures as well.

Institutional CostConsiderations

The Determinants of CostWithin broad limits defined by their

mission and circumstances, colleges havediscretion over how they expend theirrevenue and incur costs to "manufacture"their product.' The choices occur withinbroad categories representing the quantityand quality of students, faculty, facilities,and other inputs into the higher educationproduction process.' The total annual costof production for any particular institutionreflects the cost of the inputs, appropriately

amortized.-At the same time, the-pricecharged to the consumer must, in somemanner, reflect (but not necessarily equal)the total costs of producing that product.Although the same could be said for theproduction of a can of beans, a computer,an automobile, or a condominium, thereare some perplexing and challengingnuances in the case of higher education.

[Sum of the Costs of All Inputs] =Total Cost

An examination of the choicespostsecondary institutions make regardingthe quantity and quality of inputs is anessential part of the cost accountingprocess and important to the understand-

ing of both costs and prices. These choices. may be described through a production

function analysis.

The Higher Education Production Function

The quantity and quality of inputsrequired for each different amount ofoutput is usually called a productionfunction, which, for any process, ulti-mately determines the total cost (but notnecessarily price) of the product at everylevel of output.' Economists call themanner through which inputs are trans-formed into outputs the "technology." Inprinciple, some technologies mightproduce exactly the same or higher qualityand quantity of outputs with fewer orlower-cost inputs; in such a case, onecould say that those technologies weremore efficient than others. One goal ofcost accounting is to describe a productionfunction for each output within anorganization and, within those functions,to assess changing technology, the abilityto substitute inputs to achieve the sameoutputs, and the impact of changingrevenue structures upon inputs andoutputs.

The complexity and culture of highereducationcoupled with a paucity of

-necessary data=make it impossible to -even general agreement on the

elements of the production function forhigher education. However, this reportdoes not attempt to "define" a singleproduction function, but rather suggeststhat the nature of any one institution'sproduction function is a primary determi-nant of its cost structure and pricingstrategy. Despite the conceptual andtechnical difficulties, it is useful to exam-ine the production function framework forthe insight that it provides into highereducation cost structures.' Frequentlycited components of such a framework arelisted in the following table.

Thus, at the institutional level, thehigher education enterprise transforms a

As increases in state

appropriations

have failed to

match rising

costs,

public institutions

have increased

their reliance on

revenue from

tutnon

usAGroup Foundation'

T COPY AVAILAB

Each educated

studentone

output of a

postsecon

institutionis

subjected to a

range of academic

experiences that

varywidely

--m cost

from one student

to another and

from one

institution to

another.

usAGroup Foundation.

variety of inputs into some combination ofoutputs through a process (technology)that involves a broad range of activities.The cost per unit of output results fromdecisions about the quantity and thequality of inputs. To a large extent, thesedecisions depend on the mission and goalsof the institution. The production func-tion framework includes only outputs andinputs directly used or produced in theeducation processitems that impact the

courses require theaters, which cost morethan seminar rooms; and graduate educa-tion may require more specialized

equipment, smaller class sizes, and morehighly compensated professors thanundergraduate education does. As a result,each educated studentone output of apostsecondary institutionis subjected toa range of academic experiences that varywidely in cost from one student to anotherand from one institution to another.

Inputs into the processinclude the quantity

and quality of:

The activities (technology)that transform the inputs

into outputs include:

Outputs could becategorized as:

FacultySupport StaffPhysical CapitalStudents

InstructionInstitutional Support ActivitiesSponsored and Unsponsored

ResearchStudent ServicesComputer and Media Technology

Educated Students (Instruction)Knowledge (Research)Public Service

supply of higher education productsanddoes not include factors that may berelated to the demand for an institution'sproducts, such as location, studentincome, or the condition of the employ-ment market. Nevertheless, identificationof the elements of the production func-tion, and hence the elements of cost,remains complex.

Different production functions un-doubtedly exist by institutional type, or bysub-divisionsby degree program, by levelof instruction, by department or disci-pline, or even by course.' For example,research universities probably have

different production functions thancommunity colleges, and within each typeof institution, the production function forundergraduates probably differs from thatof graduate students. Because of thesedifferences, one would expect cost varia-tions to be found in each academicdivision. Within a single universityprogram, discipline, or class level (andcertainly between institutions) some unitcosts will exceed others. For instance,

laboratory sciences typically require more

resources than do the humanities; fine arts

Marginal and Average Cost

The cost of a unit of inputfaculty,staff, student, or new buildingcontrib-utes directly to the total cost of production.If the cost of each input were known, theproduction function also could be used todetermine both an "average cost" at eachunit of output and a "marginal cost" ofproducing just one more unit of output.The distinction between average and mar-ginal costs is important to an examinationof the cost/price relationship because eachconcept is used for a different purpose:

Average costs are used to measure

accounting totals or subtotals, as totalcosts equal the average cost times thenumber of units produced. This is trueirrespective of how many resources

were expended to educate eachincremental student (marginal cost).There is, of course, an infinite numberof combinations of spending on eachstudent that would give rise to thesame average cost total. At the sametime, average cost implies that acertain amount of average revenue isnecessary to cover those costs.

BEST Copy AVAILAME

[Average Cost per Unit of Output] x[Number of Units] = Total Cost

Marginal costs are the incrementalexpenses that will be incurred byadditional or curtailed production.Marginal costs are useful in predictingproducer behavior. A collegemustdecide whether to add one morestudent, one more faculty member, orone more computer terminal based onhaw much-that addition contributes

toward meeting the college's objectives.

The additional unit may add more orless than the previous unit did.

[Sum of the Marginal Cost of Each

Unit of Output] = Total Cost

Thus, there is a gap between thequestions that are asked of cost account-

.

ing and the answers that can be providedfrom the data collected through costaccounting. Cost accounting data gener-ally allow the calculation of average costs.

Special Characteristics of Higher Education

The higher education production process has at least four characteristics thatdefy the usual "production function" and cost accounting analyses.'

First, one of the outputs (an educated student) is also an input (an uneducatedstudent). This situation is similar to the case of raw wheat being processed intorefined flour, with one important distinctionthe student gets to choose the pro-duction function (institution) to which he or she will submit. Coupled with othercharacteristics noted below, this situation often leads to competitive bidding forstudents with certain identifiable characteristics and selective pricing policies thatencourage or discourage certain types of students.

Second, an inherent inequality is built into the dynamic of higher education. Otherthings being equal, greater subsidization of students attracts a greater number and/or higher quality of inputs (students). The result is a hierarchy of schools in whichquality may depend on the extent of subsidy. By itself, this is similar to privateindustry where higher-priced inputs ensure better - quality outputs. What makeshigher education different, however, is the self-perpetuating nature of the process.'The ultimate result is that a good student could pay less to attend a "good" schoolthan a mediocre student pays to attend a "mediocre" school.

Third, although market niches are not unusual in business, what is atypical in the

case of higher education is the extreme diversity in price and quality of educationalinstitutions. This diversity is exacerbated by the hierarchy of institutional prestigethat has developed, as well as allegiances of potential students to state, local, orlegacy institutions. At the same time, the inputs into the academic process (particu-

larly faculty) are much more varied than in most other production processes.Finally, higher education involves interdependent products that may be produced

in varying combinations. Joint products are not uncommon in industry: for example,the petroleum industry produces kerosene, napa, tar, and gasoline from the samecrude oil input. These industry outputs are produced essentially in fixed and knownproportions. In higher education, however, the processes used to produce eachoutput are interrelated, and institutions can choose to produce a wide range ofoutput combinations (depending on institutional missions). Higher education there-fore involves a variable input/variable output production function that is difficult to

both characterize and quantify.

EST COPY AVAUBLE 12

Higher education

involves a

variable input/

variable output

production

fiinction that is

difficult to both

characterize and

quantify.

9

10

Institutions' ratios

between

average cost

and average

tuition

reveal an

important element

of their cost/price-

structures.

usAGroup Foundation'

Because average cost suggests that a specific

level of average revenue is required,

knowledge of average costs provides theaverage tuition level that is required giventhe amount of non-tuition revenue, or viceversa.12 Average cost information also

enables insightful comparisons betweeninstitutions or similar groupings ofinstitutions. Institutions' ratios betweenaverage cost and average tuition reveal an

important element of their cost/pricestructures.

On the other hand, marginal costinformation cannot be gathered easily fromtraditional cost accounting data.13 This ismost relevant for issues of changes in

production, as well as predictions of thestudent response to institutional behavior.For example, the question of whether aprice differential between in-state and out-of-state students will attract more or fewerin-state students requires marginal costdata (along with some additional informa-tion reflecting the nature of the twocategories of students).

Average Cost PatternsDespite measurement difficulties,

average costs can be examined using theavailable data. Even within the instruc-tianal-production-functionTunit-costs (theamount higher education institutionsspend per student or per credit hour) differtremendously among institutions.

Defining Categories of Expenditure

and Instructional Costs

In the conduct of their operations,institutions incur costs for goods andservices in many different areas. Currentexpenditures (expenditures by institutionson their current operations, as opposed tocapital expenditures) may be examinedwithin the context of the U.S. Departmentof Education's Integrated PostsecondaryEducation Data System (IPEDS)." Theexpenditure data collected through theIPEDS Finance survey (NCES, 1996) have

traditionally been cost- based.15 Looking at

costs relative to the activity they support. reveals that the bulk of current expendi-: tures are called education and general

(E&G) expenditures. E&G expendituresare devoted to instruction, research, public

: service, and administration; they exclude; expenditures for self-supporting operations

or operations that are independent of theprimary missions of the institutions, suchas auxiliary enterprises, hospitals, andindependent research laboratories.'6

Institutions distribute their total E&Gexpenditures among various functionalcategories (see Figure 1, page 12). In thecase of public institutions, on average 38percent of their combined E&G expendi-

. tures went toward instruction in 1995-96;: the next highest categories were research

(12 percent), institutional support (11percent), and academic support-(9-per-

: cent). Private, non-profit institutions onaverage spent 33 percent of their combinedE&G expenditures on instruction, fol-lowed by scholarships and fellowships

: (16 percent), institutional support (13percent), and research (9 percent).

Different types of institutions vary theirdistribution of expenditures among the

: functional categories. For example,research universittectend to spend agreater proportion of their total E&Gresources on research than do other

; institutionsalmost 22 percent overall,compared with doctoral universities, thenext highest with 9 percent. Baccalaureate

. institutions spend the greatest proportion; on scholarships and fellowships, 23

percent, compared with other types ofinstitutions (NCES, 1996). In addition,

: large institutions tend to spend smallerproportions of their educational expendi-tures on support functions such asinstitutional support, student services, and

. operation and maintenance of plant; this islikely due to economies of scale (Bowen,

1980). The way in which these total

BEST COPY AV A 1 11ABLE

Cost Measurement and Allocation

If agreement were reached on the nature of the higher education production func-tion, measurement of the elements themselves would still pose substantial obstacles.None of the obstacles argue that data are impossible to gather or are meaningless.Rather, they suggest that care should be taken in the collection, presentation, and in-terpretation of cost data.

There are two basic problems with measurement. First, in the face of multiple, some-

what overlapping products, higher education must choose a unit of accountforexample, costs per unit of educated student, or per unit of knowledge, or per unit ofpublic service. At the same time, the measure must capture both the quantity and thequality of inputs. Ratios and values designed to capture these elements are difficult todefine, let alone measure. In his often cited 1980 work, Howard Bowen concludes that"there is simply no known way, except through the broad general judgment of experts,to measure the output of an institution's organized research or public service program.Cost studies, therefore, are usually confined to the educational function for which atenable measuring unit is available" (p. 5).

Even when cost analysis is limited to the educational function, however, the basisfor student units must be determined. For example, costs may be measured per full-time equivalent (FTE) student, or per student credit hour (SCH). The choice is significant,

especially when comparing costs by level of instruction. Because the course load ofthe typical graduate student is lower than that of a typical undergraduate, cost ratiosbetween graduate and lower division levels generally will be higher if they are re-ported on the basis of credit hours rather than per FTE students (Brinkman, 19891.b0

- In the absence of easily quantifiable input and output variables, proxies are oftenused. Morton Schapiro (1993) notes several measures as proxies for the output of "edu-

cated students," including years of education, test scores, size of graduating class, thepercentage of entering freshmen who eventually graduate, and the percentage of se-niors who go on for graduate education." In addition, factors such as teacher experience

and student-to-faculty ratios can be considered as proxies for the quantity and qualityof various inputs (in this case, faculty). However, such proxies can only capture a por-tion of the production function conceptsthat is, it is the quantity and quality of facultythat is relevant as an input, not the student-to-faculty ratio itself.

The second basic problem with cost measurement is that the total cost of providingthe entire educational process includes not only costs that are directly budgeted andattributable to the primary missions of instruction, research, and public service, butalso "unallocated" (indirect) costs that are accounted for separately. Indirect costs caninclude support costsadministrative expenses, security, library costs, and other costscertain costs of debt financing, and depreciation of capital. It is clear that these indirectcosts are used by some students more than others; however, it is virtually impossibleto allocate these costs among various students on any meaningful basis. The mostbasic problem of implementing cost accounting in higher education has always beenallocating portions of these costs to the various outputs and, further, to programs,disciplines, and courses. Decisions regarding the allocation of costs are always some-what arbitrary and may bias the conclusion of a study in the direction of the ad hocassignment of cost. In the case of the costs of educated students produced, for ex-ample, should the allocation be pro rata by number of students, by student credit hoursproduced, by number of faculty, or by already allocated expenses? The most commonapproach is to weight the costs by class level or program enrollment.

BEST COPY AV '1

4,

LABLE

I he most basic

problem of

implementing

cost accoun

in higher

education has

always been

allocating portions

of these costs to the

various outputs

and, fuither,

institutions vary

their

distribution

of expenditures

_among the

functional

Figure 1: Distribution of E&G Expenditures,by Institutional Control, 1995-96

(3,418 institutions)

-£;"..v`V \/ r / / I I

N. N. S. \ \ 1 1 \1 r /

1 1. 1 S. /\ s. s.

Scholarships and fellowships

Operation and maintenanceof plant

0 Institutional support

Note: Includes only those institutions that responded to every data field. Public and private, non-profitcategories include all Carnegie Classifications.

Source: NCES, 1996

expenditures are distributed impacts thecalculation of unit costs.

AlthoughtliccOsts of research andpublic service are important because theyare part of the institutional mission, it isdifficult to include them in any measure-ment of unit costs due to the lack ofappropriate measurement units. E &Gexpenditures therefore can be narrowedfurther to "educational costs," whichcomprise the portion of current expendi-tures that is directed toward the instructionand welfare of students. Educational costswere defined by Bowen (1980, 1981) asE&G expenditures minus the outlays forresearch and public service, and minus aprorated share of overhead costs attribut-able to research and public service. Whatremains is instruction, student services,institutional scholarships and fellowships,

plus a prorated share of academic support,institutional support, and operation andmaintenance of plant. Other studies havemeasured "instructional costs" even morenarrowly than educational costs.

The functional categories of E&Gexpenditures also can be divided intodirect costs and indirect costs. Threeexpenditure categoriesinstruction,research, and public servicerepresentdirect costs, as they are directly attributable(and budgeted) to an institution's mission.Other expenditure categories (academicsupport, institutional support, studentservices, and operation and maintenance ofplant) are indirect support costs, which areincurred through the production of morethan one product and can only be esti-mated and allocated on a somewhatarbitrary basis. Finally, for the purposes of

BEST COPY AVAIIALE

unit cost measurement, scholarships andfellowships are frequently treated as a netreduction in pricerather than a cost ofeducationbut otherwise they are usuallyconsidered as indirect costs."

Previous cost studies have used data onfull costsdirect plus indirect costsemploying a variety of procedures todistribute the indirect costs (see Bowen,1980; Brinkman, 1993). On the otherhand, many studies focus solely on directcosts in order to avoid issues regarding thebasis of allocation. Thus, cost measure-ment analyses may examine any

combination of the following, depending

on the desired goals: full costs or direct

costs alone; instructional costs, educa-tional costs, or E&G expenditures as awhole; and costs inclusive or exclusive of

scholarships and fellowships.21

Expected Differences in

Direct Instructional Costs

One frequent focus of cost studies isdirect instructional costs, which includeboth faculty compensation and operatingexpenses (Banziger et al., 1997). Typically,80 to 90 percent of academic departmentdirect expenditures are attributable to

Education and General (E&G) Expenditure Categories

E&G expenditures include-the following functional categories:"

Instruction: activities related directly to instruction, including expenditures forfaculty compensation, office supplies, and administration of academic depart-ments, as well as expenditures for departmental research and public service thatare not budgeted separately.

Research: activities specifically organized to produce research outcomes, com-missioned by an external agency, or budgeted separately to an internalorganizational unit.

Public service: activities established primarily to provide non-instructional ser-vices that benefit external groups and budgetedspecifically for such public service.Academic support support services that are an integral part of the institution'smission, including expenditures for libraries, academic administration, academiccomputing support, and curriculum development.Student services: admissions, registrar activities, and activities whose primarypurpose is to contribute to students' well-being and development, including expen-ditures for career guidance, financial aid administration, and student health services.Institutional support day-to-day operational support of the institution, includingexpenditures for physical plant operations, general administrative services, legaland fiscal operations, and public relations.

Operation and maintenance of plant service and maintenance related to groundsand facilities used for education and general purposes, including expendituresfor utilities, property insurance, and similar items.Scholarships and fellowships: outright grants and trainee stipends to individualsenrolled in formal coursework, including aid to students in the form of tuition orfee remission, but not including aid that is exchanged for student work.18 Prima-rily institutional aid is represented in this category, but some federal aid (such asPell Grants) and other grants that are distributed through the institutions are alsoincluded. (Scholarships and fellowships are frequently treated as a reduction in pricerather than as a cost, or at the least are excluded from many cost studies.19)

7 t

'I: _it

y=;

Scholarships and

fellowships are

frequently

treated as a

reduction

in price

rather than as a

cost, or at the least

are excluded from

many cost studies.

