Meyer 1987
Meyer 1987
ABSTRACT
The purpose of this longitudinal study was to examine the development and consequences of organiza-
tional commitment early in employees' careers. University graduates who had recently accepted full-
time permanent employment were administered the Organizational Commitment Questionnaire (OCQ)
and self-report measures of work experiences, work behaviour, and turnover intention after one, five,
and nine months on the job. Comparison of means over time revealed a significant decline in
commitment and the favourableness of work experiences during the first nine months of employment.
The results of cross-lagged regression analyses provided evidence consistent with the hypothesis that
early work experiences influence commitment to the organization, but not with the hypothesis that
commitment influences work behaviour and turnover intention.
Over the past decade there has been a steadily increasing number of studies
conducted to identify the "antecedents" and "consequences" of organizational
commitment. Most of this research, however, has been cross-sectional and, despite
repeated calls for more longitudinal research (e.g., Buchanan, 1974; Mowday,
Porter, & Steers, 1982; Porter, Steers, Mowday, & Boulian, 1974), only a few such
studies have been conducted to date (e.g., Bateman & Strasser, 1984; Mowday &
McDade, 1979,1980; Porter, Crampon, & Smith, 1976; Porter et al., 1974). If the
development of commitment is a gradual process, as Mowday et al. (1982)
suggest, more longitudinal research will be necessary to determine how this
process unfolds over time.
The purpose of the present study was to examine the development and con-
sequences of commitment during thefirstyear of employment. This early period of
employment has been identified as particularly important for the development of
commitment (Buchanan, 1974; Mowday & McDade, 1980; Porter et al., 1974).
Indeed, there is probably no other time in an individual's career when he/she is
more sensitive to organizational influence. As indicated by the high rate of turnover
during this period (Wanous, 1980), adjustment can be very difficult for many
employees. For those who have positive experiences, however, it can set the
This research was supported by a grant from the Social Sciences and Humanities Research Council of
Canada (#410-82-0439). The authors would like to thank Dr. R.C. Gardner for his helpful comments
and suggestions and Ruth Siberry and the staff at the UWO Placement Office for their assistance in the
recruitment of research participants.
Requests for reprints should be sent to John P. Meyer, Department of Psychology, The University of
Western Ontario, London, Ontario N6A 5C2.
framework for future success (Bray, Campbell, & Grant, 1974; Hall, 1976). Thus,
this is a time when an organization can have a significant impact on the commit-
ment of its employees and when proper management can have important short- and
long-term implications for both the individual and the organization.
Organizational commitment research has been complicated by the fact that
investigators have worked under differing assumptions concerning the nature of
commitment. The two approaches that have had the greatest impact on theory and
research to date are the behavioural (or social psychological) and attitudinal (or
organizational behaviour) approaches (cf. Mowday et al., 1982; Salancik, 1977;
Staw, 1977). In the behavioural approach, commitment is viewed as a tendency to
continue a particular line of activity (e.g., Becker, 1960; Kiesler, 1971; Salancik,
1977; Staw, 1977). Salancik, for example, views commitment as "a state of being
in which an individual becomes bound by his actions and through these actions to
beliefs that sustain the activities and his own involvement" (1977, p. 62). Research
in this tradition has been focused on identifying conditions (e.g., volition) sur-
rounding the initiation of a line of action that lead to its continuation, and on the
effects of such action on the development of work attitudes (e.g., O'Reilly &
Caldwell, 1981).
Advocates of the attitudinal approach view commitment as an attitude of
attachment to the organization. Perhaps the most popular definition of attitudinal
commitment is provided by Porter and his associates (Mowday, Steers, & Porter,
1979; Porter et al., 1974; Porter et al., 1976) who suggest that commitment is
comprised of: "(1) a strong belief in and acceptance of the organization's goals and
values, (2) a willingness to exert considerable effort on behalf of the organization,
and (3) a strong desire to maintain membership in the organization" (Mowday
et al., 1979, p. 226). This definition of attitudinal commitment served as the basis
for the development of a widely used research instrument, the Organizational
Commitment Questionnaire (OCQ). Researchers taking the attitudinal approach to
commitment have focused primarily on the identification of factors contributing to
the development of this attachment and its effects on organizational behaviour. It is
in this tradition that the present research was conducted.
