Original Research

Ethical leadership and staff retention: The moderating role of job resources in Uganda’s healthcare sector

Thomas S. Mayende, Ibrahim A. Musenze

Received: 15 Feb. 2018; Accepted: 08 Aug. 2018; Published: 01 Nov. 2018

Copyright: © 2018. The Author(s). Licensee: AOSIS.
This is an Open Access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

Abstract

Orientation: Retention of quality staff in a highly competitive and dynamic working environment has made retention research necessary. Current focus is on how job resources influence ethical leadership in driving staff retention. Studies investigating the moderation effect of job resources in the ethical leadership–staff retention sequence in Uganda’s healthcare sector are scarce. Thus, the establishment of the moderation effect of job resources in the ethical leadership–staff retention sequence was needed.

Research purpose: This study aimed at examining the moderating role of job resources in the association between ethical leadership and staff retention.

Motivation for the study: Staff retention in Uganda’s healthcare sector is pervasive even with several government interventions such as salary enhancement. Rising maternal and infant mortality rates, low immunisation coverage, inter alia, are some of the effects. Reversing this scenario calls for leaders to exercise ethical leadership.

Research approach/design and method: This study utilised a cross-sectional research design. Analyses were conducted by SPSS v. 21 on a sample of 214 healthcare workers.

Main findings: The results show that ethical leadership positively affects staff retention. Also, the moderation role of the composite job resources variable in the ethical leadership–staff retention sequence was significant.

Practical/managerial implications: This study demonstrated the important contribution of both ethical leadership and job resources in staff retention management.

Contribution/value-add: This moderation model offers an additional complete explanation for the moderating effect of job resources in these conditions. The study also contributes to theory by demonstrating that contrary to the previous investigations where ethical leadership has been studied as an outcome variable, it is a predictor variable of staff retention.

Introduction

Staff retention is fundamental for organisational performance and profitability because it limits loss of valuable resources and human capital assets (Grissom, Nicholson-Crotty, & Keiser, 2012; Kim, 2012; Lambert, Cluse-Tolar, Pasupuleti, Prior & Allen, 2012). Employee performance and profitability are essential factors in organisational health; consequently, they allow organisational leaders to appreciate and comprehend factors that have significant potential in predicting staff retention and subsequently shaping organisational performance (Iverson & Zatzick, 2011). Several researchers have found associations between ethical leadership and staff retention (Elçi, Şener, Aksoy, & Alpkan, 2012; Elçi, Şener, & Alpkan, 2013; Pucetaite, 2014). Thus, attraction and retention of human capital is essential for an organisation’s success (Carnahan, Agarwal, & Campbell, 2012).

The consequences of failure to retain staff in organisations globally are clear and disastrous. For instance, it leads to increased organisation expenditure (Batt & Colvin, 2011). In the USA, it costs between 25% and 500% of a worker’s annual salary to replace one employee, with total expenditures as high as $1 million as a consequence of an employee’s decision to leave the organisation (Ballinger, Craig, Cross, & Gray, 2011; Von Hagel & Miller, 2011). Replacement costs, ranging from recruitment, employee screening and hiring activities, orientation and induction costs to training, offer a substantial depletion of organisational resources (Hancock, Allen, Bosco, & Pierce, 2013; Llorens & Stazyk, 2011; Pitts, Marvel, & Fernandez, 2011; Von Hagel & Miller, 2011). Accordingly, staff retention is a fundamental facet worth considering for individuals interested in improved organisational performance (Grissom et al., 2012; Kim, 2012; Lambert et al., 2012).

According to the World Health Organisation (WHO), out of 57 countries, 36 that are grappling with serious shortages of health workers are in sub-Saharan Africa, and more than 4 million additional doctors, nurses, midwives, managers and public health workers are required to plug this gap. The African region suffers from more than 24% of the global health burden, yet access to the world’s health workers stands at a paltry 3%. In contrast to the WHO’s recommendation of at least one doctor per 5000 people, 10 African countries average one doctor per 30 000 or more people. Moreover, these statistics do not reflect the reality in the face of the rural–urban divide, as doctors concentrate more in urban areas, leaving the rural areas even more underserved. Worsened by poor economic conditions, sub-Saharan African countries, including Uganda, are struggling to retain trained health workers. Workers first migrate from rural areas to urban areas, then out of the country. According to the Center for Global Development (2006), Uganda is 22nd on the list with 45% of its doctors working abroad.

