Some antecedents of employee commitment and their influence on job performance : a multi foci study rvosK :

The challenge of using scarce and limited resources to satisfy almost limitless needs will, from a management perspective, be like the proverbial cat: it will never go away. The optimal use of human resources, in particular, remains a daunting task:. In an economic environment characterized by increasing global competitiveness, failure to reali:re this important objective could be organizationally terminal, as inefficient organizations are unlikely to survive over the long term. A variety of different measures could be used to evaluate organizational effectiveness. In this study, the individual job performance level of employees is regarded as an indicator of organizational effectiveness. It is hypothesi:red that the individual job performance of employees can be improved by enhancing employee commitment (commitment to the organization, job, supervisor, profession). In other words, the general notion is that, if employees perceive a high level of congruence between their individual objectives and those of the organization, job, supervisor, and profession, they are likely to be better performers. The empirical results showed that commitment to the profession has the strongest positive influence (p < 0.01) on job performance. The impact of organizational commitment was also positive, but only at the 5% level. Neither job involvement (commitment to the job) nor commitment to the supervisor had any influence on job performance. All the antecedents modelled exerted some influence on the different types of commitment. Internal locus of control exerts a negative influence on all of them, and career factors exert a positive influence on all of them. Both self-esteem and anticipatory socialization enhance organizational commitment and commitment to the profession, while external locus of control's influence is limited to enhancing job involvement.


Introduction
Economics could be defined as the study of the optimal use of scarce resources or, in the words of Skinner & Ivancevich (1992: 17), as the study of how a society (people) chooses to use scarce resources to produce goods and services to be distributed to people for consumption. Providing for people's multiple and often unlimited needs with limited resources requires the effective operation of, especially, business organizations. At a microlevel, business managers are thus continually grappling with the challenge of using scarce resources optimally, that is, pursuing organizational effectiveness.
One indicator of organizational effectiveness is the job performance levels of individual employees. This study focusses primarily on the management of human resources as a means to address the requirement of organizational effectiveness. Human endeavour in business organizations is typically measured by means of the job performance levels of individual employees.

Individual job performance
Job performance has at times been described as the extent to which an employee accomplishes assigned tasks (Cascio, 1992: 260), the time and energy put into a job (Sujan, 1986) and the productivity of the employee (Seigel & Ruh, 1973: 322). The latter includes the dependability of work behaviour as prescribed by the organization as well as spontaneous, innovative extra-role behaviour that goes beyond the prescribed standard (Angle & Perry, 1981: 2).
There can be little doubt that the increasing global nature of international business, given added impetus by the GATT agreement, will intensify calls for increased productivity and higher quality. In the final analysis, it boils down to a more efficient work force.
Performance appraisal systems play a crucial role in facilitating a more efficient work force. A universal objective of performance evaluation, irrespective of the circumstances, is however, the optimal utilization of this valuable resource. Appraisal permits the identification of weaknesses and ought to lead to improvements in work performance (Waldman & Kent, 1990) by, amongst others, providing employees with appropriate performance feedback (Sherman & Bohlander, 1992).
From a managerial perspective, the antecedents of actual performance are particularly important. Once the influence of antecedents on performance is known, it may be possible for managers and supervisors to create a working environment conducive to optimal performance.
Irrespective of how performance is described or appraised, it is important for the effectiveness of an enterprise (Gerber, Nel & Van Dyk, 1987: 17-18) and is one of the most widely used measures of organizational effectiveness (Sherman & Bohlander, 1992).

Employee commitment
It has been suggested that the objective of organizational effectiveness could be achieved by creating a working environment in which employees identify with their organization's values, goals and objectives, develop a positive attitude towards their jobs, identify with their superiors and identify with their occupation/professional group (Lobeland St. Clair, 1992: 1066Mowday, Steers & Porter, 1979: 237;Seigel & Ruh, 1973: 323;Tharenou, 1993: 282). This is collectively known, at least for the purposes of this study. as employee commitment. The general hypothesis is that, if employees perceive a high level of congruence between their individual objectives and those of the organization, job, supervisor and profession, they are likely to be better performers.
This study explores the relationship between different types of employee commitment (commitment to the organization, the job, the supervisor, the profession) on the one hand and individual job performance on the other hand. The influence of selected personal and work experience variables on the four types of employee commitment as a means of managing organizational effectiveness will also be considered.

