Skip Navigation
Illustration/Logo View Quarterly by  This Issue  |  Volume and Issue  |  Topics
Education Statistics Quarterly
Vol 5, Issue 2, Topic: Postsecondary Education
A Study of Higher Education Instructional Expenditures: The Delaware Study of Instructional Costs and Productivity
By: Michael F. Middaugh, Rosalinda Graham, and Abdus Shahid
 
This article was originally published as the Executive Summary of the Research and Development Report of the same name. The sample survey data are from the Delaware Study of Instructional Costs and Productivity.
 
 

Research and Development Reports are intended to
  • share studies and research that are developmental in nature;
  • share results of studies that are on the cutting edge of methodological developments; and
  • participate in discussions of emerging issues of interest to researchers.

These reports present results or discussion that do not reach definitive conclusions at this point in time, either because the data are tentative, the methodology is new and developing, or the topic is one on which there are divergent views. Therefore, the techniques and inferences made from the data are tentative and are subject to revision.

A Study of Higher Education Instructional Expenditures is an examination of higher education costs undertaken by the National Center for Education Statistics (NCES). This study of higher education costs was mandated by Congress in the 1998 Higher Education Act. The NCES response to the congressional mandate encompassed three reports: Study of College Costs and Prices, 1988–89 to 1997–98 (Cunningham et al. 2001); What Students Pay for College: Changes in Net Price of College Attendance Between 1992–93 and 1999–2000 (Horn, Wei, and Berker 2002); and this third and final report.

The first report in the congressionally mandated study drew the distinction between sticker price, i.e., the tuition that an institution charges for a college education, and cost, i.e., the fiscal resources expended by the institution to provide that education. Additionally, researchers for the first part of the study found that certain factors are associated with tuition rates. Most notable at state-supported institutions is the importance of annual budget appropriations. At private not-for-profit institutions, internal budget constraints, size of endowments, and external market competition were among factors associated with sticker price. There was little evidence indicating that expenditures for instruction were a major factor in determining tuition rates.

This report focuses solely on the issue of direct instructional expenditures, and the factors associated with the comparative magnitude of those expenditures at 4-year colleges and universities in the United States. As evident in the findings and conclusions, the factors associated with instructional expenditures are different from those associated with sticker price, as identified in the first part of the congressionally mandated study. Cost and price are not interchangeable constructs, and a strong statistical relationship between them has not been found.

The data source for this analysis is multiple cycles of the Delaware Study of Instructional Costs and Productivity, henceforth called the Delaware Study. Begun in 1992 by the Office of Institutional Research and Planning at the University of Delaware, the study has grown into a national data-sharing consortium embracing over 300 4-year colleges and universities across the United States. The foci of data-sharing activities are detailed analyses of teaching loads by faculty category, instructional costs, and externally funded scholarly activity, all at the level of the academic discipline.

Back to Top


The primary objective of this analysis of instructional expenditures is the identification of those factors that contribute to describing direct instructional costs in the colleges and universities that participate in the Delaware Study.

The study is characterized by the following factors:

  • Participation in the Delaware Study is voluntary, and is restricted to 4-year Title IV–eligible institutions only. The fact that the data population used in this study is self-selected raises the issue of nonresponse bias. For example, institutions that participate in the Delaware Study typically have enrollments of at least 5,000 students and are organizationally complex, with discrete academic departments or programs that correspond with the four-digit codes assigned to disciplines within the NCES Classification of Instructional Programs (CIP) taxonomy (Morgan, Hunt, and Carpenter 1991). In contrast, single-purpose institutions with smaller enrollments frequently have multiple disciplines grouped within a given organizational structure, e.g., Division of Social Sciences or Department of Education, and participate in much smaller numbers than their larger, more complex counterparts. In addition, because participation is restricted to 4-year institutions, findings cannot be extended to the 2-year college sector.
  • Because the population for this study is self-selected, it is, by definition, not a random sample. Descriptive statistics are applied to data from responding institutions to describe instructional expenditures for those institutions, but the findings cannot be inferentially generalized to the larger population of all Title IV–eligible 4-year colleges and universities in the United States. However, this study's findings nonetheless yield valuable descriptive information about expenditures in those institutions that participate in the Delaware data-sharing process.
  • The Delaware Study expenditure data reflect direct instructional expense, and therefore cannot be used for a full cost model. There are methodological pitfalls and inconsistencies in full cost modeling in higher education, especially with respect to allocating indirect costs (as described in the full report).

