patterns and trends in expenditurespatterns and trends in expenditures within wide limits,...
TRANSCRIPT
This PDF is a selection from an out-of-print volume from the NationalBureau of Economic Research
Volume Title: Buying the Best: Cost Escalation in Elite Higher Education
Volume Author/Editor: Charles T. Clotfelter
Volume Publisher: Princeton University Press
Volume ISBN: 0-691-02642-4
Volume URL: http://www.nber.org/books/clot96-1
Conference Date: n/a
Publication Date: January 1996
Chapter Title: Patterns and Trends in Expenditures
Chapter Author: Charles T. Clotfelter
Chapter URL: http://www.nber.org/chapters/c11279
Chapter pages in book: (p. 82 - 138)
CHAPTER 4
Patterns and Trends in Expenditures
Within wide limits, institutions can adjust towhatever amount of money they are able toraise.
Howard Bowen, 19801
THIS CHAPTER presents information on trends in real expendituresfor the sample institutions. The data were taken from detailed finan-cial information provided by each of four institutions: Duke, Har-vard, Chicago and Carleton. Although efforts have been made tocompile this information in comparable forms, the differences amonginstitutions in mission, organization, and accounting practices makeit inevitable that the presentations will not be uniform. The aim ofthe study was to collect data at five-year intervals beginning in 1976/77,but the lack of machine-readable data for the early part of the studyperiod made it impossible to collect a detailed, consistent data set forany institution. However, other data were used in an attempt to vali-date the trends observed in the detailed data. As noted in chapter 1,expenditures for professional schools and medical centers were ex-cluded from the analysis, as were the expenditures of auxiliaries.The present chapter begins by considering the source of funds andthen turns to how expenditure data may be usefully organized andpresented. Tabulations for the sample institutions follow.
WHICH EXPENDITURES ARE WE INTERESTED IN?
In considering the problem of rising outlays, all expenditures arenot of equal importance, and some changes other than in expendi-tures are as significant as the expenditures themselves. In this sec-tion, I make four arguments. First, outlays as reported are not iden-tical to costs, but for the most part, the differences are not crucial.Second, changes in internally financed expenditures are of more sig-nificance to the current concern over rising outlays than are thosefinanced by grants and contracts. Third, expenditures may rise forthree conceptually different reasons, and each has different implica-
WHICH EXPENDITURES ARE WE INTERESTED IN? 83
tions for the assessment of the increases. Fourth, some "real" effectsthat are either a cause or a by-product of rising expenditures areworth special attention.
Costs, Prices, and Outputs
According to the economics textbook theory of the firm, an enter-prise is seen as taking inputs, doing something to them, and produc-ing one or more outputs. Both inputs and outputs may be goods orservices. The cost of producing is the forgone economic benefit ofthe resources used, which, under ideal conditions, is equal to theactual outlays of the firm. This textbook equivalence is not a badstarting assumption in the case of higher education. Probably thelargest category of economic cost that is not reflected in publishedrecords of university expenditures is the opportunity cost of the landon which the campus is located. Although universities own andtherefore do not pay rent on campuses, they forgo significant in-come by using them rather than renting them out. A similar pointapplies to the physical plant, with the additional cost being the dif-ference between the economic depreciation and the amounts spentto maintain or refurbish that plant. Another element in true eco-nomic cost that always is omitted from reported expenditures is thevalue of time, including the time of students as well as that donatedby trustees and other volunteer workers.2 However, the available ex-penditure data cover most of the important economic costs of uni-versities.
Source of Funds
For the purpose of considering the causes and implications of therise in spending by colleges and universities, it is clear that not allexpenditures are created equal. They differ in both their financialimpact and their interpretation. One of the basic motivations for in-terest in expenditures is the increase in tuitions that students andtheir families must pay. Thus, expenditures financed by tuitions willhave a very different interpretation than those financed by grantsand contracts. In order to assess the importance of revenue sources,it is helpful to consider the four principal sources of college anduniversity funding: (1) unrestricted revenues (tuition, unrestrictedannual gifts, and unrestricted investment income); (2) restricted giftsand endowment income; (3) grants and contracts; and (4) fees and
84 PATTERNS AND TRENDS IN EXPENDITURES
charges for specific services. In accordance with the practice of"fund accounting" that is common in private colleges and univer-sities, every expenditure is assigned to revenue of one of these types.Because of this close correspondence between revenues and expen-ditures, it is useful to use the source of revenue as a guide to cate-gorizing expenditures.
One type of expenditure that quickly may be separated fromothers is that for auxiliaries, which are a class of activities that aredesigned to have their own dedicated fees. Examples include book-stores, dormitories, dining halls, and student health services. By con-vention, these activities normally are excluded or are treated sep-arately in analyses of higher education, largely because their functionsare easily distinguished, are not integrally academic, and may differfrom institution to institution owing simply to differences in themake-or-buy decision (the degree to which service activities are con-tracted out or to which the private market is allowed to provide ser-vices). It makes sense to follow this conventional approach in thepresent case.
Among the remaining three categories, the most important dis-tinction lies in the degree of discretion that an institution has overhow to use those revenues. In the case of grants and contracts, it hasvery little. Revenues from these sources cover direct costs and indi-rect costs (how well actual costs are covered is a different, althoughrelated, point), and institutions must record rather explicitly thepurposes for which the revenues are spent, both through budgetand grant reports and through audits of indirect cost rates. In thepresent study, the direct expenditures funded by grants and con-tracts will be referred to as externally funded expenditures. This is notto say, however, that no spillover effects exist between externallyfunded and internally funded spending. Grants and contracts prob-ably are used occasionally to pay for activities that the institutionwould have financed out of internal funds had the external fundsbeen unavailable. More obviously, external funding also generatesadministrative and infrastructure costs that are impossible to assignto outside grants. These expenditures are paid for out of generalrevenue, one source for which is the overhead recovery from allgrants and contracts. Therefore, although most administrative costsare identified as unrestricted, a portion of them certainly arises be-cause of externally funded grants and contracts.
The two categories that remain—spending financed by unre-stricted revenues and those financed by restricted gifts and endow-ments—will be referred to together as internally financed expendi-tures, on the theory that institutions exercise considerable discretion
WHICH EXPENDITURES ARE WE INTERESTED IN? 85
over their use. Institutions, by definition, have considerable latitudeto decide how to spend unrestricted revenues. Within the limits setby laws and regulations, private institutions are accountable only totheir governing boards and to the test of the market as to how theyspend unrestricted funds. Institutions have somewhat less discretionin the case of expenditures financed from restricted gifts and en-dowments. For the purpose of this study, these expenditures aregrouped together with those funded by unrestricted funds as inter-nally financed expenditures, based on the reasoning that, in practice,they are much more like unrestricted current expenditures than ex-penditures on grants and contracts. Because endowments grow outof gifts that the institution accepted and usually solicited, it does notseem unreasonable to believe that most of their conditions are con-sistent with the institution's aims. To be sure, some gifts are acceptedreluctantly, and institutional missions do change over time, but ingeneral, however, the income from most endowments is used forpurposes that continue to be fully consistent with institutional goals.Another reason why it is sensible to lump unrestricted and endowedexpenditures is that one often is a substitute for the other. Unre-stricted funds often are needed, for example, to supplement the in-come of restricted endowments established for endowed chairs;funds from the endowments attached to restricted scholarships typ-ically are used to bolster an institution's commitment to meet thefinancial need of students.
The Interplay Between Real and Financial Effects
At least as important as the increases in the level of expendituresthemselves and shifts in their composition are the accompanying al-tered nonmonetary quantities. As noted in chapter 2, colleges anduniversities have considerable latitude to respond to changes in sucheconomic circumstances as increasing costs. They have the option ofresponding to changes in the relative cost of inputs by adjusting in-puts in precisely the way that textbooks describe the behavior of thefirm, that is, by economizing on the use of one input by substitutinga less expensive one. For example, a university can minimize theneed to hire relatively expensive faculty by shifting some tasks tradi-tionally performed by faculty, such as advising or departmental ad-ministration, to other employees. Another widespread form of sub-stitution that has occurred in many business operations, includingthose in universities, is the substitution of computers for clerical em-ployees, a trend examined in chapter 6.
86 PATTERNS AND TRENDS IN EXPENDITURES
But there are other available avenues of flexibility that are nottypical of those in textbook discussions. One alternative—an optionopen to almost any economic actor—is to save money in the shortrun by undertaking actions that may well be unwise in the long run.An institution can save money in the near term, for example, bydeferring the maintenance of its buildings and other physical assets.Similarly, it can increase its revenues in the short run by raising thespending rate from its endowment and other financial assets. It alsomay have some flexibility in the extent to which it uses grants andcontracts to cover what otherwise might be considered ordinary ex-penditures, such as faculty salaries. Or it may accept gifts that willgenerate costs in excess of the additional revenue generated. An-other class of response to rising costs is, of course, simply that ofallowing quality to be degraded, such as by increasing the size ofcourses without providing concomitant improvements, by hiring lesstalented faculty, or by reducing the amount of financial aid awards.In light of these possibilities, it is useful to combine an analysis ofchanging expenditures with attention to other important changes inthe institutions being studied. Thus, to supplement the attention fo-cused on expenditures, chapters 5 through 8 examine in part non-monetary quantities.
SPENDING, BY TYPE AND DEPARTMENTAL GROUP
Basic to any evaluation of rising expenditures is information aboutwhich categories have experienced the greatest increases. In thisstudy, the two dimensions of classification used to define categoriesof expenditures are (1) the organizational entity, which correspondsroughly to function; and (2) the type of expenditure. Although acontemporary university may contain hundreds of separate adminis-trative entities, including academic departments, institutes, pro-grams, centers, service divisions, and the like, I have tried to mini-mize the number of departmental groups. Academic departments,the entities directly responsible for most teaching and research, areseparated from purely administrative entities, and the latter are fur-ther divided between academic units, such as those headed by pro-vosts and deans, and other administrative offices. Given the differ-ent functions of the various entities within a university and thecontinuing interest in the allocation of resources between adminis-trative and academic units, this functional classification is essential toany analysis of rising expenditures in universities. An additional ad-vantage of dividing expenditures by entity in this way is that it allows
PATTERNS AND TRENDS: DUKE 87
the apportioning of general university expenses. General adminis-trative functions, such as physical plant or personnel, provide ser-vices to all parts of a university, so it is reasonable to assign only aportion of these costs to the arts and sciences enterprise that is thesubject of this study.3
The other dimension of classification is the type of expenditure.Because universities are labor-intensive enterprises, the largest broadcategory of expense consists of salaries, wages, and related compen-sation. In all the expenditure tables in this chapter, fringe benefitsare added to salaries to yield total compensation.4 Moreover, becausethere are several distinct categories of labor in universities, it is use-ful to divide this compensation further between faculty and severalother classes of employees. Most, but not all, faculty are classified asregular faculty, that is, those who have tenure or are on the "tenuretrack" and who hold the title of professor, associate professor, orassistant professor. In addition to payments to labor, other types ofpayments that are distinguished include scholarships, purchases ofgoods and services, general operating costs, and capital expendi-tures. The figures used for capital expenditures are based on actualcapital outlays on structure and equipment. Averages of severalyears were used, rather than one year's outlay, because of the inher-ent lumpiness of capital spending.5 This approach is best illustratedby turning to the sample universities.