Multifoci commitment
A review of the literature on employee commitment suggests that there are two schools of thought on the nature of the concept. The first could be called the 'dimension of organizational commitment approach', and its proponents argue that certain dimensions/components of organizational commitment actually are independent types of employee commitment. Angle & Perry (1981: 4 ), for example, distinguish between value commitment (commitment to support organizational goals) and commitment to stay (commitment to retain organizational membership). Value commitment is called normative commitment by Becker & Billings (1993: 179) and moral involvement by Etzioni (1975). Commitment to stay resembles Etzioni's (1975) 'calculative involvement'.
The second approach to employee commitment could be labelled the 'foci of commitment approach' (Reichers, 1985: 465). Foci of commitment refers to the individuals and groups to whom an employee is attached (Becker & Billings, 1993: 177). Reichers (1985: 472) suggests that there are a number of foci that collectively constitute the organization. They include co-worker, superiors, subordinates, the customers and other organization-related groups such as labour unions. Following this view, one could refer to co-worker commitment, customer commitment and commitment to superiors, depending on the organizational group under focus.
The underlying notion of this approach posits that employees have multiple memberships depending on their placements within the vertical structure of the organization. Each membership makes claim to a different form of work-related commitment (Zaccaro & Dobbins, 1989: 267). Organizational commitment, according to Reichers (1985) should therefore be viewed as a collection of multiple commitments to various groups in the organization (Witt, 1993: 18) such as professional associations, labour unions, their career (profession), their supervisor, their organization, top management and their workgroup (Fukami & Larsen, 1984;Reichers, 1985;Mueller, Wallace & Price, 1992: 214-215;Becker, 1992;Gregersen, 1993).
In this study, the focus is on four foci of employee commitment: attitudinal commitment to the organization, work or job commitment, commitment to the supervisor and commitment to a profession. Collect.ively, these four foci of commitment or types of commitment will be referred to as employee commitment. S. Afr. Tydskr.Bedryfsleiding 199S 26{ 4 )

Employee commitment and job performance
Resea~ch has shown that different types of employee commitment (Mueller et al., 1992) have different effects on organizational outcomes such as job performance (Wiener & Vardi, 1980: 89;Lobel & St. Clair, 1992: 1066Gregersen, 1993: 40). The following empirical findings point to a positive relationship between different types or foci of com,. mitment and job performance: -Organizational commitment and job peformance (Mow. day et al., 1979: 237;Boshoff & Meis, 1994).
- Lobel & St. Clair (1992: 1066 have produced evidence that career identity salience has a direct effect on self· report measures of work effort. Career identity salience refers to certain identities people attach themselves to, such as, 'I am an accountant' and 'I am a lawyer'. Attaching oneself to a certain professional identity could thus have a positive influence on job performance. Based on the preceding literature review, the following hy· pothesis will be considered: H 1 : Employee commitment exerts a positive influence on job performance. Mathieu & Farr ( 1991: 130) have urged researchers to investigate the common as well as unique antecedents of different types of commitment and to examine how they jointly influence employees' work and non-work-related behaviour. In this study, the influence of four personality variables on employee commitment will be considered (protestant ethic, self-esteem, locus of control and need for achievement) as well as two work experience variables (career factors and anticipatory socialization). As very little work on multiple commitments in the working environment have been done, the literature review fo· cusses on the relationship between selected personality and work-related variables on the one hand, and organizational commitment and job involvement on the other hand. The hy· pothetical relationships depicted in Figure 1 are thus based on the work pertaining to organizational commitment and job involvement, assuming that similar relationships would be found with respect to commitment to the profession and the supervisor.

Protestant ethic
The protestant ethic refers to the individual's work orientation or internalized work values (Lodahl & Kejner, 1965: 25;Dubin et al., 1975: 417). 1be holder of the protestant ethic could be described as a person committed to the values of hard work, to work itself as an objective and to the work organization as the preferred structure within which those internalized values can be satisfied. An individual. therefore. who has a central life interest in work, will prefer the work place as the institution to carry out these activities.
It has been reported that individuals who have a strong central life interest in work Gob-oriented), tend to exhibit higher levels of organizational commitment than non-job-orientated individuals (Dubin et al., 1975;Mueller et al., 1992: 227;Pierce & Dunham, 1987: 174) as well as higher levels of commitment to their work, or job involvement (Mueller et al., 1992: 227).

