INFORMATION TO USERS

This reproduction was m ade from a copy of a m anuscript sent to u s for publication
and microfilming. While th e m ost advanced technology h as been used to pho­
tograph and reproduce this m anuscript, the quality of the reproduction Is heavily
dependent upon the quality of the material submitted. Pages In any m anuscript
may have Indistinct print. In all cases the best available copy has been filmed.
The following explanation of techniques is provided to help clarify notations which
may appear on this reproduction.
1. M anuscripts may not always be complete. When It is not possible to obtain
missing pages, a note appears to indicate this.
2. When copyrighted materials are removed from the manuscript, a note ap­
pears to indicate this.
3. Oversize materials (maps, drawings, and charts) are photographed by sec­
tioning the original, beginning at the upper left hand com er and continu­
ing from left to right in equal sections with small overlaps. Each oversize
page is also filmed as one exposure and is available, for an additional
charge, as a standard 35mm slide or in black and white paper format.*
4. Most photographs reproduce acceptably on positive microfilm or micro­
fiche b u t lack clarity on xerographic copies made from the microfilm. For
an additional charge, all photographs are available in black an d white
standard 35mm slide format.*
♦For more information about black and white slides or enlarged paper reproductions,
please contact the Dissertations Customer Services Department

TU T-\/f-T
D issertation
1 V 1 1 Inform ation S erv ice
University Microfilms International
A Bell & Howell Information C om pany
300 N. Z e e b R oad, Ann Arbor, M ichigan 48106

8700476

Harrison, R obert E.

A STUDY O F THE INFLUENCE O F SIZE AND ECONOM Y O F SCA LE IN
MICHIGAN PUBLIC COMMUNITY COLLEGES

M ic h ig a n S ta te U n iv e rs ity

University
Microfilms
International

300 N. Zeeb Road, Ann Arbor, Ml 48106

Ph.D .

1986

A STUDY OF THE IN F L U E N C E OF SIZE
AND ECONO MY OF SCALE IN MI CH I GA N
PUBLIC C O M M UN IT Y CO LLEGES

By

Ro be rt E. Har ri so n

A DI S S E R T A T I O N

Subm itt ed to
M i ch ig an State Univ er sit y
in partial fulfill men t of the req ui re me nt s
for the degree of

DO C TO R OF PH IL O SO PH Y

De p ar tme nt of Ed uc at io n al A dm i n i s t r a t i o n

1986

ABSTRA CT
A STUDY OF THE INFLUE NC E OF SIZE
AND ECONOMY OF SCALE IN MI CH IG A N
PUBLIC COMMU NIT Y COLLEGES
By
Robert E.

This

Harrison

study inv es ti ga te s the rel at io ns hi p of structural

c h a r a c t er i st ic s in or ga ni zat io ns of varying size.

A cen­

tral concern is the re la t io ns hi p between or gan iz ati on
and economy of scale.
colleges

The twenty- ni ne

in the State of Mic hi gan

public community

comprise the pop ulation

of the study which is post hoc in design.
obtained

size

Using data

from the State of Mic higan Acti vit ie s C l a s s i f i ­

ca tion Structure,
an al ysi s is used

co rre la ti o n and m u l t i v ar ia te s tat is tic al
to examine

in relation to enrollment,

the size of personnel com ponents
resource allo ca tio n in relation

to size and efficiency measures,

and resource all oc at io n in

relation to availa ble resources.

In addition

cur ri cul ar dif fe re nt i at io n on stu de nt -f ac ul ty
average class

size are in vestigated.

the effects of
ratios and

The study indicates

that economy of scale is present in redu cti on s of e x p e n d i ­
tures

per f u l l - ye ar -e qu at ed student.

ex p end itu re s
increases

per f ul l -y ear -eq ua ted

in college size.

Total operating

student de crease with

A d m in is tr at iv e expend itu re s

Robert E. Harrison

per student decline with incre ase s in college size,
even though a d mi ni s tr at iv e e x pe ndi tu res are related more
to avail abl e resou rce s than to college size.
scale

is not s i gn if i ca nt ly

influenced

by increases in

d i f f e r e n ti at io n in personnel comp one nt s
and a dm i n i s t r a t i v e
dif f er en ti at i on

professionals.

ulation

Increases

size.

These

can be made.

but due to the pop­

no inferences regarding causal

relation sh ips

In the context of the given size,

going r e l a t i on sh ip s were
that size

in curricular

findings accu ra tel y

the popu lat io n of the study,
size,

including faculty

have no adverse effects on stud ent -f acu ity

ratios or average class
de scribe

Economy of

is a causal

observed.

variable.

the f o r e ­

It cannot be concluded

Scott,

Jeffrey,

and Judith

ACKNOWLEDGEMENTS

I wou ld
regard

like

to e x p r e s s

to the f o l l o w i n g

my a p p r e c i a t i o n and high

p er son s who w e r e

important influ­

ences in my study:
To

Dr.

Jo h n M.

Eat on

to co m m i t to co n t i n u e d
for c o n t i n u e d
To Dr.

su pp ort

his p at ie nc e

study,

in w a i t i n g

for me

and to Dr. W i l l i a m M. A n d e r s o n

d ur in g my study.

R i c h a r d L.

d i r e c t i o n and

for

F e a t h e r s t o n e who

confidence

in my

e a r li es t

pr ovi ded

me

both

a s s o c i a t i o n wi t h

the

university.
To Dr.
p l a nn in g

and

To Dr.

Lou Anna
s up por t
John H.

Kirasey Simo n
throughout
Su eh r who

for a s s i s t a n c e

in p r o g r a m

the process.

first

i ns pi re d

my i n te re st

in

organizations.
To Dr.
iz a t i o n a l

A. Mo o re

P hi li p M. M a r c u s

ch a l l e n g e d

my grasp of o r g a n ­

for a lw ay s a s k i n g

the right

at the right times.

To P a t r i c i a P a c h e s n y
study

who

theory.

To Dr.
questions

Samuel

for her a s s i s t a n c e

in both my

and my work.
To Sc ot t and Jeff who

stood with w i s d o m

beyond

t hou gh

o r p ha n ed

their years

the

by study u n d e r ­

nee ds of their

father.
And
for

to Judy who

this - a better

sacrificed
day.
iii

more than an yo ne

in w ai t i n g

TAB LE OF CONTENTS

Page
LIST OF T A B L E S ...........................................

vi

LIST OF F I G U R E S .......................................... viii
C HA PTE R
I.

II.

C O N C EP TU AL F R A M E WO RK AND STATE MEN T
OF P R O B L E M ....................................
I ntr od u ct io n . . . . . . . . . . . . . . .
De f in it io n of Econ om y of Scale ............
Ap p ro ac he s to the Study of Economy
of S c a l e .................................
Sub jec t of I nv es ti gat io n ...................
Two -Y ear College Origi ns .......... « . . .
Two -Y ea r C ol leg e G r o w t h ......................
Role of the Tw o-Y ea r C o l l e g e ..............
P urpose of the I nv e st ig at io n . . . . . . .
C on tr ib ut io n of the I n v e s t i g a t i o n ..........
O v er vi ew of the S t u d y ........................

3
6
7
8
11
13
15
16

REVIE W OF THE L I T E R A T U R E ....................

18

I n t r o d u c t i o n ...............................
O r ga ni za ti on a l Size and Its Influence. . .
O r ga ni za ti on a l Size and Effi ci enc y . . . .
O r g a n i za ti on al Size and Economy of
Scale.
. . .
S u m m a r y ................
III.

M E T H O D O L O G Y ....................................
The P o p u l a t i o n ...............................
Data S o u r c e ....................................
Variab le s and Indic at ors .
..............
Hypot he ses and Ratio nal e . . . . . . . . .
Sta ti sti ca l Ana l ys is . . . . .
*

IV.

1

EM PI RI CA L FINDINGS

...........................

Hy p ot he si s I .................................
Hyp ot hes is I I ................
iv

I
2

18
18
30
37
AO
A3
A3
A3
AA
A7
52
57
57
62

V.

Hy p oth esi s I I I .....................
69
Hyp o th es is IV. ................... . . . . .
. . . . . . . .
H y p ot he si s V ..............
H y p ot he si s V I ................
.............................
Hy p ot he si s VII
S u m m a r y .................................. ....

76
82
86
90
92

SUMMARY,

93

CO NC LU SI O NS

AND R E C OM ME N DA TI ON S

.

93
O v e r v i e w ............................
Hyp o th es is I . . . . . . . . . . . . . . .
94
Hy p ot hes is II & I I I .............. ..... ......
98
Hy p ot hes is IV & V .................... 102
Hy p ot hes is V I ......................... 104
Hyp o th es is VII . . . . .
...................
107
Major O u t c o m e s ........................ ....
108
R e co m m e n d a t i o n s for Pra cti ci ng A d ­
m i ni s t r a t o r s .................................. Ill
Re c o m m e n d a t i o n s for State Decisi on
M a k e r s ..............
114
R e c o m m e n d a t i o n s for Furt her Study. . . . .
117
APP E ND IX
A.

MI CHI GA N PUBLIC CO M MU NI TY COLLEGES,
E N R O L LM EN T AND P ER SON NE L .................

B.

OP E R A T I O N A L DE FI NI TI ON S AND BASIC
STATI ST ICS OF V A R I A B L E S ........... 124

C.

DATA T R A N S F O R M A T I O N S ............... 131

B I B L I O G R A P H Y ................

121

139

v

LIST OF TABLES

TABLE
1.

2.

3.

4.

5.

6.

7.

8.

9.

Page
Fifty Years of Growth in Two -Y ear
Coll eg e Credit En ro l lm en ts 1933-34
1 9 8 3 - 8 4 ...............................

to

Size V ar iab le s of Mi c h i g a n Public
Community Co lle ge s in Fiscal Year 1984

10

.

C o r r el at io n Co ef fi ci e nt s Re lating the
Div isi on of Labor by Colle ge Size in
E n r o l l m e n t ...........................

.

14

58

T Tests for Si gn if ic an t D i f fe re nc e of
Co r r e l a t i o n Co e ff ic ie nt s in De pen den t
Sample of Groups Compare d Against FullYe a r- Eq ua te d S t u d e n t s ................. .. .

.

60

C or r el at io n C oe f fi ci en ts Relating the
A d mi ni st ra ti ve E xp en di tu re s by Component,
Colleg e Size in Enr ol lm en t and Employees,
and A d mi ni st ra ti ve E xp en di tu re s per S t u ­
dent ...........................................

64

C or r e l a t i o n Co e f f i c i e n t s of A d m i n i s t r a ­
tive C om pon en t Ex pe n di tu re s per Student
(FYES), College Size in Enrollment*
and total Op era ti ng Ex pen di tur e per
S t u d e n t ...........

66

Stepwis e M u l ti pl e Re g re ss io n of Operating
Exp en di tu re s per Student on A d m i n i s t r a ­
tive Co m po ne nt s and Enr o ll me nt (FYES), . .

68

Co rre l at io n Coe ffi ci e nt s Re lat in g Gross
Ex pen dit ur e and Colle ge Size Variables
to the Op erating Ex p en di t ur e per
S t u d e n t .........................................

70

Co rre l at io n Co ef fi ci e nt s Re la tin g E x ­
pen ditures per Student and Coll ege Size
Variables to the Ope ra ti ng Expend itu re
per Student. ; * ..............................

73

vi

10.

11.

12.

13.

14.

15.

16.

17.

18.

19.

Ste pwise Mul ti ple Reg res sio n of
Operati ng E xp en di tu re s per Student on
Com bined Ad mi n i s t r a t i v e Compon en t Ex ­
penditures, Ins tr u ct io na l Expen di tur es
and Enr oll men t (FYES).
..............

74

Mean and Standard De vi ati on of Operat in g
E xpe ndi tu re per Fu ll -Y ea r - E q u a t e d S t u ­
dent in Mic higan Public Community C o l ­
l e g e s ...........................................

76

Co rr ela ti on C oe ffi ci ent s Re lating the
Number of Courses, Sections, Full -Ye arEquated Students, and the StudentFaculty R a t i o ..................................

77

Means and Standard De vi at io ns for Number
of Courses, Sections, Ful l-Y ea r- E qu at ed
Students, and the S t u d e n t -F ac ul ty Ratio. .

79

Means and Standard Devi ati on s for Number
of Courses, Sections, Fu l l- Ye ar -E qu a te d
Students, and the Stu d en t- Fa cu lt y Ratio
with Outlie rs R e m o v e d ........................

81

Co rre la tio n C oe ff ic ie nt s Rel ating the
Number of Courses, Sections, U n d u p l i ­
cated Headcount, Fu l l- Ye ar -E qu at ed S t u ­
dents, and the Avera ge Class S i z e ........
Stepwise Mul ti p le Reg res si on of
Average Class Size on the Number of
Courses, Sections, Un dup lic at ed H e a d ­
count, and Fu ll -Y ea r- Eq ua te d Students.

. .

Co rre lat io n C oe ff ici en ts Relating
Revenue, College En ro llm en t and A d m i n i ­
strative Salary and B e n e f i t s ............
Ste pwise Mul tip le Re gr es si on of A d m i n i ­
strative Salaries on Revenue and
E n r o l l m e n t ..............
Cor rel at ion C oe f fi ci en ts Re lat ing
Physical Plant Size, Number of Users,
and Physical Plant E x p e n d i t u r e s ...........

vi i

83

.

85

88

89

91

LIST OF F IG U RE S

Fi gu re
1.

Page
R e p r e s e n t a t i o n of S i g n i f i c a n t R e l a t i o n ­
ship B e t w e e n E m p l o y e e G r o u p s ...............

viii

61

C H AP TE R I

CO N CE PT UA L FRA M EW OR K AND STA TE M EN T OF PROBL EM

Intr odu cti on

The fo undation for c o ns id er at io n of org an i za ti on s from
the stru ct ura li st

point of view was es ta bl is he d in W e b e r ’s

de f in it io n of bu reaucracy

(1947).

His

focus upon the formal

structure of or gan iz at io ns

str essed the impor tan ce

partmental

in a pyramidal h ie rar ch y of

units connected

au th ori ty reg u la te d by rules.

First in imp ort an ce in this

burea uc rat ic view of o rg an iza ti ons
dif f er en ti at io n
little

(Blau,

1970;

is the dimen sio n

Weber,

1946),

co ns id er ati on to or gan iz ati on

im pl ica tio ns

for differentiation.

of

but Weber

gave

size beyond its

Increases

in di f f e r ­

en ti ati on will co rrelate wit h increa ses in size;
the re lat io nsh ip of size to org anization
not

of d e ­

however,

e ff ici en cy is

so clear.
During

the rapid postwar

industrial

ers of the 1950's began to question
other org a ni za ti on al

variables.

expansion,

the effect

research­

of size upon

O r g a n i za ti on s were accused

of ov er - b u r e a u c r a t i z a t i o n with di sp ro po r t i o n a t e resources
al l oc at ed to a dm i n i s t r a t i v e comp one nt s
C o nt ra di ct or y
the 1950's.

fin dings were reported
Melma n

(1951)

(Parkinson,

1957).

in major res earch of

and Be ndi x (1956)

showed that

the a dm in is t r a t i v e component often de cr ea se d with increases
in org an i za ti on al

size.

Baker and Davis

1

(1954) and Terrien

2

and Mills

(1955),

however,

in the ad mi n i s t r a t i v e
creased

in size.

found d is pr op or t i o n a t e

increases

compon en t as the or ga ni za ti on

Thus

in­

the debate for econ om y of scale

began.

Def i ni ti on

of Economy of Scale

Ec o nom ist s ty pi ca ll y relate mea sur es of input to
meas ur es

of output

in order to determine

cost of pro duc tio n per unit produced
1954;

Silberston,

relates

1972;

Gold,

1981).

to pr o du ct io n cap ac ity while

e ff ici enc ie s

(Osborn,

1951;

in

Bain,

Scale in this concept
economy relates to

prod uc tio n efficiency.

Optim um economy

when m ax imu m produ ct ion

capacity is matc hed with ma x imu m

e ff ici enc y

(Bain,

of scale results

1954).

O r ga ni za ti o na l

researchers,

on the other hand,

less on cost mea sur es and more upon str uct ur al
of or gan i za ti on s

(Kimberly,

is fre qu en tl y mea su red
fe re nti ati on
herr,

1971;

1976).

in number of employees

1973).

c on si de ra ti on s

In this app roach

or co mp le xi ty of operations,
Blau,

depend

A d d i ti on al

and

the d i f ­

(Blau and S ch oen -

di me ns io ns

of scale

incl ude or ga ni za ti on a l

for ma l iz at io n c h a r a c t er i st ic s

as levels

and

of au th o ri ty

span of control.

economy of scale may be mea su red
against

by rel ating

such

In this context
scale

factors

some output meas ure con sis te nt wi t h the o r g a n i z a ­

tional mission.

O rg a n i z a t i o n a l

might also be purely th eor eti ca l
effects

scale

studies

of economy

of scale

i n ve s ti ga ti on s of the

of size on c h a r a c t e ri st ic s

of the orga niz ati on

3

itself

(Meyer,

1972).

App r oa ch es

to the Study of Econ om y of Scale

Studies of economy of scale tend to c onc ent ra te
c r o s s - s e ct io na l

ex ami n at io ns rather than

studies.

C r o s s -s ec t io ni al

a tte nti on

to changes

studies are co mpl ic ate d
time alone.

Further

by changes

for lack of

and longit ud ina l

in variab les

types

such as Blau's e x a m i n ­

ation of empl oym ent

security agencies

and co lleges

may be limited.

(1970)

and un ive rs it i es

On the other hand,

studies inclu din g a varie ty of org an i za ti on al
Haas,

or Pugh,

suffer

Minings,

1966)

or en vir on men t
1978).

ences in the oper at ion s
as compared

such as

to service

influence

in England

(1967)

of

(1968)

may

to recog niz e oc cu pa tio nal d i ff er enc es

Pfeffer and Salancik,

The

and T u rn er 's ex am in at io n

different work or ga ni zat ion s

for failure

(Rushing,

types,

and Joh ns on 's m u l t i - o r g a n i z a t i o n a l analysis

Hickson,

f ift y-t wo

of

The g e n e r a l i za ti on s of the re­

search on single o rg an iz at io n

Hall,

due to

dif fi cu lt ie s lie in the selection

o r g a n i za ti on s for study.

(1973)

long itu di nal

studies suffer

in size over time,

on

di ffe re nce s
In add it io n

of product

(Osborn,

1951;

important

differ­

producing org an i za ti on s

or ga ni zat ion s may be present.

of size has been the subject

ga n iz at io na l re se arc her s

since Mel ma n's

ma n u f a c t u r i n g industry.

Most of these re sea rc her s

gated the re lat ion sh ip between

1951

of many o r ­

org a ni za ti on al

study of the
investi­

size and the

4

size of the a dm i n i s t r a t i v e component.
has

been m e a su re d most

O r g a n i za ti on al

size

often in terms of the total number

of em pl oy ee s and the a dm i n i s t r a t i v e compo nen t has been
defin ed as a perc ent ag e or proportion

of the total number

of e mp loy ee s ass igned at or above a predefined o r g a n i z a ­
tional

level.

At issue,

as well as the divisio n
Blau holds that
sure

however,

is the definition

of size

of labor.

"in a study

of social

structure,

a mea­

of size based on people who occupy positions in the

structure

is preferable

to one based on dollars or some
i

other units"

(1973,

based on costs

28),

However,

of opera tio n

vide a d d i ti on al

co ntexts

c on si de rat io ns

or populat io ns

for in ve st ig ati ng

eff i ci en cy and economy of scale.
be better unde rst oo d

theoret ica l

co ns tru ct s of both social

In both social and fiscal
di vision of labor

requires

the division

and

is a category

definition.

the concept

in the

of

S pe ci al iz at io n is

struct ur e that

is related

to

of labor and the d is tr ibu tio n of individual

Turner,

1968).

of person nel

and Johnson,

the chief executive

Hickson,

that most broadly defined might

o r ga ni za ti o na l mi ss io n rather
Hall,

(Pugh,

The ad mi n i s t r a t i v e component

include all employees in a role

it (Haas,

economy of

and fiscal contexts.

r e s p on si bi li ti es within an or ga niz at ion
Hinings,

questions of

if it were examined

contexts

that dim e ns io n of or gan ization

served may pro­

Therefore,

scale might

of size

su pportive of the primary

than one directly related
1963).

From this view,

officer and the custodian might

to

both
be

5

included in the ad mi n i s t r a t i v e

compon en t

supp ort of an o r g an iz at i on al mission.
the ad mi n i s t r a t i v e
R u s h i n g ’s three

This

part Id en ti fic at ion

division

fall more

since

the faculty and
In the fiscal

clerical,
component
is

roles in higher education

these categories.

gories in the di vision of labor
clude

the ad mi n i s t r a t i v e

in

in studies of higher educa tio n

traditional

clearly into

of managerial,

of the a d m i n i st ra ti ve component

p ar tic ul arl y app ro pr ia te
o r ga ni za ti on s

The comp le xit y of

com po ne nt might be better add r es se d

and pr ofe ss ion al groups within
(1967).

in their mutual

Ad di ti on al

cate­

in higher educat io n in ­

the support staff.

conte xt the issue of economy of scale

involve s c on s i d e r a t i o n s of efficiency.
some ac ce p te d as su mp ti ons regar din g

Doubt was

cast on

size and ef f ic ie nc y in

a study of the relative effi cie nc y and pro fit ab ili ty of
m a nu f a c t u r i n g c or po r at io ns

(Osborn,

am in ati on of p rof it -lo ss stat em ent s
c o rp or at io ns
declared,

Follow in g e x ­

in small and

r ep re se nt in g major in dustrial

large

groups,

Osborn

’’Co ns id er ing all the u n c e r t a in ti es involved,

u ns a t i s f a c t o r y

92).

the

conc lu sio n is reached that we do not know

very much about
(1951,

1951).

re lative ef fi ci enc y in relation to s i z e ”

He found little to support the ge ne ral iza ti on

that small and m e d i um -s iz e c or p or at io ns are more efficient
than

large ones.

On the other

hand,

ef f ic ie nc y in large c or po rat ion s was

no evidence
found.

for greater

Similar

fin d­

ings were re por te d in a study of twenty m a n u f a ct u ri ng i n ­
dus tries in

1954 (Bain).

6

In the public

sector,

however,

evidence was

study of economies of scale in Iowa public
support of optimum size
per

(Cohn,

1968).

small to large
Mc La ug hl in
more than

co mpared

however,

ev al uat ed B l a u ’s conclusion
influences

that costs
this be n e ­

gained in moving

rather than from large to larger.

size,

M c L au gh li n

in effic ie nci es

in a

schools in partial

Cohn found

student de cre ase d in larger schools;

fit was most evident

found

(1974)

from

In 1979

that com ple xi ty

per student costs.

Like Blau,

found that effects of size are n eg li gib le when

to changes in st affing ratios and changes

plexity of curricula

in c om­

(1979).

Subject of Inv estigation

This

study is based on data from the tw ent y-n in e

co mm uni ty co lleges in the State of Michigan.
reason for this choice
Mic hig an
uniform

public

Since

The primary

1979 officials

in

commu nit y colleges have complied with a

statewide

cation Syste m

is twofold.

public

repor tin g system,

(ACS).

At the same

the Activ iti es

Classifi­

time they have contin ued

to ma int ai n a high degree of autonomy at the local district
level.

Therefore,

programs and services

directed by local

gov ern anc e boards vary greatly from one college
however,

cost and ef fi ci en cy co mpa ri son s can be det ermined

at the statew id e

level

due to the uniform repor tin g system.

This unique cir cu m st an ce
izational

to another;

provides an op p or tu ni ty for o r g a n ­

research in a setting in which structural

nents and dif f er en ti at io n are uniformly

co mp o­

defined in a universe

7
of organizations.
The co mmu nit y coll ege is a rel ati ve ly new organization.
A de sc ri pt io n of its emergence,
order

to prov ide a broader

sy s te m in ve st iga ted

growth,

and role

follows in

pe rs pe ct iv e in which the specific

in this

study may be viewed.

Tw o- Yea r Coll ege Origins

Pro p os al s

to create an org an iz at io n that would relieve

the un iv er si ty of freshman and sophomore
back to 1851 when Henry Tappan,
of Michigan,

president

called for the creation

(Cohen and Brawer,

1982).

ins tr uc ti on

date

of the U n i v e rs it y

of junior

colleges

Early in the twenti et h century

Wi l lia m Ra iney Harper,

president of the U ni v e r s i t y of C h i ­

cago,

of the Uni ve rs it y of. Illinois,

Edmund J. James,

David Starr Jordan,
model which extended
"higher- ord er
Harper

secondary

weak four-year

school curri cu la leaving
In addition,

that some four-year co lleges might convert
programs

into strong two-year

(Cohen and Brawer,

While

all ad vo c at ed the Eu ropean

s c h o l a r s h i p ” for the university.

proposed

programs

of Stanford,

and

junior college

1982),

some un ive rsi ti es might have thrived as upper

div isi on and gra du ate in sti tu ti on s w it hou t

lower div is ion

studies,

few gave up freshman and sophomore

general,

un i ve rs it y offici als were not inc lined to re linquish

this level of i nst ruc ti on nor were
ficials wil lin g
13 and

14.

instruction.

the se con da ry system o f ­

to accept the r e s p o n si bi li ty of adding

Nevertheless,

In

in this climate of debate

the

grades

8
o pp or tu ni ty

for the emergence of two-year col leges

the e st ab lis hm ent

of twenty junior colleges

170 ten years later.

in 1922.

junior co lle ge s increa sed

to 450,

all but five states

by 1909 and to

Th ir ty -s e ve n of for ty -ei ght

c o n t ai ne d junior col leges

(1982)

states

By 1930 the number

of

and they were present

(Cohen and Brawer,

Cohen and Brawer

led to

suggest

in

1982).

that

the un derlying

im p etu s for the a cc ep tan ce and rapid expan sio n of the twoyear college was
ucation,

the A m e ri ca n faith in the benefits of e d ­

me as u re d

in years

with eg al it ar ia n ideals
the es ta bli sh men t
cess

of schooling.

provided the ratio nal e

and growth

factor,

Therefore,

once

this

coupled

to support
Ac­

and "more than any other

acces s depends on proximity"
started

faith

of the two-year college.

to e du ca tio n was expanded,

single

This

(1982,

or gan iza ti on would

10).

flourish.

Two -Year Colle ge Growth

By 1980 more

than half of all students

tion were enr olled in two-year colleges
1982).

In

ton)

Maryland,

Ohio,

Carolina,

Illinois,

Michigan,

and Virginia)

Six of these

Pennsyl van ia,

p re di cte d en ro ll men t increases,

Florida,

(Cohen and Brawer,

1984 two -year college e nro ll men ts exc eeded

100,000 students in fourteen states.
(Arizona,

in higher e d u c a ­

Texas,

and eight

New Jersey,

states

and W a s h i n g ­
(California,

New York,

North

predi ct ed enr oll me nt decreases.

En­

r ol lme nt s na t io nw id e entered a no- gr ow th or slight ly d e c l i n ­
ing status with a likely one to two percent

decrease

9

na ti on wi de

(AACJC,

1984).

rapid expan sio n in both
credit enrollments.

Table

1 shows

by decade

the

the number of two-year co lleges and

Whi le the number of public two-year

col le ge s doubled since

1964,

en rol lm en ts in cr ea s ed

by more

(1982)

cite the work of M.J.

Cohen in

than 500 percent.
Cohen and Brawer
1972 that in ves ti g at ed

the r e l a t i on sh ip betw een popu lat io n

density and the number

of two-year colleges.

He found that commun ity colleges tended to be built
so that 90-95 percent of the state's pop ulation
lived within reaso nab le commut ing distance, about
25 miles.
When the colleges reached this ratio, the
state had a ma tu re commu nit y college system, and few
add i ti on al colleges were built.
As that s t a t e ’s
popul at ion grew larger, the colleges expanded in
enrollments, but it was no longer n ec es sar y to add
new campuses.
Cohen id en ti fi ed seven states that
in the early 1970s had mature systems:
California,
Florida, Illinois, New York, Ohio, Michigan, and
Washingt on .
In these states, the denser the pop­
ulation, the smaller the area served by each college,
and the higher the per- ca mpu s enrollment.
Applying
this formula of the relations among numb ers of
colleges, state population, and po pu l at io n density,
he showed that 1,074 public commu nit y colleges would
ef f ec ti ve ly serve the nation (Cohen and Brawer,
1982, 12).
Table

1 indic ate s

that in 1984 1,064 public two -year c o l ­

leges were established.

C o h e n ’s pr oj e cti ons would

both the st abi li za ti on in enr ol lme nts
cline in private colleges.

support

as well as the d e ­

Furthermore,

Cohen 's i d e n t i f i ­

cation

of Mic hig an as a state with a "mature system" e n ­

hances

the se lec ti on of that state's s yst em as the focus

of this study.

TABLE 1:

Fifty Years of Growth in Two-Year College Credit Enrollments* 1933-34
to 1983-84.

Year

Public

Private

Total

Public

Private

Total

1933-34

223

309

532

77,111

33,138

110,249

1943-44

261

323

584

74,853

32,954

107,807

1953-54

327

267

594

533,008

69,856

602,864

1963-64

503

268

771

913,057

120,346

1,033,406

1973-74

910

231

1,141

2,729,685

136,377

2,866,062

1983-84

1,064

155

1,219

4,799,768

148,207

4,947,975

1984-85 (est.)