PATTERNS AND TRENDS: DUKE
Detailed Expenditures
Table 4.1 presents a tabulation for Duke that uses the describedmethod of categorization. The top section shows the internally fi-nanced expenditures for Duke over the period 1983/84 to 1991/92,in constant dollars, with departments divided into 13 groups andexpenditures split into 14 different categories, by type.6 For eachdepartmental group not entirely in arts and sciences, a portion of allof the group's expenditures was allocated to arts and sciences, usingestimates provided by administrators at each university. The sameproportions for each line of the table were applied to all years.7 Al-though the question of allocation of general expenditures is impor-tant, the specific formula used is of secondary importance for thepurposes at hand. The primary aim in these calculations is to use arather simple and transparent method for allocation and to make itrelatively easy to calculate what difference it would make if other
88 PATTERNS AND TRENDS IN EXPENDITURES
TABLE 4.1Expenditures, Levels and Changes, by Departmental Group and Type: Duke
(Internally Financed Funds Only; in Thousands of 1991/92 Dollars)Type of Expenditure
HumanitiesSocial SciencesNatural SciencesEngineeringLibraryStudent ServicesPlantAdmissions and Finan-
cial AidArts and Sciences
AdministrationProvostAlumni Affairs and
DevelopmentGeneral Administration
SubtotalAthletics
Total
HumanitiesSocial SciencesNatural SciencesEngineeringLibraryStudent ServicesPlantAdmissions and Finan-
cial AidArts and Sciences
AdministrationProvostAlumni Affairs and
DevelopmentGeneral Administration
SubtotalAthletics
Total
RegularFaculty
10,0908,1709,3054,221
000
0
2,049597
03
34,4350
34,435
OtherFaculty
2,8821,0341,331
3210
700
0
80089
035
6,5610
6,561
Compensation
AdministrativeStaff
1,6431,0842,317
4693,2912,563
292
861
4,3181,537
2,9493,046
24,3712,935
27,306
NonexemptWorkers
1,3131,3212,591
5052,6991,0221,220
607
1,814231
1,1965,250
19,769456
20,226
Students
60485921
24093
1
151
26951
27117
1,13780
1,217
Changes in Expenditures 1983184 to 1991192
RegularFaculty
3,6422,4373,4311,812
000
0
890-239
- 8- 3
11,9610
11,961
Type of Expenditure
OtherFaculty
2,099629804270
0650
- 9
42029
033
4,3390
4,339
Compensation
AdministrativeStaff
854518
1,083206740990
38
144
2,728906
1,244857
10,3071,081
11,388
NonexemptWorkers
542372632
87356233109
303
70726
4201,5495,337
1485,485
Students
29- 5 0- 9 0
- 17463
- 3
147
7644
248
32213
335
ProfessionalServices
204104
6819
226830
1
39825
14940
1,315117
1,433
ProfessionalServices
10182
- 71
7025
- 1
1
246- 8
4- 1 9 4
32013
333
ContractWork
10883
1291637
135110
109
85691
624440
2,738163
2,901
ContractWork
35424311
- 1100
23
85
47377
540218
1,646- 1 4
1,633
Source: Calculations using unpublished data from Duke University.
213
PATTERNS AND TRENDS: DUKE 89
TABLE 4.1 {emit.)
Computers
4286
2539388
4072
FinancialAid
3,5132,7903,9991,477
12244
0
Supplies
333304943199241175118
General OperatingExpenses
1,2501,129
902183360112867
Capital
1401,4783,894
1843,7891,819
389
Maintenance
5415
160
20760
1.050
Residual
74
- 02600
Total
21,63917,65025,794
7,77111,1966,7854.048
16,124 67 686 22 18,842
405773
573-4622,471
82,479
3,22941
127
31,4574,052
35,508
518142
2723,0606,373
5266,900
4,91650
1,587-7,635
4,4093,5667,974
80454
5528
12,635129
12,764
13434
46333
2,017386
2,402
161
1- 5 3- 1 5
2- 1 3
20,5253,715
7,4804,229
149,67312,420
162,093
Financial General OperatingComputers Aid Supplies Expenses Capital Maintenance Residual
- 2 8 0 10,873 50 305 20
Total
- 6 2- 412360
- 4 478
2
2,2331,7572,216
72911
1970
166116173263462
- 1 8
7377
12118
186-489
- 6 4
-1091,4042,957
-1,3481,190
98577
26- 1 0
90- 3 615023
232
12- 3
9- 1 6
1- 1
0
10,3067,298
11,5841,8182,7662,331
395
11,640
-528710
393174620- 2
618
1,543- 1 6
120
19,5642,114
21,678
168111
90402
1,379142
1,522
3,007-361
236-2,458
1,244589
1,833
49426
- 1 0-2045,464-3185,146
210
126
534214747
181
1- 3 6- 1 4
-496-510
10,2431,316
2,935392
63,0243,485
66,508
90 PATTERNS AND TRENDS IN EXPENDITURES
reasonable weights were used. In all likelihood, the specific alloca-tion rules will make little difference in any assessment of overalltrends and patterns of expenditures.8
As noted in appendix 4.1, this table takes into account inter-departmental charges, such as a payment by the English departmentto the purchasing department for a box of pencils. Consequently,the totals in the far-right column of the table more nearly reflectresource use by the divisions of the university than they would iftransfers were not counted.9 Mindful of these caveats and limita-tions, one can view the tables as a rough summary of the university'sexpenditures for arts and sciences. The top part of the table givesthe breakdown of total spending for the 1991/92 academic year.Duke spent an estimated $150 million in the arts and sciences andengineering, and $162 million if athletics is included. (Recall thatprofessional schools, the medical center, and most auxiliaries are ex-cluded.) By departmental group, the natural sciences accounted forthe largest share, $26 million, or about one-fifth of the total. By typeof expenditure, scholarships and graduate assistance comprised thelargest category, at $36 million (including athletics). Of this figure,scholarships for undergraduates (excluding athletic scholarships)roughly corresponds to the entry in that column for admissions andfinancial aid, or $16 million, whereas the amounts in the variousdepartmental groups correspond to graduate assistance.10 Next insize was faculty compensation, at $34 million, followed by compensa-tion for administrative staff and nonexempt (hourly) workers. Spend-ing for computers accounted for $2.5 million, and other capital ex-penditures totaled $13 million."
The bottom half of Table 4.1 shows the changes in expenditures,in constant dollars, over the eight-year period between 1983/84 and1991/92. The subtotal on the far-right column indicates that, exclud-ing athletics, total spending increased $63 million during this period,an increase of more than 70 percent. What is apparent is that a largeshare of the total increase in spending can be attributed to just a fewcategories.12
To make it easier to examine the broad pattern of changes in ex-penditures, Table 4.2 presents levels and changes only for the subto-tals for the rows and columns in Table 4.1. Thus, total spending isbroken down once by departmental grouping and again by type ofexpenditure. For each subtotal, the table shows total spending forthe beginning and ending year of the period and other measures ofshare and growth. Columns 3 through 5 give a simple way to showthe relative importance of the various categories in contributing to
PATTERNS AND TRENDS: DUKE 91
the overall increase in spending. The contribution of any category is,by definition, the product of its share of the total, shown in the thirdcolumn, and its percentage increase, shown in the fourth.13 Columns3 through 5 in the bottom part of the table show, for example, thatfaculty salaries accounted for the largest share of spending in 1992(23 percent), and that the salaries of nonregular faculty showed thefastest growth (66 percent). The category with the largest contribu-tion, however, was financial aid, contributing 13 percent of the total42 percent increase, or more than 30 percent of the total growth.None of the departmental groups shown in the top section of thetable contributed much to the overall increase. Columns 6 and 7express these increases in terms of annual growth rates, both nomi-nal and real.
Internally Funded Versus Externally Funded Expenditures
Because the implications of changes in internally funded expendi-tures may be radically different from those applying to changes inspending from external funds of various types, it is important toinvestigate the patterns of funding underlying these expenditures.The remaining columns of Table 4.2 focus on differences betweeninternally funded and externally funded expenditures. The eighthcolumn gives, for 1986/87, the percentage of the category fundedwith external funds. The departmental group most dependent onoutside funding is the natural sciences (more than 40 percent exter-nal funding). By category, wages paid to students receive the largestshare of outside support, reflecting the work-study program that isan important part of the federal role in financial aid, but which isnot shown in the financial aid category. The last four columns showin more detail changes over the eight-year period in expenditures,by type. Overall, internally funded spending decelerated slightly be-tween the 1984-87 and 1987-92 periods, while the growth in exter-nally financed spending remained steady.
To examine further the revenue sources for these spending in-creases, Table 4.3 shows how the distribution of spending by type offunding changed over the eight-year period for which detailed datawere available. Although there was a dip in the share of federalfunding for arts and sciences at Duke in the first part of the period,the share actually increased from 1987 to 1992.14 By 1991/92, aboutone-fifth of arts and sciences spending at Duke was supported byfederal funds, 5 percent was based on sponsored research from
Q
c
9 ^
o ^cu C
I2o3 5
cu
X
w
as
ON
25 ^O OS
oo ONON -~,
0000
oo ON
ON
O
oo
CO
o>m
mmCM
COCM
O5CO
00CO
oCM
CM
00
m
m
1
o>o<y>
CO
oCOo
o>Oi
CO
COCO
O5COCM00
oCMCM
X00
oo
ooo
^
1
ooo
o
O5
CM
o1
COCO
COCO
oo>COCO
CMJ>CO1
CM
1>t
ooo
COCO
ooo
CM
mCO1
CM
COO5O>CM
CO
oCO
OCOI>
r~. co co CM o; co CM—i co o —< co o in
co -< inin CM i>
m oo —i CM CM
o o o o o© o o © © © ©
m co in t1-^ CM CM co
oo r —iP P P© © ©
CMCO
© mm co
i> m co© © ©
Tf CM•—I ©
© © © © © © ©
CO
CM©
CO
©
m©
CO CM
© 00
O200
_
X!
CM
CO©
CO
in co~-< 00^f CM
—I ©
in cot~- CO
©
CM
mCO
©
O>00
CO
m
©©
O)
CO_^
^^O>
m
©CM
• ^
COCM
COCM
CO
on
©
mm
©©
Oi
o>
CMCO
©
CM
I"—
CM
m©
00CO
© CM© Tt1
O5 CM
© " f
CO ©© ©
a3C
ac
iart
me
CO_^CM
co
^
t«
.H,-y
uman
Q
©
t/3
uc_cy'u
)cia
l S
00
mCM
CM—)H
t«CUuccu'uC/J
atur
al
Z
00
©CO
bec
agin
et
w
CM__
00
ibra
ry
00CO
m
CU
*
cu
uden
t
©
CO
ant
t/3
CO
dmis
si
0000
CM
T3
_.2'u
Fina
n
cuuca'uC/5
-o
rts
an*
<
m©CM
CO©
C0* [*"'.
-
'S
Adm
i
i>-CO
CM
rovo
st
•o
03
lum
ni
<
m
m
cCUE
c
a<50
Dev
elen
eral
O
CM
00CO
c0
istr
ati
O)
COCO00
Tot
al
aVx
mCO00
CMm
©CO00
mm
CO00
CDCM
CM
CMCOCM
^
CM
mCD
CO
m
m©
m
00co
mCD
l
©co
CO
1
CM©1
CM
1
com
CDCM
CD
1
m©CD
CMcoCM
COCMCO
COCO
m
1
co00
CO
CO
CM
<*
mCO
COCM
00CO
CM©CO
Xco©1
©©
CM
00
m
CD
CO
CM
00©
m
CO
00
CM
m
<*CM
1
i
m
CM
mCM
00CO
m©
CO
-14
00
©
00
1
00©1
CD
m1
m
CM
CM
CO
©CD
©CO
CM00
CM
oo
CO
m
CM
mCO
mo
CT5CD
CDi>
Oico
00
CM mco
CMm
m, i
CO
r-
CDCO
00
m
CMCM
CD
OCO
00
om
00CO
mo
00CD
00 CO
o oo o
CM CMCM
© © ©O -^ O CO ^q o o —i p© o o o o
—i -<f © CM© © © "*f© © © ©
mCO©
coCM©
CO
CDCD©
©o
CDCO
CM
©
CD
©
CM
CM©
coi-H
©
00
00CM©
©©
_|
CM©
©©
co
©CO©
CM©©
CM
CM©
CM©©
mCM
CMCD©
r—1
CM©
mco
CMCM©
©©
CO
00CM©
CO©©
CO
©
00©©
CDCM
CDCM©
r—t
o©
©CM
CM
©
©O_ i
I>O^
P3
DV.3Q
so
-oV
ire<uu
inr
v S—' c
oq © —;© i—ti—t
V>
'2•§T3<
S
m o
t/3
Ucu
I2 8 T5 t!u u i »
v c Bo oU U
3 P«
CT5 © CM CM
^ in co i>
_; o a.1 c CL cu be.
ta cy O !
cuu
« C feO i - wS
1
CM
CD
03
Q &
3O
C/3
13C
exX
w
CO - 1 - H oO X CO O) 00 —iCC O (O 00
O 10 IN a m o>ooO CO CD if) CT> 0 0 C O
' t to t1 m m co o rfi cool i> 000x i> ^ to 01 o o 10 Co--*1 en oi co
<yi CMCTi O COCO
m o c o o c M o o
o o
if) CO —i 00
o 10
i f) CO CTi CO CM O OOI SO CM
CM CO r-t —I O O—i CO CM
CO CMCM
©co
O —1 OCM
u bea;
C7D
-0 u OQT3
_5 0
Isrt ofa pO
PATTERNS AND TRENDS: HARVARD 95
other sources, and 8 percent came from gifts and endowment, leav-ing about two-thirds of spending funded from unrestricted sources,mainly tuition.
PATTERNS AND TRENDS: HARVARD
Detailed Expenditures
Table 4.4 presents similar tabulations of expenditures for Harvard.As in the previous presentation, expenditures have been classified byadministrative unit and type of expenditure. Although every reason-able attempt has been made to make the classes comparable amonginstitutions, several important differences remain. The most impor-tant of these, one owing to the considerably more decentralized ap-proach to accounting followed at Harvard and, therefore, to thekind of data available in comparable form, is that Harvard's figuresgenerally apply only to its Faculty of Arts and Sciences—which in-cludes Harvard College, the Graduate School of Arts and Sciences,and the major libraries—and to a number of "affiliated depart-ments," such as the museums and the massive Division of AppliedSciences, whose budgets traditionally have been overseen by Artsand Sciences. (Table A4.3, in the appendix of the chapter, lists eachadministrative unit included in the tables.)
