Professional Documents
Culture Documents
Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at .
http://www.jstor.org/page/info/about/policies/terms.jsp
.
JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of
content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms
of scholarship. For more information about JSTOR, please contact support@jstor.org.
Academy of Management is collaborating with JSTOR to digitize, preserve and extend access to The Academy
of Management Journal.
http://www.jstor.org
Definitionsof Commitment
TheoreticalFramework
The purpose of this research was to explain commitment as the result of
multiple forces including both psychological and structural (exchange)
determinants, thus working toward an integration of the two approaches
outlined above. A theoretical model that captures the complexity of the
organizational setting and includes individual factors has been advanced
by Kahn, Wolfe, Quinn, Snoek and Rosenthal (1964) and Katz and
Kahn (1966). They argue persuasively that the context of role-taking is
important for understanding how multiple factors influence organiza-
tional behavior (Katz and Kahn, 1966, p. 186). They also state that role
expectations are determined by the technology of the organization, its
policies, structure, and set of rewards and penalties. Within this model of
role-taking, a manager may use perceived benefits and costs in an exchange
paradigm to evaluate multiple influences that structure his decisions about
ongoing activities (March & Simon, 1958; Thibaut & Kelley, 1959; Hom-
ans, 1961;Blau, 1964, 1974; Jacobs, 1971).
Figure 1 presents the role-taking and exchange framework that will be
tested in this research. Given the complexity of managerial role contexts,
the model incorporates personal, role-related, and organizational variables.
This is the first attempt, of which the authors are aware, to look at these
three types of determinants of organizational commitment simultaneously.
This model also provides for the integration of the two approaches discussed
FIGURE 1
Role TakingModelof CommitmentProcess
Dependent Variables
There are several reasons for using both organizational and federal
service commitment to explore managerial commitment in this study. First,
research has suggested that occupational and organizational commitment
or identifications are compatible (Ritzer & Trice, 1969; Rotondi, 1975)
and/or complementary (Ritzer & Trice, 1969; Hrebiniak & Alutto, 1972).
Second, federal service commitment shares some characteristics of occupa-
tional commitment: transferability across organizations, socially mandated
set of expectations (Hughes, 1958), and ".. . relatively continuous patterns
of activities" (Form, 1968). Third, as Ritzer and Trice (1969) suggest,
it is possible that members may be tied to organizations through their
commitment to an occupation, since the organization provides the op-
portunity to pursue the occupation. In this sample of federal managers, who
largely occupy positions that do not have an occupational identity outside
of government service, commitment may well be generated initially to a
position or role, which then generates commitment to the employing
organization because the manager cannot occupy that position or role in
other employment sectors. Finally, the federal service is the entity which
IndependentVariables
The personal or individual variables to be examined in this study are age,
educational level, sex, job involvement, and attitude toward change. Several
studies of commitment have used the age, education, and sex variables
(Ritzer & Trice, 1969; Alutto, Hrebiniak, & Alonso, 1973), but disparate
findings have emerged. Some ambiguity also exists with regard to job
involvement, used as a correlate of a behavioral measure of commitment
(Weiner & Gechman, 1975) or as a component (Buchanan, 1974). No
published research on commitment has investigated the influence of attitude
toward change (Hage & Dewar, 1973) as a predictor of commitment.
The general arguments for using such variables to predict commitment
can be derived from both role and exchange theory. For instance, Becker's
side-bet theory suggests that advancing age, being female, certain role
characteristics, and longer tenure increase an individual's investment in the
federal service or employing organization and therefore the costs associated
with leaving.
Job involvement, as operationalized and used in this research, concerns
an individual's ego involvement with the job, i.e.,,the degree to which his
self-esteem is affected by his work performance (Lodahl &Kejner, 1965). It
follows that individuals very involved in their job will also have substantial
side-bet investments in federal service and employing organizations, which
provide the opportunity for them to act out "internalized values about the
goodness of work" (Lodahl & Kejner, 1965, p. 24).
