Antecedents of Affective Commitment of Human Resource Management Practitioners Attending a Professional Body Convention

In this paper, affective events theory (AET) is used to develop a model that can be used by organizations to enhance propitious work conditions that will encourage human resource management (HRM) practitioners to be affectively committed to their organizations. AET states that there are certain antecedents (i.e. distributive justice, job-related well-being, and employee engagement) that positively correlate with job satisfaction. According to AET, positive emotions have a positive indirect correlation between antecedents and job satisfaction. AET states that job satisfaction positively relates to affective commitment. The research design was cross-sectional and correlational, and the sample size was (n=205).From the 300 questionnaires distributed to HRM practitioners, the response rate was 68.33%. The main finding is that respondents rated the positive emotions items below the mean, and there was a significant positive relationship between distributive justice and job satisfaction (β = 0.61; p≤ 0.001). The data showed that job satisfaction positively correlated with affective commitment (β = 0.70; p ≤ 0.001).Positive emotions only mediated the relationship between distributive justice and job satisfaction (β = 0.36; p ≤ 0.001). The study results have implications for managers’ efforts to keep HRM practitioners affectively committed, as the latter were not satisfied with their remuneration. The data created awareness that when a remuneration policy is drafted, it should take cognisance that HRM practitioners who have positive emotions will be affectively committed and stay longer in organizations.


Introduction
Unlike other employees, HRM practitioners are faced with the challenge of performing a dual role (Pereira & Fontinha, 2016). Given the dualistic nature of their role, HRM practitioners tend to focus on the commitment of other employees while neglecting their own. HRM practitioners face an ongoing challenge of improving the organization's productivity and striving to find ways of encouraging employees to be more committed and to enhance levels of engagement in their organizations (Agarwala, 2003;Chew & Chan, 2008;Tladinyane & Van der Merwe, 2016). According to Fareed, Isa and Noor (2016), the current labor market requires highly capable and committed HRM practitioners to sustain long-term competitive advantage. The commitment of HRM practitioners becomes critical for realizing organizational success (Gubbins & Garavan, 2016). It is thus of paramount importance for organizations to have practices that foster commitment amongst HRM practitioners. Research shows that employees who are committed, are motivated (Mohapatra & Sharma, 2008;Nohria, Groysberg & Lee, 2008;Sharma, Mohapatra & Rai, 2013), productive (Giffords, 2009), and stay in organizations (Abbott, White & Charles, 2005;Giffords, 2009;Liou, 1995;Stallworth, 2003) (also see Meyer & Allen, 1997, for a review of benefits). Organizational commitment is defined as a psychological attachment felt by the employees of an organization (Andolsek & Stebe, 2004;Meyer & Herscovitch, 2001;O'Reilly & Chatman, 1986;Mohapatra & Sharma, 2008). Allen and Meyer (1990) proposed a three-component conceptualisation of organizational commitment; however, for the purpose of this paper, the focus is on affective commitment, which means "employees wish to remain with the organization because of emotional attachment". Research studies during the last two decades have concentrated mostly on affective commitment (Sharma, Mohapatra & Rai, 2013).
Prior to this study, no similar study has been conducted in the South African context and this is the gap this study aims to address. In this paper, the primary objective was to develop a model for testing the relationship between constructs that have been identified in the literature as antecedents of affective commitment. The secondary objectives are as follows:  To explore if there is a relationship between distributive justice, job-related well-being, employee engagement, and job satisfaction.
higher employee satisfaction (Ouweneel, Le Blanc & Schaufeli, 2011;Schiopu, 2015), and higher levels of engagement (Lyubomirsky, King & Diener, 2005). Positive emotions enhance well-being and personal success. This, in turn, benefits organizations because employees appreciate their organization's efforts to improve their well-being, which results in greater commitment (Cabrera, 2017).It can be surmised that an increase in positive emotions lead to an increase in job satisfaction, employee engagement, distributive justice, and job-related well-being. Hence the following hypothesis is advanced: H4: Positive emotions mediate the relationship between distributive justice, job-related well-being, employee engagement, and job satisfaction Positive relationship between job satisfaction and affective commitment: Affective commitment is defined as employees' attachment to the organization (Azeem, 2010;Eslami & Gharakhari, 2012;Kumar & Eng, 2011). Employees who are attached to the organization tend to be fulfilled in their work (Kanchana & Panchanatham, 2015). Rhoades, Eisenberger and Armeli (2001) mentioned that employees are attached to organizations when there is a synchronization between their values and organizations' value systems (Bilgin & Demirer, 2012;Fu, Bolander & Jones, 2009;Karim, 2008;Kaplan, Ogut, Kaplan & Aksay, 2012;Lumley, Coetzee, Tladinyane & Ferreira, 2011;Tat, Pei-Nid & Rasli, 2012;Yucel, 2012). It is believed that satisfied employees will be committed to their jobs and remains in the organization, while dissatisfied employees will intend to quit (Amos, Acquah, Antwi & Azifome, 2015). These arguments lead tithe following hypothesis: H5: A positive relationship exists between job satisfaction and affective commitment.
Based on the discussion in the theoretical section, the following theoretical model is proposed: The authors of the present study included job-related well-being and employee engagement and did not include autonomy, supervisory support, and workload as antecedents of job satisfaction, Maleka, Skosana and Lekgothoane's (2016) study.

