Research in Higher Education, Vol. 41, No. 1, 2000

COMPARING ADMINISTRATIVE SATISFACTIONIN PUBLIC AND PRIVATE UNIVERSITIES

James Fredericks Volkwein and Kelli Parmley

:::: : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : :Do administrators in public higher education experience different levels of job satisfac-tion than their counterparts in the private sector? Drawing upon the management,higher education, and public administration literature, this study examines a compre-hensive array of national data on university characteristics, state characteristics, andadministrative satisfaction. Previous analyses have explored the relationship of stateregulation to administrative satisfaction in public higher education. However, thereare few studies in higher education or public administration that have explored thedifferences in job satisfaction between the public and private sectors. We find thatthe hypothesized public/private differences are limited only to satisfaction with extrinsicrewards, and even these differences disappear when all relevant variables are con-trolled for in a regression analysis. In both sectors, job satisfaction is most consistentlylinked to work environments characterized by teamwork and low levels of interpersonalconflict.

:::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::

THE RESEARCH AND POLICY PROBLEM

Much has been written about managerial satisfaction in business and publicadministration, but most studies in higher education have examined the satisfac-tion levels of faculty rather than administrators (Austin & Gamson, 1983;Gmelch, Lovrich, and Wilke, 1984; Cotton and Tuttle, 1986; Smart, 1990;Olsen, 1993; Smith, Anderson, and Lovrich, 1995). The few studies of adminis-trative satisfaction in higher education focus primarily on understanding thenature and level of satisfaction, rather than on examining the factors producingsatisfaction and the subsequent connections to important outcomes such as turn-

This is a revised version of a paper that was presented at the Annual Forum of the Associationfor Institutional Research, Seattle, WA, May 1999.

James Fredericks Volkwein, Center for the Study of Higher Education, Pennsylvania State Uni-versity; Kelli Parmley, State University of New York, System Administration. Address correspon-dence to: James Fredericks Volkwein, Center for the Study of Higher Education, 403 S. Allen St.Suite 104, Pennsylvania State University, University Park, PA 16801-5252.

95

0361-0365/00/0200-0095$18.00/0 2000 Human Sciences Press, Inc.

96 VOLKWEIN AND PARMLEY

over and productivity, like those found in the management literature (Solo-mon & Tierney, 1977; Smart and Morstain, 1975; Blix and Lee, 1991; Glick,1992). Even so, the topic is important because many higher education scholarshave reserved an important role for student, faculty, and administrative satisfac-tion in their studies of organizational effectiveness (c.f., Cameron, 1978).

One important claim in higher education is the positive connection betweenacademic autonomy and quality (Carnegie Foundation for the Advancement ofTeaching, 1982; Gardner, Atwell, and Berdahl, 1985). The relationship betweenautonomy and effective performance is assumed to operate both at the level ofthe individual professor and at the institutional level. The autonomy, or freedomfrom regulation and control, associated with private institutions suggests thatthey generally have a distinctive and superior administrative work environment(Levy, 1992). We also know from the organizational and management literaturethat such work environments are generally associated with beneficial outcomeslike higher worker productivity and lower turnover (Tett and Myer, 1993). Thus,in this research we assume a connection between institutional autonomy, posi-tive administrative work environments, and administrative satisfaction, espe-cially among those administrators who occupy certain job titles.

While previous attempts to measure empirically the relationship betweenquality and autonomy at the institution level have proved inconclusive (Volkw-ein, 1986a,b, 1987, 1989; Volkwein and Malik, 1997), the job satisfaction litera-ture leads one to expect that autonomy may indirectly effect organizational qual-ity through gains in productivity that result from higher managerial satisfaction(Vroom, 1964; Porter and Lawler, 1968; Levy, 1992). A recent study of publicuniversities demonstrated little direct relationship between degrees of autonomyand administrative satisfaction in public universities (Volkwein, Malik, and Na-pierski-Prancl, 1998). However, the researchers found a consistent connectionbetween every measure of administrative satisfaction and the human relationsaspects of the immediate work environment. These results, however, have notbeen verified on a population of more autonomous private university administra-tors.

These findings from the higher education literature are consistent with evi-dence from public administration research indicating that public professionalsderive their job satisfaction primarily from the social aspects of their jobs, andonly secondarily from the work itself (Emmert and Taher, 1992). Public admin-istration literature also suggests that job satisfaction varies in significant waysbetween the public and private sectors (DeSantis and Durst, 1996; Steel andWarner, 1990). However, there have been no companion studies in higher edu-cation like those in the fields of business and public administration. Thus, it isboth reasonable and important to examine and compare the levels of administra-tive satisfaction in public and private universities in order to understand any

ADMINISTRATIVE SATISFACTION IN PUBLIC AND PRIVATE UNIVERSITIES 97

important differences that may exist. If there are measurable benefits from au-tonomy, they ought to be evident in the variable satisfaction levels of the admin-istrators and institutions in this study.

PURPOSE OF THE STUDY AND CONCEPTUAL FRAMEWORKS

Do administrators in public and private higher education experience differentlevels of satisfaction? Are the influences on administrative job satisfaction simi-lar or different in public versus private universities? This research examines theperceived work environments and individual characteristics of administrativemanagers in both public and private higher education. In addition to survey dataon perceived work environments, we have also incorporated university and statecharacteristics consistent with a variety of theoretical perspectives from the re-search literature: organizational theory, structural/functional perspectives, the lit-erature on university autonomy, and theories of employee satisfaction.