Of the many factors that have been linked with the development of attitudinal
commitment, experiences with the job and the organization (herein referred to as
work experiences) have been identified as particularly important (Mowday et al.,
1982; Steers, 1977). A number of such work experiences have been found to be
related to organizational commitment in previous research. These include con-
firmation of pre-entry expectations (Arnold & Feldman, 1982; Buchanan, 1974;
Steers, 1977), job satisfaction (Arnold & Feldman, 1982; O'Reilly & Caldwell,
1981; Stone & Porter, 1975), job scope and challenge (Buchanan, 1974; Marsh &
Mannari, 1977; Steers, 1977), supervisor ability/behaviour (Michaels & Spector,
1982; Morris & Sherman, 1981), participation in decision-making (Rhodes &
Steers, 1981), role clarity and freedom from conflict (Jamal, 1984; Morris & Koch,
LONGITUDINAL ANALYSIS OF COMMITMENT 201
1979; Welsh & LaVan, 1981), and organizational dependability or concern for
workers (Buchanan, 1974; Marsh & Mannari, 1977; Steers, 1977). Although these
findings are consistent with the view that work experiences influence commitment,
the research has been largely cross-sectional and has not addressed the issue of
causal priority. Alternate explanations for the relationships, including "reverse
causality" and the influence of third variables and/or response sets, are equally
plausible.
In one of the few longitudinal studies conducted to date, Bateman and Strasser
(1984) obtained measures of organizational commitment and work experiences
(along with other variables) on two separate occasions five months apart from a
sample of nursing employees. Although they found significant correlations
between commitment and work experiences within and across occasion of mea-
surement, when they employed cross-lagged regression analyses to examine the
influence of experiences on commitment (and vice versa), however, only one of the
experience variables, leader punishment behaviour, was found to influence
employees' commitment five months later. Moreover, one of the work experience
variables, job satisfaction, was found to be influenced by commitment rather than
the reverse. Although neither the cross-lagged correlations nor cross-lagged
regression parameters can be taken as evidence for causal effects, Bateman and
Strasser note several advantages (e.g., partial control for the influence of method
variance) of the regression analyses that lead to increased confidence in causal
inference. The findings of this study, therefore, provide little support for the
hypothesis that positive work experiences are instrumental in the development of
commitment to the organization.
Although many explanations might be offered to account for this lack of support,
the nature of the sample may have been a contributing factor. Participants in
Bateman and Strasser's study had an average of more than ten years of experience
at the job and almost four years with their current employer. It is likely that their
commitment to the organization was already well established and, therefore,
relatively insensitive to short-term fluctuations in work experiences (cf. Porter
et al., 1974). If this is the case, a more appropriate test of the hypothesis would be
to examine the relationship between work experiences and commitment among
new employees whose commitment is in the process of development. Mowday and
McDade (1980) measured employees' commitment and work experiences after
one day, one month, and three months on the job and found significant correlations
both within and across time (as did Bateman & Strasser). No attempt was made,
however, to address the issue of causal priority. There is a need, therefore, for more
longitudinal research to examine the influence of work experiences on commit-
ment during this early period of employment.
Research concerning the consequences of commitment has focused primarily on
turnover, with most studies reporting the predicted inverse relationship (e.g.,
Angle & Perry, 1981; Arnold & Feldman, 1982; Clegg, 1983; Ferris & Aranya,
202 MEYER & ALLEN
1983; Horn, Katerberg, & Hulin, 1979; O'Reilly & Caldwell, 1981; Porter, Cram-
pon, & Smith, 1976; Porter, Steers, Mowday, & Boulian, 1974; Steers, 1977). In
only two of these studies was the relationship between commitment and turnover
examined longitudinally with repeated measurement over time. Porter et al. (1974)
found that eventual leavers and stayers differed in their commitment to the
organization very early in their careers and that this difference increased over time.
Porter et al. (1976) found that commitment assessed on the first day on the job
predicted turnover occurring as much as a year later, and that commitment
underwent a substantial decline in the period immediately preceding the decision
to leave. Research examining the process by which commitment leads to turnover
has shown that the relationship between commitment and turnover is mediated by
turnover intention (Michaels & Spector, 1982; Mowday, Koberg, & McArthur,
1984; Stumpf & Hartman, 1984).