In Uganda, staff retention in the healthcare sector is still a challenge and the situation persists despite a series of government interventions, such as salary enhancement for health workers and provision of hard-to-reach allowances (WHO, 2006). Further, according to Orach (2008) and Afzal, Cometto, Rosskam and Sheikh (2011), Uganda is one of the 57 countries in the world facing critical staff retention problems within the healthcare sector. The effect of this has been rising maternal and infant mortality rates and low immunisation coverage, amongst others (UBOS & ICF, 2017).

Human capital loss shrinks both organisational performance and profit generation (Ballinger et al., 2011; Llorens & Stazyk, 2011). This may be occasioned by disruptions of formal and collaborative networks, as well as the deterioration of employee expertise and client information (Ballinger et al., 2011; Llorens & Stazyk, 2011). Organisational leaders use research on the significance of ethical leadership to assist in the development of programmes that address the problem of declining staff retention (Pucetaite, 2014). Continuous research on staff retention is important to organisational managers, as it enables the leaders in the identification of relations between ethical leadership and employee retention.

Studies have established that ethical leadership predicts staff retention (Elçi et al., 2012; Elçi et al., 2013; Pucetaite, 2014). Appreciating what defines an ethical leader is one of countless standpoints in leadership investigations (Brown & Treviño, 2006; Yukl, 2010). It is also essential to examine the effects of the outcomes of ethical leadership (Brown & Treviño, 2006). Specifically, by probing these effects and consequences, evidence regarding how leaders may aptly direct and manage their human resources and organisations is provided (Walumbwa, Luthans, Avey, & Oke, 2011; Yukl, 2010). Most studies investigating ethical leadership tend to concentrate on its predictors (Decker & Rotondo, 2001; Deluga, 1991; Hassan, Mahsud, Yukl, & Prussia, 2013). Limited knowledge exists on outcomes of ethical leadership, with the exception of a recent meta-analytical work by Bedi, Alpaslan and Green (2016), or on the influence of situational variables such as job resources on outcomes. The effect of ethical leadership on staff retention may be moderated by job resources. Appreciation of ethical leadership influence will require recognition of factors critical to that end (Brown & Treviño, 2006). One potential situational and composite variable is job resources.

Job resources may positively moderate the ethical leadership–staff retention relationship. Moderating effects of job resources have mainly been examined in the relationship between job demands and work engagement (Hakanen, Bakker, & Demerouti, 2005). Studies show that job resources exercise a moderating effect on the relationship between stress and performance (Khan, Shah, Khan, & Gul, 2012), job demands and burnout (Bakker, Demerouti, & Euwema, 2005) and emotional dissonance on burnout (Karatepe, 2011). To our knowledge, moderation studies examining the role of job resources in an ethical leadership–staff retention relationship in Uganda’s healthcare sector are scarce. Accordingly, consideration of job resources as a moderator allows a more precise description of the relationship between ethical leadership as a predictor variable of staff retention. If a researcher fails to consider the possibility of a moderator effect in the data, a more exact explanation for an outcome (in this context, staff retention) may be missed (Bennett, 2000). Therefore, a theorised model regarding the ethical leadership–staff retention association in Uganda’s healthcare sector embraces the investigation of job resources in the position of a moderator variable. Specifically, the introduction of job resources as a moderator will buffer the effect of the ethical leadership on staff retention.

Research purpose and objectives

The general objective of this study was to examine the relationship between ethical leadership and employee retention. Furthermore, as depicted in Figure 1, the study assessed the moderation role of job resources in the ethical leadership–staff retention relationship.

FIGURE 1: Moderated model of the effect of ethical leadership on staff retention by job resources.

Literature review and theoretical development of hypotheses

The theoretical perspective that underpinned this study was based on the conservation of resource theory (COR). COR theory was developed by Hobfoll (1989) to integrate several stress theories into a general theory of psychological stress, based on the assumption that people strive to retain, protect and build resources and that the potential or actual loss of these valued resources is threatening. COR theory identifies three main categories relevant to individuals obtaining, retaining or protecting resources: instrumental (something that serves or acts as an instrument or means of gaining a resource, say money or shelter), social (resources based on attachments to families, to the workplace leaders such as an immediate supervisor and social groups, to mention but a few) and psychological (pertaining to the mind and emotions, like self-esteem or sense of autonomy) (Hobfoll, 1989, 2010). Early research applied COR theory to the design of social interventions to ‘inoculate’ against stress in the workplace and elsewhere (Hobfoll, 2010). Given that COR theory maintains that resource losses have a more acute impact than resource gains, it follows that protecting against resource losses, or minimising these, offers a potentially powerful intervention strategy in staff retention. Also, COR theory helps to explain the value of employee involvement, to the extent that such involvement allows employees to prioritise the specific resources to be conserved under adverse conditions, which improves retention.

