Taylor AJPA

What cause employees to whistle while they work? Factors affecting internal whistleblowing in the Australian Public Service To cite this paper: Taylor, Jeannette. What cause employees to whistle while they work? Factors affecting internal whistleblowing in the Australian Public Service. Australian Journal of Public Administration. First published online: Oct 10, 2019. DOI: 10.1111/1467-8500.12401 Abstract What cause employees in the Australian Public Service (APS) to internally blow the whistle on corruption in the workplace? This research examines the impact of the nature of corruption, organisational culture, and employees’ work attitudes and actions on internal whistleblowing in the APS. The respondents were found to internally blow the whistle for most types of corruption: fraud, conflict of interest, unlawful disclosure of government information, and perverting the course of justice. Their whistleblowing behaviour was, however, unaffected by observations of theft of official assets. They were also unlikely to report observations of cronyism and nepotism. Active observers of workplace bullying (those who have blown the whistle after witnessing someone else being bullied) were three times more likely to also internally report corruption than inactive observers of bullying. Introduction What cause employees in the Australian Public Service (APS) to internally blow the whistle on corruption in the workplace? What types of corruption are they more or less likely to internally report in the APS? What types of organisational culture encourage or discourage internal whistleblowing? And what work attitudes and behaviours are significantly associated with internal whistleblowing? For example, are whistleblowers in the APS disgruntled employees or are they satisfied, committed, and productive employees? It is the aim of this research to address these questions. In the process, this research addresses several gaps in the whistleblowing literature and corruption issues in the Australian public sector. First, although corruption is multifaceted, most whistleblowing research has adopted a unidimensional approach to examine it (Andersson 2017). There have been few studies conducted on whistleblowing by type of corruption, particularly in the public sector and in a non-US setting (Callier 2017; Near et al. 2004). In addition, while whistleblowing studies by type of corruption have included fraud and other forms of corruption, most have failed to include cronyism and nepotism, two examples of corruption related to favouritism (Graycar and Prenzler 2013). Yet, cronyism and nepotism are the most prevalent forms of corruption in the Australian federal and state public sectors (APSC 2018; CCC 2017; IBAC 2017; ICAC 2018). This research examines the impact of different types of corruption, from fraud to cronyism on internal whistleblowing in the APS, particularly comparing nepotism and cronyism with other forms of corruption. Second, several investigations of corruption cases in the Australian public sector have pointed to the organisation’s culture (e.g., CCC 2017; IBAC 2017). Public sector studies examining the relationship between organisational culture and whistleblowing are, however, rare (Lavena 2016; Rothwell and Baldwin 2006). The few studies of culture on whistleblowing have also produced inconclusive findings. It is timely to examine the effects of different types of organisational culture on internal whistleblowing in the APS. Third, despite earlier research indicating that whistleblowers are not disloyal or dissatisfied employees (Miceli and Near 1988), negative stereotypes of whistleblowers continue to persist in the workplace (Wortley et al. 2008). This issue highlights the value of more research across various contexts, such as in the APS. As we gain more understanding about the characteristics of whistleblowers through more empirical evidence across different contexts, we can finally put to rest any misperceptions about whistleblowers. Further, although research on whistleblowing has identified various personal and situational antecedents (Micelli and Near 2005; Trevino and Youngblood 1990; van der Wal et al. 2016), they have not examined the effects of work behaviours on whistleblowing behaviour. This research examines whether employees who report about the occurrence of one type of wrongdoing in the workplace (witnessing someone else being bullied) is more likely to report about the occurrence of another type of wrongdoing in the workplace (witnessing someone else engaging in corruption). This article starts by defining workplace corruption and whistleblowing. It also discusses the significance of internal whistleblowing as a control mechanism against workplace corruption. Workplace Corruption and the Significance of Internal Whistleblowing According to Gong et al. (2015), corruption is ‘any behavior that violates laws or ethical standards in the eye of the beholder’ (p.464). In its annual employee census in the APS, the Australian Public Service Commission (APSC) defines corruption as ‘the dishonest or biased exercise of a Commonwealth public official’s functions’, and clarifies that ‘it involves conduct that would usually justify serious penalties, such as termination of employment or criminal prosecution’. This study defines workplace corruption as a behaviour that violates laws, rules, norms, and/or morals in the workplace. There are various typologies of corruption (Lasthuizen et al. 2011; Søreide 2014). They include the following: bribery (offering, asking, and accepting bribes); fraud; theft of resources; nepotism (preferential treatment of family members), cronyism (preferential treatment of friends); acting (or failing to act) in the presence of conflict of (private or public) interest; misuse and manipulation of information (e.g., unlawful disclosure of government information); and blackmail, perverting the course of justice. The APSC incorporates these categories of corruption, along with another – colluding with or harbouring criminals. Workplace harassment and bullying do not fall under the APSC’s definition of corruption. In this study, public employees who engage in whistleblowing in their organisation are assumed to have witnessed an organisational member engaging in any of the above corrupt behaviour. They are presumed to disclose the information to an appropriate authoritative individual within their organisation with the intention that action should be taken to