Methodology
In order to address the study's hypotheses, the research design is a survey, pre-experimental and correlational in nature. The latter assisted the researchers to test the relationship between the variables (Christensen, Johnson & Turner, 2015). Three hundred (N=300) questionnaires were distributed to HRM practitioners who attended the IPM convention in November 2016 at Emperors Palace in Kempton Park. The final sample consisted of (n=205) respondents, suggesting a response rate of 68.33%. The questionnaire comprised two sections, namely Section A: Biographical information (see Table 1) and Section B, which comprised validated scales. The job satisfaction scale was taken from Spector (1985), and for affective organizational commitment, four items were taken from Meyer and Allen (1997). For positive emotions, four items were taken from Watson, Clark and Tellegen (1998); four distributive justice items were taken from Price and Mueller (1986); and four items were taken from the employee engagement scale developed by Schaufeli and Bakker (2003). The scales had a seven-point Likert scale, where 1 = strongly disagree and 7 = strongly agree. The questionnaires stated the research purpose on the front page. Respondents were also informed on the front page that they must not include their names, that their participation is voluntary, and that they may withdraw from participating for any reason. This assisted in anonymising the respondents' identity. Prior to analysis, the data were coded in Microsoft Excel and exported into STATA version 13 for statistical analysis.

Data analysis
Discussed in this section are descriptive statistics (i.e. frequencies, means, and standard deviations) and inferential statistics (i.e. confirmatory factor analysis and structural equation modelling). Table 1are sample characteristics, with the majority (n=204) of respondents earning R10 000 and more a month, implying that they were middle and income workers.  Table 2 presents the mean scores and constructs. All the mean scores of job satisfaction were above 4, suggesting that the respondents were moderately satisfied. In terms of affective commitment, the mean scores were above 4, with the exception of the item "I would be very happy to spend the rest of my career with the organization", which had a mean score of 3.82. This can be interpreted as that the respondents moderately disagreed that they intended to spend the rest of their career in their current workplaces. Also, the mean scores of distributive justice were above 4, which can be interpreted as that the respondents perceived their rewards as being fair. Interestingly, the mean scores of positive emotions were below 4. This suggested that the respondents moderately disagreed that they were inspired by, enthusiastic about, and proud of their remuneration. The mean scores of employee well-being were above 4, suggesting that the respondents felt energetic, inspired, and excited at their workplaces. Lastly, all the mean scores of employee engagement were above 4, suggesting that the respondents were moderately engaged at their workplaces.  Reliability and validity assessment: As can be observed from Table 2, the Cronbach's alphas for each of the constructs were above 0.7, suggesting that the questionnaire was reliable (Field, 2013). The DG rho was greater than 0.7, suggesting uni-dimensionality (Ravand & Baghaei, 2016). In Table 2, the AVE scores were above 0.5 and most of the standardized loadings were >0.7, suggesting convergence validity (Ravand & Baghaei, 2016). Discriminant validity (i.e. not shown in the study) was achieved by determining which loadings of each construct (i.e. job satisfaction, affective commitment, distributive justice, job-related wellbeing, and employee engagement) "were higher than loadings on other constructs" (Ravand & Baghaei, 2016:4).