Sett-esteem
Self-esteem is the evaluation that the individual makes and customarily maintains with regard to him/herself. It expresses an attitude of approval or disapproval and indicates the extent to which the individual believes him/herself to be capable, significant, successful and worthy. In short, self-esteem is a personal judgment of worthiness that is expressed in the attitudes that the individual holds (Burns, 1979: 55) and is simply a positive or negative attitude the individual holds of him/herself. Buchanan (1974: 543) has reported that the employee's self-image and perceived personal importance to the organization are important determinants of organizational commitment.

Locus of control
Locus of control is defined as a generalized expectancy that rewards, reinforcements and outcomes in life are controlled either by one's own actions (intemality) or by other (external) forces (Spector, 1988: 335).
Pierce & Dunham (1987: 174) investigated the locus of control-organizational commitment relationship and reported that internals have a higher propensity to become organizationally committed. In a more recent study, Colarelli & Bishop ( 1990: 165) found internal locus of control to be negatively related to organizational commitment. This means that higher organizational commitment is associated with higher external locus of control (meaning lower internality).

Need for achievement
A number of different definitions of need for achievement have been suggested, of which Korman 's ( 1971: 51) 'desire to exceed some standard of behaviour' seems most appropriate.
High need for achievement individuals seem to be people who find achievement a satisfying experience; who prefer situations of moderate risk, where performance feedback is provided, and where they, not somebody else, bear the responsibility. They also seem to prefer activities in which performance is based on ability, not luck, and measured against some standard of excellence; and they focus on the satisfaction of success rather than on the embarrassment of failure (Korman, 1971: 52;Hampton, Summer & Webber, 1982: 64).
The preceding review suggests that positive perceptions of personality variables will enhance employee commitment, a contention on which the second hypothesis is based: 127 H 2 : Personality variables as measured by protestant ethic, self esteem locus of control and need for achievement exert a positive influence on employee commitment.

Work experience variables modelled as antecedents
This study focusses on two work experience variables, namely, career factors and anticipatory socialization. Career factors refer to the employee's perception of certain career rewards which include advancement, satisfying appointments, fairness of advancement policies and opportunity to participate in career planning (Jans, 1985: 386).
Anticipatory socialization means the degree to which an individual is given an accurate representation of the job or organization and congruency is established between the individual and the job (Sager & Johnston, 1989: 32).

Career factors
As new employees enter the organization with various expectations, a psychological contract is 'enacted' between the new employee and the organization (Zaccaro & Dobbins, 1989: 267). The extent to which a true picture of the organization is presented to the new employee influences the person-job fit accomplished and, eventually, the organizational commitment of the employee (Buchanan, 1974: 534).

Anticipatory socialization
Getting the best people to join the organization and how these people are brought into the organization (anticipatory socialization) have a major influence on their future relationship with the organization and their subsequent performance (Northcraft & Neale, 1990: 453 and 474;Pierce & Dunham, 1987: 174).
Poor person-organization fit causes job stress as a result of an incongruent relationship between the employee and the organization (Feldman, 1981: 309) which leads to work dissatisfaction, decreases in job performance and high levels of absenteeism and turnover (Wright & Bonett, 1991: 135). Stumpf & Hartman ( 1984 ), for example, have reported that realistic expectations and the amount of environmental exploration and information that the employee has acquired about the job and organization prior to entry, have contributed to affective organizational commitments via their effect on person-job congruence and quality of work experiences. Meyer, Bobocel & Allen (1991: 729) conclude from their findings that on-the-job experiences by the employee during the first month of employment are very important for the development of affective organizational commitment. Self-reports by the employees of the degree to which their expectations have been confirmed after one month of employment were found to predict commitment as measured six to eleven months after entry.
It appears from the preceding review as if work experience variables could influence employee commitment, a contention on which the third hypothesis is based: H3: Work experience variables, as measured by career factors and anticipatory socialization, exert a positive influence on employee commitment.