Less attention has been given to the link between commitment and other forms
of organizational behaviour, and the findings have been somewhat less consistent.
Commitment was found to be positively related to attendance behaviour in one
study (Steers, 1977) but not in others (Angle & Perry, 1981; Jamal, 1984), and
positively related to some but not all indices of individual (Lee, 1971; Steers, 1977;
Wiener & Vardi, 1980) and group (Mowday, Porter, & Dubin, 1974; Angle &
Perry, 1981) performance. The lack of attention given to the effects of commitment
on work behaviour is somewhat surprising. If it is to be argued that the commit-
ment of employees is beneficial to an organization, it must be shown not only that it
leads to decreased turnover, but also that those who remain are willing to work
toward organizational goals.
In sum, there is a need for more longitudinal research to examine the develop-
ment and consequences of organizational commitment. It is particularly important
to examine the development process early in employees' careers when attitudes are
probably most malleable (Porter et al., 1974) and declining morale and excessive
turnover have been found to be particularly problematic (Hall, 1976; Wanous,
1980). In this study, repeated measures of commitment, work experiences, work
behaviours, and turnover intentions were obtained from a sample of newly hired
university graduates. In light of existing theory and research, it was predicted that
(1) commitment to the organization would decline during the first year of employ-
ment, (2) the decline in commitment would be accompanied by a decline in the
perceived favourableness of work experiences, (3) early work experiences would
influence subsequent commitment to the organization, and (4) commitment would
influence future work behaviour and turnover intention.
METHOD
recruitment program at the University of Western Ontario were contacted in April 1983 by the university
placement office to check on the progress of their job search efforts. Individuals who had accepted
permanent full-time jobs were asked if they would be willing to participate in a longitudinal study of
work attitudes and behaviour.
Fifty-four students (33 males, 21 females) volunteered to participate. The majority of the participants
(n = 49) were between the ages of 20 and 25, all had undergraduate degrees, and eight had post-
graduate degrees. Only eight had worked previously for their employer and, of these, only one had
worked full-time. The majority (n = 41) were starting employment with large organizations (i.e., more
than 500 employees) involved in a variety of activities, the most common being insurance, accounting,
and computer sales and development.
Participants received four questionnaires, one prior to entry, and the others one, five, and nine
months after they started their jobs. Each questionnaire was mailed to the participant along with a cover
letter and stamped envelope addressed to the research office. Except for personal information, which
was obtained in the first questionnaire, the data reported here are based on responses to the question-
naires completed one, five, and nine months after starting employment. Questionnaires with usable
data were returned by 53, 47, and 37 participants, respectively.
Measures
Organizational commitment. Commitment to the organization was measured one, five, and nine
months after entry using the 15-item Organizational Commitment Questionnaire (OCQ) developed by
Porter and his associates (Mowday et al., 1979; Porter et al., 1974; Porter et al., 1976). The OCQ uses a
7-point Likert-type response format (strongly disagree to strongly agree). It has been used widely in
research and has been shown to have acceptable psychometric properties (Mowday et al., 1979). Scale
scores were computed by averaging responses across scales. The average alpha coefficient (across
administrations) for the OCQ in the present sample was .90.
Work experiences. Work experience measures were also obtained in the one-, five-, and nine-month
questionnaires. Job satisfaction was measured using the 18-item short form of the Quinn and Staines
(1979) scale plus the item "The pay is good" from the 33-item form of the scale. Responses were made
on a 7-point disagree-agree scale like that used for the OCQ. Overall satisfaction scores were obtained
by averaging across the 19 items. The average alpha coefficient (across administrations) for the total
scale in the present sample was .86.
Measures of the remaining variables consisted of 2- to 6-item scales designed to assess the extent to
which the job met the employee's expectations (confirmed expectation), the job was challenging and
exciting (job challenge), the organization expected strong loyalty from its employees (organizational
commitment norm), the organization could be trusted to behave as it said it would (organizational
dependability), the employee was involved in decision-making (participation), co-workers were close-
knit (peer cohesion), the employee believed he/she was treated equitably (perceived equity), the
employee was made to feel like an important member of the organization (personal importance), the
employee's role was made clear (role clarity), and the employee was given the opportunity for self-
expression (self-expression). The coefficient alpha for these scales ranged from .57 to .88 with a median
of .82.