Hypotheses testing:
To develop the statistical model that shows the antecedents of affective commitment of HRM practitioners (i.e. refer to Figure 1), structural equation modelling (SEM) was conducted. Before testing the hypotheses developed in the theoretical section, the authors conducted model fit indexes (refer to Table  3). The model showed an acceptable fit based on the comparative fit index (CFI) = 0.92 (Hu & Bentler, 1999, recommend from 0.90 to0.95) and root mean square era (RMSEA) = 0.10 (Hu & Bentler, 1999, recommend greater than 0.05). The other goodness of fit indicators were poor, χ2 = 567.06 with 219 degrees of freedom (p<0.001), standard root mean square (SMMR) = 0.89 (Hu & Bentler, 1999, recommend less than or equal 0.08). It is argued that χ2 is usually affected by the sample size (Kline, 2011).
The data from Table 3 indicate a strong significant positive relationship between distributive justice and job satisfaction (β = 0.61; p≤ 0.001). Therefore, Hypothesis 1 is supported. Similarly, Saadati et al. (2016) found that distributive justice positively related with job satisfaction. The data from Table 3 show that there is an insignificant low positive relationship between job-related well-being and job satisfaction (β = 0.05; p ≥ 0.001). Therefore, Hypothesis 2 is not supported. Elsewhere, it was found that as job satisfaction increased, so did job-related well-being (Faragher et al., 2005) and the result was happy employees (Bakker & Oerlemans, 2011) who enjoyed their work (Rothausen, 2013) and who were productive (Bakker & Oerlemans, 2011). As can be observed from Table 3, there is an insignificant low negative relationship between employee engagement and job satisfaction (β =-0.02; p ≥ 0.001). Therefore, Hypothesis 3 is not supported. This finding contradicts previous research where it was found that employee engagement positively related with job satisfaction (Hanaysha, 2016). Table 4indicates that positive emotions only mediated the relationship between distributive justice and job satisfaction (β = 0.36; p ≥ 0.001).Based on this finding, Hypothesis 4 is partially supported. A closer look at Table 3 shows that there is a strong significant positive relationship between job satisfaction and affective commitment (β = 0.70; p≤ 0.001). Therefore, Hypothesis 5 is supported. All the means of positive emotions were below the mid-point of 4, suggesting that the respondents were not happy or satisfied with their wages. This finding is surprising because 183 out of 300 respondents earned R20 000.00 and above per month. It also emerged from the data that there was a strong significant positive relationship between job satisfaction and affective commitment (β = 0.70; p≤ 0.001). This finding is consistent with previous research (Bilgin & Demirer, 2012;Kaplan et al., 2012;Lumley et al., 2011).

Conclusion and Recommendations
Based on the study findings, it can be concluded that HRM practitioners rated measures used in this study positively, with the exception of the positive emotions items, which were remuneration related, and as a result organizations might find it difficult to retain HRM practitioners and motivate to perform their dual role effectively. The literature and data showed that AET is an appropriate theoretical framework to determine antecedents of affective commitment. It is thus recommended that organizations should investigate the reasons why HRM practitioners rated the affective commitment items negatively and why HRM respondents would not spend the rest of their careers with their organizations. HRM practitioners implement peoplerelated policy to enhance positive employee emotions but they are employees themselves, and positive work emotions also influence their work performance. Organizational policy should therefore also focus on developing HR talent (soft skills) in the organization and not solely focus on hard-core business skills. The researchers recommend that a similar longitudinal study should be conducted, using a mixed-methods research design. Future research can also determine significant differences in terms of age, gender, and educational levels.
Based on the study results, the SEM model as developed is shown below:

Figure 2: Antecedents of affective commitment
Acknowledgements: The authors of the paper wish to thank Prof. Stuart Carr, from Massey University, for helping to conceptualise the study. We are also grateful for IPM for its unwavering support.