Not only are the professional schools excluded from these tabula-tions, as they are with the Duke tabulations, so, too, are most of thecentral administrative units that serve the entire university, coveringsuch functions as accounting, personnel, campus security, and theoffice of the president. As a result of this accounting structure, thedata for Harvard show both fewer university-produced services andfewer internal "sales" of services than is the case with Duke. Thevirtue of using this more restricted accounting base is that it greatlyreduces the necessity of relying on arbitrary rules for allocating gen-eral expenditures to arts and sciences. It does, however, add onemore difference between the data sets for Duke and Harvard. Al-though intra-institutional purchases of services should appear in ei-ther case as expenditures by academic departments, they will appearas internal transactions in the Duke data but will be indistinguishablefrom purchases of outside services in the Harvard data. Other sig-nificant differences between the data for the two institutions includemarked differences in the list of departments and different classi-fication systems used to classify expenditures by type. These differ-ences in organization and accounting practices are only the most
96 PATTERNS AND TRENDS IN EXPENDITURES
TABLE 4.4Expenditures, Levels and Changes, by Departmental Group and Type: Harvard
(Internally Financed Funds Only; in Thousands of 1991/92 Dollars)Type of Expenditure
General AcademicHumanitiesSocial SciencesNatural SciencesMuseumsLibraryStudent ServicesAdmissions and Financial AidAdministrationPlantAthletics
Total
Changes
General AcademicHumanitiesSocial SciencesNatural SciencesMuseumsLibraryStudent ServicesAdmissions and Financial AidAdministrationPlantAthletics
Total
RegularFaculty
2,16419,94412,12216,757
1,0782,357
621225
3,29214
49659,070
OtherFaculty
2715,3841,3661,073
63
2680
2630
2258,857
Compensation
AdministrativeStaff
4,3685,4912,1418,2001,9906,5354,4902,0424,258
32,667
42,185
Students
6,6193,7462,9554,535
1971,708
935502
6735
86022,160
in Expenditures, 1981182 to 1991192Type
RegularFaculty
1,6696,7915,0714,862- 1 2 6- 5 3 9
372- 7 7
1,23514
- 9 219,181
of Expenditure
OtherFaculty
- 4 0 92,543
576342
63
1130
- 9 50
- 9 82,979
Compensation
AdministrativeStaff
3,4132,5001,2254,7031,2253,9262,7461,1972,938
31,259
25,135
Students
3,5451,052
8091,424- 4 4679390246542
35257
8,935
Support
1,4692,6411,9746,6561,1097,4481,962
6941,201
6395
25,554
Support
964557550
1,26427
2,147272
45387
64
6,224
ExtraComp
- 1 1 1- 3 6 2
- 7 8- 3 4 0
514
- 1 414
87706
11
ExtraComp
472,1971,1731,601
513
374305
-5 ,64201
72
Source: Calculations using unpublished data from Harvard University.
PATTERNS AND TRENDS: HARVARD 97
TABLE 4.4 (cont.)
Operating Professional FinancialExpenses Capital Supplies Services Maintenance Aid Computers Total
6,4743,8612,9057,1343,0443,77210,5831,988
30,2606,7813,598
80,402
391,6122,0248,0623,164713
1,0827
9,944720176
27,543
246790539
5,406518
7,2011,16886262218988
17,421
1,452799388
1,616758
2,3682,048103
6,712755
1,11918,118
17100
1,798584
1,70711,545
0143
8,1521,230
25,329
7405,8044,4906,944
1898435
29,128199217
47,874
625294392
1,963155314130139198315
4,228
24,52650,00631,21769,80312,62634,23735,25334,92957,67516,68911,791
378,752
Operating Professional FinancialExpenses Capital Supplies Services Maintenance Aid Computers Total
2,5211,8401,4602,6811,3171,5034,3601,372
15,3601,4561,133
35,003
-831,2481,8211,2276492508223
6,106-1,517
-9410,432
9495278
1,258-302,048169-2550214
-1044,048
826234
-1248386
1,1561,50071
6,447445285
11,009
-6200
6413929
3,6670
1431,827245
6,527
3912,5151,9622,593-598431
13,750172217
21,927
585274294
1,932140152-27597314
3,563
13,49921,84615,09724,6093,28811,46515,21316,96327,7402,4892,827
155,036
obvious reasons to re-emphasize the significant sources of noncom-parability of figures for different institutions. Equally important arethe myriad differences in history, mission, and, indeed, quality,among institutions.
98 PATTERNS AND TRENDS IN EXPENDITURES
The resulting tabulations for Harvard contain 11 departmentalgroupings and 13 expenditure types.15 The time period covered bydetailed financial records in electronic form, which extends back tothe 1981/82 academic year, is two years longer than for Duke. Thetop section of the table presents total expenditures for the 1991/92academic year, expressed in thousands of dollars. For all of arts andsciences, broadly construed to include the affiliated departmentslisted, total expenditures were some $379 million. The largest shareof this total, roughly $70 million, is attributed to natural sciences,which, at Harvard, includes the Division of Applied Science. Admin-istration accounted for the second largest share—$58 million. Hu-manities was next, at $50 million, followed by student services, ad-missions and financial aid, the library, and social sciences. By type ofexpenditure, the broad category of general operating expenses wasthe largest, totaling $80 million in 1991/92. This category owes itslarge size relative to that in the Duke tabulation to the exclusionfrom the Harvard data of most central university support functions.Consequently, many purchases of university-produced services ap-pear as operating expenses, rather than as expenditures for the la-bor and other inputs used to produce those services. Much the samecan be said for the larger size of Harvard's maintenance and profes-sional services relative to those of Duke. The next largest categorywas compensation for regular faculty (faculty with appointmentslasting longer than one year), at $59 million, followed by compensa-tion for staff, at $42 million.16
The bottom part of the table shows changes over the decade, bycategory. Of the 143 cells shown, the largest increase at $15 million,was registered by general operating expenditures in administrativeunits. The only other item that increased by more than $10 millionwas the financial aid category of the admissions and financial aidline, which corresponds roughly to undergraduate financial aid, at$14 million. This item also registered the largest gain at Duke overthe comparable eight-year period.
Internally Financed and Externally Financed Expenditures
Table 4.5 presents figures for Harvard, comparable to those in Table4.2, for the 10-year period covered by the detailed financial data. Asin the comparable table for Duke, figures are given for the subtotalsfor departmental groups and types of expenditures. The table's sixthcolumn shows the average annual real growth rate in total expendi-tures. For all of arts and sciences, real expenditures grew at an aver-
PATTERNS AND TRENDS: HARVARD 99
age annual rate of 5.3 percent, a rate that would imply a doubling ofspending every 13 years. Comparing departmental groups, the tablereveals considerable similarity in overall growth rates. With the ex-ception of plant, athletics, and museums, the rates of growth inspending in the remainder of the groups were clustered between 4.1and 8.0 percent. By contrast, rates of growth among the varioustypes of expenditures showed much larger differences. The mostrapidly growing item, perhaps not surprisingly, was computers,which grew from a modest base almost sixfold. Next came profes-sional services, at 9.4 percent, and compensation for professionalstaff, at 9.1 percent. The extent to which the latter increase can beascribed to the unionization of Harvard's clerical workers is uncer-tain; the increase appears to be spread rather evenly between thefirst and second halves of the period. The only other above-averagerate of increase was in financial aid.
The eighth column of the table shows that the sources of fundingdiffer markedly, especially among the departmental groups. Thepercentage of funds from external funds (grants and contracts)ranged from a low of zero for plant to a high of 49 percent in thenatural science units. The share of external funding varied muchless by type of expenditure. For Harvard as a whole, about 20 per-cent of all arts and sciences spending in 1987 was funded by externalsources.
The last four columns of the table show how the growth rates ofspending differed by period and type of funding. Overall, spendingfrom internal sources exceeded that from external sources duringboth periods, the difference narrowing considerably during the lat-ter period. Focusing on the four academic departmental groupsshows a decline in the growth in internally financed spending inevery case. By contrast, external funding accelerated in the human-ities and social sciences. However, the natural sciences, the academicdivision most dependent on outside funding, did not follow this pat-tern, but instead saw a slowing in the rate of spending supported byexternal funds.
The patterns of growth by type of expenditure show considerablevariation. Double-digit growth rates were recorded in six instances(excluding extra compensation, a small residual category), and notone of the six was compensation for employees. During the first halfof the period, there were large increases in internally funded spend-ing for capital and computers. During the second half, expendituresfor both computers and for professional services grew very rapidly,and these increases drew from both internal funds and externalfunds.
X
aV0-
0
h
s- v
ate
-5cV
'%
l o3
2
w 3NJ GJQ qj
< s
"5
j . ,
"oQ
O-CM
Q
wth
,
0
O
199
1
lio
"I8
a
C/5
Nt% ON25 ^so ON0 0 ONON --H
22o
ON
22 ^-^ ONoo ONON ~^
S ps h
k, -S
as co co O] CM i> in!>• iO CO CO —< i O CM
o » o cnco eo to co
CM as © in© —J © CM
iO - H X> 00in co oo -H
o q i n i n i n o c o s C M o o o o q - ^ o qco oo oo i d oo ^ --^ © c o d e o c o
i i i T
qco
co oq • *•—i co co
•*f © © Tt1 m coCO i—i OS CM i—i -—I
CM T f CM
q -^ co ooi> o co in
CO © !>• f »^ d © a!
o q c o i n c M q q i n i n - ^ i n c o ^- ^ a s © o 6 c o i > a s © © i n c o a s
c***3 x^» c o c o f*T*i i~^ cO c o CO c o -t * c oodinco'^co^'in co'co- CMin
^^ ) ^ ^ f ** f" **> ^*^ f*^i ^ ^ ^ ^ J t*^*i ^ * i f ^ ^ ^
© d d d d d © © © © © ©
min
oo m co co coTJH oo CM co - ^
co eo oo oo co a>o —i o —' o o
as oo m -^ —i• * • * —i CM - ^
© © © © ©
as in Tf co o© —i © © ©
© © © © © © © © © © © ~H
3
2
I-J
c
a.
m' • fCM
©
ral
idem
ic
o
©©m
CM
00CM
anit
ies
|
X
CM,—,CO
rt
CO^ H
t / 3
1 Sci
enc
• 3
00
OsCO
CM
m
rice
sra
l Sc
ie)
atui
2
CO
CM»—i
CO
OS
ums
use
2
CM
~vCO
00
CMCM
ibra
CO
i nCO
©©CM
ices
;nt
Serv
If)
C
issi
ons
I
a
Os
CO
©00
anci
al >
r—
m
OsOSCM
co
inis
trat
i'
-a
t^ 00co —•
CM ©• * OS
tics
ant
thle
OH <!