Within the context of these characteristics, this study yields information about factors that contribute to direct instructional costs at an institution, and these expenditures generally compose the largest portion of the operating budget at most colleges and universities.

Back to Top


This study utilized data from multiple data collection cycles of the Delaware Study, focusing primarily on data collected during 1998, 2000, and 2001. Data were collected using an established survey instrument that requests detailed information on fall semester teaching loads by faculty category, and academic and fiscal year student credit hour production and direct expenses for instruction, research, and service activity.

Direct instructional cost per student credit hour taught is the focal dependent variable examined in this study. Patterns of dispersion and difference in cost across disciplines are examined through a series of analytical lenses that are typically assumed to be major cost factors in the literature. These include institutional mission as characterized by Carnegie institutional classification. The Delaware Study employs the 1995 Carnegie taxonomy—research, doctoral, comprehensive, and baccalaureate institutions. The study also examines the impact of other variables such as highest degree offered within a discipline, and the relative emphasis on undergraduate versus graduate instruction within a discipline.

Using appropriate statistical tools, the relationship of cost to variables such as department size (measured in terms of number of faculty), proportion of faculty who are tenured, volume of student credit hours taught, and personnel expense as a percentage of total instructional costs is examined and measured. Effects of highest degree offered in the discipline, as well as Carnegie institutional classification, are also examined. Cost factors are determined by disciplines, or where more appropriate, groups of disciplines.

Back to Top


The key finding from analysis of multiple years of Delaware Study data is that most of the variance in instructional cost across institutions, as measured by direct expense per student credit hour taught, is associated with the disciplinary mix within an institution.

A secondary factor affecting cost is institutional mission, as related to Carnegie institutional classification. This result may be associated with different faculty responsibilities at institutions with different Carnegie classifications. For example, faculty at research universities, extensively engaged in research activity, might be expected to teach fewer student credit hours at higher costs than faculty at comprehensive institutions. However, Carnegie classification accounts for less of the cost differential between institutions than the disciplinary mix factor.

Figure A reflects actual academic year 2001 Delaware Study benchmarks for 5 of the 24 disciplines analyzed in this study. The benchmarks are mean values for direct expense per student credit hour taught, as reported by participating institutions. They have been refined to correct for outliers and influential cases, and as such, are fair reflections of the average cost of instruction in those disciplines.

In chemistry, average direct expense per student credit hour taught ranges from $181 at comprehensive institutions to $264 at research universities, an $83 spread. The range in English is $28, from a low of $112 at comprehensive institutions to a high of $140 at research universities. Foreign languages range from $131 at doctoral universities to $202 at baccalaureate colleges, a $71 spread, while mechanical engineering ranges from $316 at doctoral universities to $379 at research universities, a difference of $63. And sociology ranges from $100 at comprehensive institutions to $138 at baccalaureate colleges, a spread of $38. These examples in figure A are typical of the ranges in any given Delaware Study data collection cycle.

While the foregoing discussion demonstrates that there is variation within a discipline across institution types, figure B clearly illustrates there is also considerable variation across the disciplines within an institution. Using the same disciplinary examples, at a research university, the difference in direct expense per student credit hour taught between English and mechanical engineering is $239; the difference between sociology and chemistry is $140. Comparable patterns are apparent within the other Carnegie categories as well.

These cost differentials within disciplines across institution types and between disciplines within those types lead to an overarching question. In describing the cost of instruction at higher education institutions, which is the more important factor—the designation of the institution as research, doctoral, comprehensive, or baccalaureate, or the configuration of disciplines that compose the institution?