_I

,',ustiGroup Foundation"

BEST COPY AVAILABLE

Typically, 80 to

90 percent of

academic

department

direct expenditures

are attributable

to salaries

and benefits.

14

usAGroup Foundation'

salaries and benefits (Middaugh, 1999).Direct instructional costs differ bothwithin and among institutions. Thesedifferences result from a combination offactors: student-to-faculty ratios; the typeand salary level of faculty and staff; the useof supplies and equipment; and eveneconomies of scale. Differences in costsoccur on several levels:

By discipline or field. Some disciplines,

such as natural sciences, tend to havehigher direct instructional costs thanothers. The expected differences mayderive from underlying productionrelationshipssuch as a reliance onlaboratory courses, the costs of equipment,or the use of small seminarsor may resultfrom the effects of scale, as in the case ofdisciplines that are underenrolled relative

to faculty capacity (Brinkman, 1989). Inaddition, costs vary by discipline or fielddue to differences in the number, type, andprice of faculty, equipment, supplies, andother inputs (Brinkman, 1993).

By level of instruction (lower division,upper division, and graduate). Higherlevels of instruction generally have highercosts per student credit hour. There areseveral reasons: student-to-faculty ratiostend to be higher at the lower divisionlevel; more junior and part-time faculty orteaching assistants, who tend to have lowersalaries, are used in lower division courses;

and fewer supplies and equipment tend tobe used on a per-credit-hour basis in lowerinstruction levels. In addition, scale effectsmay contribute to differences in costsforexample, smaller-scale graduate programs(with low enrollment) are likely to havehigher costs per credit hour, and vice versa(Brinkman, 1989).

By institutional type/mission (CarnegieClassification). Larger, more complexinstitutions (in terms of degree levels,degrees awarded, and emphasis on re-search) tend to have greater differences incosts by level of instruction, especially

comparing graduate with lower division

(Brinkman, 1989). Institutional missionsare relevant; some types of institutions aremore heavily involved in graduate andprofessional study than are others, andinstitutions with more higher divisionstudents will generally have higher costs(Bowen, 1980, 1981). Economies of scaleare also important in this respect. Finally,certain types of institutions may alsoemphasize higher-cost disciplines or fields,such as engineering or medicine.

By institutional control (public /private).Differences in scale and in the proportionof graduate students, both of which affectcosts, tend to be large between public andprivate institutions with similar missions

(Brinkman, 1993). Although differences inscale and proportion depend upon thenature of the production functions forgraduate and undergraduate students, theymost likely result in relatively lower unitcosts for public institutions.

These differences are not clear cut, butoverlap considerably. For example, certaintypes of institutionsthose that focus ondoctoral educationcan use doctoralstudents, who are compensated at lowerrates, to teach lower division courses.23 At

the same time, differences in costs are notstatic over time. Rather, trends in unitcosts are greatly influenced by changinglevels of faculty and staff compensation, as

well as enrollment shifts among institu-tional types. Overall instructional costs perstudent remained steady or declined in theearly part of the century, rose quickly&tiring the 1950s and 1960s with the rapidexpansion of the higher education system,and again remained steady or declinedduring the 1970s (Bowen, 1980).

Benchmark Data on Costs

One recent study of direct instructionalcosts is the University of Delaware Na-tional Study of Instructional Costs andProductivity, a data-sharing consortiumthat was established in 1992. Almost 300colleges and universities have participated

BEST COPY AVM ! ':

Trends in unit

costs are

greatlY

influenced

by changing levels

of faculty and staff

compensation, as

well as enrollment

shifts among

institutional typs.

:-USAGrOUp Foundation'

in the study since its inception by volun-teering cost data at the level of academicdiscipline. The study is intended to belongitudinal, and currently is in its sixthnational data collection cycle. It focuses ondirect instructional costs (personnelcompensation, supplies and services, anddepartmental research and service that arenot budgeted separately), collected on the

basis of student credit hours; indirect costsare excluded (see University of Delaware,1999; Middaugh, 1999).

Analysis by the University of Delaware

of the 1996-97 data led to the develop-ment of national benchmarks organizedby academic discipline and by CarnegieClassification within each discipline. Theparticipating institutions are primarily

Institutional Type and Carnegie Classifications

The most widely used classification scheme for institutional type was developedby the Carnegie Foundation for the Advancement of Teaching in 1970. The mostrecent update was in 1994, and currently includes about 3,600 U.S. institutions that

are degree-granting and accredited. The classification categories are based prima-rily on academic mission; institutions are classified according to their highest degreelevels offered, the number of degrees conferred by discipline, and the amount offederal support for research received by the institution (Carnegie Foundation, 1994).

This report uses modified Carnegie Classifications, which combine similar cat-egories. The categories used (along with their codes in IPEDS) are the following:

Research universities (codes 11 and 12): These institutions offer a full range ofbaccalaureate programs, are committed to graduate education through the doc-torate, and give high priority to research. They receive $15.5 million annually infederal support, and award 50 or more doctoral degrees each year.Doctoral universities (codes 13 and 14): These institutions offer a full range ofbaccalaureate programs and are committed to graduate education through thedoctorate. They award at least 10 doctoral degrees annually in three or moredisciplines, or 20 or more doctoral degrees in one or more disciplines.

Master's (Comprehensive) universities and colleges (codes 21 and 221:22 Theseinstitutions offer a full range of baccalaureate programs and are committed tograduate education through the master's and professional degrees. They award20 or more master's degrees annually in one or more disciplines.

Baccalaureate colleges (codes 31 and 32): These institutions are primarily un-dergraduate colleges with major emphasis on baccalaureate degree programs.This category includes Liberal Arts (Baccalaureate I) colleges, which award 40percent or more of their baccalaureate degrees in liberal arts fields and are re-strictive in admissions.

Associate of Arts colleges (code 40): These institutions offer associate degree or

certificate programs and, with few exceptions, offer no baccalaureate degrees.With the exception of public community colleges, this report generally excludesthese institutions from the analysis.

Although this classification system is useful for cost comparisons among institu-tions, it should not be considered to represent true peer groups. The institutionswithin each Carnegie Classification share important similarities, but also differ inways that can influence costs (Brinkman, 1993). The Carnegie Foundation also isreviewing its classification system for possible changes.

BEST COPY AVAL, .4k.

LE

Average

instructional costs

for one discipline

may be

three tirnes

that of another

discipline, or even

greater in

individual cases.

public colleges and universities, exceptthose institutions in the baccalaureateclassification, which are mostly privateliberal arts colleges; community colleges

are not included in the group.24 Becausethe following analysis is derived from thebenchmarks rather than individual institu-tional records, the data should be viewedonly as indicators, not precise measures."

Overall, the Delaware Study datasupport several general statements aboutdifferences in direct instructional costs:

By discipline or field Certain disci-plines have higher average direct

instructional costs than others (seeFigure 2). In fact, average instructionalcosts for one discipline may be threetimes that of another discipline, oreven greater in individual cases. Forexample, the average expenditures perstudent credit hour for protectiveservices, English and literature,

psychology, and philosophy andreligion tend to be relatively low,

whereas the averages for engineering,health professions and related sciences,

and multi/interdisciplinary studiestend to be relatively high. This patterngenerally holds within each CarnegieClassification; for example, engineer-

ing has the highest or second highestaverage expenditure per student credithour for research universities, doctoraluniversities, and comprehensiveinstitutions (it was not reported forbaccalaureate institutions) (seeFigure 3).

By institutional type /mission

(Carnegie Classification). Researchuniversities spend more to deliver astudent credit hour of instruction$198 on averagethan do doctoraluniversities and comprehensive

institutions, which spend $170 and$129, respectively, on average (seeFigure 4). Baccalaureate institutionsspend about as much per studentcredit hour as doctoral universities,$163 on average. This pattern holdstrue within many disciplines as well;-for example, research universitiesspend the most to deliver a studentcredit hour of instruction in biologicalscience, followed by baccalaureate

institutions, doctoral universities, andfinally comprehensive institutions.However, baccalaureate institutionsfrequently have the highest directinstructional costs per student credit

----hour, in such disciplines as English

Figure 2: Rank of Direct Instructional Costs per Student Credit Hour(Highest and Lowest) by Discipline and Carnegie Classification

Research Doctoral Comprehensive BaccalaureateHighest 1 Engineering Multi/Interdisciplinary

Studies

Engineering Health Professions

2 Health Professions Engineering Health Professions Visual & Performing Arts

3 Public Administration Library Science Engineering-RelatedTechnologies

Physical Sciences

Lowest 1 Protective Services Protective Services Protective Services Business Mangement

2 English Language& Literature

English Language

& LiteraturePsychology English Language

& Literature

3 Philosophy & Religion Parks, Recreation,

Leisure and FitnessStudies

Philosophy & Religion

1 VI

Psychology

Source: Delaware Study benchmarks, 1996-97

BEST COPY AVAILABLE

and literature, philosophy andreligion, and computer science.

Middaugh (1999) provides someexplanations for these cost patterns. Hisanalysis of individual records indicatedthat research universities have lower

faculty workloads (student credit hours perFTE faculty) than do doctoral universities,which in turn have lower workloads onaverage than do comprehensive institu-tions. In addition, faculty at research anddoctoral universities are less involved in -

undergraduate and lower-division instruc-tion than are their counterparts atcomprehensive institutions. Finally,Middaugh notes that baccalaureateinstitutions have fewer student credithours- per-FTE faculty than any otherCarnegie Classification, possibly becausethe participating institutions in thiscategory are largely private, selective

institutions with missions that embracesmall class sizes and few graduate teachingassistants.

Unfortunately, the national benchmarksfrom the Delaware Study cannot illustratedifferences in direct instructional costs bylevel of instruction. Examining a morespecific study of instructional costs by theArkansas Department of Higher Educa-tion can illuminate these differences. Forthe past three years, the department hascollected cost data from all of the publicfour-year and two-year institutions in thestate of Arkansas. Institutions reported

expenditures directly attributable toinstruction, research, and public service;these expenditures were allocated tocourses associated with academic depart-ments primarily on the basis of eachcourse's student credit-hour productionrelative to total departmental studentcredit-hour production. In addition,support costs not allocated directly todepartments and other indirect costs weredistributed among all courses, as well as toresearch and public service. The /Vicunas

data therefore represent full costs ratherthan direct costs. The most recent dataavailable are for 1996-97 (ADHE, 1998).

The Arkansas data also support theexpected pattern of differences in instruc-tional costs:

By level of instruction (lower division,upper division, and graduate). Totalexpenditures per FTE student differeddepending on the level of the degreeprogram, with higher degree levelsbeing more expensive than lowerdegree levels. (See Figure 5.) For

example, annual expenditures perFTE student averaged $15,950 fordoctorate degrees, $8,837 forbachelor's degrees, and $6,875 for

--associates-degreesTAverage expendi-tures per FTE doctoral student weremore than double the average forassociate's degree students.

To supplement these findings, IPEDSexpenditure data can be used to revealdifferences in costs between public andprivate institutions. Data for 1994-95were used by Winston (1997) to calculateaverage educational spending by controland Carnegie Classification. The datarepresent full educational costs for almost3,000 degree-granting institutions.26

By institutional control (public/private): Average educational spend-ing per FTE student is higher forprivate universities than for publicinstitutions (see Figure 6). While

annual educational spending per FTEstudent was $11,967 for all institu-tions, it was $9,919 for publicinstitutions and $14,172 for privateinstitutions. At research institutions,the difference was even greater:

$13,448 for public institutions,compared to $32,014 for privateinstitutions.

usAGroup Foundation'

BEST COPY AVAIIAIBLE9 nv

Figure 3: Rank of Average Direct Instructional Costs per Student CreditHour by Discipline and Carnegie Classification

On a scale of 0 to 10, with 0 being the most expensive

51.00 Health Professsions & Related Sciences

44.00 Public Administration & Services

22.00 Law & Legal Services

25.00 Library Science

40.00 Physical Sciences

26.00 Biological Sciences/Life Sciences

4.00 Architectural & Related Programs

3.00 Conservation & Renewable Natural Resources

2.00 Agricultural Sciences

1.00 Agricultural Business & Production

13.00 Education

5.00 Area, Ethnic & Cultural Studies

50.00 Visual & Performing Arts

11.00 Computer & Information Services

30.00 Multi/Interdiciplinary Studies

15.00 Engineering-Related Technologies

16.00 Foreign Languages & Literature

52.00 Business Management & Administrative Serv.

9.00 Communications

19.00 Home Economics

31.00 Parks, Recreation, Leisure & Fitness Studies

45.00 Social Sciences & History

27.00 Mathematics

42.00 Psychology

24.00 Liberal Arts & Sciences, General Studies

Philosophy & Religion

English Language & Literature/Letters

Protective Services

7--- 6---

Note: Rankings were divided by number of disciplines reported by each Carnegie Classification in

order to convert them all to the same scale. Only major CIP categories are reported.

Source: Delaware Study benchmarks, 1996-97

BEST COPY AVAILABIA. ,2 1

Thus, the available data appear tosupport the expected differences in averageinstructional costs. These differences arelikely due to a combination of factors thataffect individual institutions' production

functions, including the quantity andquality of faculty, class size, varying needsfor equipment, and different institutionalmissions.

Figure 4: Average Expenditures per Student Credit Hour,by Carnegie Classification

(weighted by number of observation at 153 institutions)

$18,000

$16,000

$14,000

$12,000

$10,000

$8,000

$6,000

$4,000

$2,000

$0

Research Doctoral Comprehensive

Note: Only major CIP categories were used for this analysis.Source: Delaware Study benchmarks, 1996-97

Baccalaureate

13 MeanMedian

Figure 5: Average Expenditures per FTE Student atArkansas Public Institutions, 1996-97

(32 two-year and four-year institutions)

$15,950

$11,875A4,$10,856 .

.4-----$8,837

$6,875s 4 J :A

i,-. ,-,

,, Se'

,41,

Associate's Bachelor's

Note: Measures full costs, not just direct costs.Source: ADHE, 1998

Master's

Type of degree

FirstProfessional

Doctorate

IPEDS

expenditure data

can be used to

reveal

differences

in costs between

public and private

institutions.

BEST COPY AVA

A college or

university's income

derives from a

number of

fungible

sources,

including tuition

and fees,

endowment

eamings, and

federal, state,

or local

appro nations.

$16,000

$14,000

$12,000

$10,000

$8,000

$6,000

$4,000

$2,000

$0

Figure 6: Differences in Average Annual EducationalSpending per FTE Student, 1996-97

(2,739 degree-granting institutions)

$14,172

$11,967

$9,919

- ..

All Institutions Public Institutions

Note: Measures full costs, not just direct costs.Source: Adapted from Winston, 1997

Alternative PricingStructures

hik many firms inprivate industry derive

their income from a numberof sources, each stream ofincome can be associated witha particular service orproduct.

A college or university's income also

derives from a number of fungible sources,including tuition and fees, endowmentearnings, and federal, state, or localappropriations Yet, these different incomestreams are not necessarily associated witha particular service or product of theinstitution.

As a result of the non-tuition sources ofrevenue, the price charged to the con-

Private Institutions

sumer (the student) can be lower than thecost of production. The pricing struc-turethe pattern of tuition and feescharged to any student or set of studentsis merely one element of the overall

revenue structure of the institution. Theactual revenue structure and its pricingcomponent depend on many factors,including the philosophy and mission ofthe individual institutiori, the philosophyof the state legislature (in the case of publicinstitutions), the prestige and size of theinstitution, and the overall makeup ofstudents and faculty.

There are four components of pricing. structures:

Tuition, which is charged to allstudents for instructional services,generally per term, per credit, or percourse, and frequently by level ofinstruction.

Mandator), fees, which are charged toall students or all students in certaingroups (for example, fees for healthservices).

EST COPY AVAILABLE A -23

Optional or user fees, which are

charged only for use of facilities andtherefore often differ by program ordiscipline (for example, fees for use oflaboratories or studios).

Tuition discounts, which reduce theprice of tuition for some students andmay come in the form of scholarships,fellowships, tuition remissions, orother price-based incentives providedby the institution.

All of these components of pricing maybe differentiated, leading institutions tocharge disparate prices to various studentsor groups of students. These are called

"price differentials," which refer to thepurposeful-variation of tuition rates inrelation to distinctions in costs, marketforces such as demand, or other factors(Yanikoski and Wilson, 1984, p. 737).

Although the concept of pricingstructures combines all four elements,

available tuition data normally show onlythe patterns of "sticker prices" (tuition andmandatory fees alone) without taking dis-counts into account." Net tuition patternsare extremely difficult to demonstrate,despite the growing prevalence of discounts.

Institutions may discount their prices forcertain categories of students (children ofalumni, residents of certain areas, orvaledictorians of their high school classes)or they may provide aid on a case-by-casebasis. As a result, many colleges and

universities charge almost as many distinctnet prices as there are students receiving

institutional aid (Jenny, 1997). Because thepattern of price discounting is so complexand most data sets do not account for suchdiscounting, it is difficult to know how itaffects any-analysis of pricing structuresand their relationship to costs."

Recognizing the difficulties of gaugingnet tuition patterns, the evaluation ofpricing structures has at least fourdimensions:

The overall level of tuition and fees,

especially relative to those of peerinstitutions or to the constituencybeing served;

The extent to which price covers

average (unit) costs;

The manner in which indirect costsfor all activities are covered by tuitionand fees; and

The incentives that certain pricingstructures have to encourage or

discourage attendance by certaincategories of students.

Data derived from cost accounting can: be quite successful in constructing

responses to the first three issues, but are

less successful in responding to the fourth.