00
00| > .co
^
COCMCM
I t
q in oq oq cocb CM cd co ^
CM
CM
CM • * i n
co in eoCM
in on CMCM
COco
_CM1
m oo
CD ini—i CO
CM
CO
coCM
o
oo
CO00CO
CDCO
om
CO OCO —1
m <*CD -H
CMCM
CDCD
CDCO
oo
CM
m coCM
CDCD
! > •
< *
00CO
CM oq q CM CM•^H CM O CD CO CO
CM
in CM cot~~ CO ^
in coCD CM
i> oqco in
CM co in co mCO CM CM —i
oq q CD q CD1 > 0 0 CM CD CD
CM
CO
qd
CO Tt4 COCM CM CM
to CD i n
CD 00 CO
b q oq oqbCO O i—i CD CDCM CM CM ^
CM 00 O
CD i—i ^ CM 0 0
co t1 CD m CM
m i—i 1 > CM CMo q q q oo d d o o
qd
oo
i> oq coin -^ CM
CD CO i—i
o o o
CO CO O CM CDi—i i—( CM
•f q H in coCD co co oo in
CO CM CD —i i—iO O O O Tf
o o o o o
CMCOo
coo
oCOo
oo
CM
o
CMCOCO o
00coo
COCM
oCO
o
COCMo
CD
O00
o o
CD
O
CD
m
CDCDCO
CM
oo
CD00
CD
m
o
CMCM
oJ>
COO
o
CMCMCM
CMCO
oo
CO
mCM
COCD
O
oo
oo
—<o1
CM
o
o00
• * !
m
oo
CM
oo
CO
moo
00
oo
CO
mCM
GO00
CO
O
CD
$
CD
mCM
oo
CM
o
o
00
378
t^
223
e
data
vel.
ishe
c92
le
3 O5
3 <*-
c <u
au
rcen
tage
inc
a.r
c
re
CO c
~L ** ^
^ 6 4g 5
re
«5 . S CJ
0 u .S 0£ u
S-,
A)
V3
CMO*JCT>»-H
"0c
„
00
•—' - o
•*" 0
cS '5w ~ £^ jZ tin^ "> r i
3 QJ
oVu3O
>^
c^V
3
C<ucx
s
5a
,to
O
s
ssos
s
12
%J
O
fo
r ^
^ *
era
ON
ONON
t N2$o
2?ON
<2o
981
*~*<
ONON
t N2o
ON* - I
<M2o00ON
<NON~~IONON"~-(
t ^2$o
QoON
<N2g•QoON^ - i
<NON
ONON
2o^oooON""•I
2g00ON
to <o o onom
H a) CM H to
Tt* rf CM oo -HC O O O O O O O m
en o i x --i CM co C X > - H C O O I >!O CM (M !fl CM - H ^ CN rn W CM
in i> J> i> to ~-'5 O CM CM • * CM - 1
o « m (D N m oi cocoooH en en w tn CM m >-H •—i
o t^ o en en —i -^ o o o o m
o c M t ^ ^ o n o o - ^ O O O C M
O C O - ^ - H O O O O O O O O C M
CM • * N e n > H P H o i n - - i o o
CM tO CM O ^ 00
rt H oj N oonocM-^m en H CM
J 5 <U
M S -T3 S
i i i i g i i r '=-°W M
u
X C/5 Z. ^ J C/3-0 rt
PATTERNS AND TRENDS: CHICAGO 103
Changes in the sources of funding are shown in more detail inTable 4.6, which gives the percentage distribution of expendituresby type of funding for the beginning, middle, and end of the periodunder study. The table clearly reveals two striking facts. First, itshows the heavy reliance on federal support in the natural sciences.About one-half of all spending in the natural sciences used federalsupport, whereas no other group depended on that source for asmuch as one-fifth of its spending. The second striking fact is theacross-the-board decrease in the importance of federal funding. Forthe university as a whole, the share of total spending supported byfederal funds fell from 21 to 15 percent. A slight offset to that de-cline was the growing importance of nonfederal sponsored research,mainly foundations, which was especially evident in the social sci-ences. With gifts and endowment remaining steady at about one-fourth of the total, the net decline in outside support resulted in anincrease in the share of unrestricted spending from 51 percent in1982 to 53 percent a decade later. Thus, unrestricted revenue, ofwhich tuition is a major component, came to have a larger impor-tance in funding the rising arts and sciences expenditures.
PATTERNS AND TRENDS: CHICAGO
Detailed Expenditures
The basic expenditure summary for the University of Chicago ispresented in Table 4.7. For the most part, the departmental groupsand expenditure classifications are similar to those used for Dukeand Harvard, with one difference: the classification of payments tostudents is combined with nonexempt compensation. Because ad-ministrative and service components from the entire university wereused in making the table, the interpretation of the columns for ser-vice-related functions, namely maintenance, professional services,and general operating expenses, is closer to the Duke case than tothe Harvard case.
For all of Chicago's arts and sciences in 1991/92, estimated expen-ditures were roughly $247 million, standing between Duke's $162million and Harvard's $379 million. As in the comparable case ofDuke, Chicago's largest single category of expenditure was financialaid, at $54 million, followed by compensation for regular faculty, at$51 million. Of the departmental groups, admissions and financialaid represented the largest expenditure, $51 million, followed bygeneral administration, at $49 million, and natural sciences, at
104 PATTERNS AND TRENDS IN EXPENDITURES
TABLE 4.7Expenditures, Levels and Changes, by Departmental Group and Type: Chicago
(Internally Financed Funds Only; in Thousands of 1991/92 Dollars)Type of Expenditure
HumanitiesSocial SciencesNatural SciencesLibraryStudent ServicesPlantAdmissions and
Financial AidProvostAlumni and
DevelopmentArts and Sciences
AdministrationGeneral Admin-
istrationOther
Total
HumanitiesSocial SciencesNatural SciencesLibraryStudent ServicesPlantAdmissions and
Financial AidProvostAlumni and
DevelopmentArts and Sciences
AdministrationGeneral Admin-
istrationOther
Total
RegularFaculty
11,39811,19417,144
000
0488
9
6,118
3,1291,062
50,543
OtherFaculty
1,416809
2,0401,935
00
3365
0
881
242676
8,096
Compensation
Administrative NonexemptStaff
962784
3,7621,280
6461,770
810356
4,355
3,550
13,5571,244
33,076
Changes in Expenditures, 1983184
RegularFaculty
1,4982,7313,003
0-120
- 7 9
- 3 8381
7
2,552
900517
11,354
Type of Expenditure
OtherFaculty
561276
1,1551,935
- 10
2255
0
535
- 9 7427
4,868
Compensation
Workers
398626
2,1352,407
3754,602
493101
1,054
1,897
10,5921,058
25,737
to 1991192
Administrative NonexemptStaff
562267
1,791-930
185943
249140
1,877
1,769
3,331442
10,626
Workers
-217-183-554
108- 3 9
3,570
- 2 650
- 2 8
-117
-915256
1,906
ProfessionalServices
530265
1,056622
865,215
408173
2,235
1,970
3,8751,307
17,743
ProfessionalServices
12983
213-630
- 1 85,024
-18694
-468
1,031
-111637
5,133
ContractWork
3400000
00
0
0
00
34
ContractWork
3400000
00
0
0
00
34
Source: Calculations using unpublished data from the University of Chicago.
PATTERNS AND TRENDS: CHICAGO 105
TABLE 4.7 (cont.)
GeneralFinancial Operating
Computers Aid Supplies Expenses Capital Maintenance Residuals Total
1101096172682237
1944
75
514
2,02226
3,862
375324
2,2661260
48,9065
47
1,953
-257
53,958
385151
1,9153,575
691,486
9825
210
838
2,464595
11,809
825467
1,59521627
3,083
429271
920
921
9,017689
18,460
62695
15,5023032124
821
75
198
2,07787
19,064
3821242537
1,288
148
21
515
1,43751
3,694
-41-164
426216196
243
151
-1
408-141556
17,05514,68248,31510,4491,595
17,701
51,2421,559
9,151
19,353
48,8196,710
246,632
GeneralFinancial Operating
Computers Aid Supplies Expenses Capital Maintenance Residuals Total
92-92-452421135
-1443
-53
270
90717
1,413
219753931
-250
30,438-144
-161
1,452
-4356
32,260
21461481
1,45923693
2921
15
523
-953242
2,808
425151
1,32434979
-149
933241
615
1,514
8,3841,967
15,831
543-45
10,401-5,640
321-37
817
66
-5
57911
6,217
213
-179-212
1,281
54
-1
316
36636
1,832
-19-108
166210165
14-2
84
36
-1-102155
4,0633,21717,997-3,066
43011,446
31,434900
1,952
9,877
11,6814,50694,436
106 PATTERNS AND TRENDS IN EXPENDITURES
million. By far the largest of the individual cells in the table for1991/92 expenditures was the cell for financial aid at the admissionsand financial line, a figure corresponding roughly to aid for under-graduates.
The bottom part of the table shows the changes in expenditures,in 1991/92 dollars, over the eight years beginning with the 1983/84academic year. During this period, the largest increase, representingan astounding one-third of the total increase in spending, was inundergraduate financial aid. No other item of increase comes any-where near that figure in magnitude.
Internally Financed and Externally Financed Expenditures
Table 4.8 presents summary measures for the column and row to-tals. Overall, internally financed spending at Chicago rose at an an-nual rate of 6.0 percent, which stands between Harvard's 5.3 percentrate and Duke's 6.8 percent rate. As in the comparable tables forDuke and Harvard, column (5) shows the contribution of each cate-gory to the total percentage increase of 38 percent over the period.Not surprisingly, of the departmental groups, admissions and finan-cial aid was the largest contributor to this increase, representing byitself a 13 percent increase in total spending, or one-third of thetotal. Large percentage increases occurred in several groups, includ-ing plant, admissions and financial aid, and the provost's areasof administration. However, the large share of total spending ac-counted for by admissions and financial aid, combined with therapid increase, was responsible for that group's major impact on to-tal spending. With respect to types of expenditures, the most rapidgrowth was recorded for general operating expenses, financial aid,and nonregular faculty.
Regarding the importance of external funding, the third entry incolumn (8) shows the importance of federal support in the naturalsciences. Over the period, externally financed spending slowed be-tween the 1984—87 period and the 1987—91 period, while the rate ofgrowth of internal funding increased. As with Duke and Harvard,the rates of growth in internal and external funding for the variousgroups over these two time periods varied greatly, looming quitelarge in cases in which the base amounts were small.
The changes in the sources of funding at Chicago are summarizedin Table 4.9. The importance of federal funding decreased, as it didat Harvard. At Chicago, this funding source fell from 22 percent of
obe
U
cvO-X
wVCL
O CM
°° G -
J C —'
OH.2
I
CVa,
CO
^o ONoo ONON -I
oo ONO N --I
1 22\5°o ON ON*"< ON
"S-
ON
oo o i; c o q(y> in oi o o o
O"! """ C M 0 0 i*""**1 f""*>00 1> TJ-I CO © ©
CO co
CM >—i i> —i o in
q q q q q od d d d d d
J> 00
^ CM
q q oo (C •*f^ ^^ j^. i co
00
CM
©
1
CO
CO
m
CO
©
CM
md1
CM
00
in
en•—'
*—i
CM
©CM
in
CO
©
d
CO
©1
00
©
©
d
CO
COCO
d
CO
d
min
00
©
175.
©
CM
q00
cod
CO
O5
CM
CM
CM
00
CM
CO
d
, |CO
md
00
in
d
CM
CO
CO
l>i
CO
d
qCO
min
O] 00
d
!>• CO
CO CM
co o—| O
Tf CM !>• O t"- iO -H XCMCMCOCMCMCO q ^
d> <zi d> d> d d> d ©
i ** co c^ *^ ^ r**© © CM © q ©d © © © © ©
t^ • * oo o « ^
TJH oq
oo co
CO COCM
co in
© COCO 00
COeodCM
o>
© CTi " ^ O*i ©
CO 00 CO CO CO
—i -<f i n CM oo© © © © co
CM in CM co eo
d © d> d d>
0.21
CM
0.01
CO
0.04
CM
0.08
0.20
00
00
0.03
CO
©©
CO
CO
ntai
artm
eG
rou]
qCO^ H
ities
uman
X
in
_Ji"H
cien
ces
jcia
l Si
CO
CO
dCO
Scie
nces
atur
al
Z
in
CO*^
t«
ibra
rie
_q
CM__;
Serv
ices
uden
t
CO
CO
d
ant
AH
ons
and
dmis
si
00
oi
cial
Aid
Fina
n
©
rovo
stA
ffai
rslu
mni
)eve
lop-
and
E
CM
men
tlie
Ad-
cade
m
m<y>
trat
ion
min
isi A
dmin
-en
eral
O
co
Co
istr
ati
CM
CM
ther
0
CM
CM
m
Tot
al
T3G
ex
co CM >n CM© 0 6 CD" t^
m CM© ©© ©"
© ©
in "*© 06m
CM CM
CD COCO
CO © CM
- H CD 00 CM r f
-^ CM TF in co
co
COCM
CDCM
COCO
©
m
inCO
in
m
©1
CO
CDCM
1
mCM
©
©©
CD00
©©
©
CO
CM
mCO
• ^
CD
©00
CO"f
CO1
m
CM
©©
©©
00
mco1
©
COCO
CM
COCM
CMCM
CM
*""
00CM
00CO
CMCDCMCO
CDCM CO t^ in
CO
minCM
CMCO-CM
00
00
©CD
COCD-CM
00CO
CMCO
CMm
m
m00
00
CO
©
CDJ>
CO
©©
©©
Tf © >—1CD in t~»
t- ^ ^
m —i co
©00CM
CO00
CD
CMCM
CO
on
r~- CD
—i o•*f co
©©-
©©-
CM © ^H CO —i© © © — ©
CO CO —i © 00
© © © © co© © © © ©
CMCM©
©CM
©CO©
.03
CMcoo
co
©©
.10
CM©
.07
©©
.00
CO©
02
©CO©
CMCM
CM©
05
CO00©
07
coCO©
08
©m©
01
00CM©
00
00CO©
00
ver
Uni
the
rom
© © © © -H
COCO m
CMm -H —i ^H
in ^^ !>• co cooo CD co © co"
CM
CMCM
COCM
©
CD-
CO 0 0 CD T f i CM
CM CM I -H © CMi—i >O
3 « s
S
C0
ZB
™G_O'StflAi0
- O
O O-t
Z80- ?
uG
_ C
0be
u
<J0
cbeC
C3O
y — c o- C CM <L) ".52 o» H u
3 CTI Q . ~
C ^ ^ t53 <*; - bebe ° T3 .CC 2 S Stn C8
. . DO 3
i: S PCO
U SU
CM
OS
oo sO> 3
3o-fsi5
~0 v
3o
en
c
I O O C O C O O O O « m o
CM —H O
a) a> o^^ ^* CM
oo oi co
ff) CO (D -H CO O CMTf CTi Tt* I> 00 i-Ht^ CD CO O) O) O tf l O) CT5 iO CT)t^I>
COCM
CM
C M C M OCM
C O O I C O C M O O OO O
C O O O O C O O O O O O O N O C)
00 N O o CM-^ o co
00 O O O O O
m CM
ltl
c
I3DC
aces
dj
—
Is' u0</3
ienc
3
z
!/3V
rvic
3 Cd
oo S
13
fi ^
.2 "t5
's |
>a;Q-oc
i« '2> S0 _3
CM CO CM CM—i CM
%%s
o -=U
< O
110 PATTERNS AND TRENDS IN EXPENDITURES
total arts and sciences funding in 1983/84 to 19 percent by 1991/92.Federal funding was of greatest importance in the natural sciencedepartments. At the same time that federal funding was becomingless important, sponsored research from other sources increased inimportance. Reflecting the trends in the other universities, there wasan increase over the period in the percentage of expenditures sup-ported by other unrestricted funds.