Hierarchical linear modeling (HLM) is a statistical tool that provides the capability to disaggregate total variance in cost by institution, and by discipline within the institution. HLM helps to explore and describe the dispersion of instructional costs across institutions, and to identify those factors that are associated with the dispersion. The hierarchical linear model constructed in this study demonstrates that most of the variance in cost is at the discipline level within an institution, ranging from 76.0 percent in the 1998 data collection cycle to 82.6 percent in the 2000 cycle.

It can be asserted that Carnegie institutional classification, as a proxy for institutional mission, is tied to at least some of the dispersion of costs at the aggregate institutional level. When Carnegie classification is taken into account in the hierarchical linear model, the dispersion in cost across institutions decreases, and the relative variance due to disciplines within an institution ranges from 81.0 to 88.0 percent.

This important finding underscores that the disciplines that compose a college or university's curriculum, not its Carnegie designation, are associated with most of the dispersion of costs among institutions. This further highlights the distinction between costs, i.e., instructional expenditures, and price, i.e., tuition. Stated plainly, price is a constant for all undergraduates at an institution; chemistry and engineering majors pay the same tuition rate as English and sociology majors. However, the cost of delivering instruction in those disciplines varies widely.

Finding that most of the variation in instructional expenditures is associated with the mix of disciplines within an institution is also important in light of the issues raised in the first part of the congressionally mandated study. Researchers found no apparent relationship between the level of instructional expenditures at an institution and the tuition rate charged by that institution. Results of this analysis of direct instructional expense underscore the difficulty in relating price to cost at the level of the academic discipline. While direct instructional expense per student credit hour taught in civil engineering is three times higher than that for sociology, it is not practical for an institution to charge engineering majors a tuition rate three times that charged to sociology majors.

Indeed, the first report in the cost study found that institutional tuition rates at public institutions are determined largely by state appropriation levels, while competitive market forces shape tuition at private institutions. Neither of these external factors has anything to do with what it costs to deliver instruction in a discipline. Price (i.e., tuition) and cost (i.e., institution expenditures) are not interchangeable constructs.

While the foregoing discussion described the forces that are associated with instructional cost within an institution, the study also focused on those factors that impact expenditures within a discipline. In The Economics of American Universities (Brinkman 1990), Paul Brinkman postulated that the behavior of marginal and average costs can be associated with four dimensions: size (i.e., quantity of activity or output), scope of services offered, level of instruction (for instructional costs), and discipline (for instructional costs).

The analyses in this study determined that 60 to 75 percent of the variation in cost within a discipline or groups of disciplines is associated with specific cost factors consistent with those identified by Brinkman. While the association of a given variable with cost, as measured by direct expense per student credit hour taught, may vary from discipline to discipline, the following general patterns are consistently observed:

  • The volume of teaching activity, as measured by total student credit hours taught, is a major cost factor. Cost decreases as volume increases.
  • Department size, as measured in terms of total number of faculty, is a consistent cost indicator. The larger the department, the higher the cost.
  • The proportion of faculty holding tenure is a cost factor. The higher the proportion of tenured faculty, the higher the cost.
  • The presence of graduate instruction in a discipline increases costs, although the measured effect of this variable on direct expense in this study is smaller than teaching volume, department size, and faculty tenure rate.
  • Similarly, the extent to which expense is associated with personnel costs, as opposed to equipment costs, has less impact on total direct instructional expenditures within a discipline than do teaching volume, department size, and tenure rate.

Figure A. Direct expense per student credit hour taught: Institution type within discipline, 2001
Figure A. Direct expense per student credit hour taught: Institution type within discipline, 2001

NOTE: Data for mechanical engineering at baccalaureate institutions are not applicable.

SOURCE: University of Delaware, The Delaware Study of Instructional Costs and Productivity, 1998–2001.