Objectives
The broad objective of this study is to investigate whether employee commitment (as measured by the employee's organizational commitment, job involvement, commitment to the profession and commitment to the supervisor) influences job performance. The study will explore the influence of selected personality variables and work experience variables, as antecedents, on employee commitment. The hypothesized relationships are graphically depicted in Figure I.

Sample
Three occupational groups were surveyed during this study: chartered accountants, teachers and office administrative personnel.
A random sample of one thousand professionally registered chartered accountants were selected. A questionnaire, a stamped, self-addressed envelope and a covering letter were mailed to each respondent. An identical questionnaire was mailed to 500 teachers and 500 office administrative personnel of three academic institutions.
A total of 382 useable questionnaires were returned which represents a response rate of 19.1 %. The response rate per individual sub-sample was: 12.8% (128) for chartered accountants, 30.2% (151) for teachers and 20.6% (103) for administrative personnel.

Measuring instruments
As far as possible, measuring instruments with confirmed validity and reliability were used to measure the different variables included in the study. Organizational commitment was measured using the shortened, nine-item version of the Mowday et al. (1982) Lodahl & Kejner (1965). Job involvement, in this study, was used as a measure of 'commitment to the job'. As a consequence of our definition, a self-developed scale was used to measure commitment to the profession, using selected items from the OCQ scale of Mowday et al. ( 1982). A self-developed four-item scale was used to measure commitment to the supervisor by using the same four items extracted from the OCQ scale (Mowday et al., 1982) for the construction of the commitment to the profession scale. The word 'organization' was replaced by 'supervisor' in the phrasing of the items. Respondents were requested to mark their responses to the items on a seven-point Likert-type scale.
The 19-item Mirels & Garret (1971) scale was used to measure protest and ethic. The items were linked to a sevenpoint Likert-type scale while the Rosenberg (1965) scale was used to measure self-esteem in this study.
Need for achievement was measured with a sub-scale of the Manifest Needs Questionnaire (MNQ) of Steers & Braunstein (1976). The seven-item scale of Jans (1985) was used to measure perceptions of career rewards such as promotion, satisfying senior appointments, fairness of advancement policies and participation in career planning.
The instrument used to measure anticipatory socialization was originally developed by Sager & Johnston (1989: 39) and measures the congruency between the employee and the job/ organization as well as the reality of the individual's expectation with respect to the organization.
Four approaches are typically used to measure job performance: supervisory evaluation, self-appraisal, peer evaluation and some form of objective measure such as sales figures. It appears as if most scholarly studies of job performance have used ratings by either supervisors or self appraisal (Steers, 1977: 50;Huselid & Day, 1991: 383;Tharenou, 1993: 274  All types of individual job performance measures have, however, been subjected to some form of criticism. In this study, self-appraisal was used to measure job performance. A self-devised self-report measure of job performance was used in this study which produced a Cronbach alpha of 0.93 in previous use (Boshoff & Meis, 1994). Respondents were asked to respond to a seven-point Likert-type scale, rating themselves on how well they do their jobs, how successful they are on their jobs and to what extent they think their job performance is above average.

Internal reliability
The first step in the data analysis procedure was to assess the internal reliability of the measuring instruments by means of Cronbach alpha coefficients using the computer programme SAS (SAS Institute 1988). The results reported in Table 1 show that all the instruments returned alpha values of more than 0.7 except for need for achievement (0.610) and job involvement (0.603). The item-to-total correlation of each item was then considered to establish whether the alpha values could not be improved by deleting items with low (below 0.35) item-to-total correlations. The removal of low item-to-total items improved the internal reliability of four instruments. Removing three items from the job involvement scale (INVOLVl,INVOLV2 and INVOLV6) improved the coefficient from 0.603 to 0.705. In the case of the organizational commitment scale, the removal of OC 1 and OC8 marginally improved the alpha value of the organizational commitment scale to 0.874. The coefficient of commitment to the supervisor increased from 0.757 to 0.831 when SUP-COM3 was removed. By deleting four locus of control items (LOCUS 5,7,8 and 12), the coefficient improved to 0.817. Although the removal of one need for achievement item (NACHl) with a low item-to-total correlation(< 0.35) did not really alter the initial alpha value, it was nevertheless removed to remain consistent.
After the removal of these items, only need for achievement remained with an alpha value below 0.7. Although some may argue that this value justifies its complete removal, Nunnally ( 1967) suggests values of 0.6 and even 0.5 are accept- able for preliminary research. It was therefore decided to retain need for achievement in the theoretical model.