PATTERNS AND TRENDS: CARLETON
Detailed Expenditures
Not only is Carleton College quite different from the three researchuniversities covered in this study, the financial information availablefor it was distinctive as well, in that detailed disaggregated data werenot available in machine-readable form even for the last 5 years ofthe period of study. The next best alternative was to use tabularsummaries of expenditures contained in annual financial reportspublished by the college. Although these reports offer the advantageof consistent coverage for the entire 15-year period of the study,they present breakdowns only for six categories of expenditures, twoof which (student work and travel) are quite small. Moreover, thesedata do not permit the separation of internally funded and exter-nally funded expenditures. The classification by departments wasconsiderably more detailed than in the three sample universities.Therefore, the departmental groups could be rearranged to besomewhat comparable to the presentations for the research univer-sities. The components included in the departmental groups listedare noted in appendix Table 4A.5. It also was possible to add twoadditional "departmental" categories that actually reflect differenttypes of expenditures or sources of funding. The category for schol-arships in effect adds one important additional expenditure type,and the organized research category can be viewed as largely com-prising outside grant and contract funding.
Table 4.10 presents the division of expenditures for Carleton in1991/92. Out of a total of $43 million spent during that year (whichincludes externally financed expenditures), fully one-half was ac-counted for by the nearly all-inclusive compensation category. Ofthe 98 individual cells in the table, the largest category of spendingwas for scholarships, at $7.4 million, very much in line with the highcost of undergraduate aid in the research universities. The bottom
SUMMARY 111
section of the table shows changes in real spending for all the catego-ries over the decade 1982-1992. Here again, financial aid is the larg-est single category, with an increase of $4 million, or about one-fifthof the total increase.
Table 4.11 summarizes the changes in spending for Carleton overthe entire 15-year period in a form similar to that used for the re-search universities. Overall spending doubled over this period, foran average annual growth rate in real terms of 4.7 percent. For thedecade 1982-1992, the annual rate was 5.7 percent, which compareswith 6.8 percent for internally funded spending at Duke, 5.3 percentat Harvard, and 6.0 percent at Chicago. The departmental groupsshowing the most rapid growth over the decade were computing andorganized research, but the line with the largest contribution wasscholarships, owing to the relative importance of this item (17 per-cent of all spending) and to its strong rate of growth (7.9 percent).Variation in growth rates among the types of expenditures showedmuch less variation; equipment expenditures rose the fastest duringthe 1982-1992 period.
SUMMARY
This chapter presents data on expenditures for the four sample in-stitutions. It focuses on internally financed expenditures, that is,spending that is not specifically tied to outside grants and contracts.For each institution, expenditures are divided by type and depart-mental group and are compared over time to determine both thedistribution of spending of different types and the growth in thatspending. In line with the aggregate trends noted in chapter 1, ex-penditures at the sample institutions grew rapidly in real terms fromthe early 1980s to 1991/92. The annual growth rates in total spend-ing were: Duke, 6.8 percent; Harvard, 5.3 percent; Chicago, 6.0 per-cent; and Carleton, 5.7 percent. Among the components of thisoverall increase, financial aid was conspicuous for its growth; all theinstitutions experienced rapid growth in this category. Another reg-ularity among the three universities was a decline over the period inthe importance of federally supported expenditures. It is worth not-ing a proviso stated earlier. Despite the attempts made to arrangethe expenditure data in comparable ways for all four institutions, thedifferences among institutions in organizational structure, account-ing conventions, and functions mean that comparisons among insti-tutions inevitably will be problematic. For this reason, the present
112 PATTERNS AND TRENDS IN EXPENDITURES
TABLE 4.10Expenditures, Levels and Changes, by Departmental Group and Type:
Carleton (In Thousands of 1991/92 Dollars)Type of Expenditure
HumanitiesNatural SciencesSocial SciencesGeneral, Other Academic DepartmentsOrganized ResearchLibraryAcademic AdministrationComputingStudent ServicesAdmissions and Financial AidScholarshipsGeneral AdministrationAlumni, Development, Public RelationsPlant
Total
Changes in Expenditures, .
Compensation
5,6643,1361,8981,386
589784435565
1,534777
01,9141,7201,242
21,645
StudentWork
133157278116
1675076
34326
0834950
1,258
1981182 to 1991192Type of Expenditure
HumanitiesNatural SciencesSocial SciencesGeneral, Other Academic DepartmentsOrganized ResearchLibraryAcademic AdministrationComputingStudent ServicesAdmissions and Financial AidScholarshipsGeneral AdministrationAlumni, Development, Public RelationsPlant
Total
Compensation
2,1891,215
590576462267210329617274
0849784336
8,699
StudentWork
7221102013113945
107110
3923
0411
Supplies
1271515966942883
298235286
0182879
562,545
Supplies
3644312241
145
26810477
0109422
31,204
Source: Calculations using data in Carleton College, Report of the Treasurer, 1976/77,1981/82, 1986/87, and 1991/92.
SUMMARY 113
TABLE 4.10 (cord.)
Equipment TravelOther
Expenses Total
49165
1016
35580852
69096
60
323679
2,394
13122
1161031070182741360
1602184
1,136
1114829
1,837162593982475956
7,3941,036328
2,15814,400
6,0973,6702,0253,5031,3191,8561,0891,6713,2401,2887,3943,4073,2313,589
43,378
Equipment TravelOther
Expenses Total
-5912571
31326638
53870- 4010
-38-3
1,263
-16-74-163591213
1406507872-1401
791
-1278527-12751538016
4,02362920621
6,859
2,3021,332625
1,468916545610
1,2081,419439
4,0231,7141,283955
18,837
ao
Z
U
p
a,3
o
O
13pcu£
3 o-3 Q
Pi
oO
oO
CD
3
•3pa.w
C/5
1
I
C5
I-
00
I N ON
ON
92 ^~-< ONoo ONON "-I
pcu
aQ
CM CO - ^
CO CO CO
m coc> 9 90 0 0
d
,097
CO
enCM
i n
795
CO
746
od
,670
co
,389
CO
338
CM
138
od
,025
CM
,857
"—'
401
223
00 -*co in
CM
CO
co
o
inCM
en ini—l CO
CM —1o q0 0
en00CO
CMCO
CM
CM
r- coin oo
CM 00 00
00 CM i r i
9o o o
00CO
dcoCO
dco
CM
denCO
denCM
dcom
CM
d
000d
,503
co
,470
CM
035
CO0
d>
,319
,423
403
Tt1
0
,856
^ ^
,692
^ ^
311
co0d>
,089
695
479
qd
,671
^
,354
463
0
d
,240
CO
,395
CM
822
- H 00co en
' 0 cu 0 -g .a CS.2
porti -
nisi 3
aS0
U
cj
rvi
CUC/3
PCU
-o3c/5
- H 00co en
cn
CM cn
r-1 enq qd d
CO
o
COin
qd
ooo
^H 00 CMCM CM COco t i m
q q tor-» in i>
COCMCO
o d d
co
dmd
omd
o• ^
dCM
d
co
d
b X Oo o o
© Cn CO ^ CO ^ COJ> iri 00 CT> CO 00 1>
—i 1> CM CO J>co TJH i n i n - ^
>—• o "^ in T
m •'f co i> T
o <—i co co <-
t1 in i>f co m
H co co
199
-ocftr-'00cB00cn
o co x> co in oo cod> d d d> d> d> d>
o co co co co co oin q q q q co qd d d o o d —
,288
F - H
,104
849
,394
,912
370
,407
co
,465
CM
692
,231
CO
,160
CM
948
,589
CO
COco
co"
635
,378
CO
,181
co
541
,645
CM
,481
946
,258
,083
847
,545
CM
,980
341
,394
CM
,014
co
132
,136
765
735
,400
,857
cn
541
.378
co
.181
co
541
—i CM
co m co CM -H cn -^cn CM CM oo CM oo cob to o ^ m ^ wT ^ ^ co" i-Z
^ H CM
c 2(« S J_tn rt rt
g . S o-rt fe J3
cuC Qo
^ 2 §IJ fa £C jg 3
C "cuC O JrCU 'w 0
^ cu "' ^
•a J3 H jj c 5 35 o. u
P £ "S a£ -
H
CCUO HX
CU O
o
beC
o
116 PATTERNS AND TRENDS IN EXPENDITURES
study focuses on changes in spending over time for the same institu-tion. One cannot expect to discover, for example, which institutionspends the highest percentage on natural sciences or administration.In the next chapter, these expenditure data are combined with otherobservable quantities in an effort to find explanations for the spend-ing increases.
Appendix 4.1
Dealing with InterdepartmentalTransfers and Recharges
THE KIND of fund code accounting practiced by the universities stud-ied here involves a significant amount of internal transfers and re-charges. The first aim in dealing with these transfers and rechargesis to avoid double counting. Beyond that, however, the way these aredealt with will have implications for the kind of information that canbe gathered. Consider, for example, a simple kind of rechargewhereby one unit of the university (let us say, the copy center) per-forms a service for another (in this case, the physics department).The copy center purchases inputs and "sells" the output to the aca-demic department. Suppose that, during a year, it buys $500 worthof paper, pays $250 for machine rental, and pays workers $250 tophotocopy material for the physics department. The end-of-yearledger will show expenditures for these items under the copy center(general administration); a recharge receipt of —$1,000, also underthe copy center; and a recharge expenditure of $1,000, under thephysics department. If all recharges are ignored, the university's ex-penditures are reflected correctly, but the portion accounted for byphysics is understated and the portion accounted for by general ad-ministration is overstated. Stated another way, the summary tablefor the university would look different if physics had purchased thephotocopying service directly from an outside vendor. If the re-charges are counted, as they are in the chapter, it is possible to re-flect both the university's expenditures on inputs and the use of theresulting goods and services within the university. However, this ap-proach necessitates the inclusion of the recharges, which appear as anegative entry in the general operating columns of the origin-desti-nation expenditure tables.
It is useful to illustrate the approach taken with a simple example.Table 4A. 1 shows the categorization of expenditures for a hypotheti-cal university having three departmental groups (all arts and sciencedepartments, a professional school, and general administration) andthree types of expenditures (salaries, services provided by one unitfor another, and other operating expenses). Many services are pro-vided by one unit for another—for example, security, housekeeping,
118 DEALING WITH INTERDEPARTMENTAL TRANSFERS
TABLE 4A.1Accounting for Shared Costs and Internally Provided Services:
A Simple Example
Departmental Group Salaries
Type of Expenditure
ServicesProvidedInternally
Entire University
Arts and Sciences Departments 60Professional School3 50General Administration 40
Total 150
1020
- 3 00
OtherOperatingExpenses
2030
100150
Total
90100110300
Arts and Sciences Only—Allocating 50% of General Administrationto Arts and Sciences, Counting Internal Transactions
Arts and Sciences Departments 60 10 20 90General Administration11 20 —15 50 55
Total 80 - 5 70 145
Arts and Sciences Only—Allocating 50% of General Administrationto Arts and Sciences, Ignoring Internal Transactions
Arts and Sciences Departments 60General Administration6 20
Total 80
205070
3070
150aOmitted below.Expenditures for arts and sciences assumed to be 50 percent of actual levels shown
in the top section.
repairs, and photocopying. Although practices differ, institutionstypically make accounting entries for services that are easily attribut-able; these services are the ones measured in the second column.The top part of the table presents expenditures for the university asa whole. It shows that the units of the general administration spent atotal of $140 on salaries and other operating expenses but "sold" $30worth of services to academic units. Including the column for inter-nal transfers makes it possible to attribute all identifiable expendi-tures to the units that are the ultimate beneficiaries. Thus, the $30worth of services provided by the general administration appropriatelyis assigned to the arts and sciences departments and to the professionalschool. These services are counted toward the ultimate user in thesame way as if they had been purchased outside the university.