Job Satisfaction

At the core of this study is the job satisfaction literature. There is generalagreement in the literature that job satisfaction is multidimensional. Herzberg’sTwo Factor Theory (1966) draws our attention on the one hand to intrinsic jobcontent factors (such as feelings of accomplishment, recognition, and auton-omy), and on the other hand to extrinsic job context factors (such as pay, secu-rity, and physical working conditions). Several studies have examined the intrin-sic and extrinsic dimensions of job satisfaction in higher education (Olsen, 1993;Austin and Gamson 1993; Hackman and Lawler, 1971; Kalleberg, 1977; Haged-orn, 1994).

In addition to agreement in the literature that job satisfaction is multidimen-sional, most studies conclude that satisfaction is influenced by a complex arrayof personal and situational circumstances (Austin and Gamson, 1983; Hoppock,1977; Mumford, 1972; Bruce and Blackburn, 1992). Research has shown thatseveral work related variables exert positive and significant influences on ad-ministrative satisfaction—a supportive organizational culture, teamwork, rela-tionships with colleagues and superiors, worker autonomy, and self-fulfillment(Berwick 1992; Bensimon and Neumann, 1993; Austin and Gamson, 1983;Boone, 1987; Lawler, 1986; Rigg, 1992; Volkwein et al., 1998). Job and work-load stress exert negative influences on satisfaction and are almost always in-cluded in studies of job satisfaction (Blau, 1993; Blix and Lee, 1991; Glick,1992; Olsen, 1993; Hagedorn, 1996; Volkwein et al., 1998).

98 VOLKWEIN AND PARMLEY

Organization and Environment Perspectives

The organizational literature generally leads us to expect that an array ofcampus characteristics exert significant influences on the workplace. Perspec-tives from organization theory emphasize the importance of the organization’sstructure and its environment (Hall, 1995; Lawrence and Lorsch, 1967; Aldrich,1979; Pfeffer and Salancik, 1978). Studies of colleges and universities havedemonstrated that campus mission, size, wealth, complexity, and selectivity ex-ert significant influences, ranging from small to large, on a variety of internaltransactions and outcomes (Austin and Gamson, 1983; Hall, 1995; Pascarellaand Terenzini, 1991; Volkwein et al., 1998). Researchers and accrediting bodiesalike believe that effective organizations produce satisfied organizational mem-bers. Increasingly, administrative satisfaction is used as an indicator of organiza-tional effectiveness (Cameron, 1978).

Individual Characteristics

Consistent with research in other organizations, studies of managers in col-leges and universities suggest that a variety of personal and organizational vari-ables exert potential influences on their job satisfaction. Among these personalcharacteristics are physical and mental health (Spector, 1997), age and retire-ment proximity (Austin, 1985; Hagedorn, 1994; Lee and Wilbur, 1985; Solomonand Tierney, 1977), sex (Austin, 1985; Hagedorn, 1996), level of education(Martin and Shehan, 1989), length of service (Bamundo and Kopelman, 1980),administrative rank (Austin and Gamson, 1983), administrative area (Glick,1992), and levels of stress (Gmelch et al., 1984; Smith et al., 1995; Sullivanand Bhagat, 1992).

RESEARCH METHODS

Guided by these various perspectives and concentrating on variables poten-tially important to job satisfaction, we created the analytical database for thisstudy drawing information from a wide range of different sources. Approxi-mately half the nation’s Carnegie research and doctoral universities agreed toparticipate in the study. We collected almost 1200 survey responses from uni-versity administrators at 120 public and private institutions. We added data re-flecting campus organizational and environmental characteristics, including datafrom both the Integrated Postsecondary Education Data System (IPEDS) andguidebook sources, and then engaged in data reduction using principal compo-nents analysis and scale building techniques. The resulting variables and scales

ADMINISTRATIVE SATISFACTION IN PUBLIC AND PRIVATE UNIVERSITIES 99

form the basis for examining the dynamics of administrative satisfaction in pub-lic and private universities.

Target Population, Data Sources, and Variable Summary

This study examines the major organizational features of 120 doctoral univer-sities, their administrative work environments as perceived by 1191 respondents,and the individual characteristics of the participants in the study. On each cam-pus we directed a survey instrument to managers with 12 specific job titles(ranging from vice presidents to deans and directors)—a total of 1400 adminis-trators. This survey contains 7 questions about the respondents’ background,and 44 items assessing their satisfaction, stress, and working conditions. Usingfollow-up procedures that guaranteed respondent anonymity, we eventually re-ceived a response rate exceeding 85%.

Table 1 summarizes the variables used in this study. The first group is campuscharacteristics reflecting organizational mission, size, wealth, quality, complex-ity, and autonomy/flexibility. These are modeled after the factor analytic andscale building procedures described in Volkwein and Malik, 1997. Sources forthe data include NCES/IPEDS, the National Research Council study of doctoralprograms (1995), the Graham and Diamond national study (1996), and theguidebook information contained in Barron’s and US News. The scale of auton-omy and flexibility measures the nature of the external environment at eachinstitution. For public institutions, we use the three-category (high/medium/low)scales developed by Volkwein and Malik (1997). For the separate administrativeand academic autonomy dimensions, each campus is classified as high (onestandard deviation above the mean), low (one standard deviation below themean), or medium. This forms the first three categories of a four-category con-tinuum. All private universities are classified as falling in the fourth (or highest)autonomy category.

The second variable group includes self-reported respondent characteristics(age, ethnicity, sex, rank, administrative area, and personal and financial health).The third variable cluster in Table 1 lists their reported working conditions,teamwork, funding, facilities, work pressure, and interpersonal conflict. Theseare self-reported on a series of survey items with 5-point response scales.