1,064

155

1,219

4,720,000

150,000

4,870,000

*Does not include an estimated 4.5 million noncredit enrollment
Data Source: AACJC Letter, 1984

*

11

Role of the Two-Year College

The

"junior

college" has been defined as "an i n s t i t u ­

tion off ering two years of ins tr uct ion of strictly co llegiate
grade"

(Bogue,

As s oc ia ti on

1950,

p.xvii);

however,

that

1922 Ame ric an

of Junior Co lle ge s def in it io n has since been

broadly expanded.
college" was

In the

1950s and

1960s the term "junior

commonly used in reference

to church supported

two-year colleges and to uni ver si ty branches
lo we r-d iv isi on instruction.
came

into

The term "com mun it y college"

being as an id e nt if ica ti on of publicly supported

tw o-year colleges with c om pr eh ens ive
Brawer,

programs.

(Cohen and

1982).

Co m pr eh en si ve
year college.
programs

programming exp an de d the role of the two-

In add it ion

designed

university,

to provid in g

for students inten din g

traditi ona l academic
to transfer

to the

the co mp reh ens iv e commun ity college also provides

vocational and technical education.
signed

providing

to provide

through vocational

These

students with entry

level

programs are de­
job skills

training in c omb in ati on with general

gained
ed­

ucation requirements.
The

role of commu nit y education

the co mp re hen si ve commun ity college.
gan summarizes
anything,

this role:

"We will

at any time whenever

p. 215).

Cultural,

dimens io n of

A Texas college's slo ­

teach anyone,

anywhere,

there are enough people

interested in the program to justify
1950,

is another

its offering"

recreational,

(Bogue,

and hob bicraft

12

co u rse s

are

co ll ege s.
signe d

co mmo n n o n c r e d i t
In ad di tio n,

to pr ov id e

continuing

licensure

commonplace

th r o u g h
and

he al th

involved

te c h n i c a l

of c re di t

assistance

are

pha rm ac is ts ,

and

C o m m u n i t y c o ll eg es

development activities

in ec o n o m i c
and

de­

toward

requirements

as realt ors ,

in e c o n o m i c

co u r s e s

to be ap p l i e d

practitioners.

t r ai ni ng

in most c o m m u n i t y

education

or c e r t i f i c a t i o n

participation

t h ro ug h

continuing

for such group s

oth er a l l i e d
are also

units

offerings

development

consulting

p r og ra ms

ser vic es

d es ig ne d

c o un ci ls
in di rect

for b u si n es s

and

industry.
Finally,
mission

be ca us e

ins tit ut io n ,

previous academic
grams are
l e ar ni ng
and

success,

labs,

study

sk i ll s n e e d e d

ve l o p e d

st u d e n t B

r e me di al

Bas ic

skills

s ki ll s

co lleges,

comprehensive

public

is us u a l l y

of
pr o ­

tut orials,
cou ns el in g,

are d e s i g n e d

and

applied

is no doubt

that

in part

to the c o m p r e h e n s i v e

college.

developmental

special

Since

(Cohen

the ra pid

regardless

ad­

to

the p r e r e q u i s i t e

p ro g r a m co m pl eti on.

p rog ram s.

" c o m m u n i t y ” c o ll eg es

an open

courses,

to a cq uir e

for s u c c e s s f u l

t wo -y ea r

due

or

centers,

s tu de nt s

"community college"
and

college,

assistance activities

lo w a b i l i t y

Most

a d m i t s all

c o m m o n pl ac e .

peer-group

enable

the c o m m u n i t y

private,
the 19 70 s
to both

and Brawer,

gr owth

have d e ­
term

"ju n io r"

1982).

in t w o- ye ar
role of the

the

There

colleges
c o m m un it y

is

13

Purpo se

of the Inv es ti ga ti on

In this study hy p ot hes es will be tested to inv es tig ate
the influe nc e of size

on eff ic ien cy and economy of scale

in

the Mi chi gan c om mu nit y coll ege system.
The com mu ni ty
is a comp lex
grams

college s yst em in the State of Mic h ig an

or ga ni za ti on

in opera tio n and size.

of the tw en ty -n in e com mu ni ty

colleges

vary from tr a di ti on al ac ade mi c study
serving

specific

in Alpena and

year

in the sy st em

to uniqu e programs

po pul at ion s such as concrete te chnology

ski lift m ai nt ena nc e in Gogebic.

Va r ia ti on s
fiscal

The pr o ­

in size are also

1984 general

low of $2 ,6 25 , 15 2. 00

great

(see Table 2).

fund e xp en di tu re s ranged

In

from a

in Glen Oaks Com mu ni ty Col leg e

to

a high of $ 39 ,3 99 ,7 04 .0 0 in Macomb C o m mu ni ty College.
Student

enr ol lm en ts

equat ed students
15,831 FYES
credit hours

ranged from a low of 753 full- ye ar-

(FYES)

in Glen Oaks C.C.

in Oakl and C o mm un it y College,

P er son ne l emp loyed were as

few as 73.5 f u l l - t i m e - e q u a t e d

positi on s at Glen Oaks C.C.

and as high as 1,283 at Ma co mb C.C.

size is

Phy s ic al

plant size

129,000 gross square feet at Glen Oaks C.C.

to 1, 64 5, 74 8 at Oakla nd C.C.
Another

and fiscal year

gener ate d range d from 23,363 at Glen Oaks C.C.

to 490, 777 at O ak la nd C.C.

varied from

to a high of

dimension

(Grobe & Root,

of the di ve rs it y and co m pl e xi ty

the d if fe re nce in me asu res

among the twe nt y-n in e

1985).
in

of op er a ti ng eff ici en cy

organizations.

For example,

cost of

14

in st ru ct io n per FYES ranged from $1 ,1 82 .0 0 in Sou th we st er n
M ic hi ga n Coll ege to $1,998 .0 0 in W a s ht en a u C om mu nit y C o l ­
lege.

Energy costs

TABL E 2:

in physic al

plant

operation

ranged

Size Va ri ab le s of Mic hig an Pu blic Commu nit y
Co lle ges in F isc al Year 1984

Variable

Mean

Fu l l- Y e a r - E q u a t e d

St ude nt s

(FYES)

S t .D e v .

(n=a29)

4285.34

4171.74

216.59

180.40

12 ,6 36 ,7 43

10 ,946,361

Ope r at in g E xp en d it ur es per FYES

3155.09

436.68

I n s t r u ct io na l Ex p e n d i t u r e s
FYES

1580.21

218.72

132,862

129,327

464,066

361,607

F u l l - T i m e - E q u a t e d Person ne l
FTE
Ope r at in g Ex pen d it ur es

per

Credi t Hours G e ner ate d
Physical Plant
Source;

(square

feet)

Grobe and Root,

1985

from a low of $1 2 5, 33 9. 00

in North Central

Colle ge

to a high of $ 2, 03 6, 75 2. 00

& Root,

1985).

munity

A complete

in Oakl and C.C.

list of M i ch ig an

public

(Grobe
com­

col leges may be found in A p pe nd ix A.

In view of these differe nce s
operation,
support

Co mm uni ty

a funding app ro ach

rates and

costs of

that pro vides eq uitable

given the mix of programs

A d d i t io na l vari ati on s

in size and

is difficult

in tax rates,

political en vi ro nm en ts

tax bases,

further

to achieve.
tuition

comp lic at e

15

de v el opm ent of an a ppr opr ia te
of this

funding formula.

study is to in ves tig at e

o rg an iz at io na l varia ble s

related

Is there an optimum economy

the influence

The

purpose

of size on

to the problem of funding.

of scale in the Mic hi gan

co m ­

munity college system?

What size and resource a ll oc at ion

factors cont rib ut e most

to the a ch ie ve me nt of greater e f ­

ficiency?

The purpose of this study is limited to q u e s ­

tions of ef fi cie ncy without

regard for progr am e f f e c t i v e ­

ness or quality of achievement.

It cannot

be assumed

m ax i m u m eff ic ie n cy equates to m ax im um effectiveness;
ever,

the results

that
how­

of this study may provide data for e d u ­

cation d e c i s i o n -m ak er s

dealing with

problems of efficiency.

C o n t r i bu t io n of the In ves ti ga ti on

Since

1979 the legisl atu re

co o pe ra ti on with

of the State

of Mi chi gan

the De pa rt men t of Ed uca ti on and the D e ­

partment of M a n a g em en t and Budget has been seeking a
formula
State

funding model

that will effect

support of M i c h ig an 's

leges at the same time
d iff ere nc es

parity in the

twenty -ni ne co mmu nit y c o l ­

that it reco gn ize s existin g

in size and local resources.

severely co mpl ica te d

by v ar ia ti ons in size among the

col leges and the assu me d effects of those
The results of this
better

This task is

study may provide

de cision making with

respect

variations.

empirical data

to issues

of economy

of scale.
Or g an iz a t i o n a l

theory might also

for

be ad va nc ed

by

in

16

providing
Blau

further

(1973,

1974)

support
and more

that d i f f er en ti a ti on
affect

for the theories ad vanced by

student and

Stomm el' s

mo dels

or g ani zat io n

ad di ti on al

Scott,

than size will

finding

(1985)

that

public

community colleges.

co nt rib ut ion s

(Perrow,

per

size in land-grant uni versities

(Pfeffer & Salancik,

wo r k theories

(1979)

rel at io ns hi p exists between expenditure

will be tested in Mic h ig an
Finally,

by M c L au gh li n

and comp le xit y more

economy of scale.

a cu rvi li nea r

recently

1979;

to resource de pendence

1978) and environment and n e t ­
Pfeffer & Salancik,

1978;

1981) may also be realized.

Ov e r v i e w of the Study

This study is o r g a ni z ed into five chapters.
ground,

the theoretical

framework,

an t i c i p a t e d co ntr ib ut io ns
Chapt er

I.

gations

of o rg an iz ati on al

the purpose,

of the study were

A review of the literature

provide d in Chapter

II.

Chap te r

Op er at io na l def in it io ns

of the study and
in

and me th od ol og y are

Fi ndi ngs and i n t e r p r e t a ­

tests are rep orted in Chapter IV,

and co nc lu sio ns regard in g
of scale in Mic hig an

to i n v e s t i ­

each hypothesis are set forth

in Chapt er III.

tions of st atistical

presented in

related

The hy po th es es

under ly ing

also co nt a in ed

and the

size and economy of scale is

the rationale
III.

The b a c k ­

the influence of size and economy

public

commun ity colleges are pr e ­

sented in Chapter V along with im pl ica ti ons
college a dm i n i s t r a t o r s and state

level

for community

decision makers.

17

Ch a p t e r V c on c l u d e s wi th
sea rc h .

recommendations

for fu ture

re­

CHAP TE R IX

REVI EW OF THE LIT ER AT UR E

Intro duc tio n

A review of the lit era tu re in three broad cat egories
related

to this study is presen ted in this chapter.

c a te go rie s are or ga ni za ti on a l
ganiza ti ona l

size

O r ga ni za ti on al
of the organ iz ati on

e xam ina ti on

Size and Its Influence

has been the subject of much
inquiries was Melman's

industries in 1951.

that org ani za t io n

inquiry.
study of

Melma n con c lu de d in his

size and

the or ga ni za tio nal

a d mi ni st ra t iv e component were inversely related.
total number of personnel Increased,
mi n is tr at iv e component
personnel.

of

size and its influe nce on the structure

the first of these

m a nu fa ct ur in g

and the re lat io n sh ip

or­

to economy of scale.

Org a ni za ti on a l

Among

size and its influence,

size and efficiency,

or ga niz ati on al

These

As the

the number in the a d ­

decreased as a perc ent ag e of total

Size in Mel man 's

study was mea su red in terms of

the total number of employees.
Using the same measure of size,
ployees,
fornia

Terri en and Mills

schools

found in a 1955 study of C a l i ­

that the ad m in is t r a t i v e compo nen t

in large schools.
m e nt ar y school

total number of e m ­

Their

districts,

findings were
high school

18

consistent
districts,

increased
in e l e ­
and

19

unified and city school districts.
found increases in the mean
a d mi ni st ra ti v e component.
increase was from
17.6% in large

In high school districts

(1955,

schools

13).

Or g an iz ati on al

the

to as much as

However,

this study is m e a su re d in terms of total
regard for the number

they

perc ent ag e of employees in the

11.4% in small

schools

In each group

school size

personnel without

of st udents enrolled.

size was re c og ni ze d in two dim ensions

in An de r so n and W a r k o v fs study of the ad m i n i s t r a t i v e
ponent
sured
than

in hospit al s

(1961).

Or ga ni za t i o n a l

by the annual avera ge daily
the total ho spital

force;

patient

however,

found

to be es sen ti all y equivalent.

tant,

Anderson and Wa rk ov

variable,

org an iz a ti on al

Or ga ni za ti on a l

complexity,

ultaneously.

co m ­

size was m e a ­

load rather

both mea su res were

Perha ps more i m p o r ­

int rod uc ed a second

indepe nde nt

into their

study.

com ple xit y was me asu re d in terms of the

number and variety of tasks perfor med
m ax imu m number

in

as well as the

of sp ec ia li ze d tasks that might occur s i m ­

In r el at ion shi p

to org a ni za ti on al

size and

co m p l e x i t y An derson and W ar ko v found:
1.

The rel ative size of the a d m in i st ra ti ve component
decreases as the number of persons pe rforming
identical tasks in the same place increases.

2.

The relative size of the ad mi n i s t r a t i v e compon ent
increases as the number of places at which
work is perfo rme d increases.

3.

The relative size of the ad mi n i s t r a t i v e component
increases as the number of tasks performed at
the same place increases (or as roles become
in c rea si ngl y sp ec ial ize d and dif fer en t ia te d )
(1961, 27).

20

In view of these findings,
stood

Terr ien and Mills may

be u n d e r ­

to conf irm the second pr op os it io n without nu ll if yin g

the first
school

pr op os it ion

since in their

districts mul tip le

the stru ct ura l

study of Ca l if or ni a

sites were commonplace.

Clearly

di men sio n of co mp le xi ty was e st ab lis he d in

this study as an im po r ta nt c o n s i d e ra ti on of size.
Str uctural

c o ns id er at io ns

in the work of Haas,
of the size

Hall,

comp on ent "

str a ti ve co mpo nen t
compon en t
directly

still

and J oh nso n in their

of the su pp or tiv e compon en t

zations of vari ous types.
" s up po rt ive

are expanded

further

1963 study

in thirty o r g a n i ­

In their study the term

is used in place of the a d m i n i ­

in pre v io us studies.

The s up por tiv e

is a larger group in cl u si ve of all person ne l not
re la ted to the o r g a n i za ti o na l

lieved

that this

broader

category made

unlike

o rg an iz at io n s more applicable.

goals.

Th ey be ­

co mp ari so ns among
The varying size of

the su pp or ti ve

co mp one nt was exa mi ned

o r g a n i za ti on al

ch ar a c t e r i s t i c s such as total em ployee

number of ope ra ti ng

sites,

tion and function.

The work of Haas,

clearly

and or ga n iz at io na l
Hall,

to other
size,

differentia­

and Johnson

tests the co nc lu si on s of both Terr ien and Mills and

And er son and Warkov;
The

in rel ation

however,

imp o rt an t di ffe re nce s exist.

previous studies exa mi ne d a more narrowl y defined

employee c om po nen t w it h in
Haas, Hall,

and John son

a single o rg an iz a t i o n a l

sought g e n e ra li za ti o ns

to a variety of o rg an iz a ti on s

a p p li ca bl e

and the com pa ris on

"s up por ti ve co mpo ne nt, " was inclus ive

of more

type.

group,

personnel

21

ranks than
Haas,
em pl oy ee s
za ti on al

in studies
Hall,

of p r ev io us

and Jo h ns on

size

incre ase d

the s u p p or ti ve

profit

found

(1963).

og an iz at lo ns

type;

compo nen t

zation,
causal

suggested.

that o r ga n i z a t i o n

however,

person nel

size

(1963,

and n o n ­

than in private,
16),

of personnel

On the other hand,

and they

in an o r g a n i ­

occupy mu lti pl e

than d i f f e r e nt ia ti on

a co n cl us io n in con fli ct with

Wa r k o v

they also found

sites.

No

they found

size in f lu e nc ed the size of the s u p p o r ­

tive com po ne nt more
age,

the number

the more likely it will
link was

d e cre ase d as o r g a n i ­

This re la ti on sh ip was

had even less

that the larger

the pe r ce nt ag e of

in go ve rn me nt al

profit or ga n i z a t i o n s of similar
found

that

in the s u p p o rt iv e component

found across or ga ni za ti on
that

researchers.

or org an iz at io n

that of A n d e r s o n and

(1961).

Haas,

Hall,

and Johnson,

were the first to note that

the r e l a t io ns hi p between or g a n i z a t i o n
the su pp or ti ve

size and the size

compo nen t may be curvilinear.

that s ca tt er gr am s

showed larger

of

They found

p ro po rti on s of pe rsonnel

in su pp or ti v e ac tiv it y in small and large org an iz at io ns
than in me dium
ever,

that

size organ iza ti ons .

Later

how­

due to the small number of m edi um size o r g a n i ­

zati ons in their
should

They concluded,

be taken

study "the c urv il ine ar

as sug ge st iv e only at this time"

res ear ch was
The first

r el at io ns h ip
(1963,

14).

to find this r e l a t io ns hi p more important.

study of or ga ni zat io nal

size and its influence

on higher e d u c a t i o n o rg a n i z a t i o n s was rep orted

by Hawley,

22

Boland,

and Boland

in 1965.

O rg an iz at io na l

size in their

study was defin ed as the number of full-time and part-time
faculty

reduced

to full- tim e equivalents.

Dep en d en t v a r i ­

ables inc lud ed the number of p ro fes si ona l adm inistrators,
the number of departments,
found

and the op era tin g budget.

that facu lty size and budget

a d mi n i s t r a t i v e
by the number

size more

They

size inf lue nce d the

than d if fe re n t i a t i o n as mea su r ed

of departments.

This finding would appear

to

be i nc on si st en t with An der son and Warkov regard ing the
influe nc e of dif fe re nt ia ti on
larger number

or complexity;

however,

of faculty re sul ti ng from Increased

the

differ­

entia ti on may over ide the effect of di ff e re nt ia ti on alone
in Hawley,

Boland,

and Boland.

This

study,

however,

affirms

Anderson

and Wa rko v in finding that

strators

to faculty tends to decline as size of faculty

inc re ase s"
son,

(1965,

253).

"the ratio of a d m i n i ­

Again as in Haas,

Hall,

slight att en ti on is given to a tendency

linearity,
plotted

but it "disapp ea rs when a scatter

on a log-log grid"

Studies
employee

(1965,

of the influence

groups within

by di ffe re nce s

size

size on

the o rg an iza tio n may be compli cat ed

in oc cu pat ion s as well as by employee c l a s ­

the a dm i n i s t r a t i v e

Rushing

(1966)

postu­

component may vary by o c c u p a ­
or dec rease with o r g a n i z a ­

due to o cc upa ti ona l

or ga ni za ti on al

diagram is

of o r g an iz at i on al

tion so that it might increase
tional

toward curvl-

253).

sif ic ati on within the organization.
lated that

and J o h n ­

size differences.

di ff ere nce s rather
Div idi ng the

than

23
a d mi n i s t r a t i v e
ical,

compon ent

professional,

"firm size has

into five groups - managers,

sales,

and

service - Rushing

found that

quite diffe ren t effects on diffe ren t c om­

ponents of a d m i n i s t r a t i v e pers onn el "
findings called "into question
a d mi ni st ra ti ve

cler­

(1966,

106).

His

the popular criticism that

personnel uniformly and sy st e ma ti ca ll y i n ­

crease di sp r op or t i o n a t e l y as bur ea ucratic or g an iza ti ons
increase

in size"

may explain

(1966,

in part

107).

Rushing's obse rva ti ons

the incons ist en t

findings

of previous

research which varies greatly in definition of the a d m i n i ­
strative component.
Rushing

tested further

the findings of And er son and

Warkov in light of his c on si der at ion s
strative component

(1967).

tries Rushing examined
division of labor
personnel.

regarding

In a study of forty-one

the effects of industry

on the rel ative number

He defined industry

of pro duction personnel,

clerical

personnel

strative personnel

not total

personnel.

The a d m i n i ­

professionals,

personnel.

plexity or division of labor was measured
Martin formula

size and the

of a d mi ni st ra ti v e

was me asu re d as a ratio

to pro duction

indus­

size in terms of the number

strative component inclus iv e of managers,
and

the a d m i n i ­

of a d m i n i ­

Ind ustry co m­

by the Gibbs-

based on the di str ib uti on of personnel

among the different

depart men ts in the organization.

Rush in g found:

1.

The relative number of a dm in is tr ati ve personnel is
directly related to the division of labor among

24

pr o du ct io n personnel, but inversely related to the
total number of pro duc ti on personnel.
This is
true for all a d m i n i s t r a t i v e groups.

Thus,

2.

Co nt ro ll in g for the effects of size and division
of labor does not usua lly cause the relati ons hi p
between the other variable and the re lat ive size
of the a dm i n i s t r a t i v e groups to disappear.
In
fact, control for the division of labor Increases
the r el at io ns hi p between industry size and a d m i n i ­
stration.
To this extent, the effects of the two
va ri abl es are independent.

3.

At the same time, however, the variables Interact.
The effe cts
of the division of labor are greater in
smaller industries, while the effects of industry
size are greater in indu st rie s where the division of
labor is high.
This finding, however, is not true
for the relative number of pr ofe ssional personnel.

4.

Finally, the ratio of cl erical and professional
perso nne l increa ses as the division of labor in­
creases (1967, 293-294).

the hy po th es es

however,

of And e rs on and Wa rk o v are affirmed;

they do not account

for all of Ru shi ng 's

The r el at io ns hi p of o rg an iz at io na l
is ques tio ne d in the study
(1967).

of Hall,

size to com plexity

Haas,

and Johnson

In their ex am in at io n of a wide variety of both

product and service

or ga niz at ion s

they found

ity and f o rm al iz a ti on were not ne ce ss ar il y
size,

findings.

but rather

dep e nd en t

on the nature

and purpose.

the less

that

size produces

ch ara ct er is ti cs

formal the structure

Little ev idence was

called into

of

the more profes si ona l

found

regardless

in support of a causal

link between size and o rg an iz at ion al
therefore,

con ditions

of the or ga n iz at io n' s work force

They suggested

the work force,
of size.

they were structural

that c o m p l e x ­

structure,

and they,

question the theory that expanding

greater d i f f e r e n ti at i on

(1967).

r

25

Hall*
Hawley,

Hass,

Boland,

and J oh ns on r e in for ced the finding of
and Boland regarding ec ono mi c in fl ue nc es on

an orga ni zat io n.
budget

size,

Hawley,

Boland,

and Boland

more than complexity,

found that

may have in fl ue n ce d

size of the ad m i n i s t r a t i v e compo nen t

(1965).

Hall,

Haas,

and John son ar gued that both size and co mp le xi ty may
depend ent on av ai la bl e resources and eco nom ic input
In 1970 Peter M. Blau set forth his formal
d i f f e r e n t i a t i o n in or gan izations.
tions were deduced

(1967).

theory of

Two major g e n e r a l i z a ­

bureaus.

by Blau and Sc h oe nh er r

ge n e r a l i z a t i o n was

be

based on e mp iri ca l fin dings in a study

of gove rn men t emp lo ym en t
reported

the

"incr easing

Complete results were

in 1971.

B la u's first

size genera tes

structural

di f f e r e n t i a t i o n in o r g a n i z a t i o n s along va r iou s di me ns io ns
at de ce le ra ti ng
tion was
larges

rates"

"s tructural

Haas,

(1970)

(1967)

labor.
(1967)

(1975)
On

The

reported above.

vary

in de n ot at io n among researchers.

used the term "d i ff er en ti at io n "

guished wi thin

213).

is in direct conflict with that of

nu mber of structural com po ne nt s

Scott

(1970,

be noted that the terms "d if fer en ti at io n"

"compl ex ity "

Blau

The second g e n e r a l i z a ­

d i f f e r e n t i a t i o n in o rg an i za ti on s e n ­

and John son

It should
and

204).

the a dm i n i s t r a t i v e co mp o ne nt "

second g e n e ra li za ti on
Hall,

(1970,

to refer to the

that are formally d i s t i n ­

the or ganization.

For both Blau

(1970)

and

d i f f e r e n ti at io n refe re nce d the di vis io n of

the other hand,

Hall,

used the term "com ple xit y"

Haas,

and Johnson

to re fer en ce the div isi on

26

of labor and
number

"d if fe re n t i a t i o n "

of hi er ar ch ia l

"compl ex ity "

levels.

to re fe r en ce the

Scott however,

ar gued that

of work should not be con fused with " d i f f e r e n ­

tia tion" of roles.
the second

is used

The

first is an aspect of technology;

is a facet of structure

Differentiation

(1975).

is B l a u ’s prim ary concern.

He found

that d i f f e r e n ti at io n at first grows rapidly with increases
in o r ga ni za ti on al
is realized.
large

size but more

Blau argued

o rg an iz ati on s

w i th in

slowly when very large size

that economy of scale occurs

due to increases

in the span

the hi era ch y of the or ganization.

creases in the wo rk group size result

in

of control

That is,

in­

in a decrease in the

p r op ort io n of man age rs as a part of the total work force,
thereby

result in g in economy of scale in ma nagement.

finding is in direct contrast
who

found that

school distr ict s

however,

and un iv er sit ies

study

com­

in large
exceed

These econo mie s decline

with c o nt in ue d exp an di ng

In a second major

sum ma riz in g

increa se s

Blau argued that ec on om ie s of scale

ex pe ns es of di f fe ren tia ti on.
idly,

(1955)

the perc ent ag e of the ad mi n i s t r a t i v e

ponent in Ca li for ni a
districts.

to Terri en and Mills

This

size

his earlier

rap­

(1970).

based on 115 Ame ri can

Blau aff ir med

the

colleges

findings.

In

the process of d if f e r e n t i a t i o n Blau g en era te d

three "theorems":
1.

I nc re asi ng org a ni za ti on al size gener ate s
entiation at de cli nin g rates.

differ­

2.

Inc rea sin g org a ni za ti on al size reduces a d m i n i ­
strative ratios at de cl in in g rates.

27

3.

S tr u ctu ral d i f f e r e n ti at io n in org an i za ti on s
enlarges a dm i n i s t r a t i v e ratios (1973, 259-260).

Bl au's me as ur e of size was
The

third of these three

context
though

of as sum pti on s

the number of faculty members.

theorems must

be un derstood in the

regarding economy of scale.

di ff ere nt iat ion

Even

incre ase s a d m i n i s t r a t i v e ratios,

the

o r g a n i za ti o n realizes econ omy of scale ben efits due to
overall

increa se s in the total work force.

Re c og niz in g the co nt ra s ti ng
size r ep re sen ted
found minimal
of Blau

(1970)

by Hall,

on the effects

and Joh nso n

of

(1967) who

effects as so ci at ed with size and those views
and Blau and Scho enh er r

great impo rt anc e
effects

Haas,

views

to size,

Meyer

(1972)

of size in a causal model.

(1971) who att ached
chose to test the

He co ncl ud ed that size

cannot

be u n de re st im at ed

in its impact on other o r g a n i z a ­

tional

cha ra cte ri sti cs.

In his study of 194 city,

and state dep ar tme nts
apparent

of finance,

r el at io nsh ip s among

vanished when

Meyer det erm in ed

org a ni za ti on al

size as a variable was

M e y e r ’s view was

the u n id ir e ct io na l

za tio nal c h a r ac te ri st ic s

cause

by Hol daw ay

and Blowers

(1971).

organi­

were

They found

school systems

Perhaps more noteworthy,

finding regar din g

of other

(1955)

strative ratios decline d with increa ses
system.

Size in

(1972).

ex am in at io n of forty-one urban

school

that

ch ar ac te ri st i cs

controlled.

The findings of Terr ie n and Mills
again

county,

tested
in their

that a d m i n i ­

in the size of the
however,

the p ro fe ssi ona l group within

is their

the

28

administrative
increases

in s y s t e m size.

c h o l o gi st s,
showed

c o m p o n e n t wh i c h

social

workers,

o ff ic e m a n a g e r i a l

tionships
(1966,

(1971).

and

group,

1967)

in

Still

their

furt her

fi n d i n g s

in K a s a r d a ' s

s t ra ti ve

component

reported

that

to sy st em

in c r e a s e s

Base d

on this

his a r g u m e n t
c o mm on ly

r es ea rch ,

demonstrated

that

per

and c u r r i c u l a r
per student.
economy

number

institutional
how ever,

and

professional

have

differentiation.

groups

(1970,

model,

1971,

in 1979

in s t af fi ng

far gr ea te r

by

ef f e c t s

e ff ec t
ratios
on cost

of size in p ro d u c i n g
er as ed

Costs

and d e g r e e s

and

McLaughlin

size has a direct

can be c o m p l e t e l y

in

c o mp one nt.

p r ov id e d

changes

he

groups.

to be c or rec t

f i nd in gs

The p o s i t i v e e f fe ct s

of c u r r i c u l a

He

and Blowers,

in size of pe r s o n n e l

for B l a u ' s

complexity

of scale

c o m p l e x i t y and

like H o l d a w a y

a path a n a l y s i s

student;

systems.

c o m p o n e n t was n e g a t i v e l y

d i f f e r e n t i a t i o n were

Using

v a r i a t i o n was

of the a d m i n i ­

schoo l

in the a d m i n i s t r a t i v e

McLaughlin.

the v a r i a t i o n s

component.

R u s h i n g ap p e a r s

va r i a n c e s

s up por t

regarding

on cost

but

rela­

affirmed Rushing

examination

in 178 C o l o r a d o

size,

included

Fu r t h e r

(1974)

size while

negative

this o b se rv ed

in the c l er ic al

for

showed

a pe r s o n n e l

the m a n a g e r i a l

found

organization

regarding

for

psy­

t ea ch in g c on su lt a nt s,

p er s o n n e l

support

provided

re l a t e d

including

H o l d a w a y and Bl ow er s

in size of gr ou ps w it hi n

the

This

p os i t i v e r e l a t i o n s h i p s with

ce n t r a l

1973)

did show i nc r e a s e s with

by i n c re as ed

can be r ed uce d

can be held

if

c o ns ta nt

29

while enrollment

is incre ase d

(McLaughlin,

1979).

