Cornell University ILR School
DigitalCommons@ILR
Articles and Chapters
ILR Collection
1985
Institutional Environments and Resource
Dependence: Sources of Administrative Structure
in Institutions of Higher Education
Pamela S. Tolbert
Cornell University, pst3@cornell.edu
Follow this and additional works at: http://digitalcommons.ilr.cornell.edu/articles
Part of the Higher Education Commons, Labor Relations Commons, Organizational Behavior
and Theory Commons, and the Work, Economy and Organizations Commons
Thank you for downloading an article from DigitalCommons@ILR.
Support this valuable resource today!
This Article is brought to you for free and open access by the ILR Collection at DigitalCommons@ILR. It has been accepted for inclusion in Articles
and Chapters by an authorized administrator of DigitalCommons@ILR. For more information, please contact hlmdigital@cornell.edu.
Institutional Environments and Resource Dependence: Sources of
Administrative Structure in Institutions of Higher Education
Abstract
Two theoretical perspectives are combined to explain the pattern of administrative offices in public and
private institutions of higher education. The first perspective, resource dependence, is used to show that the
need to ensure a stable flow of resources from external sources of support partially determines administrative
differentiation. The second perspective, institutionalization, emphasizes the common understandings and
social definitions of organizational behavior and structure considered appropriate and nonproblematic and
suggests conditions under which dependency will and will not predict the number of administrative offices
that manage funding relations. The results of the analyses indicate that dependence on nontraditional sources
of support is a strong predictor of administrative differentiation and demonstrate the validity of integrating
these two theoretical perspectives.
Keywords
higher education, administration, resources, institutionalization
Disciplines
Higher Education | Labor Relations | Organizational Behavior and Theory | Work, Economy and
Organizations
Comments
Suggested Citation
Tolbert, P. S. (1985). Institutional environments and resource dependence: Sources of administrative
structure in institutions of higher education [Electronic version]. Administrative Science Quarterly 30(1), 1-13.
Required Publisher Statement
Copyright held by Cornell University.
This article is available at DigitalCommons@ILR: http://digitalcommons.ilr.cornell.edu/articles/425
InstitutionalEnvironments and Resource Dependence: Sources of
AdministrativeStructure
in Institutionsof Higher
Education
Pamela S. Tolbert
? 1985 by Cornell University.
0001-8392/85/3001-0001/$1 .00.
Helpful comments by Paul DiMaggio, John
Meyer, Jeffrey Pfeffer, Robert Stern, William Stevenson, Ray Zammuto, Lynne
Zucker, and the ASO reviewers are
gratefully acknowledged. Special thanks
are owed to George Jakubson for his extraordinarilygenerous assistance with the
computer program and analytic
procedures.
Two theoretical perspectives are combined to explain the
pattern of administrative offices in public and private institutions of higher education. The first perspective, resource dependence, is used to show that the need to
ensure a stable flow of resources from external sources of
support partiallydetermines administrative differentiation.
The second perspective, institutionalization,emphasizes
the common understandings and social definitions of
organizational behavior and structure considered
appropriate and nonproblematic and suggests conditions
under which dependency will and will not predict the
number of administrative offices that manage funding relations. The results of the analyses indicate that dependence
on nontraditionalsources of support is a strong predictor
of administrative differentiation and demonstrate the
validity of integrating these two theoretical perspectives.
Inexplainingthe formalstructureof organizations,classic
organizationtheoryemphasizes problemsof coordinationand
controlof workactivities(e.g., Taylor,1911; Weber, 1946;
Simon, 1956). Reflectingthis tradition,researchon sources of
growth inthe administrativecomponent of organizationshas
typicallyfocused on factors such as size and complexitythat
are assumed to impede the efficientsupervisionof tasks
(Terrienand Mills,1955; Bendix,1956; AndersonandWarkov,
1961; Pondy,1969; Blau,1970; Hsu, Marsh,and Mannari,
1983).
Recent workon this problem,however, places greateremphasis on environmentalrelationsand influencesthanon internal
relationshipsas determinantsof administrativestructure(Freeman, 1973; Meyerand Brown,1977).A numberof different
perspectives on organizationalenvironmentsand the way in
which environmentsaffect organizationalbehaviorand structure have emerged. Inone, the environmentis conceptualized
interms of otherorganizationswith which the focal organization engages indirectexchange relations(Levineand White,
1961; Thompson,1967; Pfefferand Salancik,1978). Administrativestructure,fromthis perspective, reflects efforts to
ensure a stable flow of resources and to manage problemsand
uncertaintiesassociated with exchange transactions.Increasing dependence on exchange relationshipsproducesadministrativedifferentiationas organizationscreate offices and positions to manage these relationships.