According to COR theory (Hobfoll, 1989), job resources boost staff retention (Bakker, Schaufeli, Leiter, & Taris, 2008). We argue that ethical leaders offer job resources (for instance, role clarification or emotional support) and, therefore, ethical leadership links positively to staff retention. Furthermore, several scholars argue that the effectiveness of leadership varies according to the situation (Podsakoff, MacKenzie, & Bommer, 1996; Walumbwa et al. 2011; Yukl, 2010). Brown and Treviño (2006) argue that there is a need to identify contextual variables that explain when ethical leadership matters most and can enhance staff retention. Based on the above, this study appropriates the key concepts from COR theory, which are job resources and ethical leadership – in this case, ethical leadership as a predictor of staff retention.

Ethical leadership

Theorists also consider ethical leadership as ‘the demonstration of normatively appropriate conduct through personal actions and interpersonal relationships, and the promotion of such conduct to followers through two-way communication, reinforcement, and decision-making’ (Brown, Treviño, & Harrison, 2005, p. 120).

Several consequences are positively associated with ethical leadership. These include employee retention, organisational citizenship behaviour, organisational commitment and job satisfaction (Brown & Treviño, 2006; Walumbwa et al., 2011). Ethical leadership is also associated with staff retention (Elçi et al., 2012; Elçi et al., 2013; Pucetaite, 2014) and related factors such as employee commitment (Siegel, 2013) and supervisor support (Niemeyer & Covazotte, 2016). Kim and Brymer (2011) found that ethical leadership was positively related to job satisfaction. Studies show that job satisfaction improves staff retention (see, Shields & Ward, 2001). Therefore, given that ethical leadership is related to job satisfaction, yet job satisfaction predicts staff retention, it is likely that a relationship exists between ethical leadership and staff retention. In this circumstance, we hypothesise that:

H₁: Ethical leadership is associated with staff retention.

Moderation role of job resources on ethical leadership–staff retention relationship

Employees who believe that their organisation supports them with job resources may develop positive attitudes toward other facets of their employment and thus consider staying with the organisation (Blackmore & Kuntz, 2011). Shahkarami, Shafiee and Hamedani (2016) found that job resources were significantly and positively related to ethical leadership. Relatedly, Walumbwa et al. (2011) established that workers, who perceived the organisation’s support in delivery of job resources as a function of leader choice and ethics, appreciated the support more highly than if the support was secured through the organisation’s code of conduct.

Conversely, a relationship has been established between job resources and job satisfaction and subsequent staff retention (Hsiu-Jen, 2015). Building on this argument, Karatepe (2013) argues that lack of job resources culminates in job dissatisfaction, leading to general displeasure with the worker’s career and low staff retention. Perceived job resources also mitigate some negative organisational effects and increased job demands. Bakker, Demerouti, de Boer and Schaufeli (2003), Xanthopoulou, Bakker, Demerouti and Schaufeli (2009) and Hsiu-Jen (2015) established that aspects of job resources like a people-centred culture, social support, role clarity, job autonomy and subordinate involvement in decision-making lessened the physical and emotional demands of the job, resulting in increased staff retention. Studies show that many workers have a penchant for participative decision-making, a factor in job resources, and studies are in agreement with this argument (Wood & Fields, 2007). This is likely to explain the degree of this specific facet of job resources. An investigation of workers at Midwestern University in the USA recognised that job resources moderated the effects of job demands (Ilies, Dimotakis, & De Pater, 2010). Therefore, it is likely that job resources can buffer the negative effects of extreme physical, psychological and behavioural responses to stressors (Cullen, Silverstein, & Foley, 2008), which may increase staff retention.

Studies also support a moderating role for perceived organisational support (POS) – job resources. For instance, Simosi (2012) found POS that is a job resource had played a moderation role in the relationship between two sets of organisational supports (namely, perceived supervisor support and training transfer, and perceived colleague support and employee affective commitment) on organisational commitment. These findings sustain the utility of job resources as a possible moderator.

Moderating effects of job resources have mainly been examined in the relationship between job demands and work engagement (Hakanen et al., 2005). It has been established that job resources exert a moderating effect on the relationship between stress and performance (Khan et al., 2012), job demands and burnout (Bakker et al., 2005) and emotional dissonance and burnout (Karatepe, 2011). To our knowledge, there have been no studies investigating a moderation role of job resources in the ethical leadership–staff retention relationship.