Finally, the survey contains 25 satisfaction questions where responses aresolicited on a 5-point Likert-type scale from very satisfied (5) to very dissatis-fied (1). Our principal components and data reduction techniques produced fourscales in addition to the single “overall satisfaction” measure: intrinsic satisfac-tion, extrinsic satisfaction, satisfaction with working conditions, and satisfactionwith colleagues and other people one works with. The alpha reliabilities onthese four multi-item scales are shown in Table 1 for the two populations. They

100V

OLK

WE

INA

ND

PA

RM

LEY

TABLE 1. Measures Used in the Multivariate Analyses

Cronbach’sConstructs and Variables from Alpha Cronbach’sHigher Ed Regulation and Job Private Alpha Example of Key itemsSatisfaction Literature Nature of the Measure Univ Public Univ in Each Scale

Campus Characteristics These campus variables are assem-Campus Size bled from a variety of sources andCampus Wealth modeled after procedures in Volk-Faculty Quality ein & malik (1997) and VolkweinUndergrad Quality et al. (1998)Has Medical school/hospitalHas Agriculture/Land Grant statusPublic Flagship in the stateConstitutional recognitionCampus AgePercent students in dormsCampus rural environmentPercent Minority studentsAutonomy/FlexibilityAdministrative Categories ranging from least flexi-

ble (1) to most flexible (4).Private U. = 4

Academic Categories ranging from least flexi-ble (1) to most flexible (4).Private U. = 4

Administrator CharacteristicsAge 4 categories from survey Under 30; 30 to 44; 45–59; 60+

AD

MIN

IST

RA

TIV

ES

AT

ISF

AC

TIO

NIN

PU

BLIC

AN

DP

RIV

AT

EU

NIV

ER

SIT

IES

101

Female Dichotomous variable; from surveyHighest Degree Self identified from survey High school diploma; Bache-

lors; Masters degree; Earneddoctorate

Academic Rank Self identified from survey Do you have an academic rank?Administrative Rank Self identified from survey and cat-

egorized in five ranksAdministrative Division Self identified from survey and Academic Affairs; Business &

translated into dummy variables Finance; Human Resources, etc.Personal Financial stress 1 item from survey—5 point Likert Personal or family financial

scale indicating the extent to which problems.this contributed to stress

Personal Health stress 1 item from survey—5 point Likert Personal or family health prob-scale indicating the extent to which lems.this contributed to stress

Perceived Work ClimateRegulatory Climate 1 item from survey—5 point scale This institution experiences

more external regulation thanmost other campuses.

Controlled Work Environment 1 item from survey—5 point scale The work environment here ismore controlled than at mostuniversities.

Administrative Teamwork 2 items from survey—5 point scale .74 .67 This campus is characterized byan unusual degree of administra-tive teamwork.

Inadequate Funding 1 item from survey—5 point Likert Inadequate funding.scale indicating the extent to whichthis contributed to stress

102V

OLK

WE

INA

ND

PA

RM

LEY

TABLE 1. (Continued)

Cronbach’sConstructs and Variables from Alpha Cronbach’sHigher Ed Regulation and Job Private Alpha Example of Key itemsSatisfaction Literature Nature of the Measure Univ Public Univ in Each Scale

Inadequate Facilities 1 item from survey—5 point Likert Inadequate facilities.scale indicating the extent to whichthis contributed to stress

Pressure of Workload/Time 2 items from survey—5 point Lik- .80 .74 Time pressure; overwhelmingert scale indicating the extent to work responsibilitieswhich this contributed to stress

Interpersonal Conflict 7 items from survey—5 point Lik- .79 .76 Conflict with “difficult” individ-ert scale indicating the extent to uals; interpersonal relationshipswhich this contributed to stress with colleagues; staff turnover

Dependent Satisfaction VariablesOverall Satisfaction 1 item from survey—5 point Likert

scale indicating the overall level ofjob satisfaction

Intrinsic Satisfaction 8 items from survey—5 point Lik- .90 .89 Autonomy & independence; in-ert scale indicating the level of sat- tellectual challenge; feelings ofisfaction with this aspect of job accomplishment

Extrinsic Satisfaction 3 items from survey—5 point Lik- .79 .77 Salary & fringe benefits; oppor-ert scale indicating the level of sat- tunities for advancementisfaction with this aspect of job

Work Conditions 5 items from survey—5 point Lik- .79 .74 Hours; job security; workert scale indicating the level of sat- pressureisfaction with this aspect of job

Relationships with Others 5 items from survey—5 point Lik- .81 .80 Professional relationships withert scale indicating the level of sat- colleagues; relationships withisfaction with this aspect of job students; social status or recog-

nition

ADMINISTRATIVE SATISFACTION IN PUBLIC AND PRIVATE UNIVERSITIES 103

range from .74 to .90. These four scales plus the single “overall” item are thedependent measures in this study. While the original scales were constructedusing a public university population, the alpha reliabilities of the work climate,stress, and satisfaction scales from private university administrators are in mostcases even stronger than those in public universities. More information aboutthe psychometric characteristics of these measures can be obtained from the firstauthor.

Table 2 shows the number of respondents from public versus private universi-ties by functional area and rank (cases with missing rank or function are ex-cluded). We targeted these managers for the study not because they necessarilyrepresent all university administrators, but because they occupy particular jobtitles that are potentially influenced by the autonomy that accompanies privateuniversity control. These 120 universities are the ones that agreed to participatein the study and they represent approximately half of all the nation’s doctoral-granting universities. The totals across the tables are not completely congruentbecause not every respondent completed every item on the survey. The greatest

TABLE 2. Survey Respondents by Sector, Division, and Rank

Administrative Rank

Administrative Rank Rank Rank Rank RankSector Division 1 2 3 4 5 Total

PublicAcademic 88 90 6 4 5 193Business & Finance 65 42 40 180 29 356IR & Planning 3 2 7 56 6 74Human Resources 3 11 14 75 12 115Student Service 1 7 3 137 7 155Physical Plant 2 10 16 52 3 83Public Sub-Total 162 162 86 504 62 976

PrivateAcademic 18 19 1 38Business & Finance 18 11 9 36 3 77IR & Planning 3 1 5 3 12Human Resources 4 2 19 25Student Service 1 4 2 27 34Physical Plant 2 1 2 9 14Private Sub-Total 46 37 14 97 6 200

Total 208 199 100 601 68 1,176

Note: Rank 1 = President, Vice President, and Provost; Rank 2 = Dean and Associate Vice President;Rank 3 = Assistant Vice President; Rank 4 = Director; Rank 5 = Assistant to.