Ina second approach,the environmentis conceptualizedin
terms of understandingsand expectations of appropriateorganizationalformand behaviorthat are sharedby members of
society (Zucker,1977, 1983). Such normativeunderstandings
constitute the institutionalenvironmentof organizations.Organizationsexperience pressureto adapttheirstructureand
behaviorto be consistent with the institutionalenvironmentin
orderto ensure theirlegitimacyand, hence, theirchances of
survival(Meyerand Rowan,1977; DiMaggioand Powell,
1983).
This researchintegratesthese two perspectives to explain
administrativedifferentiationin colleges and universities.A
centralpremise of this approachis that dependency relationshipscan, over time, become sociallydefined as appropriate and legitimate.Itis hypothesizedthatwhen relationsare
1/AdministrativeScience Quarterly,30 (1985): 1-13
1
Theintense debateoverthe bail-outof the
ChryslerCorporation
bythe federalgovernment is a highlyvisibleexampleof the
institutionalized
natureof dependencies.
Criticismsof the government'sactions
clearlyreflectednot onlylong-runcost concerns, butcommonconceptionsof the
appropriateness
andlegitimacyof public
supportof a privateorganization
(see
Chapman,1979; Friedman,1979; Samuelson, 1979).
institutionalizedinthis way, variationsamong organizationsin
actuallevel of dependencywill be unrelatedto the numberof
administrativeoffices associated with the managementof the
relations.When an organizationenters intoan exchange relationshipthat runscounterto institutionalizedpatterns,however, the maintenanceof this relationshipwillgenerallyrequire
intensiveadministrativeeffort. Itis hypothesizedthat when
relationsare not institutionalized,increasingdependence will
of administrative
be directlyassociated with the proliferation
offices to manage the relationship.Thus,the institutionalenvironmentdefines the conditionsunderwhich increaseddependency leads to administrativedifferentiation.Thisresearch
investigates these hypotheses, focusing on patternsof dependence among highereducationinstitutionsand the number
of administrativeoffices that manage these dependencies.
Sources of Administrative Structure
Dependency relationsthat are not institutionalizedare generally less predictable,more uncertain.As these relationsbecome
an increasinglyimportantsource of supportfor the organization, the numberof administrativeoffices and positionsassociated with the managementof the relationsis likelyto grow. In
this case, the magnitudeof dependency will predictadministrativedifferentiation.Formaloffices are created to serve a
directlyfunctionalrolein negotiatingand managingthe demands and problemsaccompanyingthe relationship.They
serve a symbolicroleas well, since theirpresence can act as
an indicator,or signal,of the organization'scommitmentto the
exchange relationship(Spence, 1973, 1974; Meyer, 1979).
Thus,the institutionalization
of dependency relationsdetermines whether or not increasingdependence willdirectly
affect the proliferation
of formaladministrativeoffices within
organizations.
ADMINISTRATIVE
DIFFERENTIATION
AND PATTERNSOF
DEPENDENCE
Inapplyingthese ideas to an examinationof administrative
differentiationin colleges and universities,distinctionsmust be
made between publicand privateinstitutions.The two types of
institutionshave a long-standingtraditionof drawingon different sources of financialsupport.Publicinstitutionshave typicallyreliedheavilyon governmentalsources of support,especiallysupportfromstate legislatures,while privateinstitutions
have receivedtheirincome primarily
fromtuition,endowments, and gifts and grantsfromprivatedonors.
These patternsderivehistoricallyfroma majorlegaldecision
addressingthe issue of state controlover institutionsof higher
education.Duringthe colonialperiod,state governments provided substantialsubsidies to privateinstitutionswithintheir
jurisdictions.In1819, however, a Supreme Courtrulingestablishedthe autonomyof privatehighereducationinstitutions
fromgovernmentsupervisionand control.Followingthis decision, states rapidlywithdrewsupportfromthese institutions
(Rudolph,1962; Brubacherand Rudy,1965). Whileprivate
institutionsbecame increasinglyrelianton tuitionand privately
sponsored endowment funds as primarysources of revenues,
an ideologydeveloped to accompanythis independence,
promotedlargelyby the institutionsthemselves. As Rudolph
(1962: 189) noted, "Beforelong, college presidentswould be
talkinglikePresidentEliot[of HarvardUniversity]as spokesmen for ruggedindividualism,forthe virtuesof independence
and freedom fromstate support."
Overtime, differences in dependency relationsfor publicand
privatecolleges and universitieshave become institutionalized; dependence on differentsources of supportis viewed as
an appropriatedifferencebetween the two types of organizations. Patternsof supportfor publicand privatecolleges and
universitiesprovideevidence of the institutionalization
of these
dependencies, as shown in Table1.