A problem arises, however, in allocating services that are not re-
DEALING WITH INTERDEPARTMENTAL TRANSFERS 119
corded by means of internal transfers of this sort. General opera-tions, ranging from groundskeepers and road repair to the presi-dent's office, are not easily allocated. In the present study, whichexamines only arts and sciences, a portion of these general functionsis attributed to arts and sciences, where the specific proportions arebased on the rough estimates and educated guesses of administra-tors at each institution. In the example shown in Table 4A.1, thatproportion is 50 percent. The middle section of Table 4A.1 showsthe resulting summary of expenditures for the arts and sciences por-tion of the university only: the line for the professional school isomitted, and only 50 percent of each entry for general administra-tion is included. The problem is that, after some entities have beenomitted and a proportion applied to general administrative expendi-tures, the internal transfer column no longer necessarily sums tozero—it may be positive or, as it is in this case, negative. One ap-proach that avoids this ambiguity is to ignore internal transfers alto-gether, as is done in the third section of Table 4A.1. By ignoringinternally provided services, however, this approach overstates theportion of the university's total expenditures accounted by generaladministration. Furthermore, to the degree that various componentsof the academic division use differing amounts of such services, theapproach illustrated in the table's third section would misstate theirrelative sizes as well. Thus, the approach taken in the present studyis that illustrated by the second section: a portion of general admin-istrative units is assigned to arts and sciences, and internally pro-vided services are recorded where possible. In practice, these trans-fers usually will come close to netting out. The gain in reflecting therelative size of departmental groups appears to be worth the loss inaccounting tidiness.
To summarize, the approach taken here implies that the costs ofsome activities that are performed by a service unit in the university(such as photocopying) will be reflected in three places. The totalcost will show up as a general operating cost under the departmentthat ordered it; the cost of inputs will show up in the appropriatecolumns in general administration; and a negative entry (the re-charge) will appear in the general operating column for administra-tion. As a consequence, the arts and sciences totals along the bottomof the table will reflect the actual use of resources, and the totalsalong the right-hand side will reflect total spending by the divisionsof the university.17 As noted in the text, the accounting data forDuke and Chicago include the expenditures of general administra-tion units. Thus this approach will make the most pronounced dif-ference in tabulations for those institutions.
120 DEALING WITH INTERDEPARTMENTAL TRANSFERS
A related issue arises in measuring expenditures for purchases ofcapital, which includes such items as computers and scientific equip-ment. At most institutions, data on operating expenses in the gen-eral ledger do not include expenditures for construction or for somemajor capital expenditures. At Duke, these capital items are paid forout of separate plant accounts. Those accounts receive their fundseither by way of transfers from operating accounts, which are readilymonitored, or by way of borrowing. In the case of debt finance, simi-lar transfers from operating accounts are used to retire the debt. AtDuke, the funding of capital expenditures is treated differently inadministrative units and academic units. Administrative units em-ploy depreciation ("betterments") codes, wherein current funds aretransferred into plant accounts, from which expenditures for suchitems as computers, renovation, and furniture are made. Academicexpenditures for these items are often made from current accounts,however, and are designated here as computers or capital expendi-tures. Thus, although I ignore most transfers, I include transfers toplant accounts for academic units because they reflect capital expen-ditures that have taken place or that will take place.
Appendix 4.2
Categories Used to Create Expenditure Tables
To EXAMINE IN some detail changes in expenditures over time forindividual universities, it was convenient to categorize expendituresby departmental group and by type of expenditure. These catego-ries are only approximately comparable among institutions, how-ever, in part because the available data differ in a variety of ways.The data from Harvard, which are from the Faculty of Arts andSciences, provide virtually no detail on a number of nonacademicdepartments. In contrast, the information for Duke and Chicagocovers all departments in the university with equal detail. Anotherreason why the categories are not strictly comparable is simply thedifference in the functions that each institution carries out: one uni-versity may offer programs that are simply not found in anotheruniversity. Thus, "natural sciences" represents an amalgamation ofdepartments and other administrative entities that has a differentweighting of disciplines, not to mention individuals, from one insti-tution to the next.
The categories used in this study are summarized in this appendixin some detail for each of the institutions examined. In most cases, itseems fairly clear how to categorize a program or a type of expendi-ture, although some choices may be debatable. For the most part,"area studies" programs were placed under humanities, althoughthere is undeniably a social science element in much that goes on insuch programs. Another ambiguous discipline is history, which I alsogrouped with humanities.18
Tables 4A.2 through 4A.5 give detailed descriptions of the entitiesand types of expenditures used in defining the departmental group-ings and types of expenditures in the tables presented in the chap-ter.
Q
ohe
a;Q
™ <u -3 60 C
J £°.£•3 -i 3
-C O
'5b
LC C 3 £C <U
ii '•" "n 3
I £ g\^ S
>-, C 3
3 .u u .JJ >-£ J3 P3 „ C-J ""1 -a w £_ « : « 3'j U i; u uC ^ « . i o
a
>-- <
•ss 2 S>- bo Q
OH S >r--c
c c0 <U
H3 w3s 5U '^-3 3
a >-5 bee O3 "o
- h. « >- • -
ro ?< •',JH .a —*qj tn W^ « C3 rr\ .«
be
•y- 'S ^ s |
(« O
I £
2 'a £
q 2
•- id
be.2 _- (j :C T! -O W
K i
J 3 "J £ ^- i< u 3 •,(j , M «
r U 3 ibog3
(£ J5 ^
2 -n £3 JS Q
« uQ o
u
S S3
dj 3
„ o
^ • ^
3
< is
-5 T5 u -3
. 1) ce hr 3
O
"S-
CS
5
^~
T33PS
<
©""^i-OCMCM
O
• ^
CMCM
©CM• ^ H
CMCM
©"
COCMCM
©~
^H
CMCM
b
©"©CMCOCM
©CO©COCM
m• ^
CMCM
©"©CDCMCM
©~00
CMCM
3
u<->-J
^inCOCOCM
COmcoCOCM
CMinCOCOCM
_,"inCOCOCM
©"00CMCOCM
faaV
-a
©"CM©COCM
O—H©COCM
CM00CMCM
o "
CDCDCOCM
©'CM00COCM
3
ubecOH
©"in^^
CM
©CM
COCM
COCM
©'©
CM
O"
©COCM
a3be
_r^XCOCM
inCOCOCM
CMCOCOCOCM
CO"COCMCOCM
_TCOCMCOCM
.a cI; £
£ 0< 0<
co"CO^
CM
00COCMCOCM
CM00CMCM
o "
5CDCOCM
m"CM©COCM
3 .u
wbe P« 53c -s
1
CJ —
^ o
3 W
g £be t;
bb £ cro p ;j
JH - ./"
3 ctc« O
s 2i- rbe >
a -
C 3
5 c«
be *j
c a•S Ic 0.
(A 0
© © m• * <o ©1> © CMCM CO COCM CM CM
i iI> ©CM CMCM CM
© m c M c o © — O C MC M C M C O C O C O C O C O C OC M C M C M C M C M C M C M C M
©
CMCMCM
©"
©CMCM
©"CM©CMCM
CMCM
CMCM
m"
CMCM
mmCOCOCM
^'
CMCOCM
CMCOCMCOCM
.a 3 p
B c o3 S .a c/3
bcO3 N .£ OH •£
© © .CD CD CDCO CM XCM CO CDCM CM CO„ . CM
© m I© co ©m CM xCM CO CDCM CM CO
CM© © - ©X CD CO ocCO © <f> -4CM CO 5S CDCM CM gj CM
©" ©" CM— ©CM COCM CM CM
© © a ©co x 5; coCM CM 00 CO
CM CM CM CM©" -H" — CD"in CD CD x© CD CM CDCM CM CO COCM CM CM CM
"S-S c
c'Soc
1a.
3 .2 ^ .2 S
-o- re
£) Q- 3re TJ
TJ * Wo 2 re
T3 S
2 £C TJ
• 2 - H<H <u
^ C
re rTi
t <J
1 0U PQ < •>
53 tc
- o
II3 .a
t/5 re
S aTJ C'35
£ £<U 5 C «
cre oC T,
3 re
CJ X
b> - °U C J=V O ^
• 3 u ( / lT2
be C
Cu I S
S S S TJ
j2 ^ XJ be c
- - £ re.a ua 2
re G . 2 —"
•S -0 y 1
o _;OJ re'° "Cid uO (^
.5 u ^ re &.-0
5 I*H u
£ <•" c
O o o
I "P S .5 > c 35H u e "> n v C
unse
0O
c>"cD
u0<u<J
0
cE
Fi-
ess
.£3
be
Vic
e
(J
0<u
OS
cre
eral
G<U
O
3~
nen
0Cu
QXJ ffi
0J qj
Q
'« - s .fa O53 J*- , ^ .-
TJ .G
ii fi-£O
- & <
•x <j
O c < cq .S h £ > Q
o ^^ ' 3:> a
,-a £ £
R .- «
.5 y
s 3.22 PuG ^
E C
s- be, 0 re
fi|TJ «i
-0crebec'Sc
ua,
'35 E - ^
u . -
c .«
c tx:
o oV VJ
—
© "CMCO
CO^ ^CO0 0en
COCM
© "1 - ^
COCO
—1 CM© —CO CO—1 00—1 en
© coCO COCO - H
m
1 Om coen co
£ ^| 2 ©2 ° CO COg ] CM CM
i o ©(2 O) *CM CM CM
• S 0 0 CM
ai © <crt M tOCM CM CMA © ©o en ©© en Tf—< CM coCM CM CM
in inm ©CM 00
© ©-H ©
oooCO
om©
©©©©
©©moo
7©©©oo
[ant
Cu
TJ
re
c.2
dmis
j
<
TJ<
_rec
Fina
i
C
TJ
its a
n
0
ati
„
TJt ^ - E Q
CM <
ETJ<
"re;_DG
C.0reu
o
•f-1
>c
•3cen
/ )OCJ
c"0
His
t
c"o
truc
t
s
bec
X !CJi_
wCU
res
eg
•oG
C
s.sti^rS1_O
o-a
0Cu
:tors
,
.s
rati1
tn
'5'ET3
&5en
OJCJ
CaQ
13
ican
c
>
H en
I
c o y
S °. I"-O bC 3
< .E o
c °°o be' £
-S t
rtCuCD
-o
eral
subebec
oaCU
en
&cu
>~>* J
c,oCu
exce
xTus_CUentUl_
_ ^
5
men
a,
>2
CUCU
X I1
nce-
Oa0
nts
t
wE
enk.
_O"CU
ou
nj
CJ
CUen3O
X !
