The Effect of Career Management on Public Employee Job Satisfaction in Jimma City: Taking Career Development as Mediator

The main purpose of this study was to examine the effect of career management practices on public employees’ job satisfaction in Jimma Cit y taking career development as mediating variable. Survey questionnaire was used to collect primary data from 224 employees of public institution in Jimma City and structural equation model using medsem Stata 14 command was applied to analyze mediation analysis. The findings revealed that career management elements such as career planning, succession planning and career policy has small effect on employee’s job satisfaction and career development has no mediation effect. These study findings contribute to a better understanding of career management practices and job satisfaction for public employees and provide empirical evidence on the relationship between two variables in public institutions.


Introduction
During the past three decades, a number of different debates have been going on about the role of human resource in organizational success and it is said that the only sources of sustained competitive advantage is HR which can create value in a unique way and which can neither be imitated not substituted (Armstrong, 2006;Armstrong and Taylor, 2014). Thus, people are the most valuable resource in contemporary organizations. However, in the global business, changes are constantly being observed on day-to-day basis requiring organizations to restructure and downsize processes, mergers and acquisitions, make technological advancements, and other actions to deal with the dynamic nature of globalization (Greenhaus, Callanan & Godshalk, 2000;Appelbaum, Ayre & Shapiro, 2002;Baruch, 2004as cited in Bola, 2011. Thus, such a constant changes at the organizational level have elevated the importance of managing people at work, and in particular, the planning and managing of their careers (Baruch, 2004as cited in Bola, 2011 through provision of development opportunities. Past studies, regarding the effect of career management elements (such as career planning, succession planning, career counseling, and career development), shows far reaching influence on various outcomes including job satisfaction (Koech, 2003;Bola, 2011;Ismail, Madrah, Aminudin & Ismail, 2013;Navern, Quraisha & Anis, 2015;Adamu, Daddie & Ebikeseye, 2017;Lydia & Nzulwa, 2018). These studies all agree that career management elements have both organizational and personal influence if done effectively and efficiently. However, particularly in the developing world, public institutions did not place any emphasis on the development and implementation of career management programmes or practices (Johannes, 2002). The employees' career paths were mainly uttered by the prescripts and actions applicable to a variety of professional classes in the public institutions. Consequently, as noted by Lydia & Nzulwa (2018), low employee performance and productivity led to decline in the customer satisfaction level. In Ethiopia, a comprehensive civil service reform programme was launched in 1994 with the aim of creating an efficient, transparent, accountable and professional civil service that responds to the newly adopted decentralized governance architecture. The programme paying attention on seven major areas including strengthen human resource management and control; improve performance and service delivery; improve transparency and accountability in the federal civil service; construct policy and institutional governance capacity in the four of emerging regions; advance governance of financial resource management practice and control; fortify reform coordination structure and strengthen top management of the federal civil service (Institute of Leadership and Good Governance, 2013). However, little is known with regard to the contribution of such reforms in improving career practices and job satisfaction in public institutions which in turn could affect performance and quality of service delivery. Research on career management emphasizes the importance of career dynamics and development in matching individual needs with an organizational need (Armstrong and Taylor, 2014). In the extant career development literature it is widely acknowledge that career management aims to provide employees with opportunities for developing their talents which helps to satisfy future aspirations. As noted by Luthans (2008) job satisfaction is the outcome of employees' perceptions of how their job provides something that is consider important. Therefore, this study aims to extend such as insight in the career development literature through providing theoretical and practical contributions as discussed in the following few paragraphs below. This study is believed to contribute in the following ways for career development literature. First, several studies (Rafique, 2014;Navern et al. 2015;Tobing, 2016;Abdul, Mahfudnurnaj, Syahrir and Baharuddin, 2016;Lydia & Nzulwa, 2018) were conducted so far for examining the effects of various elements of career management on employee job satisfaction. But these studies failed in revealing the context in public institutions. The studies were conducted in banks (Koech, 2003;Bola, 2011;Sobia, Saira, Faisal and Ishtiaq, 2013;Tobing, 2016), universities and tech companies (Tser-Yieth, 2003) and manufacturing firms (Navern et al., 2015). Hence, they may not reflect the realities of public institutions due to contextual differences between various organizations. As noted by Susana & Anna (2018) cultural values could influence the perceptions of job satisfaction and its determinants, so public employees from different cultures may display different levels of job satisfaction, though the literature is open to doubt on this topic. Therefore, this study will contribute to understanding on the effect of career management practice on job satisfaction in public institutions and provides perspective from developing countries like Ethiopia. Second, few studies (e.g., Bola, 2011;Ismail, Madrah, Aminudin & Ismail, 2013) conducted on the mediation role of career development in the relationship between career management elements and job satisfaction. However, they have methodological flaw mainly in the mediation analysis. Both studies used regression analysis for mediation analysis. As noted by Iacobucci et al. (2007 as cited by Mehmetoglu, 2018) using the regression (REG) technique suffers from a serious drawback by consistently producing larger standard errors for the path coefficients than does the Structural Equation Model (SEM) technique. Hence, Mehmetoglu (2018) suggested the structural equation framework using medsem Stata command for two benefits it provides in the mediation analysis. One is that SEM helps to make mediation estimation in optimal, more proper and complete way. Secondly, SEM can facilitate mediational analysis using observed or latent variables as well as combinations of observed and latent variables. Therefore, this study used SEM technique using medsem Stata command for analysis of mediating role of career development in the relationship between career management elements and job satisfaction in public institutions found in Jimma City. By doing so, this study believed to contribute for wider literature on how career development mediates career management and job satisfaction in public institutions. Third, this study believed to contribute to existing human resource literature in general and understanding of career management elements or constructs in particular. There is great deal of confusion in the literature on the building blocks of career management. For example, Bola (2011) and Ismail et al. (2013) synonymously used career planning and career management as distinct constructs their respective studies. In contrary, Armstrong and Taylor (2014) suggested a career management process comprised of career management policy, talent audit, career planning, career counseling, succession planning, and career development. Thus, this study adopted the model by Armstrong and Taylor and provided empirical evidence on the constructs of career management in public institutions.