address it (Lewis et al. 2014). This is because internal whistleblowing is the main avenue for APS employees (APSC 2017). The benefits of internal whistleblowing are multiple. Corruption is difficult to detect and address through formal channels, particularly when public organisations are becoming increasingly complex and skilled at hiding corruption (Bartos 2013; Miceli and Near 2005). Brown et al. (2008, p.44) state that, ‘the unique position of employees within organisations gives them a strategic role as quality information sources’. Employees who engage in internal whistleblowing provide organisations with a unique opportunity to manage the corrupt situation before external actors are involved. According to Smith (2010: 717), ‘If they mishandle this opportunity, further investigations are unlikely to lead to better outcomes’. In addition to supporting an ethical climate in their organisation, public employees who internally blow the whistle on workplace corruption can minimise any potential damages to their organisation, such as its reputation, loss of public trust, and disruptions to service delivery (Miceli and Near 1985; PwC 2016; Victor and Cullen 1988). Determinants of Internal Whistleblowing Type of Corruption Most forms of corruption, such as fraud, bribery, theft of official assets, and unlawful disclosure of government information, are likely to be positively related to whistleblowing. Many employees who have directly observed acts of fraud and bribery generally know about the severity of the behaviours and how to label them, which in turn facilitates whistleblowing. This is because whistleblowing rises with the perceived severity and cost of a corrupt act (King 1997; Miceli and Near 2005; Near et al. 2013). The severity of a corruption case can also encourage employees to gather quality evidence because the onus of proof is often heightened with the severity of the case, such as to meet a legal standard (Near et al. 2004). Having access to strong evidence can increase whistleblowing (Miceli and Near 1985). Whistleblowing also occurs when a corrupt behaviour is perceived to be unambiguous and where the observer can attribute more blame to the perpetrator’s personal characteristics than contextual factors (Near et al. 2013; Robinson et al. 2012). For example, theft of official assets shows an individual’s intention for personal gain. Similarly, people who have witnessed fraud are more likely to blame the perpetrator than contextual factors (Harman 1999). The same can apply to unlawful disclosure of government information, particularly when the APS Values and Code of Conduct in Practice (APSC 2017) devote an entire section to discussing the gravity of this type and other types of corruption. As a result of these contextual differences, there are likely to be some variations in the magnitude of the relationships between various types of corruption and whistleblowing. The direction of such relationships are, however, likely to be positive, as explained above. It is expected that H1a: Most types of corruption are positively related to internal whistleblowing. Exceptions can however apply to nepotism and cronyism. Here, the relationships with whistleblowing can be negative. Compared to the above types of corruption, acts of cronyism and nepotism are often harder to pin down. Brown et al. (2008) had earlier stated that, ‘some types of perceived wrongdoing can be by definition be easier to identify and remedy, while others can be more subjective or “political” in nature’ (p.51). Nepotism and cronyism may fall under the latter. According to Stephen Bartos (2013), a former senior APS bureaucrat, nepotism and cronyism is one of the toughest forms of corruption to investigate and monitor: ‘Selection and promotion “on merit” is not objective, because it depends on an interpretation of merit. This is difficult territory’. Miceli and Near (1985) had stated that observers of a corrupt act were more likely to blow the whistle when ‘they had convincing evidence of the wrongdoing’ (p. 525). It would be challenging for observers of nepotism and cronyism to get easy access to incriminating evidence. Research published by the New South Wales’s (NSW) Independent Commission Against Corruption (ICAC 1994) shows that workplace corrupt behaviours that are too difficult to prove are often not reported by state employees. One of their study respondents said that, ‘A person in my position could not get any proof. ICAC insist on “documented proof”’. Similarly, the 2018 research conducted by Victoria’s (VIC) Independent Board-based Anti-corruption Commission (IBAC) indicates that the most common reason for not reporting corruption is ‘I’d need to have evidence to back up the allegation’. Bartos (2013) argues that ‘it is much more common for people to see what looks like corrupt behaviour but have no “smoking gun” direct evidence. Without evidence, they can feel they have nowhere to go’. Perceptions that these acts of favouritism are widespread, and committed by management can also deter whistleblowing. These issues have been raised by NSW’s ICAC (1994); comments collected from NSW state employees included the following: ‘Why bother reporting situations like this, the department runs on cronyism and jobs for the boys’; ‘this happens in our [organization] so it would be no use … and [the manager is the] worst offender’; and ‘It’s always a question – “jobs for the boys” and when it’s the executive and managers that are doing it, who wants to go to the police’. Queensland’s (QLD) Crime and Corruption Commission’s (CCC 2017) audit of 43 complaint files involving allegations of corrupt conduct in recruitment and selection activities in state agencies found that a majority of the perpetrators occupied managerial and higher positions; 37 percent were managers, 30 percent were directors, and 17 percent were senior executives. Employees may stay silent based on a belief that no effective action would be taken in response to their report (Near et al. 2004; Wortley et al. 2008). Research published by Australian state anti-corruption bodies have consistently noted employees’ unwillingness to report corruption when they lack confidence in: (1) their organisation’s openness to reports of corruption, (2) their organisation’s capacity to take appropriate action in response to the reports, and (3) their organisation’s capability to offer anonymity and protection from victimisation (e.g., IBAC 2017; ICAC 2018). They may fear that