In a c om pr eh en s iv e review of eighty empirical
on org a ni za ti on al

size and its str uctural influence,

berly noted that while

the number

the most common measu re

of employ ee s was

of org a ni za ti on al

d ef ini ti ons of size are lac king"
v ar ia tio ns in mea su res of size
size,

(1976,

based

concept of size is ambiguous

574).

inquiry

role.

Among

pacity,

Because

the aspects

concepts of
The

it is too
by li m ­

of size cited are physical

personnel,

inputs and outputs,

able d is cr et io n ar y resources.

From these more

ated aspec ts of size,

r el at io nsh ips

structure and aspects

of its size might

co nt ra di ct io n s

A second major

differenti­

(1961)

to resolve

in Terri en and
might

then be

than on

"m et h­

(1976).

prob lem in the study of o r ga ni za ti on s is

the variety of or ga niz at ion types,
there are different

o rg an i zat ion al

influence or gan iz at io na l
type to another.

Efforts

r e f or mu la ti on " rather

refinement"

and a v a i l ­

be tested with

such as those found

(1955) and And er son and Wa rkov

based on "theoretical

ca­

between o r ga ni zat ion al

greater co nsi st enc y among researchers.

odo log ica l

of

to some aspect of size and its theore tic al

av ail abl e

empirical

clearly

to compare.

simply because

Kim­

"conceptual

Kim berly ca lled for greater sp eci fi ca ti on

iting

Mills

size,

on varied

results of studies are difficult

broad.

studies

Ki mb er ly

argued

that

types and that size may

structure di ff er en tl y from one

The role of size may be in dep en den t of

or ga ni za ti on a l type,

and aspects of size may vary as a

30

functio n of or ga n iz at io na l
previously,
supp or tiv e

Haaa,

Hall,

component

tions was smaller

type

(1976,

and Johnson

in government

than in private,

Kim be rly concl ude d

(1963)

in org an iz at i on al

type are diffic ul t and l i k e ­
Im pli cat io ns

the influence

have sought to

of o rg ani za tio n size upon

This

as well as how they might be defined.

strong evidence is provide d in support of the

of d i f f e r e nt ia t io n as a key size factor

all other

An early
iency and

influe nci ng

comp one nt s within the organization.

O r ga ni za t i o n a l

Size and E f f i ci en c y

inquiry into the relation of size to e f f i c ­

pro fi t ab il it y was conduc te d

by Os bo rn

exa mi ned then ac ce pt e d gen er al iz at io ns
small

itself.

the

is co mpl ic ate d by both the vari et y of o r g a n i ­

za ti ona l comp one nts

role

of dif ­

type are ignored.

various com pon ent s wi thi n the or gan i za ti on

However,

organiza­

profit organizations.

In summa ry or gan iz ati on re searchers

ex a mi na ti on

found that the

and service

ly to continue as long as th eoretical

determine

As noted

from his lite rat ur e re view that problems

of sa mpling and org a ni za ti on al

ferences

594),

(1951)

who

that me di um and

size cor po ra ti on s may be more effici en t and more

profitable

than large corporations.

He c o n s id e re d e f f i c ­

iency in terms of costs of production and costs of
m a r ke ti ng

goods or services

Efficiency,

then,

in a c o mp et iti ve environment.

is relative to com parable co nd it io ns

business operations.

Dif fe r en ti al s in geo gra ph ic

of

location,

31

similarity of product,
c o mp lic at e cost
tions Osborn

labor costs,

comparison.

found that,

in pr of ita bi lit y

and other variables

In light

of these c o n s i d e r a ­

in large corporations,

tend to be narrow er

ations.

He did not find s ubs tan ti al

pothesis

that

than in small c o r p o r ­
support for the hy­

large co rpo rat io ns are less

less efficient;

however,

the ranges

he did conclu de

prof it abl e or
that large b us­

inesses freq ue ntl y trade off increa sed p r o f i t a b i l i t y for
Incre as ed

stability.

He was unwill in g

They grow c on ser va tiv e

to co nclude that much is known about

relati ve e ff ici en cy in relation
Other econ om ist s
in ves ti ga ti on

but stronger.

to size

reached similar

of twenty ma nu f a c t u r i n g

sought to de ter mi ne the
lowest unit costs"

"minimal plant

(1954,

18).

Based

(1951).

conclusions.
industries,

In an
Bain

size r eq uis it e
on data

gained

for
in

qu e st io nna ir es es pe ci al ly design ed for each of the r e ­
spondents,

Bain

mul t ip la nt

firms varied among industries.

empiri ca l

found

found ati on

in con si d er at io ns
market entry,
un im por ta nt

that economies of scale in large

for economies

No consistent

of scale was

found,

and

regarding product c o nc en tr at io n or

the issue of economy of scale was g en er all y

(1954).

Bain' s fin dings are sup po rt ed

fu r t h ­

er by Shepher d who ex ami ned econo my of scale as a d e t e r m i ­
nant

of corporate

structure.

"resea rc h into economies
(1979,

259).

question

Sh eph erd con cl ud ed that

of scale teaches one mo des ty "

Em pir ica l findings are few and open to

(Shepherd,

1979).

32

Eco no mic

an aly ses

of economy of scale suffer for lack

of clari ty in the basic concept of scale

(Gold,

After r ev ie win g a vari ety of concepts of scale,
for a r ed ef in i ti on
level

of planned

extent

to wh i ch

su b di vi si on

spe c ia li za ti on

of the component
(1981,

in terms of "the

has

been appl ie d to the

tasks and fa cilities of a

15).

On the basis of this more

concept of scale a variety of models

econ om y of scale might
af fe cte d

Gold argued

pro du ct io n capacity which de t er mi ne s the

unified o pe rat io n"
precise

of the scale concept

1981).

be developed.

for assessing

V ar ia tio ns in scale

by tec hno lo gy and o r g a n i z at io na l com pl ex it y might

be ac c o u n t e d for by co n si de r a t i o n s
the di vis ion

of s p ec ia liz ati on and

of labor as o r g an iz at io na l comp one nt s a f f e c t ­

ing pr od uct io n cap aci ty
E c on om ist s

(Gold,

1981).

such as Osborn,

Bain,

and Gold have

co n ce rn ed for economy of scale and industrial
Little att en ti on
as edu ca ti on al

efficiency.

is given to no np r of it o r g a n i za t io ns such

inst it uti ons .

ers cited pre vio us ly
or g a n i z a t i o n s

been

Few,

if any,

of the r e s e a r c h ­

in this revie w examin ed edu ca ti on al

for op era tin g efficiency.

Until

the 1970s

a hist ory of strong support for e du ca tio n existed in
A m er ic an

society;

however,

with changes in student attitudes

re fl ect ed in the re be ll io ns of the 1960s mixed with d i m i n ­
ishing resources

in support

e f f ic ie nc y in ed uc at i on al
O rg a n i z a t i o n s

of education,

for

co ns um pt io n of resources arose.

such as the National

tion M a n ag em en t Syst ems

co ncerns

(NCHEMS)

Center

for Higher E d u c a ­

and West ern In te rst ate

33

C om m i s s i o n

for Higher Edu ca ti o n

In 1970 the staff
papers

came into existence.

of W I C H E pu bl ish ed a report

of a na tio na l res ear ch

puts of higher education.
justify e xp end it ure s
puts.

(WICHE)

inclu din g the

tra ining seminar

Pa rt ic ip an ts

on the ou t­

came together

in higher ed uca tio n in me as ur ab le o u t ­

I ro nic all y the focus of the papers

presented was

upon cost ef fi ci en ci es and more upon il l- def in ed
cost

ef fec tiveness.

more on variab les

M e as ur es

advoc ate d

tended to depend

that were d if fic ul t to define and control

Weather sb y,

Bowen and Dougl as

and Patterson,
(1971)

time

1970).

were more part icu lar

c o n s i d e ra t io n of e du c at io na l efficien ci es.

in their

They asserted

that small co lleges could real ize greater ef fic i en ci es
a ve rag e

class size with in cr ea se s in enro llm en t

the cur ric ul ar
noted,

however,

options were
that

typica lly re ali zed

less

issues of

such as added value in student earning capacity over
(Lawrence,

to

held constant.

in practice

in

so long as

They also

these econom ie s are not

because with increa se s in size,

colleges

tend not only to expand cu rr icu lu m but also to upgrade
fac i li ti es

and staff

(1971).

In a cost ana ly sis of c o ll eg e curri cul a Meeth found
that there was a high co rr ela tio n betw een avera ge
size and in st itu tio n
class

size

(1974).

He found

incre ase s in

size until enr ol lm en t of 850 were reached

On the other

hand,

Meeth

found

is not relat ed to i n st it ut io na l
Schools in his

study

spent

that

(1974,

30).

"the cost per student

size"

accor din g

class

(1974,

42-43).

to resources available

34

rather

than accord ing

student.

to an optim um eff ici en cy

Ex pen di tu re s were

in cost

per

driven by ad mi ni s t r a t i v e c o n ­

siderat io ns not ne ces sa ril y related

to cost per student

(1974).
The problem of me as u ri ng ed u ca tio nal

p ro du cti vit y was

ar ti cul at ed very clear ly by Priest and Pickelman
Citing

such a uth ori ti es as Ben Lawrence,

Nationa l Center

of the

for Higher Ed u ca ti on M an a g e m e n t Systems,

and Pat ric k Haggerty,
Texas Instruments,

former C hai rpe rs on of the Board of

Priest and P ic ke lma n a ck now le dge d

pr o du ct iv it y mea sur es in e du cat io n were
a "p rimitive

Director

in 1976.

state."

"imper fec t" and in

They proposed c on si der at ion

number of in st it uti on al cha ra ct er i st ic s

that

of a

in a simple p r o ­

ductivity

model

most part

they appealed for the de ve lop men t of sc ientific

models rather

of output and input ratios,

than offered such models.

and a mb ig uou s nature
has been mind
not detract

but for the

"The elusive

of the whole concept of pro ductivity

boggling"

(1976,

35),

but this fact should

from dealing with the problem

(1976).

Among the first more de fi ni tiv e methods of cost a n a l ­
ysis in higher
1978.

education was

He combined an activity c l as si fi ca ti on system with

program variables

defined by i ns tr uc ti on al wo rk load

mat ri ces and d ep ar tme nta l
format

that descri bed by Gamso in

personnel data forms.

pro vided a cost ana ly si s model

on subj ect iv e

that

This

depended less

judgment and more upon q ua nti ta tiv e data.

Through ma na ge me nt

inform ati on

systems and

quant it ati ve

35

data based mo dels of ma na ge me nt science,
be made with predictable

results.

McClenney

that since 1970 the Na tio na l Center
Mana ge men t

Systems p ub li she d more

on pla nning and man agement

formation
those

systems,

in Gamso.

(1980)

noted

for Higher Education

in higher education.

on cla s si fi ca ti on

could

than seventy documents

on the m eas ure me nt of educat ion al
pub li cations

decisions

Em phasis

outcomes was evident in

structures,

and cost analysis

man ag em en t i n ­

techniques

Mc Cle nne y also noted

such as

that community

co l­

leges in particular were re sp o ns iv e to NCHEMS models and
that they had taken the leadership
agement

sciences

(1980,

role in adv a nc in g m a n ­

19).

A c o m p r e h en si ve re view of higher education
studies was conducted
1985.

ef ficiency

by Paul T. Br in km an and re ported in

Br inkman examined fifty years of research

of scale

in higher education.

grate and synthesize

the results of empirical

the size-cost r el ati ons hi p
After noting

His purpose was

on economy

"to i n t e ­
studies of

in higher educ at ion "

(1985,

1).

the di ffi c ul ti es of dealing with the a m b i g u i ­

ties su rro un din g costs,
ies add res sin g

Brinkman conclu ded

that most

stu d­

the issue of eff ic ie n cy were really c o n ­

cerned with de te rm in in g if ef fi ci en ci e s vary with size.
That

is,

do e xp end it ure s

per student

creases

in size,

or in other words,

scale?

Bri nkman limited his search

ed uc at io n in the United States.
the early

1920s,

decrease with in
is there economy of
to studies of higher

He found documents from

but he indica te d that

"almost all of the

36

usable

studies" were

pro duced since 1950

(1985,

6).

Using en rollment Interv al s and per cen ta ge changes
cost,

Brinkman

vari ous

studies

grouped into
four-year

sought

to st a nd ar di ze

the results of the

for c om pa ra ti v e purposes.

three major

colleges,

categories:

in unit

The

studies were

two-year colleges,

and res ear ch un iversities.

Brinkman

found six co nc lu sio ns based on his c o mp r e h e n s i v e review;
1.

Tw o- ye ar and four-year colleges, on average,
expe rie nce positive returns to size.

do

2.

Su bs t an ti ve siz e- re la te d ec onomies are most likely
to occur at the low end of the enro llm en t range.

3.

The en ro ll men t range over wh i ch such economies
are likely to be found differs by type of i n s t i ­
tution .

4.

The extent of such econo mie s differs by function,
with the a dm i n i s t r a t i v e area typica lly e x p e r i e n c ­
ing the greatest re du cti on in unit cost and i n ­
struction the least.

5.

For ed uc at i on al and general expenditures, the
br oadest category, a three- to four-f old d i f f e r ­
ence in enro llm en t among small in s ti tu ti on s is
ac c om pan ie d by a di f fe r en ce in cost per student,
at the mean, of 25 percent for two-year i n s t i t u ­
tions and 23 perce nt for four-year institutions.

6. The extent to which s c a l e - r e l a t e d economies or
d is ec on om ie s are de mo ns tr at ed by a given set of
i ns tit ut ion s depends on variations among them in
the scope and vari ety of the pr ograms and services
they offer, the salaries they pay, and the general
di s p o s i t i o n of their res o ur ce s (1985, 28).
In general Bri nkman
ured by the number

found that ins ti tu ti on al

size as m e a s ­

of students enr ol le d is only one i n ­

fluence among a vari ety of inf lue nc es upon the cost per
student.

That

influence,

however,

found in small i ns ti tu ti o ns

of less

is more likely

to be

than 1000 enr ollment

37

than

in larger i ns ti tu ti on s
In summary,

(Brinkman,

the ap pro ach

ence of or ga ni za ti on

to i nv es ti ga ti ng

mists seek to measu re o r g a n i z at io n al
In ad dit io n

w it hi n the organization,
ma rket

entry and

O rg a n i z a t i o n a l

Cle arl y Br ink man

per student.

In this

that this app ro ac h is really
than

effic­

of def ining an output.

Size and E co nom y of Scale

(1985)

for the concept

that the concept

effec ts

studies in e du ca tio n are sev erely e n ­

cu mbe red by the prob lem

support

of resources

the con seq uen t

than not in terms of cost

Ef f i c i e n c y

organi­

e ff i c i e n c i e s in e du ca tio n

an i nv es ti g at io n of res ource a ll oc at ion rather
iency.

in a ma cro -

concerned with

to the di str ib ut i on

be argued

Ec o n o ­

On the other hand,

i nv es ti ga ti ng

i nv es ti ga t io n

latter case it might

pro duc ti v it y

they are also

within a single system and examine
more often

researchers.

to a com ple x of va ri abl es

con ce ntration.

za tio nal re se ar ch er s
tend to limit

the i n f l u ­

size on ef f ic ie nc y varies d r a m a ti ca ll y

between ec on om ist s and o r g a ni za ti on al

ec ono mic approach.

1985).

found

su bst an tia l

of economy of scale.

empirical
He also

found

of econ omy of scale itself is not a clear

cut one and i d e n t lf ed

four "s ources of amb igu it y" a s s o c ­

iated with the concept.
First,

the defi nit io n of scale

and .practice.

Clas sic de f in i ti on s

tive capacity,

but most

terms

of or g a n i z a t i o n

empir ica l

is co nf us e d in theory
relate scale to p r o d u c ­

studies

define scale in

size or quantity of output.

The

38

re l at io ns hi p betw ee n quantity of outputs and a v er ag e cost
per unit

is the focus of many

(Brinkman,

economy of scale

studies

1985).

Second,
"short -ru n"

economy of scale
versus

may influ enc e

studies may be influ en ced

"long-run"

results

behavior.

S ho rt -ru n effects

in c r o s s - s ec ti on al

studies.

other

hand short- ru n effectB might be m i n im iz ed

dinal

studies

based on performance

o rg an iz at io ns
tions

over

time.

variables

Educati on al
fluctua­

such as enrollment.

student may at first decline with enr oll me nt

On the

in l o n g i t u ­

may be pa rt ic ul a rl y prone to wide

in import ant

Costs

incre ase s

later rise with the con seq ue nt add iti on of staff and
ices.

by

per
but

se rv ­

Re su lt s of studies of ed uc at io na l o rg an i za ti on s may

be distorted
meyer,

by these flu ct ua ti on s

(Brinkman,

1985;

1982).

Third,
sistency

empiri ca l

of economy of

to what

(Brinkman,

1985).

scale lack c o n ­

is to be held co nstant when

the re l at io ns hi p between scale

a v er ag e cost"
economy

studies

"with resp ect

es ti ma ti ng

(or size)

The bro adest d e f i ni ti on

In studies of ed uc at io nal

o rg an iz at io ns

broad i nt er p ret ati on would al lo w for co mp ar is on
tions even though they dif fered
as s ist an ts

of

Fourth,

of i n s t i t u ­

to u n de rg ra du at e

st u­

1985).

economy

factors

this

in ratios such as teaching

to prof ess or s or graduate

(Brinkman,

me ntal

and

of scale allows for variances in input and output

measures.

dents

Dick-

of scale can be in fl u en ce d by e n v i r o n ­

that are r eg ula tor y in nature.

Funding

39

formulas wi thin some

states,

str ic tio ns on In st it ut io na l
control ex pe nd it ur e

for instance may impose r e ­
operations

patterns.

that,

In addition,

in fact,

rules

based

funding may result in ex pen di tur es mad e on the basis of
resourc es a va ila bl e rather
vided.

than on need

State c oo rd in at in g age ncies may also

de v el op me nt

by politi ca l or dem ogr ap hic

beyond the control

may be si g ni fic ant

(Brinkman,

studies,

cannot

The in fluence

1985).
be avoided

ent irely in h i g h ­

but they need to be acknowledged.

Bri nkm an found that in many

studies,

being asked is not about

scale or size on unit costs,
in s tit ut ion s

program

on economy of scale co n si de rat io ns

These am b ig ui ti es
er educat io n

direct

co ns ide rat io ns

of local institutions.

of re gu la tor y ag en ci e s

tion

or service pro­

spend less

however,

"the

ques­

the indepe nde nt effects
but simply whether

per student

of

large

(or per credit hour)

than do small i ns ti tu ti on s witho ut regard to in te rv eni ng
factors"

(1985,

5).

Uni for m support
not found

for a U-s hap ed ave rag e cost

by Brinkman.

In two -ye ar colleges

curve was

econo mie s of

scale are likely to be greater in small than in large
In stitutions;

however,

is for the most part

the r el at io n sh ip

linear.

of economy

In four-year

to scale

i ns tit ut ion s U-

shaped cost curves were found in large in sti tu ti on s
re sea rch er s
1985).

Most

(Jenny and Wynn,
researchers,

1970; Maynard,

however,

found

ec on om ie s realized in movin g from small

1971;

by some

Stommel,

that after initial
to large

that the

40

curve

tended

to remain flat

nating Council
Brinkman,
typical

for Higher

(Metz,

Education,

1984 and 1985).

Ca li for ni a C o o r d i ­

1969;

Carlson,

On the other hand,

finding of per pupil

su pports

1964;

school

1972,

the more

costs in high schools

the U- sha pe d avera ge cost curve (Fox,

1981).

In

his re v ie w of more than thirty economy of scale

studies

for schools and school

" es s e n t i a l ­

ly all of the studies

districts Fox concluded,
suggest

for large

size schools,

U - sh ap ed "

(1981,

that

dis economies will occur

so the average

cost curve appears

286).

The influ enc e of the larger en vir on men t of educational
o rg an iz at io ns
fied

cannot be ignored.

Bri nkman

(1985)

identi­

the influe nc e of state re gu la to ry ag encies on a v a i l ­

able re so ur ce s in public institutions.
i nv es ti ga te d the influ en ce

Tol ber t

of two th eoretical

resour ce dep end en ce and i ns ti t ut io nal iz ati on,
strati ve structure

in higher education.

(1985)

perspectives,
on a d m i n i ­

These p e r s p e c ­

tives are well gr ounded in the wo r k of Pfeffer and Sa lancik
(1978),

P er r o w

(1978),

both or ga ni za t i o n a l
fluenced

and Scott

(1981).

From this view

size and econ omy of scale may be in­

by decisi ons made in set tin gs outsi de the i n s t i t u ­

tion .

Summary

The role
by Blau
ining

of d i ff e r e n t i a t i o n

(1970) would appear

first

formally ad van ce d

to be a key factor in d e t e r m ­

the influence of size on economy of scale.

Studies

41

affirm

the fi n d i n g

stant,

greater

increases

plexity .

1971;

Economy
(1963)

r at io s

declined with

Ru shing,

1974;

Mc L a u g h l i n ,

1967;

Mos t

ference

moving

Brink ma n,

1984 and

stud ies

and Boland,
1971;

in d e s c r i b i n g

small

Ka sarda,

the a v e r a g e

Tendencies

to war d

in s t ud ie s of Haas,

Boland,

and Bo land

economies

to large

and

of scale

less

from

large

to larger

S c h o en he rr ,

1971;

Blau,

dif­

1973;

Ot he rs

have

found

suggesting

that

the a v e r a g e cost

(Jenny and Wynn,

Stommel,

com­

1985).

found g re a t e r

1985).

organizations

1981;

in most

but d i s m i s s e d

from

Hall,

Administrative

Bo la nd

and Hawley,

Blau and

is U - s h a p e d

and

studies.

in o r g a n i z a t i o n s m o v i n g
1967;

Fox,

noted

(1963)

(Rushing,

in la rge

of scale

researchers

in o r g a n i z a t i o n s

by Haas,

in size

p e rs is ts

1985).
com­

in size.

Hawley,

Blau

organizational

of the s u p p o r t i v e

B r i nk ma n,

inconsistency

and J o h n s o n

Brink man ,

H o l d a w a y and Blo wers,

1979;

c u r v i l i n e a r i t y we re

curve

1961;

1970 and 1973;

1979;

found

in c r e a s e s

cu r ve in e co nom y

(1965).

was

from in c r e a s e s

1965;

Hall,

M cL a u g h l i n ,

and Warkov,

Some

(Blau,

in the r e d u c t i o n

resulting

( A n de rs on

size

of scale

is held c o n ­

scale are to be r e a l i z e d with

are a s s o c i a t e d wit h

po n en t

cost

of

in o r g a n i z a t i o n

f i nd in gs

Johnson

if d i f f e r e n t i a t i o n

economies

and Sc h o e n h e r r ,
Si m i l a r

that

1985).

1970;

F i n d i n g s on

diseconomies

Ma yn ard ,

1971;

this i s s u e remain

unclear.
Environmental
e co nom y

of scale

influences

are

on o r g a n i z a t i o n a l

considered

in mo r e

re cent

size

and

studies.

42

D i s t o r t i o n s may
sca le

result

in o r g a n i z a t i o n s

Educational
instanc e,
to scale
tional
ati on

in c o n c l u s i o n s
influenced

organizations

m a y ex pen d
(Brinkman,

n e t w o r k s may
(P feffer and

1985).
also

funding

according

Political

i nf l u e n c e

Salan cik ,

e co n o m y

of

by r e g u l a t o r y ag enc ie s.

in fo r m u l a

resources

regarding

states,

to rules

unrelated

c o n t e x t s and o r g a n i z a ­

organizational

1978;

for

Perrow,

1979;

configur­
Scott,

1981).
Thes e
The

su m m a r y c o m m e n t s are

study of o r g a n i z a t i o n a l

size and

e co n o m y

of scale is c o m p l i c a t e d

in this

review.

Methodology

ers who op e r a t e

primarily

than a t h e o r e t i c a l
researchers

one

(1970)

employment

agencies

(1973).
fied

(1971)

s tu d i e s

Consequently,
to study

fa c tor s

va r i e s w i d e l y

a mo ng

1976).

base

Amon g

on

identified
research­
ra th er
the ma ny

Blau e s t a b l i s h e d a t h e o r e t i c a l

w h i c h wa s

in terms r e le va n t

gated.
study

In most

by ma n y

(Kimberly,

general.

its i n f l u e n c e

from an e m p i r i c a l

r e v i e w e d only

foundation

by n e c e s s i t y very

then te st ed
and

c ol l e g e s

in s tu di es
and

universities

the v a r i a b l e s e x a m i n e d

to a p a r t i c u l a r
the s y n t h e s i s

is very limited.

of

are s p e c i ­

problem investi­

of g e n e r a l i z a t i o n s from

C H AP TE R III

M E T H O DO LO GY

The Po pulation

In this study of o r g a n i za ti on al
scale

size and economy

the researcher has chose n to work within

imposed by a c r o ss -s ec ti on al
type and has

selected

A co mplete

to anal yze the Mic hi gan

list of Mic hi gan

colleges and their size
f u l l - y e a r -e qu at ed

commun it y

public commun it y
public

in fu ll -t im e - e q u a t e d

student

the limits

study of a single org an iz at io n

college system inclu siv e of t we nt y- ni ne
colleges.

of

commun it y
person nel and

enr oll me nt may be found

in

Ap pen dix A.

Data

The primary
of Education

Source

source of data is the Mi ch i ga n De p ar tm en t

that gathers

i nfo rma ti on

commu nit y colleges as dete rmi ned

by the Mi chi gan Com mu ni ty

Co lle ges Ac t iv it ie s Cl as s i f i c a t i o n
and

the Manual

been selected as the period
the most

recent

The use of 1984 data
of state officials,
greater

S t r u c t u r e . 1981

(ACS)

for U n if or m Fi na nci al R e p o r t i n g : Michiga n

Public C om mun it y C o l l e g e s . 1 9 8 1 .

sents

from all Mic hi gan

Fiscal year

1984 has

for an aly si s since this r e p r e ­

data ava il ab l e

through

is justified in that,

this

source.

in the judgment

the val id ity of unif orm r ep or tin g

in this coll ect io n

period
43

than

in any

previous

is

44

period
data

(Root,

1985).

encourages

In ad dit io n the high quality of the

r ep li ca tio n

studies

in the future

that will

use 1984 as the base year.

Variab le s and Ind icators

O r ga ni za ti on al
two di stinctly

size in this study is inv es tigated

different

contexts.

The first

is more con sis ten t with org a ni za ti on al

in

examination

research

that de­

pends on st ructural c o n s i d e ra ti on s such as the number

of

employ ees and the d i f f er en ti at io n or complexity of op er a­
tions.

The second is more con sistent with economic

terests in opera tin g

costs and eff iciencies.

in­

A descrip­

tion of the varia ble s examin ed in this study follows.
Ad d it io na l

operati ona l

de fi nit io ns and in dicators may be

found in App en di x B.
In the first
are di st rib ute d

context,

in a five

execu ti ve administra tor s,
clerical

Btaff,

service

is related

f ul l- ti me -e qu at ed

part division
a d m i n i st ra ti ve

other support

ulty group includes all
whose

total

staff,

personnel,

of labor including
professionals,

and faculty.

to the provision of direct

librarians and cou nse lor s are sometimes

the faculty category,

The

in­

basis.

included

the a d mi ni st ra ti ve

group along with all other a dm i n i s t r a t i v e pro fe ssi ona ls
serve

in

in this study they are included in

the ca tegory of pro fe s si on al s within

who do not

fac­

full and part-time,

struction and is com p ut ed on a f u l l -t im e- e qu at ed
Wh i le

employees

in an executive

capacity.

45

Placing

librarians and counselors

tive profess ion al s

group

in the a d m i n i s t r a ­

frees co ns id er at i on s

faculty ratios from di st or ti on s

produced

of student-

by inclusion of

personnel not f unc tio ni ng within the cl as sro om setting.
S t u d e n t -f ac ui ty
part-time

ratios are based on the

faculty as m e as ur ed

pool of full and

by a ful l-t i me -e qu at ed

formula.
The execut iv e ad mi ni str at ors
a count of pe rsonnel

listed

cation Di re cto ry with

Life Officer,

is det ermined by

in the HEP 1985 Higher

Edu ­

the following title or function:

Chief Execu tiv e Officer,
Aca demic Officer,

group

Executive Vice President,

Chief Business Officer,

and Deans.

These

cons is ten t with those d es cr ibe d

Chief

Chief of Student

titles or functions are
by k_ Manual

for Budgeting

and Acc oun tin g Ma npo we r Resour ce s in Pos ts e co nd a ry
E d u c a t i o n , Techni ca l Report Number
P. Jones of the National Center

84

(pp.

for Education

addi ti ona l

executive

by Dennis

for Higher Education

Mana gement Systems and The odore H. Drews
Center

47-54)

of the National

Sta tistics and are inclu siv e of
titles such as chancellor,

president,

and provost.
The remaining
staff

groups

and the clerical

to staff personnel
services,

staff.

The support staff is limited

in maintenance,

and other

in other groups.

to be defined are the support

support

security,

food

functions not clearly ide ntified

The support

services might otherwise

custodial,

group includes

be co ntracted

personnel whose

rather than employed.

