Presentations

Deconstructing Delta: Explaining Educational Costs through Analysis of the Instructional Portfolio

Categories
Published
of 12
All materials on our website are shared by users. If you have any questions about copyright issues, please report us to resolve them. We are always happy to assist you.
Related Documents
Share
Description
Deconstructing Delta: Explaining Educational Costs through Analysis of the Instructional Portfolio Peter M. Radcliffe Executive Director Office of Planning and Analysis University of Minnesota
Transcript
Deconstructing Delta: Explaining Educational Costs through Analysis of the Instructional Portfolio Peter M. Radcliffe Executive Director Office of Planning and Analysis University of Minnesota Abstract: Assessing the cost of higher education has become a critical policy priority. However, this effort is frustrated by limited information and standards. The Delta Cost Project has attempted to address this by assembling data on costs across sectors and levels of higher education, but their reports create as much confusion as clarity. Analysis shows that the overwhelming majority of variance in educational spending between institutions can be explained by differences in the disciplinary and degree level portfolio. Accounting for these differences allows institutions to benchmark themselves and overseers to evaluate the performance of their institutions. Prepared for presentation at the Association for Institutional Research Annual Forum, June 2-6, 2012, New Orleans, LA Deconstructing Delta: Explaining Educational Costs through Analysis of the Instructional Portfolio The escalation of the cost of higher education has captured the attention of educational leaders, institutional governing bodies, government officials, media commentators, and the public at large. Over the past decade, average published tuition at public four-year institutions has increased by 5.6 percent annually after controlling for inflation (Baum and Ma, 2011, p. 13). Only 22 percent of the general public believes that a college education is affordable (Pew Research Center, 2011). Against this backdrop, President Obama issued a warning to higher education institutions in his 2012 State of the Union Address, saying to applause, so let me put colleges and universities on notice: if you can t stop tuition from going up, the funding you get from taxpayers will go down (Obama, 2012). Archibald and Feldman (2011) note that higher education is actually wrestling with two related challenges: increasing prices and increasing costs. In economic terms, price is the amount paid by consumers for a product or service, while cost is the amount paid by suppliers to produce that good. In any economic activity, there is generally some difference between price and cost. Because public subsidies in a number of forms provide a significant portion of that difference in higher education, however, the relationship between price and cost is an important public policy issue. Since most higher education in the United States is provided by non-profit organizations, whether publicly or privately owned, the difference between price and cost is not due to profitseeking, but rather costs exceed price for most institutions, with the difference made up from state appropriations, private gifts, investment earnings, and other sources. Increasing prices, therefore, are driven by both increasing costs and insufficient subsidies. The discussion that follows will be focused exclusively on understanding differences across institutions in costs. A vigorous public debate is underway regarding the reasons for the increasing costs of higher education. The Higher Education Price Index, produced by Commonfund, which tracks prices on a market basket of goods purchased by higher education institutions, just as the Consumer Price Index tracks goods purchased by typical consumers, shows that over time, the prices colleges and universities pay for their inputs have risen at a rate higher than general inflation (Commonfund). Archibald and Feldman (2011) argue that this is driven by higher education s heavy reliance on highly educated labor, and the inability to increase productivity inherent in face-to-face education delivery without continuously increasing class sizes. As productivity rises elsewhere in the economy, higher education must compete with those industries for access to this pool of labor, driving up wages despite the lack of productivity growth. This is the cost disease phenomena identified by William Baumol and William Bowen. Others, such as Richard Vedder (2004) argue that institutions themselves are at fault for increasing spending in non-instructional areas. A thoughtful assessment of whether institutions are excessively expensive, however, is frustrated by limited information and standards. One attempt to explore the causes and consequences of rising costs in higher education is the Delta Project on Postsecondary Education Costs, Productivity, and Accountability, known as the Project Cost Project. Funded by the Lumina Foundation for Education, the Delta Cost Project produces a longitudinal aggregation of data on college revenues, expenditures, and