The Determinants of Price

Multiple philosophies and actors areusually involved in setting tuition levels,"

especially for public institutions. On thestate level, explicit policies may rangefrom keeping tuitions as low as possible toencourage student access, to a "hightuition/high aid" policy in which tuitionlevels are allowed to rise while taxpayerfunds are funneled into grants to needystudents. In addition, tuition levels maybe indexed to the levels at peer-groupinstitutions or regional averages, or thedecisions may be made in accordance withinstitution-level policies or budgetaryneeds. The authority to set tuition atpublic institutions usually involves

multiple agencies, such as state legisla-tures, coordinating boards, governingboards, and/or the institutions themselves.Obviously, private, non-profit institutionshave more autonomy in making tuitiondecisions than do public institutions, butthe decisions still involve many partici-pants at the administration and trusteelevels.

24

Many colleges and

universities charge

almost as many

distinct

net prices

as there are

students receiving

institutional aid

1

22

A four-year college

degree is analogous

to an all-day

ticket at

Disney World:

a student pays a

price that entitles

him or her to

partake of the

entire academic

experience.

usAGroup Foundation

Many states and institutions use infor-mal or formal guidelines in theirdetermination of tuition rates. Tuitionincreases may be linked to external 'eco-

nomic variables or internal cost factors,including consumer price indices, statepersonal income levels; and state highereducation appropriation levels. Somelinkage to the cost of providing instructionis also common: in 1997, 12 states linkedthe total cost of education to tuition levelsin the four-year sector using a direct or

_indexed relationship, anc27 states tookthe cost of education into account moreindirectly (Lenth, 1993; Christal, 1997).Average full-time tuition may be set as a

total costs are, themselves, a variable in thebalancing process. But individual statelegislatures and institutions may have some

: inclinationand in some cases a legislative. mandateto alter tuition to cover the

costs that are not covered by other finan-cial resources or, alternatively, to alter

: appropriations to cover a specific propor-tion of total costs.

Third, there is the question of an: institution's willingness to "price discrimi-: nate." Pricing strategies in industry usually

relate-to attempts to segregate the con-

sumer market into distinct segments andthen charge different prices to each

; subgroup in order to maximize profitabil--§pecific ratio to average instructional costs, ity. In this process, the subgroup that isfor example, or states and institutions may . least sensitive to price is generally chargedset tuition rates for certain subgroups of : the relatively higher price. Examplesstudents in relation to instructional costs. abound in shopping malls, -v here clothing

Several questions immediately arise manufacturers sell different lines ofregarding the nature of these decisions. merchandise. Examples also are foundFirst, there is a basic question of the unit of : within regulated industry, such as whenoutput for which the price is chargedfor commercial water rates are less thanexample, per credit hour, per credit hour in- residential rates or bulk postage is less thana particular program of study, per term, per regular postage. In higher education, theyear, or per degree. Proponents of cost combination of many objectives, variedaccounting in higher education tend to . philosophies of pricing, and alternativebelieve that price should be in alignment sources of revenue leads to a variety ofwith the smallest unit cost that is practi- . pricing structures that have little to docal--a-particularlevel-ofstudy-orprogram----with-profit-maximization, however, andof study. The argument could easily bemade, however, that a four-year college

degree is analogous to the all-day ticket atDisney World: a student pays a price thatentitles him or her to partake of the entireacademic experience. The rides he or shetakes, and the curriculum he or shepursues, are up to the student, but what isbeing sold is the overall experience.

Second, the multiplicity of revenueresources gives rise to the question ofwhich financial resource is calculatedfirsttuition and fees, or other sources ofrevenueand which is the remainder.Undoubtedly, the actual process is asimultaneous movement of multiplerevenue sources to cover total costs. The

the practice of price discrimination hasmore to do with the mission of theinstitution.

Average and Differential Prices

In examining pricing structures, it isimportant to make a distinction betweenaverage prices and marginal prices. Similarto the distinction in cost analyses, averageprice represents the average amountcharged per unitto all students or to allstudents in a certain categorywhereasmarginal price represents the amount thatis charged to each additional student. Thisreport focuses on average prices in examin-

ing the relationship between pricing andcost patterns. The average price level

EST COPY AVAMAIka 25

determines the total amount of revenue aninstitution can collect from tuition, given acertain level of enrollment."

[Average Tuition per Student Unit] x[Number of Student Units]

Total Revenue from Tuition

Differential prices can be viewedessentially as charging varying average

prices for the same educational product todistinct groups of students, based onspecific characteristicsa form of pricediscrimination (see below). However,

many differential price patterns can resultin the same average price for studentsoverall. For example, one institution maycharge all students $2,000, while anothermay charge -in -state students $1,000 and

out-of-state students $3,000; if the secondinstitution has an equal number of in-stateand out-of-state students, then bothinstitutions have an average price of

$2,000. Therefore, the overall average pricedoes not reflect the different levels oftuition charged to specific subcategories ofstudents within the overall group.

The average tuition payment by acategory of students is useful in assessing

the extent to which the total costs of aparticular program, discipline, or courseare offset by tuition revenueor, from theopposite perspective, what proportion ofcosts must be covered by revenue fromother sources. Average tuition levels arealso useful in describing the extent towhich price differentials exist betweeninstitutions, categories of students, pro-grams, or disciplines.

As previously mentioned, chargingdifferent prices for the identical product(which has the same average and marginalcost because it is the identical product) is

termed "price discrimination." Becausecategories of students can be identified andstudent demand for any particular institu-tion, program, discipline, or course has atleast some sensitivity to the price charged,

postsecondary institutions regularlypractice price discrimination. Pricediscrimination in higher education leadsto various tuition differentials. The mostobvious example of price discriminationin higher education occurs in the case ofin-state and out-of-state students, butother examples exist. Price discrimination

policies are illustrated in Figure 7, whereaverage program costs for two students (ortwo categories of students) are identicalbut the level of tuition charged to eachdiffers.

Price Patterns and Institutional Goals

The impact of each component ofpricing strategy operates within thecontext of each potential student's de-wand for education. Average price datacan provide some insight into the effectsthat differentials will have on studentresponses, especially when the data areviewed over time. The most that can besaid, however, is thatother things beingequalthe group of students receiving thelower price is encouraged and the groupreceiving the higher price is discouraged.For example, a college might find a

change in overall student demand when itincreases or reduces tuition across theboard.3' A college might also find thatstudents with specific characteristicsresidents of the state, graduate students,minority students, students seeking toenroll in specific programs, academicallyoutstanding students, and so onaredrawn to the university when they arecharged different average prices. Still otherstudents may be charged a higher pricedue to their greater "willingness to pay."

Nevertheless, more information isnecessary to evaluate student responsesfully. In addition, average data cannotpredict the probable success in whichvarious categories of institutions employdifferential pricing patterns to achievetheir objectives. These objectives range

from enhancing tuition and fee revenue,

Differential prices

can be viewed

essentially as

charging varying

average prices

for the same

educational

product to distinct

groups of students,

based on specific

characteristics

a form of price

scrimination.

A college might

find that

students with

specific

characteristics

are drawn

to the

university

when they are

charged different

average prices.

Figure 1: Illustration ofDiffering Tuitions with

Identical Program Costs

$7,000

$6,000

$5,000

$4,000

$3,000

$2,000

$1,000

$0

Student 1 Student 2

a Average subsidy n Average tuition

to attracting more (or fewer) out-of-statestudents, to maximizing prestige byattracting top students, to achievingdiversity goals by enrolling minoritystudents. If there is a conscious policy onthe part of an institution to encourage ordiscourage enrollment of certain categoriesof students, then evaluation of the effec-tiveness of the policy would require threeadditional pieces of information: (1) howmuch and -in- what manner each- individualin the specified category is charged; (2)what the student's alternative choices are; .

and (3) how sensitive that potentialstudent is to price differentials. Eventhough the pricing strategy of any institu-tion may address each element of potentialstudents' subjective demand, in the finalanalysis the matriculation decision is up tothe student.

Data on Pricing PatternsTuition and fees play an increasingly

important role in the overall revenuestructure of most institutions. This relativeimportance has directed attention towardthe way in which pricing strategies actuallydiffer among institutions.

BEST COPY AVAILABre

Sources of Revenue

Higher education institutions' sourcesof revenue include the following:32

aTuition and fees: charges assessedagainst students for education pur-poses, including tuition and feeremissions or exemptions as well astuition and fee revenue that is remit-ted to the state."

27

Government appropriations (fed-eral /state /local): all revenue received

through acts of a legislative body thatis for meeting current operating ex-penses, not for specific projects orprograms.

Government grants and contracts(federal/state/local): all revenue re-ceived from government agenciesthat is for specific research projectsor other types of programs.

Private gifts, grants, and contracts:all revenue received from private do-nors for which no legal considerationis involved, and private contracts forspecific goods and services providedto the funder in exchange for thefunds."

Endowment income: total incomefrom endowment and similarfunds."

Other sources of revenue, including:

sales and services of educationalactivities incidental to the conduct ofinstruction, research, or public ser-

vice; revenues generated by auxiliaryenterprises such as residence hallsand college stores; income receivedor generated by hospitals operatedby the institution; revenues associ-ated with independent operations;and other sources of revenue not cov-ered elsewhere. 3

Defining Categories of Revenue and

Pricing Structures

Higher education institutions receiverevenue from various sources for theircurrent operations, and revenue structuresdiffer considerably among institutions.(See Figure 8.) For example, private, non-profit institutions as a whole received a

larger proportion of their total revenuefrom tuition and fees-41 percent,compared with 18 percent for publicinstitutions. Public institutions, on theother hand, receive a substantial propor,tion of their total revenue from state

appropriations-32 percent, comparedwith less than one percent for private,

non-profit institutions (NCES, 1996).In addition, reliance on certain revenue

sources has shifted over time. Both publicand private institutions have seen a

reduction in the relative importance ofpublic revenue sources and an increase in

100

90

80

70

60

50

40

30

20

10

0

the importance of tuition revenue. In1980-81, for example, taxpayer revenuefrom all sources accounted for 63 percentof total revenue in public institutions, butfell to 51 percent by 1994-95; over thesame period, tuition and fee revenue as ashare of total revenue rose from 13percent to 18 percent (The Institute forHigher Education Policy, 1999). This shifthas been reflected in the rapid rise inaverage tuition levels in recent years.

As institutions' reliance on tuitionrevenue relative to other sources of

revenue has increased, the methods andpolicies used to set tuition levels have

received increasing attention. The differ-ences in tuition and fees charged to

various groups of students are essential to: determining -what proportion of instruc-

t tional costs is being paid by students andtheir families, as opposed to publicsupport.

Figure 8: Distribution of Sources of Revenue,by Institutional Control, 1995-96

(3,418 institutions)

./.:07.....,<,/..%.. +:::,......4 4..4+.....X.X+

.1.1...-;.; ///:. / <

. : : : : : : . : ; ' ;: :

M11111111I

11111

....

AllPublic

AllPrivate, non-profit

Note: Includes only those institutions that responded to every data field. Public andprivate, non-profit categories include all Carnegie Classifications.

Source: NCES, 1996

Independent operations

Other sources

o Hospitals

Auxiliary enterprises

I Sales and services

ErEndowment income

Private gifts, grants andcontracts

b Grants and contracts

Appropriations

Tuition and fees

Tuition and fees

play an

increasingly

important role

in the

overall revenue

structure

of most

-institutions;

usAGroup Foundation'

2 8

Both public and

private institutions

have seen a

reduction in the

relative

importance

of public revenue

sources and an

increase in the

importance of

tuition revenue.

Expected Differentials in

Tuition and Fee Levels

Pricing structures vary among states andinstitutions and there are many explicitrationales used to justify differential prices.

For example, some argue that a portion ofthe higher costs of certain programs shouldbe passed on to students. Furthermore, insome fieldssuch as first-professionalprogramsfuture earnings are likely to behigh enough to allow students to pay ahigher proportion of their costs. Finally,when programs face high enrollmentdemand, higher tuition rates can signalquality and selectivity (Yanikoski and

Wilson, 1984). Some of the possible typesof tuition differentials follow:

By residency status. The most commondifferential, made in virtually all publicinstitutions, is higher tuition levels forstudents who are not residents of the state.In fact, this differentiation is written intostate guidelines in 34 states. Non-residenttuition rates may be two to three timeshigher than in-state rates. In many cases,non-resident tuition rates are set at a levelequivalent to some calculated measure offull costs, or a higher proportion of thosecosts than is paid by resident students(Lenth, 1993).

---By-number-ofcredit hours.-In-someinstitutions, tuition is charged on a percredit hour basis up to a maximum cutoffof credit hours. Other institutions, on theother hand, may have a flat charge perstudent regardless of hours enrolled.Tuition structures also could consist of amixture of these options (Troutt, 1983).

By discipline or field. Some institutionscharge higher tuition rates for specializedhigh-cost programs or programs that are inhigh demand. However, it appears that thetechnical complexity involved in settingdifferentiated tuition levels by disciplinemay limit the use of this strategy (Lenth,1993).

By level of instruction (lower division,upper division, and graduate). Many

institutions charge higher rates of tuitionto postbaccalaureate students, and somecharge slightly higher tuitions to upperdivision undergraduates compared withlower division students. These may beloosely based on the recognition that thecosts of educating upper division andgraduate students tend to be higher thanthe costs of educating lower division

students; however, because many institu-tions do not separate out the costs ofhigher level programs, the tuition differen-tial may just result from applying someincidental fees or an incremental amountto graduate programs. "Only in profes-sional fields...do tuition rates approach theactual per student costs of the program"(Lenth, 1993, p. 19).

By institutional type/mission (CarnegieClassification). Tuition differentials alsomay exist across types of institutions. Thedifferences may be part of an explicit

policy in public systemstuition levels atresearch universities are generally set

higher than rates at comprehensiveinstitutions, and rates at communitycolleges tend to be set at considerablylower levels, likely reflecting differences in

the public provision of subsidies. However,the differentials also tend to exist acrossprivate; now-profit institutions. Thereasons for these differences are not clear,but could range from lower costs due tothe less expensive equipment and facilities(and, possibly, faculty) that are required atcomprehensive or community collegeinstitutions, to reaching greater economiesof scale or other factors.

By institutional control (public/private).Because most public institutions areheavily subsidized with state taxpayer

resources, they generally are able and

inclined to offer lower rates of tuition thancan private, non-profit institutions. Thistype of tuition differential is due todiffering revenue structures, rather thanspecific policies implemented by institu-tions or states.

BEST COPY AVAILABLE

National Averages on Pricing

Data from IPEDS can illustrate howtuition rates are differentiated on averagefor all U.S. institutions. The IPEDSInstitutional Characteristics survey collectsannual information from institutionsregarding tuition and required fees forundergraduates, graduate students, andfirst-professional students (NCES, 1997).Institutions are asked to report averagetuition and required fees for full-timestudents for the full academic year."Tuition and fee levels also must be re-ported for in-state and out-of-statestudents at each level of instruction. Usingthe typical number of credit hours takenby full-time, full-year students by level ofinstruction, the average tuition and fees

per credit hour can be calculated for eachlevel of instruction, for residency status,for institutional control, and for CarnegieClassification.37

The tuition and fees per credit hourcalculations for 1996-97 reveal tuitiondifferentials, on average, for severaldifferent categories:

By residency status. For public

institutions, out-of-state students paidconsiderably higher tuitions onaverageby two to three timesthan

did in-state students, at every level ofinstruction (see Figure 9).

By level of instruction (undergraduateand graduate). For most types ofinstitutions, graduate students paidhigher levels of tuition on averagethan did undergraduates. However,the ratio of graduate to undergraduatetuition varied widely by institutionaltype and control: it tended to behighest at research universities andlowest at baccalaureate institutions,and higher at public institutions thanat private, non-profit institutions (seeFigure 10). In fact, at private, non-profit comprehensive and baccalaureate

institutions, there was virtually no

difference in the average tuition perhour paid by undergraduate

and graduate students. Althoughtuition and fee rates vary widely forspecific programs, on average first-

professional students paid more percredit hour than did both graduatestudents and undergraduates."

By institutional type/mission

(Carnegie Classification). On average,tuition per credit hour tended to behighest for research universities and

Figure 9: Ratio of Out-of-State to In-State Tuition per Credit Hour at PublicInstitutions, by Carnegie Classification and Level of Instruction, 1996-97

(1,433 institutions)

5 3.50

3.0

2.5c2

2.0

.G0 1.5

p 1.09

0.500

4 0.0Research Doctoral Comprehensive Baccalaureate Associate of Arts

1:1 Undergraduate tuition and fees C Graduate tuition and fees

Note: Each bar reoiesents the ratio of out-of-slate to rn -state tuition per credit hour Dy level of instruction at me given institution.Source: NCES. 1997

BEST COPY AVAILABLE:. a .3 0

Because most

public institutions

are heavily

subsidized with

state taxpayer

resources,

they generally are

able and inclined

to offer lower rates

of tuition than can

private, non-profit

Although tuition

and fee rates vary

widely for specific

programs, on

average first-

professional

students paid more

per credit hour

than did both

irate

students and

lowest for associate of arts colleges (see

Figure 11). This was true for bothpublic and private, non-profit institu-tions, undergraduates and graduatestudents, and in-state and out-of-statestudents. (Baccalaureate institutionshad the highest rate for in-stateundergraduates when public andprivate were combined, largely becausethere are far more private, non-profitsin this category than publics.)

By institutional control (public/private). For most subsets of students,the average tuition per credit hour was

higher at private, non-profit institu-tions than at public institutions (seeFigure 12). In fact, tuition rates forstudents at private, non-profit institu-tions were up to four times the averagerates paid by in-state students at publicinstitutions. However, out-of-stategraduate students paid similar levels oftuition on average at the two types ofinstitutions.

Other tuition differentials are notreflected in national data, but may exist atsome institutions (see following page).

By level of instruction (lower division

and upper division). Some institutionscharge higher tuition and fees forupper division undergraduates thanfor lower division students. Forexample, for the fall 1998 semesterMichigan State University charged$144 per credit hour to in-statefreshmen and sophomores, but $160per credit hour to in-state juniors andseniors (MSU, 1999).