Unethical leaders often display traits of dishonesty and exhibit lack of trustworthiness and are, therefore, unconcerned about others. Consequently, they are not approachable to employees with individual problems and concerns (Brown & Mitchell, 2010). Consequently, followers may ignore a leader whose behaviour is inconsistent with his or her ethical pronouncements (Yussen & Levy, 1975), as cited by Brown and Mitchell (2010). Studies show a positive correlation between unethical leadership and oppressive, abusive and manipulative tendencies, actions which are perceived as intentional and harmful to the employees and organisations (Tepper, Moss, Lockhart, & Carr, 2007). Similarly, studies reveal that unethical leadership negatively influences employees’ attitudes (Brown & Mitchell, 2010; Tepper, Henle, Lambert, Giacalone, & Duffy, 2008; Tepper et al., 2007), task and extra-role performance (Obicci, 2015), resistance behaviours (Brown & Mitchell, 2010), psychological well-being (Yukhymenko-Lescroart & Brown, 2015; Tepper et al., 2007) and personal lives (Yukhymenko-Lescroart & Brown, 2015). Moreover, unethical leadership positively influences deviant and unethical work behaviour amongst employees (Tepper et al., 2008).

In spite of the massive support established for the robust direct ethical leadership and job resource’s influence on staff retention, fairly sparse studies have been done on the moderating role of job resources on the ethical leadership–staff retention relationship. Contributions by Brown and Mitchell (2010), Tepper et al., (2008) and Tepper et al. (2007) suggest that job resources can play a buffering role in the prevention of unethical leadership practices such as dishonesty, leader distrustfulness and oppressiveness, abusiveness and manipulative tendencies in organisations. These unethical leader practices demoralise employees and are likely to lower staff retention. In the circumstances, job resources – for example, through social support, employee decision latitude and challenge, amongst others – are essential in buffering such low ethical leadership practices. It therefore follows that job resources might moderate the ethical leadership–staff retention relationship. Thus, it is expected that job resources could buffer the adverse impact of job demands on turnover intentions.

In the light of positive psychology we tested a new moderator hypothesis, that is, whether job resources moderate the association between ethical leadership and staff retention. On the one hand, job resources buffer the negative effect of low ethical leadership on staff retention, and on the other, job resources facilitate the mobilisation of ethical leadership, thus enhancing the experience of staff retention. This hypothesis fits well within the COR theory (Hobfoll, 1989). In line with this theory, different kinds of resources are likely to accumulate and thus lead to more positive outcomes. The possibility of moderating effects of job resources in the ethical leadership–staff retention relationship was previously recognised (for instance, Simosi, 2012) but not tested empirically. In light of the preceding theoretical settings, we formulate our final hypothesis:

H₂: Job resources moderate the relationship between ethical leadership and staff retention.

Research design

Research approach

A quantitative research design was used to test the two hypotheses. A survey method was used to collect quantitative data. Descriptive and inferential statistical methods were employed for data analysis.

Research methods

The study relied on a quantitative survey-based research method to examine the association between ethical leadership and staff retention, and using the moderation effect of job resources as a situational variable. The study used a cross-sectional study design to enlist data on the exogenous variables of ethical leadership and job resources and the endogenous variable of staff retention. The use of the cross-sectional survey design was found suitable for this study because it is one of the fastest methods of collecting survey data within a short period of time, for example, less than 2 years (Sarantakos, 2005; Sekaran, 2003; Zikmund, Babin, Carr, & Griffin, 2010).

Research participants

A sample of 214 respondents was drawn from a total population of 460 workers from Iganga and Bugiri District Referral Hospitals using guidelines established by Yamane (1967). In order to select the sample of 214 participants from the study population, we relied on a simple random sampling technique. This sample is consistent with the recommendations of Tabachnick and Fidell (2007), who propose that a sample above 200 is desirable for multivariate analysis. With a response rate of approximately 94%, the study met the recommended benchmark. The use of the simple random sampling technique allowed us to deal with the danger of study bias (Zikmund et al., 2010). The majority of respondents were male (67%). Of the respondents, approximately 30% were younger than 29 years, 40% were within the age range of 30–39 years, whilst 30% were aged between 40 and 49 years. Those with diploma qualification accounted for 36%, 50% had a bachelor’s degree and those with a master’s degree constituted 14%. In regards to working experience, those with less than 5 years of working experience accounted for 15%, 34% of the workers had served their respective organisations for a range of 6–10 years, 20% between 11 and 15 years, whilst 31% had working experience spanning 16 years.