Work behaviour and turnover intention. A number of self-report measures of work behaviour were
obtained in the questionnaires administered five and nine months after entry. Specifically, participants
indicated the frequency of absences per month (absences) and the number of times per month they
arrived at least ten minutes late for work (tardiness). They also rated their performance relative to others
with a similar amount of experience (performance), the frequency with which they performed tasks they
judged to be "above and beyond the call of duty" (extrarole activity), and how motivated they were to
perform their jobs well (motivation). Finally, they indicated the likelihood that they would actively
search for work in a different organization within the next year (intent to search), and the likelihood that
they would leave the organization within the next year (intent to leave). With the exception of absence
and tardiness, all responses were made on 7-point Likert-type scales.
204 MEYER & ALLEN
RESULTS
Data analyses
Analyses reported here were performed only on the data obtained from participants
(n = 37) who completed all four questionnaires. Although comparison of the
continuing and discontinuing participants revealed no major differences, these
comparisons are difficult to interpret because those who left the study did so for a
variety of reasons. Differences would have been expected in some cases (e.g., for
comparisons involving voluntary leavers) but not in others (e.g., for comparisons
involving those who lost interest or relocated). Consequently, the failure to find
differences may simply reflect the effects of averaging. None of the subgroups of
discontinuing participants was of sufficient size to justify more fine-grained
comparisons.
The first step in the data analysis was to examine the stability and change in
commitment, work experiences, work behaviour, and turnover intention during the
first year Relative stability was determined by correlating scores on the same
measures across occasions of measurement; change was determined by comparing
means across the three time periods.
The second step was to examine the relationships between commitment and
work experiences within and across time. The static relationships were determined
by computing correlations between the variables obtained concurrently. The
relationship over time was examined for each work experience variable indepen-
dently by means of cross-lagged correlations and cross-lagged regression anal-
yses. Although measures were obtained on three separate occasions, for present
purposes the data were treated as two separate two-wave, two-variable longitudinal
panels, with the one- and five-month, and five- and nine-month data constituting
the two panels. The cross-lagged regression analysis was similar to that employed
by Bateman and Strasser (1984) and involved the use of structural regression
models in which the influence of one variable, x, on another, y, is represented by
the regression parameter of the path from x at an earlier time (t) to y at a later time
(t + 1) after the influence of the prior measure of y is removed. Similarly, the
influence of prior y on subsequent x is represented by the regression parameter for
the path from y at an earlier time to x at a later time after the influence of prior x is
removed. The specific regression equations in raw score form are as follow:
where the parameters, p, and -y, represent the time-lagged influence of a variable
on itself, and (32 a n d I2 represent the time-lagged influence of one variable on the
other (see Rogosa, 1980). Under the usual assumptions of linear regression
analysis, the influence of one variable on the other is indicated by non-zero values
LONGITUDINAL ANALYSIS OF COMMITMENT 205
for the relevant cross-lagged regression parameter. Note that this analysis permits
the detection of influence of one or both variables on the other over time. Thus, if
commitment were the y variable and job satisfaction the x variable, a significant
value for f$2 would indicate that satisfaction at t influences commitment at t + 1 ; a
significant value for 7 2 would indicate that commitment at t influences job
satisfaction at t + 1. Where both of the cross-lagged regression parameters are
significant, reciprocal influence is indicated (i.e., x influences y and y influences x
over the same interval of time).
It should be noted that the term "influence" is used above, and in subsequent
discussion, for purposes of simplicity. These analyses are "exploratory" in nature
(see James, Mulaik, & Brett, 1982) and cannot be taken as proof of causality.
Nonetheless, as Bateman and Strasser (1984) suggest, they do provide greater
confidence in causal inference than do correlation coefficients obtained within or
across occasions of measurement.
In the final step in the analyses, the same procedures described in Step 2 were
performed to examine the relationships between commitment and the work
behaviour and turnover intention variables measured after five and nine months on
the job.