Literature Review and Hypotheses
According to Armstrong (2006), career management refers to providing opportunities for people to progress and develop their careers and ensuring that the organization has the flow of talent it needs. On the other hand, Greenhaus et al (2000) defined career management as a process by which individuals develop, implement and monitor career goals and strategies. Thus, career management is a continuous process of work life. Martin, Romero, Valle & Dolan (2001) argue that the career management initiative is a nexus between organizations and individuals where organizations endeavor to match individual interests and capabilities with organizational chances through a premeditated programme including activities like career systems, career counseling, job rotation and other career management tools and resources. In addition, Brown (1998) argues that career management activities must support individuals in their efforts to develop the knowledge, skills, and behaviours that will enable them to be successful. The combination of both individual career planning and organizational career management practices results in career development when the individual reaches the career goals set.
The extant literature provides various but very similar definitions of job satisfaction. According to Robbins and Judge (2007), job satisfaction is a positive emotion regarding the work of an employee who comes into view from the assessment of his/her characters. Similarly, others (e.g., Nelson and Quick, 2006;Luthans, 2008) also noted job satisfaction as a positive emotional state resultant from evaluating one's job experiences. They describe it as how an individual is feeling with his or her job. Job satisfaction is employees' sense of achievement and success and generally seems to be straightforwardly linked to productivity as well as to personal wellbeing. Thus, a person with a high level of job satisfaction holds a positive feeling about the job, while a person who is dissatisfied with his/her job holds a negative feeling about the job. Career management elements and job satisfaction: According to Armstrong and Taylor (2014), career management process starts from setting career management policies and development process. It involves career management policies, talent audits, succession planning, career planning and career development. In relation these elements, the extant research confirms significant effect on job satisfaction as briefly discussed in the following few paragraphs. The conceptual framework presented in figure 1 below shows the relationship between variables of this study. Career management policy and job satisfaction: According to Khaled (2017), if employees notice that promises of a professional development policy are inconsistent with their expectations, in consequence, the employees neglect their job and only perform the responsibility that they are formally assigned to perform. In addition, job satisfaction is connected with pay, work-related anxiety, strengthening, organization and company policy, accomplishment, self-awareness, association with others, and the general working state (Rahman, Waheed and Saad, 2017). H1: Career management policy has significant effect on job satisfaction of public employees in Jimma City. Career planning and Job satisfaction: Hall and Associates (1986) tried to define career planning as a purposeful process for becoming awake of self opportunities, restrictions, selections and consequences, and identifying career linked goals, and programming for work, education, and related developmental experience to provide the direction, timing and sequence of steps to attain a particular career goal. Leibowitz, Farren & Kaye (1986) argue that individuals are responsible for initiating their own career planning as well as identify their skills, values and interests, and seek out individual career choices in order to lay down goals and set up individual career plans. The extant research confirmed the significant effect of career planning on satisfaction. For instance, Bola (2011) showed that the significant link between the variables of career planning and career management, and career development, and in turn, with job satisfaction and career commitment. Rafique (2014) also supported Bola's finding by concluding significant impact of career planning and development on job satisfaction, organizational commitment and employee retention in Pakistani banking sector. H2: Career planning has significant effect on job satisfaction of public employees in Jimma City. Succession planning and job satisfaction: As noted by Iles and Preece (2010), succession planning can be seen in terms of identifying successors for key posts and then planning career moves and/or development activities for potential successors. H3: Succession planning has significant effect on job satisfaction of public employees in Jimma City.
Career development and job satisfaction: As defined by Nadler (2007), career development is an organized learning experience in a given time period to increase the likelihood of improvement in growth performance of sense of duty, in accord with the education, training and environmental change in an activity of the organization. Additionally, career development is the lifelong process of performing learning, work, leisure, and transitions aiming to move toward a personally firm and growing preferred future (Armstrong, 2009). Therefore, it affects job satisfaction. H3: Career development has significant effect on job satisfaction of public employees in Jimma City. Mediating role of career development: Past studies provided insight on the mediating role of career development in the relationship between career management elements and job satisfaction. For instance, Bola (2011) found significant link between the variables of career planning and career management, and career development, and in turn, with job satisfaction and career commitment. Similarly, Ismail et al. (2013) revealed that career development does act as an important mediating variable in the relationship between career programs and personal outcomes like job satisfaction and career commitment. H4: Career development significantly mediates between career management elements and job satisfaction of public employees in Jimma City.