By number of credit hours. Some

institutions charge higher rates oftuition for students enrolled withmore credit hours, or for studentsenrolled full-time instead of part-time.For example, in 1998-99 DePaul

University charged undergraduates inits Liberal Arts and Sciences school

$285 per credit hour if they wereenrolled for 11 credit hours or less,but $294 per credit hour if they wereenrolled with 12 or more credithours" (DePaul University, 1999).

By discipline or field. Some institu-

tions charge different rates of tuitiondepending on the particular program

Figure-10:-Watio-Of Giatratetd-Undeigia-difite-Tiiition per Credit Hour,by Residency and Institutional Control, 1996-97

(4,560 institutions)0

1.8

er)

c 1.40

a)

01.0

rn

)

0

0rn

00co

cc

1.6

1.2

0.8

0.6

0.4

0.2

0.0

All public

U In-state

All private, non-profit

Out-of-state

Note: Public and private, non-profit categories include all Carnegie Classifications. Each bar representsthe ratio of graduate to undergraduate tuition per credit hour by residency.Source: NCES, 1997

31 BEST COPY AVATEAlum.

Figure 11: Tuition and Fees per Carnegie Classification, 1996-97(2,654 institutions)

$600

$500a) 0

$400'651,

8 $300

$200

$100

$0In-state

undergraduateOut-of-state

undergraduateIn-state graduate Out-of-state

graduate

OResearch Doctoral /Comprehensive / Baccalaureate SI Associate of Arts colleges

Note: Baccalaureate and Associate of Arts colleges categories include some private, for-profit institutions.Source: NCES, 1997

Figure 12: Tuition and Fees by Institutional Control, 1996-97(4,560 institutions)

In-state Out-of-state In-state

U Public a Private, non-profit

Note: Public and private, non-profit categories include all Carnegie Classifications.Source: NCES, 1997

or school in which the student isenrolled. For example, at the Univer-sity of Illinois at Chicago, resident

undergraduate engineering studentswere charged $2,379 in tuition andfees for the 1997-98 fall and springsemesters, compared with the $2,179charged to other resident undergradu-ates (UIC, 1999)."

As the data demonstrate, tuition ratescan differ by some of the same characteris-tics that reflect varying instructional costs.Within institutions, price differences byresidency statuswhich are only appli-cable for public institutionsmay be the

Out-of-state

widest. Although nationally representativedata cannot support the assertion, pricingstrategies appear to be relatively flatwithin a certain level of instruction, withlittle difference by discipline. There are

clear differences in average prices amonginstitutional types, on the other hand,supporting the assertion that the missionand control of an institution are impor-tant determinants of average prices.

Tuition

differentials

vary not only

nationally but also

by individual

*tutions.

ea

v.4;4.,..1' Vol!

I I S 1. S

EST COPY AVAII;OLE 32

Even in private

enterprise the

actual relationship

between

cost and price

is an obscure one.

30

usAGroup Foundation

The RelationshipBetween Priceand Cost

n 1973, the CarnegieCommission on Higher

Education recommend-ed a reevaluation of tuitionpolicy to gear it more to theactual costs of education.

Specifically, the Commission favored

allowing gross public tuition levels to riseto about one-third of educational costs.The stated reasoning was to bring publictuition more in line with private tuitions;to provide greater equity between studentswith financial means and those without(with increases in grants and scholarships

to low-income students); and to anticipate

rising rates of tuition in line with increasesin real incomes. Since then, otherpolicymakers on many levels have sug-

gested proportional goals for therelationship between- costs-and-prices.

Even in private enterprise, however, the

actual relationship between cost and priceis an obscure one. On the one hand, pricesalways "reflect" costs. Even the most

flagrant monopolist charges a price that iscognizant of cost, but that usually exceedsaverage cost. But, on the other hand, thecausal relationship depends upon thenature of the industry in which anyparticular producer operates and theunique conditions facing that producer. Inthe case of higher education, tuition is onlyone element of revenue, and colleges canuse non-tuition resources such as endow-ments to "unlink expenditures and pricesin the short run" (McPherson and Win-ston, 1993, p. 75). The availability of

non-tuition revenue suggests there is evenless reason to expect a consistent relation-ship between price and cost across allinstitutions or groups of institutionseven though any one institution couldexhibit a consistent relationship.

The Expected Relationship

The relationship between the price ofany product and the costs of producing itis not a simple one.4' Given certain

production functions and demands for aproduct or service, the producer decideshow much to produce, and consumersdecide how much they are willing to payfor that level of output. The price may beequal to average cost or some multiple ofaverage cost, depending upon the charac-teristics of production and demand, thetime period over which the examination ofcosts and revenues is conductedand thelegal environment in which the enterpriseexists. In the case of higher education, eventhese relationships are obscured.

As already noted, colleges and univer-sities have multiple sources of revenuein addition to tuition and fees. Thisfact allows costs to exceed the tuitionand fee component of revenue, and

-permits institutions to distribute nontuition resources in a way that furthererodes any relationship between pricesand costs. The difference between

average tuition and fees and theaverage educational cost is the "sub-

-sidy."42 Public institutions are

subsidized primarily through statetaxpayer funds, while subsidies atprivate institutions are based onvoluntary gifts as well as governmentfunding.

Higher education institutions are notusually seeking a profit, per se, but areseeking to optimize some alternativeset of goalsfor example, reputation,

BEST COPY MAILABLE

the quality of the institution, the sizeof the institution, or political objectives.

These considerations suggest that costsdo not impact prices to a predictableextent in cross-sectional data. In otherwords, the specific characteristics of higher

education appear to preclude a cost/pricerelationship that can be applied to allinstitutions, or even to all institutions witha certain mission. If one institution setstuition as a constant proportion of appro-priations, and another indexes tuitionaccording to the consumer price index,and still another feels compelled to keeptuitions low by increasing other sources ofrevenue, it is unlikely that an overallcorrelation between prices and costs couldbe found.

Bowen (1980) examined the relation-ship from the other direction, by observingthat the amount of available revenueand, by extension, pricesmay affect costsper student, given a certain level ofenrollment:

Within wide limits, institutions canadjust to whatever amount of moneythey are able to raise. When resourcesare increased, they find uses for thenew funds, and unit costs go up.When resources are decreased, theyexpress keen regret and they protest,but in the end they accept the inevi-table, and unit costs go down. This setof generalizations might be called therevenue theory of cost... (p. 15).

Yet Bowen recognizes this "revenue

theory of cost" as a short-run possibility. Inthe longer run, colleges are forced to raisesufficient funds to meet fixed operatingexpenses, and are compelled to adjustvariable costs so as to compete with otherinstitutions. At the level of an individualinstitution, the relationship between costsand prices cannot be completely random,

even though cross-sectional data maymake it appear that way. With enoughinformation about an institutionincluding the nature of demand, themultiple sources of revenue, the produc-tion function, and other factorsonecould discover a relatively consistentrelationship between costs and prices atthat institution. Such a relationship wouldprobably be somewhat fixed over time,barring a sudden shift in legislativepolicy."

The Institution-Level Identity

With only average cost and price data, acausal relationship between cost andpricing structures across all or part of thehigher education sector cannot be de-scribed. Nevertheless, a loose associationbetween average costs and average pricescan be expressed for an institution orgroup of institutions, given certainassumptions. Non-tuition sources ofrevenue are limited, even if different limitsapply to different institutions." Thus, thelevel of non-tuition revenue restricts theextent to which costs can exceed prices;institutions reaching the outer bound mayneed to raise tuition levels to accommo-date increasing costs. At the same time,other objectives cannot be pursuedwithout consciousness of the "bottomline"even in non-profit organizations.

In other words, if an institution's totalcurrent revenues are equal to its totalcurrent expenditures" and currentrevenues are derived from both tuitionand other sources, then average costs areequal to the sum of average tuitionrevenue and average non-tuition revenue.The average non-tuition revenue can bethought of as the average subsidy at thatinstitution. This holds for an institutionor a group of institutions in total, but notfor one component of an institution, suchas an undergraduate program alone.

The specific

characteristics of

higher education

appear to

preclude a

cost/price

relationship that

can be applied to

-all institutions,

or even to all

institutions with a

certain mission. !!

1: EST COPY AVAIPA§LE

Effective analysis of

financial data

should reveal the

amount of revenue

needed to

supplement tuition

and fees at any one

institution or type-

of institution.

Average Costs per Student Unit[Average Tuition Revenue per

Student Unit] + [Average Non-Tuition

Revenue per Student Unit]

Examining average cost and tuition datacan address an important question: Whatis the amount of revenue needed to

supplement tuition and fees at any oneinstitution or type of institution? Inaddition, the analysis can be taken furtherto look at the pattern of intra-institutionalsubsidization given a specific level of non-tuition revenue.

Non-Tuition Income and

Infra-Institutional SubsidizationWithin each program, course, or

discipline, total costs do not necessarilyequal total revenues, due to transfers ofrevenue within institutions. These trans-fers lead to "intra-institutional" subsidies

that vary among programs, disciplines,and courses; such subsidies can be thoughtof as the amount of non-tuition revenuethat has been allocated to a particularprogram, discipline, or course. From thisperspective, the amount of subsidy perstudent is determined by the average cost

per-student of-that programidiscipline;-or-course compared with the average amountof tuition and mandatory fees paid perstudent." In other words, if the averagetotal cost of an undergraduate program is$5,000 and the average price paid byundergraduates is $2,000, then the subsidyfor that program is $3,000. By definition,

therefore, an intra-institutional subsidy iscovered by some amount of non-tuitionrevenue.

If one student's total cost (for a pro-gram, discipline, or course) is matched bymore non-tuition revenue than some otherstudent's total cost, then the first studenthas been subsidized to a greater extentthan has the other student. One instanceof this pattern is illustrated in Figure 13,

NJZc

where average tuition is identical for twostudents (or two categories of students) butaverage program costsand thereforeaverage subsidiesdiffer.

Such a pattern is often referred to as a"cross- subsidy," although the term is

generally inappropriate in the case-ofhigher education. "Cross-subsidy" impliesthat the price charged to an identifiablesubset of the student body exceeds theaverage costs of providing education tothat subset, and that the "excess" revenue isused to pay a portion of the average costsof educating some other identifiable subsetof students. At most postsecondary

institutions, however, prices do not exceedaverage costs for any subset of studentsall students are subsidized to some degree.Only in situations in which average tuitionand fees make up a very high proportionof average total revenue is true "cross-

subsidy" possible. Otherwise, patterns suchas the one described are probably cases ofdifferential "intra-institutional" subsidiza-tion, where a greater subsidy goes to onesubset of students than to another.

Figure 13: Illustration ofProgram Costs with Identical

Tuition Levels

$7,000

$6,000

$5,000

$4,000

$3,000

$2,000

$1,000

$0

Student 1 Student 2

U Average subsidy U Average tuition

35

Average data (average cost of a program,

discipline, or course compared withaverage tuition paid for a program,discipline, or course) can be used toexamine the size of intra-institutionalsubsidies. Differential patterns betweentypes of institutions can be examinedwith cautionas well.

Comparisons between Price and Cost

Descriptive analyses of average price

and cost data generally include thepresentation of price-to-cost ratios, whichreframe both types of subsidies in terms ofthe proportion of the costs of educationthat is met by tuition, on average, forvarious subgroups of students. For ex-

ample, one might compare the proportionof educational costs paid by in-statestudents to the proportion paid by out-of-state studentsespecially if it is a policygoal for the latter group to pay a higherproportion of costs than the former group.

In fact, comparisons between averagecosts and prices for groups of institutionsremain both common and relevant, largelybecause linkages between the two are

made at the public policy level. As dis-cussed above, many colleges and statelegislatures have instituted policies thateither directly or indirectly link price-andcost (Lenth, 1993).47 Average tuitionoverall may be mandated as a specificproportion of average costs," or tuitionrates for certain subgroups of students may

be set in relation to instructional costs.This can mean that tuition differentials arebased on specific percentages of average

costs, or that differentials are loosely scaled

according to cost differences. In this case,decisions are made as to which groups ofstudents should pay higher proportions ofthe cost of educating them, and whichgroups should pay relatively lower propor-

tions. From the opposite perspective, somestudents will receive relatively highersubsidies because they are paying a lower

proportion of their costs.

11"4_,ST COPY AVAILABIA

Figure 14: Illustration ofDiffering Program Costs and

Differing Tuition Levels

$7,000

$6,000

$5.000

$4,000

$.3.000

$2,000

$1,000

$0

Student 1 Student 2

Average subsidy U Average tuition

Price-to-cost ratios can be examined in

at least three ways: as benchmark data(categorized by institutional characteristics,such as Carnegie Classification); as intra-institutional comparisons (within aninstitution, such as between programs ordisciplines); and as inter-institutionalcomparisons (between institutions). Thechosen method depends on the purposefor which the comparison is being made.However, price-to-cost ratios are limited in

the extent of the story they tell. The sameprice-to-cost ratios may mask differentabsolute levels of prices, costs, and subsi-

dies, which also have implications forpolicy. For example, in Figure 14, average

tuition amounts for student 1 and student2 make up the same proportion of respec-tive average costs, yet total tuition, totalcost, and total subsidy are vastly different.Price-to-cost ratios, therefore, tell only a

portion of the total story.

Relating the Data on Average

Costs and PricesIt is clear that establishing a relationship

between cost and price across institutions,or finding a link between pricing struc-

36

Comparisons

between average

costs and prices for

groups of

institutions remain

both common

and relevant,

largely because

linkages between

the two are made

at the ublic

Average data are

sufficient to

calculate the

average subsidy

provided each

type of student,

differences in

average subsidy

between types of

institutions, and

changes in subsidy

over time.

usAGroup Foundation'

tures and institutional objectives, requiresinformation beyond average data. Yet

average prices and costs tend to vary insomewhat predictable ways. In addition,average costs and prices for each categoryof students can be compared in terms ofprice-to-cost ratios, towardevaluating theirapparent agreement with certain policygoals. Finally, average data are sufficient tocalculate the average subsidy provided eachtype of student, differences in averagesubsidy between types of institutions, andchanges in subsidy over time.

Cost and Price Patterns

The U.S. General Accounting Office(GAO) (1998) attempted to answer thequestion of whether there was a relation-ship between cost and tuition byexamining IPEDS data via regressionanalysisincluding the rate of increase intuition and required fees for full-time, full-year undergraduates over a five-year period

(1989-90 to 1994-95), with 15 differentfactors that might account for thechanges." Results included the following:

In the case of public four-year institu-tions, the change in governmentappropriations and the change ininstructional- expenditures- were-the

two most important factors in explain-ing the variation in tuition increases.

In the case of private four-year

institutions, the most important factorin explaining the variation in tuitionincreases was the change in revenue

from grants, contracts and othersources.

Thus, changes in other sources of revenueappear to be correlated with tuitionincreases in addition to changes in costs.As the study concludes: "Costs may beonly one of several factors considered when

tuition-setting decisions are made" (GAO,1998, p. 3). This finding supports the

assertion that the cost/price relationship inhigher education is complicated. At thesame time, however, the GAO's analysisfocused primarily on overall tuitionaverages, which do not reflect the differen-tial prices charged to groups or levels ofstudents within certain types of institu-tions, and the study's cost data did notreflect differences in instructional costs bydiscipline or by level of instruction.

The data presented in this report revealthat price and cost patterns vary insomewhat predictable ways. For example,data from the Delaware Study and theArkansas Department of Higher Educationsuggest specific differences in average costs:

By level of instruction (lower division,

upper division, and graduate), witheach level having successively highercosts;

By discipline or field, with suchdisciplines as engineering, healthprofessions, and multi/interdiscipli-nary studies having relatively highercosts and English and literature,psychology, and protective serviceshaving relatively lower costs; and

By institutional type and control, withresearch universities having relativelyhigher costs and comprehensive

universities having lower costs, andprivate institutions having relativelyhigher costs than public institutions.

At the same time, national aggregatedata (primarily from IPEDS) indicatetrends in average price differentials:

By residency status, with out-of-statestudents generally paying higher ratesof tuition at public institutions;

By level of instruction, with graduate

and first-professional students paying

BEST COPY AVAILABLE 37

slightly higher tuition per credit hourthan undergraduates; and

By institutional type and control, withprivate, non-profit institutionscharging higher prices than publicinstitutions, and research universitiescharging higher prices than commu-nity colleges.

Nevertheless, these price differentials do

not appear to reflect cost differencesexactly, even in the cases in which costsand prices vary in the same direction.Holding all else equal, in some cases,pricing patterns appear to differ in thesame direction as instructional costpatternsfirst-professional student costsand prices-are-higher than those of under-graduates, and costs and prices both tendto be higher at research universities than atcomprehensive institutions. However, thedifferences often do not appear to besignificant to the same extent. At the sametime, differences-do not appear to be in thesame direction in other cases. For example,where public tuition rates seem to differconsiderablythat is, by residencystatusthey are not necessarily reflectingcost differences, but rather the decisions ofstates to make non-residents bear a largershare of the costs of educating them.Furthermore, although instructional costsdiffer widely by discipline, it appears thatmost institutions have not chosen todifferentiate their tuition by discipline,preferring instead to allow students to buyinto a broader education "package."

Meanwhile, the aforementioned formsof differential pricing may be superim-posed by a pattern of net pricesthroughscholarships and grants for particularlymeritorious or minority scholarsthat isfrequently decided on a case-by-case basis.Because the pattern of price discounting isso complex, it is difficult to know how itaffects any analysis of price patterns andtheir relationship to costs.

BEST COPY AVM A 1r

Price-to-Cost Ratios

As theory predicts, it appears that tosome extent the decision to differentiatetuition rates rests on conscious choices bystates and institutions regarding whichgroups of students should pay more (andwould be willing to pay more) of theireducational costs, rather than reflectingdifferences in instructional costs. Becausetuition levels do not vary exactly withinstructional costs, different students paydifferent proportions of those costs. Theimportant question here is who is bearingwhat share of educational costs. In otherwords, what proportion of the costs ofeducation is met by tuition, on average, forvarious subgroups of students?