Privateinstitutionshave regularlyderivedapproximately10
percent of theirincome fromprivategifts and donations,
whereas publicinstitutionshave typicallyreceivedonlyabout 2
percent of theirincome fromthis source. Publicinstitutions,on
the other hand,have been able to depend heavilyon state and
3/ASQ, March 1985
Table 1
Percentage of Revenues of Public and Private Colleges and Universities from Major Sources of Support,
1939-1974*
1939-49
1953-54
1959-60
1961-62
1963-64
1967-68
1970-71
1971--72
1972-73
1973-74
State
Public
Federal Tuition
61.9
65.6
59.4
44.0
42.9
44.5
47.5
46.3
47.1
47.9
13.5
16.0
21.6
18.6
19.1
17.8
11.7
11.8
12.1
11.1
20.3
13.3
13.2
10.4
11.2
11.6
13.1
13.7
13.4
12.9
Gifts
State
Private
Federal Tuition
Gifts
1.9
3.1
3.4
2.4
2.2
.6
1.9
1.9
2.0
2.0
3.1
2.6
2.1
1.7
3.0
1.3
2.2
2.2
2.3
2.7
.1
19.1
25.9
23.2
26.5
28.2
12.1
11.8
11.5
11.0
13.8
16.6
15.6
10.6
10.6
7.6
9.5
9.7
9.4
9.5
57.0
47.0
43.8
32.4
30.4
33.6
35.8
35.8
35.6
35.8
*Data for the years 1939-1960 are from O'Neill (1973); data for the remaining years are from U.S. Department of
Health, Education and Welfare, National Center for Education Statistics, FinancialStatistics of Higher Education.
locallegislaturesfor support.This is clearlynot the case for
theirprivatecounterparts.Moreover,the proportionof revenues derivedfromeach of these majorsources shows remarkablylittlevariationwithineach set of institutionsover the
20-yearperiod.Onlyrevenues fromthe federalgovernment
show any substantialchange, reflectingwaxingand waning
supportfor academic research.Thisoverallstabilitysuggests
stronglyinstitutionalizedpatternsof dependency relations.Private donorsand alumnitarget privatecolleges and universities
as appropriatebeneficiariesof theirlargesse, while government agencies favorthose inthe publicsector as appropriate
recipients.
Consequently,publicinstitutionstypicallyfindit more difficult
to obtainsignificantfinancialsupportfrom privatefunding
agencies and alumnithando privateinstitutions.The latter,in
state and
turn,often findgovernmentalbodies, particularly
locallegislatures,extremelyreluctantto offer financialsupport.
A small but noteworthyindicationof this is the move by many
privateinstitutionsto relabelthemselves as "independents."
Thisaccompanies efforts by these institutionsto attractpublic
sources of funding(Brenemanand Finn,1982).
Ithas been arguedthat it is onlywhen dependency relations
are not institutionalized
that the degree of dependency will
predictadministrativedifferentiation.Thus, it is hypothesized:
Hypothesis 1: Dependence on publicsources of supportwillstrongly
predictthe numberof administrativeoffices that managepublicfundingrelationsamong privateinstitutions.
Hypothesis la: Dependence on publicsources of supportwillnot
predictthe numberof public-funding
offices among publicinstitutions.
Hypothesis 2: Dependence on privatesources of supportwillstrongly predictthe numberof administrativeoffices that manageprivatefundingrelationsamong publicinstitutions.
Hypothesis 2a: Dependence on privatesources of supportwill not
offices among private
predictthe numberof private-funding
institutions.
Ifdifferentiationis linkedsimplyto the managementof particulartypes of dependency relations,there should be no interaction effects of controland dependency as predictorsof dif4/ASQ, March 1985
Sources of AdministrativeStructure
ferentiation.Ifadministrativedifferentiationis affected by the
of the relationship,dependence on particuinstitutionalization
larsources of supportshould show an interactiverelationship
with controlin predictingdifferentiation.
METHOD
Data
Dataused in the analyses are fromthe HigherEducation
GeneralInformationSurvey(NationalCenterfor Education
Statistics, 1980) and other secondarydata sources for 19751976. The survey is conductedannuallyamong two- and
four-yearinstitutionsof highereducationinthe UnitedStates
and gathers a varietyof informationon these institutions,
includingenrollment,sources of revenues, and expenditure
patterns.Inaddition,the surveyasks respondentsto list key
administrativeofficers andtheirpositions; each positionis
then assigned a common job code, based on the descriptionof
the associated administrativeresponsibilities.These data have
two advantages. First,they are the only publiclyavailabledata
on college and universityadministrationat the nationallevel.
Second, because common job codes are assigned, there is no
problemwith functionallyequivalentoffices havingdifferent
jobtitles.