Figure B. Direct expense per credit hour taught: Discipline within institution type, 2001
Figure B. Direct expense per credit hour taught: Discipline within institution type, 2001

NOTE: Data for mechanical engineering at baccalaureate institutions are not applicable.

SOURCE: University of Delaware, The Delaware Study of Instructional Costs and Productivity, 1998–2001.

Back to Top


While the first report in the congressionally mandated study of expenditures in higher education provided evidence that the price that students pay for an education is largely associated with factors external to the institution, the analyses in this report suggest that the direct cost of providing that education is more associated with internal institutional decisions and priorities.

The mix of disciplines that compose an institution's overall curriculum is associated with direct instructional expense at that institution and, to a smaller extent, its designation as a research, doctoral, comprehensive, or baccalaureate institution. Costs vary more substantially across disciplines within a given institution than they do across institutions within a given discipline.

Within the individual disciplines at an institution, economies of scale have the greatest impact on instructional costs. When given a faculty of fixed size, the more student credit hours taught, the lower the unit cost. Increasing the size of that faculty without a concomitant increase in student credit hour production raises instructional expense. Increasing the proportion of tenured faculty—that cadre of faculty who are better compensated and are essentially a "fixed cost"—will increase instructional expense. And to a lesser extent, introducing or increasing the level of graduate instruction raises instructional costs.

While the data analyzed in this study reflect cost patterns for those 4-year colleges and universities participating in the Delaware Study of Instructional Costs and Productivity only, they nonetheless provide a clear and measurable understanding of cost behaviors within those institutions. These are fresh data, collected at the academic discipline level of analysis, and lend themselves to descriptive statistics that illuminate and clarify cost patterns within those institutions that elect to belong to this data-sharing consortium.

A college or university's tuition rate is tied to what competing institutions charge, i.e., marketplace conditions, and what state legislatures provide as an operating subsidy. Instructional expenditures are tied more to fixed-cost factors, i.e., the mix of disciplines in place at the institution, and within those disciplines, student credit hour production, department size, and tenure rate. This study suggests that depending upon their magnitude, these variables constitute a baseline level for instructional costs within a discipline, and these costs vary less by discipline across institutions than they do among disciplines within an institution.

Most higher education institutions have multiple revenue streams, tuition being but one, to cover instructional costs. It is evident from this study that the factors that are associated with instructional costs are very different from the factors that are associated with tuition prices.

Back to Top


Brinkman, P.T. (1990). Higher Education Cost Functions. In S.A. Hoenack and E.L. Collins (Eds.), The Economics of American Universities: Management, Operations, and Fiscal Environments. Albany, NY: State University of New York Press.

Cunningham, A.F., Wellman, J.V., Clinedinst, M.E., and Merisotis, J.P. (2001). Study of College Costs and Prices, 1988–89 to 1997–98 (NCES 2002–157). U.S. Department of Education. Washington, DC: National Center for Education Statistics.

Horn, L., Wei, C.C., and Berker, A. (2002). What Students Pay for College: Changes in Net Price of College Attendance Between 1992–93 and 1999–2000 (NCES 2002–174). U.S. Department of Education. Washington, DC: National Center for Education Statistics.

Morgan, R.L., Hunt, E.S., and Carpenter, J.M. (1991). Classification of Instructional Programs (NCES 91–396). U.S. Department of Education. Washington, DC: National Center for Education Statistics.

Back to Top

Data source: University of Delaware, The Delaware Study of Instructional Costs and Productivity, 1998–2001.

For technical information, see the complete report:

Middaugh, M.F., Graham, R., and Shahid, A. (2003). A Study of Higher Education Instructional Expenditures: The Delaware Study of Instructional Costsand Productivity (NCES 2003–161).

Author affiliations: M.F. Middaugh, R. Graham, and A. Shahid, University of Delaware.

For questions about content, contact Aurora D'Amico (aurora.d'amico@ed.gov).

To obtain the complete report (NCES 2003–161), call the toll-free ED Pubs number (877–433–7827) or visit the NCES Electronic Catalog (http://nces.ed.gov/pubsearch).


back to top