Development of commitment
The correlations between commitment and work experiences measured con-
currently are presented in Table 3 and those between measures obtained on
different occasions are reported in Table 4. These correlations provide fairly strong
evidence for a relationship between work experiences and commitment both within
and across time. The strongest and most consistent relationships were obtained for
206 MEYER & ALLEN
TABLE 1
Stability coefficients for commitment antecedent, and consequence variables
Variable '•is
TABLE 2
Comparison of means over time
Means ANOVA
Variable 1 month 5 months 9 months F df P
TABLE 3
Concurrent correlations between commitment and antecedent variables
Concurrent correlations
Variable 1 month 5 months 9 months
TABLE 4
Cross-lagged correlations for commitment and antecedent variables
Cross-lagged correlations
Variable f
c 5 a. V:, r
c9a5 V5
Confirmed expectation 58*** 59*** 68*** 32*
Job challenge 59*** 60*** 38* 54***
Job satisfaction 65*** 65*** 67*** 38*
Org. commitment norm 36* 39** 23 20
Org. dependability 61*** 55*** 41** 47**
Participation 28 17 52*** 21
Peer cohesion 08 12 02 20
Perceived equity 58*** 57*** 43** 29
Personal importance 56*** 57*** 55*** 40*
Role clarity 24 29 43** 08
Self-expression 50*** 08 24 15
TABLE 5
Cross-lagged regression parameters for organizational commitment and antecedent variables
TABLE 6
Concurrent correlations, cross-lagged correlations, and cross-lagged regression parameters for
commitment and consequence variables
Standardized
Concurrent Cross-lagged regression
correlations correlations parameters
r r r Pc9b5
Variable c5b5 C 9 b9 c9b5
DISCUSSION
The fact that commitment scores were relatively stable from the first to the ninth
month of employment is consistent with the suggestion of Mowday et al. (1982)
that individuals entering an organization differ in the propensity to become
committed. This propensity may reflect personal characteristics of the employee
and/or differences in the conditions surrounding their decision to work for an
organization. Although the data reported here do not address these possibilities,
there is evidence from prior research that personal characteristics such as need for
achievement (Mowday & McDade, 1979; Steers, 1977) and a strong work ethic or
central life interest in work (Dubin, Champoux, & Porter, 1975) are related to
commitment, and that commitment is greater when the individual chooses to work
for the organization of his/her own volition (Mowday & McDade, 1979; O'Reilly
& Caldwell, 1981).
210 MEYER & ALLEN
satisfaction. We found, consistent with the theoretical work of Porter et al. (1974),
that commitment was influenced by job satisfaction whereas Bateman and Strasser
found the reverse. This discrepancy may, as anticipated at the outset, be due to
differences in the samples studied. The focus of the present study was on newly
hired employees because it was felt that commitment would be most sensitive to
differences in the quality of work experiences for these individuals. Although an
explanation based on differential sensitivity seems reasonable, however, other
explanations cannot be ruled out. There is clearly a need for more longitudinal
research to examine the development and maintenance of commitment at various
points in employees' tenure in the organization.
It is noteworthy that work experiences were found to predict commitment only in
the five- and nine-month lag. This may be due to the fact that most of the
participants in this study reported going through some form of orientation/training
during the first few months of employment. Consequently, they might have
recognized this period as being atypical of their ultimate experiences and therefore
"reserved judgement" on their feelings about the organization until they had the
opportunity to begin the work for which they were hired.
The fact that ratings of confirmed expectation at five months were influenced by
commitment at one month, although contrary to prediction, is consistent with
Salancik's (1977) view that commitment can influence perceptions. An employee
who has a strong attachment to an organization might view objectively similar
experiences within that organization quite differently than one who does not share
that attachment. The fact that, logically, commitment can both influence and be
influenced by perceived work experiences complicates research efforts and under-
scores the importance of using analytic procedures that are sensitive to the
direction of influence. More attention needs to be given in future research to
commitment's role as both an antecedent and consequence of perceived work
experiences.
A final purpose of this study was to examine the relationships between commit-
ment and self-report measures of work behaviour and turnover intention. Signifi-
cant concurrent and cross-lagged correlations were obtained for several of the
consequence measures, most notably, motivation, performance, intent to search
for alternate employment, and intent to leave the organization within the next year.