104 VOLKWEIN AND PARMLEY

representation of administrative respondents is from the “Director” rank and the“Business and Finance” functional area. Additionally, two-thirds of the respon-dents are between ages 45 and 60 and 75% are male. Most have a doctoral ormaster’s degree.

Analytic Procedures

The study uses principal components analysis to collapse the survey itemsinto a smaller number of scales that reflect the concepts in the literature and theresults of previous studies. The internal consistency of each scale is tested usingCronbach’s alpha. Once the scales and individual measures were identified, weexamined the data using descriptive statistics and compared the responses fromthose in public versus private universities. In order to assess any differences insatisfaction between public and private administrators, separate OLS stepwiseregression equations were constructed for each of the five satisfaction measuresfor public and for private universities. This enabled us to examine and comparethe influences on each satisfaction dimension identified in the research literatureand empirically derived from our survey data.

RESULTS

Do administrators in public and private universities experience different lev-els of satisfaction? Table 3 compares the public and private university responseson each of the five aspects of administrative job satisfaction. On four of the fivesatisfaction dimensions, there are no significant differences in the mean re-sponses and standard deviations of the public and private sector respondents.Both groups are most satisfied with the intrinsic rewards of their positions (3.9to 4.0 on a 5-point scale). Both groups are least satisfied with their extrinsicrewards and work conditions (3.2 to 3.5 on a 5-point scale). The largest andonly significant difference between the two groups of administrators occurs onthe extrinsic rewards dimension (salary, benefits, promotion), where the privateuniversity group is significantly more satisfied than the public. Thus, we finddifferences between the two groups that are smaller than one might expect fromreading the literature.

Moreover, when all the relevant variables are entered into an OLS Regressionmodel, these small differences in satisfaction between the public and privatesectors all prove to be insignificant, including those for extrinsic satisfaction.Table 4 shows the regression models for each of the five satisfaction measures.Controlling for all other variables, the private/public dichotomy does not exerta significant influence in any of the five models. Thus, we conclude that auton-omy in general, and type of control in particular, do not account for meaningfuldifferences in university administrative job satisfaction.

ADMINISTRATIVE SATISFACTION IN PUBLIC AND PRIVATE UNIVERSITIES 105

TABLE 3. Mean Differences in Satisfaction Between Publicand Private University Administrators

Dependent Variable(s) MeanSatisfaction Measures Public Private Difference

Overall SatisfactionMean 3.90 3.93 −.03Std Dev .80 .87N 992 202

IntrinsicMean 3.94 4.00 −.06Std Dev .72 .74N 975 198

ExtrinsicMean 3.23 3.47 −.24*Std Dev .90 .90N 983 196

Work ConditionsMean 3.37 3.35 .02Std Dev .74 .83N 972 203

Relationships with OthersMean 3.70 3.68 .02Std Dev .65 .67N 976 203

*t-test significant at the .001 Level.

We tested for possible colinearity between the private/public variable and thetwo measures of autonomy, which included private universities as the most ex-treme case of autonomy. When we excluded the two autonomy measures fromthe analysis, we observed trivial differences in the regression results.

The first model in Table 4 shows the beta weights from the OLS regressionusing overall satisfaction as the dependent measure. The model is moderatelyrobust with an R-square of .29. Neither the campus characteristics nor the ad-ministrator characteristics are significant, but four of the workplace measuresare. Controlling for all other variables, an absence of interpersonal conflict andan atmosphere of teamwork account for most of the explained variance in over-all satisfaction. These are three to five times more influential than workloadpressures and an overly controlled work environment, which exert negative in-fluences on overall satisfaction.

The second model in Table 4 reveals that only six of the 33 variables in theanalysis exert an influence on intrinsic satisfaction. Interpersonal conflict has a

TABLE 4. Regression Beta Weights for Adminstrative Satisfaction Measures

Dependent Variables

Overall Intrinsic Extrinsic Relation-Satis- Satis- Satis- Work ships with

Independent Variables faction faction faction Conditions Others

Campus CharacteristicsCampus Size −0.07**Campus Wealth .07*Faculty QualityUndergrad Quality .16*** 0.06**Has medical school/hospital .08**Has agriculture/Land Grant status 0.07**Public Flagship in the StateConstitutional recognitionCampus agePercent students in dormsCampus rural environmentPercent minority studentsAutonomy/Flexibility

AdministrativeAcademic

Private = 1, Public = 0 ns ns ns ns ns

Administrator CharacteristicsAge −.09** 0.08**Female −.05* −.09** −0.08**Highest degreeAcademic rank 0.14***Administrative Rank −.11*** −.15***Administrative Division

Academic Affairs .10**Business & FinanceIR & PlanningHuman ResourcesStudent Services

Personal Financial stressPersonal health stress

Perceived Work ClimateRegulatory climateControlled work environment −.09** −.17***Adm. Teamwork .26*** .26*** .15*** .20*** 0.20***Inadequate funding −.11*** −.07**Inadequate facilitiesPressure of workload/time −.07** −.36***Interpersonal conflict/stress −.35*** −.36*** −.19*** −.35*** −0.30***

Total R-Squared .29 .33 .19 .45 0.23

* p = <.05** p = <.01*** p = <.001Non-significant Beta weights not shown.