1 | U 2 53 cu I) u U « • =
en cj g o (G (MS -73 '> 3 =3 3 "0
•3 rs
ra C TZ c/i
I 111
e« cu
« C 3
CJ Cu cu
§ 8 I
E & ccu "OCu be 3x c stn cu «-j
CJ O f^
1•2 c•S o
2 ^a 3
W CU.S <*- cj
I- c
V 3 -W C -S cu -S
8 s -s g s ien ^ 3 rt
UH 3
« c S " ° *cutE 2o E
3 be p
c c ccu OCU'3
._: «.5 <u £
^ 3 CU
n<« 10 u
cj X!C iri
B J «
^ 5 !2-o a'Sc cu «o3
>a. t, -Q Cu -D
u E JS c cu =» o W W 3 3 .2 u 71 Cucy) Xi :
13 o .
c1 EO CU
.i! o2i £ x
<*-i beo c
X! -ObC'en
•33 - a
O "3
a 5 ^oa; .E ^ o(j t- 3 o
cx: CL, 3 be
©©CO
co"©©CO
co"
©CO
©©CO
__l >COCO
11
©mCOcom"
©CO
©©©CO
00'
©CO
00'CD©CO
CD'
©CO
CM COCO J>© ©CO co II
© 00OO 00 CMCO O CDCO CO CO
—1 ^ ©
co r^© ©co CO o ^
CO cO
1 > © CDCD t ^ CM© © CMCO c o c o
m CMm i>© ©CO co
to ^~© ©CO «O
©" I>CD r^CO CO
© CD 00—' CO CMCM © CMCO co CO
CM COm co© ©CO CO
CD CO© ©CO CO
COCD©CO
|COCD©CO
©CO
© "
©CO
i nCMCMCO
_<"CMCMCO
1
CDCO
I >CMCDCO
Tf CDCD CO© ©CO CO
00" CD"CO •<*© COCO CO
00" ©in co© coCO co
CO COI
CO
CM • * CO00 m coCD CD inco co co
-H 00 CD00 CO CDCD CD mCO co co- . I
© Tt< CO00 —i •*CD CD inCO CO CO
— < i n © r ^ - r ^ C D © C M c om c o c j i i n o o c D c M C M ^ H© © © © © C O C M C D C Dc o c O c O c O C O c O c O c O c O
- H CM© COCO CO
. ICM - H
—' in co CD -fr^ co co 00 coi> in CD CD mco i> co co co. . . I I
i> Tf in co ©m TJ< co 00 ©t> ^H m CD inco i> co co co
" ©" CO" Tf"x> 00 d•^f CD CDCO CO CO
1*3
ccaenCCU
aUIO
0
3Ua
UH^H
C
"3beHai
caenCCU
ZJV
(2uV
cux:0u
0
c<nenCCU
aUIO
0 u
rs
ens
aUIO
U
2ccu
- 0
_O' e n
cuc3O
eu
u
vie
0
CJ
uC0
U
en
ute
cuE0
0
—' —
CU <U
0 £_ CU
tfl a .
tu re
CMentO
entoCM<—1
Cnto
oto
m"ento
Tt<"
into
tO
tO
•>*<
to
o
to
COentO1
00CMento
to"CMento
©
to
mtoento
toento1
tOmCDto
CM
ento
^
ento
CO
ento1oento
ooooto• * "
ooentot^
ento1
CM
to
to
in
O
*-H
I>
©"CM
©©
enento
oo
mto!—»
1|
CO
toto
en
©toin
en
|oinJ>
©"m
©
oooo-»
COen•>CMenen
©to00
_rin00
©moo
intom
"CM
CM
to
oo©in
oo"totoj ^
totO
m ©en —ien to
toto
f-"entoto1oototo
m"
tototCCMtoto
©"©toto
en©mi^
COto1en
COtoto
i-H
tot^
enen00to1
©CMooto
to'
ootoi
iCM©ooto
£ be
a. g
V ac o4i a .
2
3 .a
«J 0
1 iT3 .2
u1 « l
If
re
ac
obo
W
T3
're
Q
it
re
£V
o
£
'£V
use
0
s-Tre£
0CQ
qj
<U
Qj£re*JboG
'2,re<uH
_£
O
£
tio
re< j
Ed
ubo.£'3£
£re
£
Fre
£_3~3u
:_3UV;_0u
Q
"53to'—H
o
'C
£O
cati
3
to1
_O
£
Xto1
>
ive
£
ter.
£
U
Vo00
1g
c P
'Sil i•£ T3to c
re 3
u 3
3 ^• - re
Gre
bore
- "0
qj ^
"§ C^ 50
be 3
bo
"0 re E?- ~ OJ
re « £
bo Ore -c .3
' 3 £ >bo re cS <L> to1 iTJ re°^ §
* 1 > e0J~ C/3 ^ to
- " " («_ „ P qj
S s-3 §u
~a £ re3 re OHv '35 re
u —3 £
.a - £u £ re
J o
t- .S w
^ W D
bore
1 1 Sf. (U £i ^ be OJ
^ 3° £ 5S ^ g£
2 °PH £
-£ rere N
.i! bojj
^ £ - a
£ .re <
t« re re
w qj
£ 3£ o
re
o.2 £
£ i-•- ^ -2
5.2 < ^< bo - .
•ti x£ wo . 2 bo1
U _bo revT 15 befl pi £o ,,
3 > £ c
£ Q
re ^
C £ .SiJJ
OJ £bo O
T3 qi^3 qjO «
5
§23
3 re i/ £u O
1to i i o
^ S OJ c^ § ^
u bo -
bo aj
.2 « H
u 3
O B c
re ^a- E
II§ EO V
£ 3
£ "reu
'S.a^ 3
"re "§
§ ; ^=3
- ££ §£ "
£ re
• 3 - ^ "- £4
fi.I£ -2
^I
rebov
« £CM wCO O H
^ OJ
s *co «i
rebo
j _ , co
O CM
3 3re re ^ re <2bo b o " ^ bo Qqj qj co <u rr
i i , CO qj f~l qj
n i> o -O o3 —i u 3 u
U
oT3O
CD O cojC m o CM
° to" CM CMSC 00 I > O-2 o co —i
rebo
£_£
O00
£
_£
CO
O —iOH"*1 0 -00 t/5
5 O "3 o re
O O S
re
ene
a
g-a
£re£3X
.9*~0 0303 b o"c3 ° «
CU
- E4J J
a uo c'So 50 fe^ 03
CU
p ,
~ fi
1*CD ^
03 — H
- o ., Q
=?, ta <o1-£w c0T<cu —<
a,« 03
| 83 OJ
,9 rt2 £S §'£ 3o wfi Cw
8 °
sicQ 03
boqj . -
*-5—. 03o3 rqj'0 ho i
CD I03
^ .asi3 CD03 03fi 'G- o-o ^
§ e03 3
O £
s5u
* 5°
bo-fi
OH <
cu T3 qj
fi 3 Q« w «U - c
• i f Iv o C^ ^ §
" fi £<v V C
IS3
I I•-H O
3 fe
1 ^ 1 S £cSI .S -11 £ a
Q .2
£ P
-r 2 ^ S fi
bc^ «c 03 r1
CU
c QO r/,
=§ „-1 "i <S « I.S-?P H
fi
5 cj 03 c/J .'S
2 1 1 1 1g § a s «2 ^ ^ fi 03
C-M fi t« ^^- O •—i CU —j
l " f i b e l l) Q
3c qjbe y
IIrtbO.O Q
03
3 fc- («"
s_, CD -fi -aqj > ^ 03 j j
fi .U
oi ^PQ
o bI
s w >-QJ
fi'.y
S 3 C <U
3 a.2 "OH fi .22£ o .> .O U Q bo
<u fi . 2 xi5 o c a «
15 c -o . 2 «
3 iJ CD 3fi 5 fe "o. g^ c
be.so >-
II c o03 X>
v. § So 3qj CD
fi .y•« £
> M
—< ^ 03o3 qj b o(H <-> Si
« fi O.v u «r
O cu - cu
u £
a o
« u cu3 0 3 *
ho 2!- « —
be 1
c fi -fi <O P O- fi
{-4 p-H PJM <—Hqj OIo o. S O U
' U Q* 2
*-« -fi,< CD
- 2 CQ
UCJ 3 ^aO
rt ^
«.2 5e-a pq ^fi - o03 >p 5
N
^H SM
IO
3CD
CD*"
<C D
o11
'>cu
Pi
-a03>u03
en"
CUufiOJ
'GCD
- ^oo
pqC/3
n.
illi
]
PH
fl. 0
03
QJ
0CD
<
O
X
c003
<
irm
s
oQQJbecu
"or j
fl0303
<u033O
cu
O CD
"O CD
2 O
H 03 3
.-2 ^ .fe
qj rt «*Jbe 3 <
£ JJ T33 S3 o3 II
-Q30303bo(U
£
o
P CO
03
be
CO
o*
'-' £oo" fi
OO
- Q303
gf5g
5; ^ oo01 CO en „3 OH 3 Tt< &«? § « ? § « «
b C ^ bO a , ^
^ CD S Q CD
a,vQ
- 0300 beCO QJ
O O„ -o. «S^ 3 0-
CDcu
CD
oCD
2-S5 3J CD
wh-l
S
O <
re 3Sfio
.-H U 3oj ft , en
P i - <+••T3 o
£ S SOJ u re
3 IsG1 a3 3
CQa *O J ~ en <u
en Q'So -
U
h£ |
N 1O 00
3 .S3 S
2fi-tf
en . 3(U -Pi S ^
.3 OG =3 O X
5/3 U ^ g
5 ^ O rS .2CQ > Q U en
o S- re
m beCM OJ
^ oen <U
3enbo
s
H ^ CMen 5
J3 2en qj
0GreenGO
"5'<re
'0G
a- rs
*J G
c .2
. o
^ e re
glS•a « OJre .y t i
FH — re-2 JO 'Cg a o
s H t o 1f j . Qj "2 en C«
00
j 3OJ
«
0
<
bojy
O
U
G
^23
V>
~ en
315
2 I - 3 S2 c
be.2
- S GG
1
en u— re
•.i§> u
O OJ
-are "re
boG
V "3
o
o
. o
O S
be
S CM
11
P3be
oCMo00*
o
rebe
3HPH
O H30
0
O500
10000en
OHV
-a
en
0000
00
en
0-
en
m00
aQJQ
G J]
GVaG
"S0-are
_G
:ion
retnG(UOH
a0u
on
e
Gre
OJ
O
a
GOJO H
a0u
1G
tio
CJre
T3 3 ^ g
Jiff<U G >en . - CUaj fc en
^ s-2 " Lp, frj c*-i
g CL ^d* " cl
i23
03en en
G 35O *J
CL
£
g « 8
eoCM
00
o
03uots
enCU u
OI - i
sa-
Gi aj« CL
§ sn 0U
Stu-
I1G
_0
Vo
"0G
GCD
CL
S G
^ > " o 3
3
b p _03 C ^5 J3
G
w V3
o o9
cGaj
bec
^ G 2 S Saj 3 03 03 O
^ g ™ fe aG 2i ^ 2 §
_O CL g -Q ^CJ ^ ^ ^ P ^ en
-° bC C J3 "53 o3 ™ i ) -»CL tn *->
"3 o. bo
$ sg a
. £
oo
G03
s 2
v"i G
bc-53 U 2 Hj
-aS
Gv
.9"cr yaj * ^
G tj)03
aj .5 i—
j2 o~ft cu *iS £ Q
o3 *C
i o '£ ^.£ '5 £ fc
U - j e H r i . '
^ -5 0 SH^ i:
eo ^fCO CM
O * CMCM CM
co co~- CM
en03
u
en
U
en03
u
en CTi03 CO
U u
CL
0
X
beG
U W 003
u
beG
so3CLG 13
-2 cj
G o3
• ccu aj
S |V GG -0
be"rt
I 2'3 3
siCLX!
equ
0
ecti
op
03
aj
03
ten
_G
G
lars
oX,CJ
ants
bo
oho
O
O o
03cj en
CU OCJ • £in IM
c t;O 033 CL
"o-6CU . 5
aOcjaj
-a
"oG
.Gcj
U OCL q,;
CJen (_g
" o
bo
"o3 O G
00CO
COCO
CMco
oCO
CM
•— £ en 1- bO
2 "S
CO
CL U CL
V5 C/2 C/3 V3C/5 ^ 5 C/3 C/3
u u u
O5 O
C c« C
^ . 2 8 g
IP I3 s- * -H
03
U03
U
aou
obey
IS
i-J
3
cOJex
JJ
Q
Q j
-a «re S
c <S «o S P *II
!!_ <jre • «
re £ QJ aj"77 v to uO rh, re 3
T3 >
"3 c75
re «
-C QJbe h
I C[re j
tyj re
= re x re c
re en C
w be<u re
rebe
lQJ reW 32 bC
| | l a l/C C 3 QJ ••-
3 J
£ S c
I 1 'IC o °
O ^CM P
QJ
QJ
CM
1 00
oCO
SO
oCO
a; p-
U 3G • - -
•- ?u —
t^ "0 —Si
' 0 -Oo 3/3 ^
§ beu o
w ,<£
. obe go o
"3 ua. B
be Co o
CJ ' ^ _O
ath
3
0
* ^cw
rce
0
oJPi
(- O
8 S3J3" be-a ^
33•-- 3 re . y
33 C
u
II
£ £QJ QJ-C £o o33 S
C CO 0
5 g:g0! UJ •=•
O en
QJ <
^ 1 3QJ 3
cC c/5
S =
t -aQJ C
S SQJ en
" O QJ3 .y
bcaicccre
Xoooo
N
reX—o
,QJ U "O
— QJre tyjco c a !