Figure 1: The conceptual framework of the study
This study was conducted in all Sectorial public institutions found in Jimma City. Cluster sampling technique was used to select sample respondents categorized based on Sectorial orientation and sample size was determined by using the method developed by Carvalho (1984). Survey questionnaire with five point lickert scale was collected from 224 employees in selected public institutions. The questionnaire consists of items about career planning, succession planning, career policy, career development and job satisfaction. All the constructs were based on reflective items (Diamantopoulos, Riefler and Roth, 2008 as cited in Susan & Anna, 2018) using Lickert scales (from 1 to 5). The individual questions for each construct were obtained and adapted from the literature, where possible, to fulfill with content validity. Three types of validity assessment were conducted such as internal, external, and construct. The content was assured well explanation of the independent variables directly responsible for the effect of the dependent variable in the literature review section while external validity was reached using well known sampling technique and collection instrument like questionnaire. In addition, validity was attained through a clear operational definition of the independent variables of the study. Moreover, the questionnaire used in the study was tested for consistency, in the pilot study, using Cronbach's Alpha test. According to Nunnally (1998), Cronbach's alpha test values which are greater than 0.7 are used as a threshold to measure the reliability of the questionnaire. Thus, as presented in the tables 1 below, the Cronbach's Alpha test result for all variables are greater than 0.7 and hence, accepted. The data was using Spearman rho and Structural Equation Model (SEM). First, Spearman rho assessment was used to examine the relationship between independent and dependent variables of this study. In the cases like collecting Likert scale data, experts have contended that the Spearman rho assessment should be used for analysis instead of parametric tests(e.g., t tests, analysis of variance, Pearson correlations, regression), which, strictly speaking, require interval data (Jamieson, 2004). Second, SEM was conducted to investigate the mediating role of career development in the relationship between career management practices and employee job satisfaction. Stata built-in command for Structural Equation Modeling (SEM) called medsem was employed. This method (medsem) is a post-estimation command that is run after estimating a mediation model. It provides a post-estimation command testing mediational hypotheses using Baron and Kenny's (1986 as cited by Mehmetoglu, 2018) approach modified by Iacobucci et al. (2007) and an alternative approach suggested by Zhao et al. (2010as cited by Mehmetoglu, 2018. As suggested by Mehmetoglu (2018), benefit of medsem is that it can contribute to conducting a proper and complete mediational analysis based on even very complex models (including observed and/or latent variables and with multiple mediators) due to the simultaneous estimation capability of structural equation modeling (SEM) technique.