ADMINISTRATIVE SATISFACTION IN PUBLIC AND PRIVATE UNIVERSITIES 107

significant negative influence and teamwork has a significant positive influence.Of lesser importance is a controlled work environment and three respondentcharacteristics—age, female sex, and administrative rank. The model is moder-ately robust with an R-square of .33.

The third column of beta weights in Table 4 reflects the results for extrinsicsatisfaction. With an R-square of .19, this is the weakest of the five regressionmodels. Again, interpersonal conflict is significantly negative and teamwork issignificantly positive. As one might expect, extrinsic satisfaction is lower atuniversities with perceived inadequate funding and higher at universities withgreater campus wealth, with higher undergraduate quality, and with a medicalschool or hospital. Females are less satisfied with their extrinsic rewards. Thosewith a higher rank (lower number) and those in academic affairs are more likelyto have higher extrinsic satisfaction.

The fourth column in Table 4 is the most robust of these models (R-square =.45). Only four of the 33 variables prove significant, and they are all associatedwith the work environment. Teamwork is a positive influence on satisfactionwith work conditions, and three workplace measures have a negative influ-ence—interpersonal conflict, workload and time pressure, and inadequate fund-ing. None of the respondent or campus characteristics are influential.

In the last model, eight variables achieved significance in explaining adminis-trator satisfaction about their relationships and interactions with others. As weexpected, interpersonal conflict has a strong negative influence, and teamworkhas a strong positive influence. Campus size and being female also has a nega-tive but weaker influence on this type of satisfaction. Four other measures havea positive association with this type of satisfaction—undergraduate quality, landgrant status, age, and academic rank.

Thus, we see that all five regression models in Table 4 reveal prominent butopposite roles for teamwork (positive influence) and for interpersonal conflict(negative influence). Most of the other workplace variables effect one or twotypes of job satisfaction, but not the other types. Workplace pressure and anoverly controlled work environment do tend to lower overall satisfaction, aswell as a satisfaction subcomponent. Most campus and administrator character-istics exert influences that are generally small and focused on a particular kindof job satisfaction, but do not influence overall satisfaction. Those with higheradministrative rank are more satisfied with the intrinsic and extrinsic rewardsof their jobs. While females have the same level of overall satisfaction as males,they are significantly less satisfied with their intrinsic rewards, extrinsic re-wards, and job relationships.

While public versus private has no significant influence on satisfaction in anyof these five models, we wondered if the patterns of influence on satisfactionmight be different in the two populations. Thus, we repeated the OLS Regres-sion analysis on each separate population for each of the five satisfaction mea-

108 VOLKWEIN AND PARMLEY

sures. Are the influences on administrative job satisfaction similar or differentin public versus private universities? Table 5 shows the results of this additionalmultivariate analysis. For each of the five dependent measures the beta weightsfor the public and private regression models are displayed side-by-side.

The first two columns show the beta weights from the OLS regression modelsusing overall satisfaction as the dependent measure. In both the public and theprivate models, the largest influence on overall satisfaction is interpersonal con-flict, followed by teamwork. In other words, controlling for all other variables,an absence of interpersonal conflict and an atmosphere of teamwork account formost of the explained variance in overall satisfaction among both populations.The private university model is especially robust with an R-square of .39. In thepublic university model (R-square = .27), two other variables exert a lesser butsignificant negative influence on satisfaction—a controlled work environmentand workload/time pressure.

The second pair of columns in Table 5 presents the results of the intrinsicsatisfaction models where the public university model obtained a total R-squareof .31 and the private university model obtained an R-square of .40. In bothcolumns the largest beta weight is the negative influence of interpersonal con-flict. In the public university model, intrinsic satisfaction is associated addition-ally with teamwork (.26) controlled work environment (−.17), age (−.11), andrank (−-.09). This indicates that respondents are more satisfied if they are inwork situations without interpersonal conflict, if they are involved in teams, ifthey occupy less controlled work environments, if they are younger, and if theyhave higher administrative rank. In the private university model, intrinsic satis-faction is additionally associated with rank (−-.23) with teamwork (.21), andwith female (−.18). Thus private university respondents are more satisfied ifthey are involved in teams, are male, and have a higher administrative rank.

The third pair of columns in Table 5 presents the results of the extrinsicsatisfaction models where the public university model obtained a total R-squareof .17 and the private university model obtained an R-square of .23, results thatare less robust than the other models in the table. The private university modelis the least complex in that satisfaction with extrinsic rewards is significantlyassociated with only two variables: interpersonal conflict and rank. Among theserespondents, the absence of interpersonal conflict (−.40) is almost twice as im-portant as a high rank (−.26). In the public university model, extrinsic satisfac-tion is also associated with rank (−.17) and interpersonal conflict (−.16), butseveral other variables play a significant role. Among these are undergraduatequality (.12), having a medical complex (.10), female (−.08), and highest degree(−.08), being in academic affairs (.13), teamwork (.15), and campus funding(−.12). This indicates again that respondents are more satisfied if they are inwork situations without interpersonal conflict, if they are involved in teams, andif they have higher administrative rank.

ADMINISTRATIVE SATISFACTION IN PUBLIC AND PRIVATE UNIVERSITIES 109

The fourth pair of columns in Table 5 presents the results of the models forsatisfaction with work conditions where the public university model obtained atotal R-square of .42 and the private university model obtained an R-square of.54. These are the most robust models in our analysis. In the private universitymodel, satisfaction with working conditions is most strongly associated withinterpersonal conflict (−.44), followed by workload and time pressure (−.36)and teamwork (.19). The same variables are significant in the public universitymodel with similar beta weights, but one other variable exerts a lesser but signif-icant negative influence on satisfaction—campus funding. However, the bottomline here is that both public and private university respondents are more satisfiedif they are free of interpersonal conflict and undue workload pressure, and areinvolved in teams.