It is also logical to assume that an individual who is favorably disposed
toward change would incur less personal costs in leaving an organization as
a response to increasing pressures. If role conflict or overload increased, the
change-orientedmanager'sexchange equation would permit easier exit from
the situation. Positive attitudes toward change should thus be negatively
related to commitment.
Recent research indicates that role-related factors are important influ-
ences on commitment (Alutto & Hrebiniak, 1975; Steers, 1977), involve-
ment (Patchen, 1970), organizational identification (Lee, 1971), or-
ganizational socialization (Buchanan, 1974), and organizational attraction
(Dubin et al., 1975). The role-related variables to be used in this study
are: work overload, managerial level, organizational tenure, positional
tenure, task characteristics, and perceptions concerning the importance
for promotion of performance, seniority, and technical skills. Role variables
include more dynamic aspects of the job situation that may make staying
with a given organization or the federal service more or less attractive at a
given point in time: Work overload would be perceived as a cost and
negatively affect commitment. Too much time in any one position may be
perceived as career stagnation and have an adverse effect on commitment,
TABLE 1
Basic Statistics for Independent and Dependent Variables
Hypothesized
Relationship
with
Variable Mean sd Cases Commitmenta
Personal Attributes
1. Sex of supervisor .162 .369 620 +
2. Age of supervisor 3.069 1.019 623 +
3. Change attitude 4.354 .852 625
4. Job involvement 2.545 .399 627
5. Education in years 15.314 2.369 611
Role-Related
6. Manageriallevel 4.740 1.017 632 +
7. Work overload 2.140 .790 616
8. Years in organization 10.504 8.636 614 +
9. Years on position 4.573 4.672 612
10. Skill level of subordinates 3.609 1.326 614 +
11. Performance in promotion .008 .768 612 +
12. Technical skill in promotion .000 .999 613 +
13. Seniority in promotion .000 1.000 612
OrganizationalFactors
14. Organizationalsize 338.082 298.565 634 +
15. Union presence .821 .383 632
16. Percent of supervisionb 14.470 5.429 634 +
17. Centralizationof authority 1.189 4.440 634
Commitment
18. Organizational commitment 3.618 .937 608
19. Federal Service commitment 4.143 .922 610
a A + indicates a positive and a - indicates a negative relationship with both forms
of managerialcommitment.
b A measure of average span of control.
from decision making autonomy, create costs for the manager, drain re-
sources, and decrease commitment.
Independent and dependent variables, basic descriptive statistics, and the
direction of hypothesized relationships are presented in Table 1.
Statisticsand Measurement
The principal method of data analysis used in this study is multiple
regression because of the multivariate relationships specified in the re-
search model. The primary strategies used to construct scales from multiple
item measures were factor analysis and standardization of measures [see
Stevens (1976) for further details].
To measure commitment, managers were asked if they would leave the
installation or the federal service given no, slight, or large increases in pay,
freedom, status, responsibility, and opportunity to get ahead. They re-
sponded by checking one of three categories: would definitely leave, un-
decided, or would definitely not leave. The object in constructing a scale
from such items was that respondents who would stay with certain induce-
ments were more committed than those who would leave with the same or
less inducements.
Because of controversy in the literature, two scales were used: The
Ritzer and Trice (1969) (R-T) scale composed of the five inducement
items (replicating their original scaling procedures) and the "slight change"
only items, favored by Hrebiniak and Alutto (1972).2
The operationalizations of most independent variables were based on the
responses of supervisors. Variables 3, 4, 7, 11, 12, and 13 (Table 1) were
multiple-item scales based on factor analytic results. Centralization was
based on a decision matrix approach to measuring influence in decision
making (Beyer & Lodahl, 1976). Further details on the operationalization
of these variables are available from the authors.