The frequency counts for qualifications in rows and frequency counts for experience of the respondents before joining healthcare employment in columns reveal that 65% of the respondents had prior experience before joining the healthcare employment service. Of these respondents, 22% had diplomas, 38% had undergraduate degrees and 0.05% had postgraduate degrees. The results indicate a significant difference between groups regarding qualifications of the respondents and their work experience before joining the healthcare employment service (χ² = 35.23; df = 3; p = 0.01).

Measuring instruments

Job resources: This was evaluated using a scale by van Veldhoven, de Jonge, Broersen, Kompier and Meijman (2002). It is a five-dimensional factor scale comprising autonomy (five items; e.g. ‘Are you allowed to decide by yourself how to perform your work?’), social support from supervisor (five items; for instance, ‘If necessary, can you ask your superior for help?’), social support from colleagues (five items, such as, ‘Can you count on your colleagues when you encounter difficulties in your work?’), decision latitude (six items, such as, ‘Are you free to decide what work activities to enrol in?’) and challenge (five items, for instance; ‘My work is very interesting’). Items were anchored on a six-point Likert-like rating scale, ranging from 1 (never) to 5 (always). The index score for job resources was computed as the total of the five subfactor scores.

Ethical leadership: This variable was evaluated using De Hoogh and Den Hartog’s scale developed in 2008. Some modifications were done relevant to the study context. Specifically, previous facets like despotic leadership and optimism about the future were excluded. Their exclusion was on account that they were best suited for political surveys (Stevens, 2007). The adopted scale comprised three subscales: morality and fairness (six items; e.g., ‘My supervisor makes sure that his/her actions are always ethical’), role clarification (five items; for instance, ‘My supervisor explains who is responsible for what’) and power sharing (nine items; e.g., ‘My supervisor allows subordinates to have influence on critical decisions’). These items were anchored on a six-point Likert-like rating scale that ranged from 1 (disagree very much) to 6 (agree very much). The composite score for ethical leadership was computed as the total of the three subfactor scores.

Staff retention: This was evaluated using Döckel, Basson and Coetzee’s scale of 2006. It comprises six dimensions of: compensation (18 items, such as, ‘I am comfortable with my overall level of pay’), job characteristics (four items; for instance, ‘The job denies me any chance to use my personal judgment in carrying out the work’), supervisor support (six items; e.g., ‘I feel undervalued by my supervisor’), training and development (11 items; for instance, ‘I am satisfied with the training I have received so far’), career opportunities (11 items; e.g., ‘It would be easy to find a job in another department’) and work–life balance (nine items, such as, ‘My work schedule is often in conflict with my personal life’). The items were anchored on a six-point Likert-like rating scale, ranging from 1 (disagree very much) to 6 (agree very much). This scale has been widely used and produced robust results in retention studies. See the recent works of Potgieter, Coetzee and Ferreira (2018), Deas (2017) and Dyk, Coetzee and Takawira (2013). The global score for staff retention was computed as the total of the six subfactor scores.

Covariates: Past studies indicate that covariates such as age, gender and educational level might be relevant predictors of job resources and staff retention (Agyeman & Ponniah, 2014; Govaerts, Kyndt, Dochy, & Baert, 2011). Therefore, this study incorporated age (years), gender (1 = male, 2 = female) and educational level (1 = diploma, 2 = bachelor’s and 3 = master’s) as covariates.

Research procedure and ethical considerations

Human resource officers for the two district referral hospitals of Iganga and Bugiri in eastern Uganda were approached and requested to serve as contact persons in this study. Following the granting of permission from the respective medical superintendents, the study commenced. A letter explaining the objective of the survey and requesting participation in the survey was then sent to employees within these hospitals. Questionnaires were personally distributed by hand by the researchers to participants who indicated their willingness to participate in this study.

Statistical analysis

The statistical analysis was carried out using SPSS software, version 21. The first step involved descriptive statistics where the means, standard deviations, skewness and kurtosis were determined to describe the data. A cut-off point of 2.00 was set for skewness and 4.00 for kurtosis to ensure that the data were normally distributed. The reliability and validity of the scales were also determined by means of Cronbach’s alpha coefficients, as well as the content validity index. The reliability analyses for job resources, ethical leadership and staff retention were 0.82, 0.79 and 0.86, respectively. These statistics are consistent with the recommendations of Nunnally and Bernstein (1994), who propose 0.70 as an acceptable guideline for determining the internal consistency of an instrument. It is essential for a research scale to have an acceptable alpha coefficient, which is important for decision-making in the determination of whether the items in the study tool measure a similar construct. The second and last stage of analysis utilised inferential statistics. At the initial phase, the Pearson’s product-moment correlation coefficient was used to determine statistically significant relationships between ethical leadership, job resources and staff retention as detailed in Table 2. A moderated regression analysis (MRA) was performed to establish the degree to which the variance in staff retention can be attributed to ethical leadership and job resources.