Discriminant validity
The second phase of analysis involved an assessment of the discriminant validity of the measuring instruments. For this purpose, two factor analysis procedures were conducted using the computer programme BMDP4M (Frane, Jennrich & Sampson, 1990). Maximum likelihood was specified as the method of factor extraction and a Direct Quartimin oblique rotation of the original factor matrix was used (Jennrich & Sampson, 1966) in both instances. The first factor analysis involved the antecedents depicted in Figure 1 and the second the various types of commitments.
In the first instance, the extraction of six factors was specified as modelled in Figure I. It was surmised that each of the six variables modelled (protestant ethic, self-esteem, locus of control, need for achievement, career factors and anticipatory socialization) are separate and distinct constructs but that their 'separateness' needs to be empirically verified. The resultant empirical evidence did not, however, support this contention. After considering various options, including four, five, six, seven and eight factor solutions, it had to be concluded that the instrument used to measure protestant ethic did not demonstrate sufficient evidence of discriminant validity. As a result, protestant ethic was removed from the model and was not incorporated in subsequent statistical analyses.
The most interpretable factor structure was the one reported in Table 2. Table 2 shows that factor 1 consists of all the items expected to measure self-esteem (except STEEM2 which did not reach the cut-off point of 0.4) as well as three of the fouritems expected to measure need for achievement. In this study, respondents thus did not distinguish between the two constructs. The three need for achievement items (NACH2, NACH3 and NACH4) were thus regarded as additional measures of the construct self-esteem. As NACH5 and STEEM2 did not load to a significant extent on any factor, they were excluded from subsequent statistical analyses.
Seven items which are regarded as measures of external locus of control loaded on factor 2. Factor 3 consists of three of the seven items expected to measure career factors (the others, CAREER 1, 2, 4 and 5, did not load on any factor and were subsequently excluded) and factor 4 of the three items expected to measure anticipatory socialization. Four items are regarded as measures of internal locus of control (Factor 5).
The second factor analysis considered the four types of commitment modelled in Figure 1. Table 3 shows that all the items expected to measure the various constructs loaded on a separate factor, confirming the discriminant validity of the instruments used to measure the various constructs in the theoretical model (Figure 1 ).
From the two factor analyses described (and the resultant Tables 2 and 3), Table 4 identifies the items which were regarded as measures of the individual latent variables retained in the theoretical model. Only these items were used in all subsequent statistical procedures.

Empirical results
The various relationships depicted in Figure 1 were tested using a series of multiple regression analyses procedures  According to Table 5 (and Figure 2), commitment to the pro-(I) Loadings greater than 0.4 were considered significant fession is significantly related to job performance (p < 0.01).
This finding supports the positive relationship suggested by available on the computer programme SAS (SAS Institute, Lobel & St Clair ( 1982) and implies that the stronger the employee's commitment to his/her profession, the higher the 1990). The first regression analysis considered the influence job performance. The results also indicate a significant posiof employee commitment (the four types of commitment as independent variables) on job performance. The other regres-tive relationship between organizational commitment and job sion analyses had the four types of commitment as dependent performance (p < 0.05) which implies that highly organivariables and the modelled personality and work experience zationally committed employees are likely to be better pervariables as independent variables.
formers. As pointed out earlier, this relationship has generally The empirical results are reported in Tables 5 and 6 and been found to be positive but not always to a statistically graphically depicted in Figure 2. significant extent (Mowday et al., 1979). Job involvement and commitment to the supervisor, on the other hand, have no significant relationship with job performance. Employees who psychologically identify with their daily work Uob) and their supervisor are thus not necessarily better job performers.
The hypothesis (H 1 ) that employee commitment exerts a positive influence on job performance, is accepted with respect to professional and organizational commitment but, rejected for job involvement and commitment to the supervisor. As far as job involvement is concerned. this finding supports those reported by Siegel & Ruh (1973) and Huselid & Day ( 1991) but contradicts those by Wiener & Vardi ( 1980) who reported a positive relationship with work effort. The latter was measured as the commission income of an insurance sales force.
The four types of employee commitment listed in Table 5 explain 36.9% (R 2 -value) of the variance in job performance.