RESULTS
The intercorrelation matrix of all of the variables to be used in subse-
quent analyses is given in Table 2. In these results, seven of the 17 hypoth-
2 The
operationalizations,scale construction, and tests of reliability and validity for the
independent and dependent variables are discussed in detail in Stevens (1976). In addition,
questions regarding the appropriatenessof measures used by Alutto et al. (1973) to assess
the R-T scaling procedures are discussed in Beyer, Stevens, and Trice (1977).
TABLE
Correlation a Matrix for
1 2 3 4 5 6 7
1. Sex
2. Age -.061
3. Change .098 .027
4. Job involvement -.047 .177 .130
5. Education -.206 -.179 .004 .117
6. Managerlevel -.148 -.007 -.094 .108 .255
7. Overload -.008 -.093 .113 -.012 .118 -.082
8. Years/organization .040 .442 .008 .044 -.198 -.112 -.022
9. Years/position -.083 .377 -.015 .018 -.103 -.058 -.032
10. Skill level -.164 -.078 -.049 .020 .454 .184 .033
11. Promotion/performance .020 -.014 .043 .054 -.090 .049 -.155
12. Promotion/technical -.025 .113 -.009 .043 -.109 -.060 -.055
13. Promotion/seniority .020 .160 .048 .122 -.145 -.109 .042
14. Organization size .072 .025 .037 .004 -.035 -.286 -.015
15. Unions .051 .044 .061 .031 -.053 -.124 .062
16. Percent of supervision --.117 .059 -.030 .026 .074 .018 .001
17. Centralization -.025 -.011 .036 .026 .044 .184 .019
18. Organizationcommitment .006 .117 -.074 .159 .045 .061 -.163
19. Federal commitment .062 .107 -.081 .042 -.204 -.037 -.250
a With 600 df, when r = .08, p = .05; r = .11, p = .01; r = .14, p = .001.
2
Commitment Study Variables
8 9 10 11 12 13 14 15 16 17 18
.446
-.104 -.083
-.045 -.039 -.039
.073 .042 -.022 .483
.055 .030 -.162 .052 .233
.204 -.019 -.022 .045 -.064 -.108
.114 .072 -.108 .124 .060 .031 .233
-.106 -.003 .158 -.098 .074 .071 -.395 -.262
.018 .020 -.021 .008 .031 .116 -.160 -.058 .116
.184 -.009 .104 .090 .087 -.014 .029 .032 -.032 -.023
.129 -.037 -.096 .172 .134 .012 .044 .078 -.065 -.014 .404
Influenceson Commitment
In Table 3a, five role and personal variables are significant and another
personal variable is a marginally significant predictor of organizational
commitment, as measured by the R-T scale. Years in the organization
emerge as the best positive predictor and work overload as the largest
negative predictor, giving support to the exchange or side-bet approach to
commitment. Job involvement is also a strong positive predictor, underlining
the importance of psychological predispositions toward work. The findings
here generally support the hypotheses and usefulness of the composite
research model. In addition, the positive influence of skill level of sub-
ordinates suggests that job content can also be an important influence on
organizational commitment.
The finding that positional tenure is negatively related to organizational
TABLE 3
Regression Results for Alternative Measures of Managerial Commitment
on Personal, Role, and Organizational Variables
A. OrganizationalCommitment(n = 611)
StandardizedCoefficients
CommitmentMeasures Evaluated
Var.
Influence Cat. R-T Scale Slight
Work overload R -.15**** -.13***
Years in the organization R .24********
Job involvement P .15** 3***
Skill level of subordinates R .11* .14**
Positive change attitude P -.08* .01
Years in current position R -.12*** -.12***
Manager'sage P .04 .09*
R2 .13 .13
B. Federal Service Commitment(n = 610)
StandardizedCoefficients
CommitmentMeasures Evaluated
Var.