Ethical consideration

The ethical considerations that were suitable for this study comprised the following: voluntary participation, informed consent and confidentiality. Voluntary participation refers to a situation where a respondent is not forced to participate in the study, whilst the potential respondents were also fully informed about the procedures and risks involved in this study and agreed to participate. Lastly, the respondents were assured that identifying information would not be made available to anyone who was not directly involved in the research (Saunders, Lewis, & Thornhill, 2009).

Results

Primary analysis was done to determine the degree of normality and multicollinearity of the data. Skewness and kurtosis was in the range of −0.677 and 2.727 for all study constructs as presented in Table 1. These statistics met the suggested benchmark of 2 or more and/or kurtosis of 7 or more, as recommended by Finney and DiStefano (2006) and West, Finch and Curran (1995).

Further, lack of symmetry (skewness) and pointiness (kurtosis) are two major ways in which a distribution can deviate from normal. The values for these parameters should be zero in a normal distribution. These values can be converted to a Z-score as follows, using the following formulae (Equations 1 and 2):

An absolute value of the score more than 1.96 or less than −1.96 is significant at p < 0.05, whilst more than 2.58 or less than −2.58 is significant at p < 0.01, and more than 3.29 or less than −3.29 is significant at p < 0.001. In small samples, values more or less than 1.96 are sufficient to establish normality of the data. Nonetheless, in large samples (200 or more) with small standard errors, this criterion should be changed to ±2.58, and in very large samples no criterion should be applied (i.e. significance tests of skewness and kurtosis should not be used) (Field, 2009; Ghasemi & Zahediasl, 2012). An analysis of the Z-scores for skewness and kurtosis reveal that scores of all the variables were in the range −1.230 to 5.509. With a sample of 214, the significance level is placed at p < 0.001, implying that the variables met the normality test (Field, 2009; Ghasemi & Zahediasl, 2012).

Multicollinearity of the data was assessed through the variance inflation factor (VIF) and the tolerance statistics. Menard (1995, as cited by Field, 2009) suggests that the VIF should be less than 10 and tolerance statistics above 0.2 in order to rule out multicollinearity problems. Results showed that all the VIFs for the study variables were less than 3.181 and the tolerance statistics were all above 0.314, implying tolerable levels of multicollinearity.

Table 2 presents mean scores, standard deviations and correlations between the study variables. Ethical leadership and job resources were moderately correlated (r = 0.481, p < 0.01). This means that as leaders become more ethical in their managerial operations, the allocation of job resources is likely to be conducted in a just and fair manner. Similarly, both ethical leadership and job resources correlated with staff retention (r = 0.265, p < 0.01 and r = 0.468, p < 0.01, respectively). This implies that changes in ethical leadership bring about corresponding changes in staff retention, just as job resources do. None of the correlations was > 0.50, confirming that the variables did not overlap to a considerable extent.

The statistical analysis applied in the identification of moderator variables in regression models is based on the recommendations of Arnold (1982) and Sharma, Durand and Gur-Arie (1981). Their method posits that in order to identify a moderator, and what type of moderator variable it is (in this case, job resources), one should construct a series of nested models.

The Moderated Regression Analysis (MRA) procedure began with Model 1, which included only the covariates. The covariate variables had no significant effect on the model and marginally explained 0.6% of the total variance in staff retention in organisations, as presented in Table 3 (R² = 0.006; p = 0.742). The ethical leadership variable was then entered into Model 2. The predictor variable of ethical leadership explained 2.7% of the total variance in staff retention (R² = 0.0271; p ˂ 0.05). Based on these results, H₁ (ethical leadership is associated with staff retention) is supported.

The interaction term for ethical leadership × job resources was entered into Model 3. As explained by Sharma et al. (1981), if the job resources variable is statistically significant, it can be classified as a predictor variable. As pointed out earlier, the interaction term for ethical leadership × job resources was entered into Model 3. The interaction was statistically significant (R² = 0.245; p < 0.01). This indicates that the impact of ethical leadership on staff retention is moderated by the job resources. On account of this, H₂ (job resources moderate the relationship between ethical leadership and staff retention) is supported. The drive for reliance on both Model 2 and Model 3 was to assess whether the job resources variable is a ‘pure’ moderator, with no direct association with the criterion variable (staff retention), or a ‘quasi’ moderator that also has a direct association with the criterion variable.