Sample
The sample was drawnfromthe set of highereducation
institutionsclassifiedby the CarnegieCommissionas either
or as "comprehensive."Thisincludedall
doctorate-granting
majorresearchinstitutions,universitiesgrantingat least 20
Ph.D.'syearly,and institutionsofferingprofessionalprograms
in additionto a basic liberalarts curriculum.These institutions
were stratifiedby publicand privatecontrol,and a random
sample was drawnwithineach stratum.The resultingsample
contained281 institutions,of which 167 were publicand 114
were private.
Measuring Administrative Differentiation
Fromthe list of coded administrativepositions,six were
selected as havingmajorresponsibilityfor managingrelations
with externalsources of financialsupport.These were used to
measure the dependent variables,the numberof offices with
primaryresponsibilityforthe managementof private-funding
sources and the numberresponsiblefor dealingwith publicfundingsources.
Administrativepositionswith responsibilityfor managingprivate fundingincludethose of chief development officer,director of alumnirelations,and directorof admissions. The job
descriptionof the duties of the chief development officer
includesobtainingfinancialsupportfromalumniand other
organizations,coordinatingvolunteerfund-raisingactivities,
and managinggeneralpublicrelationsactivities(Jones, 1977).
The duties of the directorof alumnirelationsare very similar,
though obviouslytargeted specificallyat alumnigroups.The
directorof admissions is responsiblefor the recruitment,
selection, and admissionof students. Since tuition'feesare
often a significantsource of private-sectorsupport,how the
recruitmentand selection processes are managedcan have an
importantimpacton the institution'ssupportbase.
5/ASQ, March 1985
Responsibilityfor dealingwith sources of publicfundingis
assigned to the directorof the informationoffice, the chief
planningofficer,and the directorof institutionalresearch.The
duties of the directorof the informationoffice center on
providinginformationaboutthe institutionto students, faculty,
and the generalpublic.Thisincludespreparingand reviewing
news releases and informationbulletins,and managingrelations with the news media.The directorof institutionalresearch carriesout researchon the institutionitself and disseminates this informationto appropriatesources (e.g., legislative or governmentalagencies). The responsibilitiesof the
chief planningofficerincludemonitoringand managingstate
and federalrelations,as well as long-rangeplanningto allocate
institutionalresources. Itshouldbe noted that the office of
each serves a varietyof functions,not alldirectlyinvolving
fundingrelations.
Table2 shows the percentageof publicand privateinstitutions
that reporthavingeach of these administrativepositions.The
percentageof privateinstitutionsreportingthe presence of a
Table 2
Percentage of Public and Private Institutions Reporting Selected
Administrative Positions for Funding Relations
Administrative positions
Public
Private
Chief Development Officer
Director of Admissions
Director of Alumni Relations
Chief Planning Officer
Director of Information
Director of Institutional Research
47.9
66.5
50.3
43.7
43.1
45.5
83.3
86.8
61.4
28.9
25.4
24.6
chief developmentofficerand a directorof admissions is
significantlyhigherthanthat of publicinstitutions.A noticeably
higherproportionof privateinstitutionsalso reportshavinga
directorof alumnirelations,althoughthe differencedoes not
reachsignificance.Publicinstitutions,on the other hand,report
significantlymore often havingthe positionsthat manage
relationsassociated with sources of publicfunding.
These resultsaccordwith expectations derivedfrom both the
institutionalization
and resourcedependence perspectives. In
the firstperspective,structuralpatternsare expected to be
associated with institutionalized
dependencies. Thus,the
offices among private
greaterfrequencyof private-funding
offices among publicinstituinstitutionsand public-funding
tions is predicted.Likewise,froma resourcedependence
offices would be expected to
perspective, private-funding
occur more frequentlyamong privateinstitutionsthat presumablydepend more heavilyon privatesources of support,while
offices.
the reverse would be expected for public-funding
However,the resourcedependence perspectivesuggests that
increasingdependence should predictan increase insuch
offices, regardlessof the type of organizationalcontrol.If,in
contrast,the creationof these offices is affected by the
of dependency relations,increaseddepeninstitutionalization
dence on publicfundingshould be a better predictorof the
offices in privateinstituoccurrenceof the three public-funding
tions than in publicones. Similarly,the creationof the three
6/ASQ, March 1985
Sources of AdministrativeStructure
offices shouldbe more stronglypredictedby
private-funding
dependence on privatefunds in publicuniversities.
Measuring Dependence
Dependence was measured by the proportionof the institution's total revenues derivedfromfourprimarysources of
governmentgrantsand
support:governmentappropriations,
contracts,gifts and grantsfromprivatesources, and "selfgenerated"funds -tuition, fees, and endowment income.
The firstcategory,governmentappropriations,is a source of
viewed as a mainstayof publicinstitusupporttraditionally
tions, while the last two categories are supportthat is conventionallyassociated with privatestatus. Althoughpublicand
privateinstitutionshave both receivedgovernmentgrantsand
contractsfor researchsince WorldWar11(Nelson, 1978),
governmentgrantswill be treated here as technicallya public
source of support.