Results and Discussion
This section presents findings of the study and discussion by comparing with past studies. Table  2 below show the characteristics of the sample respondents who participated in this study. Source: Survey questionnaire (2020) The study conducted factor analysis technique called Factorial Analysis (FA) in order to examine the structure of items in the questionnaire. According to Van Kesteren & Kievit (2020), factor analysis a method to explain patterns of correlations within a set of observed variables. That means factorial method helps to identify sets of highly correlated variables and infer an underlying factor structure. The factorial objective is to reproduce a data structure, as well as possible only using a few factors in the final model. The factorial analysis (table 3) resulted four constructs namely career planning, succession planning, career management policy, and career development. Initially, the study proposed two independent variables namely career planning and succession planning and one mediating variable called career development which are expected to have effect on employees' job satisfaction. However, the factorial analysis (table 3) shows that items ranging from 37 through to 40 explain one construct resulting in new fourth factor. The researcher examined these items and named the factor as career management policy. Therefore, the study employed three independent variables (namely career planning, succession planning, and career management policy) and one mediating variable (career development) in final model to examine their effect on employees' job satisfaction. This result is rewarding to the extant literature in management field in general and HR in particular in that career management policy is a separate construct in career management which has effect on job satisfaction. The result also provides empirical support to theoretical model of career management process proposed by Armstrong and Taylor (2014). The study also sought to examine the relationship between career planning, succession planning, career policy, and career development with public employees' job satisfaction in Jimma City. The study findings indicate that career policy (r=.69, p<0.05) has strong association with employees' job satisfaction of public institutions in Jimma City while career planning (r=.31, p<0.05) and succession planning(r=.30, p<0.05) have moderate relationship. However, the correlation result revealed that career development has weak association with employees' job satisfaction of public institutions in Jimma City. Generally, career planning, succession planning, career management policy and career development have significant and positive relationship with job satisfaction in public institution of Jimma City. This result is consistent with past studies (Bola, 2011;Sobia et al., 2013;Ismail et al., 2013;Naway and Haris, 2017) which also found that career planning, succession planning, career management policy, and career development has significant and positive association with employees' job satisfaction.   Mehmetoglu, 2018). The benefit of medsem is that it can contribute to conducting a proper and complete mediational analysis based on even very complex models (including observed and/or latent variables and with multiple mediators) due to the simultaneous estimation capability of structural equation modeling (SEM) technique (Mehmetoglu, 2018). For analyzing the mediation (indirect) effect, three ways are suggested in the literature. First is to examine the completely standardized coefficient (table 6). Kenny (2016) suggested that a small effect size would be 0.01, a medium effect would be 0.09, and a large effect would be 0.25. Hence, as per the results in the table 6 below, all independent variables (career planning, succession planning, and career management policy) of the study have small effect on employees' job satisfaction. This is indicated by coefficient value of 0.024, 0.015, and 0.003 for career management policy, respectively. Second is to take the ratio of the indirect effect to the total effect (RIT). The RIT ration is interpreted as that a mediated effect explains percentage of the total effect of the independent variable (X) on a dependent variable (Y) or that percentage of the effect of the independent variable (X) on (Y) is mediated by the mediator variable (M) (MacKinnon, 2008). Thus, RIT of career planning is determined as 0.024 / 0.179 = 0.134 meaning that about 13.4 % of the effect of career planning on employees' job satisfaction is mediated by career development. In similar calculation, 25.4% and 0.4% of the effect of succession planning and career management policy on employees' job satisfaction is mediated by career development, respectively. Third, to measure the effect size of an indirect effect is, to take the ratio of the indirect effect to the direct effect (RID) and the result is interpret as the mediated effect is about RID value times as large as the direct effect (MacKinnon, 2008). The RID result shows that 0.2 times (0.023 / 0.151= 0.151) as large as the direct effect of career planning on employees' job satisfaction and succession planning indirect effect has 0.2 times (0.016 / 0.083 = 0.197) a large as the direct effect while career management policy has 0.0 times (0.003 / 0.807) = 0.003 as large as the direct effect on employees' job satisfaction.  As either STEP 1 or STEP 2 (or both) are not significant, there is no mediation! As either STEP 1 or STEP 2 (or both) are not significant, there is no mediation! As either STEP 1 or STEP 2 (or both) are not significant, there is no mediation! Source: Own computation (2020) Iacobucci et al. (2007 as cited by Mehmetoglu, 2018) modified the BK approach and proposed a series of steps for conducting mediation analysis via SEM. First step is to evaluate if either X -> M or M -> Y is not significant (or both are not significant) and hence, there is no mediation and the researcher should stop there. Medsem result under table 6 above shows that in case of career planning and succession planning X->M is significant while M->Y is insignificant and for career management policy both X->M and M->Y are insignificant. This result revealed that career development has no mediation in the relationship between career management practices (such as career planning, succession planning, and career policy) and employees' job satisfaction. This result is inconsistent with Bola (2011) who concluded that career management elements have effect on career development and career development, in turn, affect job satisfaction. In addition, as opposed to the finding of this study, the study by Ismail et al. (2013) shows that career development does act as an important mediating variable in the relationship between career program and job satisfaction.

Conclusion
The relationship between career management practices and job satisfaction has been attracting attention of HRM and particularly career development researchers and practitioners. The practice of career management has much implication on employee outcomes like job satisfaction in public organization; it has received little attention, particularly, in public institutions. Career management importance to enhance employees' job satisfaction increases when taking some intervening variable which in turn may affect their performance because of unsatisfied employees not contributing to the creation of public value. The findings of this study have the following theoretical and managerial implications for career development literature.