others would misinterpret their complaint as ‘just sour grapes on the part of an unsuccessful applicant’ (Bartos 2013). They may think that others would not view their judgment as legitimate (Near et al. 2013). Even when employees have access to evidence, they may be hesitant about blowing the whistle because of fears of retaliation, particularly when such acts of favouritism are often committed by their superiors (CCC 2017). Although retaliation is uncommon in the Australian public sector, whistleblowers who see themselves as being treated badly are twice more likely to blame management than coworkers (Brown and Olsen 2008). In their study, Wortley et al. (2008) reported that many of the respondents who indicated a reluctance to report workplace corruption because of fears of retaliation from the perpetrators were actually referring to perpetrators who were higher than them in the management chain. According to Bartos (2013), ‘If corruption involves senior managers, they can not only hide evidence but also persecute anyone who might reveal it’. It is thus hypothesised that H1b: Nepotism and cronyism are negatively related to internal whistleblowing. Type of Organisational Culture An organisation’s culture is believed to affect ethical behaviour, including whistleblowing behaviour (Barnett and Vaicys 2000; Lavena 2016; Near et al. 2013). Research has often differentiated between different types of organisational culture. Cameron and Quinn (2011), for example, distinguish between four types: a clan culture that focuses on the wellbeing of the individual employee, including cooperation, teamwork and group cohesion (support orientation); an adhocracy culture that focuses on creativity, experimentation, and openness to change (innovation orientation); a market culture that focuses on tasks, accomplishment, and efficiency (goal orientation); and a hierarchy culture that focuses on respect of authority, hierarchy, and rationality of procedures (rules orientation). Studies on the effects of a clan culture on whistleblowing are mixed. On one hand, Lavena (2016) finds that such a culture, or more accurately a respectful, cooperative, and fair work environment is negatively related to whistleblowing. Such organisational conditions may produce low levels of corruption, and consequently low levels of whistleblowing. Organisational members in such a culture may also view corruption as a group responsibility, inhibiting whistleblowing by individuals. Further, under such conditions, some organisations may be more tolerant of corruption and less tolerant of whistleblowers. Fear of retaliation of whistleblowing might intervene with employees’ perceptions of the work environment as respectful, cooperative, and fair to reduce whistleblowing. On the other hand, a culture that prioritises teamwork can predict whistleblowing intention (Rothwell and Baldwin 2006). Employees may be more likely to blow the whistle in an organisation that is viewed as being ‘permissive of employee voice in a decision-making process’ (Miethe and Rothschild 1994: 124). On this basis, it is hypothesised that H2a: A clan organisational culture is positively related to internal whistleblowing. Findings on the effects of a hierarchy culture on whistleblowing are inconclusive. Although Rothwell and Baldwin (2006) argue that whistleblowing intentions and actions are related to cultures which emphasise following organisational rules (rules climate), their study concludes that such cultures generally fail to predict whistleblowing consistently. Nevertheless, it is likely that a culture that focuses on rules and procedures would facilitate the observers of a corruption case to blow the whistle in accordance to their organisation’s rules and procedures on reporting wrongdoing in the workplace. At the same time, it is equally likely that such a culture can deter whistleblowing when the rules and procedures for reporting are perceived to be ambiguous and/or cumbersome. The fact that government bureaucracies are often characterised by red tape (Bozeman 1993) would suggest a negative impact on whistleblowing, particularly when employees do not trust their organisation’s capability to handle the whistleblowing process well (Taylor 2018; Wortley et al. 2008). H2b: A hierarchy organisational culture is negatively related to internal whistleblowing. There is a lack of empirical evidence for the other two types of culture (adhocracy and market) on whistleblowing. In the absence of such evidence, a null hypothesis is put forward. H2c: Adhocracy and market organisational culture are not related to internal whistleblowing. Employee Work Attitudes and Actions Rather than viewing it as a desirable behaviour, all too often whistleblowing is perceived as a deviant behaviour committed by disloyal employees (Lavena 2016). Whistleblowers are often labelled snitches or traitors (Akerstrom 1991). Whistleblowing is associated with more positive job responses (Brockmann 2017; Miceli and Near 1988). Whistleblowers are more satisfied than inactive observers (Near et al. 2013). Compared to inactive observers, whistleblowers are also more likely to have high levels of commitment and job performance (Brewer and Selden 1998; Near and Miceli 1996). Some studies have failed to find a significant relationship between whistleblowing and job satisfaction or organisational commitment (Mesmer-Magnus and Viswesvaran 2005; Sims and Keenan 1998). For example, Mesmer-Magnus and Viswesvaran’s (2005) meta-analytic study reported an average correlation of zero between whistleblowing intention and job satisfaction as well as that between whistleblowing intention and organisational commitment. Nevertheless, most studies have indicated that satisfied and committed employees who have witnessed a workplace corruption would blow the whistle because they care enough about their organisation to undertake the challenge (Near et al. 2013). On this basis, H3a: Desirable work attitudes (job satisfaction and organisational commitment) and behaviours (job performance) are positively related to internal whistleblowing. In addition to the above work attitudes, self-efficacy, which is ‘judgement of one’s capability to accomplish a certain level of performance’ (Bandura 1986: 391) has been reported to raise the propensity for whistleblowing (MacNab and Worthley 2008; Miceli and Near 1992). Employees with higher self-efficacy levels can feel more confident about their capabilities with identifying and