enrollments, as well as reports based on that data. The dataset assembled by the Delta Cost Project is based primarily on the Integrated Postsecondary Education Data System (IPEDS), managed by the National Center for Education Statistics in the federal Department of Education, supplemented with other descriptive and economic data. Some alterations in the data have been made by the Delta Cost Project to simplify comparisons between public and private institutions, and to adjust for definitional changes over time. While these changes make working with the data set easier, they come at some cost to transparency and to efforts to link the data set to additional information about institutions. In addition, methodological and reporting decisions made by the Delta Cost Project create as much confusion as clarity. While these decisions may seem logical on the surface and appear intended to make their assembled data accessible and relevant for public policy decisions, they create distortions that have serious consequences for the conclusions drawn and the decisions informed by those observations. IPEDS tracks institutional spending across a set of categories, defined by federal accounting standards (Governmental Accounting Standards Board, or GASB, for public institutions, and Financial Accounting Standards Board, or FASB, for private and some public institutions). Those categories include spending in the primary mission areas of instruction, research, and public service, along with support categories attached to the delivery of the mission, including student services, academic support, institutional support, and the operation and maintenance of plant. IPEDS reports total and per full-time equivalent (FTE) student spending in each of these categories by institutions. These measures are somewhat limited, however, because they do not connect the support costs to the individual mission activities. Addressing this limitation through the creation of a fully loaded version of instructional costs is one of the core goals of the Delta Cost Project. The primary variable of interest in the Delta Cost Project database is education and related costs, or E&R. The concept of E&R expenditures is a fully loaded measure of educational costs, including instructional and student services expenditures and an estimate of the share of other support costs that can be attributed to the educational mission. It is extremely challenging for an individual institution to segregate support expenditures by mission activity. For example, does a campus system support teaching or research? Does landscaping support instruction or public service? In practice, of course, some overhead can be directly attributed to a specific mission activity, but much of it is shared. If these challenges are extraordinary for an individual institution, attempting to find a straightforward means to make these judgments across institutions is simply impossible. The task for analysis then is to devise a relatively simple, consistent methodology for approximating the appropriate share of overhead costs to attribute to each mission function for an institution, without needing extensive and currently unavailable financial details. The Delta Cost Project s intuitive solution is to add together total expenditures coded as instructional or for student services, and split the remaining support expenditures by the relative proportions of expenditures for education (instruction plus student services), research, and public service. Each mission area s share of the total overhead, therefore, is determined by the relative amount of spending in that area. While this will obviously not be precise for any institution, it is explainable, replicable, consistent, and feasible with existing data. The Delta Cost Project reports E&R spending in three primary ways: as a raw total, per FTE student, and per degree or completion (combined certificates and degrees). Spending is obviously tied to the size of an institution, so reporting by FTE student or degree granted are simple means to attempt to standardize expenditures and make comparisons about relative efficiency, or at least resource intensiveness. These standardization attempts, however, are only effective to the extent that all FTE students (or degrees, or completions) are either highly similar to each other, or are distributed similarly across institutions. If the costs associated with educating some types of students or with providing some instructional programs are different from each other, and those programs are unevenly distributed across institutions, those standardization approaches will create false equivalencies, inaccurately implying that some institutions are more or less efficient than they truly are. Providing a student with a full year s academic load obviously consumes more educational costs than one taking a single course. Representing this difference is the purpose of adjusting for FTE enrollments. The standard FTE calculation also recognizes the differences in the intensity of undergraduate and graduate study, setting the full-time bar for undergraduates at 30 credits annually while lowering it for 20 credits for graduate and professional