ControlVariables
Inadditionto the measures of dependency, the analyses
includedtwo other independentvariablesas controls:size and
a measure of researchorientation.Size, measured here by
total student enrollment,has often been linkedto administrative differentiation(Terrienand Mills,1955; Blau,1970; Hsu,
Marsh,and Mannari,1983). Thismeasure was logged to
correctfor skewedness. The complexityof administrationin
institutionswith a strong researchorientation,measured by
the percentage of total expendituresallocatedfor research
activities,mightalso be expected to produceadministrative
differentiation.Thus,these variableswere includedto ensure
that the observed relationshipswere not merely reflectionsof
differences in size or between researchand nonresearch
institutions.
ANALYSIS
To test the effects of resourcedependence on administrative
differentiation,two dependent variableswere created, one
offices reportedand
based on the numberof public-funding
offices reported.
one based on the numberof private-funding
Table3 shows the means and standarddeviationsof the
variablesby type of institution.
Table 3
Means and Standard Deviations of Resource-Dependence
Variable
Size (in 1 000's)
% Expendituresforresearch
% Revenues
Government appropriations
Government grants
Private gifts
Self-generated
Number of public-funding offices
Number of private-fundingoffices
Variables by Type of Institution
Public (N= 167)
S.D.
Mean
Private (N= 114)
S.D.
Mean
16.511
9.133
11.214
8.691
8.722
8.344
6.556
8.822
48.044
11.708
2.462
15.496
1.396
1.718
13.659
7.506
2.242
8.322
.876
.938
2.345
13.642
7.837
51.034
.876
2.389
4.920
10.217
5.275
19.289
.877
.760
Ineach case, the dependent variablecan assume one of only
fourpossible values (0-3). Because multivariatelinearanaly7/ASQ, March 1985
ses are based on assumptions requringan intervallevel of
measurementforthe dependent variable(Maddala,1983), the
use of ordinary-least-squares
techniques is inappropriate.2
Instead, an orderedprobitmodel was employed. Thisis an
extension of a dichotomousprobitmodel that is applicableto
analyses involvingordinaldependent variables(McKelveyand
Zavoina,1975; Winshipand Mare,1984). Withthis analytic
procedure,a distinctionis drawnbetween an underlying
theoreticaldependent variable,which has an intervalscale of
measurement,and the observed dependent variable,which is
ordinal.Inthe present case, the latenttheoreticalvariablemay
be thoughtof as the amountof pressurefor administrative
differentiation;the observed variableis the presence of zero,
one, two, or three offices.
Moreformally,if Y*is the underlyingtheoreticalvariable,and Y
is the observed variable,then it is assumed:
Y* = OX+ u,
(1)
where u -
N(O,&2). To set the scale of measurement,(2_1
(see Joreskog and Sorbom,1981).
Then:
OifY*<ao
1 if ao < Y* < a
Y_
a
(2)
2ifota <Y* < a2
3 if U2< Y*
The alphasinthis equationrepresentthe threshholdpoints in
the distributionof Y*at which the observed Ytakes on a
differentvalue.The relationshipbetween the latentvariable,
Y*,and the observed Yis depicted in Figure1.
f(Y*)
L<
'~~~~~~~~~Y
1
2=
Y= 0
Y= 3
Figure 1: Relationship between observed ordinalvariable (Y)and latent continuous variable (Y*).
2
Linearregressionmodelsarebased on the
assumptionthatthe distribution
of data
pointsaroundthe regressionlineyieldsa
set of errortermswitha meanof zeroand
a constantvariance.Whendealingwithan
ordinaldependentvariable,these assumptionsaregenerallynotvalid(McKelveyand
Zavoina,1975).
Combining the first two equations implies that:
P(Y= OIX)=
P(Y*< OLX)=
P(PX+ u< OL)=
P(u < O0 - AX)=
8/ASQMar1
8/ASQ, March 1985
(3)
Sources of Administrative Structure
where F(o0 - PX)signifies a standard normal cumulative
density function(see Figure1).Thiscan readilybe extended to
the othervalues of Y.Because orderedprobitmodels take into
account ceilingand floorrestrictionson probabilities,they are
highlypreferableto linearmodels when the observed Yis
markedlyskewed (Winshipand Mare,1984).
Maximumlikelihoodmethods (Berndtet al., 1974; McKelvey
and Zavoina,1975; Chow, 1983)were used to obtainestimates of the parametersof the model. Separateanalyses
were carriedout for each dependent variable,using the full
sample of institutions.These analyses were then repeated,
usingthe subsamples of publicand privateinstitutions.