reporting corruption, and persuading authorities to correct it (Miceli et al. 2012). Self-efficacy can be facilitated by an effective whistleblowing policy i.e. one that is easy to comprehend, and provides clear procedures and advice to employees on the types of corruption to report, appropriate reporting points for the different types, and the level of information required in a report (Roberts et al. 2011). Such a policy enable employees to determine how to identify corruption, how to report it, and the consequences of failing to report it. Rothwell and Baldwin (2006), however, did not find a correlation between such a policy and whistleblowing. They proposed that unless the policy mandates employees to report corruption, employees are unlikely to blow the whistle because of the high cost associated with such a behaviour. Since most studies have reported about the importance of self-efficacy for whistleblowing, it is expected that H3b: Self-efficacy in identifying and reporting workplace corruption is positively related to internal whistleblowing. Finally, like workplace corruption, workplace bullying/harassment represents an organisational wrongdoing. Employees who have witnessed someone else being bullied in the workplace are generally expected to assist the victim by internally reporting the incident. Lee et al’s (2004) application of a whistleblowing framework to examine sexual harassment found results that were similar to those reported by studies of whistleblowing about other types of wrongdoing, such as corruption. It implies that the antecedents of whistleblowing of different types of wrongdoing may share some similarities. Although the process of reporting bullying typically differs from that of reporting corruption in an organisation, both processes deal with standing up against wrongdoing in the workplace. Employees’ willingness to report wrongdoing is affected by their understanding of the reporting process and their level of comfort in reporting the wrongdoing to someone in their organisation (IBAC 2017). It suggests that past organisational experience matters. If a negative whistleblowing experience can reduce one’s trust in the organisation (Roberts et al. 2011), then a positive experience in a whistleblowing process can improve one’s trust in the organisation. When employees blow the whistle in response to having witnessed someone else being bullied, they gain experience from speaking up against wrongdoing in the workplace. This can affect their levels of confidence and self-efficacy to speak up against future situations involving the wrongdoing or other types of wrongdoing. It can also affect their trust in their organisation to address any wrongdoing. It implies that employees who have previously blown the whistle after witnessing someone else being bullied may do the same in response to witnessing a corrupt act, another type of wrongdoing, because of the self-confidence and trust in their organisation gained from the previous incident. Conversely, those who choose to stay silent in response to having witnessed someone else being bullied may also remain silent for workplace corruption. Employees who blow the whistle in response to observations of workplace bullying and workplace corruption may also be public-spirited individuals or have high levels of public service motivation (PSM) (Lavena 2016). Their concern for the common good and social justice imply that they are likely to report observations of wrongdoing, such as bullying and corruption. In short, employees who internally blow the whistle because they have witnessed someone else being bullied may do the same for corruption because of PSM and/or increased confidence, self-efficacy, and trust in organisational processes. H3c: Internally blowing the whistle on workplace bullying is positively related to internally blowing the whistle on workplace corruption. Methodology Data Source and Subjects This study uses the 2016 APS Employee Census, which is carried out across APS agencies with at least 100 employees. The sample is stratified by level, organisational size, organisation, and location. Only those respondents who reported having witnessed corruption in their organisation in the last 12 months are included in this study. The proportion is small – 3.6 percent (N = 3,495 out of 96,762). A majority of the respondents were females (57.6%), aged below 45 years (51.8%), did not have a university degree (51.3%), held a job position below middle management (73.2%), and were in permanent (94.8%), full-time (89.2%) employment. Most of them had worked in large (i.e. containing at least 1001 employees; 88.5%), and operational (75.3%) agencies. They had worked in their agency for less than 10 years (55.9%). Measures The measurement of internal whistleblowing is a two step-process. The respondents were initially asked to provide a yes/no response (represented by 1 and 0 respectively) to the question: ‘Excluding behaviour reported to you as part of your duties, in the last 12 months have you witnessed another APS employee in your agency engaging in behaviour that you consider may be serious enough to be viewed as corruption?’ Those who responded ‘yes’ to the question proceeded to another dichotomous question: ‘Did you report the potentially corrupt behaviour [in accordance to your agency’s policies and procedures]?’ This item is used as the dependent variable. The independent variables are listed below: Type of corruption. Ten types were investigated: nepotism, cronyism, acting or failing to act in the presence of an undisclosed conflict of interest, fraud or embezzlement, theft or misappropriation of official assets, unlawful disclosure of government information, perverting the course of justice, bribery, blackmail, and colluding, conspiring or harbouring criminals. Respondents who had witnessed them in their agency were asked to identify them. Observations were coded ‘1’, while non-observations were coded ‘0’. Type of organisational culture. Four types were examined. The categories are quite similar to Cameron and Quinn’s (2011). A hierarchy culture places an emphasis on procedures, a clan culture focuses on employees, particularly morale and team cohesion, an adhocracy culture prioritises innovation and development of new ideas, and a market culture emphasises productivity and goal achievement. Each category is measured with four items. The item wording for these measures and other multi-itemed measures are shown in the appendix (available online). The multi-itemed measures uses a 5-point Likert scale, with ‘1’ for ‘strongly disagree’, to ‘5’ for ‘strongly agree’. Cronbach’s alpha value (α) of the adhocracy culture, hierarchy culture, clan culture, and market culture measures are .88, .81, .86, and .82 respectively. Job performance. A single item measures employee perception of their overall job performance in the last fortnight: ‘In the last fortnight, please rate your overall job performance on a scale of 1 to 10, where 1 means your worst performance ever at your job, 5 means your usual performance at your job and 10 means the best you have ever worked in your job’. Satisfaction with organisational rewards and conditions. Five items measure employee satisfaction with monetary and non-monetary rewards/conditions in their organisation, such as pay and work-life balance. α = .78. Organisational commitment. Three items measure the respondents’ affective attachment to their organisation. α = .90. Self-efficacy. Three items measure employee knowledge of policies and procedures for managing ethical violations, and their confidence in identifying and reporting such violations in their organisation. α = .76. Reporting of workplace harassment or bullying. A single item asks the respondents whether they have reported their observation of someone else being subjected to bullying or harassment in their workplace in the last 12 months. Like the dependent variable, it has a dichotomous yes/no response (represented by 1 and 0 respectively). Data Analysis The multi-itemed variables, which rely on self-reported data from a single source (APS employees) raise the risk of common source bias (CSB) (Podsakoff et al. 2003). Results from two tests – Harman’s single-factor test, and the common latent factor technique – fall within the thresholds, suggesting that CSB may not be a major issue in this study. Concerns about CSB should also be balanced with the appropriateness of a self-report methodology for this study. This study taps into the respondents’ perceptions of their workplace, such as commitment to their organisation. Although employee perceptions of organisational situations may not always be accurate, they are real to the individuals holding these views, and can affect their behaviours (Lind 2001), which is the focus of this whistleblowing study. Favero and Bullock (2015) note that ‘the organizational context in which an individual is situated may be related to how he/she processes and reports information about his/her workplace’ (p.288). The Australian government had also taken measures to lessen evaluation apprehension in the census by assuring confidentiality of results, and anonymity of respondents, including not tracking individual responses over time. Further, a confirmatory factor analysis (CFA) provides evidence of construct validity of the measurement model containing the multi-itemed scales. The Comparative Fit Index is .91, Tucker-Lewis Index is .90, Standardized Root Mean Square Residual is .04, and Root Mean Square Error of Approximation is .06. The CFA fit statistics fall within the thresholds, suggesting an acceptable model fit (Fornell and Larcker 1981; Hu and Bentler 1999). Factor loadings of the items range from .58 to .92. The convergent and discriminant validity of the constructs are also demonstrated by the average variance extracted (AVE) and the construct reliability (CR) results for the constructs. AVE, which ranges from .52 to .72, is above the desirable threshold of .5, and CR, which ranges from .73 to .90, is above the .7 threshold (Fornell and Larcker 1981). Discriminant validity is further demonstrated with no correlation between any pair of the latent variables greater than .78 (between adhocracy culture and clan culture), as shown in table 1 in the next section. Findings Descriptive Statistics and Correlations Table 1 presents the descriptive statistics and correlations of the study variables. Internal whistleblowing appears to be quite low, with a mean value of 0.33. About 33 percent of respondents who had witnessed corruption in their organisation indicated that they had engaged in whistleblowing. <insert table 1 here> On average, cronyism is the most common corruption witnessed (M=.64, N= 2251). This is followed by nepotism (M=.25, N= 878), conflict of interest (M=.21, N= 727), and then fraud (M=.14, N=499). Number of observations of theft of official assets, unlawful disclosure of information, and perverting the course of justice are lower, at 230, 141, and 99 respectively. The least common forms are bribery (N= 90), blackmail (N=35), and conspiring with or harbouring criminals (N=39). These types of corruption are dropped from subsequent analyses; the small N and in turn lack of sufficient degrees of freedom suggest that meaningful analyses cannot be conducted. <insert figure 1 here> The frequency distribution of corruption by type is summarised in figure 1. Cronyism and nepotism were the most commonly witnessed corruption in 2015 and 2016. Table 1 also shows four prominent features. First, internal whistleblowing is positively related to most corruption (e.g., fraud and unlawful disclosure of information), but it is not related to some types of corruption (bribery, blackmail, conspiring with criminals). It is negatively related to nepotism and cronyism. Second, internal whistleblowing is positively associated with most types of culture (e.g., clan), but it is not related to hierarchy culture. Third, whistleblowing is associated with employee satisfaction (r=.13), commitment (r=.13), and performance (r=.06). Finally, whistleblowing is related to self-efficacy (r=.31), and reporting of workplace bullying (r=.43). Respondents who blew the whistle appeared to be satisfied with organisational rewards and conditions, were committed to their organisation, and believed that they were productive. They were confident about identifying and reporting workplace corruption. They had also reported observations of workplace bullying over the period surveyed. Comparative Trends by Agency Type Since the respondents work in different agencies, the findings on internal whistleblowing and corruption type are next analysed by type of organisation. The APS Employee Census provides four categories of agency: operational, specialist, regulatory, and policy. Table 2 presents the analysis of variance results. Only significant findings are reported. <insert table 2 here> Table 2 shows two notable findings. First, internal