The interaction terms for the MRA were formed by multiplying the mean-centred ethical leadership variable by the mean-centred job resources variable. Mean centring minimises collinearity between the independent variables and the interaction term (Jaccard, Choi, & Turrisi, 1990). The main effect terms for ethical leadership and job resources were not based on mean-centred variables, as this is unnecessary and complicates their interpretation. The results of the MRA are shown in Table 3.

Discussion

Starting from the proposition that establishing how organisations retain intelligent and knowledgeable human capital is central to those organisations in achieving their goals, this study aimed at extending knowledge about the effect of ethical leadership on staff retention by introducing job resources to discover a moderation model. Specifically, the current investigation had two primary objectives. The first was to determine the predictive value of ethical leadership on staff retention. The second was to assess the ability of job resources to moderate the association between ethical leadership and staff retention. Results at individual level analysis demonstrate that ethical leadership predicts staff retention. This augments the earlier findings that staff retention is a function of morality and fairness, role clarification and power sharing (Elçi et al., 2012; Elçi et al., 2013; Pucetaite, 2014).

Ethical leadership is regularly used as an endogenous construct in organisational studies (Amundsen & Martinsen, 2014). Research usually examines constructs that influence the ethics of leaders such as employees’ psychological empowerment (Fathi, Javanak, Taher, & Shohoudi, 2014), organisational innovativeness (Pucetaite, 2014) and organisational cynicism (Mete, 2013). The importance of this study is the adoption of ethical leadership as a predictor of staff retention translating into support for a significant and positive association between ethical leadership and staff retention.

Findings of this study provided support for the moderation of job resources in the ethical leadership–staff retention relationship. Therefore, in this study we attempted to increase the knowledge of the moderating role of job resources (social support from supervisor, autonomy, social support from colleagues, challenge and decision-making latitude) in the staff retention process. The results confirm that, in general, job resources buffer and enhance the ethical leadership–staff retention relationship. These results are in tandem with an earlier study by Walumbwa et al. (2011), who found that employees who viewed the organisation’s support in provision of job resources as a matter of the leader’s will, or pleasure, and ethics, treasured the support to a greater degree, culminating in increased staff retention, rather than if the support were mandated by the organisation’s code of conduct. Studies also support a moderating role for job resources. For instance, Simosi (2012) found Perceived Organisational Support (POS), which is a job resource, had moderated the relationship between binary sets of organisational supports (i.e. perceived colleague support and employee affective commitment, and perceived supervisor support and training transfer) on organisational commitment.

These findings sustain the relevance of job resources as a possible moderator. Contributions by Brown and Mitchell (2010), Tepper et al. (2008) and Tepper et al. (2007) suggest that job resources can play a buffering role in the prevention of unethical leadership practices such as dishonesty, leader distrustfulness and oppressiveness, abusiveness and manipulative tendencies in organisations. These unethical leader practices demoralise employees and are likely to lower staff retention. In the circumstances, job resources – for example, through social support, employee decision latitude and challenge amongst others – are essential in buffering such low ethical leadership practices. It therefore follows that job resources could moderate the ethical leadership–staff retention relationship. Therefore, job resources buffer the adverse impact of job demands on turnover intentions.

The moderation role of job resources has largely been examined in the relationship between job demands and work engagement (Hakanen et al., 2005). It has been recognised that job resources exert a moderating effect on the relationship between stress and performance (Khan et al., 2012), job demands and burnout (Bakker et al., 2005) and emotional dissonance and burnout (Karatepe, 2011). To our knowledge, there are limited studies probing the moderation role of job resources in the ethical leadership–staff retention relationship within the healthcare sector, with the exception of recent work by Musenze and Mayende (2018). Within the prism of positive psychology, and consistent with previous investigations (Bakker et al., 2005; Hakanen et al., 2005; Karatepe, 2011), this study found support for job resources as a moderator in the ethical leadership–staff retention relationship. Job resources buffer the negative effect of low ethical leadership on staff retention. Also, job resources facilitate the mobilisation of ethical leadership, thus enhancing the experience of staff retention.

Research by Chiang and Birtch (2006) suggests generally that those demographic variables such as age, sex and marital status affect reward preferences, and in so doing also affect staff retention. In line with this, the analysis controlled for gender, but the results did not show any support for this contention. Lack of gender influence on staff retention offers support for the finding by Kajungu and Mugisha (2015), in a study of the role of remuneration in health workforce retention in a countryside district setting in Uganda, that gender had little influence on staff retention. Kajungu and Mugisha (2005) suggested that demanding parenting roles of young female staff could be one of the causes undermining staff retention.