RESULTS
Table4 presents the results of the orderedprobitanalysisof
the numberof public-funding
offices, using the fullsample of
institutions.The beta coefficients representthe slope coefficients inthe latentregression,or the incrementin Y*brought
about by a unitchange inthe independentvariable.The first
alphacoefficient is automaticallyset to zero for normalization.
The othertwo alphas representthe remainingcutpointsin the
distribution of Y*. The Xis a test of the overall significance of
the model, based on a comparisonof the presented model
with one inwhich the betas are constrainedto be zero.
Table 4
Ordered Probit Analysis of Public-Funding Offices in Total Sample
Independent variables
B
S.E.
t-value
-1.16
3.450
-.50
1 .70
1.12
1.17
-1 .95g
-1 .77
12.00a
16.170
Constant
Size
% Research expenditures
Control (public)
% Government appropriations
% Government grants
Control x appropriations
Control x grants
-.231
-.006
.653
.038
.016
-.032
-.028
.200
.774-5
.012
.363
.033
.013
.016
.016
atn
1.137
2.183
.095
.135
OL2
Log L
X
.267-4
-333.07
44.44000
*1p<.10; *61p<.05; *00tp<.01.
Thereare clearinteractiveeffects of controland dependence
on governmentgrantsand appropriations.
Apartfromsize, the
strongest and only significantpredictorsof the numberof
offices are publiccontroland the interaction
public-funding
terms.
InTable5 the analysisis repeated, using the numberof
offices as the dependent variable.
private-funding
Dependence on privategifts and self-generatedsources of
fundingare not, by themselves, strong predictorsof the number of offices. The interactionterm for controland dependence
on privategifts does emerge as significant,however, along
with the terms for controland for size.
9/ASQ, March 1985
Table5
OrderedProbitAnalysis of Private-FundingOffices in Total Sample
B
Independentvariables
Constant
Size
% Researchexpenditures
Control(public)
% Privategifts
% Self-generated
Controlx gifts
Controlx self-generated
1.678
.171-4
.008
-1.210
-.003
.013
.061
.013
1.996
2.050
at1
OL2
S.E.
t-value
.499
.763-5
.009
.492
.022
.013
.034
.013
3.36g
2.251
.85
-2.460
-.14
.57
1.87
.98
.113
.133
8.790
15.410g
-333.65
53.180
LogL
X
*p<. 10; *6p<.05; *00tp<.01.
Bothanalyses suggest, then, that the effect of dependency on
administrativedifferentiationvariesfor publicand privateinstitutions.To examine this more closely, the analyses were
repeated, usingthe subsamples of each type of institution.
offices by public
The orderedprobitanalyses for public-funding
and privateinstitutionsare shown inTable6.
Table6
OrderedProbitAnalysis of Public-FundingOffices in Publicand Private Institutions
B
Independentvariables
Constant
Size
% Researchexpenditures
% Governmentappropriations
% Governmentgrants
Ot
1
aL2
-.005
.005
-.009
.343
.895-5
.013
.005
.015
1.156
2.207
.122
.167
.187-4
-209.49
6.30
Log L
X
*rx.10
.521
Public
t-value
S.E.
16" *r
nn
101
*
1<
1.52
2.080
-.34
.97
-.64
9.470
13.23a
B
Private
S.E.
-.469
.680-4
-.020
.037
.018
.240
.241-4
.030
.018
.023
-1.960
2.83g
-.68
1.001
2.223
.253
.177
6.780
11.38a
t-value
1.9400
.78
-120.46
18.62a
.
Among publicinstitutions,onlysize emerges as a strong
predictor.Increasingdependence on publicsources of support
has no substantialeffect on the numberof public-funding
offices. Overall,the low value of Xindicatesthat these variables add littleto the predictionof administrativedifferentiation. Among privateinstitutions,on the other hand,dependence on governmentalappropriationsis significantlyrelated
to differentiation.The test statisticforthe overallmodel is also
significantinthis case. These results,then, are consistent with
the firsthypothesisand its corollary.
offices by
Table7 presents the analyses for private-funding
type of institution.As hypothesized,increasingdependence on
privategifts, is a much
privatesources of support,particularly
strongerpredictorof differentiationin publicthan in private
1O/ASQ, March 1985
Sources of Administrative Structure
Table 7
Ordered Probit Analysis of Private-Funding Offices in Public and Private Institutions
Independent variables
Constant
Size
% Research expenditures
% Private gifts
% Self-generated
.475
.10o-4
.016
.059
.018
.957
2.060
Ot
1
aL2
1
.250
.834-5
.012
.030
.012
.112
.153
-212.87
13.92a
Log L
X
*ne.
Public
S.E.
B
bn
t-value
B
1.567
Private
S.E.