whistleblowing has occurred most frequently in operational agencies. Post-hoc comparisons using the Tukey HSD test indicate that the mean score for operational agencies is significantly different from that for specialist agencies and policy agencies, but not that for regulatory agencies. Second, most cases involving conflict of interest and cronyism have occurred in specialist agencies (cronyism was also most prevalent in policy agencies). A majority of cases involving nepotism are witnessed in operational agencies. Post-hoc tests indicate that the significant differences are largely confined between (1) regulatory and specialist agencies for matters relating to conflict of interest; (2) regulatory and operational, regulatory and specialist, and policy and operational agencies for nepotism; and (3) operational and specialist, and operational and policy agencies for cronyism. Factors Related to Whistleblowing Since internal whistleblowing is measured with a dichotomous variable, a hierarchical logistic regression is used. Model I focuses on contextual variables (type of corruption and organisational culture), and model II on individual variables (job performance, job satisfaction, organisational commitment, self-efficacy, and reporting of workplace harassment). The final model explains 42 percent of the change in whistleblowing within APS agencies. In addition to the regression coefficients and standard errors, the odds ratios are reported. An odds ratio greater than one suggests that the likelihood of blowing the whistle rises when the independent variable increases by one standard deviation (SD), and an odds ratio of less than one suggests that the likelihood of blowing the whistle drops when the independent variable increases by one SD. The results are displayed in table 3 and summarised below. <insert table 3 here> Type of corruption. Most types of corruption are positively related to whistleblowing: fraud, conflict of interest, unlawful disclosure of information, and perverting the course of justice. Using model I results as an example, 1 SD increase in a respondent’s observation of perceived fraud raises the odds of whistleblowing by 40 percent. The finding for theft of official assets is non-significant. H1a of a positive relationship between whistleblowing and most types of corruption is thus rejected. Cronyism, and to a lesser extent, nepotism are inversely associated with whistleblowing. One SD increase in a respondent’s observation of perceived cronyism decreases the odds of whistleblowing by 38 percent. The negative relationship between whistleblowing and these two types of corruption supports H1b. Type of organisational culture. Whistleblowing is positively related to clan culture in model I but not model II. In model I, one SD increase in a respondent’s perception of a clan culture raises the odds of being a whistleblower by 18 percent. Since a clan culture does not consistently predict whistleblowing, H2a of a positive relationship between clan culture and whistleblowing is rejected. Whistleblowing is negatively related to a hierarchy culture, which supports H2b. One SD increase in perception of a hierarchical culture reduces the odds of being a whistleblower by 16 percent. The remaining two types of organisational culture – adhocracy and market – produce non-significant findings, which support H2c. Individual characteristics. The results in model II show that the respondents’ perceived job performance, satisfaction, and commitment to their organisation are not significantly related to whistleblowing. The results reject H3a of a positive relationship between whistleblowing and these three factors. Only self-efficacy and reporting of workplace bullying show significant findings. A SD increase in a respondent’s self-efficacy raises the odds of blowing the whistle by 85 percent, which supports H3b. Similarly, a SD rise in the respondent’s reporting of workplace bullying increases the odds of their reporting of workplace corruption by almost three-fold, which supports H3c. Discussion This research examines the impact of the nature of corruption, organisational culture, and employees’ work attitudes and actions on internal whistleblowing in the APS. It supports previous studies that whistleblowing varies by type of corruption (Caillier 2016; King 1997; Near et al. 2004). Most types of corruption are found to be positively related to internal whistleblowing. APS respondents who had observed situations involving fraud, conflict of interest, unlawful disclosure of government information, and perverting the course of justice had generally blown the whistle in their organisation. There are exceptions to the above findings. First, the respondents’ whistleblowing behaviour were not affected by observations of theft of official assets. The findings in NSW’s ICAC (1994) report may provide some answers to the non-significant relationship between theft and whistleblowing. The report indicated that most workplace thefts were not reported because employees had failed to view the thefts as severe. Some employees also differentiated corruption from theft; they saw theft as a dishonest behaviour rather than a corrupt behaviour. Further, some employees felt justified in engaging in small scale theft (e.g., taking pens home) based on their contribution of other material goods to their organisation (e.g., paying for their own fares while out on business). Employees may thus not report theft of official assets because of a host of reasons. Lack of details about the nature of thefts reported in the APS dataset warrant further studies to be conducted for improved understanding. Second, nepotism and cronyism are negatively related to whistleblowing. While the respondents were willing to blow the whistle for most types of corruption, they were reluctant to do the same for nepotism and particularly, cronyism. They may perceive these forms of corruption as being more subjective and harder to identify (Bartos 2013; Brown et al. 2008). They may find it hard to get access to evidence (IBAC 2018; ICAC 1994). They may have low levels of trust in the capability of their organisation and senior management to handle the whistleblowing process and come up with an effective remedy (Taylor 2018; Wortley et al. 2008). They may be fearful of reprisals, especially when acts of favouritism are often committed by more senior organisational members (Brown and Olsen 2008; CCC 2017). These reasons may explain why nepotism and