Theoretical implications

Firstly, although previous researchers have shed some light on the exigency factors that might change the relationship between ethical leadership and staff retention to some extent (Elçi et al., 2012; Elçi et al., 2013; Pucetaite, 2014), the effect of the moderation model in our study offers a much more complete explanation for the moderating effect of job resources in these conditions. The effect of the moderation model can also provide evidence for the applicability in this area of the job resources interaction perspective. Specifically, having operationalised job resources as a conditional variable in our proposed model, we demonstrate that this variable is able to moderate the direct relationship between ethical leadership and staff retention. Therefore, the theorised model of the association between employee attitudes about ethical leadership and staff retention incorporates the investigation of job resources.

Secondly, the majority of studies examining ethical leadership concentrate on the predictors of this leadership issue (Decker & Rotondo, 2001; Deluga, 1991; Hassan et al., 2013). Limited knowledge exists on what outcomes result from ethical leadership, with the exception of the recent works of Bedi et al. (2016), and the influence of situational variables such as job resources on outcomes. The study contributes to the theory by demonstrating that contrary to previous studies where ethical leadership has been studied as an outcome variable, ethical leadership is a predictor variable of staff retention within the healthcare sector. Therefore, this study reinforces the limited literature of Bedi et al. (2016) on the outcome of ethical leadership.

Practical implications

This study showed that ethical leadership could be a feasible solution to improve staff retention in organisations, specifically within the healthcare sector. Organisations should therefore develop ethical leaders internally to sustain and improve staff retention. This could be done through ensuring that morality and fairness prevail within the organisations’ standard operating procedures. Specific ethical leadership development programmes can in addition be developed and implemented to equip leaders with ethical and moral values, qualities, capabilities, principles and practices.

Secondly, employees perceive that their supervisors are the agents of the organisation (Eisenberger, Huntington, Hutchison, & Sowa, 1986). Consequently, effective and continuous training programmes should also be arranged for supervisors so that they can learn how to provide assistance to junior employees in coping with the ethical leadership requirements vital for staff retention. This is crucial, because employees should receive signals from supervisors that the organisation values their contributions and cares about their well-being. Such an implication is important for the healthcare sector. That is, supervisors in authority should feel that they are obliged to provide moral support, justice and trust to their subordinates, and such employees may reciprocate through increased workplace retention.

Lastly, the findings suggest that staff retention may arise from the provision of job resources, such as social support from colleagues in the workplace, supervisory support, employee decision-making latitude and challenge. Intervention programmes could therefore focus on coaching employees on how to better manage job demands, as well as promoting the resources available at work in an innovative manner to address the issue of rising staff turnover.

Limitations and recommendations

One of the limitations of this study was related to its design. The study adopted a cross-sectional design, which might have confounded the causality between the studied variables. Prospective studies could undertake longitudinal studies to surmount this evident limitation. Secondly, all respondents were nurses, medical doctors or clinicians employed by two hospitals in Uganda. Prospective studies could collect data from respondents employed in other sectors in diverse settings to scrutinise the generalisability of these findings. Furthermore, in the model we proposed, the possibilities of other explanations cannot be ruled out, therefore, future researchers could explore the impact mechanisms of ethical leadership on staff retention by introducing other moderators, which will lead to a better understanding of why and how ethical leadership can reduce staff retention.

Conclusion

The general aim of this study was to examine the moderating effect of job resources on the ethical leadership–staff retention sequence in Uganda’s healthcare sector. Quantitative surveys were used to enlist data and several statistical methods were used to analyse the data. The results showed that job resources moderated the ethical leadership–staff retention relationship. The results sustain the view that ethical leadership can be relied on in the healthcare sector to improve staff retention. Ethical leaders provide the necessary job resources required by organisational employees to mitigate the effects of burnout, thereby increasing staff retention in organisations. This study contributes to the existing moderation literature by aligning job resources as a significant moderator in the ethical leadership–staff retention relationship in Uganda’s healthcare sector, an area having insufficient empirical evidence.

Acknowledgements

The authors declare that they do not have any financial or personal relationships that may have inappropriately influenced them in generating this article.

T.S.M. was the project leader and was responsible for experimental and project design. He also made conceptual and theoretical contributions, data collection, analysis and report (manuscript writing). I.A.M. made conceptual and theoretical, data collection, analysis and report writing contributions.

References