Although the Delaware Study representsa group of specific institutions andtherefore is not generally applicable,instructional cost data for these institu-tions can be matched with tuition datafrom IPEDS to show average tuition ratesas a proportion of average instructionalcosts for specific subsets of students." Indoing so, certain characteristics areexamined holding all the others constant;one "slice" of the data is consideredwithout taking into account the manner inwhich other "slices" cut across it. Forexample, average price-to-cost ratios byresidency status can be compared byassuming that instructional costs are thesame for both in-state and out-of-statestudents. In reality, those costs may differbetween the two groups of students if in-state students are more likely to attend acertain type of institution or take coursesin specific disciplines. Keeping in mindthis complexity, price-to-cost ratios cannevertheless point to broad trends.5'

Briefly, tuition-to-cost ratios for theDelaware Study group of institutionsreveal the following, on average:

Students at research universities tendto pay a lower proportion of theirinstructional costs than do their

0.

Although

instructional costs

differ widely by

discipline, it

appears that most

institutions have

not chosen to

differentiate their

tuition by

discipline,

preferring instead

to allow students Ito buy into a

broader education-cc __

What proportion

of the

costs of

education

is met by tuition,

on average, for

various subgroups

counterparts at other types of institu-tions. Students at baccalaureate

institutions appear to pay the highestproportion; however, this may berelated to the fact that the institutionsin this group are mostly private, non-profits, whereas the other groups are

comprised mainly of public institu-tions (see Figure 15).

At public institutions, out-of-statestudents appear to pay a higherproportion of their instructional coststhan in-state students, at both the

Figure 15: Comparisons of Price to Cost, 1996-97Examples from the Delaware Study Group of 153 Institutions

Holding all else equal:

Average

instructional

cost per

credit hour

Average

tuition per

credit hour

Subset

By Carnegie Classification (public and private institutions): See Note 1

Price-to-

Cost

Ratio

Research universities $198 $338 171%Doctoral universities $170 $349 205%__Comprehensive institutions $129 $308 239%Baccalaureate institutions $163 $608 373%By residency status (public institutions only): See Note 2

Undergraduate tuition, in-state $198 $103 52%Undergraduate tuition, out-of-state $198 $299 151%Graduate tuition, in-state $198 $159 80%Graduate tuition, out-of-state $198 $403 204%

By level of instruction (public and private institutions): See Note 3

Undergraduate tuition, in-state $198 $170 86%Undergraduate tuition, out-of-state $198 $237 171%Graduate tuition, in-state $198 $338 120%Graduate tuition, out-of-state $198 1449- 227%By discipline (public and private institutions): See Note 4English and literature $104 $220 212%Computer science $153 $220 144%Engineering $317 $220 69%

By discipline within Carnegie Classification (public and private institutions): See Note 5Research universities, English and literature $110 $170 155%Research universities, computer science $173 $170 98%Research universities, engineering $333 $170 51%

1) Using out-of-state undergraduate tuition for all categories, to eliminate the effects of discounts for state residents and lessen thedifferences by institutional control.

2) Using average costs for research universities, which cannot be disaggregated by level of instruction or by control (although 85percent of research universities in this group are public). Assumes costs do not differ by residency status. Using tuition for publicresearch institutions only.

3) Using average costs for research universities, which cannot be disaggregated by level of instruction, and using tuition for allresearch institutions. Assumes costs do not differ by residency status.4) Using average costs by discipline for all institutions, and using undergraduate in-state tuition for all institutions.Assumes tuition does not differ by discipline.5) Using average costs by discipline for research universities alone, and using undergraduate in-state tuition for allresearch universities. Assumes tuition does not differ by discipline.

Note: Many of these examples are for research universities, but the trends also hold for other CarnegieClassifications in almost all cases.

Source: Delaware Study benchmarks, 1996-97 (for instructional costs); NCES, 1997 (for tuition and fees)

BEST COPY AVAILAFtr.P. 39

undergraduate and graduate level(assuming instructional costs do notdiffer between the two groups ofstudents).

When instructional costs are assumedto be the same for the two groups,graduate students appear to pay aslightly higher proportion of theirinstructional costs than do under-graduate students. However, otherevidence has indicated that instruc-tional costs do differ by level ofinstruction. As the Delaware Studybenchmarks cannot be disaggregatedby student level, it is impossible tomore accurately compare the propor-tions paid by graduate students versusundergraduates using these data.

When tuition is the same, students inlow-cost disciplines pay a higherproportion of their instructional coststhan students in higher-cost disciplines.

In addition, Winston (1997) usedIPEDS data from 1994-95 to calculate

average cost and price data by control andCarnegie Classification.52 These data canbe used to determine average price-to-costratios by type of institution, recognizingthat they do not take into account priceand cost differentials within each institu-tional type. The following tendencies canbe drawn from these data:

Students at private institutions tend topay a higher proportion of theireducational costs than do students atpublic institutions. Average tuitionand fees covered 69 percent of averagecosts at private institutions, but only23 percent at public institutions (seeFigure 16).

At most public institutions, studentsat schools with different missions tendto pay similar proportions of theireducational costs, ranging from 29 to34 percent. An exception- is studentsat community colleges, who pay alower proportion of their costs-20percent on averagethan do studentsat public universities and other four-

Figure 16: Sticker Price as Percent of Educational Spendingper FTE Student, 1994-95

(2,739 degree-granting institutions)

Public Institutions

Research

Doctoral

Comprehensive

Liberal Arts (Baccalaureate)

Two-Year

Private Institutions

Research

Doctoral

Comprehensive

Liberal Arts (Baccalaureate)

Two-Year

0% 10% 20% 30% 40% 50% 60% 70% 80% 90%

23%

34%

69%

..

34%

9%

30%

20%

53%

77

%

81%

6

74%

Note: Measures full costs. not just direct costs.Source: Adapted from Winston, 1997

ki,' 4;y 40

Students at

baccalaureate

institutions

appear to pay the

highest proportion

of their

instructional costs.

Students at

community

colleges pay a

lower

proportion

of their costs-

20 percent on

ave do

students at public

universities and

9

year colleges. At private institutions,

the proportions vary more widely, withtuition and fees covering 53 percent ofcosts on average at research universitiesbut 81 percent at comprehensiveinstitutions.

In general, these trends hold for netprice-to-cost ratios as well. However,average net prices are particularly difficultto draw conclusions from regarding theeffects on categories of students, because

_students within a particular subgroup donot pay "average" net prices (Jenny, 1997).

Average Subsidies

The flip side of price-to-cost ratios isthat the proportion of costs paid by non-tuition revenueincluding taxpayermoneyalso differs among groups ofstudents. Average subsidies cannot becalculated for the Delaware Study group,due to the need to use distinct data sets forcost and price data and the lack of data onindirect costs. Thus, it is difficult to get asense of how average subsidy differs by

level of instruction or by discipline (inother words, intra-institutional subsidies)

without more detailed analysis. However,Winston (1997) used the IPEDS data onaverage costs and prices to calculate

average subsidy by control and CarnegieClassification. Although the data cannotprovide insight into differential pricingstrategies, they do reveal the importance ofsubsidy at every type of institution. Theresults support the expectations mentionedin the previous sections. Holding otherfactors constant:

Subsidy makes up a higher proportionof educational spending for studentsat public institutions-77 percent, onaveragethan at private institutions,31 percent. This is not surprising,given the fact that private institutionstend to rely more on tuition and feerevenue, and less on non tuitionrevenue such as government appro-priations, than do public institutions(see Figure 17).

Similarly, subsidy differs by institu-tional mission. In particular, privateresearch universities appear to have alarger subsidy as a proportion of

Figere1T. General Subsidy as Percent of Educational Spendingper FIE Student, 1994-95

(2,739 degree-granting institutions)

Public Institutions

Research

Doctoral

Comprehensive

Liberal Arts (Baccalaureate)

Two-Year

Private institutions

Research

Doctoral

Comprehensive

Liberal Arts (Baccalaureate)

Two-Year

0% 10% 20% 30% 40% 50% 60% 70% 80% 90%

77%

1 66%

66%

71%

BO%

70%

31%

147%

1

23%

%

26%

32%

Note: Measures full costs, not just direct costs.Source: Adapted from Winston, 1997

BEST Copy AVAILARnz.:= 41

educational costs-47 percentthanother private institutions, such ascomprehensive colleges, 19 percent.Subsidy proportions at public institu-tions are relatively greater, regardless ofmission; students at communitycolleges have the greatest subsidies as a

proportion of educational costs(despite comparatively lower subsidiesin absolute terms).

As expected, calculation of full sub-sidyusing net prices rather than stickerpricesreveals similar trends, with theaverage full subsidy being higher across the

board. Again, calculations involvingaverage net price cannot say much aboutwhich students actually receive the highersubsidy.

Implications

his report provides a

comprehensive

description of the cost andpricing structures associatedwith higher education.

It explains the various factors that deter-mine costs and prices, the differences incosts and prices within and across institu-tions, and the potential for interactionbetween cost and price averages. Much canbe gained through an understanding ofthese aspects of higher education finance,for both internal institutional evaluationand broader public policy purposes,especially in an era of increasing pressure

for financial accountability.

Broadly, several observations about costand price structures can be made:

Alternative sources of revenue subsi-dize virtually all students to someextent.

Because institutions have enormousflexibility in selecting their productioninputs, each institution has a uniqueset of costs and highly differentiatedoutputs.

Differences in program costs withinany single institution imply that somestudents are subsidized more thanothers.

Despite differences in missions andresources across institutions, categoriesof institutions exhibit somewhatconsistent pricing and cost patterns.

Measurement of costs and relatedpricing data can be adapted by aninstitution and used to compare itselfwith other institutions in order toprovide insight into performance andcost effectiveness, while simultaneously

recognizing differences in missions,resources, and other conditions.

Within individual institutions,differential tuitions among differenttypes of students can be used to alterthe composition of the student body.

In drawing these observations, theauthors of this report highlight the vitalrole of institutional and state missions andgoals in the differentiation of costs andprices. An institution's mission influencesthe mixture of products (instruction,research, and public service) it strives toproduce, while the production process itemploys determines its ability to fulfill itsmission. Meanwhile, differing patterns ofprices impact the characteristics of stu-

Future financial

constraints will

demand a

deepening

understanding

of the various facts

that determine

costs and prices in

hider education.

usAGroup Foundation'

BEST COPY AVAIIABLE:it'''. 4.2

Rather than

differentiating

pnces,

an institution

might decide

to shut down

certain

programs

if unit casts are

excessive and

40

usAGroup Foundation

dents who attend the institution and theamount of non-tuition revenue that isrequired to support a range of course

offerings, thereby affecting an institution'sability to conform to its mission andobjectives. Patterns of differential pricingalso may reflect the decisions of state

policymakers about spending priorities, atleast at public institutions.

Given the essential role of institutionaland state objectives, the observationsregarding cost and price structures raisesome fundamental questions:

For an institution or group of institu-tions, are the mission and goals thatare incorporated into cost structuresthe same as those incorporated intopricing patterns, or are they, in fact,operating at cross purposes?

Is the relationship between costs andprices in alignment with institutionalor state goals?

Is that relationship reflected in price-to-cost ratios and subsidy patterns thatare also consistent with an institution'smission or a state's spending priorities?

Is the inequa/ity-of-prices-and-costs

among different groups of studentsinequitable in any sense?

The answers to such questions haveimplications for the evaluation of costefficiency, institutional decisions onpricing, and public policy.

Internal Cost Evaluation

Fundamentally, an understanding of therelationship among costs, revenues, andprices allows administrators to follow theflow of funds within the institution,evaluate the cost efficiency of academicprograms, and measure the institution'sperformance against historical benchmarksor peer institutions. Such self-evaluations

continue to influence the future opera-tions of individual institutions, especiallythose that must choose between raisingprices, cutting costs, or a combination ofboth. Cost accounting techniques may bedifficult and complicated to adopt, butmany institutions clearly feel it is worth-while to learn from the effort involved.

The cost evaluation process differs foreach institution. For example, manyinstitutions have decided it is important tomaintain non-differential prices bydiscipline or programin other words, tomaintain the situation in which a studentpays a certain price regardless of thecourses chosen. In this case, knowledge ofthe courses and programs actually chosenby various students and the costs attachedto those courses is highly relevant forfuture curriculum planning, budgeting,and, if necessary, cost-cutting initiatives.

Rather than differentiating prices, aninstitution might decide to shut downcertain programs if unit costs are excessiveand student demand is low. Regardless ofthe decisions an individual institution maymake as a result of its evaluation, however,

access to these kinds of data is essential ina period of increasing financial pressures.

Pricing- PolicyWith an understanding of the relation-

ship among costs, revenues, and prices andsufficient information to construct price-to-cost ratios, individual institutions alsocan evaluate the alignment of subsidypatterns with their missions and goals. Ifthey find a lack of alignment, they maydecide to adjust the levels and patterns oftuition to conform with their missions andgoals. In addition, institutions may use

price-to-cost ratios in combination withdemand information to maximize tuitionincome or to alter the composition of theirstudent bodies through the differentiationof prices.

BEST COPY AVAILABU 43

Although the individual steps will varyconsiderably for each institution, someexamples can demonstrate this process:

A baccalaureate institution might havethe primary mission of providingundergraduate education. At thisinstitution, undergraduate studentsmay pay a lower proportion of theircostsand therefore receiving aproportionally larger subsidythangraduate students. If cost and pricedata revealed the opposite, institu-tional leaders might decide todifferentiate tuition so that under-graduate students pay less andgraduate students pay more, while stillreceiving the same amount of totalrevenue from tuition.

An institution might have the primarymission of providing access fordisadvantaged populations, anddisadvantaged students may be payinga lower proportion of their costs thanother students. If they are not, theinstitution's net tuition structure mightbe altered to reflect this focus by usinginstitutional aid to reduce the nettuition paid by disadvantaged stu-dents.

An institution might have the goal ofenhancing tuition revenue. To accom-plish this goal, students who arerelatively less price-sensitive may needto pay a higher proportion of theircosts than students who are moreprice-sensitive. Thus, institutionalleaders might decide to segment thetuition structure to charge highertuition and fees to individuals who areless sensitive to price.

Institutions also may analyze theirpricing strategies within the context ofstudent demand. Students whose tuitioncovers a relative lower proportion of their

educational costs than other students mayperceive they are getting a better "value"for their money, and vice versa. Suchperceptions may impact the patterns ofenrollment across (competing) institu-tions." If institutions have enoughinformation on potential enrollmentshifts, they may try to counter them bytargeting a specific ratio between costs andprices or by choosing a pricing strategythat minimizes negative demand effects.

To succeed in the process of evaluatingcost, price, and subsidy patterns, themission of each institution or group ofinstitutions needs to be defined clearly."For one thing, a lack of clarity in missionmakes it difficult to evaluate the confor-mity of subsidy patterns with thatmission. This is especially true in a case of"mission creep." For example, wheninstitutions spend greater amounts ofmoney on research, graduate study, andother prestige-oriented activities in orderto "expand" their missions, they must beaware of.the accompanying shifts in costand subsidy patterns. Clearly definedinstitutional missions also are importantbecause decisions about whether tomanage tuition income or the composi-tion of student bodies through the use ofdifferential pricing strategies depend uponthe mission and goals of the individualinstitution.

State PolicyState policymakers have an interest not

only in decisions made at the institutionallevel, but also in trends in costs, revenues,and prices over entire public highereducation systems. Students at publicinstitutions tend to pay relatively lowerproportions of their costs than students atprivate, non-profit institutions, as a resultof states' direct appropriations to publicinstitutions. In 1995-96, state govern-ments accounted for 36 percent of the$124 billion of the current revenue ofpublic institutions.55 Along with this

BEST COPY AVAIWLE

When institutions

spend greater

amounts of money

on research,

graduate study,

and other

prestige-oriented

activities in

order to

"expand" their

missions,

they must be

aware of the

Some states,

however, have

allowed forms of

``tuition

engineering,

or the individual

manipulation

of cost/price

relationships, at

public institutions.

public investment, however, comes a needfor accountability. State legislators,governing boards, coordinating boards,and other policymakers must ensure thatstate funds are being spent in alignmentwith public priorities. Like institutionalpricing strategies, the objectives of statefunding may differfor example, publicsubsidies may be targeted on undergradu-ate students, on in-state residents, onstudents attending community colleges,on students enrolled in a particularcurriculum, or on some combination ofthese factors.

One method of accountability involvesthe way in which public tuition levels andsubsidies are set. As this report describes,many states have mandated fixed relation-ships of some sort between costs andprices. In doing so, policymakers haveattempted to ensure that subsidy patternsconform to their objectives. Other states,however, have allowed forms of "tuitionengineering," or the individual manipula-tion of cost/price relationships, at publicinstitutions.56 Most state policymakers

likely choose some combination of theseand other techniques.

Regardless of the process used to setprices and subsidies, state policymakerscan examine the-resulting-price:to-costratios for various groups of studentsattending public institutions in order toevaluate whether the distribution ofsubsidies actually reflects their states'

funding goals. Using this type of analysis,policymakers may decide to change thepatterns of the subsidy they provide, bylowering tuition levels for students atinstitutions with specific institutionalmissions or with other characteristics, andraising tuition levels for other groups ofstudents. Alternatively, policymakerscould alter the relative levels of fundingprovided to different types of publicinstitutions.

Future Concerns

To construct price-to-cost ratios andevaluate the alignment of cost, price, andsubsidy patterns with goals and missions,institutional and other leaders must havecomplete, meaningful data that reflectpricing and cost differences across andwithin academic institutions. However, thecollection and use of such data face anumber of challenges that make highereducation different and more idiosyncraticthan private sector cost accountingsystems."