1.90g
1.20
1.33
1 .92g
1.51
-.021
.006
.001
.695
.297-4
.020
.025
.008
7.810
13.470
1.166
2.179
.292
.307
.686-4
t-value
2.540
2.3100
-1.06
.24
.06
3.99a
7.090
-117.20
9.00
no 1011p.l
grind.05
institutions.Inthe lattercase, the measures of dependence on
each source have virtuallyno relationshipto the numberof
offices. Thus,these results providesupportfor
private-funding
the second set of hypotheses as well.
As expected, dependence on publicor privatesources of
of administrativeoffices only
fundingpredictsthe proliferation
when the dependencies are not alignedwith traditionalpatterns. Inotherwords, dependence on a given exchange relationshipmay or may not lead to the creationof offices and
positionsto managethose relationships.Since institutionalized
dependencies are often accompaniedby the institutionalization of structuralcomponents, organizationscharacterizedby
such relationsadoptthose components ceremonially,independent of actuallevels of support.Thus, increasingdependence
does not necessarilyproduceadministrativedifferentiation.It
is onlywhen dependency relationsare not institutionalizedthat
increasingdependence is stronglyassociated with the development of separate administrativeoffices to manage them.
DISCUSSION
perspective
These analyses suggest that an institutionalization
defines conditionsunderwhich hypotheses generated by a
resourcedependence perspectivewill hold.Thereare, however, alternativeexplanationsand modificationsof this argument
to be considered.First,it could be arguedthat the results of
the analysissimplyreflectthe limitedvarianceof the independent variableswhen dependencies follow traditionalpatterns.
Examinationof the means and standarddeviationsof the
variables,shown inTable3, substantiallyweakens the plausibilityof this argument.None of the measures of traditional
among public
dependencies (e.g., governmentappropriations
institutions,gifts and donationsamong privateinstitutions)has
a markedlyrestrictedvariance.The rangeforthe measure of
among publicindependence on governmentappropriations
stitutionsis from 10 to 84 percentof total revenues; the range
for dependence on privategifts is from 1 to 33 percentamong
privateinstitutions.Thus,this alternativeis not supportedby
the data.
Itmightalso be arguedthat the relationshipbetween dependence and administrativedifferentiationshould be reciprocal.
11/ASQ, March 1985
Whilea plausiblecase may be made forthis, the central
concernof the present analysisis to demonstratethe link
between two generaltheoreticalperspectives that focus primarilyon the sources of administrativedifferentiationrather
thanon the consequences. The latter,however, remainsan
interestingproblemfor furtherresearch.
CONCLUSIONS
Contemporaryresearchon organizationshas produceda variety of theoreticalperspectives, each pointingto differentexplanatoryfactors.Althoughthe complementarityof different
perspectives is sometimes acknowledged(Aldrichand Pfeffer,
1976; Ulrichand Barney,1984), empiricalresearchtypically
draws on a single theoreticalapproachin explainingparticular
cases of organizationalbehaviorand structure.Positiveoutcomes are interpretedas evidence of the validityof that
perspective.
Itis clear,however, that most of the perspectives that currently guide researchare not trulycompetitive,such that support
for one underminesanother.Instead,they are more likely
simplyto be applicableunderdifferentconditions,as was the
case inthis study. By combiningresourcedependence and
institutionalization
perspectives, a much fullerexplanationof
the process of administrativedifferentiationwas providedthan
could have been providedby either perspective independently.
phenomenaare much too complex to be deOrganizational
scribedadequatelyby any single theoreticalapproach.Current
researchon organizationscould benefit greatlyif researchers
were to pay closer attentionto specifyingthe points of intersection of differenttheoreticalperspectives and to combiningthese perspectives to providemore complete explanations
of the behaviorsthey study.
REFERENCES
Aldrich, Howard, and Jeffrey
Pfeffer
1976 "Environments of organizations." InA. Inkeles (ed.),
Annual Review of Sociology, 2:
79-1 05. Palo Alto, CA: Annual
Reviews.
Anderson, Theodore, and Seymour
Warkov
1961 "Organizationalsize and functional complexity: A study of
administration in hospitals."
American Sociological Review,
26: 23-38.
Bendix, Reinhard
1956 Work and Authority in Industry
New York:Wiley.
Berndt, E. K., B. H. Hall, R. E. Hall,
and J. A. Hausman
1974 "Estimation and inference in
non-linear structural models."
Annals of Economic and Social
Measurement, 3: 653-666.
Blau, Peter
1970 "A formal theory of differentiation in organizations." American Sociological Review, 35:
201-218.
Breneman, David, and Chester
Finn, Jr.
1982 "An uncertain future." In D.
Breneman and C. Finn (eds.),
Public Policy and Private Higher Education: 1-62. Washington, DC: Brookings.
Brubacher, John, and Willis Rudy
1965 Higher Education in Transition.
New York: Harper& Row.