Functio ns such as snow removal

or rem ode lin g of facilities

are examples of personnel ac tivity in cluded in this group.
The

clerical group,

on the other hand,

based personnel whose
ad m i n i st ra t or s
taries,

performance

or a dm in is tr at iv e

receptionists,

and switchb oa rd

includes office

is in support of executive
professionals.

bookkeepers,

Secre­

data entry operators,

operators are jobs typical of the clerical

group.
In addition
economy of scale

to structural co ns i de ra ti on s

involves co ns ide ra tio ns of efficiency.

Eff i ci en cy is generally
in rel ation

the issue of

defined as the analysis of costs

to output(s).

In higher education eff iciency

mea su res

typically include c on si de ra tio ns of costs

student,

costs

per square foot,

member

( M e e t h , 1974).

ponent

in relation

plored

by many researc her s

Hall & Johnson,
1971;

Blau,

1973).

The size of the a dm in is t r a t i v e c om­

(Terrien & Mills,

Rushing,

1966;

students

Perso nne l variab le s included
personnel,

professionals,
sonnel,

Haas,

Ho ld a wa y & Blowers,

(FYES)

executive administration,

personnel.

is limited to personnel
on official ACS state

faculty,

total

a dm in i st ra ti ve
clerical

per­

Each of these variables

as described

records.

size

to 15,831 F Y E S .

in this study are;

fu ll- ti me -e qu i va le nt

and support

1955;

colleges varying in enro llm en t

753 fu ll -y e ar -e qu at ed

college

size is an issue e x ­

In this study eff ici en cy measures are

ex amined in commun ity
from

and costs per faculty

to organ iza ti on

1963;

per

above and as reported

k1
Ex pen dit ur e variab les
personnel
records

included

in this study are

total

e xp end itu re s for each college as reported in ACS

for ins tr u ct io n al

tutional admi ni str ati on ,
varia ble s are

pooled

support,

student

and instruction.

in order to examine

tional e xp end it ure s in contrast

services,

inst i­

In addition,

some

total noninst ruc -

to instruc tio na l e x p e n d i ­

tures.
Add iti ona l

variables

di ff er en ti at io n

include me asures of curricular

as indicated

by the number of different

courses and course sections available.

Ef f ic ie nc y me asures

such as stu de nt- fa cu it y ratios and average

class

size are

includ ed as well as resource dependence variables
property
tuition

tax revenue,
and

state ap pro pr ia t io n revenue,

fee revenue.

Finally physical

including gross square feet and gross
examined

in comparison

size is measured
budget,

such as

to enrollment

plant

student

variables

cubic feet are
size.

Org anization

in terms of total personnel,

and fu l l- ye ar -e qu at ed

and

the operating

enrollment.

H yp ot hes es and Rationale

Based on the concepts re viewed in Chapter II seven
hy potheses are tested.

H yp ot hes is I:

a.

The proportion

of executive a d m i n i s t r a ­

tors decreases

in midsize and

large

colleges and is highest in small
leges .

col­

48

b.

The pro por ti on of the remai nin g a d m i n i ­
strative co mpo ne nt increases
tionately

from small

propor­

to midsi ze

to

large colleges.
c.

The

pr oportion

constant
d.

The

of faculty remains

reg ar dle ss of college size.

proportion

of clerical

remains constant
colleges

personnel

in small and midsize

but increases

in large c o l ­

leges.
e.

The proportion

of support

remains co nstant

personnel

reg ardless

of college

size.

Co r re la ti on

ana ly sis may show pro po r ti on a te

d i st ri bu ti on s accord in g

to size.

of economy of scale in schools,
port compon ent
a co nse qu enc e

includes

sions

In much of the literature
the a d mi ni st ra ti v e or sup­

all n ont eac hi ng

of different

the no nt ea ch ing component.

Conclu­

regarding incre ase s in the a dm i n i s t r a t i v e component

may be distor te d
of labor.
quently

for failure

Argume nts

to provide

based on ass um pti ons

from this assu mpt io n

addit io nal

for this division

for economy of scale are also

no nte ach in g

fre­

that limit the adm in i st ra ti v e

compon en t to execu tiv e officers only,
result

personnel and as

does not provide for comparisons

employee groups within

personnel

for failure

A distortion may
to recognize

profess ion al s may swell

that

49

ad m i n i s t r a t i v e

ranks while the number of execut iv e

rema in s re la ti vel y constant.
person ne l from clerical
resting vari at ion s
no n cl er ic al

Sepa rat ion of the support

personnel may

provide

since many of the services

some in t e­
provided

support perso nne l may be contr ac ted

outside age ncies

rather than

officers

by employment.

by

through

Election

of

this option may be a factor of size.

H y p ot he si s

II:

a.

In st itu tio na l ad mi n i s t r a t i v e e x p e n d i ­
tures

decrease as college

size in­

creases.
b.

In str uc t io na l

support and student

services ex pe nd it ur es

increase with

incre ase s in colle ge enrollment,
c.

The ex p en dit ure per student
ad m i n i s t r a t i v e
three

components

enr ol lme nt

T e st in g
penditures

including

in relation

colleges.

to c ol leg e size,

e xp e nd it ur es

required

by functi on s within

for college ad mi n i s t r a t i o n

cost r eg ard les s of size,
de c li ni ng

cost as size increases.

test also examines
services

and therefore,

the premise

that

On

ex­

but division of

provides support for the premise that

agem en t

these

is greater in large

this concept examin es a dm i n i s t r a t i v e

a d mi n i s t r a t i v e
ponent

cost

for total

this co m ­

the m a n ­
is a uniform

it will

likely be a

the other hand,

pro fessional

this

support

increase with in creases in college size at a rate

50

great enough to result

Hy p ot he si s

III:

In lost economy

Ex pen d it ur es
student

of scale.

per fu ll - y e a r - e q u a t e d

(FYES)

are greater

than small colleges,

in large

but the Increase

is depen den t upon n on in s t r u c t i o n a l
ex pe ndi tu re

increases rather

than f a c ­

ulty expenditures.

If one of the compon en t
account s

be indicated.

since

it is suspec te d that i n s t r u c t i o n ­

services more than i n s ti tu ti on al a d ­

m in i s t r a t i v e expenses will account
per student

IV:

increase

A finding of this sort would be par ­

al support and student

H yp ot hes is

than any other

an influence ac co u nt in g for ex pen di tur e

ticula rly important

tures

cons ist en tly

for d is p r o p o r t i o n a t e va riance more

variable,
might

variables

for

increased e x p e n d i ­

in large colleges.

The s t u d e n t -f ac ui ty ratio is independent
of college size when course

dif fe re nt i at io n

and number of sections are taken into
consideration.

S tu d e n t - f a c u i t y ratios are
ined

by college

in s ti tu tio nal

size and more

philosophy.

less likely

likely to be Inf lue nc ed

the wi d es pr ea d

by

C o mm un i ty col leges in particular

have claimed a dv ant age s atten dan t
Still another major

to be d e t e r m ­

influence

to small class

on s t u de nt -f ac ui ty

size.
ratios is

presence of coll ect iv e barg ai nin g in the

51

State of Michigan.

Hy p ot he si s V:

a.

Avera ge class size remains
college enro llm en t

b.

increases.

Increases in course

di ffe re nt ia ti on

the number of mu lti pl e

c.

Und up li ca te d

increases

headcount

student he adc ou nt

differentiation,

ficiency in economy of scale.

to be a measure

If course

increases in aver age

in economy of scale,

however,

to be linked

to un dup lic at ed

declines.

Aver ag e class size is assumed

size are likely

sections offered
size.

but FYES in proportion

result

d i f fe re nt ia t io n may also

class size

in course d i f ­
Increased

increase headco un t without

FYES may result

dent services expend it ure s

Hy p ot hes is VI:

I nc rea se d he ad cou nt but

in declining ef fi ci en cy

in st u­

due to increases in costs of

storage,

and retrieval.

Adm in i st ra ti ve

e xp end it ure s

increase as ava il ab le

per FYES

re sources increase

in de pen de nt of college en ro llm en t
creases.

could

incre ase s in college

to increases

pr o po rti ona te increases in FYES.

record maintenance,

of e f ­

dif fer en ti at io n

fe re ntiation with no gain in economy of scale.

de cl ini ng

and

effects con stant average class

with incre ase d course

remained constant,

uncha nge d as

in­

52

It Is an t ic ip at ed

that services beyond direct i n s t r u c ­

tion are dependent upon availa bl e colle ge resources more
than the size of the en rollment

served.

firmation of this e xpe cta ti on would
an eq ua liz at ion
dent

Findings in con­

suggest the need for

factor in the state ap pr op ria tio n

populations

if stu­

are to be served uniformly or according

to need.

Hyp othesis VII:

M ain te nan ce and energy expen di tur es
are

functions of physical

plant size

i nd epe nd ent of enr oll men t and personnel
size.

In the design of the physical
fecting opera tin g ef fi ci en ci es

plant,

decisions a f ­

are set in place as fixed

costs of operation i nde pe nde nt of enrollment.
volume of use eff ected

by personnel and students may have

some influence on custodial
is negl ig ibl e in relation
college plant.

co ns id er at io ns
pa rt icularly

expenditures,

to physical

this influence

features of the

Economy of scale c o nsi der at ion s

plant are likely

While the

to have

in physical

been ov e rr id de n by aesthetic

in the development of the physical

plant

since most commun it y college c on str uc tio n was

comple te d in a less energy co nsc io us era.

St at ist ic al Analysis

The basic

design

of this study is to investi gat e the

53

relat io nsh ip of stru ctu ra l c h a ra ct er is ti cs
of varying

size.

A central

in orga ni zat io ns

concern in volves

the i n v e s t i g a ­

tion of the r el ati on shi p between o rg an iz at io n size and
economy of scale if,

in fact,

such a r el ati on shi p exists.

While hyp oth es es have been stated

in directional

based on theore tic al concepts,

the nature

more e xpl or ato ry

toward

assumptions.

than disposed

St ati st ic al

explore questions

terms

of this study is

the de mo n s t r a t i o n of

tests have been designed

to

such as:

What changes in the co mp os it ion of personnel
comp on ent s are effected in commu nit y colleges of
varying sizes?
What changes in ad mi n i s t r a t i v e ex pen di tur es are
eff ected in commu nit y colleges of varyi ng size?
Will exp en di tu re s per student be greater
colleges than in large colleges?

in small

Is the s t u de nt - fa cu it y ratio inf lue nc ed more by
college enr oll men t or by program di ffe ren ti a ti on ?
What is the influence of coll ege enro ll men t and
program di ff e r e n t i a t i o n on average class size?
What is the re lat io nsh ip between av ai la bl e r e ­
sources and ad m i n i s t r a t i v e ex pen dit ur es per
student ?
What is the r el at ion shi p between fixed costs of
physical plant operat ion and the number of physical
plant users?
Answers

to these

questions depend on a^ priori

of the rel at i on s hi ps

of variables.

predictive models can be e ffe ct ive ly
lation and mu lti pl e
Pearson

judgments

D es cr ip ti ve

as well as

tested through c o r r e ­

regression analysis.

product moment co rr ela tio n c oe ffi cie nt s were

54

calcu la ted

to examine

the r el at io ns hi ps

exi sting

the size of person ne l comp on ent s and or ga n iz at io n
M u lt ip le

regre ss ion

ana lysis with stepwise

between
size.

selec tio n of

in d ep en de nt

varia ble s was used to i nve sti ga te r e l a t i o n ­

ships among

struc tu ral

colleges

of varying

S ta tis ti cal

c h a ra ct e ri st ic s in community

size.

tests were

com pu ted using

Package

for the Social

Sc iences

1984).

The research is post hoc

the St ati s ti ca l

for the IBM PC / XT

in design and is free of

pr oblems of miss ing data or es timated data.
selection

of in de pe nd en t variables

an alysis was elected
influence

in order

of in de pe n de nt

variab le s

on the dependent

provided

that summarize

by step

and no ns ta nd a r d i z e d

reg re ss i on

and the constant.

selected,

the

coefficients,

the co eff ic ien t

The defa ult

of

limits of

for entry or p<.10 for rem ova l of in de pen den t

variab le s
order

T abl es are

the variable

the sta ndard error of the estimate,

p<.05

re gr es sio n

to de te rmi ne the relativ e

in the order of their influence.

determination,

The stepwise

in mul tip le

variable

stand ar diz ed

(Norusis,

in the regr ess io n equ ati on have been accepted in

to excl ude

meeting

from the equ ation

these levels

It is assumed

of significance.

that as su m pt io ns

normal it y may be violated
this study.

The

universe of o rg an iz at io ns

of l i ne a ri ty and

due to small

population

sequences of these

those var ia b le s not

does,

popu lat io n

however,

under study,

and

size in

comprise
therefore

the
co n­

violations may not be severe if the

55

resu lt s are

limited to de s cr ip ti ve rather than inferen ti al

conclusions.
re ga rdi ng
however,

the se rio us nes s

in sta ti sti ca l

of violat in g

at least some re s ea r ch er s

argue that
fluenced

D is a g r e e m e n t exists

"pa rameter estim ate s

by vio lations

regression assumptions,

(Kerlinger and Pedhazar)

are not m ea ni ng fu ll y

of as su mp ti o ns "

(Lewis-Beck,

In view of these con cerns regar din g
as s um pt io ns

u nd er ly ing mu lti pl e

tailed format

for re po rti ng

above has been selected.

regre ssi on

ables on the d ep en den t variable.

regar din g

on the variin the
vari­

In some cases the in­

to a simple

variable of s ign ifi ca nt

the d e ­

the spe­

of indepe nde nt

t e rpr et ati on may be highly descriptive,

single

30).

tests described

O b s e r va ti on s

of the influence

the results may be limited

1980,

re gr ession analysis,

in the eq uat io n may be important

final i n t er pr e ta ti on

in­

the data and the

cific effects of the entry of a new variable
lables alre ady

li terature

while in others
id en ti fi c a t i o n of a

influ enc e about which little

else might be said.
The inclu sio n of the n o n s ta n da rd iz ed
efficient,

B,

i nde pen de nt
metric

permits m ea s u r e m e n t of the Influ enc e of the

variable on the dependent

terms of the variables.

coefficient,

Beta,

of the ind epe nd ent
st andard
SE/B,

score

allows

variable

regression

for rec og ni ti on of the influence

va ria ble on the d e pen den t va ri abl e in

form.

regression

in the

The st an dar di zed

The sta ndard error of the estimate,

is a useful measu re of sig ni f ic an ce

ardized

regr es sio n c o ­

coefficient.

of the n o n s t a n d ­

If the n o n st an d ar di ze d

56

r eg res sio n c oe ff ic ie nt
standard

error

is greater

three

times the

we can be 95% certain

of the estimate,

a value of zero

than

that

lie withi n the 95% confidence

does not

interval and the reg res sio n c oe ff ici en t is significant;
slope est im ate is s i gn if ic an tl y different

P<.05 level.
garding

The con stant

si d er at io ns

negative

consta nts may

ope r at io n or overhead.
na ti on R^,

Finally,

is an i mpo rta nt

ance

in the depen den t

dent

variable.

ex am ine d

signal

in for ma tio n re­
In economic c o n ­
built

in costs of

the co eff i ci en t

indicator

of d e t e r m i ­

of the amount of v a r i ­
for by the indpen-

in each reg res si on model were

first

for high i n t e r c o r r e l a t i o n s between indepe nde nt

variables.
culated,

value.

var iable ac co unt ed

Var i ab le s inc lud ed

from zero at the

provid es useful

the trend of the intercept

the

and

Pearson product moment

c or rel at ion s were c a l ­

to reduce the potential

for m ul ti co ll i n e a r i t y

re g re ss io n models were re d es ig ne d when int er co r re la ti on s
above r = .70 were found.
in de pen de nt
models
ables

of highly correlated

va riables were tested in alter nat e

in order

to inv es ti ga te

in different

plots and mea s ur es

of central

regression

the influence of the v a r i ­

configura ti ons .

vi o la ti on s of as su mp ti on s
analysis.

Subsets

In addition,

sc atter-

ten dency were examin ed

for

un derlying mul tip le regression

C H AP TE R IV

E M PI RI CA L FINDINGS

This chapter

presents

of seven h y p ot he se s
scale in Mic hig an

the results

of st ati sti ca l

regar din g effi cie nc y and economy

tests
of

public co mmu ni ty colleges.

H yp ot hes is I

Re s ea rc he rs

have shown that

stru ct ura l characteristic,

Blau,

1973;

a

varies in relation to o r g a n i z a ­

tion size (Terrien and Mills,
1967;

the division of labor,

1955;

and Mc Laughlin,

Anderson and Warkov,

1979).

In this study

c o mmu ni ty c ol leg e personnel were ass ig ned to one of five
groups:

execu tiv e ad mi ni str at ors ,

sionals,

faculty,

clerical,

part-t im e employees,

a d mi ni st ra ti ve

and support.

full-t ime

profes­

In the case of

eq uiv al ent s were

determined.

In the case of employees with re s po ns ib il i ti es in more than
one group,

their time was alloc ate d

full-t im e equivalents.

The number

in fractional values
in each

of

group were e x ­

amined in a c orr ela ti on matrix agai ns t the ful l-y ea r- e qu at ed
student

enr oll men t as a measu re of college size.

sults of this test are report ed

The re­

in Table 3.

The results in dicate ten positiv e c or re lat ion s at the
p<.001

level

p<.01 level.
(r=.96)
and,

of si gni fi can ce and one co r re la t io n at the
The re lat io ns hi p between enro ll men t and faculty

is the highest

as expected,

co rre la tio n related

to enrollment,

the execut iv e adm in is t ra to rs
57

group

(re=.40)

58

TAB LE 3:

C o r r el at io n C o e f f ic ie nt s Rel a ti ng the Div ision
of Labor by College Size in Enr ollment

Mean

Variables

VI:

College Enro llm en t in Full
Year Equat ed Students

(n=29)

St.Dev.

4172

4285
6.1

1.9

58.4

60.6

216.6

180.4

Clerical

70.0

78.3

Support

29.3

26.9

V2:

Executive A dm in ist ra tor s

V3:

A d mi ni st ra ti ve P ro fe ss io na l s

V4:

Faculty

V5:
V6:

Zero -order C or re lat ion s

VI
V2

V2

V3

.40

.91**
.41

V3
V4

V5

V6

.96**

.84**

.90**

.46*

.38

.27

.91**

.90**

.74**

.87**

.86**

V4 c.

.72**

V5

* p < .01;

( P e a r s o n 1s r )

**p<.001

59

is not si gni fi ca nt ly related to enrollment.
execu ti ve a d mi ni st ra to rs
any other

group except

are not si gni fi ca nt ly

faculty

ship is weak in comparison
clerical and support
other

(r=.46),

and

related

this r e l a t i o n ­

groups are highly

corre lat ed

groups except execut ive administr at ors .

groups

but executive adm inistrators.

s i gn ifi ca ntl y corre lat ed
The primary

concern

influe nc e of enrollment

of employee

further

in order

between

these

highly

indicate

however,

groups.

that only

is the

While a number

in the matrix,

groups

the co r­

were examined

if significant

correlated

all

the faculty are

groups and enrollment

to determine

Admini­

groups.

in this study,
on employee

The

to all

cor related with
Only

to all other

of high in t er c or re la ti on s are found
relations

to

to other Intercor rel ati on s.

st rative p ro fe ssi on als are also highly

would

As a group

differe nce s

existed.

The ma tr ix

the executive a d m i n i st r at or s were

not highly corr el ate d with enrollment.
Di ff ere nc es
coeff ic ien ts

between two population

using dependent

samples can

correla ti on
be tested for

signif ic anc e using a t test where:
t
xy r xz r yz^
The

underlying

distribution

with n-3 degrees
227).

of freedom

In order to increase

is the students
(Hinkle,

t-dist rlb ut ion

Wiersma,

Jurs,

1979,

the robustness of this test,

the numer ato r value was modified

to include an n-5 factor

60

in place of n-3
again st

to reflect

the enro ll men t

the inc lu si on

variable.

value was se lected at the

of five groups

In ad dit ion

.01 level.

the critical

The results

of this

test are re por te d in Table 4,

TABLE 4:

T Tests for Si gn if ic ant D if f e r e n c e of
Co r re la ti on Co ef fi ci en t s in D ep en den t Sample
of Group s Com p ar ed Agai nst F ul l- Ye a r - E q u a t e d
Students.

Va ri abl es

t

Com par ed r

Facu lt y and A d m i n i st ra ti ve
P r of e s s i o n a l s

.96 -

.91

2.106

Facul ty and Support

.96 -

.90

2.318

Facu lt y and Clerical

.96 -

.84

4.123*

Ad m i n i s t r a t i v e P r o f e s si on al s
and Cle rical

.91 -

.84

1.870

Support and Cle ri ca l

.90 - .84

1.163

Cl er ica l and Execut iv e
Ad m in is t r a t o r s

.84 -

3.617*

*p<.01;

critical value = 2.797

The results
was

indicate

found between

faculty,

that no signif ica nt

professionals,

to enrollment.

clerical

the a d m i n i s t r a t i v e

groups.

was in de pe nd e nt

Only

of the

and the support

A si gn i fi ca nt

found b et we en the faculty and the clerical

not between

difference

the co rre l at io n co ef fi ci en ts

the ad m i n i s t r a t i v e

group in re lation
was

.40

pro fess ionals,

difference
group but

support,

the ex ecu tiv e ad m i n i s t r a t o r

of all other

groups.

These

and

group

rel ati on shi ps

61

are presented
The

in Fi gur e

re la tio ns hip s

1.
depicted in Figure

wi t h in cre as es in enrollment*
increase.

Figure 1:

A D M .P
t
SUPPORT

support

group

R
O
/
C L E RI CA L < --- /---> EXEC. ADM.

R ep r e s e n t a t i o n of S ig ni fic an t Re l a t i o n s h i p s B e ­
tween Emp lo yee Groups.

personnel will

increase witho ut

along with faculty;

however*

groups and the clerical
increase may be held

likely

faculty will most certa inl y

A d m i n i s t r a t i v e p r o f e s si on al s and

FAC^ L

That is,

1 suggest that

group

be enc um be re d
Increases.

is least likely

suggests

that their rate of

by cl eri ca l demands.

professionals,

to require clerical

to clerical

the link between these two

back sl igh tly

a dm i n i s t r a t i v e

number may

si gn ifi can t dif fer enc e

in particular,

support and inc re as es

are

in their

by economic co n st ra in ts a tt end an t
The exe c ut iv e a d m i n i st ra to r group

to in cre ase with in cre as es in enrollment.

They are

in dep en den t of other p er so nne l groups and only

sli ghtly

correlated with enrollment.

Ex a mi na ti on

of sc at te rp lo ts

failed

tions of the as su mpt io n of linearity.
r e la ti on sh ip s were evident;
leges over 6000
mon.

Some

fu ll -y ea r- eq ua te d

tendency

students.

Def in ite linear
out li ers among c o l ­

students were not u n c o m ­

toward cu rv i l i n e a r i t y was ob se r ve d in

the plot of a dm in is tr at iv e
equa te d

however,

to reveal v i o l a ­

The line

p ro fe ss ion al s with fu ll-yearpeaked at 11,250 FYES and fell

62

slightly with two cases above 13,50 0 FYES.
beyond

general

to small

observation,

popu la tio n

however,

is in co nc lu s iv e due

size.

Hy po th es is

E ff ic ie nc ie s

II

in a dm i n i s t r a t i v e e xp en di t ur es may be

gained with incre ase s in or ga ni za ti on size
Warkov,

1961;

Hawley,

H o ld aw ay and Blowes,
1985).

Int er pr et at i on

Boland and Boland,
1971;

On the other hand,

Kasarda,

(Anderson

1965;

1974;

and

Rushing,

1967;

and Brinkman,

incre ase s in or ga n iz at io na l

c o m p le xi ty and d i f f er en ti at io n may have a neg ati ve effect
on economy of scale
Schoenherr,
The

1971;

(Blau,

1970 and 1973;

McLaughlin,

second hy po th es is

1979;

in this

Blau and

and Brinkman,

study was designe d

1985).
to measure

the influ enc e of ex pe nd i tu re s in a dm i n i s t r a t i v e co mponents
on the a dm i n i s t r a t i v e
A match

costs

per student.

between perso nn el

varia ble s in hy po the si s

(e.g.

a dm i n i s t r a t i v e FTE co mponents)

ables

in hy pothesis

and ex pe nd it ur e va r i ­

II was not possible

of the Act iv it ie s Cl as s i f i c a t i o n

due to the format

Struc tu re

(ACS)

data.

exe c ut iv e ad mi n i s t r a t i v e

compon en t in hyp oth es is

execut iv e a d m i n i s t r a t o r s

from i n st ru ct io na l support,

dent se rvices
however,

support,

and in sti tu ti on al

salary e xp en di tu re s

I

I includes
stu­

admi nis tra ti on;

for the ex e cut ive s in each of

these areas are pooled in ACS reports with all other
administrative
Therefore,

perso nne l

The

in the same

service

exe c ut iv e salary i nfo rm ati on

component.

exclus ive of all

63

other a dm i n i s t r a t i v e

salaries

is not available.

This pro b­

lem will be ad dre sse d further in Chapter V.
With these qu ali fi cat ion s
tested.

Ins ti tut io nal

the follo win g hypo the ses were

a d mi ni st ra ti ve

as college enro llm ent increases.
includes,

among

others,

personnel component

Ins tr u c ti on al

These a d m i n i st ra ti ve comp one nts

tive in st ru ct ion al and student
the ex p e nd it ur e
inclu din g

life officers.

these

three components,

includ ing a dm in is tr a ti ve

in str uc ti on al

Finally,

is greater in

colleges.

To test these hypotheses a multiple
designed

include e x e c u ­

per student for total adm in is tr at iv e

large enr oll men t

was

support and

service ex pe nd itu res increase with increases in

enrollment.

costs,

support,

regr es sio n model

ex pen dit ur es

student services,

total a d m i n i s t ra ti ve

expenditure

depen den t variable.

Also included as me asures

s tit uti on al

in Table

variables

however,

of
faculty

due to the high i nt er ­

found among the I nde pen de nt va ria bl es e v i ­
5.

The results found
reasons.

of in­

students and fu l l- ti me -e qu at e d

This model was rejected,
co rre la ti on s

and the

per student as the

size were the indepe nde nt

fu l l- ye ar -e q ua te d

for

and i n s t i t u ­

tional a d mi n i s t r a t o r s as i nde pe nde nt variables,

dent

decrease

the chief exe c ut iv e officer and

the chief business officer.
student

This

ex pen ditures

Wh i le

in Table 5 are important

for two

high in te rc or re la ti on s among the in de ­

pendent varia ble s are evident,

very low co rre lations

64

TABLE 5;

C o r r e la ti on Co ef fi ci en ts Re lat in g the A d m i n i ­
strative E xp en dit ur es by Component, College
Size in En r ol lm e nt and Employees, and A d m i n i ­
st rative E xpe nd itu re per Student.

Varia bl es

(n= 29)

Mean

St.Dev.

De p en de nt Variable:
VI:

1,066.4

Ad m i n i s t r a t i v e E x p e n d i ­
ture per FYES

253.8

In de pen de nt Variables:
V2:

In st ru ct io na l Support

1,531,376

1,624,145

V3:

Stude nt Ser vices

1,3 59,325

1,252,198

V4:

Ins ti tu ti on al
stration

1,393,477

1,406,794

4,285

4,172

V5:

V6:

Admini­

Full Year Equated
Students (FYES)
Full Time Equated
Employees (FTE)

Zer o- ord er

VI
V2

Cor rel at i on s

180.4

(Pearson* s r )

V2

V3

V4

V5

V6

.03

-.18

.06

-.28

-.26

.85**

V3
V4
V5
* p < .01;

216. 6

.92**

.86**

.88**

.88**

.98**

.95**

.85**

.83**
.96**

**p<.001

65

with the intended

depend ent var iable are

neg ati ve co rr el at io ns
inverse

present,

in the dependent var iable

but three

suggest an

re lat io ns hi p be tw ee n en ro ll men t and ex pen di tu re s

con sis ten t with

the hypothesis.

Second,

due to the high

in t e r c o r r e l a t i o n s betw een the i nde pe nde nt variables,
mu lti ple re gr es si o n an aly si s would be too strongly i n f l u ­
enced by m u l t i c o l l i n e a r i t y

to provide

conc lus iv e support

for the hypothesis.
To adjust

for these pr oblems

of poten ti al m u l t i c o l l i n ­

earity a second mul tip le reg r es si on model was built with
a dm i n i s t r a t i v e
student.

ex pe ndi tur es

In ad dit ion

to express

e x pre sse d as e xp end itu re s

per

the d e p e n d e n t va ria bl e was adj usted

total operat in g

ex pe nd it ur es per student.

Cal­

culating

these v ar ia ble s as functi ons of fu ll- y ea r- eq ua te d

students

pro vided

some adju stm en t

served m ul ti co ll i ne ar it y.

for the previously o b ­

Co r r e l a t i o n s re su l ti ng from

these ad j us tm en ts are indic ate d in Table
The results ind i ca te d in Ta b le
exami nin g

ind epe nd ent

variables

6.

6 provided a basis for

expres se d as functions

per student costs in a m u lt ip le reg res si on model.
ar dizing

the ind epe nd ent

the chances

ity in the reg res sio n model.

enrollment,
The

Ne gat ive

the enro llm en t var i ab le

rating ex pe n di tu re s

Stand­

var i ab le s reduced the cor re la ti on s

to a level that dimi nis he d

ated with

of

for m u l t i c o l l i n e a r ­
co rre l at io ns a s s o c i ­

(FYES)

suggest that op e ­

per student decline with increases

an ob se rva ti on

in

in support of economy of scale.

si gn ifi can t co rre la tio n

of r= -.50 between

66

TABLE 6:

Cor re lat io n Co ef fic ie nts of Ad mi ni str at ive
Compon en t Exp en di tu re s per Student (FYES), C o l ­
lege Size in Enrollment, and Total Operat ing
Ex pe ndi tu re per Student.