In addition to methodological andtechnical problems, cultural resistance tocost accounting and other data collectionmechanisms continues to exist at theinstitutional level. Concern is expressedthat a comparison of ratios among institu-tions or even among departments will missimportant aspects of higher education thatare uniquely captured by that institutionor department. Many activities that are notnecessarily valued by specific consumersfor example, unfunded researchareessential parts of the higher educationenterprise. As a result, many fear theconsequences of revealing these patternswithout an accompanying understandingof the higher education "package." Theseconcerns should inform thneed -forgreater financial transparence, but do notpreclude such accountability.

It also is important to re-emphasize thecomplicating factor of net prices. From theperspective of the institution, patterns ofinstitutional aid affect the distribution offull subsidies within and among institu-tions. In addition, from the perspective ofthe student, student aid from state, federal,and other sources has an impact on theprice they actually pay. Knowing howstudent aid patterns compare with statedcost and pricing structures ultimately isessential to the formulation of publicpolicy on many levels. The availability ofnet price data would allow policymakers toaddress several issues:

45

In theory, student aid programs areintended to "equalize" the burdensfaced by students. In practice, how-ever, it currently is almost impossibleto determine whether disadvantagedstudents pay a greater or lesser propor-tion of their educational costs thanother students. This issue is compli-cated by the concentration ofdisadvantaged students in certain typesof institutions and programs.

The-cambinaton of greater institu-tional reliance on tuition revenue andgreater student-reliance on federalloans impacts the differentiation ofsubsidies in various ways. It might beargued that grants reduce the propor-tion of educational costs paid bystudents, while loans do not. Increas-ing the amount of federal grants,therefore, would substitute public debtburden (through taxes) for studentdebt burden.

Patterns of net tuition may havegrown so complex as to affect thequality of higher education. From aninstitutional budget point of view,expenditures may be being reallocatedaway from instructional programs andtoward institutional aid. Withoutknowing the interaction between shiftsin instructional costs and patterns ofinstitutional aid, it is difficult toanticipate the potential effects onacademic quality.

Despite these considerations, it is clearthat there is much to be gained by institu-tions themselves, the higher educationcommunity as a whole, publicpolicymakers, and students by collectingcost and price data. An improved under-standing of cost and price structures isonly the starting point. Ultimately, thehigher education community should usethis knowledge to assess the distribution of

BEST COPY AVAIL01[41E

subsidies among various groups of stu-dents and help ensure that the actualworkings of institutional financingstructures contribute toward its statedmissions and goals.

46

An improved

understanding of

cost and price

structures

is only the

starting point

ADHE. See: Arkansas Department'of Higher Education.

Allen, Richard and Paul Brinkman. 1983. Marginal Costing Techniques for Higher Educa-tion. Boulder, CO: National Center for Higher Education Management Systems(NCHEMS).

Arkansas Department of Higher Education (ADHE). 1998. Arkansas Academic CostAccounting: A Strategic Management Tool for Higher Education Planning and Campus

Decision-Making, Executive Summary. Little Rock, AR: Arkansas Department ofHigher Education, January.

Banziger, George, Anne Marie Wagner, and Thomas Watts. 1997. "On the Road toRatios." Business Officer. National Association of College and University BusinessOfficers (NACUBO), November. From the NACUBO website (www.nacubo.org).

Bowen, Howard R. 1980. The Costs of Higher Education: How Much Do Colleges and

Universities Spend per Student and How Much Should They Spend? San Francisco, CA:Jossey-Bass Publishers, Inc.

. 1981. "Cost Differences: The Amazing- Disparity Among Institutions of HigherEducation in Educational Costs per Student." Change. January/February, pp. 21-27.

Brinkman, Paul T. 1989. "Instructional Costs per Student Credit Hour: Differences byLevel of Instruction." Journal of Education Finance. Vol. 15 (Summer), pp. 34-52.

. 1993. "The Cost of Public Higher Education in California: Concepts and Institu-tional Expenditure Data." Unpublished paper prepared for the CaliforniaPostsecondary Education Commission, February.

Carnegie Commission on Higher Education. 1973. Higher Education: Who Pays? WhoBenefits? Who Should Pay? New York, NY: McGraw-Hill, June.

Carnegie Foundation for the Advancement of Teaching. 1994. "A Classification ofInstitutions of Higher Education." From website (www.carnegiefoundation.org/cihe),downloaded April.

Christal, Melodie E. 1997. State Tuition and Fee Policies: 1996-97. Denver, CO: StateHigher Education Executive Officers (SHEEO), March.

DePaul University. 1999. "1998-99 Tuition Rates." From DePaul University website(www.depaul.edu/admission/factfigs/tuition.htm), downloaded March 24.

Dundar, Halil and Darrell R. Lewis. 1995. "Departmental Productivity in AmericanUniversities: Economies of Scale and Scope." Economics of Education Review. Vol. 14,No. 2 (June), pp. 119-144.

4.

Fischer, David. 1996. "Rolling the Dice: Ohio's Muskingum College Cuts Tuition Ratesand Increases Enrollment." U.S. News &World Report. Vol. 121, No. 23 (December9), p. 24.

Fortune, Jim C. 1993. "Why Production Function Analysis is Irrelevant in Policy Delib-erations Concerning Educational Funding Equity." Education Policy Analysis Archives.Vol. 1, No. 11 (November 2). From website (olam.ed.asu.edu/epaa/v1n11.html),downloaded April 20..

GAO. See: United States General Accounting Office.

Hanushek, Eric A. 1986. "The Economics of Schooling: Production and Efficiency inPublie-.Schools Journal ofEconomic Literature. Vo1.24 (September), pp. 1141-1177.

Hodas, Steven. 1993. "Is Water an Input to a Fish? Problems with the Production-Function Model in Education." Education Polity Analysis Archives. Vol. 1, No. 12(November 2). From website (epaa.asu.edu/epaa/v1n12.html), downloaded April 20.

Hopkins, David S.P. 1990. "The Higher Education Production Function: Theoretical

Foundations and Empirical Findings." In Stephen A. Hoenack and Eileen L. Collins,eds. The Economics of American Universities, Management, Operations, and Fiscal

Environment. State University of NeW York, Frontiers in Education Services, Vol. 10,

Jenny, Hans H. 1997. "The Decile Discount: It's the Structure of Discounts, not.AverageDiscounts...Stupid." Business Officer. March. From NACUBO website(www.nacubo.org).

Kramer, Martin. 1993. "Changing Roles in Higher Education Financing." In NationalCommission on Responsibilities for Financing Postsecondary Education (NCRFPE).Background-Papers and Reports.-Washington, Der(NeRFPE);-Arit---

Lenth, Charles S. 1993. The Tuition DilemmaState Policies and Practices in Pricing PublicHigher Education. Denver, CO: State Higher Education Executive Officers (SHEEO),December.

Lewis, Ethan G. and Gordon C. Winston. 1997. "Subsidies, Costs, Tuition, and Aid inU.S. Higher Education: 1986-87 to 1993-94." Discussion Paper No. 41r, WilliamsProject on the Economics of Higher Education, Williams College, MA, April.

McPherson, Michael and Morton Schapiro. 1998. The Student Aid Game. Princeton, NJ:Princeton University Press.

McPherson, Michael S. and Gordon C. Winston. 1993. "The Economics of Cost, Price,and Quality in U.S. Higher Education." In McPherson, Schapiro, andWinston, eds.,1993, pp. 69-107.

46

46

usAGroup Foundation

McPherson, Michael S., Morton Owen Schapiro, and Gordon C. Winston, eds. 1993.Paying the Piper: Productivity, Incentives, and Financing in U.S. Higher Education. AnnArbor, MI: The University of Michigan Press.

. 1996. "The Economic Analogy." Discussion Paper No. 37, Williams Project onthe Economics of Higher Education, Williams College, MA, June.

Michigan State University (MSU). 1999. "Michigan State University Tuition, Fee, Tax,and Housing Rates to Be Effective Fall Semester 1998." From MSU website(pilot.msu.edu/ unit /ctlr /sa_cashr /tuitio98.htm), downloaded March 24.

Middaugh, Michael F. 1999. "How Much Do Faculty Really Teach?" Planning for HigherEducation. Vol. 27 (Winter 1998-99), pp. 1-11._

Monk, David H. 1990. Education Finance: An Economic Approach. New York, NY:McGraw-Hill.

. 1992. "Education Productivity Research: An Update and Assessment of Its Role inEducation Finance Reform." Educational Evaluation and Policy Analysis. Vol. 14, No.4, pp. 307-332.

MSU. See: Michigan State University.

National Commission on the Cost of Higher Education. 1998. Straight Talk About CollegeCosts & Prices. Phoenix, AZ: Oryx Press.

NCES. See: United States Department of Education. National Center for EducationStatistics.

Picus, Lawrence 0. 1995. "Does Money Matter in Education? A Policymaker's Guide." In-National Center for Education-Statistics (NCES). Selected Papers in School Finance--Washington, DC: Government Printing Office, pp. 16-33.

Schapiro, Morton Owen. 1993. "The Concept of Productivity as Applied to U.S. HigherEducation." In McPherson, Schapiro and Winston, eds., 1993, pp. 37-68.

The Institute for Higher Education Policy. 1999. The Tuition Puzzle: Putting the PiecesTogether. Report prepared for the New Millennium Project on Higher EducationCosts, Pricing, and Productivity. Washington, DC: The Institute for Higher Educa-tion Policy.

Troutt, Marvin D. 1983. "Deciding Tuition Structure with Linear Programming."Research in Higher Education. Vol. 18, No. 3, pp. 359-71.

UIC. See: University of Illinois at Chicago.

University of Delaware. 1999. National Study of Instructional Costs and Productivity.Information from website (www.udel.edu/IR/cost/brochure.html).

}I y jr, I. lir

49

University of Illinois at Chicago (UIC). 1999. "UIC (The University of Illinois at Chi-cago): Undergraduate Catalog 1997-99." From UIC website (www.uic.edu/ucat/vatalog/SC.html), downloaded March 24.

United States Department of Education. National Center for Education Statistics(NCES). 1996. Integrated Postsecondary Education Data System (IPEDS). FinanceSurvey Fiscal Year 1996.

. 1997. Integrated Postsecondary Education Data System (IPEDS). InstitutionalCharacteristics Survey 1996-97.

. 1998. Digest of Education Statistics 1998. Washington, DC: Government PrintingOffice.

United States General Accounting Office (GA0).1998. Higher Education: Tuition Increasesand Colleges' Efforts to Contain Costs. GAO/HEHS-98-227. Washington, DC:Government Printing Office.

Walberg, H.J. 1982. "Educational Productivity: Theory, Evidence, and Prospects."Australian Journal of Education. Vol. 26, No. 2, pp. 115-122.

Winston, Gordon C. 1997. "College Costs: Subsidies, Intuition, and Policy." In NationalCommission on The Cost of Higher Education, 1998, pp. 117-127.

. 1998. 'A Guide to Measuring College Costs." Discussion Paper No. 46, WilliamsProject on the Economics of Higher Education, Williams College, MA, January.

Winston, Gordon C., Jared C. Carbone, and Ethan G. Lewis. 1998. "What's BeenHappening to Higher Education? A Reference Manual, 1986-87 to 1994-95."Williams Project on the Economics of Higher Education, Williams College, MA,March.

Yanikoski, Richard A. and Richard F. Wilson. 1984. "Differential Pricing of Undergradu-

ate Education." Journal of Higher Education. Vol. 55, No. 6 (Nov/Dec), pp. 735-50.

Zemsky, Robert, ed. 1999. Policy Perspectives. Vol. 8, No. 4 (April).

BEST COPY AVAI :113LE3 0

The University of DelawareNational Study of InstructionalCosts and Productivity

The University of Delaware

National Study of InstructionalCosts and Productivity is a data-sharing consortium that was

established in 1992. The study is intendedto be longitudinal, and currently is in itssixth national data collection cycle."Almost 300 colleges and universities haveparticipated in the study, and are aggre-gated into research universities (Research Iand II), doctoral universities (Doctoral Iand II), comprehensive institutions(Master's/Comprehensive I and II), andbaccalaureate institutions (Baccalaureate/Liberal Arts I and II). In the 1996-97cycle, 153 institutions participated,including 48 research universities, 35doctoral universities, 57 comprehensiveinstitutions, and 13 baccalaureate institu-tions. The participating institutions areprimarily public colleges and universities,

except in the classification of baccalaureateinstitutions, which are mostly privateliberal arts colleges. (See Figure 18.)

Therefore, the Delaware Study group ofinstitutions is not necessarily representativeof the national universe of postsecondaryinstitutions, or even of all four-yearinstitutions.

Participants volunteer cost data at thelevel of academic discipline; in the 1996-97 cycle, 29 academic disciplines wereexamined for research universities, withslightly fewer disciplines from the otherCarnegie Classifications. The study focuseson direct instructional costs (personnelcompensation, supplies and services, anddepartmental research and service that arenot separately budgeted), collected on thebasis of student credit hours and disaggre-gated into salaries, benefits, andnon-personnel costs. Due to the difficultyin determining their correct allocation,indirect costs are excluded from theDelaware Study. In addition, the studydoes not capture non-instructional facultyactivity that is funded through departmen-tal instructional funds, such as academicadvising and supervision of student

Figure 18: Comparison of Institutional Universes

Number of cases Percentage of the totalPercentage of each CarnegieClassification

ControlCarnegieClassifications

DelawareStudybenchmarks

REDSnational dataon four-yearinstitutions

DelawareStudybenchmarks

IPEDSnational dataon four-yearinstitutions

DelawareStudybenchmarks

IPEDSnational dataon four-yearinstitutions

Publicprivate, non-profitAll

ResearchResearch

Research

41

748

85

40125

27%5%

31%

6%3%9%

85%15%

100%

68%32%

100%PublicPrivate, non-profitAll

DoctoralDoctoralDoctoral

278

35

6444

108

18%5%

23%

5%3%8%

77%23%

100%

59%41%

100%Publicprivate, non-profitAll

ComprehensiveComprehensiveComprehensive

4116

57

275245

520

27%10%37%

20%18%

38%

72%28%

100%

53%47%

100%PublicPrivate, non-profitAll

BaccalaureateBaccalaureateBaccalaureate

2

11

13

86535628

1%7%8%

6%39%45%

15%85%

100%

14%85%

100%PublicPrivate, non-profitAll

AllAllAll

111

42153

510864

1381

73%27%

100%

37%63%

100%

73%27%

100%

37%63%

100%

Note: For the purposes of this comparison, the IPEDS national data includes only institutions in the research, doctoral,comprehensive, and baccalaureate classifications. In addition, institutions that charge tuition by programprimarilyprivate, for-profit Institutionsare excluded.Source: Delaware Study benchmarks, 1998-97; NCES, 1997

research. The study separately collects

direct expenditure data on research andpublic service. For further definitions andmethodology, see University of Delaware(1999) and Middaugh (1999).

National benchmarks for 1996-97,organized by academic discipline and byCarnegie Classification within eachdiscipline, were developed by the Univer-sity of Delaware and were used in thisreport. In particular, this report uses thenational benchmarks on direct instruc-tional expense per student credit hourtaught. The benchmarks are actually"refined means," in which statisticaloutliers (data more than two standarddeviations above or below the mean) havebeen excluded. In addition, one shouldnote that the benchmarks do not differen-tiate between public and privateinstitutions, even though instructionalcosts are likely to differ somewhat bycontrol. This report's analysis of directinstructional costs is derived from thebenchmarks rather than individualinstitutional recordsin other words, thereported means are usually averages ofaverages, weighted by the number ofobservations. Thus, the means cited inthis report may differ slightly from theones reported in Middaugh, 1999, andelsewhere.

The Integrated PostsecondaryEducation Data System (IPEDS)

The U.S. Department of Education'sNational Center for Education Statistics(NCES) uses the Integrated PostsecondaryEducation Data System (IPEDS) to collectannually information from approximately10,000 postsecondary institutions in theUnited States. IPEDS surveys gather datain various areas, including revenues andexpenditures, fall enrollment, degreecompletions, faculty and staff, and otherissues of interest. Although IPEDS is themost comprehensive data set on institu-tional financing available, it is important

to keep in mind its various limitations,including the following: institutions maydiffer in how they report certain activities;the cost reporting structure is not verydetailed, and distinctions between certaintypes of costs (for example, sponsored

research versus "departmental" research)may be somewhat vague; and tuition andcredit hour data are reported for the"typical" student, masking the significantdifferences that likely exist.

The Finance survey data provide eachinstitution's current fund revenues bysource, and current fund expenditures byfunction, assets and indebtedness, andendowment investments. Data from theFinance survey for Fiscal Year 1996 are themost recent data available as a final release.

Preliminary data are available for morerecent years, but cannot be used to generate

nationally representative figures. For thisreport, the 1995-96 data were derived onlyfrom those institutions that had filled outall fields in the revenue and expenditurecategories (slightly more than 4,000institutions overall, 3,400 of which werepublic or private, non-profit institutions).Data also were broken down by institutional

control and by Carnegie Classification.

Beginning in the Fiscal Year 1997

collection, the Finance survey fOr privateinstitutions was changed to reflect newFinancial Accounting Standards Board(FASB) guidelines; for example, expenses

will be measured according to accrualrather than cost accounting methods(specific changes are noted throughout thetext). For public institutions, the Govern-mental Accounting Standards Board(GASB) recently issued a new accountingrule that will be phased in during the fiscalyears beginning after June 15, 2001.Although it is currently unclear how thenew standard will affect the IPEDSFinance survey for public institutions,changes similar to those for privateinstitutions are expected in the future.

usAGroup Foundation'

50

usAGroup Foundation"

In addition to other broad institutionalinformation, the Institutional Characteris-tics survey collects annual data frominstitutions regarding their tuition andrequired fees for undergraduates, graduatestudents, and first-professional students.According to the IPEDS survey instruc-tions and glossary, required fees are the

"fixed sum charged to students for itemsnot covered by tuition and required ofsuch a large proportion of all students thatthe student who does NOT pay is anexception." This implies that only manda-tory fees are included, not optional or userfees. Institutions are asked upon what basisthey charge tuition. Institutions thatcharge by program are asked to reporttuition and fees for the six programs withthe largest enrollments. These institutionswere excluded from the analysis in thisreport, as most institutions charging byprogram are private, for-profit institutions.Thus, the tuition data provided in thisreport include only institutions thatreported charging tuition by term, by year,or by credit hour (almost 6,400 institu-tions in total, of which about 4,600 werepublic or private, non-profit institutions).The report also focuses primarily onresearch, doctoral, comprehensive, bacca-laureate, andassociate of arts institutions.