Modelled antecedents of employee commitment
Two groups of antecedents of employee commitment were modelled in this study, namely. personality variables and work experience variables. Following the reliability and validity tests, the following personality variables were modelled as antecedents: self-esteem. internal locus of control and external locus of control. The work experience variables. used as independent variables, were career factors and anticipatory socialization.
To test H 2 and H3. a multiple regression analysis using the computer programme SAS (SAS Institute, 1990) was again used. The results arc reported in Table 6 and graphically depicted in Figure 2.  Table 6 highlights the importance of self-esteem, locus of control, career factors and anticipatory socialization as determinants of employee commitment. The coefficients of determination (R 2 ) show that the modelled personality variables (self-esteem and locus ofcontrol) and work experience variables (career factors and anticipatory socialization) explain a considerable 39.6%. 15.0%, 49.6% and 20.9% of the variance in organizational commitment, job involvement, professional commitment and supervisor commitment respectively.

Sett-esteem and employee commitment
The empirical results show that self-esteem has a significant (p < 0.01) influence on both professional commitment and organizational commitment. The relationship is a positive one in both cases, indicating that high self-esteem employees are more committed to their professions and employing organizations than are low selfesteem employees. This finding supports those reported by Buchanan (1974 ), and Morris & Sherman (1981).) The results thus suggest that satisfying self-esteem and higher order needs, such as the employee's need for achievement and personal importance, do not influence the employee's commitment to the job (job involvement) or supervisor. The absence of a relationship between job involvement and self-esteem is in accordance with the results of Cook & Wall (1981). S.Afr.Tydskr.Bedryfsleiding 1995 26(4) Locus of control and employee commitment Individuals with high external locus of control are more job involved. This relationship is positive (p < 0.05). External locus of control is not significantly related to professional commitment, organizational commitment and commitment to the supervisor.
The complete opposite has been found with respect to the effect of internal locus of control on employee commitment. Internal locus of control is negatively related to all four types of employee commitment (p < 0.01). In other words, the more employees believe that their job situation is determined by their own abilities and efforts, the less they will be committed to their organizations, their jobs. their supervisor and their profession.
The negative relationship between internal locus of control and organizational commitment contradicts those reported in earlier studies. Colarelli & Bishop ( 1990: 165) reported a positive correlation between internal locus of control and organizational commitment, and Pierce & Dunham (1987) reported a positive correlation between internal locus of control and organizational commitment propensity.
The hypothesis (H2) that personality variables, as measured by self-esteem and locus of control exert a positive influence on employee commitment, is rejected with respect to internal locus of control (all negative) and external locus of control (no relationship) except for job involvement. The absence of a relationship between self-esteem and job involvement, and between self-esteem and commitment to the supervisor, also leads to the rejection of H 2 in these instances.

Career factors and employee commitment
Significant positive relationships (Table 6 and Figure 2) were recorded between career factors and all four of the foci of commitment. With the exception of professional commitment (p < 0.05), the exceedance probabilities are all below O.oI.
The hypothesis (H 3 ), that work experience variables as measured by career factors exert a positive influence on employee commitment, is therefore accepted. These findings are in line with the majority of those reported in the past including Sager & Johnston (1989) and Huselid & Day (1991). Table 6 and Figure 2 also show significant relationships between anticipatory socialization on the one hand and organizational commitment and professional commitment on the other hand (p < 0.01). Establishing congruency between the employee and the job/organization, as well as giving a realistic job preview to the employee when joining the organization, have a significant influence on the employee's professional and organizational commitment. The relationship is particularly strong between anticipatory socialization and organizational commitment (p < 0.01 ). The relationship is also in a positive direction, indicating that healthy early socialization processes are associated with higher levels of professional and organizational commitment. These findings are in agreement with those reported by Steers (1977), Sager & Johnston (1989) and Zaccaro & Dobbins (1989).
The relationships between anticipatory socialization and commitment to the job and supervisor proved to be non-significant. The hypothesis (H 3 ), that work experience variables. as measured by career factors and anticipatory socialization. exert a positive influence on employee commitment, is thus accepted in the case of career factors. It is also accepted in respect of the influence of anticipatory socialization on organizational commitment and commitment to the supervisor. It is rejected in respect of job involvement and commitment to the supervisor.
The latter confirms the findings of Sager & Johnston ( 1989) who could not find a relationship between anticipatory socialization and managerial satisfaction but contradicts those by Huselid & Day (1991) who found a negative relationship between anticipatory socialization and job involvement.
The employee's perception of promotion opportunities, satisfying senior appointments and the fairness of the organization's advancement policies are strong determinants of the employee's commitment to the organization, profession, job and supervisor. Positive perceptions of career factors are thus associated with increased employee commitment.