Variables

De p en de nt
VI:

Mean

(n= 29)

S t .Dev.

Variable:

Opera ti ng Expenditure
per Student (OE/FYES)

3,155

437

361

115

Student Services
(SS/FYES)

341

94

In sti tu ti on al A d m i n i s t r a ­
tion (I A / F Y E S )

364

112

4,285

4,172

Indep endent Variable:
V2:

V3:

V4:

V5:

Inst ru cti on al
(IS/FYES)

Support

Full Year Equated Students
(FYES)

Zero- order Co rre la ti on s

VI

r)

V2

V3

V4

.45*

.46*

.53*

-

.50*

.31

.55*

-

.03

.44*

-

.27

V2
V3

V5

- .37

V4
*p<01;

(Pearson's

**p<.001

67
operat in g ex pe nd it ur es
the enro ll men t

per fu l l- ye a r- eq ua te d

var iable

(FYES)

shows

student and

that ad mi n i s t r a t i v e

ex p en di tu re s decline with enr oll men t increases.
ma i ni ng ne gat ive c or rel at ion s

suggest that

tive e xp en di tu re s most inf lue nce d
(V5) are in st it ut io n al
services

(r= -.27)

ex p en di tu re s
re g ar dl es s

Mu lti ple
these
The

I ns tr uct io nal

and student
support

size as indicated

by the r= -.03

coefficient.
reg res sio n analysis was used to test further

obs ervations.

The results are reported in Table

stepwise method

a dm i n i s t r a t i v e
variable

(r« -.37)

per student appear to remain fairly constant

of en ro ll m en t

c or re la ti on

the a d m i n i s t r a ­

by changes in enro llm ent

a dm in i st ra ti on

expenditures.

The r e ­

first iden ti fie d

ex pe n dit ure

of gr eatest

f ul l - y e a r - e q u a t e d

7.

the in sti tu ti on al

per student as the in dep en den t

influence.

With

the add ition

of the

student va riable the p<.05 limits were

reached.
The results
gested

found in Table 7 confirm

by the co rr e la ti on s in Table 6.

m in i s t r a t i v e e xp en dit ur e
the operating ex p en di tu re
penditure

per student

su g­

The In sti tu ti o na l a d ­

is highly c o r re la te d with

per student;

however,

op er a ti ng e x ­

per student declines with incre ase s in enrollment.

The re g re ss io n results indicate
ences,

the findings

the influence

s tru cti on al

that beyond

these two i n f l u ­

of a dm in is tr at iv e ex pe nd it ur es

for in­

support and student ser vices is not significant.

The pooled var ia nce
va ri ab le s acc epted

(R^a.39) acc o un te d for by the in de pe nd ent
in the regr es sio n indicates

that a

68

sub st ant ia l amount of the operating
is not accou nte d

TABLE 7:

Stepwise Mu ltiple Regression of Operating E x ­
pen ditures per Student on Adm in is tr a ti ve
Comp on ent s and Enrollment (FYES).

Va ria ble

1

In sti t ut io na l Adm.

B

Beta

2.06

.64

.53

.28

2,403.94

243.34

In s ti tut io nal Adm.

1.54

.65

.39

.28

FYES

-.04

.02

-.35

.39

2,752.19

280.70

(c o n s t a n t )
.05 limits reached

p<.05

The hyp ot he si s as first stated in this
suppo rt ed

by the altern ate

in variables

in the model.

regression model
The

for 28%

student;
tive.

(R2 ».28)

however,

Th erefore,

influence
fairly

of the operating

support

due to changes
that

per student ac­

(FYES)

enrollments may diminish

per

is ne g a ­
the

of institut io nal a d mi ni s tr at iv e expenditures.
small

standard deviation

a d m i n i s t r a t i v e ex pe n di tu re s
deviation

is not

expend itu re

the enrollment variable
high

test

findings do indicate

i ns ti tut io nal ad m i n i s t r a t i v e e xp en dit ure s

The

R2

SE/B

(c o n s t a n t )

count

per student

for in this model.

Step

2

e xp en dit ure s

(112)

in f u l l -y ea r- e qu at ed

the likelihood

extremes would

that

have greater

in institut io nal

and

the large standard

student en rollment

(A,172)

the effects of enrollment
influence

than institut io nal

69

a d m i n i st ra ti ve expend itu re
support is provided

extremes.

Therefore,

for the hypo th esi s

partial

that i ns ti tu tio na l

a dm i n i s t r a t i v e e xp end it ure s will decrease as college
en rollment increases.
support and student

Fi nd in gs

regarding instruc tio na l

service e xp end it ure s are inconclusive.

No support was found

for the hypo th esi s that

strative expendi tu re

per student will

total a d m i n i ­

be greater

in large

enrol lm ent colleges.

Hy p ot he si s

In testing hypo the sis

III

II the effects

of d i f f e r e n t i a ­

tion limited to ad mi n i s t r a t i v e expend itu re s was i n v e s t i ­
gated.

In hypo the si s III the ex pen dit ur e per fu ll-year-

equated student was in ves tig at ed in terms of the total
college

operating expenditure.

In str uc tio nal

salary and

benefit ex pen d it ur es and the total no ni ns tr uc ti on al
penditures were two v ar iab le s added
com ponents
tested is:

tested

previously.

ex pen d it ur es

to the a d mi ni st ra ti ve

S pe cif ic all y

the hypo the sis

per fu l l- y ea r- eq ua te d

greater in large enrollment

colleges

ex­

student are

than in small colleges,

but that the increase is dependent upon non in str uc ti on al
ex pe ndi tu re

increases rather than

faculty expenditures.

The co rr ela tio n matrix of these variab le s is reported
Table

8.
The results

ceptable number

reported in Table 8 indicate an u n a c ­
of high int er correlations.

i nd epe nd ent variables were

All of the

in ter co rr el at ed at r*=.83 or

in

70

TABLE 8:

Co rr el a ti on Coeff ic ien ts Relating Gross E x p e n d i ­
ture and College Size Variab les to the O pe ra tin g
Ex pe ndi tu re per Student.

Variables

Dependent

Mean

Variable:

O p era ti ng Ex pe nd itu re
per Student (OE/FYES)

VI:

St.Dev.

(n=29)

3,155

A37

6,588,160

5 , 5 8 3 , 2A9

Indepe nd ent Variables:
V2:

No n in st ru ct io n al

V3:

In st ruc ti ona l Exp.

6 , 0 A 8 ,58A

5 , AA6.806

V4:

Student Service Exp.

1 ,359,325

1,252,198

V5:

I n st ru ct io na l S u p ­
port Exp.

1,531,376

1,624,145

Ins ti tu ti on al A d ­
m in ist rat io n Exp.

1 , 3 9 3 , A75

1,406,794

Full Year Equated
Students

A , 285

4,172

V6:

V7:

Zer o- order Co rre la tio ns

VI

V2

V3

-A1

- . A2

V2

.97**

V3
VA
V5
V6
* p < . 01;

Exp.

(Pearson's r)
VA

-.44*

V5

V6

-.37

.36

V7
-.50*

.97**

.92**

.90**

.98**

.97**

.83**

.81**

.98**

.85**

.88**

.98**

.92**

.86**
.85**

**p<.001

71
above with a pr ob abi lit y of p<.001.
relations result

These high i n t e r c o r ­

from at least two factors.

the pre vious test,

First,

as in

the depen den t variable was based on a

per e qu ate d student ex pen di tur e while the independent
v a ria bl es are mea sur es of ex pen dit ur es witho ut a per s t u ­
dent c ons ide ra tio n.

Second,

the non i ns tr uc ti on al

expendi­

ture va ria ble is highly correlated with a d mi ni st ra ti ve
component

variab le s

ad m i n i s t r a t i v e

compon ent

co n sti tut e about
itures.

because an ex ami na tio n of the mean
ex pe nd it ur es

that they

65% of the mean n on in st ru c t i o n a l

Re g ar dl es s of these problems,

relat io ns in the dependent variable
inverse

indicates

expend­

the ne gat iv e c o r ­

strongly

suggest an

re l at i on sh p between each of the in dep e nd en t v a r i ­

ables and
designed

the dependent
to test

An alter nat e model was

these r el at io nsh ips

A second model was
de ve lop ed

variable.

designed

in hy po the si s II.

further.

based on the alt er na te model

A d mi n is tr at iv e com po nen ts were

combined and entered as a function of fu ll- ye a r- eq ua t ed
students.

In like manner,

variable was modified.
by compon en t had already
hypothesis,

the ins tr uct ion al

Since a d m i n i st ra ti ve

expendi tur e
expen di tur es

been ex amined in the previous

it was now ap pr op ri at e to look at the a d ­

dition of the in st ru ct ion al exp en dit ure

va riable in order

to determine its influence on operating expenditures.
college

size m ea sur e in fu ll -y ea r - e q u a t e d stu dents was

retained.
omitted

The

The n o n in s tr uc ti on al

since about

e xpe ndi tu re

variable was

65% of no ni ns tr u c t i o n a l expe ndi tu res

72

are expres se d

in the combined a dm i n i s t r a t i v e expenditure.

The co rr el at i on ma t ri x of the variables

in this c o n f i g u r a ­

tion are reported in Table 9.
The results
va ri abl es

in Table

9 indicate

in terms of f u l l -y ea r- eq ua te d

co r rel ati on s above r=.70.
va ri ab le
exist

Ex ami na tio n

in this ma trix indic ate s

between

ins tr u ct io na l

a d mi n i s t r a t i v e
rating

these correlations,

c oe ff ici ent s

(r= -.50)

students

does

enrollment

between

in a t test of cor ­

sample

(t =

.3956;

e xp en dit ure s

not support the hy po t he si s

increases,

and

di fference

found

and operating

colleges.

(r=.55)

The ne gat iv e co r re la ti on

per student will be greater
than in small

corre la tio ns

to o p e ­

in a dependent

critical value = 2.479).
y e ar -e qu at ed

was

eliminated

in relation

No signi ficant

operat in g

of full-

per FYES

that exp en di tu re s

in large enrollment

These

the

of the dependent

si gnificant

(r=.61)

however,

st an dar diz in g

students

e x pe ndi tu res

exp end it ur es

expenditures.

relation

that

colleges

cor re la ti on s suggest
ex penditures

that as

per student

for

in st ruc ti on and a dm i n i s t r a t i o n decline.
These

o bs er va ti on s

were tested
cluding

further

based on c or re lat ion s

in an altern ate

the four varia ble s

indicates

the results

comp on ent s
able.

r eg res si on model

iden ti fie d

of the mu lti ple

Data in Table 10 indicate
had very little

that

influence

When ex ami ne d by compon ent

in Table

in Table 9.

in ­

Table

10

regression analysis.

the a d m i n i s tr at iv e
on the de pe nd e nt

in hypothesis

i n st itu ti ona l a dm i s t r a t i v e compo nen t

9

va r i ­

II the

alone accoun te d

for

73

TA B LE

9:

Co rr el ati on Co ef fi ci e nt s Re lat in g Expen dit ur es
Per Student and Coll ege Size Variables to the
Op er ati ng Ex pen di tur e per Student.

Varia bl es

Depend en t
VI:

Mean

(n=29)

St.Dev

Variable:

O p er at in g Ex pe nd it ure
per Student (OE/FYES)

3,155

437

1,444

209

A d m i n i s t r a t i v e Exp.
(AE/FYES)

1,066

253

Full Year Equated
Stu dents (FYES)

4,285

4,172

In de pen de nt Variables:
V2:

V3:

V4:

Inst ru cti on al
(IE/FYES)

Exp.

Zero- order Co rr el at io ns
V2
VI

.55*

V2
V3
* p < .01;

(Pearson 's r)
V3

V4

.61**

-.50**

.06

-.17
-.28

**p<.001

74

TABLE

10:

Ste pwi se Mu lti pl e R eg res si on of O p e r at in g E x ­
pend itu res per St ud en t on C o m bi ne d A d m i n i s t r a ­
tive Co mp on en t Exp enditures, I ns tr uc tio nal
Ex p en di tu re s and En r ol lme nt (FYES).

Step

Vari a bl e

B

SE/B

Beta

1

A d mi ni st ra ti on

1.05

.26

.61

.37

2,040.64

288.04

A d m i n i st ra ti on

1.10

.18

.64

.37

In s tr uct io n

1.22

.22

.58

.71

216.61

385.48

(constant)

2

(constant)

3

R2

.97

.17

.57

.37

Instruct io n

1.12

.21

.54

.71

FYES

-.03

.01

-.26

.77

606.25

384.79

Admi n is tr at io n

(constant)
.10 limits

reached

p<.10

75

28 percent
iture

(R^=.28)

of the variance

per student.

is 37 percent
tional

The pooled influence

(r 2=.37).

expenditure

34 percent

and in str uc tio nal

leaving only

10 further
exp enditures

fluence on operating
sta nd ardized

the small

that

account

per student.

deviation

of enrollment

per student.
students,

per FYES would

Their

(Table 9) in
in enrollment

(FYES)

is greater

exp enditures
In colleges

reductions

on

exceeding

in op erating

decline while a dm in is tr ati ve and

expe ndi tu res would remain relatively constant.

findings

in conjunction with those of hypotheses

I and II strongly support
incre ase s result

the conclusion

in economy of scale.

II further conf ir m these

that enrollment

Data found in Table

findings.

Data in Table 11 Indicate

that

the mean operating e x ­

per student dec lines with increases

In ad dition

in­

The n o n ­

are significant.

ins tructional

10,000 fu ll -y e ar- equ at ed

instructio na l

for the major

exp en di tu re s

standard

and

operating expen di tur es

ex p end itu re s

that a dm in is tr ati ve

standard de viation associated with a d ­

the influence

than a d mi ni st ra ti ve

penditure

indicate

times the standard error in each case;

to the large

These

29% of the variance

expe ndi tur es

m i n i s tr at iv e and ins tr uct ion al

suggests

for an additional

regression coe ff icients

values exceed three

comparison

9

for.

Data in Table

however,

found in Table

The in tro duction of the i n s t r u c ­

variable ac counts

( r 2 s .34)

u n ac co un ted

in operating e x p e n d ­

the decline in standard

deviations

in enrollment.
indicates

76

that
per

the range
st ud en t

of d i f f e r e n c e s

also

diminishes

in o p e r a t i n g e x p e n d i t u r e s

with in c r e a s e s

in e nr o l l m e n t

size.

TA B L E

11:

Mean and S t a n d a r d D e v i a t i o n of O p e r a t i n g E x ­
pe n d i t u r e per F u l l - Y e a r - E q u a t e d S tu den t in
M i c h i g a n Pu bli c C o m m u n i t y Co l l e g e s

College Enrollment

Less

than

(FYES)

n

2500

St.Dev.

Mean

15

3,2 84. 87

4 55. 60

Between

2500-6000

6

3, 1 8 8 . 6 7

4 18. 59

Greater

than

8

2, 88 6. 38

320.28

Base d

6000

on

No supp or t

these o b s e r v a t i o n s

is found

to i n d ic at e

full-year-equated

student

will

rollment

than

small

wh ile

co l l e g e s

in

the n o n s t a n d a r d i z e d

enrollment
deviation

v a ri ab le

differences

in o p e r a t i n g

that

t ia tio n

were

In h y p o t h e s i s
investigated
nu mber

II and

ex a m i n e d

( A , 172)

To

the high

makes

efficiencies

III

for

the

st an d ar d
i m p or ta nt

on en rol lm ent .

IV

of

of o p e r a t i n g

differen­
e x p e n d i tu re s.

differentiation

in terms of c u r r i c u l a r
and c our se

b as ed

en­

the co ntrary,

pos si ble

the e f f e c t s

in terms

per

in large

colleges.

(-.03),

IV the e f fe ct s of

of co u r s e s

expenditures

be gr e a t e r

Hypothesis

In h y p o t h e s e s

III is rejected.

regression coefficient

is small

in e n r o l l m e n t s

hypothesis

va r ia ti on s .

s e c ti on s

were

were
The

examined

for

77

the influence of these variables
ratio.

In a d di ti on

tained as a measu re
hypot he sis tested

on the st ude nt- fa cu lt y

ful l- ye ar -e qu at e d

students were r e ­

of coll ege size.

Specifically,

is:

the

the s tu de nt - f a c u l t y ratio is in­

dependent of coll ege enrollment

size when course

differ­

entia ti on and number of sections are taken into c o n s i d e r a ­
tion.

The results

of this inv es tig ati on are rep o rt ed in

Table 12.

TABLE 12:

Co rr e la ti on Co e ff ic ie nt s Relating the Number
of Courses, Sections, F ul l- Ye ar - E q u a t e d S tu­
dents, and the St ud en t- Fa cu lt y Ratio (SFR).

Variable

Course s
Sections

(CO)

SEC

FYES

SFR

.87**

.66**

.34

.92**

.55**

(SEC)

F u l l - Ye ar -E qu at ed
Students (FYES)
* p < .01;

.63**

**p<.001

Data in Table
y ea r-e qu ate d

12 indicate

students is more highly co rr elated

with the s t u d en t- fa cu lt y
courses or sections.
co r re la ti on

between

the st ud en t- fa cu l ty
between

that en rollment

ratio

(r = .63)

than either the number

of

The data do indicate a signifi ca nt
the number
ratio.

of sections

(r=,55) and

No significant dif ference

these corr ela ti ons was

lation c oe ffi ci ent s

in full-

(t = 1.317;

found

in a t test of c o r r e ­

cr itical value = 2,479).

78

The number

of sections

clearly has some re la ti o ns hi p with

the stu de nt- fa cu lt y ratio;
the st ude nt- f ac ul ty

however,

it is also clear

ratio is not in dependent

Due to the high

i n t e r co rr el at io ns

theless,

to problems

in the interest

might have the greatest
inv es tig at e

influ enc e

the

significant

of this test

F value

for

for

the enr ol lme nt

The

selection limits of

(FYES)

(R^=

.395),

was

of cu r r i c u l u m in
The

17.68 with a s i g n i f ­

co ef fic ie nt of d e te rm i na ti on
and the

.05 were reached with

Sc a tt er pl ot s did reveal two cases

accoun te d

stepwise

step one.

that deser ve c o m ­

The plot of courses with s t u d e n t -f ac u lt y ratio

contained two outliers.

In the first case one ins ti tut io n

(Lansing Co mm un it y College)
That number
of courses
outlier

the

of courses or se ctions offered.

40% of the va riance

ment.

variable

the enr oll men t va ria bl e was

icant F = .0003.

in di ca te d that

inf lue nc ing

than d i f f e r e nt ia ti on

terms of the number

test was run to

selection in the re­

of the three in dep en den t variables

more

Never­

of de ter mi nin g whic h va riable

The results

s t u d e n t -f ac ul ty ratio,

12

ratio var iable

of mu l ti co ll in ea ri ty .

the step one variable

gression model.

of enrollment.

seen in Table

mult ip le re gr ess ion on the st ud e nt -f ac ul ty
would be subject

that

is nearly
(1,175).

re por te d offering

three times the second larg est number
A co m pa ri si on of means exclu din g

case results in a decli ne from a mean

to a mean of 609.

2,933 courses.

the

of 689 courses

This case is an i nf lu ent ia l anomaly.

This case has not been excluded,

however,

because

of the

79

small

popu lat io n

ac curate

size and

an

de s cr ip ti o n of the population.

A second outlier
faculty ratio.
lege)

the in te gri ty of ma in t a i n i n g

is a ss oc iat ed with

One in sti t ut io n

the student-

(Oakland C om mu nit y C o l ­

reports a ratio of 30.2:1 which excee ds the second

high es t ratio

(23.03:1)

for this variabl e
lies more

by 7.17:1.

The standard d evi ati on

is 3.37 j therefore,

than two

the case in question

sta ndard dev iations

from the mean.

Again this case was ret ai ne d to ma int ai n

the i n t eg ri ty of

the po pu la ti on d es cr ipt ion .

statistics

va ri ab le s in

TA BL E

13:

this

test are

De sc ri pt iv e

reported in Table 13.

Means and St and ard De vi at io n s for Number of
Courses, Sections, F u l l - Y e a r - E q u a t e d Students,
and the S t u de n t- Fa cu lt y Ratio,

Mean

Variable

Courses

(n»29)

S t .D e v .

689

531

Se ctions

2,642

2,172

Full Time Eq ua te d Stu de nts

4,285

4,172

18.6:1

3.4:1

St u d e n t - F a c u l t y Ratio

Based on results of these
great

for

tests

it is ev id ent

di ff e re nt ia ti on in curric ula was

present.

that

The mean

number of courses off ere d was 689 with a ’standard dev ia ti on
of 531.

The number

varied greatly.

of m u lt ip l e

Nevertheless,

found to be the variable

sections

offered also

the enro llm en t variable was

of greatest

influence on the

80
stu d en t- fa cu lt y
both

ratio.

the signifi can t

This ob ser va tio n

correla ti on

12 and by the an aly s is
reg r es si on analysis.

of r=,63

of va ri an ce
Therefore,

is suppor te d

by

found in Table

provided

by mul tiple

no support was found

for

the hy po th esi s that the st ude nt -f ac ul ty ratio is i nd e p e n ­
dent of college enrollment
tiation

is considered.

suggests

size when curricular

differen­

C or rel ati on and reg ression analysis

that the enrol lm ent

variable more than

the c u r ­

ricular d i f f er en ti at io n variab les influ enc e s t u d en t- fa cu lt y
ratios.
high

These results may be spurious,

i nt er co rr e l a t i o n s between
While outliers were

outliers.

due to the

independent variables.

retained to mai nta in the integrity

of the popu la tio n description,
run a second time

however,

the regr ess io n analysis was

to examine the effect of removing the

High in te rc o rr e l a t i o n s between the independent

variables

remained;

however,

co rre la tio ns of each of the

in de pen de nt va ri a bl es with the depend ent

variable declined.

C o rr el at io ns

ratio and both the

multiple

the s t u d e n t - f ac ul ty

section and

s ign ifi ca nt
nificant

between

levels,

levels

enrollment

(FYES)

variables fell

r=,55 and r=,63 respectively,

of r«.41 and r=.40 respectively.

from

to i n s i g ­
In the

re gr es si on

equation

from 16.41

in the first model to 16.66 in the second model;

however,
tiple

the constant was virtu all y un cha nge d

the stepwise

selection metho d id e nt if i ed

section var ia bl e as the influent ial

than the e n ro llm en t variable.
( r 2=,166)

was accou nte d

the m u l ­

variable rather

Only 17% of the variance

for in these cases,

and the

81

stepwise

select ion

limits

of p<.05 were

reached with

step

one.
These

findings are important.

of curr icu la r d i f f e r e nt ia ti on
but perhaps more
stant

exa mi nat ion

that

variables.

The

constant

TABLE 14:

some change

deviations

but the change

are extremely

that

the c o n ­

unchanged.

of the de scr ipt iv e statistics

of means and standard
are removed,

in this model,

important is the finding

Table 14 indicates

remain

is detect ed

in the regr es sio n eq uation remains

addition,

values

Some small influence

In
in

in the rel ative value

results when

the outliers

in the stu de nt -f ac ul ty

small by comparison

to the other

stu de nt -f a cu lt y ratio value appears
regar dl ess of changes

ratio

to

in the independent

Means and Standard Devi ati on s for Number of
Courses, Sections, F ul l- Ye ar - E q u a t e d Students,
and the St ud en t- Fa cu lt y Ratio with Outliers
Removed.

Va ri abl e

Mean

Courses
Sec tions
F u ll - Y e a r - E q u a t e d

Students

S t u de nt -F ac u lt y Ratio

variables.

The

possib ili ty

influenced,

if not controlled,

(n=27)

S t .D e v .

591

306

2257

1669

3618

3397

18.1:1

2.6:1

is raised that this ratio is
by a variable not

included

in the model.
Politi ca l

co nst ra int s

such as the influence of

82

collective
present

b a r g a i n i n g may

in the de p e n d e n t

al a s s i g n m e n t s
agr e em en ts ,
m in or

are

controlled

of i n d e p e n d e n t
from

only.

in the very

variable.

the i n d e p e n d e n t

influences

su l t i n g

result

variables

If f ac ult y i n s t r u c t i o n ­

by c o l l e c t i v e

va ri ab l es

The

small v ar i a n c e

in

b ar g a i n i n g

this model

loss of s i g n i f i c a n t

with

the e l i m i n a t i o n

the

de pe nd en t

of ou tl i er s

may

be

correlations

va ria ble

s u pp or ts

re­

this

observation.

Hypothesis

Curricular
vestigated

differentiation

for their

this test.

in f l u e n c e

The h y p o t h e s i s

and e n r o l l m e n t were
on av e r a g e

tested

re m a i n s

unchanged

as c o l l e g e

c re ase s

in c our se

differentiation

of fe red

a ff ect s

Unduplicated

stud ent

he a d c o u n t

differentiation,

in p r o p o r t i o n

unduplicated
students.

va r i a b l e s

h ea d c o u n t

c o ns ta nt

he a d c o u n t

students.

lationship

betwe en

clas s

incr ea ses .

the nu mber

size
In­

of m u l t i ­

a ve r a g e class

were e s t a b l i s h e d

eq u ate d

Correlations

size in

size.

in c r e a s e d
stu dents

h e a d c o u n t declines.

than e n r o l l m e n t m e a s u r e d
Correlations

in Ta ble

in ­

in

terms

of

as well as f u l l - y e a r - e q u a t e d

Unduplicated

r e po r te d

average

in c r e a s e s with

fl u enc e

are

and

class

but f u l l - y e a r - e q u a t e d

to u n d u p l i c a t e d

Enrollment

is:

enrollment

ple s e c t i o n s

course

V

may

be a s t r on ge r

in­

in terms of fu l l - y e a r b et we en

these

v a ri ab le s

15.

in Ta ble

15 ind ic at e

headcount

(r=.59),

a si g n i f i c a n t

re­

full-year-equated

83

students

(r=.63)

and av e rag e class

con tr ar y to the hypo the sis

size.

stated above.

re la ton sh ip between number of courses

TABLE 15:

This

finding is

No significant

(r=.26),

number of

Co rr ela tio n Co ef fi ci en ts Relating the Number of
Courses, Sections, U nd up li ca te d Headcount, FullY e ar -E qua ted Students, and the Average Class
Size.

Variabl e

Courses

SEC

(CO)

Sections

UH

.87**

(SEC)

Ave.
Class Size

FYES

.78**

.66**

.26

. 97**

.92**

.41

.98**

.59**

Un d up li ca te d HeadCount (UH)
F u l l - Y e a r -E qu at ed
Students (FYES)
* p < .01;

sections

**p<.001

(r*=.41),

the other hand,
cantly
courses

.63**

and aver ag e class

un duplicated

size was found.

student headcount

co rr ela te d with di ff er en ti a ti on
(r=.78)

and sections

in f u l l - y e a r- e qu at ed
he ad co un t was found.
the curricular
and sections,

is s i g n i f i ­

variables

of both

but no decline

students in proportion

to un dup l ic at ed

The positive re la ti ons hip between

di f fe re nt ia ti on

variables,

number of courses

and the headc oun t va ria bl e suggests

courses and sections relate
addition,

(r=.97),

On

that more

to increa se s in students.

the c or re lat io ns of u nd upl ic ate d headcount

In
and

f u l l - ye ar -e qu at ed

students with aver age class size de mo n­

strate rela tiv el y

no loss

in avera ge

class

size efficiency.

84

No si gn ifi can t
co r r e l a t i o n s
Pro ble ms

dif fe re nc e was

(t = 1.31;

critica l

value =

of these

2.479).

of m u l t i c o l l i n e a r i t y would again be present

in mu lti ple reg res si on
in hy po th esi s

found in a t test

IV,

due to high

however,

i nt er co rre lat io ns.

int erest

in the stepwise

As

s el ec ­

tion in the re gr es ss io n model was reason enoug h to run the
test.

The results

in Table

of the regre ss ion an al y si s are reported

16.

As in the previo us model,
dent va ria bl e was
a v er ag e class

size.

in ave rag e class
sec ti ons

found

courses was entered.
en te rin g

un af fe c te d

ficients.

step

headc oun t

of

of

variable.

in this an aly si s is re la ­

in f ul l- ye ar - e q u a t e d
in each

of di ff e r e n t i a t i o n

of courses

the number

by n on s t a n d a r d i z e d reg res si on

two is ne u t r a l i z e d with

number

on

The p< .10 limits were reached before

tains a positi ve influe nce
influe nc e

influ enc e

step the number

and in step three

es ta bli she d

Increa se s

st u ­

for 40% of the variance

In the second

the un du pl i ca te d

The con st ant
tively

to have the gr eatest

It ac co unt ed

size.

was entered,

the f u l l - y e a r - e q u a t e d

st udents r e ­

step.

in number

coef­

The neg ati ve

of sections

the po si t iv e influe nce

in step three.

The as su mp ti on

in
of the

regarding

the in fl ue nc e of u n d u p li ca te d hea d co un t as a greater in­
fluence

than

fu ll -y e ar - e q u a t e d

These o bs er va ti on s may,

however,

lems of mu lt ic ol lin ea rit y.
straints

students was not satisfied.
be m i s le ad i ng

In addition,

due to p r o b ­

political c o n ­

such as those noted in h yp oth es is

IV may exert

85

TABL E 16:

Stepwise M u lt ip le Re g re s si on of Ave rag e Class
Size on the Number of Courses, Sections, U n ­
dupli ca ted Headcount, and F u l l -Y ea r- Eq ua te d
Students.

Variable

Step

1

14.47

(constant)

.10 limits reached

R2

. 63

.40

.74580

1.64

.40

-.00178

.00052

-1.10

.58

15.52

.70164

.00214

.00032

2.54

.40

- .0 045 4

.00093

-2.81

.58

.00672

.00199

1.02

.71

CO
(constant)

.00013

Beta

.00027

FYES
SEC

SE/B

.00138

FYES
SEC

3

.00053

FYES
(constant)

2

B

14.93
p<.10

.61791

86

an important influe nce
Based on these

on avera ge class

tests

size.

the hypothesis

that average class

size remains unchanged as college enro ll men t
rejected.