Institutions that charge tuition by credithour, by term, or by year were asked toreport tuition and required fees for full-time students for the full academic year.Tuition and fee levels also must be reportedfor in-state and out-of-state students ateach level of instruction. It should benoted that the tuition charges reflect thetypical amount charged to various catego-ries of students, and are averages of thevarious prices charged by institutions in acertain category. Differential prices chargedto particular students in certain targetedcategories are lost by averaging.

At the same time, institutions wereasked to report the typical number ofcredit hours taken by full-time, full-year

students, by level of instruction (under-graduate, graduate, and first-professional).From these data, we calculated the averagetuition and fees per credit hour for eachlevel of instruction, for residency status,for institutional control, and for CarnegieClassification. (Institutions that did notreport the typical number of credit hourstaken by full-time, full-year students wereexcluded from the analysis.) In reality, ofcourse, the number of credit hours takencould vary widely among students. For asummary table of the IPEDS national dataon tuition per credit hour for 1996-97, seeFigure 19 on page 53.

Comparing Costs and PricesAlthough the Delaware Study data were

provided in the form of benchmarks ratherthan individual institutional records, theparticipating colleges and universities wereknown. Thus, it was possible to use thisselected group of institutions for thepurposes of comparison with IPEDS dataon tuition and fees for the same year. Inderiving tuition data for the Delaware Studygroup, two assumptions were utilized:

If a secondary campus for a publicsystem was not specified in the list ofparticipating institutions, data were--drawn only for the main campus ofthe system.

In the few cases in which reportedCarnegie Classification codes differedbetween the Delaware Study and theIPEDS data, the code was changed toremain consistent with the DelawareStudy cost data.

For a summary table of tuition and feesper credit hour for the selected group ofinstitutions, see Figure 20 on page 54. Anyanalysis of this data should keep in mindthe fact that the selected institutions arenot necessarily representative of the fulluniverse of four-year institutions.

Winston (1997) used a broader universeof institutions in his comparison of costsand prices. Data from the 1994-95 IPEDSFinance survey were used for the collegesand universities in the 50 states thatreported positive expenditures, FTEenrollments of more than 100 students, ofwho 20 percent or more were undergradu-ates. These requirements led to a group of2,739 degree-granting institutions. Asummary table adapted from his results isshown in Figure 21 on page 55. Accordingto Winston's definitions, the "generalsubsidy" is equal to educational spendingless the stated price of tuition (stickerprice), whereas the full "subsidy" is

educational spending less the net price(sticker price reduced by institutionalfinancial aid). Thus, the full subsidyincludes both the general subsidy to allstudents and an individual subsidy in theform of student aid.

Additional Data NeedsFor state policymakers, price-to-cost

ratios and subsidy patterns that hold otherfactors constant for groups of institutions,such as the ratios revealed in this report,can give some sense of broad trends in thecost/price relationship. For decision-making at the institutional level, a morecomplex analysis that takes into accountlower levels of analysis, cross-cutting

categories of students, and demand factorsshould be accomplished. In the latter case,the theory of the cost/price relationshipcurrently exceeds the ability of the data tosupport it. Nevertheless, the advantages tobe gained by institutions, students, and thepublic in general make additional datacollection efforts worthwhile.

The analysis presented in this report hasrevealed some of the problems in measur-ing cost data in particular. Clearly,modified cost accounting techniques areneeded in higher education, in order toencourage the collection and disseminationof more detailed cost data. Thus, cost

accounting techniques must be appliedwith reference to the varying missions ofpostsecondary institutions, the multiplesources of revenue used by most institu-tions, and the range of units ofmeasurement that exist. Meaningfulcomparisons and benchmarks require:consistent cost and revenue data; similardata between public and private institu-tions; and further breakdowns of data bydiscipline and program. Although theIPEDS data, the Delaware Study cost data,and other datasets each meet some of theseobjectives, none satisfies all of the needs.At the same time, existing data do notgenerally make it possible to distinguish

between the sticker price and the net pricefor categories of students. In a period inwhich universities are expanding their useof differential pricing to attract students ofmerit, minority students, and so forth, thisdifference is becoming increasinglyimportant.

As mentioned, changes are currentlybeing made in the IPEDS. Finance surveythat may address some of these concerns.

For example, tuition and fee revenue forprivate institutions is now reported (as ofFY1997) net of price discounts, with anoffsetting adjustment to the expense side.Similaf--:bift not identical changes areexpected for public institutions in the nextfew years. However, revenue and expendi-

ture information will still not be availableat a more disaggregated level, and as aresult of these changes researchers will

have more difficulty in comparing publicand private institutions.

Consistent measurement of costs andprices is also essential. Currently, the dataare not only kept inconsistently across

institutions, but also are difficult tointerpret across institutions even whenthey are compatible. The terms used inhigher education finance require precisionif comparisons are to be made between

institutions, between years, or, for that

usAGroup Foundation'

matter, between research efforts. Anyexamination of cost/price data as it appliesto institutions, and programs or disciplineswithin institutions, involves choices thatmust be explained: (1) how to groupdifferences in costs per program, discipline,or course; (2) how to spread indirect costsamong programs, disciplines, or courseswithin an institution; and (3) how tomeasure units (per student, credit hour,total enrollment, or other unit). In addi-tion, careful delineation of terms such as"subsidy" (particularly "cross subsidy") or"total cost" can avoid misunderstandings.However, even similar ratios or other coststatistics must be evaluated in light of theunique mission and structure of eachinstitution.

Ultimately, it is important to carefullycollect and use cost and price data torespond to the questions that are asked.For example:

Time series and longitudinal data mustbe collected using a consistent meth-odology and must be utilized torespond to questions that refer tochanges over time.

Average data can be used to respond toquestions dealing with totalKtutalcosts, tuition revenues, or subsidy, for-instanceor totals within specificcategories. It is important to rememberthat summary data, such as averages orratios, run the risk of masking impor-tant details. The same average costs donot directly correspond to identicalproduction functions; one universitymay use greater faculty input and less

technology, while another may usegreater technology and less faculty.

Along similar lines, benchmarks andratios can be used in selective and

appropriate circumstances, especiallywhen they do not mask vastly differentunderlying totals.

Marginal data, such as the cost ofeducating an additional student or theprice charged to an additional student,must be used to answer questionsregarding the relative incentives ofspecific pricing policies to targetedcategories of students. The relation-ship between price and cost is onlycomplete if demand characteristics areknown for any institutionthat is,the extent to which various types ofstudents are sensitive to the pricesbeing charged.

Of course, the technical aspects of costmeasurement and data collection compriseonly one of the obstacles to greater use ofcost accounting in higher education.Ultimately, the cost accounting movementwill need to address the cultural resistanceto change as well. The clear interest ofvarious state legislatures and competinguniversities to have information regardingcost/price relationships, and the danger ofusing that information in a way thatreduces student options and choices, couldbe resolved by publicly providing onlyaverages for groups of institutions ratherthan individual institution data. Mean-while, institutions should collect moredetailed data for their own internalanalyses.

55

Fig

ure

19: N

atio

nal A

vera

ge D

ata

on T

uitio

n an

d F

ees

per

Cre

dit H

our,

by

Leve

l of I

nstit

utio

n,R

esid

ency

Sta

tus,

and

Car

negi

e C

lass

ifica

tion,

199

6-97

(Inc

lude

s on

ly th

ose

inst

itutio

ns th

at c

harg

e tu

ition

by

cred

it, b

yte

rm, o

r by

yea

r)

Und

ergr

adua

te tu

ition

and

requ

ired

fee

s pe

r cr

edit

hour

:G

radu

ate

tuiti

on a

nd r

equi

red

fees

per

cre

dit h

our:

Gra

duat

e/un

derg

radu

ate

tuni

cdi

ffer

entia

l, pe

r cr

edit

hour

:Fi

rst-

prof

essi

onal

per

cred

it ho

ur:

Ave

rage

,in

-sta

te

tuiti

on

Ave

rage

,ou

t-of

-sta

te

and

fees

Rat

io, o

ut-

of-s

tate

toin

-sta

te

I I C

ontr

olC

arne

gie

Cla

ssif

icat

ion

Num

ber

ofin

stitu

tions

In-s

tate

Out

-of-

stat

e

Rat

io, o

ut-

of-s

tate

toin

-sta

teIn

-sta

teO

ut-o

f-st

ate

Rat

io, o

utof

-sta

te t

,In

-sl

at

Rat

io, g

radu

ate to

unde

rgra

duat

e,In

-sta

te

Rat

io, g

radu

at t ,

unde

rgra

duat

eou

t-of

-sta

Publ

ic

I

Res

earc

h85

$115

$323

3.0

$178

$436

2.6T

1.6

1.4

$221

5509

2.5

Doc

tora

l64

$113

$310

2.8

$152

$374

2.5

1.5

1.3

8201

$426

2.2

Com

preh

ensi

ve27

5S9

552

522.

8$1

4853

302.

41.

51.

382

22$5

652.

6B

acca

laur

eate

86S9

6$2

372.

5$1

27$2

912.

41.

31.

2A

ssoc

iate

of

Art

s92

3$6

3$1

463.

2Pr

ivat

e, n

on-p

rofd

Res

earc

h40

$720

$7220

$900

$900

1.01

1.2

1.2

5892

$892

1.0

Doc

tora

l44

5470

$470

1.0

$524

$524

1.0;

1.1

1.1

1.1

$598

$610

1.0

Com

irre

hers

ive

245

$421

$421

1.0

$379

$379

1.0

I i1.

01.

0$5

27$5

271.

0B

acca

laur

eate

535

$450

$450

1.0

$386

$386

1.01

1.0

1.0

$344

$344

1.0

Ass

ocia

te o

f A

rts

148

$250

$253

1.0

All

Res

earc

h12

552

83$4

352.

453

6855

582.

211.

51.

4$3

61$5

892.

2D

octo

ral

IOR

$241

$367

2.2

$298

5433

1.9

1.3

1.3

$437

$535

1.5

Com

preh

ensi

ve52

0$2

45$3

302.

0$2

5653

531.

71.

31.

2$4

58$5

36-

1.4

Bac

cala

urea

te62

8$3

94$4

151.

2$3

3953

681.

21.

01.

0$3

44$3

441.

0A

ssoc

iate

of

Art

s12

73$1

01$1

632.

6Pu

blic

All

200i

-57

651

842.

951

5053

44L

41.

51.

3$2

1654

912.

4Pr

ivat

e, n

on -

prof

itA

ll25

567

$354

S355

1.0

,53

6553

681.

01.

01.

053

6453

691.

0T

he B

acca

laur

eate

and

Ass

ocia

te o

f A

rta

Car

negi

e C

lass

itiat

ions

incl

ude

som

e pr

ivat

e,or

-pr

ofit

inst

itutio

ns.

The

'all

publ

ic"

and

'all

priv

ate,

non

-pro

fd"

cate

gori

es in

clud

e so

me

Car

negi

e C

lass

ific

atio

ns th

atar

e no

t sho

wn.

Ave

rage

s fo

r fi

rst-

prof

essi

onal

tuiti

ons

are

for

only

thos

e pr

ogra

ms

offe

red

by s

peci

fic

Car

negi

e C

lass

ific

atio

ns, a

ndar

e av

erag

es o

f av

erag

es.

Not

e: S

ome

scho

ols

did

not r

epor

t the

typi

cal n

umbe

r of

cre

dit h

ours

take

n by

thei

r st

uden

ts. T

hese

inst

itutio

nsw

ere

excl

uded

, as

tuiti

on p

er c

redi

t hou

r fi

gure

s co

uld

not b

e ca

lcul

ated

.So

urce

: NC

ES,

199

7

as

56B

EST

CO

PY A

V L

AB

LE

,T

vzia

s

r.

Fig

ure

20: T

uitio

n an

d F

ees

per

Cre

dit H

our,

by

Leve

l of I

nstit

utio

n, R

esid

ency

Sta

tus,

and

Car

negi

e C

lass

ifica

tion

for

the

Del

awar

e S

tudy

Gro

up o

f Ins

titut

ions

, 199

6-97

Und

ergr

adua

te lu

ltion

and

requ

ired

fee

s pe

r cr

edit

hour

:

-1G

radu

atei

unde

rgra

duat

eG

radu

ate

tuiti

on a

nd r

equi

red

fees

per

cre

dit h

our:

tuiti

on d

iffe

rent

ild, p

er c

redi

tho

ur:

Firs

t-pr

ofes

sion

al tu

ition

and

fee

spe

r cr

edit

hour

:

Con

trol

Car

negi

eC

lass

ific

atio

nN

umbe

rIn

stitu

tloIn

-sta

te

I Out

-of-

stat

e

Rat

io, o

ut-

of-s

tate

toin

-sta

tein

-sta

teO

ut-o

f-st

ate

Rat

io, o

ut-

of-s

tate

toIn

-sta

te

Rat

io,

grad

uate

toun

derg

radu

ate

,In

-sta

te

Rat

io,

grad

uate

toun

derg

radu

ate,

out

-of-

stat

eA

vera

ge,

In -

stat

e

Ave

rage

,ou

t-of

-st

ate

Rat

io, o

utof

-sta

te t

in-s

lat

blip

Res

earc

h41

$103

I $2

993.

1$1

59$4

032.

71.

51.

4$1

86$4

392.

6D

octo

ral

27$1

09$3

102.

9$1

53$3

832.

51.

51.

4$2

08$4

352.

2co

mpr

ehen

sive

41$1

041

5254

2.6

51 6

4$3

422.

21.

51.

3$2

74$5

642.

0B

acca

laur

eate

2$7

7I

$265

3.7

$130

$374

3.4

1.6

1.4

vide

, non

-pro

fit

Res

earc

h7

$561

$568

1.1

$756

$756

1.0

1.0

1.0

$1,3

32$1

,332

1.0

Doc

tora

l8

$471

$471

1.0

$501

$501

1.0

1.1

1.1

$678

$678

1.0

Com

preh

ensi

ve16

5451

$451

1.0

5421

$421

1.0

0.9

0.9

$385

$385

1.0

Bac

cala

urea

te11

$676

$677

1.0

$491

$491

1.0

1.1

1.1

IR

esea

rch

4$1

7053

382.

8$2

37$4

492.

51.

51.

3$3

00$5

282.

4D

octo

ral

35$1

97$3

492.

4$2

35$4

102.

21.

41.

3$4

69$5

701.

5C

ompr

ehen

sive

57$1

991

5308

2.2

$232

$363

1.9

1.4

1.2

$341

$457

1.4

Bac

cala

urea

te13

$576

I $6

081.

553

88$4

581.

71.

31.

2Pu

blic

All

111

$104

$285

2.9

$159

$376

2.5

1.5

1.4

$194

$446

2.5

Priv

ate,

non

-pro

fit

All

4255

32$5

341.

0$5

15$5

15I

1.0

1.0

IT. O

T$7

76$7

761.

0A

llA

ll15

3$2

20$3

522.

4$2

43$4

092.

11.

41.

3$3

39$5

292.

1

Ave

rage

s fo

r fi

rst-

prof

essi

onal

tuiti

ons

arc

for

only

thos

e pr

ogra

ms

offe

red

by s

peci

fic)

Car

negi

e C

lass

ific

atio

ns.

Not

e: S

onia

sch

ools

did

not

rep

ort t

he ty

pica

l num

ber

of c

redi

t hou

rs ta

ken

by th

eir

stud

enti.

The

se in

stitu

tions

wer

e oc

clud

ed, a

s tu

ition

per

cred

it ho

ur f

igur

es c

ould

not

be

calc

ulat

ed.

Sour

ce: N

CE

S, 1

997

JEST

CO

PYA

VA

RA

BL

E

Fig

ure

21: S

ubsi

dies

, Cos

ts, P

rices

and

Aid

by

Con

trol

and

Car

negi

e C

lass

ifica

tion,

199

4-95

(in d

olla

rs p

er F

TE

stu

dent

)

Num

ber

ofIn

stitu

tions

Ave

rage

Edu

catio

nal

Spen

ding

Ave

rage

Stic

ker

Pric

e

Stic

ker

Pric

eas

Per

cent

of

Edu

catio

nal

Spen

ding

Ave

rage

Gen

eral

Subs

idy

Gen

eral

Subs

idy

asPe

rcen

t of

Edu

catio

nal

Spen

ding

Ave

rage

Net

Pri

ce

Net

Pri

ce a

sPe

rcen

t of

Edu

catio

nal

Spen

ding

Ave

rage

Full

Subs

idy

Full

Subs

idy

as P

erce

nt o

fE

duca

tiona

l

Spen

ding

1A11

Ins

titut

ions

2,73

9$1

1,96

7$5

,919

49%

$6,0

4851

%$3

,770

32%

$8,1

9768

%

Publ

ic I

nstit

utio

ns1,

420

$9,9

19$2

,272

23%

$7,6

4777

%$1

,233

12%

$8,6

8688

%R

esea

rch

83$1

3,44

8$4

,571

34%

$8,8

7766

%$3

,150

23%

$10,

298

77%

Doc

tora

l63

$11,

155

$3,7

7634

%$7

,379

66%

$2,6

5624

%$8

,499

76%

ompr

ehen

sive

271

$9,9

33$2

,907

29%

$7,0

2671

%$1

,816

18%

$8,1

1782

%ib

eral

Art

s (B

acca

laur

eate

)w

o-Y

ear

80$9

,389

874

i$8

,022

$2,8

57$1

,623

30%

20%

$6,5

32$6

,399

70%

80%

$1,5

80$7

1417

% 9%$7

,809

$7,3

0883

%91

%Pr

ivat

e In

stitu

tions

1,31

9$1

4,17

2$9

,846

69%

$4,3

2631

%$6

,502

46%

$7,6

7054

%es

eam

h39

$32,

014

$16,

975

53%

$15,

039

47%

$11,

646

36%

$20,

368

64%

tora

l43

$18,

736

$14,

417

77%

$4,3

1923

%$1

1,01

659

%$7

,720

41%

ompr

ehen

sive

245

$12,

903

$10,

484

81%

$2,4

1919

%$7

,399

57%

$5,5

0443

%ib

eral

Art

s (B

acca

laur

eate

)52

6$1

5,42

5$1

0,50

568

%$4

,920

32%

$6,3

6141

%$9

,064

59%

wo-

Yea

r24

3i$1

0,22

7$7

,565

74%

$2,6

6226

6/$5

,063

50%

$5,1

6450

%G

ener

al S

ubsi

dy =

Edu

catio

nal S

nend

ino

I pea

Stir

lepr

13.