The fifth and last pair of columns in Table 5 presents the results of the satisfac-tion models for relationships with others. The public university model obtained atotal R-square of .22 and the private university model obtained an R-square of.26. The private university model is the least complex in that satisfaction withthe people one encounters on the job is significantly associated with only twovariables: interpersonal conflict (−.40) and teamwork (.22). Among these re-spondents, the absence of interpersonal conflict is almost twice as important asa working atmosphere of teamwork. In the public university model, satisfactionis again associated most strongly with interpersonal conflict (−.29) and team-work (.19), but several other variables play a significant role. Among these arecampus size (−.17), faculty quality (.10), undergraduate quality (.10), having anagricultural college (.08), being in academic affairs (.13), and perceiving a cli-mate of regulation (−.08).

In general then, the public and private models display roughly similar results,but the public models have a greater number of significant variables due to thelarger number of cases.

DISCUSSION AND CONCLUSIONS

This is the first study in higher education to compare administrative satisfac-tion among public and private university managers. We collected survey datafrom 1191 administrators ranging from presidents and vice presidents to direc-tors and assistants, and found empirical support for five aspects of job satisfac-tion. We found more similarities than differences between the two populations.Both groups on average are above the midpoint of the 5-point response scale onall five satisfaction measures, and both groups report being most satisfied withthe intrinsic rewards of their jobs and least satisfied with the extrinsic rewardsand work conditions. There are no statistically significant differences betweenpublic and private university administrators in global satisfaction, satisfactionwith intrinsic rewards, satisfaction with working conditions, and satisfaction

110V

OLK

WE

INA

ND

PA

RM

LEY

TABLE 5. Regression Beta Weights for Public and Private Administrative Satisfaction

Dependent Variables

Overall Intrinsic Extrinsic RelationshipsSatisfaction Satisfaction Satisfaction Work Conditions with Others

Independent Variables Public Private Public Private Public Private Public Private Public Private

Campus CharacteristicsCampus Size −.17**Campus WealthFaculty Quality .10*Undergrad Quality .12*** .10**Has medical/hospital .10**Agricul.Land Grant status .08*Public Flagship in the StateConstitutional recognitionCampus agePercent students in dormsCampus rural environmentPercent minority students

Administrator CharacteristicsAge −.11**Female −.18* −.08*Highest degree −.08*Academic rankAdministrative Rank −.09* −.23* −.17* −.26**Administrative Division

Academic .13** .15***

AD

MIN

IST

RA

TIV

ES

AT

ISF

AC

TIO

NIN

PU

BLIC

AN

DP

RIV

AT

EU

NIV

ER

SIT

IES

111

TABLE 5. (Continued)

Dependent Variables

Overall Intrinsic Extrinsic RelationshipsSatisfaction Satisfaction Satisfaction Work Conditions with Others

Independent Variables Public Private Public Private Public Private Public Private Public Private

Business & FinanceIR & PlanningHuman ResourcesStudent Services

Personal Financial stressPersonal Health stress

Perceived Work ClimateRegulatory climate −.08*Controlled work environ. −.09** −.17***Adm. Teamwork .25*** .26*** .26*** .21* .15*** .20*** .19* .19*** .22***Inadequate funding −.12** −.06*Inadequate facilitiesPressure of workload/time −.09** −.37*** −36***Interpersonal conflict/stress −.32*** −46** −34* −.49*** −.16*** −.40*** −.33*** −.44*** −.29*** −.40***

Total R-Squared .27 .39 .31 .450 .17 .23 .42 .54 .22 .26

* = <.05** = <.01*** = <.001Non-significant Beta weights not shown.

112 VOLKWEIN AND PARMLEY

with the people they come in contact with in their work. Only in the area ofextrinsic rewards do private university managers report significantly more satis-faction, but even these differences disappear when all relevant variables arecontrolled in an OLS Regression Model.

While public administration research suggests that there are significant differ-ences between the public and private sectors with respect to levels of satisfaction,there is little evidence in this university population apart from pay and benefitsthat confirms this hypothesis. Even the satisfaction differences for extrinsic re-wards (3.23 versus 3.47 on a 5-point scale), while statistically significant, arehardly dramatic and do not translate into differences regarding overall satisfaction.

These results may be more surprising to those outside the academy than in-side. Scholars have long observed that there are more similarities than differ-ences in organizational culture among research universities, and such organiza-tional differences between a public and a private university are certainly not asgreat as those between a manufacturing plant and a government agency. Never-theless, many higher education observers and policy makers draw distinctionsbetween public and private, often assuming that the private sector is somehow“better managed.” Certainly the national rhetoric tends to overemphasize thedifferences between public and private universities, and thus holds public uni-versities more accountable and more subject to regulation and audit.

This study may have implications for administrative productivity, turnover,mobility, and other job market characteristics. The larger organizational litera-ture suggests that job satisfaction is associated with an array of such workerbehaviors. Based on the similar responses among private and public universityadministrators in this study, we predict that other researchers will find few dif-ferences among them that are not based upon the immediate work environment.We also suspect that administrative leaders moving between public and privateuniversities may find the transition to be an easy one.

In the multivariate analysis we examined the causes of satisfaction and dissat-isfaction. Our analysis suggests that workplace relationships and an atmosphereof teamwork are almost universally important contributors to every dimension ofadministrator satisfaction. This finding supports the scholarship that emphasizesteamwork and cooperative work arrangements (Bensimon and Neumann, 1993;Frost, 1998). It is also consistent with Hagedorn’s recent research (1996) show-ing that interpersonal relationships positively influence job satisfaction and alsolessen job-related stress.