cronyism are the most commonly witnessed corruption (figure 1). Just as low whistleblowing levels for cronyism and nepotism can contribute to high levels of these types of corruption (Somers and Casal 2011), high levels of cronyism and nepotism can dampen whistleblowing behaviour, for example, through the bystander effect: ‘The fact that everyone knows about a particular type of misconduct may actually make people less likely to report it’ (ICAC 2018: 29). This two-way interaction can create a perpetual cycle, each reinforcing one another, leading to high levels of cronyism and nepotism, and low whistleblowing levels. The finding that whistleblowing varies with type of corruption would suggest that public managers need to develop a deeper understanding of how different types of corruption are perceived by employees, and why they are (and are not) reported. We cannot assume that public employees will report all types of corruption just because they have reported some types of corruption. Two obvious examples are cronyism and nepotism. Most respondents admit a general reluctance to speak out against these types of corruption. Cronyism and nepotism may be more tolerated than other types of corruption in many APS agencies. In addition to the possible reasons discussed earlier, there may be implicit cues in many agencies’ culture about what is (and is not) considered ‘real’ or ‘serious’ corruption. Accordingly, a clear and comprehensive code of conduct that is supported with clear, fair, and transparent remedial processes when corruption is reported can facilitate whistleblowing by helping to clear up any mixed messages (Somers and Casasl 2011). Whistleblowing is found to be less prevalent in a hierarchy organisational culture. Respondents who perceive that rules and procedures govern what people do in their organisation are less likely to blow the whistle. This finding contradicts that in Borry’s (2017) study. Here, the respondents follow organisational rules when their organisation’s culture encourages them to follow procedures and rules in order to resolve a wrongdoing. So why did so few respondents in this study blow the whistle in such a culture? For one, whistleblowers are often more concern about the outcomes of their reports than getting the organisational processes right (Smith 2010), which are at odds with the characteristics of a hierarchy culture. There may also be some issues in the formal whistleblowing procedures, rules and/or policies in these organisations. Brown et al. (2011) state that the legality and adversarial nature of the formal reporting process in the Australian public sectors can intimidate employees. They noted separate sets of procedures for related matters in many organisations, such as a set for the reporting process, and another set dealing with the investigation process. They also identified various issues with the process, such as a lack of feedback during the reporting process, and inadequate confidentiality and support provided to people who report corruption. Active observers of workplace bullying (those who have blown the whistle after witnessing someone else being bullied) were three times more likely to also internally report corruption than inactive observers of bullying. They may have high levels of PSM (Lavena 2016), and/or increased confidence, self-efficacy, and trust in organisational processes as a result of reporting the former wrongdoing. Although no two whistleblowing situations are identical, individuals who report wrongdoing often have knowledge about appropriate internal reporting channels, and a feeling of responsibility for reporting the wrongdoing (Near et al. 2013). The results of this research should be viewed in light of three limitations. First, the study sample size was restricted given that the data were segmented into fine groupings, such as by type of corruption. Although the sample size for a few types of corruption was relatively large (e.g., N=878 for nepotism), it was small for many types of corruption, such as perverting the course of justice (N=99). The extent to which the findings can be generalised to other samples and employees in the Australia public sectors might thus be limited so the results must be interpreted with care. Second, the concepts in this study are measured from a large government dataset that is not designed to provide a comprehensive measure of these constructs; they suffer from limitations in terms of breadth and depth of the measures. The measures also included single-itemed measures (e.g., job performance and reporting of corruption), and covered limited timeframes (e.g., the last fortnight for job performance). The types of corruption examined were also quite restricted because they were limited to those listed in the APS Census. Some corruption types were dropped because of too small sample sizes. Future studies should use established multi-itemed measures of the concepts. Observations or actions measured should cover reasonable and comparable timeframes. An examination of corruption type should ideally cover a wide rather than a limited range. Further, although it is the aim of this study to examine APS respondents’ perceptions of workplace corruption, future studies could more closely corroborate the respondents’ self-reports with objective evidence, say at the organisational type level, or organisational level. Third, this study has relied on cross-sectional data. In an effort to protect employee anonymity, the APS Employee Census does not track individual responses over time. With cross-sectional data, this research examines the strength and direction of the relationships between the study variables. Future studies with a longitudinal design can test for causality. A diary study methodology can also be used to track employee reactions over a defined time period of change. This research supports past studies that whistleblowing varies by type of corruption. Future whistleblowing studies should consider type of corruption as a potential moderator (Somers and Casal 2011). In addition, the relatively high incidence of cronyism and nepotism in the APS may be due to specific practices. In the 2018 survey of its members, the Community and Public Sector Union had discussed concerns about the erosion of the APS merit-based employment principles through initiatives such as the creation of talent pools. Individuals were often nominated by a senior organisational member to join a talent pool, making merit