Chapman, Stephen
1979 "No-fault capitalism." New Republic, 181 (November 17):
11-14.
Chow, Gregory
1983 Econometrics. New York:
McGraw-Hill.
12/ASQ, March 1985
DiMaggio, Paul, and Walter Powell
1983 "The iron cage revisited: Institutional isomorphism and collective rationalityin organizational fields." American
Sociological Review, 48: 147160.
Freeman, John
1973 "Environment, technology and
the administrative intensity of
manufacturing organizations."
American Sociological Review,
38: 750-763.
Friedman, Milton
1979 "Are jobs the issue?" Newsweek, 94 (September 10): 66.
Hsu, Cheng-Kuang, Robert Marsh,
and Hiroshi Mannari
1983 "An examination of the determinants of organizational structure." American Journal of
Sociology, 88: 975-996.
Sources of AdministrativeStructure
Jones, Dennis
1977 A Manual for Budgeting and
Accounting for Manpower Resources in Postsecondary
Education. Washington, DC:
National Center for Education
Statistics.
National Center for Education
Statistics
1980 LongitudinalFile of Financial,
Enrollment and Faculty Statistics for Institutions of Higher
Education: 1974-75 to 197778. Washington, DC: NCES.
Joreskog, Karl, and Dag Sorbom
1981 Analysis of LinearStructural
Relationships by Maximum
Likelihoodand Least Squares
Methods. Uppsala: University
of Uppsala.
Nelson, Susan
1978 "Financialtrends and issues."
In D. Breneman and C. Finn
(eds.), Public Policy and Private
Higher Education: 69-142.
Washington, DC: Brookings.
Levine, Sol, and Paul White
1961 "Exchange as a conceptual
framework for the study of interorganizational relationships." Administrative Science Quarterly, 5: 583-601.
O'Neill, June
1973 Sources of Funds to Colleges
and Universities. Washington,
DC: Carnegie Foundation.
Maddala, G. S.
1983 Limited Dependent and Qualitative Variables in Econometrics. Cambridge: Cambridge University Press.
McKelvey, Richard, and William
Zavoina
1975 "A statistical model for the
analysis of ordinal level dependent variables." Journal of
Mathematical Sociology, 4:
103-1 20.
Meyer, John, and Brian Rowan
1977 "Institutionalized organizations: Formal structure as
myth and ceremony." American Journal of Sociology, 83:
340-363.
Pfeffer, Jeffrey, and Gerald
Salancik
1978 The External Control of Organizations. New York: Harper&
Row.
Pondy, Louis
1969 "The effects of size, complexity and ownership on administrative intensity." Administrative Science Quarterly, 14:
47-61.
Rowan, Brian
1982 "Organizationalstructure and
the institutional environment:
The case of public schools."
Administrative Science
Quarterly, 27: 259-279.
Rudolph, Fredrick
1962 The American College and University. New York: Knopf.
Meyer, Marshall
1979 "Organizationalstructure as
signaling." Pacific Sociological
Review, 22: 481-500.
Samuelson, Paul
1979 "Judging corporate handouts."
Newsweek, 94 (August 13):
58-61.
Meyer, Marshall, and M. Craig
Brown
1977 "The process of bureaucratization." American Journal of
Sociology, 83: 364-385.
Scott, Richard, and John Meyer
1983 Organizational Environments.
Beverly Hills, CA: Sage.
Simon, Herbert
1956 Administrative Behavior. New
York: Macmillan.
Spence, Michael
1973 "Job market signaling."
QuarterlyJournal of Economics, 83: 355-374.
1974 Market Signaling. Cambridge,
MA: MITPress.
13/ASQ, March 1985
Taylor, Fredrick
1911 Principles of Scientific Management. New York: Harper&
Row.
Terrien, Fredric, and Donald Mills
1955 "The effects of changing size
on the internal structure of
organizations." American
Sociological Review, 20: 1113.
Thompson, James
1967 Organizations in Action. New
York: McGraw-Hill.
Tolbert, Pamela, and Lynne Zucker
1983 "Institutional sources of
change in organizational structure: The diffusion of civil service reform, 1880-1930."
Administrative Science
Quarterly, 23: 22-39.
Ulrich, David, and Jay Barney
1984 "Perspectives in organizations:
Resource dependence, efficiency and population."
Academy of Management Review, 9: 471-481.
Weber, Max
1946 Essays in Sociology. New
York: Oxford University Press.
Winship, Christopher, and Robert
Mare
1984 "Regression models with
ordinalvariables." American
Sociological Review, 49: 512525.
Zucker, Lynne
1977 "The role of institutionalization
in cultural persistence." American Sociological Review, 42:
725-743.
1983 "Organizations as institutions." In S. Bacharach (ed.),
Research in the Sociology of
Organizations, 2: 1-47. Greenwich, CT: JAI Press.