The

influence of enrollment is evident

c orr ela ti on and re gression coefficients.
influence

increases

of dif fe re nt ia ti on

cated in regression analysis;

in both

Some negative

in course sections
however,

is

is i n d i ­

these results are

spurious due to the unknown effects of m ul ti c ol lin ear it y.
Co r r e l a t i o n coeff ici ent s did not support
regar din g

increases

the hypo th esi s

in un du pl ic at ed headcount;

therefore,

it is rejected also.

Hypo the sis

VI

This hyp oth esi s was des igned to test the influence of
revenue sources on a d m i n i s tr at iv e salaries.
tested is;

a d mi ni st ra ti v e e xp en di tu r es

equa ted student

The hy pothesis

per full-year-

vary as a function of available

resources

rather than college enrollment.
In dep end en t va ria bl es in clu din g property tax revenue,
state aid revenue,
amined

and tuition and fee revenue were e x ­

for their influence on a d mi ni st ra ti ve

benefit e xp en dit ure s per fu ll - ye ar- equ at ed

salary and

students.

a measure of colle ge size the fu ll -y ea r- eq ua t ed
enr oll men t was also

included

student

as an ind epe nd ent variable.

The intent was to in ves tig at e whether
direct

As

services

beyond

instruction were depen den t more upon av ail ab le r e ­

sources or on the size of the enrollment

served.

Each of

87

the

re ve nu e

equated

variables

s tu den t

The c o r r e l a t i o n
of these

basis

means,

correlation

sa lary was

p r op er ty

tax and other

tuition,

is s i g n i f i c a n t

and

the

state

aid

crease,

An a d d i t i o n a l

b et wee n

state

aid and

s tu den t

contribution

in e n r o ll me nt .

relationship

wo uld

f i nd in gs

tricts
would

hold

true

omits

tive sa l a r i e s
mi n or

r e po rt e d

i nf l u e n c e

variable

in the

linki ng

to have ei th er

18.

as a factor

Property

This

The

direct.
high
paid

than d i s ­

observation
size.
regres­

regression

influencing

(TAX)

the

higher

by m u l t i p l e

enrollment
tax

but

than

s e rv ic es

per

c ol le g e

(r= -.19),
rather

found
The

d i s t r i c t s wi th

su p p o r t e d

and w e i g h t s

de­

i n c r ea se s

of co l l e g e e n r o l l m e n t

in T a b l e

(.02).

s e le ct ed

share

students.

found

tax revenue.

are

(TU)

(SAL),

was

s a la r ie s

likely

regardless

t u it io n

tuition

administrative

observations

sion a n a l y s i s

ex pen di tu re s;

to be in ve rse

are

or more

aid and

r e l a t i o n s h i p was

that c ol leg e

with lo w pr o p e r t y

The se

tion

a pp ea r

tax r e v e n u e s

administrators

the

be tw ee n

state aid d ec li ne s wit h
e v i de nc e

s ug ges t

tax and a d m i n i ­

state

full-tirae-equated

L it tl e

deviations

As tax r e ve nu e

sa lary

significant

e n r o l l m e n t and a d m i n i s t r a t i v e

p r op er ty

sources,

negative.

data.

17.

pr o p e r t y

share and

of

the

the r e l a t i o n s h i p

r ev enu e
and

fu l l - y e a r -

st and ard

in Tab le

so do a d m i n i s t r a t i v e

however,

These

and

b et w e e n

found,

on a per

to s t a n d a r d i z e

is r e p o r t e d

strative

increases

f a c to re d

in order

matrix,

va ri a bl es

A high

was

administra­

(FYES)
is the

r e g r e s s i o n an aly sis ,

equa­

and

as a
first
its

88

TABLE

17:

Co rre la ti o n C oe ff ic ie n ts Re lating Revenue,
College Enrollment, and A d m i n i s tr at iv e Salary
and Benefits.

Variable

Mean

S t .D e v .

(n=29)

De pe nde nt Variable:
VI:

Ad mi n i s t r a t i v e Salary
and Bene fi ts/ FYE S

773.34

186.38

Ind ep end en t Variables:
V2:

Pro per ty Tax/FYES

V3:

State Aid/FYE S

V4:

Tuit ion

V5:

FYES

.69**

V2

271.46

939.66

176.17

4,285

(Pearson's r)

.01

-.41

19

-.46*

21
48*

•

V5

i

V4

CO

V4

4,172

V3

-.44*

V3

*p<.01;

1 ,247.38

o
•
1

VI

531.95

& Fees/FYES

Ze r o- ord er Co rr el a ti on s
V2

883.38

* 29

**p <.001

89

TA BL E 18:

Step

1

Stepwise Mul ti p le Reg re s si on of A d m i n i s t r a t i v e
Sal aries on Rev enu e and Enrollment.

Va ria bl e

B

SE/B

Beta

R2

.24

.05

.69

.48

558.83

49.89

Pro pe rty Taxes
(constant)

2

Pr operty Taxes

.30

.05

.86

.48

State Aid

.27

.09

.39

.60

173.36

143.97

Pro perty Taxes

.37

.05

1.05

.48

State Aid

.44

.12

.64

.60

Fu l l- Ye ar -E qu at ed
Student

.02

.007

.35

.67

(constant)

3

(constant)

.05 limits reached
Note:

-166.09

203.55

p<.05

An adj usted R2 of .62 indica te s that little if
any bias is present due to the ratio of v a r i ­
ables to cases.

90

signif ic ant
and

i nfl uen ce is Indica te d by both the low SE/B

the high Beta

SAL «

- 166.09

(1.05).

+

re i nf or ci ng
but negative

equation

.37 (TAX)

The ne gat iv e constant
y ea r- eq ua te d

The

+

pr ovided is:

.44 (AID)

+

.02 (FYES)

is pr odu ced when enro ll men t

students is taken into the formula
the o bs e rv ati on
in influence.

(.05)

that enr oll me nt

in full-

further

size is small

The negative con stant indicates

an a dm i n i s t r a t i v e overhead expendi tur e

Is present

even at

zero enrollment.
The

co ef fi ci ent of determin ati on

the o bse rv ati on
fluence

that enrol lme nt acc ounts

in salary.

of the f u l l -y ea r- eq ua te d

by only

students

students

7%, from r2=.60 to R^=.67.

thermore the beta weight of property
times as great as

for little In­

The per cen tag e of variance acc ou n te d

for with the inclus io n
variable increases

also supports

tax (1.05)

Fur­

is three

the beta weight of fu ll- ye a r- eq ua te d

(.35).

Based on these finding s this hypothesis
A d mi ni st ra ti ve

is supported.

e xp e n d i t u r e s per fu ll -y ea r- eq ua te d

vary as a function
lege enrollment.

of avail abl e resources rather
Local

the stron ges t influ enc e

property

student
than

col­

tax in par tic ul ar exerts

on ad mi n i s t r a t i v e expenditures.

Hy p ot he sis VII

This hyp oth esi s was designed to In vestigate
ence of c ha r a c t e r i s t i c s

of physical plant

the i n f l u ­

size on college

91

opera ti ng exp enditures.
tenance and energy

The hypot he sis

tested is:

main­

ex pen dit ur es are functions of physical

plant size in d ep en d en t of enr oll me nt

and personnel

size.

Due to l imi tat io ns in avail abl e

data physical

plant

size mea sur es were
total square

limited to a single va ri ab le based on

feet of building interiors.

miles of roadway,

and miles of wa lk w ay s were not available.

The f u l l - y e a r - e q u a t e d
equated

students variable and

positi on s var ia ble were combined

va riable as a measu re of the user
plant.
their

Data on acreage,

press

the ful l-time-

in a single
on the physical

These in de pen den t varia ble s were ex amined
influe nce

maintenance,

on the physical

energy,

The results

and custod ial

for

requirements.

correlated.

Co rr el at i on C oe ff ici ent s Relating Physical
Size, Number of Users, and Physica l Plant
Ex p en di tu re s (PPE)

Var iable

US

Square Feet

(FT)

Nu mber of Users
* p < .01;

e xp en dit ur es

rep or te d in Table 19 Ind icate that all

three va ria bl es are very highly

TABLE 19:

plant

for

PPE

.95**
(US)

Plant

.96**
.97**

**p<.0 01

Cor r el at io n c oe ff ic ie nt s
be made.

provided no basis upon which

d is ti nc ti on s

could

St epwise mul tip le re gression

was run only

to determine which

variable was first

selected.

92

The user va riable was

selected with an F value

with a signif ica nt F = .00005,
te rm in a ti on

att ached

of 394,93

The co ef fic ie nt of d e ­

to this va ria ble was r 2=,94.

to pro blems of m u l t i c o l l i n e a r i t y , however,

Due

these o b s e r v a ­

tions are spurious.
This hyp oth esi s was neither

accepted nor rejected.

The results are inconclusive.

Summary

The findings

of statistical

tests regar din g e f f i ­

c i en cy and economy of scale in Mic hi g an
co lleges were

presented

in this chapter.

m u l t i c o l l i n e a r i t y were en c ou nt er ed
cases models were redes ign ed
a v ail ab le
ardize
mined

data.

data

of m ea su re me nt

to better

of

In some
test the

it was ne ces sa ry to s t a n d ­

by treating them in terms

by the fu ll - ye ar - e q u a t e d

Pr obl ems

frequently.

in order

Most frequ ent ly

public co mmu ni ty

of an aver age d e t e r ­

students.

Thus the units

in the in de pen den t varia ble s were

st an d­

ardized,

and results were cal cu la te d with a common base.

Further

in te rp re ta t io n of the findings are provided with

conc lu sio ns

in Chapter

V.

CHAPTER V

SUMMARY,

C O N C L US IO NS AND REC OM ME ND AT IO NS

Ove rview

This research invest ig ate d
characteri st ics

the i nf lue nc e of structural

of or ga niz ati on size on measures of e f f i c ­

iency and economy of scale.

Based upon Blau's theorems

on the process of differentiation,
recent studies,

and the findings of more

this research examined

components in relation to enrollment,

the size of personnel
resource al location

in rel ation to size and effi cie ncy measures,
allo ca tio n in relation to av ail ab le resources

and resource
in the

twenty-nine public commun it y colleges of the State of
Michigan.

In addition

the effects of cur ric ul ar d i f f e r ­

entiation on s tu den t-f ac uit y ratios and average class
size were investigated.
In this theoretical frame wor k hyp ot he se s were tested
using cor re lat ion ana lys is and multiple regression an a l ­
ysis.
ered

Variables were based on fiscal year 1984 data g a t h ­
by the Michigan Depa rtm en t of Ed uca ti on in the format

of the Mic h ig an Community Colleges Act ivi tie s C l a s s i f i ­
cation S t r u c t u r e . 1 9 8 1 .

The research design is post

hoc and descriptive.
This chapter summarizes
gation,

the results of the i n v e s t i ­

interprets the findings,

mend at ion s for further study.

93

and concludes with re co m­

94
Hyp oth esi s I

Findings

The literature of o rg ani za tio n re search on economy of
scale con tains conflicting fi ndings regarding
of size on the division of labor.

the in fluence

A key problem exists in

that the units of me a su r em en t within
are not cons is ten t from one study

the division of labor

to another.

T he ref or e

ge ne ra li za ti on s regarding the effects of size on a personnel
component must

be carefu lly qualified.

A central issue in the d eve lop me nt of a funding model
for Mic h ig an public commu nit y colleges
economy of scale.

is the issue of

Small college re pr es ent at ive s argue that

ad min is tr at iv e costs are
enrollment colleges.

p r op ort io nat ely higher in small

They argue

further that large en­

rollment colleges realize in str uc ti on a l e ff ic ien cie s

based

on larger average class size and higher stude nt -fa cu ity
ratios.
In this study
examine

the first hypo th esi s was designed

the influence of size on personnel comp on ent s in

Mic higan public commu nit y colleges.

The results of c o r ­

relation an aly si s support the ar gument
scale.

to

for economy of

The number of execu tiv e ad min is tr at or s were not

highly corr ela te d with enrollment,
dent from other

they were in d e p e n ­

personnel groups that were found

highly cor related with enrollment.
hy pothesis I-a:

and

The

to be

Findings support

proportion of executive ad min is tr at or s

95

de c r e a s e s

in m i d s i z e and large colleges and is highest in

small colleges.
The linear reg r e s s i o n of e x ecutiv e a d m i n i s t r a t o r s on
full-year-equated

students results in an intercept of -1107

and a slope of 883.

This finding s uggests

increase s of nearly

1000 f u l l - y e a r - e q u a t e d

r equired

before

influenced.

that e n r o llment
students are

the number of executive a d m i n i s t r a t o r s is

The negative in te r c e p t sets the m i n i m u m o p ­

er ating level
enrollment.

at 1.25 e x ecuti ve a d m i n i s t r a t o r s with zero
S u b s t anti al e c o n o m i e s of scale may be a s s o c ­

iated with this

personnel group and increases in e n r o l l ­

ment.
High c o r rela tions

of other

rollment inc r e a s e s support
or g a n i z a t i o n

the pos ition that changes in

co mplexity and d i f f e r e n t i a t i o n may diminish

gains in economy of s c a l e .
esis 1 - b :

Suppor t was

found for h y p o t h ­

The remain ing a d m i n i s t r a t i v e c omponent increases

pro p o r t i o n a t e l y
While

personnel groups with e n ­

from small

the sc atterpl ot

to mi dsize

to large colleges.

suggests that i ncr e a s e s in this

comp onent may stabilize in large enro llment colleges,
mo ve m e n t from small
highly corr elated
pr ofessionals.

to midsize enr ollment was

found

the

to be

(r=.91) with i n crea ses in a d m i n i s t r a t i v e

The slope of the linear re g r e s s i o n was found

to be 62.40 which suggests
are strongly influen ced

that i n crease s in this component

by i nc rementa l e n r o l l m e n t increases

of fewer than 100 f u l l - y e a r - e q u a t e d
tion was observed

to hold

students.

This c o r r e l a ­

true for e nrol lments up to 11.000

96
full-year-equated
were observed

students.

No Increases In this component

in the two largest colleges with enrollmen ts

of 14,606 and 15,831 f u l l - y e a r - e q u a t e d
bilization may ind icate
tures,

the saturat ion

rendered

students.

This

sta­

that in community college s t r u c ­

point for d i f f e r e n t i a t i o n of services

by a d m i n i s t r a t i v e profe s s i o n a l s is reached at or

slightly above e n r o l l m e n t s of 10,000 full- ye a r - e q u a t e d
students.
The strongest c o r r e la tion
faculty
pothesis

and enrollment.
I-c:

The

this

these

Result s strongly

size.

support h y ­

is most

is 22.28 indicating

however,

ference

between increase s

support

personnel and

sionals comp onent

resulted

is ignored

if economy of scale is measure d

creased

therefore,

support and

Increase s in

to credit hour

in the ma nner of faculty

e f f i c i e n c i e s effected

credit gen eration may

tiation of noncre dit
creased

that are not related

the enrollm ent measure,

expenditures;

profes­

p r o f e ssional s and support personnel

result in e x p e n d i t u r e s
generation,

professionals,

The a d m i n i s t r a t i v e

in terms of executive a d m i n i s t r a t o r s only.
both a d m i n i s t r a t i v e

for s i g n i f i ­

in no sign ificant d i f ­

in a d m i n i s t r a t i v e

faculty.

that

sensitive to changes in

An exam ination of c o r relati ons

cant difference,

between

The slope of the linear re­

two variables

pers onnel c o mpon ent

enrollment.

was found

propor tion of faculty remains constant

re g a r d l e s s of college
gr ession of

(r=.96)

through i n ­

be c onsumed in the d i f f e r e n ­

genera ting servic es rend e r e d by in­
professio nal

staff.

97

The c lerical c o mpo nent is also highly correlate d with
enrollment;

however,

sc a t t e r p l o t and r e g r ess ion data are

not clearly d e f i n i t i v e in de s c r i b i n g the clerical

relation­

ship as linear or curvilinear.

Some tendency in support

of h y p o t hesis I-d is provided:

The

pro portion of clerical

personnel rema ins constant in small and mids ize colleges
but i n crease s in large colleges.
indicates that

the c o n c e n t r a t i o n of the cleric al component

tends to cluster

tightly in colleges of less than 5000

fu l l - y e a r - e q u a t e d students.
the trend

S c a t t e r p l o t examin ation

line conti nues

Beyond this e n r o llme nt level,

but with gre ater diff u s i o n in

higher enrollm ent colleges.

Due to li m i t a t i o n s in p o p u ­

lation size it is not clear what inf l u e n c e is present on
clerical

personnel in college s with enrollment

greater

than 5000 f u l l - y e a r - e q u a t e d students.
The support
(r=.90)

staff component is also highly correlated

with enrollment.

Sc a t t e r p l o t ex amin a t i o n indicates

a linear r e l a t i o n s h i p in support of h y p o thesis I-e:
p ro p o r t i o n of support staff
c ollege size.

The

remains constant regar dless of

Two outl i e r s with e nrollme nts above

10,000

fu l l - y e a r - e q u a t e d students do lie below the slope of the
linear reg ression of the support co mponent in relation to
fu l l - y e a r - e q u a t e d students.

Again it is not clear due to

l imita t i o n s in popu lation size what influence is present on
s upport personnel in colleges with enr o l l m e n t greater
10,000 f u l l - y e a r - e q u a t e d students.

than

98
C o n c l usions

This i n v e s t i g a t i o n of the influence of size on p er­
sonnel comp onents provides ev id e n c e in support of economy
of scale p arti c u l a r l y wi thin the e xecutiv e admi n i s t r a t i v e
group*

Other personnel

componen ts remain pr oportionately

constant with increases in o r g aniz ation size,

but no c o m ­

ponent was found to inc rease d i s p r o p o r t i o n a t e l y with in­
creases in size.

Hyp o t h e s e s II and III

Findings

These hypothese s were designed

to invest igate college

e x p e n d i t u r e s ass ociated with personnel compone nts in re­
lation to college size.
I suggested
co mponents

Conc lu s i o n s reached

in hypothesis

that economy of scale may be present in some
but dimin ished in others*

An exami nation of

ex p e n d i t u r e s a s s ociate d wit h personnel compo nents was de­
signed to test further those conclusions.
Prob lems associat ed with
re definition of the varia bles

the data source required
tested in these hypotheses.

The Activity C l a s s i f i c a t i o n S t ructur e format for reporting
f u l l - t i me-equa ted positions makes

possible the d e t e r m i n a ­

tion of the number of employees in all major categories
of employment.
category

The a d m i n i s t r a t o r / s u p e r v i s o r y / t e c h n i c a l

(ACS 4 C32)

visory roles

lists all a d m i n i s t r a t i v e or s u p e r ­

by area of activity

such as i n s t ructi onal

99
support,

student services,

i n s t itution al support,

plant and indep endent operations.

physical

From that total it is

possible to det e r m i n e the number of n o n - ex ecutive a d m i n i ­
strators

by su btracting

strators identif ied

the number of execut ive a d m i n i ­

by coll ege in the HEP 1985 Higher

Education D i r e c t o r y .

In this manner the e xecutive a d m i n i ­

strator and the a d m i n i s t r a t i v e pro fession als variables in
hyp o t h e s i s I were established.
General fund e x p e n d i t u r e s are reported

in ACS 3,

The

major ca t e g o r i e s for reporting expenditur es are in struction
(1.00),

public service

( A,00),

student services

tion

(6.00),

(3,00),

instruct ional support

(5,00),

I n s t itut ional a d m i n i s t r a ­

physical plant opera tion

pendent op e r a t i o n s

(9.00).

(7.00),

and i n d e ­

Each of these ca tegories may

have a d m i n i s t r a t i v e or supervis ory personnel included.
Some categori es contain specific line
a d m i n i str ative expenditures,
possible to identify

item entries for

but in no category is it

the incl usion of a specific a d m i n i ­

strator or a specific a d m i n i s t r a t o r ' s salary.
based on ACS data it is not possible
admi n i s t r a t o r salary e x pendi tures
strator

salary expendi tures.

Therefore,

to separate executive

from the total of a d m i n i ­

Consequ ently,

admini strativ e

expend iture variables in hypo theses II and III cannot

be

matched wit h a d m i n i s t r a t i v e personnel compon ents expressed
as f u l l - t i m e - e q u a t e d po sition variables in hypothesis I.
For purposes of this study the distor tions resulting
from these m o d i f i c a t i o n s in variables are not important.

100
Interest in e x p e n d i t u r e s within a d m i n i s t r a t i v e comp onents
exceeded int erest in the costs of exec u t i v e ad mi n i s t r a t o r s
above;

however,

cauti on in relating the findings of h y ­

potheses II and III to f indings of h y p o t h e s i s I is n e c e s ­
sary .
Hypo thesis

II was designed

to i n v e s tig ate the i n ­

fluence of e x p e n d i t u r e s in three a d m i n i s t r a t i v e components
on the total operating ex p e n d i t u r e
student.

per full-y e a r - e q u a t e d

C o r r e l a t i o n coef fi c i e n t s and regres sion c o e f f i c ­

ients support the con c l u s i o n that operating e x pend itures
per fu l l - y e a r - e q u a t e d student decline with incr eases in
college enrollment.

E x p enditur es

in the insti t u t i o n a l

a d m i n s t r a t i v e c o mpon ent are more closely ass o c i a t e d with
the decline in operating

e x penditu res

than are e x p e n d i ­

tures for i nstr u c t i o n a l support or student
Results rega rding
conclusive;

services.

these latter two co mponent s are not

however,

c o r r e l a t i o n c o e fficien ts suggest

that student service ex pend i t u r e s
cline more than i nstr u c t i o n a l

(r= -.27)

may d e ­

support expendi tures

(r= -.03).
Hypothes is

III was design ed

to in v e s t i g a t e further

the infl u e n c e of e x p e n d i t u r e s in a d d itiona l

personnel

co mponent s on the total ope r a t i n g ex penditu re

per student.

As in the previous test no support was found for the
hy pothesi s that operati ng e x p e n d i t u r e s

per student were

greater in large than small en r o l l m e n t colleges.
sion analys is supports

Regres­

the con c l u s i o n that operating

101
e x p e n d i t u r e s per student decli ne with increase s in college
enrollment.

The combined inf l u e n c e of all a d m i n i s t r a t i v e

e x p e n d i t u r e s and i n s t r u c t i o n a l e x p e n d i t u r e s accounted
71% of the var iance in operat ing expen ditures .

for

While c o ­

e f f i c i e n t s of d e t e r m i n a t i o n in a reg r e s s i o n model clearly
are not additive,

the inc rease in an

of

.37 ass ociated

w i t h a d m i n i s t r a t i v e e x p e n d i t u r e s e x c l u s i v e l y to an R^ of
.71 a s s o c iated with the combined

influence of a d m i n i s t r a ­

tive and i n s t r u c t i o n a l e x p e n d i t u r e s
on operating

e x p e n d i t u r e s may be balanced

tional and n o n i n s t r u c t i o n a l
.57 (admin istrati on) and
obse rvation.

suggests the i nfluence

influences.

.54

between i n s t r u c ­
Beta values of

(instruction)

support this

Fu rther r e i n f o r c e m e n t for this observati on

is provided in view of these
e x p e n d i t u r e s exclusively.

variables ref l e c t i n g personnel

Expenditures

for instru ctional

m a teria ls and sup plies were not c o n s idered

in this test.

Conc lusions

This
personnel

i n v e s tigatio n of e x p e n d i t u r e s in relation
c o m p o nents provides

of economy of scale.

further evidenc e in support

I n s t ituti onal a d m i n i s t r a t i v e

pen d i t u r e s de crease as colleg e size increases,
operating e x p e n d i t u r e s
in college enrollment.

to

ex­

and total

per student decline with increases
Noninstructional administrative

e x pendi tures do not rise d i s p r o p o r t i o n a t e l y with increases
in in str u c t i o n a l

faculty expenditures.

is not s i g n i f i c a n t l y inf luenced

Economy of scale

by increased

dif f e r e n t i a t i o n

102
in personnel c o m po nents I nclud ing faculty and a d m i n i s t r a ­
tive professionals.

Hy p o t h e s e s IV and V

Fi ndings

These hypot heses were de signed to inv est i g a t e the
influe nce of c u r r icular d i f f e r e n t i a t i o n on e f f i ciency
m easur es in terms of s t u d e n t - f a c u i t y ratios and a v e r age
class

size.

Ho support was found

for the h yp o t h e s i s

that

the s t u d e n t - f a c u i t y ratio is indepe ndent of college e n ­
rollment
ered.

size whe n c u r ri cular d i f f e r e n t i a t i o n is c o n s i d ­

Enro llment was found to be more infl uential on

the stu d e n t - f a c u i t y

ratio

than the number of courses or

mult i p l e course sections in spite of s ubstant ial d i f f e r ­
ences in these measu res of diffe rentia tion.
courses offered ranged

The number of

from a low of 242 to a high of 2933

and the number of s ections ranged from 515 to 8759.
theless,

the s t u d e n t - f a c u i t y ratio was

e n r ollmen t

influenced

Never­

more

by

than curr icular diff erentiation.

Contrary to findings in studies of u niver s i t i e s
(McLaughlin,

1979;

Stomrael,

1985)

the effects of c u r ric ular

d i f f e r e nt iation in comm unity colleges do not result
e f f i ciency when measured
Further insight into this
the data a second
outliers.

in lost

by the s t u d e n t - f a c u l t y ratio*
phen omena was gained

by testing

time with the omissi on of two observed

Some small

influence of d i f f e r e n t i a t i o n

in place

103

of enrollme nt was realized;
ch anges in independent

however,

more import antly

variables produced no signifi cant

change in the d e p endent variable.

The influ ence of a

vari a b l e not a c c ounted for in the model was suggested.
Political c onst raints within the o r g a n i z a t i o n may account
for

the lack of vari ance in the dependent variable.

pervasive i n fluen ce of c ol l e c t i v e bargaining
may accou nt for a defined

The

in M ich i g a n

s t u d e n t - f a c u i t y ratio more than

structural c h a r a c t e r i s t i c s within the o rganization.

Fur­

ther researc h is needed to confirm this observation.
Hyp o t h e s i s V was designed

to in v e s t i g a t e the effects

of c u r r i cular d i f f e r e n t i a t i o n on avera ge class size.
relation c o effic ients

indicated

that enro llment more than

number of courses or course sections was related
erage class size.
low coefficents;

R e g r ess ion analy s i s resulted
however,

Cor­

to a v ­
in very

the influence of the enrollment

coefficient was greatest due to the larger

standard d e v i ­

ation in enr o l l m e n t size than in course or section number.
Some favorable influe nce on average class size may be
realized without a s ubs tantial i n flue nce on the studentfaculty ratio.
section may
affectin g

Greater eff iciency in enr o l l m e n t per course

be present in high enr o l l m e n t colleges without

the stud e n t - f a c u l t y

class size,

ratio even though m a x imum

like the s t u d e n t - f a c u l t y ratio,

may be defined

by co l l e c t i v e bargaining agreements.
T he mean stu d e n t - f a c u l t y ratio was 17.30:1
with less than 2500 FYES,

18.61:1 in col leges

in colleges

between 2500

104

and 6000 FYES,
6000 FYES.
was

found

further

1 7 . 23 :1

in c o l l e g e s w i t h g r e a t e r

On the ot h e r hand,

the mea n a v e r a g e

in c o l l e g e s

than 6000 FYES.

the c o n c l u s i o n s s t a t e d

No e v i d e n c e was fo u n d in s u p port
that I n c r e a s e s
increases

in d i f f e r e n t i a t i o n

These

and

fin d i n g s

below.
of the h y p o t h e s i s

result

in d i s p r o p o r t i o n a t e

in u n d u p l i c a t e d h e a d c o u n t in r e l a t i o n

year-equated

than 2500

b e t w e e n 2500 and 6000 FYES,

in c o l l e g e s g r e a t e r
support

than

class size

to i n c r e a s e fro m 15.49 in c o l l e g e s less

FYES to 16 . 0 8
to 19.56

and

to full-

students.

Conclusions

T e s t s of the i n f l u e n c e of c u r r i c u l a r d i f f e r e n t i a t i o n
indi cate

that

influenced

the s t u d e n t - f a c u l t y

by i n c r e a s e d

curricular differentiation.

age class siz e i n c r e a s e s wi t h
it is not

significantly

tiation.

The s t u d e n t - f a c u l t y

m ay be i n f l u e n c e d

r a t i o is not s i g n i f i c a n t l y

increases

influenced

Aver­

in en r o l l m e n t ,

and

by c u r r i c u l a r d i f f e r e n ­

ratio and a v e r a g e class

by r e g u l a t o r y c o n s t r a i n t s

size

of c o l l e c t i v e

bargaining.

Hypothesis

VI

Findings

In t e s t i n g the i n f l u e n c e of e n r o l l m e n t
s o u r c e s on a d m i n i s t r a t i v e

salar y

and r e v e n u e

expenditures,

erty tax r e v e n u e was s i g n i f i c a n t l y

only

c o r r e l a t e d with

prop­

105
a d m i n i s t r a t i v e salaries.

R e g r essi on anal y s i s resulted

in

the id en t i f i c a t i o n of both property tax and state aid as
the variables acc o u n t i n g for the greates t inf l u e n c e on a d ­
m i n i s t r a t i v e salaries.

En rollmen t and tuition variables

were not influential.
These

findings suggest that college d i stri cts with

high property tax revenues are likely to have eithe r higher
paid a d m i n i s t r a t o r s or more a d m i n i s t r a t o r s than d istricts
with low property tax.