students. As a means for equating educational spending per student, however, this approach only works if graduate and professional credits cost 50 percent more on average to provide. That is, 20 credits of graduate education cost roughly the same as 30 credits of undergraduate education. If not, differences between institutions in the composition of their student bodies will appear as differences in the per-student cost of education. To see this, consider two identically sized institutions, each with a single undergraduate program and a single graduate program, where the costs per FTE student are the same across institutions at the same level, but the costs of education at the two levels are different from each other. For simplicity, assume each institution has 100 FTE students, a per- FTE undergraduate cost of $1,000, and a per-fte graduate cost of $2,000. At Directional State University, a predominantly undergraduate institution, we assume 75 undergraduate FTE and 25 graduate FTE, for our total of 100. At University Institute, a more graduate-focused institution, the FTE students are split evenly, with 50 undergraduate FTE and 50 graduate FTE. As shown in Table 1 below, even though costss at the program level are identical, the differing distribution of students at the two institutions means that overall average costs at University Institute are 20 percent higher than at Directional State University. Table 1 Illustration of impact of differing enrollment distributions on average costs UG UG Grad Institution Directional State University University Institute FTE Cost/FTE $1,000 $1,000 FTE Grad Cost/FTE $2,000 $2,000 Total FTE Total Cost Total Cost/ /FTE 100 $125,000 $1, $150,000 $1,500 % Difference: 20% We can examine how this theoretical issue plays out in real institutions by contrasting the E& &R spending per FTE of public research universities with thee percentagee of their student body that is at the undergraduate level. What appears is a clear trend that replicates the theoretical example of the table above. As the percentage of studentt FTE at the undergraduate level increases, the E&R expenditure per FTE declines, as shown in Figure 1 below. Figure 1 Public Research University E&R Expenditures 2008 vs. Undergraduatee FTE Percentage Measuring E&R expenditures per degree or per completion likewise suffer from the same limitations. For example, an associates or masters degreee program is expected to take significantly less time than a bachelors or doctoral program. So long as the distribution of degree levels is the same across institutions, this would not be a serious obstacle. However, in practice the degree offerings of institutions by level vary widely. To control for this, in the model below, total E&R expenditures were regressed on total degrees awarded by level for all public four-year colleges and universities that have a Carnegie Classification (Basic, 2010) as a research, doctoral, or masters institution (classification numbers 15 through 20). Although simple, the model accounts for nearly 87 percent of the variance in total E&R spending across this set of institutions (adjusted R-squared value of.8656). Table 2 Total E&R expenditures by public research, doctoral, or masters institutions by degrees awarded Variable Coefficient Std Error t Score P t CI Lower CI Upper Associates $20,558 $28, $(35,286) $76,401 Bachelors $39,248 $4, $31,074 $47,423 Masters $16,671 $11, $(5,589) $38,932 Doctorate $588,288 $47, $495,548 $681,028 1st Professional $227,888 $37, $153,874 $301,901 Constant $26,600,000 $6,768, $13,300,000 $39,900,000 As with the distribution of degree programs by level, if degree programs in different fields were all equally expensive to offer, knowing the distribution of those programs by field at an institution would not be necessary to understand the relative efficiency of an institution in delivering its educational mission. As it happens, that question has been examined. In 2009, the State Higher Education Executive Officers, published a study of field and level-specific enrollments and expenses from four states (Florida, Ohio, Illinois, and New York). The study found marked differences in the distribution of expenses and the distribution of enrollments, as measured by student credit hours, the base for determining FTE totals. For example, while graduate and professional enrollments accounted for only 20 percent of the total student credit hours, they accounted for 34 percent of the total instructional costs. Likewise, the proportion of total instructional costs related to health science education vastly exceeded the proportion of total credit hours taken in those fields, while most liberal arts fields represented a smaller share of expenses than they did of enrollments. Clearly, both the level of instruction and the field of instruction are critical drivers in educational costs. To address this, it is necessary to examine the portfolio of degree programs at an institution. The ideal measure of student demand for instruction and support services would most likely include intensity of enrollment, level, and field for all students, but that information is not available through public sources. What is