The cross-lagged regression analyses, however, provided no evidence for the
influence of commitment on these behaviours/intentions over time.
The fact that commitment was not found to influence turnover intention is
somewhat surprising in light of the fairly strong support obtained in previous
research (e.g., Porter et al., 1976; Porter et al., 1974). It is possible that the failure
to detect an influence was, at least in part, a function of the length of the lag
involved. If commitment during this early period of employment is particularly
volatile, as has been suggested (e.g., Mowday et al., 1982), it might be a better
predictor of turnover intention in the short term than in the long term. The same
212 MEYER & ALLEN
change not only in the approach taken to commitment research (e.g., more
longitudinal research), but also to the way commitment is conceptualized (Allen &
Meyer, 1985).
RESUME
Le but de cette etude longitudinale etait d'examiner le developpement et les consequences de l'implica-
tion ("commitment") organisationnelle tot dans la carriere d'employes. Des etudiants ayant termine
l'universite et qui avaient recemment accepte un emploi permanent a temps plein repondirent au
questionnaire sur 1'implication organizationnelle (OCQ) et des mesures d'auto-iapport sur leurs
experiences de travail, leur comportement au travail et leur intention de quitter leur emploi ("turnover")
apres un, cinq et neuf mois a leur emploi. Une comparaison des moyennes sur une base ternporelle
revcla un declin significatif de l'implication et I'aspect favorable de leurs experiences de travail durant
les premiers neuf mois de travail. Les resultats d'analyses de regression a travers le temps ("cross-
lagged regression analyses") apporterent despreuves en accord avec l'hypothese que les experiences de
travail initiales influencent l'implication dans l'organization; par centre ces preuves etaient en desac-
cord avec l'hypothese voulant que l'implication influence le comportement de travail et I'intention de
changer d'emploi.
REFERENCES
Allen, N.J., & Meyer, J.P. (1985, June). Organizational commitment: Development and assessment of a
model. Riper presented at the annual meeting of the Canadian Psychological Association, Halifax,
Nova Scotia.
Angle, H., & Perry, J. (1981). An empirical assessment of organizational commitment and organiza-
tional effectiveness. Administrative Science Quarterly, 26, 1-14.
Arnold, H.J., & Feldman, D.C. (1982). A multivariate analysis of the determinants of job turnover.
Journal of Applied Psychology, 67, 350—360.
Bateman. T.S., & Strasser, S. (1984). A longitudinal analysis of the antecedents of organizational
commitment. Academy of Management Journal, 27, 95—112.
Becker, H.S. (1960). Notes on the concept of commitment. American Journal ofSociology, 66, 32-42.
Bray, D.W., Campbell, R.J., & Grant, D.L. (1974). Formative years in business. New York: Wiley-
Interscience.
Buchanan, B. (1974). Building organizational commitment: The socialization of managers in work
organizations. Administrative Science Quarterly, 19, 533-546.
Clegg, C. (1983). Psychology of employee lateness, absence, and turnover: A methodological critique
and an empirical study. Journal of Applied Psychology, 68, 88-101.
Dubin, R., Champoux, J.E., & Porter, L.W. (1975). Central life interests and organizational commit-
ment of blue-collar and clerical workers. Administrative Science Quarterly, 20, 411-421.
Ferris, K.R., & Aranya, N. (1983). A comparison of two organizational commitment scales. Personnel
Psychology, 36, 87-98.
Hall, D.T. (1976). Careers in organizations. Santa Monica, CA: Goodyear.
Horn, P. W., Katerberg, R., & Hulin, C.L. (1979). Comparative examination of three approaches to the
prediction of turnover. Journal of Applied Psychology, 64, 280—290.
Jamal, M. (1984). Job stress and job performance controversy: An empirical assessment. Organiza-
tional Behavior and Human Performance, 33, 1—21.
James, L.R., Mulaik, S.A., & Brett, J.M. (1982). Causal analysis. Beverly Hills, CA: Sage.
Kiesler, C A . (1971). The psychology of commitment: Experiments linking behavior to belief. New
York: Academic Press.
214 MEYER & ALLEN