This ob s e r v a t i o n holds true

regardle ss of col lege enrollme nt size.
C o r r e l a t i o n c o e f f i c i e n t s indicate a significant
rela tionship

between property

tax revenue

(r«.69) and

a d m i n i s t r a t i v e sal aries

but virtua lly no re lati o n s h i p be­

tween state aid revenue

(r=.01) and a d m i n i s t r a t i v e salaries.

Co e f f i c i e n t s of d e t e r m i n a t i o n in the re gressio n equation
attr ibute 48% ( r 2=,48) of the varia nce in a d m i n i s t r a t i v e
salaries to property

tax revenue and 60%

(R^e.hO)

property tax and state aid revenue combined.
7% ( r 2=.67)

is at tribute d

to

An additional

to the enr ollment variable,

but

the tuition variable is not admitted to the equation.
lege enr ollment

size is not a strong influence on a d m i n i ­

strative salaries.

The enroll ment c orrela tion coe ffi c i e n t

is ra -.19 and the reg ression c oef ficient is
standard error of

.007.

On the other hand,

co e f f i c i e n t s for both property
are significant;

Col­

.02 with a
regr ession

tax and state aid revenue

each exceeds three times its standard error.

106

Conclusions

Administrative
student

deduction
The

two d e d u c t i o n s

gross

In view

of one mill

per c r e d i t

to d e t e r m i n e

First,
times

hour

with high

greater
aid.

revenue

T h e r e fore,

college revenue
a nc e

in the

v a l u a t i o n are

b e n e f i t is
per credit.
if

pr operty

lost

state
from

equalized

local

rests

realiz ed

a

is also
by these
of

to r e c e i v e

variable

tax r e v e n u e
lower

taxes

is very
rates

colleges

is low,

substantial

budget.

Dis­

than f r o m state

for g r e a t e s t

property

in t u i t ion

low

v a l u a t i o n may r e a l i z e

property

in the

Consequently,

if t u i t i o n is

Seco nd,

only d i s t r i c t s with

l ik ely

the p o t e n t i a l

tuition

the state

the state aid sha re

s tate aid as a p r o p o r t i o n of the o p e r a t i n g
tricts

the

need.

of these d e d u c t i o n s

equalized

by c o l l e g e

o p e r a t i n g need is a d j u s t e d

in o r d e r

the g r o s s o p e r a t i n g

factors.

the c o l l e g e di s t r i c t .

of $ 2 5 . 0 0

be a c ­

in the c u r r e n t state

o p e r a t i n g need

by two e q u a l i z a t i o n

v a l u a t i o n of

requir ed.

s tate

The gross

requires a deduction

equalized
tuition

to this c o n c l u s i o n may

by p r o v i s i o n s

formula.

is a d j u s t e d
formula

l e a ding

for in part

funding

r e s o u r c e s r a t her

enrollment.

The r e s u l t s

aid

per f u l l - y e a r - e q u a t e d

vary as a f u n c t i o n of a v a i l a b l e

than c o l l e g e

counted

expenditures

v a r i a n c e in

tax v a riable .

Vari­

small

since

abo ve

or below $ 25 .00

r e c e i v e more

and while

litt le

state aid

state aid may

than $ 2 5 . 0 0 per credit,

be

no state

107

aid is gained if tuition

is grea ter than $25*00.

Hyp o t h e s i s VII

Findings

D ata ava i l a b l e from the A c t i v i t i e s Clas s i f i c a t i o n
Structure are not su fficien t
physical

to test the r e l a t i o n s h i p of

plant o p e r a t i n g e xpend i t u r e s and enro llment and

personnel size.

Several

important c h a r a c t e r i s t i c s a s s o c i ­

ated with phy sical plant operation need to be identified.
Is the c o l l ege a m u l t i - c a m p u s college?
from one or many

Does it operate

buildings? What p er c e n t a g e of exterior

surfaces are glass?

What geograph ic

such as v a r i a t i o n s in snowfall?

influences are present

Are m a i n t e n a n c e intensive

units such as swimming pools present?

Each of these c h a r ­

a c t e r istics may acc ount for i mpo rtant differ ences in
physical

plant opera ting

expenditures.

Conclusions

The simple test of c o mpa ring
total number of users

to physical

tures was inconclusive.
however,

It was

total square feet and
plant operating e x p e n d i ­

found in hyp othesis

I,

that increases in the support personnel component,

which was limited

to physical

in d e p e n d e n t operators,
in enrollment.
id entifie d

plant p e r sonnel and

were dire c t l y

related to i ncreases

R e g r e s s i o n a n a l y s i s in hy p o t h e s i s VII also

the user var iable as the more influential.

108

A p p arentl y physical plant opera ting e x p e n d i t u r e s are linked
more strongly to users
plant;

however,

than fixed features of the physical

further r e s e a r c h wo uld be necessa ry

to s u p ­

port this observation.

Major Outcomes

The results of this study provide strong support for
the the orems of Blau

(1973).

Due to l i m i t a t i o n s in p o p u ­

lation size and the pres ence of outl iers p a r ticular ly in
colleges with e n r o l lments greater
equated students,
however,

than 10,000 f u l l-year -

con c l u s i v e evide nce is not provided;

no evid ence in c o n t r a d i c t i o n of Blau's theories

was found.
Declin ing rates of in crease

in some personnel c o m ­

ponents were obser ved as o r g a n i z a t i o n s increased
Administrative
sion until

p rofe s s i o n a l s

in size.

increased in a linear p r o g r e s ­

e n r o l l ments in excess of 10,000 f u l l - y e a r - e q u a t -

ed st udents were reached.

The clerical comp onent was

clearly def ined in a linear pr o g r e s s i o n until enrollmen ts
exceeded

5,000 f u l l - y e a r - e q u a t e d students.

are co n s i s t e n t w ith
o r g a n i z a t i o n size,
rates

(Blau,

the theory

These results

that with increas es in

d i f f e r e n t i a t i o n increases at declining

1973).

While e x e u c t i v e a d m i n i s t r a t o r s r emain relatively
constant with inc r e a s e s in c o l lege size,
sonnel co m p o n e n t s

increase.

p r o f e ssional s are strongly

all other p e r ­

Increases in a d m i n i s t r a t i v e
related

to i n creas es in faculty

109

and support personnel.

Increases in the divis ion of labor

remain constant in faculty and support personnel com ponents
while increases in a d m i n i s t r a t i v e pr ofessional and clerical
comp onents are less clearly defined in large colleges.
Even though support was found indicating

that a d m i n i s t r a ­

tive salary e x p e n d i t u r e s are dependent on a v ail able re­
sources,

a d m i n i s t r a t i v e e x p e n d i t u r e s per f u l l - y e a r - e q u a t ­

ed student are less in high enroll ment c olleges
low enrollment colleges.

than in

These results are c onsistent

wit h the theory that with increa ses in o r g anizati on size,
a d m i n i s t r a t i v e ratios are reduced at declining rates
(Blau,

1973).

Increases in college size result in economy of scale
benefits.

Oper a t i n g e x p e n d i t u r e s per f u l l - y e ar -equate d

student are less in large colleges than small colleges.
Decr e a s e s in the a d m i n i s t r a t i v e e xp enditur e per student are
most r esponsi ble for increased

operating

efficiency

spite of increases in the number of a d m i n i s t r a t i v e
fessionals.

in
pro­

Economy of scale is not diminished as a

co n s e q u e n c e of increa sed d i f f e r e n t i a t i o n in curriculum.
In fact,

aver age class size increa ses with increases in

en r ollm ent even though curri cular diff e r e n t i a t i o n as
measured

by the number of different courses offered may

vary greatly.
unaffected

In ad dition the stud e n t - f a c u l t y ratio remains

by curricul ar dif ferentiation.

Th ese findings are consi stent in part with those of
Stommel

(1985)

in his study of land grant institutions.

110

While little compar ison
possible,

between the study pop ulation s Is

econ omies of scale were found in both u n ivers ities

and comm unity colleges.

S ubstant ial economies of scale

are ass ociated with a d m i n i s t r a t i v e e x pendit ures while
inst r u c t i o n a l e x pe nditure s show little sign of decline with
increases in size.

Larger instruc tional and research e x ­

pend itures in the university graduate level may drive op­
erating e x p enditur es per student upward,
of this specialization,

but in the absence

operat ing e x p enditu res per student

decline in the university as they do in community colleges
with increase s in size.

Some support for the U-shaped

curvil inear

between size and economy of

rel ationsh ip

scale found in Stom mel's study of land grant Institut ions
is provided in this study;

however,

l im itation s in the

small populat ion of this study render this comparison
inconclusive.
This res earch tested
that

the theories of Blau by proposing

the notion of economy of scale would not be satisfied

when the effects of dif f e r e n t i a t i o n in personnel co mponents
beyond execu tive a d m i n i strator s wer e investigated.
increases in the total adm i n i s t r a t i v e component,
ical component,

With

the c l e r ­

and the support component efficienc y would

diminish and economy of scale would not be realized.

Tra­

ditional efficiency measu res of s t u den t-facul ty ratio and
a verage class size would
ricular d ifferentiation.

be negative ly influenced

by c u r ­

Based on the results of this study,

these hypot heses are rejected.

Ill
Beyond these c o nclusio ns are major o utc o m e s not
a nticipated

in the design of this study.

Col l e c t i o n of

p e r f o r m a n c e data in the format of the Ac t i v i t i e s C l a s s i f i ­
cation Structure

provides a useful

data base for ranking

college p erf ormance for com p a r i s o n purposes.
hand,

On the other

the l i m i t a t i o n s of ACS data are such that little can

be det e r m i n e d to ac count for the r esulting co mparative
rankings.

College size is an important c h a r a c t e r i s t i c a c ­

c ounting for some co m p a r a t i v e rankings;
c onstr aints not present

however,

operating

in ACS data limit the expla natory

power of the data.
The community college o r g a n i z a t i o n exists in a network
of constraints.
programs.
tices.

D e m ograph y controls college

C o l lectiv e b a r gain ing regulates

Local

population and

p ersonnel

prac­

tax bases are highly correl ated with d i s ­

cr etionar y resources,

and at the state level provisions

of the funding formula direct college a d m i n i s t r a t o r s in
their d i rect ion of college
state support.
ACS data.

performance to gain m a x i mum

Few of these cons traints are recognized in

All of these c onstra ints influence college

performance.

R e c o m m e n d a t i o n s for Practic ing A d m i n i str ators

P r a cticin g a d m i n i s t r a t o r s need

to be aware of ope r ­

ating cons traints and resource dependencies.

In the r e ­

source de p e n d e n c e model of P f e f f e r and S alancik
ad ministr ator has two roles.

First,

as an

(1978)

"active

the

112

m a n i p u l a t o r ” the a d m i n i s t r a t o r must influence the e n v i r o n ­
ment in w ays that may

benefit the org anization.

as a " p r o c e s s o r ” the a d m i n i s t r a t o r must adjust

Second,
the o p e r a ­

tion of the o r g a n i z a t i o n to comply with the const raints
imposed.
In this latter role the a d m i n i s t r a t o r must accu rately
assess

the college district demography.

high school completi on projections,

In add ition to

knowledge of levels of

acad emic prepar ation in the adult popul ation is important
in iden tifying needed programs and

services.

The numbers

of unemployed and their skill levels identify base points
for retr aining programs.

Know ledge of business and i n d u s ­

trial employm ent needs and hiring p r actice s is essential if
programs are to match training with successf ul

placement.

To what extent is the local economy dependen t on agriculture,
retail sales,

resort activities,

the local quality of life?

and

basic services?

What

Are cultura l programs needed

enhance the quality of life?

What leisure

is

to

time act ivities

and person al interest learning o p p o rt unities are needed?
Direction for a d m i n i s t r a t i v e
lie in de m o g r a p h i c data.

decisions regarding

these issues

So do the constraints.

One of the most severe c onstrai nts found in this study
is the influence of the local
high property

tax base.

D i str icts with

tax r evenues have greater d i s c r e t i o n a r y re­

sources than low tax revenue districts.

State formula

a djust ments for low tax di s t r i c t s are not equal
benefits in high tax districts.

to the

The college president

113
must culti vate

local support through public services,

and

at the same time p ers u a d e or lobby for l egisl ative r e c o g ­
nition and support of pubic services whether

these services

are designed for programs of local skill d e v e l o p m e n t or for
programs of i n t e rnat ional awa r e n e s s and und erstanding.
W i t h out local and state support a d m i n i s t r a t i v e creativity
or c reative leadersh ip may be limited to i n ve ntive coping
amidst re source constraints.
can be cultivated,

If local and state support

the potential for in n o v a t i o n beyond

coping is possible.
Administrators

need

to inform le g i s l a t o r s

the communi ty col lege o r g a n i z a t i o n is c ommunity
its services.
include
ing.

better

that

based in

The variety of community se rvices needed

but exceed

t r a d i tion al acade mic and career t r a i n ­

The c o mmu nity college is a college of proxim ity

depends upon access for its success.
lows public service.
access

that

Public access fol­

If services are limited,

so will

be limited.

Finally,

the practicing a d m i n i s t r a t o r must

tive to constr a i n t s imposed

by co l l e c t i v e

be s e n s i ­

bargaining.

ternal o p e r ations of the o r g a n i z a t i o n s are regulat ed
personnel prov isions
agreements.

In­
by

set forth in both formal and informal

Terras of these agr e e m e n t s can restrict i n ­

ternal resource all o c a t i o n as well as struct ural changes
aimed at impro ving o r g a n i z a t i o n a l efficiency.

Me asures

such as s tud e n t - f a c u i t y ratios and average class size vary
little as a c o n s e q u e n c e of adop ting common working

114

conditio ns within the netw ork of c ommuni ty colleges.
Changes in o r g a n i z a t l o n l structure may have to begin at
the bar g a i n i n g table.

R e c o m m e n d a t i o n s for State Decision Makers

A c t i vities C l a s s i f i c a t i o n S t ructu re data are limited
in ex p l a n a t o r y

power but strong

in d e s c r iptive power.

Col­

leges can be ranked a c c o r d i n g to p er formanc e data reported,
but little can be learned from the data to account for the
ranked differences.

This limi tation result s from the fact

that variances in the data are dependen t more on constraints
not a c c o u n t e d
organizational

for in the data than on actual di f f e r e n c e s in
performance.

On the basis of this study size cannot
causal variable,
dition.

be said to be a

but it is an ex c e e d i n g l y important c o n ­

The range of coll ege programming

is so const ricted

by size that just as some p rograms cannot be maint ained
until

sufficient size is reached,

be el iminate d once a given

other

size has

programs may not

been reached.

Local

d i s c r etionar y author ity w ithi n college d i strict s appears
by com p a r i s o n

to many states

to be liberal;

however,

in

reality a net work of c onst raints at the local and state
levels r estrict college autonomy

severely.

Many of the co n s t r a i n t s have already

been described in

terms of the college a d m i n i s t r a t o r in the previous section.
At the state level the funding formula and the Activi ties
Classification

Structur e are a d d i tional sources of important

115
operating

constraints.

The f o r m u l a d i f f e r e n t i a t e s

instructional

five c a t e g o r i e s of i n s t r u c t i o n .
tion are d e t e r m i n e d
twenty-nine
then

instruction within
This

colleges,

set based upo n a v e r a g e

as it has

been pr a c t i c e d ,

accommodated

costs.

costs reported
f u n ding

by the

l e v e l s are

Colleges

t hen p r o v i d e

set by the formula.

the c o n t i n u a t i o n of i n s t r u c t i o n
but i n s t r u c t i o n a l

in the past c a n n o t

future w i t h i n

needs not

be a c c o m m o d a t e d

in the

this c o n s t r aint.

Differentiation

presently

a p p r o p r i a t e and s h o u l d
to d e t e r m i n e what
as what

and

the l i m i t a t i o n s

p r o ce ss a c c o m m o d a t e s

based on

A v e r a g e c o s t s of i n s t r u c ­

from a g g r e g a t e

community

need

it has

contin ue;

howe v e r ,

in the f o r mula
study

Perhaps

colleges

of i n s t r u c t i o n

coul d

by p r o g r a m

is

is needed

the cost of i n s t r u c t i o n s h ould

been.

to p r o p o s e cost s

recognized

be as well

be c alle d upon

based on d e m o n ­

s t r a b l e o p e r a t i n g needs.
Further
and

r e v i e w of the i n f l u e n c e of t u i ti on

tax e q u a l i z a t i o n p r o v i s i o n s

present formula
district
credit

l ittle i n c e n t i v e

to be a b o v e or

hour.

below

colleges.

A de fact o

through

s u p p o r t mu s t

is p r o v i d e d

provision

revenue

if p arity

tax e q u a l i z a t i o n ,

be h i g h e r

than

In the

for a c o l l e g e

r e s ults

l e v e l s amo ng

t u i t i o n rate is set

In the case of tax e q u a l i z a t i o n ,
is s o ught

needed.

the $ 2 5 . 0 0 d e d u c t i o n

The e ffec t of this

l i t t l e v a r i a n c e in t u i ti on

is also

deductions

in very

community

by the formula.
among c o l l e g e s

guaranteed

the f o r m u l a

per

l evels of

presently

provides.

116

ACS data p r o v i d e a basis
however,

since

categories
subject

for p e r f o r m a n c e c o m p a r i s i o n s ;

these c o m p a r i s o n s

of f u n ding

are used

and l evels

of fundin g,

to p r o b l e m s of m e a s u r e m e n t error.

w ill s t r i v e
funding

to r e p o r t

benefits.

performance

credibility
however,
continued
State
ACS data.

favorable

to r e a l i z e m a x i m u m

to the e x t e n t

d e c i s i o n m a k e r s must
T he c o m m u n i t y

college

ACS data al o n e do not

and

only a l i m i t e d

of fundi ng can be d e d u c e d

basis

Finally,

access

was c o m plete d.

to ACS da t a should

recog­

colleges

based needs

also

of
and

be recogn ized.

be broadened.

to the date

that a s u c c e s s f u l

R e s e a r c h e r s and c o m m u n i t y

d e v e l o p i n g an i n t e r a c t i v e
promises

levels

t welv e m o n t h s from the date a c cess

have d i r e c t a c c e s s

high p riorit y.

Further

Community

c o n s t r a i n t s must

study was d e l a y e d

to data w a s g r a nte d

for a p p r o p r i a t e

p a r t i c u l a r l y among

2500 F Y E S , is needed.
bargaining

state funding.

from ACS data.

nition of size c o n s t r a i n t s ,

in a n e t w o r k

p r o v i d e the data

equity

and a c c e s s

condition,

in the use of

is e m b e d d e d

Beyond

s h ould

T hi s

be c a u t i o u s

to s o l v e the p r o b l e m of e q u i t a b l e

This

that

than h aving no data,

n eeded

collective

guidelines

c o l l e c t i o n of ACS data is r e c o m m e n d e d .

of c o n s t r a i n t s .

than

are

Organizations

of the ACS da t a is qu e s t i o n e d .

is more

the data

I n t e r p r e t a t i o n of r e p o r t i n g

may vary f r o m c o l l e g e to c o l l e g e

less

to d e t e r m i n e

to ACS data.
s y s t e m for

to r e l i e v e this

Decision making

data run

college officials

Current

streamlined

progress

in

data input

p r o b l e m and dese r v e s

p r o c e s s e s at the local

117

community college level as well as at the state level will
be greatly enhanced with c o m p a r a t i v e data from sister
i n s t itution s readily available.

Rec o m m e n d a t i o n s for Further Study

This study does not purport to be an effe ctivene ss
study.

The results are limited to co n c l u s i o n s regarding

the influence of struc tural ch aracte r i s t i c s of organization
size on measu res of effi ciency and economy of scale.
ertheless,

some links

between

vations on quality should

Nev­

these c o nclusio ns and o b s e r ­

be noted.

Wit h i n creas es in college size curri cular d i f f e r e n ­
tiation also increases.

The range in number of different

courses offered varies from a low of 242 to a high of 2933.
With more course options,

overall quality of programs a-

vailable may be enhanced

by the inclus ion of courses that

could not be supported in low enrollment colleges.
vanced

study options at the sophomore level may

Ad­

be limited

or nonexistent in small enrollment colleges with the r e ­
sultant effect of the college
based,

one -year college.

becoming primarily a freshman

Further study is needed

to d e ­

termine the va lidity of this observation.
A second con nection

to quality may exist in the f i n d ­

ing that a d m i n i strativ e salary e x penditu res are dependent
more on available resources

than enrollment.

If high tax

revenue districts have high a d m i n i s t r a t i v e salary e x p e n d ­
itures due to added

professional s providing more support

118
services,

the quantity of instructional

student services

provided may affect the quality of student

support activities.

Some support for the position that

this personnel component
found

in hypoth esis

decrea ses

support and

I,

increases with enrollment was
Hypothe sis III demons t r a t e d

that

in o perating expend itures per f u l l - y e ar-e quated

student are
e x pendi tures

Influenced most
per student.

by decreas es in a d m i n i stra tive
Therefore,

if redu ctions in

a d m i n i s t r a t i v e expen d i t u r e s per student are realized
co n j u n c t i o n with i ncreases in the a d m i n i strativ e
group,
iture

it is reasonable

to suspect

for a d m i n i s t r a t i v e

salaries

tricts is due to increased

in

personnel

that the higher e x p e n d ­
in high tax revenue d i s ­

personnel

rather

than higher

wages.
On the other hand,

high

tax revenue districts may be

a s s ociate d with high adm i n i s t r a t i v e e x pendi tures due to
higher salaries

rather

zations are connected

than more

personnel.

All o r g a n i ­

to e n v i r onmenta l dependencies,

in the context of resource dependence models
resources

linked with

resources may lead
pr ogram support.

local autonomy

to dispar ities

and

local district

in the al l o c a t i o n of

in wages as well as

If student populatio ns are to be served

uniformly or acc o r d i n g to need,

more study

is needed

to

d eterm ine appropr iate equali z a t i o n factors in state
appropriations.
The small population of this study makes conclusions
regard ing large colleges

problematic.

Only eight of the

119

tw e n t y-nine colleges have en r o l l m e n t s above 6000 fully e ar-eq uated

students and only four have enroll ments

above 10,000.
c olleges

(Fewer than

in the United

6000 FYES.)

Linear

20% of the public community

States have en r o l l m e n t s grea ter than

r e l a t ion ships are clearly evident in

colleges

less than 6000 students;

of cases

beyond

however,

this level deviate

gres sion with greater

frequency and may be i n d i c a t i v e of

cu r v i l ineari ty were observed,

would

Some tend encies

but instances are

to be concerned

associat ed with the r e por ting format required
Act i v i t i e s C l a s s i f i c a t i o n Structure.
discussi on of hypothe ses II and III,
to a c c o m m o d a t e

Usef ulness of ACS data would
ificity

in salary

thermore,

Influenc es

should

As noted in the
some m o d i f i c a t i o n of
the data source.

be enh anced

if greater spec-

.

Fur­

to be aware of the limitatio ns

be included

in economic and

study is needed

in future researc h designs.

college o r g a n i z a t i o n is deeply

political constraints.

to examine more closely

of scale in communi ty colleges.
that

in the

C o n s i d e r a t i o n s of e n v i r o n m e n t a l and political

The M i c h i g a n co mmunity
e mbedded

for problems

e xpen diture data were present.

researc hers need

of ACS data.

basis

large college trends.

Futu re r esearch ers need

variab les was required

toward

too few to

Further research on a longi t u d i n a l

help to clarify

number

from the line of re­

the presence of h e t e r o s c e d a s t i c i t y .

be definitive.

the small

the effects

Further

the issue of economy

Thi s study dem onstrat es

of scale infl uence more o r g a n i zation al

120

characteristics
Pres ently

than a d m i n i s t r a t i v e overhead alone.

the funding formula for Michig an community c o l ­

leges co ntains only one compo nent
of scale.

in r e c o g n i t i o n of economy

Importan t va riations in cur ricular

ation are not considered.
with physical
factors need

Too many unkn o w n s are a s s o ciated

plant operations.
to be re-examined.

Local

support

study of economy

tax equa li z a t i o n

If decision makers at the

state level seek to establi sh a funding
equita ble

differenti­

formula

for colleges of varying size,
of scale in Mic higan public

leges is recommended.

that provides
further

community c o l ­

APPENDICES

APPENDIX A

M I C H I G A N P U B L I C C O M M U N I T Y COLLEGES,
E N R O L L M E N T AND P E R S O N N E L

Thi s
colleges

a p p e n d i x li sts
included

the M i c h i g a n

in this

study,

their

by fi sc al year 1984 f u l l - y e a r - e q u a t e d
full-time-equated

(FTE)

pe rs onn el .

121

public

c o mm un it y

size as m e a s u r e d
s t ud en ts

(FYES)

and

122
FYES

FTE

Alpena C om mun it y College

1419

115.4

Bay de Noc Com m un it y College

1262

114.1

C.S.

6413

668.1

6840

724.2

753

73.5

Gog ebi c Commun it y College

1179

107.4

Grand Ra pids Junior Co ll ege

6428

436.0

Henry Ford Com mu n it y College

7958

672.4

Highland Park Co mmu ni ty College

2203

181.4

Jack son C om mun it y College

3676

372.7

Kalam azo o Va lley Com mu n it y College

3971

353.2

Kello gg C om m u n i t y College

2764

219.2

Kir tland Co mmu ni ty College

1004

125.7

Lake Mi ch i ga n C o m mu ni ty College

1680

197.9

Lansing C o m mu ni ty College

10,765

1111.5

Ma comb Co mmu ni ty College

14,606

1283.0

1180

128.2

1723

176.0

M o n t c a l m Co mm u ni ty College

785

78.1

M u sk eg on Co mm un it y College

2741

279.0

885

87.6

2489

260.3

15,831

967.2

2488

252.6

5155

527.2

C o mmu ni ty College

Mott Co mm un it y College

Delta College
Glen Oaks C o m mu ni t y Co ll eg e

Mid Mi ch ig an C omm un ity

College

Monroe Count y C om mun it y College

North Central Com mu ni ty College
No r th we st er n Mic h ig an College
Oakland Commu nit y College
St.

Clair Com m un it y College

Sc hoo lcr af t Colle ge

123
Co mm uni ty Colle ge

FYES

FTE

So ut hwe st ern Mi ch i ga n Coll ege

1819

170.9

W a s h t e n a u C o m mu ni ty Co ll eg e

4581

468.3

10,917

805.0

760

101.3

Wayne County Com mu ni ty Col leg e
West Shore Co m m u n i t y Colle ge

APP EN DI X B

OP E R A T I O N A L DE FI NI TI ON S AND
BASIC ST AT IS TI CS OF VARIABLES

Thi s appendix
all

provides the op e ra ti on al

primary va ri ab le s used

De pa rt me nt
(ACS)

The ab br ev i at io n

St ruc tu re

provided as i nd ic ato rs of the variables.
for the var iab le

the mean and standard

is provided.

Secondary

is co nta ine d in p a r e n t h e ­

devia tio n for each variable

variab les

of primary varia ble s are defined
pot heses

The Mi ch ig an

of Educat ion Act ivi ty Cl a s s i f i c a t i o n

codes are

ses and

in this study.

de fi nit ion s of

based on s t a nd ar di za ti on s
in the co nte xts of h y ­

in which they were derived.

124

Mean

F u l l - Y e a r - E q u a t e d Students
measu re

(FYES)

A

Standard
D e via ti on

4285.34

4171.74

216.59

180.40

6.10

1.88

58.40

60.60

of colle ge size based on the

total credit hours gener ate d divided
by a f u l l - ti me - st ud en t
31 credit hours.

equivalent of

Indicator:

ACS 6

Item 7
Full-Time-Equated
A measure

positions

(FTE)

of college size based on

full-t im e equ iva le nts

of full and

part-t im e employ ees includ ing
on sabbatical.

Indicator:

those
ACS 4

Item 36
Ex ec ut iv e A d m in is tr a to rs
The
tors

number

(EXEC ADM)

of execu tiv e a d m i n i s t r a ­

listed in the HEP 1985 Higher

Edu c at io n D i r ec to ry wit h the title
Chief Ex e c u t i v e Officer,
Vice President,

E xe cu tiv e

Chief Academic Officer,

Chief B us in es s Officer,
dent Life Officer,

Chief of St u ­

and Deans.

A d m i n i s t r a t i v e P r o f e s si on al s
(ADM PROF)

This number has

been de te rmi ne d by subtrac tin g
the number of ex ecu ti ve

126

Mean

a dm i n i s t r a t o r s

(EXEC ADM)

co mbi ned totals

St andard
De vi at io n

from the

re ported in ACS 4

Item 30 I ns tr u ct io na l Assistance,
Item 32 A d m i n i s t r a t i v e / S u p e r v i s o r /
Technical,

and Item 33 C ou ns elo rs /

Librarians.
5.

Fac ult y
sonnel

(FAC) Total

teaching

per-

in full-time equ iva le n ts

216.59

180.40

70.02

78.31

in­

cluding ACS 4 Item 27 Full-time

6.

Faculty,

Item 28 Pa rt- ti me Faculty,

and Item

29 Ov er l oa d Faculty,

Cl eri ca l

personnel

(CL)

This

number is dete rmi ne d by sub­
tracting from

the total reported

in ACS Item 34 O ff ic e / C l e r i c a l / T r a d e
those employ ee s
Physical

in ACS Unit 7.00

Plant and those in ACS Unit

9.00 In dep en den t Operations.
7.

Support

personnel

of phy sical
ope rator

(SU)

The number

plant and in dep end en t

personnel.

Indicator:

ACS 4 Item 34 less clerical
personnel.

(CL)

29.34

26.95

127

Mean

8.

In st ruc ti ona l Ex p end itu re s

(I)

Total salaries and benefits
faculty.
1.00
9.