Theoretical Contribution
The theoretical contributions of this study, to career development literature and beyond, are stated in the following ways. First, this study believed to contribute existing literature by showing examining the effects of various elements of career management on employee job satisfaction in public institutions. Past studies (Rafique, 2014;Navern et al. 2015;Tobing, 2016;Abdul et al., 2016;Adamu, Daddie & Ebikeseye, 2017;Lydia & Nzulwa, 2018) were conducted in non-public organizations. That means they may not reflect the realities of public institutions due to contextual differences between various organizations. As noted by Susan & Anna (2018) cultural values could influence the perceptions of job satisfaction and its determinants, so public employees from different cultures may display different levels of job satisfaction, though the literature is open to doubt on this topic. Therefore, this study may contribute to understanding on the effect of career management practice on job satisfaction in public institutions. Second, the finding that career management elements such as career planning, succession planning and career policy has small effect on employee's job satisfaction and career development has no mediation effect may trigger a new insight in the field. This is in contrary to past studies (Bola, 2011 andIsmail et al., 2013) which found that there is significant link between the mentioned study variables. However, this study provided empirical evidence that this is not always true, particularly, in pubic organizations. Third, the current study may contribute to scientific community by applying advanced and improved methods like medsem for mediation analysis in business and management fields. Past studies used regression for mediation analysis. However, Iacobucci et al. (2007 as cited by Mehmetoglu, 2018) argued that using the regression (REG) technique suffers from a serious drawback by consistently producing larger standard errors for the path coefficients than does the Structural Equation Model (SEM) technique. Hence, Mehmetoglu (2018) suggested the structural equation framework using medsem Stata command for two benefits it provides in the mediation analysis. One is that SEM helps to make mediation estimation in optimal, more proper and complete way. Another is that SEM can facilitate mediational analysis using observed or latent variables as well as combinations of observed and latent variables. Therefore, this study used SEM technique using medsem Stata command for analysis of mediating role of career development in the relationship between career management elements and job satisfaction in public institutions found in Jimma City. Fourth, through running factorial analysis, this study will contribute to existing human resource literature in general and understanding of career management elements or constructs in particular. There is great deal of confusion in the literature on the building blocks of career management. For example, Bola (2011) and Ismail et al. (2013) synonymously used career planning and career management as distinct constructs their respective studies. Through adopting career management process model by Armstrong and Taylor (2014), the factorial analysis result of this study contributes that career management in public organization of Jimma City consists of elements like career management policy, career planning, succession planning, and career development. Thus, this study provided empirical evidence for Armstrong and Taylor's career management constructs in public institutions.

Practical Contribution
Besides the theoretical contribution, this study has the following managerial implication. First, human resource managers at public organizations should consider career management policy as part of career management process. The career management process model by Armstrong and Taylor (2014) outlines that it starts at formulating career management policy and ends with career development programmes. The career policy dictates the other activities or elements of career management such as career planning, succession planning, and career development. The second managerial implication is that managers of public organization should focus on the direct effects of career management elements such as career planning, succession planning and career management policy on job satisfaction of public employees. Unlike past studies, the current study findings revealed that career development has no mediation role in the relationship between career management elements and job satisfaction of public employees. This finding is so important to managers that help them to direct limited public resources which may be spent for improving job satisfaction of their employees in greater efficiency and effectiveness.

Study Limitation and Future Research Recommendation
This study has the following pitfalls. First, due to the similarity of overall HRM practices as well as career management in public institutions, the sample size of this study is fairly small(n=224) compared to very large number of public employees in Ethiopia. Although this study sample is larger than the samples used in related longitudinal studies (e.g., n= 140 employees, Ismail et al., 2013;n= 175 employees, Tobing, 2016; n= 178 teachers, Naway & Haris, 2017; n = 120 employees, Lydia & Nzulwa, 2018), we recommend future research to collect even larger samples to enhance the generalizability of findings. Second, this study followed the career management process model but was not able to include two elements: talent audit and counseling. Of course this is due to no practice found in public institutions found in Ethiopia. Hence, future studies may also include career management elements such as talent audit and counseling of public institutions found in other developing countries like Ethiopia. This helps to validate or critic the findings of the current study Third, the finding of this study represents public institutions but private institutions were not included. Because career management practices in private institutions may have significant contribution to the understanding by scientific community and practitioners. Thus, future studies should conduct comparative study between private and public organizations with respect to examining the effect of career management elements on job satisfaction.