Influence Cat. R-T Scale Slight
Work overload R -.21**** .20****
Years in the organization R .13*** .06
Positive change attitude P -.08 -.03
Years in current position R -14*** -.15**
Level of education P -.14* -.10*
Importanceof performance
for promotion R .10** .07
R2 .15 .10
*p < .10
** p < .05
*** p < .01
**** p < .001
DISCUSSION
The results of both regression analyses indicate that factors with benefit
and cost implications are the most powerful influences on managerial com-
mitment. While some personal variables, such as job involvement and
change attitude, were important, other personal characteristics were not.
Unlike a previous study (Hrebiniak & Alutto, 1972), sex did not emerge as
a significantpredictor. The relationship may have been due to concentration
of one sex in the sample. In the earlier study, the sample consisted primarily
of female teachers and nurses; in this sample, only 16 percent of the federal
managers were females. Another possible interpretation is that sex differ-
ences may be less salient for commitment for the managerial as opposed to
the nonmanagerial employees, implying that female managers assess cost-
benefit exchanges similar to males. Samples with more equal concentrations
of males and females may be needed to adequately test this issue.
The major discrepancies between the hypotheses and findings were for
the organizational variables in the model. Organizational size, centraliza-
tion, and percent of supervision, were not found to be important influences
on managerial commitment. Though no direct relationshipswere found, it is
reasonable to suppose that the influence of organizational factors may be
moderated by role-related variables such as work overload.
Another consistent finding was that a positive attitude toward change was
a negative influence on both organizational and federal service commitment.
This finding may indicate that the value placed upon change as a strategy for
adaptation to societal or organizational expectations may be an important
variable in predicting job changes. It is likely that change values affect per-
sonal cost-benefit ratios by decreasing the psychological costs of leaving a
particular role, job, organization, et cetera and increasing the benefits of
leaving. It is also conceivable that the salience of certain perceived benefits
or costs varies with the manager's predisposition toward change. For
example, managers who tend to favor change or innovation may not be
committed to occupations or organizations where inflexibility or overt
formalization prevails. Other managers may self-select themselves for more
formalized organizations and away from roles calling for innovation.
Although traditionally considered a predictor of commitment, age was
not significantly related to either type of commitment in multivariate anal-
yses and only weakly (r = .12 and .11) in zero-order correlations. Other
variables may moderate the effects of age, especially job involvement (r -
.18 with age, p < .01) and educational level (r = -.18 with age, p < .01).
Work overload is the most significantnegative influence on organizational
and federal service commitment. Federal managers who perceive them-
selves as having too little authority to fulfill their responsibilities, bothered
by work overload, and having to finish the work of others have low
commitment. This finding is actually ratherencouraging, because it involves
a variable that is easier to change then some others. Management can
conceivably alter structural or other factors causing overload before its
SystemConsiderations
In addition to the variables examined here, other factors may influence
commitment. The level of the multiple correlations obtained suggests that
some important variables may have been omitted in this research: in-
terpersonal influences emanating from peers or organizational socialization
processes may influence commitment to the organization or federal service
through commitment to the immediate work group or organizational unit.
Current conditions in the job market and economy can be expected to
affect costs of leaving or staying.
Also, commitment may be composed of multiple elements, some of
which may be causally or temporally antecedent to others (Stevens, 1976).
Other results suggest that a theoretical framework for ordering the pre-
cedence of various forms of commitment may be a valuable approach to
building a comprehensive theory of commitment.3 Such a theory would
ideally specify the relationships between elements or forms of commitment
and their determinants. A review of the relevant literature exposed both
overlap and ambiguity with regard to competing concepts of commitment.
Terms such as professional commitment, occupational commitment, or-
ganizational loyalty, organizational attraction, organizational identification,
organizational involvement, role commitment, job involvement, or job com-
mitment have been used interchangeably or with no clear differentiation
with regard to related constructs. These concepts could theoretically sub-
sume or complement the elements of commitment studied here or the
REFERENCES
1. Alutto, J. A., and L. G. Hrebiniak. "Researchon Commitment to Employing Organiza-
tion: Preliminary Findings on a Study of Managers Graduating from Engineering and
MBA Programs" (Paper presented at the Thirty-fifth Annual Meeting, Academy of
Management,New Orleans, 1975).