Indicator:

6,048,584

5, 44 6,8 06

1,531,376

1,624,145

1,359,325

1,252,198

1,393,477

1,406,794

1,066.36

253.78

for

ACS 3 Unit

Cl + C2.

Ins tru ct ion al
tures

(IS)

benefits
port

Standard
D ev iat io n

Support Ex pen d!-

Total

for in st ruc ti ona l

personnel

deans,

salaries and

(instructional

directors,

librarians,

s up ­

coordinators,

etc.)

Indicator:

ACS 3 Unit 4.00 Cl + C2.
10.

Student Ser vices E xp end it ure s
(SS)

Total

salaries and ben e­

fits for student
sonnel.
Unit
11.

services

Indicator:

per­

ACS 3

5.00 Cl + C2

I ns ti tu tio na l Adm in is tr at i on
(IA)

Total

salary and b e n e ­

fits for in st itu ti ona l a d m i n i ­
stration

(president,

trustees,
Indicator:

business
ACS 3

board of

office,

etc.)

Unit 6.00

Cl + C2
12.

Total A dm i n i s t r a t i v e Exp en diture

(TAE)

Combined

salaries

128

Mean

Standard
Devia tio n

and benefits of ins tr uct ion al
support,

student

services,

and in st it ut ion al a d m i n i s t r a ­
tion

per FYES,

Indicator:

Total of varia ble s 9,

10, & 11

divided by FYES
13.

Operating Ex pe nd it ur es
(0E)

Total

14.

by FYES,

Indicator:

3 Item 08 minus ACS 3

15.

5,583,249

1,730,354

1,566,688

4502.63

4345.36

ex­
ACS
Unit

Cl + C2

Physical Plant Operat in g Expenditure
tures

6,588,160

Indi­

Ex pen di tur es

Total fiscal year

penditures.

1.00

436.60

ACS 3 Item 08

N on in st ru c t i o n a l
(NI)

3,155,03

fiscal year e x p e n d i ­

tures divided
cator:

per FYES

(PP0) Total e x p e n d i ­

for operation

of the

physical plant inclu din g en­
ergy expenditures.

Indicator:

ACS 2B Item 110 and ACS 7 Item
066
16.

Total Institution Popu la tio n
(PER) Com bi n ed FYES and FTE,
variables

1 and 2

129

Mean

17.

Number

of Co ur se s

(CO) Total

number of differ ent
the college
or:
18.

S tan dar d
D e via ti on

689.34

530.65

2641.55

2172.04

47 ,20 0. 90

44, 555.37

18.58

3.37

16.74

3.51

courses in

curricula.

Indicat­

ACS 6 Item 09

Numbe r of Sec tio ns

(SEC)

Total number of course sections
offered.

Indicator:

ACS

6

Item 10
19.

U n d u p li ca te d H ead co unt
Total number
student
Item

20.

(UH)

of different

enrollments.

ACS 6

26

S tu d e n t - F a c u l t y Ratio
Total d iff er ent

(SFR)

en r ol lme nts

divided by total f ul l -ti meequated

faculty.

Var iab le 19 (UH)

Indicator:
divided by

Va ri abl e 5 (FAC)
21.

Avera ge Class
different
divided

(SEC)

(AC) Total

student en ro l lme nts

by the number of course

sections.
19 (UH)

Size

Indicator:

divided

Va ria bl e

by Var iab le 18

130

Mean

22.

Revenue:

P r op er ty Taxe s

St andard
D e vi at io n

3,325,366

3,204,877

4,815,169

4,320 ,36 0

4,279,464

4,168,026

464,066

361,607

6,661 ,1 16

5, 729,719

(TAX) Total reven ue gained
from In- di st rl ct
taxes.

property

Indicator:

ACS

2B

Item 101
23.

Revenue:
Total

State Aid

(AID)

revenue gained from

state appro pr iat ion s.
Indicator:
24.

Revenue:

ACS 2B Item 102

T ui tio n and Fees

(TU) Total

revenue gained

from tuit ion and
Indicator:
25.

fees.

ACS 2B Item 100

Gross Square Feet

(F2) Total

square feet in physical
interior.

Indicator:

plant
ACS 7

Item 070
26.

Gross Cubic Feet

(F3) Total

cubic feet

in phy si c al plant

interior.

Indicator;

Item 071

ACS 7

AP PE ND IX C

DATA T R A N S F O R M A T I O N S

This a p pe nd ix reports

the influ enc e

f o rm at ion s on the results of this study.

131

of data

trans­

132

The fin dings of this
data

study were tested further using

tra n sf or ma ti on s in order to adjust for problems

m u l t i c o l l i n e a r i t y and skewed distributions.
quadratic,
produce

squared,

tical tests

fi nd ing s re por te d in Chapters

a.

log,

failed to

differ enc es in the results of s ta ti s­

reported in this study.

H y po th esi s I:

Natural

and cubic tr an sf or ma ti on s

si gnificant

of

The

Further support for

IV and V are listed below.

proportion of execut iv e a d m i n i ­

strators

decreases in midsi ze and large

colleges and is highest

in small co l ­

leges .

Ana ly sis revealed a s ign if ica nt
cated by the change in
squared FYES

beta

finding affirms

as tested

(t = 4.103,

quadratic

trend

by the t test for the

df = 27,

p < ,0004).

the conc lus io n that the proportion

utive adm in is tr at or s

does decrease

colleges and is highest

in di ­

This
of ex ec ­

in midsi ze and large

in small colleg es and

proportion of execut ive adm i ni st ra to rs

that the

and FYES

is curvi-

linearly related.

Hyp o th es is

I:

b.

The

proportion of the remaining a d m i n i ­

strative compon en t Increa ses p r o p o r ­
tionately

from small to m id siz e

to

large colleges.

An al ys is re vealed a s ign ifi ca nt linear trend,
ins ig ni fi ca nt

quadratic and cubic trends.

with

These analyses

133

suggest

that the pro por ti on of ad mi ni st r a t i v e professi on als

increases with incre ase s in FYES follow ing a linear

Hyp ot hes is

I:

c.

The prop or tio n of faculty remains
constant

A signifi can t
proportion

trend.

regardless of coll ege size.

first order c or re la tio n between

of FTE faculty and FYES

(r «=* -.1111,

the

df = 28,

p < .283) sug gests the proportion of FTE faculty does re­
main constant rega rdl es s of college size.

Hyp o th es is

I:

d.

The proportion of the clerical co m­
ponent remains constant
midsize

colleges

in small and

but increases in large

colleges.

The nature

of this hypo the sis

re la tio nsh ip where the function
college size is a constant
FYES and increa ses
cor re la ti on s

for small and mids ize values of

e.

Firot

the proportion of clerical

These
personnel

of college size.

The pro portion
remains

order

of FYES

component were not significant.

constant regar dl ess

I:

personnel with

for large values of FYES.

analyses suggest that

Hypot he sis

of clerical

between squared and cubed values

with the cl erical

remains

suggests a curvilinear

of support

personnel

constant reg ard les s of college

size.

A si gnificant

first order

c orr ela ti on between the

134

propo rt ion
df o 28,

of support

p < .19)

perso nne l and FYES

suggests that

(r = - .1691,

the pro portion of support

personnel does remain co nstant re ga rdl es s of college

H yp ot hes is

II:

a.

Ins ti t ut i on al

size.

a d m i n i s tr at i ve

ex p en di tu re s dec rease as college

size

increases.
b.

I ns t ru cti on al

support and student

services ex pe nd it ur es
incre ase s
c.

in college enrollment.

The ex pen di tur e

per student

ad m i n i s t r a t i v e cost
three comp on ent s
enro llm ent

An aly sis
data set:

suggests

three

1) non- no rma l

increase with

for total

includ ing

these

is greater in large

colleges.

specific problems with the

dis tribution,

2) high m ul t i c o l l i n -

earity among i n d e pe nd en t variables,

and 3) small number

cases reducin g

Co rre l at io ns

degrees of freedom.

total a d m i n i s t r a t i v e e xp en di tu re s and squared
values of FYES were not significant,
co r re la ti on
and

of

between

and cubed

while a signif ica nt

between natural log tr ans fo rme d values of FYES

total ad mi n i s t r a t i v e

ex pen dit ur es was found

- .3792,

df a 28,

gression

for total ad mi n i s t r a t i v e e xp e nd it ur es with i n d e ­

pendent variables

p < .021).

A subs eq uen t

(r =

in log values re vealed

amount of variance was ex pl ai ne d when

stepwise

that

re­

the greatest

the log values of

FYES and i ns tr uct ion al support were entered.

Student

135

services and i n st it ut io na l a d m i n i s t r a t i o n va ria bl es were
not ente red because the pr ob ab i li ty
sa ti sfi ed

(p < .05).

One i nt e r p r e t a t i o n
that i ns tr uc ti on al
are

to enter was not

si gn ifi can tl y

expenditures,

of these analyses might

support e x pe nd it ur es
co -r el a te d to total

and c ol leg e size

a d m i n i st ra ti ve

wh ere total a d m i n i s t r a t i v e

crease with in cr e as es

suggest

in c o ll eg e size.

e xp en di tu re s d e ­
In cr ea se s in total

a dm i n i s t r a t i v e

ex p e n d i t u r e s are related more to i n s t r u c ­

tional

e xp en di tu re s

support

than are student

e xp en di tu re s or i n st it ut io na l ad mi n i s t r a t i v e
once

the effect

III:

e x p e n d it ur es

of colle ge size has been partialed

total a d m i n i s t r a t i v e

Hyp o th es is

services

out of

expendit ur es.

E xp e n d i t u r e s
large
crease

per FYES are greater

than small colleges,
is depend ent

ex p en di tu re

in

but the in­

upon no nl ns tr u c t i o n a l

i nc re ase s rather than faculty

expenditures.

All i nd ep en de nt

va riables were

ated with e xp en di tu re s

s ig n i f i c a n t l y c o r r e l ­

per student and in te r c o r r e l a t i o n s

between in d ep en de nt

variables were high.

ed an i ns ig ni fi ca nt

quadratic

student with FYES.

A stepwise

sion reveal ed

reveal­

for e x pe nd it ur es

per

univ ari at e mul tip le r e g r e s ­

that FYES ac co un t ed

tion of variance
penditures

trend

Analysis

in e xp end it ure s

for the greatest

propor­

per student and that e x ­

per student decreased as FYES

increased.

136

N on - i n s t r u c t l o n a l

ex pe nd it ur e s acc ou nt ed for the most v a r i ­

ance in e xp en di tu re s per student of all the remai nin g in ­
depende nt v ar ia ble s after FYES had been entered,
penditures

and e x ­

per student inc re a se d as no ni n s t r u c t i o n a l

ex p en di tu re s increased.

I ns tr uc tio na l

support ex pen d it ur es

acc o un te d for the greatest amount of va riance in e x p e n d i ­
tures

per student after

entered,

the first two va ri abl es

and ex pe nd itu res

st ruc tio na l

Hy po th es is

per student

had been

de cre as ed as in­

support ex pe nd itu res increased.

IV:

The s t u d e n t -f ac ul ty

ratio is indepen den t

of college size when course d i f f e r e n t i a ­
tion and number

of sec tions are taken into

consideration.

The

section variable acc o un te d for the greatest

amount

of var ia nce in the s t u d e n t - fa cu lt y ratio when courses and
sections were entered via a forward r eg res sio n method.
change in

was not si gni fi can t at the p < .05 level when

FYES was ent ere d with sec tions already

Hyp o th es is

The

V:

a.

parti ale d out.

Avera ge class size remai ns

unchanged as

en r ol lm en t Increases.
b.

In cr e as es in course d i f f e r e nt ia ti o n and
the nu mber of mu lti ple

sections offered

effects constant ave rag e
c.

U nd up li ca te d

student

class size.

headcount

creases with increased course

in­

137

d iff ere nt iat ion ,

but FYES

in pro-,.

Portion to un d u p l i c a t e d headcount
declines.

An alysis

showed an in si gni fi can t quadratic

a v er ag e class
the greatest

size with FYES.
prop ort io n

trend for

FYES was found to explain

of variance

fol lowed by sections and courses.

in av er ag e class size,
Un d u p l i c a t e d

was never entered into the equation

because

headcount

it was highl y

corr ela ted with FYES and never explained more variance in
ave rag e class
found

size than

FYES.

to inc rea se as college

Avera ge class
size Increased,

number of courses and sections

size was
and the

both affected co nstant

a v er ag e class size with the number of sections decreasing
with a v er ag e class size and
ing as aver age class

Hy p ot he si s VI:

the number of courses

increas­

size increased.

A d m i n i s t r a t i v e e xp en dit ure s

per FYES

in­

crease as a v ai la bl e r es ou rce s increase i n ­
dependent of colle ge enrollment.

None

of the ind ep end ent v ar ia ble s were

found to be

si g ni fic ant ly cor re l at ed with ad mi ni st r a t i v e ex pe nd itu res
at the p < .05 level.
order
-

Tu it io n obt a in ed

the highest first

co rr el a ti on with a dm i n i s t r a t i v e e xp e nd it ur es

.2248,

df = 28,

p C.121).

There were high i nt e r c o r ­

re la ti on s between all ind ep end ent
state aid and property

(r =

varia ble s except for

tax revenue.

The univariate

138

mul ti ple reg res si on was not attempted
si gn ifi ca nt

first order co rr el at io n s

because of the in­
between

the independent

variables with a dm i ni st r a t i v e expenditures.

Hyp o th es is

VII:

Ma i nt en a nc e and energy ex pen dit ur es are
funct io ns of physical plant size
ind ep end en t of enr ol lme nt and

personnel

size.
None of the independ en t va ria ble s were sign ifi ca ntl y
corre la ted with physical
the p < .05 level.

plant opera tin g exp en di tu re s at

High c or re lat ion s

pendent variab les were present.
an al ysi s was not condu cte d

between the in de ­

The univ ar iat e re gression

because of the in significant

first order c or rel at ion s with physical

plant operating

expenditures.

No important d iff er enc es in the major findings of
this study result as a co n se qu en ce of using data t ra n s f o r ­
mati on s .

BIBLIOGRAPHY

BI BLI OG RA PH Y

Ame ri can As soc ia tio n of Commun it y and Junior Colleges
AACJC L e t t e r * Washington, D.C.: Am er i ca n As ­
sociat io n of Co mm u ni ty and Junior Colleges,
Septem be r A, 1984,
______ .

1985 C o m m u n i t y . T e c h n i c a l . and Junio r C o l ­
lege D i r e c t o r y . Washington, D.C.:
American
A sso cia ti on of Community and Junior Colleges,

1985.

Anderson, Th eodore R. and Warkov, Seymour.
Organiza­
tional size and functional complexity:
a study of
a dm in is t r a t i o n in hospitals.
Ame ri c an Soci olo gi cal
R e v i e w . 1961, 26, 23-28.
Baker, A. W. and Davis, R. C.
Ratios of Staff to Line
Em ployees and Stages of D i f fe re nt i at io n of Staff
F u n c t i o n s . Columbus, Ohio:
Bureau of Business
Research, Ohio State University, 1954.
Bain,

Joe S.
Ec on om ie s of scale, concent rat io n, and
the condit ion of entry in twenty m an uf act ur ing
industries.
Am eri ca n Eco nomic Review, 1954,
44 (1), 15-39.

Bendix, Reinhard.
Work and Au thority in I n d u s t r y .
New York:
Wiley, 1956.
Berry, Wi ll ia m D. and Feldman, Stanley.
Multipl e
Reg r es si on in P r a c t i c e . Beverly Hills:
Sage
Publications, 1985.
Blau,

Peter M.
A formal theory of dif fe re nt ia ti o n in
organizations.
Ame ri can Soc io l og ic al R e v i e w .
1970, 35 (2), 201-218.

______. On the Nature of O r g a n i z a t i o n s .
John Wiley & Sons, 1974.

New York:

______. The Or g an iz at io n of Aca d em ic W o r k .
York:
John Wiley & Sons, 1973.
Blau,

New

Peter M. and Schoenherr, Richard A,
The
Struct ur e of O r g a n i z a t i o n s . New York:
Basic
Books, Inc., 1971.

139

140

Bogue, J. P.
The Co mmu ni ty C o l l e g e .
McGra w- Hil l, 1950.

New York:

Bowen, Howard R. and Douglas, Gordon K.
Ef fi ci en cy in
Liberal E d u c a t i o n . New York:
Mc Gra w- Hi ll Book
Company, 1971.
Brinkman, Paul T.
Ec on o mi es of scale in higher e d u c a ­
tion:
fifty years of research.
A paper prepared
for the 1985 annual conv ent io n of the National
Center for H ig he r E du cat io n M a n a g em en t Systems.
Cal i fo rn ia Co or di n at in g Council for Higher Education.
M e et in g the Enr ol lm en t Demand for Pu blic Higher
Ed uc ati on in C a l if or ni a through 1977 - the Need
for Ad di ti on al C ol le ge and U n i v er si ty C a m p u s e s .
Sacramento:
Co o r d i n a t i n g Cou nci l for H ig he r E d ­
ucation, 1969.
Cameron, Kim.
M es ur in g or ga ni zat io nal eff ec tiv ene ss
in in st i t u t i i o n s of higher education.
Admini­
strative Science Q u a r t e r l y . 1978, _23 (4),
604-629.
Carlson, D, E.
The Pro du ct io n and Cost Be hav io r of
Higher Ed uca tio n I n s t i t u t i o n s . Berkeley:
Uni­
vers it y of California, 1972.
Child, John and Mansfield, Roger.
and or g an iz at io n structure.
6, 369-393.

Te chnology, size,
S o c i o l o g y . 1972,

Cohen, Arthur M. and B r a w e r , Flo re nce B.
The
A m er ic an C o mm un it y C o l l e g e . San Francisco:
Jo s s e y - B a s s Pu blishers, 1982.
Cohn,

Elchanan.
Econo mie s of scale in Iowa high school
operations.
The Journ al of Human Resources. 1968,
3 (4), 422-434.

Dickmeyer, N.
Small i nd ep end en t col le ges and economies
of scale.
R es ea rc h in Higher E d u c a t i o n , 1982,
12, 51-67.
Fox, W i l l i a m F.
Re vie wi ng economies of size in educa tio n
Journ al of Edu ca ti on F i n a n c e . 1981, fi,
273-296.
Gasmo, Gary.
An^ Appr o ac h to Cost Studies in Small C o l ­
leges . Boulder, Colorado:
National Center for
Highe r Educa tio n M a n a ge me nt Systems, 1978.

141

G e o r g o p o u l o s , Basil S. and Ta nnenbaum, Arnold S.
A
study of or ga ni zat ion al eff ectiveness. American
So c io lo gi ca l R e v i e w . 1957, 22, 534-540.
Gold,

Bela.
Cha ngi ng pe r sp ec ti ve s on size, scale,
returns:
an i n t e r pr et iv e study.
J ou rna l of
Ec ono mic L i t e r a t u r e , 1981, XIX, 5-33.

and

Grobe, Loret ta L. and Hoot, Ronald, eds.
Mic hi g an
C o m mu ni ty C o l le ge s FY 1983-84 Ac t i v i t i e s C l a s s i f i c a ­
tion St ructure (ACS) Data B o o k . Lansing, Michigan:
De p ar tm en t of Education, 1985.
Haas,

Eugene, Hall, Rich ard A., and Johnson, Norman J,
The size of the su p po rt i ve compo nen t in o r g a n i z a ­
tions:
a m u l t i - o r g a n i z a t i o n a l analysis.
Social
F o r c e s . 1963, 43, 9-17.

Hage,

Jerald and Aiken, Michael.
R el at io ns hi p of c e n t r a l ­
iz a tio n to other str uct ur al properties.
Administra­
tive Science Q u a r t e r l y , 1967, 1_2 (1), 72-92.

Hall,

Richa rd H., Haas, J. Eugene, and Johnson, Norman J.
O r ga ni za ti on al size, complexity, and formalization.
Am eri can S oc io lo g ic al R e v i e v . 1967, 32 (6),
903-912.

Hawley, Amos H., Boland, Walter, and Boland, Margaret.
Pop u la ti on size and ad mi n i s t r a t i o n in in sti tu ti o ns
of higher education.
Am er ica n So cio lo gic al
R e v i e w . 1965, 30, 252-255.
The HEP, 1985 Higher Educat io n D i r e c t o r y . W a s h ­
ington , D.C.:
Hi ghe r E duc at ion Pu blications,
Inc., 1985, pp. 152-163.
Hinkle, Dennis E., Wiersma, William, and Jurs, Stephen
G . A p pl ie d St at is ti cs for the Beh avi or al
S c i e n c e s . Boston: Hou ght on M if fli n Company, 1979.
Holdaway, Ed wa rd A. and Blowers, Th om as A.
A d m i n i st ra ti ve
ratios and or ga n iz at io n size:
a l on gi tu di na l e x a m ­
ination.
Am eri ca n S oc io log ica l R e v i e w . 1971,
36, 278-286.
Jenny, H. H. and Wynn, R. G.
The Turn ing P o i n t : A
Study of Income and E x p e nd it ur e Growth and D i s t r i ­
bution of Fo r t y - e i g h t P ri vat e Four Year L ib era l Arts
Co lle ges 1 9 6 0 - 1 9 7 0 . Wooster, Ohio:
The Colle ge of
Wooster, 1972.

142
Kasarda, J. D.
The st ructural im pli c at io ns of social
system size:
a th re e-l ev el analysis.
Ame rican
So c io lo gi ca l R e v i e w . 1974, ^9, 19-28.
Kerlinger, Fred N.
F ou n d a t i o n s of Beha vi ora l R e s e a r c h .
New York:
Holt, Ri nehart and Winston, Inc., 1973.
Kiesling, Herbe rt J.
M ea su rin g a local government
service:
a study of school distr ict s in New York
state.
Re view of Economics and Statistics, 1967,
49, 356-367.
Kimberly, John R.
Or ga ni za t i o n a l size and the s t r u c ­
turalist perspective:
a review, critique, and
proposal.
Ad m i n i s t r a t i v e Science Quarterly,
1976, 21 (4), 571-597.
Lawrence, Ben, Weathe rsb y, George, and Patterson, Virginia
W. eds.
Outp uts of Hi gh er E d u c a t i o n : Their
I d e n t i f i c a t i o n . M e a s u r e m e n t . and E v a l u a t i o n .
Boulder, Colorado:
Weste rn Int ers tat e C o m m is si on for
Higher Education, 1970.
Lerner, Joseph, Haaland, Gordon A. and Kraus, John D.
The
use of the student credit hour to quantify higher e d u ­
cation productivity:
some new techniques.
Planning
for H ig h er E d u c a t i o n . 11 (4), 10-12.
Lewis-Beck, Micha el S.
Ap pl ie d Reg re ss io n An I n t r o ­
duction . B ev erl y Hills:
Sage Pu blications, 1980.
Mach ad o Da Silva, Clovis Luiz.
O rg an iz at io n effectiveness:
a study of school di str ic ts in Michigan,
Doc t or al
dissertation, M i ch ig a n State University, 1980.
Maynard, J.
Some M ic ro ec o n o m i c s of Hi gher E d u c a t i o n .
Lincoln:
U n i v er si ty of Ne bra ska Press, 1971.
McClenney, Byron N.
M a n ag e me nt for P r o d u c t i v i t y .
W a s h i ng to n D.C.:
Am er i ca n As s oc ia ti on of Co mm un it y
and Junio r Colleges, 1980.
Mc Laughlin, Gerald M.
Size and efficiency.
Paper pre­
sented at the annual forum of the As s oc ia ti on for
I n st itu ti ona l Research, San Diego, California,
May 13-17, 1979.
Meeth, L. Richard.
San Francisco:

Qual ity Educat ion for Less M o n e y .
Jo ss ey -B ass Publishers, 1974.

143
Melman, Seymour.
The rise of a d m i n i s t r a t i v e o v e r h e a d in
the m a n u f a c t u r i n g i n d u s t r i e s of the U n i t e d States,
189 9-1947.
Ox fo rd E c o n o m i c P a p e r s . 1951, 3_,
62-112.
Metz,

G. E.
C ur r e n t Fund E x p e n d i t u r e s . Atlanta:
C o m m i s s i o n on Colleges, S o u t h e r n A s s o c i a t i o n
C o l l e g e s and Schools, 1964.

of

Meyer, M a r s h a l l W.
Size and the s tr u c t u r e of o r g a n i z a ­
tions:
a ca u sa l a na lys is .
American Sociological
R e v i e v . 1972, J37, 434-441.
Norusis, M a r i j a J.
Chicago :
SPSS

SP S S / P C for
Inc., 1984.

the IBM P C / X T .

O'Neil, June.
R e s o u r c e Use in H i g h e r E d u c a t i o n :
T r e n d s in Ou tp ut s and I n p u t s . 1930 to 1 9 6 7 .
C a r n e g i e C o m m i s s i o n on Hi gh er Ed uc at ion , Be rkeley,
Ca l if or ni a, 1971.
Osborn, Ri chards, C.
E f f i c i e n c y and p r o f i t a b i l i t y in
r e l a t i o n to size.
H a r v a r d B u s i n e s s Review, 1951.
29, 82-94.
Par k in so n, C. N.
P a r k i n s o n 1s Law and Othe r St u d i e s
A d m i n i s t r a t i o n . Ne w York:
Bal la nti ne , 1957.

in

Perrow, Ch arles.
C om p l e x O r g a n i z a t i o n s . G le nv iew ,
Illinois:
Scott, Fo r e s m a n and Co., 1979.
Pfeffer, J e f f r e y and Salancik, G er al d R.
The E xt e r n a l
C o n t r o l of O r g a n i z a t i o n s : A. R e s o u r c e D e p e n d e n c e
P e r s p e c t i v e . New York:
Ha rp er & Row, P u b l i sh er s,
1978.
Price, Ja m e s L.
The study of o r g a n i z a t i o n a l ef f e c t i v e n e s s .
Th e S o c i o l o g i c a l Q u a r t e r l y . 1972, 1_3, 3-15.
Priest, Bill J. and Pi ck elm an , Jo hn E.
Increasing Pro­
d u c t i v i t y in the C o m m u n i t y C o l l e g e . W a s h in gt on ,
D.C.:
A m e r i c a n A s s o c i a t i o n of C o m m u n i t y and J u n i o r
Colle ge s, 1976.
Pugh,

D. S., Hic ks on, D. J., Hi nings, C. R .; and T u r n ­
er C.
The c o n t e x t of o r g a n i z a t i o n s t ru ct ure s.
A d m i n i s t r a t i v e S ci en ce Qu ar te rl y, 1969. 14
(1), 91-113.

______ . D i m e n s i o n s of o r g a n i z a t i o n str uct ure .
Admini­
st r a t i v e S c i e n c e Q u a r t e r l y . 1968, 3JJ (1),
65-106.

144

Ri ch ar ds on , Ro b e r t H.
A c o m p a r a t i v e a n a l y s i s of the
st r u c t u r a l c h a r a c t e r i s t i c s of school o rg a n i z a t i o n s .
D o c t o r a l d i s s e r t a t i o n , M i c h i g a n Stat e U n i v e rs it y, 1978.
Ru sh ing , W i l l i a m A.
The e ff e c t s of i n d u s t r y size and
d i v i s i o n of lab or on a d m i n i s t r a t i o n .
Admini­
s t r a t i v e S c i e n c e Q u a r t e r l y . 1967, 12^ 273-294.

______ . O r g a n i z a t i o n a l size and a d m i n i s t r a t i o n :
the
p r ob le ms of ca us al h o m o g e n e i t y and a h e t e r o g e n e o u s
catego ry .
P a c i f i c S o c i o l o g i c a l Review. 1966.
9, 100-108.
Scott, W. Ric hard.
Organizations: Rational.
N a t u r a l , and O p e n S y s t e m s . New Jersey:
P r e n t i c e - H a l l , Inc., 1981.
______ . O r g a n i z a t i o n a l st r u c t u r e .
S o c i o l o g y . 1975, JL, 1-20.
Shepherd, W i l l i a m 6.
The E c o n o m i c s
g a n i z a t i o n . E n g l e w o o d Cliffs,
Hall, 1979.

A n n u a l R e v i e w of

of I n d u s t r i a l O r ­
N. J.:
Prentice-

S i l b e rs to n, Aubrey.
E c o n o m i e s of scale in th eo ry and
p r act ic e.
The E c o n o m i c J o u r n a l . 1972, 82
(3 2 5 S ) , 369 -3 91.
Stommel, Man fr e d.
Ef fe cts and c o n s e q u e n c e s of o r g a n i z a ­
tion size:
a study of l a n d - g r a n t in st i t u t i o n s .
Doc­
toral d is s e r t a t i o n , M i c h i g a n State Un i v e r s i t y , 1985.
Ter ri en, F r e d e r i c W. and Mills, Do na ld L.
The effect
of ch a n g i n g size upon the i n t e r n a l s tr u c t u r e of
organizations.
A m e r i c a n S o c i o l o g i c a l Review,
1955, 20, 11-13.
Tolbert, P a m e l a S.
I n s t i t u t i o n a l e n v i r o n m e n t s and r e ­
so u rc e de p e n d e n c e :
s ou r c e s of a d m i n i s t r a t i v e
st r u c t u r e in i n s t i t u t i o n s of higher edu cat ion .
A d m i n i s tr at ive S ci e n c e Q u a rt er ly . 1985. 30
(1), 1-13.
Weber, Max.
E s s a y s in S o c i o l o g y .
U n i v e r s i t y Press, 1946.

New York:

Oxford

______ . T he T h e o r y of S oc ia l and E c o n o m i c O r g a n i z a ­
tion . New York:
O x f o r d U n i v e r s i t y Press, 1947.

145

Yuchtman, E ph ra im and Seashore, Stanl ey E.
A
system re source app ro ach to o r ga ni za ti on al
e ffe cti ve nes s.
A m e ri ca n So ci ol og ic a l R e v i e w ,
1967, 32, 891-903.