Other variables in the study are associated with specific dimensions of satis-faction. For example, holding a higher administrative rank is significantly asso-ciated with both intrinsic and extrinsic satisfaction in both populations. Thepressures of workload and time significantly reduce satisfaction with workingconditions in both populations. In public universities, campus size tends to re-

ADMINISTRATIVE SATISFACTION IN PUBLIC AND PRIVATE UNIVERSITIES 113

duce administrator’s satisfaction with the quality of their relationships with oth-ers in the workplace, and undergraduate quality tends to promote satisfaction intwo of the models.

In general, however, few of the campus characteristics seem to influenceadministrative satisfaction for either population and neither of the two autonomymeasures exerts an influence anywhere in the study. It appears to us that theconnection between institutional autonomy and the autonomy of the individualadministrator in the work setting is quite weak. This finding contradicts thebusiness and public administration literature on organizational autonomy, but itis entirely consistent with the previous university study by Volkwein, Malik,and Napierski-Prancl (1998). Doctoral-granting universities may be uniqueenough, or alike enough, so that they do not conform to the management dynam-ics observed in other types of organizations.

Moreover, the results suggest that one’s supervisor and work associates exertcrucial influences on one’s satisfaction. This has important implications for uni-versity management. Reducing interpersonal conflict and promoting teamworkshould rate high on the list of priorities for academic managers.

This study does not explore the relationship between job satisfaction andother important outcomes, such as turnover, productivity, and organizationalsuccess. However, understanding the levels and dynamics of job satisfaction inhigher education is an important first step to further exploration. Managerialsatisfaction remains an important ingredient to organizational health in the pub-lic and private sectors alike; within higher education, satisfaction is viewed asan important indicator of organizational effectiveness. The growing literature onthis topic tends to examine faculty and students more frequently than administra-tors, so there is ample room for fruitful research.

Higher education administrators and policy makers need to consider the im-plications for job satisfaction as various reforms and management techniquesfind their way into the work place. Clear and consistent evidence from thisstudy and a few others in higher education, indicate the importance of a workenvironment that is team oriented and free from interpersonal conflict. The re-ported sources of administrative stress and conflict in this study include turn-over, job security, interpersonal relations with supervisors and colleagues, andlack of personal respect. Efforts to “rightsize” and “restructure” higher educationinstitutions without regard for the effect on work units or teams have potentiallydire consequences for the levels of satisfaction, on all dimensions, experiencedby university managers.

Acknowledgments. The authors acknowledge with appreciation the earlier data collec-tion and database development provided by Shaukat Malik, Office of Institutional Re-search, SUNY–Stony Brook

114 VOLKWEIN AND PARMLEY

REFERENCES

Aldrich, H. E. (1979). Organizations and environments. Englewood Cliffs, NJ: PrenticeHall.

Austin, E. A. (1985). Factors contributing to job satisfaction of university mid-leveladministrators (ERIC Document Reproduction Service No: 259626).

Austin, A. E. and Gamson, Z. F. (1983). Academic workplace: New demands, heightenedtensions. ASHE-ERIC Higher Education Report No. 10. Washington DC: GeorgeWashington University.

Bamundo, P. J. and Kopelman, R. E. (1980). The moderating effects of occupation, ageand urbanization on the relationship between job satisfaction and life satisfaction.Journal of Vocational Behavior 17: 100–123.

Bensimon, E. M. and Neumann, A. (1993). Resdesigning collegiate leadership: Teamsand teamwork in higher education. Baltimore: Johns Hopkins University Press.

Berwick, K. R. (1992). Stress among student affairs administrators: The relationship ofpersonal characteristics and organizational variables towork related stress. Journal ofCollege Student Development 33: 11–19.

Blau, G. (1993). Further exploring the relationship between job search and voluntaryindividual turnover. Personnel Psychology 46: 313–330.

Blix, A. G. and Lee, J. W. (1991). Occupational stress among university administrators.Research in Higher Education 32(3): 289–302.

Bluendorn, A. C. (1982). A unified model of turnover from organizations. Human Rela-tions 35: 135–153.

Boone, C. W. (1987). The relationship between job characteristics, role conflict, roleambiguity, internal locus of control, and job satisfaction of college and universityadministrators. Dissertation Abstracts International 47: 2676-A.

Bruce, W. L. and Blackburn, J. W. (1992). Balancing Job Satisfaction & Performance.Connecticut: Quorum Books.

Cameron, K. S. (1978). Measuring organizational effectiveness in institutions of highereducation. Administrative Science Quarterly, 23(4): 604–629.

Carnegie Foundation for the Advancement of Teaching. (1982). The control of the campus:A report on the governance of higher education. Princeton: Princeton University Press.

Cotton, J. L. and Tuttle, J. M. (1986). Employee turnover: A meta-analysis and reviewwith implications for research. Academy of Management Review 11: 55–70.

DeSantis, V. and Durst, S. (1996). Comparing job satisfaction among public and private-sector employees. American Review of Public Administration 26(3): 327–343.

Emmert, M. and Taher, W. (1992). Public sector professionals: The effects of publicsector jobs on motivation, job satisfaction and work involvement. American Review ofPublic Administration 22(1): 37–48.

Frost, S. H. (Ed.). (1998). Using teams in higher education: Cultural foundations forproductive change. New Directions for Institutional Research, No. 100. San Francisco:Jossey-Bass.

Gardner, J. W., Atwell, R. H., and Berdahl, R. O. (Eds.). (1985). Cooperation and con-flict. Washington, DC: Association of Governing Boards.

Glick, L. N. (1992). Job satisfaction among academic administrators. Research in HigherEducation 33(5): 625–639.