2. Alutto, J. A., L. G. Hrebiniak, and R. C. Alonso. "On Operationalizingthe Concept of
Commitment,"Social Forces, Vol. 51 (1973), 448-454.
3. Baldridge, J. V., and R. A. Burnham. "OrganizationalInnovation: Individual, Organi-
zational and Environmental Impacts," Administrative Science Quarterly, Vol. 20 (1975),
165-176.
4. Becker, H. S. "Notes on the Concept of Commitment,"American Journal of Sociology,
Vol. 66 (1960), 32-40.
5. Becker, H. S., and J. Carper. "The Elements of Identification with an Occupation,"
American Sociological Review, Vol. 21 (1956), 341-347.
6. Beyer, J. M., and T. M. Lodahl. "A Comparative Study of Patterns of Influence in
United States and English Universities," Administrative Science Quarterly, Vol. 21
(1976), 104-129.
7. Beyer, J. M., J. M. Stevens, and H. M. Trice. "On the AppropriateApplication of Tests
of Reliability: a Research Note on MeasuringCommitment" (Mimeo paper, School of
Management,State University of New York at Buffalo, 1977).
8. Blau, P. M. Exchange and Power in Social Life (New York: Wiley, 1964).
9. Blau, P. M. On the Nature of Organizations (New York: Wiley, 1974).
10. Buchanan, B., II. "BuildingOrganizationalCommitment:The Socialization of Managers
in Work Organizations," Administrative Science Quarterly, Vol. 19 (1974), 533-546.
11. Dubin, R., J. E. Champoux, and L. W. Porter. "CentralLife Interests and Organizational
Commitment of Blue-Collar and Clerical Workers," Administrative Science Quarterly,
Vol. 20 (1975), 411-421.
12. Form, W. H. "Occupationsand Careers," in D. L. Sills (Ed.), International Encyclo-
pedia of the Social Sciences, Vol. 11 (New York: MacMillan and the Free Press, 1968),
pp. 245-254.
13. Hage, J., and R. Dewar. "Elite Values Versus OrganizationalStructurein Predicting In-
novation," Administrative Science Quarterly, Vol. 18 (1973), 279-290.
14. Homans, G. C. Social Behavior: Its Elementary Forms (New York: Harcourt Brace
World, 1961).
15. Hrebiniak, L. G., and J. A. Alutto. "Personal and Role-Related Factors in the Devel-
opment of Organizational Commitment," Administrative Science Quarterly, Vol. 18
(1972), 555-573.
16. Hughes, E. Men and Their Work(Glencoe, Ill.: Free Press, 1958).
17. Ivancevich, J. M., and J. H. Donnelly, Jr. "Relation of OrganizationalStructureto Job
Satisfaction, Anxiety-Stress, and Performance," Administrative Science Quarterly,
Vol. 20 (1975), 272-280.
18. Jacobs, T. 0. Leadership and Exchange in Formal Organizations (Alexandria, Va.:
Human Resources Research Organization (1971).
19. Kahn, R. L., D. M. Wolfe, R. P. Quinn, J. D. Snoek, and R. A. Rosenthal. Organiza-
tional Stress: Studies in Role Conflict and Ambiguity (New York: Wiley, 1964).
20. Kanter, R. M. "Commitmentand Social Organization:A Study of Commitment Mech-
anisms in Utopian Communities," American Sociological Review, Vol. 33 (1968),
499-517.
21. Katz, D., and R. L. Kahn. The Social Psychology of Organizations (New York: Wiley,
1966).
22. Kochan, T. A. "Determinantsof the Power of Boundary Units in an Interorganizational
Bargaining Relation," Administrative Science Quarterly, Vol. 20 (1975), 434-452.