available through IPEDS, however, is the number of degrees awarded at each level and in each field, broadly defined. Degree field is tracked in IPEDS using the Classification of Instructional Program (CIP) taxonomy. CIP codes have six digits, in three pairs. Fields are identified through these codes in the CIP hierarchy at three, nested levels: twodigit, four-digit, and six-digit. At the two-digit level, the broadest grouping, there are thirty-eight separate codes, listed in Table 3 below. Table 3 CIP 2010 two-digit codes CIP # Program Description 1 Agriculture, agriculture operations, and related sciences 3 Natural resources and conservation 4 Architecture and related services 5 Area, ethnic, cultural, and gender studies 9 Communication, journalism, and related programs 10 Communications technologies/technicians and support services 11 Computer and information sciences and support services 12 Personal and culinary services 13 Education 14 Engineering 15 Engineering technologies/technicians 16 Foreign languages, literatures, and linguistics 19 Family and consumer sciences/human sciences 22 Legal professions and studies 23 English language and literature/letters 24 Liberal arts and sciences, general studies and humanities 25 Library science 26 Biological and biomedical sciences 27 Mathematics and statistics 29 Military technologies 30 Multi/interdisciplinary studies 31 Parks, recreation, leisure, and fitness studies 38 Philosophy and religious studies 39 Theology and religious vocations 40 Physical sciences 41 Science technologies/technicians 42 Psychology 43 Security and protective services 44 Public administration and social service professions 45 Social sciences 46 Construction trades 47 Mechanic and repair technologies/technicians 48 Precision production 49 Transportation and materials moving 50 Visual and performing arts 51 Health professions and related clinical sciences 52 Business, management, marketing, and related support services 54 History Data on degrees awarded by field are not included in the Delta Cost Project database. However, they are tracked in the IPEDS system from which the institutional expenditure data is drawn, so it is possible to supplement the Delta Cost Project database with this data. As noted previously, because the Delta Cost Project s adjustments to the underlying IPEDS data are difficult to replicate from the outside, this analysis is based on data directly from IPEDS, applying the Delta Cost Project s approach to calculating E&R spending. In IPEDS, the degree award data is grouped at five levels associates, bachelors, masters, doctoral, and professional and across the thirty-eight two-digit CIP classifications at each level. This produces 190 possible combinations of field and level. However, there are many combinations at which no institution awarded a degree in , the last year for which the data is available. In addition, there are several combinations where so few degrees are offered that they cannot be modeled cleanly. To address these problems, combinations where no degrees were awarded we dropped from the model, and combinations where fewer than ten degrees were awarded in that level and field were combined with degrees awarded in the same field at the nearest degree level. These two adjustments reduced the total number of combinations in the model to an even one hundred. Regressing total E&R expenditures on these degree award level-field combinations explains over 95 percent of the variation in institutional spending among public research, doctoral, and masters institutions. While conceptually the coefficients represent the expenditures associated with a single degree awarded in the relevant field at the relevant level, in practice each institution produces a diverse portfolio of degree awards in any given year, and it is the combination of all of the programs that produce those degrees that lead to a particular spending level for that institution. The value of the model, therefore, is not in attempting to isolate the cost of instruction in any particular program (that would be more effectively accomplished through datasharing arrangements like the Delaware Instructional Cost Study), but rather in providing an estimate of the expected level of expenditures for an institution given their mix of degree programs. The table below shows the actual and predicted E&R expenditure from the model for the University of Minnesota Twin Cities and its comparison group (for the full model output, see the Appendix). Table 4 Predicted vs. actual E&R expenditures for U of Minnesota Twin Cities comparison group Institution Actual E&R Predicted E&R Variance % Var Penn State U $1,366,000,000 $1,210,000,000 $156,000,000 13% U of California Los Angeles $1,537,000,000 $1,380,000,000 $157,000,000 11% U of Washington Seattle $1,284,000,000 $1,180,000,000 $104,000,000 9% The Ohio State
Search
Similar documents
Related Search
We Need Your Support
Thank you for visiting our website and your interest in our free products and services. We are nonprofit website to share and download documents. To the running of this website, we need your help to support us.

Thanks to everyone for your continued support.

No, Thanks