Summary of empirical findings
The empirical findings show that both organizational commitment and commitment to the profession exert a positive influence on job performance. Neither job involvement nor commitment to the supervisor exert any influence on job performance. The influence of all the antecedents on the various types of commitment which are statistically significant are positive except for internal locus of control. Organizational commitment and commitment to the profession are positively influenced by self-esteem, career factors and anticipatory socialization but negatively influenced by internal locus of control. Job involvement is positively influenced by external locus of control and career factors and negatively influenced by internal locus of control. Commitment to the supervisor is negatively influenced by internal locus of control and positively influenced by career factors.

Managerial implications
Organizational commitment can be improved by enhancing the self-esteem of employees, ensuring that individuals perceive rewards such as promotions and appointments as fair, have the opportunity to participate their career planning (career factors) and ensuring that they are given an accurate representation of the job/organization prior to employment, as well as ensuring that the individual is successfully matched with the job/organization (anticipatory socialization). At the same time, attempts need to be made to reduce the extent to which they believe they are solely responsible for their job outcomes (internal locus of control).
The same applies to commitment to the profession and, except for self-esteem, also for job involvement. A reduction in the negative impact of internal locus of control and enhancing career factors is likely to improve commitment to the supervisor.
Professional commitment could be enhanced by giving a group of employees doing the same job a professional image, attaching to the group an occupational title, consulting them as an occupational group when decisions which concern them have to be made, enhancing social interaction among the group members and increasing their skills and interest in their chosen profession. 133 The positive relationship between self-esteem and both organizational commitment and commitment to the profession shows that efforts to improve the self-worth of employees will be rewarded with improved organizational effectiveness. These feelings could be enhanced through, amongst others, improved competence and skills through training, challenging job assignments, membership of a selected or prestigious organizational group and formal recognition and/or rewards.
Rewards and reward systems play an important role in the employee's commitment to the organization and withdrawal patterns. To make a useful contribution, reward systems need to be seen, however, as equitable by all employees, directly tied to actual performance, and tailored to the needs of employees (Robbins, 1992: 250).
The significant influence career factors have on all four components of employee commitment justifies the emphasis that management theorists have placed on satisfying careers for employees since the mid 1970s. Career development programmes formed a central part of this approach once managers realized that they are likely to benefit both the organization and the employees. From a managerial perspective, these programmes result in the better use of employees' skills and knowledge whilst providing employees with a personalized career path based on the needs of both the employee and the organization (Sherman & Bohlander, 1992: 240).
Internal locus of control exerts a strong negative influence on all four types of commitment. The negative impact of internal locus of control suggests that the more employees perceive themselves to be in control of their destiny the less likely they are to be committed to the organization, their jobs, their profession and their supervisor. The perception that what happens to employees in their working environment is largely determined by their own behaviours needs to be reduced by, for instance, emphasizing the virtues of teamwork, joint problem solving and collective effort. Rewarding team or group rather than individual performance may go a long way to realizing that objective.

Limitations and future research
It was unfortunate that the protestant ethic construct had to be removed from the theoretical model. An individual's work values or job interests should be important determinants of commitment to the job, organization and even profession. It could be one of the main reasons why job involvement has 'performed' weakly in this study. The inclusion of the protestant ethic in future studies should be considered but, as one reviewer pointed out, it should be contextualized for South African circumstances. In other words. the social, cultural and political conditions peculiar to South Africa need to be reflected in a future study.
This study did not examine the impact that various types of commitment could have on each other and the issue of conflicting commitment also needs to be addressed in more detail. Commitment to a labour union, for instance, is believed to be in conflict with organizational commitment. It could be useful to investigate how job performance, for example, would be affected if a 'conflict' commitment is included in the theoretical model in a similar study.