Gmelch, W. H., Lovrich, N. P., and Wilke, P. K. (1984). Sources of stress in academe:A national perspective. Research in Higher Education 20: 477–490.

Graham, H. D. and Diamond, N. (1996). The rise of American research universities:Elites and challengers in the postwar era. Baltimore: Johns Hopkins University Press.

ADMINISTRATIVE SATISFACTION IN PUBLIC AND PRIVATE UNIVERSITIES 115

Hackman, J. R. and Lawler, E. E. (1971). Employee reactions to job charactersitics.Journal of Applied Psychology 55: 259–286.

Hagedorn, L. S. (1994). Retirement proximity’s role in the prediction of satisfaction inacademe. Research in Higher Education, 35(6): 711–728.

Hagedorn, L. S. (1996). Wage equity and female faculty job satisfaction: The role ofwage differentials in a job satisfaction causal model. Research in Higher Education37(5): 569–598.

Hall, R. H. (1995). Organizations: Structures, processes, and outcomes (4th edition).Englewood Cliffs, NJ: Prentice Hall.

Herzberg, F. (1966). Work and the nature of man. New York: The World PublishingCompany.

Hoppock, R. (1977). Job satisfaction. New York: Arno Press.Kalleberg, A. L. (1977). Work values and job rewards: A theory of job satisfaction.

American Sociological Review 42: 124–143.Kezar, A. (1998). Trying transformations: Implementing team-oriented forms of leader-

ship. In S. H. Frost (Ed.), Using teams in higher education: Cultural foundations forproductive change. New Directions for Institutional Research, No. 100. San Francisco:Jossey-Bass, pp. 57–72.

Lawler, E. E. (1986). High involvement management. San Francisco: Jossey-Bass Pub-lishers.

Lawrence, P. R. and Lorsch, J. W. (1967). Organization and environment. Cambridge,MA: Harvard University Press.

Lee, R. and Wilbur, R. E. (1985). Age, education, job tenure, salary, job characteristics,and job satisfaction: A multi-variate analysis. Human Relation 38(8): 781–791.

Levy, D. (1992). Private institutions of higher education. In B. Clark and G. Neave(Eds.), The Encyclopedia of Higher Education. Oxford: Pergamon Press, pp. 1183–1194.

Mumford, E. (1972). Job satisfaction: A study of computer specialists. London:Longman.

National Research Council. (1995). Research-Doctorate Programs in the United States:Continuity and Change. Washington, DC: National Academy Press.

Olsen, D. (1993). Work satisfaction and stress in the first and third year academic ap-pointment. Journal of Higher Education 64(4): 453–471.

Pascarella, E. T. and Terenzini, P. T. (1991). How colleges affect students: Findings andinsights from twenty years of research. San Francisco: Jossey-Bass.

Pfeffer, J. and Salancik, G. R. (1978). The external control of organizations: A resourcedependence perspective. New York: Harper and Row.

Porter, L. W. and Lawler, E. E. II. (1968). Managerial attitudes and performance. Home-wood, IL: Richard D. Irwin, Inc.

Rigg, M. (1992). Increased personal control equals increased individual satisfaction. In-dustrial Engineering 24(2): 12–13.

Smart, J. C. and Morstain, B. R. (1975). Assessement of job satisfaction among collegeadministrators. Research in Higher Education 3: 1–9.

Smart, J. C. (1990). A causal model of faculty turnover intentions. Research in HigherEducation 31(5): 405–424.

Smith, E., Anderson, J. L., and Lovrich, N. P. (1995). The multiple sources of workplacestress among land-grant university faculty. Research in Higher Education 36(3): 261–282.

Solomon, L. C. and Tierney, M. L. (1977). Determinants of job satisfaction among col-lege administrators. Journal of Higher Education 48(4): 412–431.

116 VOLKWEIN AND PARMLEY

Spector, P. E. (1986). Perceived cotnrol by employees: A meta-analysis of studies con-cerning autonomy and participation at work. Human Relations 11: 1005–1016.

Spector, P. E. (1997). Job satisfaction: application, assessment, cause, and conse-quences. Thousand Oaks, CA: Sage Publications.

Steel, B. S. and Warner, R. L. (1990). Job satisfaction among early labor force partici-pants: Unexpected outcomes in public and private sector comparisons. Review of Pub-lic Personnel Administration, 10: 4–22.

Sullivan, S. E. and Bhagat, R. S. (1992). Organizational stress, job satisfaction and jobperformance. Journal of Management 18(2): 353–374.

Tett, R. P. and Myer, J. P. (1993). Job satisfaction, organizational commitment, turnoverintention, and turnover: Path analysis based on meta-analytic findings. Personnel Psy-chology 46: 259–293.

Volkwein, J. F. (1986a). State financial control of public universities and its relationshipto campus administrative elaborateness and cost: Results of a national study. The Re-view of Higher Education 57(5): 510–528.

Volkwein, J. F. (1986b). Campus autonomy and its relationship to measures of universityquality. Journal of Higher Education 57(5): 510–528.

Volkwein, J. F. (1987). State regulation and campus autonomy. In J. C. Smart (Ed.),Higher Education Handbook of Theory & Research (Vol III). New York: AgathonPress, pp.120–154.

Volkwein, J. F. (1989). Changes in quality among public universities. Journal of HigherEducation 60(2): 136–151

Volkwein, J. F. and Malik, S. M. (1997). State regulation and administrative flexibilityat public universities. Research in Higher Education 38(1): 17–42.

Volkwein, J. F., Malik, S. M., and Napierski-Prancl, M. (1998). Administrative satisfac-tion and the regulatory climate at public universities. Research in Higher Education39(1): 43–63.

Vroom, V. H. (1964). Work and motivation. New York: John Wiley and Sons, Inc.