O enC C>s O
U
3-~ B> o
u tH
3, £
3 w
0 i-i
- Kre re
§ U
' 3 enre ' '
0 QJ.c be^3 re3 c
O ^ ^ R
>, beoZ
c = .2
"O •
^ o« « £ M 3£ £ "2 c "«O O re W re
o CM <*^ ^ co"ti u COa. v CMQJ XQ ui O Q
oO CO
agQ uj
COCD
1i
mCOCO
_TCOCO
60,
CD
Q.QJ
Q
-
CMmCD
49,
CO
a.QJ
Q
CO"O^ *CO"0
CMO
CO
3 IaOJ be-o a3 ~
a c
il
.2 O'
to T3
re «
3.2•2 .S
U O c n
Pro
cial
OJ
Q -
CD
'ouO
ice,
Of!
aOJ
a3ao
13
for
eOJ
U
a-OJ
_CJ
•—
<
nal
_o
'>5<vV
U Q OJ
c w
OJ OJbo u.SiSa Oo -o £
re G OJ
<13
9 f s «OJ
0
OJ
IIo oOJ OJ OJ
o •- £ '££ Z £
,2 >T~ 3O O O < -a X
CD P
^ T 3
•3 <
11E .2o «
•o 'Q
toQ
.£ c3 - 2 ^a re -oE a ?o .2
<*•> B . ^o re Q. be>..£ 3 .c
be o >-,U c -o .S
. 3 OJ
OJ M
CJ £
o ytn W
O >T
- B OJ bc'5T3 <
£ c -ot O OJ
.o -a ^
Q 2-f,
U -3
a in ^ j ^ ^ S j J < n 3 f i
S ^ U .2c 3 * 2
i ^ z «"3 CH ^ .2"> ° a -S
•5b «i 1 -3v re c a
3 3 o OJ
.a - O .aOJ
c-3 '3
o a?, «
H 3 fi— -^ .2
OH rt
"33 B
II§ i -s c g
, OJ <+H
£ •£
3 3 CDO..2 U
£ 3
- i! 5 o S j
S z S l 9 . 2 r § ^
S a im JS c « •; S
„ E CD
c < •r0 — S•a re ore c '33a .2 •>
u u c
§ § ' § f .• a "33 CD O r
«r -^ c a OJ
§ OJ >. J ^ P H
i- .S b c OJ
1- a -a a OJ
OJ OJ a OJ u£ £ re Z £
I fl f °1 &s 1 1 |— —• £ a n
o 2 -3 o OJ
tn 3 re
.£ 0
O ^
OJ „ _ _
33 c 2
S-2-S I 3 .a isce « & S
« £ U r| S!
£ pq O o
a.D
se,
e
p l
>n~0 0
0 0
Ot-4—1
OOJCJ
a.CDU-t
0OJ
te
OJCD
£OJ
Q—<
obo u
3-o
be
1a (« vD IS •>
1*O
OJ ^
6 .2
- re
sOJ -be o
o aOH
T3 re « 3
s is
re y pq
c 5 ^ u1
3 O
=? £ s - :2 .a a
a^OO is 3
" CD 00_r •* o2 r-» —•
^"2CD
mCD
i n
OJ
Q
-CD
0 "
CD ^
H ^D-oOOJ
Q
0CD
CJOJX
to1
rCD
eo"
CD
a,OJQ
00CM
m"0
aOJ
a
^^CM0CO
^CDCD
m"c^
>n"co0 "muOJX
w
to0oo" ^
IT* in"^ 0t>T • *
OJ
Q
622
Oico10a>CD
711
0i >
m"0i >
00CDCM
0 0i nCM
G .2
-a<"re
inan
ci
to
vost
0
z.
'S3
<[UU
I
3<
_O
>OJ
Q-aare m
enl
_!_'0CD
-aare
;_,
'E
nee:
OJ
a_g
stra
l
. —' c «OJ
O
•£•
wPQ
S
9*
o
Vbeoi
-ac03
V
"C-3"!3C/i
"w3
om
4Jbe03
-acOS
°n_oS"oSen
OJbe03
T3G03
c/:pJJ'C.3"oiIT,
be03
• a
OS
tnU
j t"o3IT)
co'c3
be bo ^S S SVs sa. a. t!S £ S
o o r. 3
^ D
3 a.
"3 o-
c c ^ S*G o " -i H
1 SIss c ^
3 O 503 " 3 >
•• S 'c
co U Z
-a ^j
£ . 2" § 1T3 r" «i'(3 D ^
2 S J
S I IS 1il^7
OS -T-f ^ "Z2
3Q f2 •-
•lilt-a•s s-s.
uC M •" U S
•3 £ XI X! .N
a u u a; tiU pi c/i **S CO 2 U 03 Su a-oi 3 -a
II
II
ocn
1COo
enenen"?oooo
enenCOoo•
or - H
CO
X3
CO
en I
§7§§
en en<x cooo <o
I Io oo o0 0
CO
00- 1 CM
„ OO GOO rt
2~ ^ enc c 2P3 3O 0 O_ _ O03 03 00X X t~-3 3
CO CO
= So _Lb 203 00
X - ^3
CO
o q£ -I M £o oU U
1
ca0
a
G0
z0
'•3
03
COidtu
V
O
fess
i
0
ces
ervi
CO
Co(JX3co
3a£0
U
e[BU
I
e s
CO © CD
§©o>"
© CD© CMz; ooCD CD
„ Iin ©oc ©CD 00oc CD
© CD© CD
CD —'00 CD
CD" ooCD ©
00 CDI
© CM© ©CD - *f ^ CD
I©
<£) CDCD
CD" 0 0CD 00CM ICD ©
§ «fSo<J ©« CO
X l CO3
CO
1©©COCO
CD"CD
1©©
c3OCJ( j
CD©CD
|
©CJ)
c30t j
X)3
2 u
° g"•g w<u beC CuO
a 3
u
134 CATEGORIES USED TO CREATE EXPENDITURE TABLES
TABLE 4A.5Expenditure Categories, Carleton
Departmental Group
Humanities
Natural Sciences
Social Sciences
General, Other Aca-demic Departments
Organized ResearchLibraryAcademic Administra-
tionComputingStudent Services
Admissions and Finan-cial Aid
ScholarshipsGeneral Administration
Alumni, Development,Public Relations
Plant
American Studies, Archaeology, Art and ArtHistory, Arts Studies, Asian Languages and Lit-erature, Asian Studies, African/African-American Studies, Classical Languages, Cogni-tive Studies, English, Media Studies, Germanand Russian Languages, History, Judaic Stud-ies, Latin American Studies, Linguistics, Music,Philosophy, Religion, Romance Languages,Russian Studies, Women's Studies, ColloquiaPerformance ProgamBiology, Chemistry, Geology, Mathematics andComputer Science, Physics and AstronomyEconomics, Educational Studies, Political Sci-ence, Psychology, Sociology and Anthropology,Urban StudiesIntegrated General Studies, Physical Education,Studies in Technology and Public Policy, Learn-ing Disabilities Program, Science and Ethics,Other Educational Programs
Instructional support, less computer center;registrarComputer center; data processingStudent services, less admissions, financial aid,registrarFinancial aid; admissions and minority admis-sions; student financial services in 1992Student aid, less faculty and staff scholarshipsAdministration; general institutional, less dataprocessingPublic services and information (1977); devel-opment (other years)
Type of Expenditure
Compensation
Student WorkSuppliesEquipmentTravelOther Expenses
Salaries plus staff benefits allocated in propor-tion to salaries; for faculty, also includesfurloughs and faculty-staff financial aid
Includes maintenance and new equipment
Source: Carleton College, Report of the Treasurer, Schedule B—2, "Detail of CurrentExpenditures," 1976/77, 1981/82, 1986/87, and 1991/92.
Appendix 4.3
Trends in Duke Expenditures from 1976/77 to 1983/84
As NOTED IN chapter 4, machine-readable financial data were notavailable for Duke or Chicago before 1983/84 or for Harvard before1981/82, making it impossible to conduct a full analysis of trends inspending at any of the three universities over the entire 15-yearstudy period. Almost no machine-readable accounting data for Carle-ton were available, but consistent tabular information covering theentire 15-year period had been published. As a partial remedy forthis deficiency (mainly as a check on the representativeness of themost recent period), accounting records for Duke extending back to1976/77, stored as microfiche, were examined. Because it was im-practical to undertake the kind of full examination of expendituresby unit and type made possible by computers, aggregated data in theform of fund codes were collected for a selected number of depart-mental units. In order to use these data to get an idea of trends inthe earlier period, two questions were posed: First, using the fundcode data, how did the growth in spending in the earlier periodcompare with that in the period 1984-1992? Second, how do thetwo types of data compare for the period for which detailed infor-mation exists? If both types of data yield similar results for the morerecent period, then one can more confidently rely on the fund codeinformation for the earlier period.
Fund codes are the basic accounting unit used in most private col-leges and universities. In the case of academic departments at Duke,a single fund code often covered an entire department over the en-tire 15-year period in question; this comparability allows at least agross analysis of trends in spending over time. On the administrativeside, however, fund codes appear to have proliferated in accordancewith changes in organizational changes, making it necessary in mostcases to add together a number of individual fund codes to approxi-mate the total expenditure for any one administrative area.19
Table 4A.6 presents a comparison between fund code data anddetailed accounting data (for unrestricted expenditures only) for se-lected academic department groupings and administrative areas.20
The first two columns compare the 1992 dollar amounts of expendi-tures, giving some idea of how closely the total from the selected
T3OU-oC3
Q
Pi
oO
8 17^
00
SI" 7 ^
i O ••—' C I O^ ^
^ d> d> ~ d>
t» N - (O lflco -^ 06 -^ TJH
CS -1 O) Oin r^ CM eo
d o o coo in CM co o i> —1co •* Tt< d iri —< CM
o co TJ<
CO CO OI I
10 ^ q in"^H T^H ^ _ | l£} —< eo —• 06 06 i r i co
CO CO co coCO CM CM —<
• ^ O CM 00CO CM CM i r i co —1 CO <X> o i CO
o 00 co coCM d 0 ©
ao
S uS 3u «
a. 2
u o
•5 I
•s ^5 "C «J 5 s
E w3 3 g « • =
•3 J J 'S c2 z s M u< < I i i D O
—i - oo oo—; i CD -<f
^ E
J2 c
•X "-' -a i- TP
•5 E
g ^ 7 Sco o 5 "0
C O .5 (C =
0
00 c o * -=O 3 —c rt .2
t - .!3 - 3
*2 "0 IO
"0 T3 Tf
« "- *^ »n m
E o
a "^ 4; s
TRENDS IN DUKE EXPENDITURES 137
fund codes approximates the total unrestricted spending as calcu-lated using the detailed financial data.21 Comparisons are made foracademic departmental groupings (the same ones used in Table 4.2,specific administrative areas, and several categories of financial aid.In general, the amounts shown in the first two columns are similar inmagnitude. The similarity suggests that, in most cases, the fundcodes chosen provide reasonably similar coverage to the tabulationsbased on the detailed financial data. The largest differences are inthe financial aid categories. Other than simply providing less cover-age, one reason why the total based on fund codes is smaller is theomission of salaries to graduate student instructors, which are addedto financial aid in the total unrestricted numbers. However, thereare other categories in which the differences in totals suggest thatthe two sets of data are not comparable.
Turning to the growth rates for these categories of spending, it isinstructive to begin by comparing the last two columns, in order toassess the similarity of growth rates calculated for the period 1984-1992 using the two methods of measuring expenditures. In a major-ity of categories, the two growth rates calculated for the period arewithin one percentage point. Where the rates diverge significantly, itis for a category in which the two measures differ in coverage. Insummary, even if two different bases are used to collect data, thegrowth rates for the 1984—1992 period appear to be fairly compara-ble in most cases. This comparability gives us some confidence thatusing the fund code information to detect trends over a longerperiod will provide useful supplemental information on rates ofgrowth in expenditures.
With this background, it is now possible to compare the growth inexpenditures based on fund codes. Was the period 1977—1984 oneof slower or faster growth compared to the 1984—92 period forwhich detailed data are available? Looking first at the four majordepartmental groups, there is little doubt as to the answer: real ex-penditures accelerated rapidly in each division after 1984. Engineer-ing led the way in both periods. Marked by the hiring of a numberof prominent professors beginning in 1983, the humanities divisionshowed the most dramatic jump in its rate of spending increase. Theadministrative units shown in the next section of the table generallyexhibited faster rates of growth, especially in the earlier period.Whereas three of the four academic groups had growth rates of lessthan 1 percent from 1977 to 1984, only one among the selected ad-ministrative units did. But these rates of growth generally subsidedduring the second half of the period, falling from an average of 5.7to 4.2. The unit displaying the most rapid increase over the entire
138 TRENDS IN DUKE EXPENDITURES
period is the university counsel.22 There was a similar racheting-upof spending in the selected financial aid categories. No doubt theitem with the largest impact was undergraduate financial aid. (Recallthat these expenditures are taken from unrestricted funds and thusdo not include endowed scholarships or government-financed aid.)After growing at "only" 2.5 percent per year in real terms before1984, this item exploded after 1984, growing at a 14.4 percent rate.Moreover, although the magnitudes are smaller, there were similarjumps in three of the four categories of graduate aid, with the larg-est increase recorded for humanities, perhaps indicating the degreeto which support for graduate students is considered a necessaryingredient to improvement in academic departments. Overall, thesecomparisons suggest that the growth in administrative expendituresslowed after 1984 (although continuing to increase in real terms),whereas academic budgets generally grew significantly only duringthe later period.