40,.

Full

Subs

idy

= E

duca

tiona

l Spe

ndin

g L

ess

Net

Pri

ceSo

urce

: Ada

pted

fro

m W

inst

on, 1

997.

1. For discussion of the impact of prices and costs on "quality," see McPherson and Winston, 1993.

2. "Pricing" in this report refers to the amounts charged to external consumersprimarily studentsratherthan "internal" pricing strategies involving various academic departments. Cost accounting would allow,

indeed would require, that an institution trace and impute internal costs, as well.

3. This section discusses costs to institutions, not costs to students, which would consist of both price andforegone income.

4. In other words, an input is any resource the institution uses in its production process.

5. For further discussion of the special characteristics of higher education, see various papers produced

under the Williams Project on the Economics of Higher Education, Williams College, MA(www.williams.edu/Mellon); Bowen, 1980; and Allen and Brinkman, 1983.

6. With multiple revenue sources: (1) relatively high subsidy attracts high quality students; (2) high qualitystudents perform well in school and beyond; (3) persons who perform well beyond school earn higherincomes, add significantly to the institution's prestige (and ability to attract gifts, grants and donations), andgive larger donations; (4) larger gifts, grants, and donations provide larger subsidies, and so on.

7. Studies of the production functions of various commodities generally attempt to answer one or more ofthe following questions in a predictive manner: 1) If all inputs are doubled, tripled, or multiplied by someamount, will the output be increased by more or less than that multiple? In other words, are there "econo-

mies of scale" in production? 2) Can the proportions of the inputs be altered and still produce the sameoutput? Or, conversely, is output for the given set of inputs being maximized? 3) Subsequently, how do thecosts of the inputs alter the proportions of inputs used?

8. Various papers have described and estimated production functions for higher education (for example, seeAllen and Brinkman, 1983; Hopkins, 1990; Schapiro, 1993; Dundar and Lewis, 1995) and K-12 education(Walberg, 1982; Hanushek, 1986; Monk, 1990 and 1992). Debate over the usefulness of the approach hascentered on the ability of the functions to capture all aspects of achievement (for example, see Fortune,

1993; and Hodas, 1993).

9. The appropriate breakdown depends upon the question that is being asked of the accounting data.

10. Though somewhat ad hoc, student units also may be weighted per level of student to reflect increasingcosts in upper-level courses (Bowen, 1980).

11. It should be noted that just because a measure correlates strongly with an output variable, this does notimply that the measure is an input variable or a proxy for an input variable, and certainly not that therelationship constitutes a production function. It is a correlation. For example, the discussion regardingwhether expenditure impacts student outcomes at levels K-12 is a question of correlation, not the K-12production function, as is often suggested. For example, see the discussion in Picus, 1995.

12. Whether it is tuition or non-tuition revenue that is actually treated as the residual by policymakers is aquestion that cannot be answered here.

13. Various authors have suggested constructing marginal data by looking at cost accounting data over time,which could reveal costs at differing levels of matriculation or instruction; for example, see Allen and

Brinkman, 1983.

14. For a discussion of the problems of accounting for capital costs, see McPherson and Winston, 1993. Inaddition, some studies of the "cost burden" on students argue that other types of costs, such as the costs of

student subsistence (room and board) and foregone student earnings, should be included in a broaderassessment of the costs of higher education (see Kramer, 1993).

BEST COPY AVAHAtfLE 62

15. Methods for IPEDS Finance data collection are changing. For private institutions, survey forms havebeen changed to reflect new Financial Accounting Standards Board (FASB) standards (116, 117, and 124).

Under the new forms, which became effective in FYI997, expenses will be measured according to accrual

rather than cost accounting methods. Similar changes are expected for public institutions in the future, as

the Governmental Accounting Standards Board (GASB) recently issued a new government accounting rule.

16. Under the new IPEDS form for private institutions, functions are called "expenses" rather than currentexpenditures, and there is no E&G expenditures category.

17. The balance of E&G expenditures also includes mandatory and non-mandatory transfers of funds outof the institution's current funds accountfor example, for the amortization of debt.

18. In the new IPEDS form for private institutions, this category includes only student aid that is recog-nized as an expense in the institution's financial statements; tuition remissions are reported as a netreduction of revenue.

19. Some cost studies (for example, Winston, 1998; McPherson and Schapiro, 1998) suggest subtracting all

scholarships and fellowships from total educational costs because they represent a reduction in the pricefaced by students (and therefore in the net tuition receipts of the institution) rather than a cost. However,this approach cannot be applied at lower levels of analysis (for example, by level of instruction). At the sametime, some institutional aid may legitimately represent a cost rather than simply a discount in priceforexample, improved quality of students (inputs) at selective institutionsand it is difficult to separate out thetwo types of uses (Winston, 1998). As a result, price discounts may be treated as expenditures in cost studies

(a "generally accepted accounting practice" in private industry) rather than as reductions in tuition revenue,and in fact are treated as expenditures in data collected for IPEDS (for private institutions, before the FASB

changes). Because of the nature of the federal data, scholarships and fellowships are effectively describedas acost in this report, although they are excluded from instructional costs.

20. As mentioned in the previous note, some cost studies suggest treating scholarships and fellowships as areduction in the price faced by students. Others disagree (see Bowen, 1980, 1981). In this report, scholar- .

ships and fellowships are effectively described as an indirect cost, but they are excluded from the discussionof instructional costs.

21. In addition, cost studies have various ways of dealing with capital costs and other issues.

22. In this report, this category is referred to simply as comprehensive institutions.

23. In fact, Brinkman (1993, p. 4) has called this use of upper-division students to teach lower-division

students an example of "cross-subsidization," although the term does not conform to the definition of cross-

24. See appendices for more detailed explanation and description of how this group of institutions differsfrom the national distribution.

25. For example, means calculated by Carnegie Classification and reported here differ from the ones

reported in Middaugh (1999), which are presumably calculated from the individual records. The rank orderis the same, however, with the exception of baccalaureate colleges.

27. Optional fees also may be excluded because they are charged upon use.

28. Some examinations of higher education finance (for example, McPherson and Schapiro, 1998) subtractinstitutional aida subcategory of the scholarships and fellowships categoryfrom total tuition revenue inorder to arrive at net tuition revenue. However, this approach cannot be applied at lower levels of analysis.In other words, we could achieve an average net tuition figure, but not figures representing the net pricesfaced by various categories of students. In addition, many price discounts are based on the total price of

BEST COPY AVAILABLE

attendance, including room and board, rather than on tuition and fees alone (Jenny, 1997). Given theseconsiderations, the tuition data presented in this report generally do not take into account institutional aid.

29. Note that "tuition" as used in this report means tuition and fees, and is not net of financial aid unless

30. Ideally, this calculation would be net of tuition discounts and other forms of institutional aid. If so, theequation would be: [Average Net Tuition per Student Unit] x [Number of Student Units] = Total NetRevenue from Tuition.

31. Muskingum College president Samuel Speck has claimed that after annual tuition was slashed by$4,000 (about one-third), the institution's gross revenues increased by 6 percent, the freshmen class climbed

by 35 percent, and the number of transfer students grew by 50 percent without sacrificing the quality ofeducation; see Fischer, 1996, p. 24.

32. Under the new IPEDS form for private institutions, all revenue includes both current and non-currentrevenue; the previous form included only current revenue.

33. Under the new IPEDS form for private institutions, tuition and fees are to be reported net of studentaid that was applied to tuition and fees. One should also keep in mind that some forms of government-provided student aidsuch as student loansand some private scholarships that are awarded directly tostudents may show up as tuition revenue. Other forms of student aid show up in other revenue categories;for example, Pell Grants are reflected in the federal grants and contracts category.

34. The new IPEDS form for private institutions includes a new category: contributions from affiliatedentities. Such as fund-raising foundations and booster clubs.

35. Under the new IPEDS form for private institutions, this category is called "investment return" anddiffers slightly in its definition.

36. The data that follow include only institutions that reported charging tuition by term, by year, or by

credit hour. See appendices for more details.

37. Note that tuition charges reflect the amount charged to various categories of students, and tuition percredit hour is calculated using the typical number of credit hours taken by all full-time, full-year students.

See appendices for more details.

38. For a summary table of national price averages that shows these trends, see Figure 19 in the appendices.

39. The latter category was directed only toward students who had been enrolled prior to summer 1996;full-time students who enrolled since that time paid a package rate of $14,070 for 12-18 credit hours.

40. These rates are for students registered for 12 or more semester hours.

41. In a private market with highly differentiated producers, economic theory suggests only the following:1) the average revenue must equal or exceed the average costs if the enterprise is to stay in businessabusiness can't operate at a net loss; 2) if the cost of producing each additional unit increases as production

increases, then the producer's profitability is maximized when the cost for the additional unit equals therevenue generated by the additional unitin other words, when the marginal cost equals the marginalrevenue (in private industry, profitability must be the prime consideration); and 3) the price that can be

charged does not depend directly upon costs, but rather depends upon the price that consumers are willingto pay for the level of production that the producer has decided is the optimal level to produce.

42. Winston (1997) and the National Commission on the Cost of Higher Education (1998) call this the"general subsidy." The other part of subsidy is institutional scholarships and other price discounts; thus, the

full student subsidy would be measured as average educational cost less average net price. However, since the

currently available data measure stated tuition and fees rather than net tuition, "general subsidy" rather than"full subsidy" is used primarily in this report.

43. Information derived through cost accounting for an individual institution could provide a considerableportion of the data required to describe the cost/price relationship for that institution. Over time, costaccounting data could even outline the demand characteristics associated with the institution.

44. In fact, "...any given level of subsidy resources can support an infinite variety of cost and price strategiesso long as the difference between costs and price matches those non-tuition resources" (McPherson,Schapiro, and Winston, 1996, p. 8).

45. This statement also assumes that auxiliary enterprises and other independent operations break even,and does not take into account capital expenditures. Furthermore, it assumes that the institution is notsaving

,,revenue.

46. Again, this description of subsidy does not take into account financial aid and other price discounts. Inaddition, costs here refer to total educational costs, not total E&G and other expenditures.

47. In these instances, a college takes on many of the price/cost characteristics ofa regulated utility, whereprices are restricted to average cost by regulationnot by market force. Generally, the relationship inregulated utilities relates price not to average cost but to some rate of return on invested capital that theregulators find acceptable. In the case of higher education, the rate may be indexed, established as a constantproportion of cost, or set as a fixed percentage of appropriated amounts.

48. Overall average tuition ignores the varying tuition rates charged to non-residents, graduate students,and other groups, and does not incorporate any estimate of tuition net of discounts or student aid.

49. Tuition and all financial variables were adjusted for inflation. The 15 possible explanatory variablesincluded the change in: government appropriations; revenue from grants, contracts and other sources;instruction expenditures, student services expenditures; institutional scholarships, fellowships and grants;other student-related expenditures; research expenditures; other non-student related expenditures; the ratioof in-state undergraduate tuition to average tuition received per FTE student; the amount of tuition usedfor non-current fund purposes; the market value of the endowment fund; theamount by which revenuesexceeded expenditures; and the surplus or deficit from independent operations, auxiliary enterprises, andhospitals. The absolute ratios of in-state undergraduate tuition to average tuition received per FTE studentand the level of in-state undergraduate tuition in 1989-90 also were used as explanatory variables.

50. See appendices for more detailed methodology.

51. Because prices are being compared to direct instructional costs rather than to full costs, the price-to-cost ratios frequently are greater than 100 percent.

52. See appendices for more details.

53. The concept of a better value for the money has been raised by McPherson and Winston (1993) and byWinston, Carbone, and Lewis (1998), who also suggested that enrollment in the public sector may havebeen shifting to community colleges because of the relatively low increase in the prices students pay for adollar's worth of education. It also appears that consumers are paying higher proportions of their educationalcosts now than they did in the past. Although it is beyond the scope of this report, it is possible to constructtrends in price-to-cost ratios and in average subsidies over time using IPEDS Finance data, at least byinstitutional type. Lewis and Winston (1997), Winston, Carbone and Lewis (1998), and others haveattempted to do this, with the resulting conclusions that students are paying more for the education they get.

54. As Robert Zemsky (1999) suggests, especially in public systems, the goals of each type of institutionshould be more distinct from each other.

55. Federal government funds accounted for 11 percent of the funds used by public institutions. Stategovernments accounted for only 2 percent of the funds for private institutions, while the federal govern-ment accounted for 14 percent. (NCES, 1998, p. 193).

BEST COPY AVAILABLEt.,

56. The choice between the extremes of mandating fixed price-to-cost relationships or allowing tuition

engineering has several consequences. Fixed relationships between prices and costs have the danger oflessening any institution's ability to: (1) compete with other institutions in the same or similar market nichefor students with particular characteristics; (2) maximize tuition income in order to reduce reliance onalternative income sources; and (3) structure cost/price relationships that accomplish other objectives of theinstitution. On the other hand, tuition engineering has two major problems: (1) it produces a complexstructure that is not transparent to the student or parent who must plan for the college education, and (2) itruns the risk of substituting ad hoc policies selected by colleges for open public policies selected through a

political process. Furthermore, tuition engineering can create a dynamic of competition for gifted studentsthat can ultimately lead to a situation in which good students pay less at good schools than poorer studentspay at lesser schoolsthe result being an increasingly hierarchical structure of students and institutions.

57. Ideally, such data should be collected in a well-defined and consistent manner across institutions and

programs within institutions. However, recognizing the various methodologies that exist and the fact thateven consistent data do not necessarily mean the same thing for each institution, the available data may still

be useful in terms of its application to institutional operations and public policy. For more details regardingadditional data needs, see appendices.

58. Prior to 1997, all of the data were collected in a single survey, whereas beginning with the 1997 data,collection cycles were divided into two phases. The change in data collection time frame resulted in the

absence of fall 1995 teaching load data and academic/FY 1995-96 cost and productivity data (seeMiddaugh, 1999). The data presented in this report reflect the 1996-97 academic and fiscal year.

BEST COPY AVAELABLE

usAGraup Rnindatiorr New AgendaSeries'

e purposeof the USA Group'Foundatiorris.to carry out they,-:.

research-and philanthropic goals;

of USA Group; Inc:, with special

emphasis on improving higher education:

Specifically, the Foundation supports

original and sponsored research, and

sponsors selected educational activities.

The Foundation also supports educational

and civic nonprofit endeavors primarily

in the communities where USA Group

employees live and work.

The Foundation encourages originalsponsored research, which is typically

commissioned by the Foundation to

address critical issues about the

improvement of higher education.

Believing that published research may

have the longest-term impact on higher

education, the Foundation publishes and

disseminates articles, research reports

and books. Topics and approaches to

subject matter that are more practical

than theoretical, and which emphasize

pragmatic tools that will assist institutions

and public policy makers, are preferred.

Additional information about the

Foundation may be obtained from:

Dr. Robert C. Dickeson

Senior Vice President

USA Group Foundation

P.O. Box 7039

Indianapolis, IN 46207-7039

317/951-5755

317/951-5063 fax

rdickeso@usagroup.com

Additional information about USA Group

may be obtained by visiting the website

at www.usagroup.com.

The USA Group FoundatiorrNew Agenda.Series7

Robert.C..DickesovExecutive Editor

Jean B. Rose

Managing Editor

Natasha Swingley

Design & Production

Printed by Design Printing

USA Group FoundationEditorial Advisory Board

Martha D. Lamkin

Susan O. Conner

Robert R Murray

Patricia M. Scherschel

Sara Murray-Plumer

The USA Group Foundation

New Agenda Series' is published

periodically by the USA Group

Foundation, P.O. Box 7039, Indianapolis,

Indiana 46207-7039.

Copyright © 1999, USA Group, Inc.

All rights reserved.

POSTMASTER: Send all address changes

to USA Group Foundation, P.O. Box 7039,

Indianapolis, Indiana 46207-7039.

67BE

A o 0

BESTCaPYAVAILABLE

I

68

U.S. Department of EducationOffice of Educational Research and Improvement (OERI)

National Library of Education (NLE)

Educational Resources Information Center (ERIC)

NOTICE

Reproduction Basis

ERIC

This document is covered by a signed "Reproduction Release(Blanket)" form (on file within the ERIC system), encompassing allor classes of documents from its source organization and, therefore,does not require a "Specific Document" Release form.

This document is Federally-funded, or carries its own permission toreproduce, or is otherwise in the public domain and, therefore, maybe reproduced by ERIC without a signed Reproduction Release form(either "Specific Document" or "Blanket").

EFF-089 (3/2000)