selection a formality and locking out others from the opportunity. In addition to nepotism and cronyism, problems associated with talent pools include unconscious bias and secrecy (CPSU 2018). VIC’s IBAC (2018) had also discussed about the practice of ‘double-dipping’, in which state employees set up companies in order to use recruitment agencies to hire relatives and friends back into public agencies at an inflated fee. There may be other practices that have allowed these types of corruption to thrive unchecked. Future studies can identify organisational and individual practices that enable these and other types of corruption to occur, including those that hinder their reporting. Finally, it is unclear if the outcomes of whistleblowing vary by type of corruption. For example, are there some types of corruption, such as cronyism, that are more susceptible to retaliation than others? Near et al. (2004) found that type of corruption affected the comprehensiveness of retaliation. And what are the other outcomes of whistleblowing for different types of corruption? Future studies can examine and compare the positive and negative outcomes of whistleblowing by type of corruption. Conclusion Whistleblowing is a key resource for managing corruption in the public sector (Brown et al. 2008; Treviño and Victor 1992). Yet, this study found that many APS respondents who had observed cronyism and nepotism were unlikely to blow the whistle. Cronyism and nepotism have been the most commonly witnessed corrupt behaviours in the APS over the past few years (see Figure 1 and APSC 2018). Clearly something needs to be done about these trends, especially when public trust in the government continues to decline (APSC 2018). Too often, cronyism and nepotism are overlooked in public sector corruption studies in favour of high profile corruption cases, such as those involving fraud. Yet, employees who have witnessed cronyism and nepotism in the public service often feel cheated, resentful, demoralised, and in turn have no/little motivation to work hard (Graycar and Prenzler 2013). This can ultimately undermine public sector productivity and good governance (Feeney and Kingsley 2008; Khatri and Tsang 2003). General citizens who learn about cronyism and nepotism in the public service through the mass media and other modes of public communication often lose trust in the government’s capability to govern. Both public administration academics and practitioners should therefore pay as much attention to addressing cronyism and nepotism in the public service as other types of corruption because of their impact on governance.  References Åkerstom, M. 1991. Betrayal and Betrayers: The Sociology of Treachery. New Brunswick, NJ: Transaction. Andersson, S. 2017. ‘Beyond Unidimensional Measurement of Corruption’. Public Integrity 19(1):58–76. Australian Public Service Commission (APSC). 2017. APS Values and Code of Conduct in Practice. Canberra: APSC. Australian Public Service Commission (APSC). 2018. State of the Service Report: State of the Service Series 2017–18. Canberra: APSC. Bandura, A. 1986. Social Foundations of Thought and Action: A Social Cognitive Theory. Englewood Cliffs, NJ: Prentice-Hall. Barnett, T. and C. Vaicys. 2000. ‘The Moderating Effect of Individuals’ Perceptions of Ethical Work Climate on Ethical Judgments and Behavioral Intentions’. Journal of Business Ethics 27(4):351–362. Bartos, S. 2013. ‘Our Costly Complacency on Corruption’. Sydney Morning Herald, Mar 5. Borry, E. L. 2017. ‘Ethical Climate and Rule Bending: How Organizational Norms Contribute to Unintended Rule Consequences’. Public Administration 95(1):78–96. Bozeman, B. 1993. ‘A Theory of Government “Red Tape”’. Journal of Public Administration Research and Theory 3(3): 273-303. Brewer, G. A. and S. C. Selden. 1998. ‘Whistle Blowers in the Federal Civil Service: New Evidence of the Public Service Ethic’. Journal of Public Administration Research and Theory 8(3):413–439. Brockmann, J. 2017. ‘Unbureaucratic Behavior among Street-Level Bureaucrats: The Case of the German State Police’. Review of Public Personnel Administration 37(4):430–451. Brown, A. J., E. Mazurski, and J. Olsen. 2008. ‘The Incidence and Significance of Whistleblowing.’ In Whistleblowing in the Australian Public Sector: Enhancing the Theory and Practice of Internal Witness Management in Public Sector Organisations, ed. A. J. Brown. Canberra: ANU E Press, 25–52. Brown, A.J. and J. Olsen. 2008. ‘Whistleblower Mistreatment: Identifying the Risks’. In Whistleblowing in the Australian Public Sector: Enhancing the Theory and Practice of Internal Witness Management in Public Sector Organisations, ed. A.J. Brown, Canberra: ANU E Press, 137–164. Caillier, J. G. 2017. ‘Public Service Motivation and Decisions to Report Wrongdoing in U.S. Federal Agencies: Is This Relationship Mediated by the Seriousness of the Wrongdoing’. American Review of Public Administration 47(7):810–825. Cameron, K. S. and R. E. Quinn. 2011. Diagnosing and Changing Organizational Culture: Based on the Competing Values Framework. Cheshire: John Wiley and Sons Ltd. Community and Public Sector Union (CPSU). 2018. Community and Public Sector Union Principal Submission: Independent Review of the APS. Available at https://www.cpsu.org.au/system/files/cpsu_submission_to_review_of_aps.pdf Crime and Corruption Commission (CCC) Queensland. 2017. Queensland Public Sector Responses to Corrupt Conduct Incidents in Recruitment and Selection Activities: Summary Audit Report. Brisbane: CCC. Favero, N. and J. B. Bullock. 2015. ‘How (Not) to Solve the Problem: An Evaluation of Scholarly Responses to Common Source Bias’. Journal of Public Administration Research and Theory 25(1):285–308. Feeney, M. K. and G. Kingsley. 2008. ‘The Rebirth of Patronage: Have We Come Full Circle?’ Public Integrity 10(2):165–176. Fornell, C. and D. F. Larcker. 1981. ‘Evaluating Structural Equation Models with Unobservable Variables and Measurement Error’. Journal of Marketing Research 18(1):39–50. Gong, T., S. Wang and J. Ren. 2015. ‘Corruption in the Eye of the Beholder: Survey Evidence from Mainland China and Hong Kong’. International Public Management Journal 18(3): 458–482. Graycar, A. and T. Prenzler. 2013. Understanding and Preventing Corruption. New York: Palgrave Macmillan. Harman, G. 1999. ‘Moral Philosophy meets Social Psychology: Virtue Ethics and the Fundamental Attribution Error’. Proceedings of the Aristotelian Society 99: 315–331. Hu, L. and P. M. Bentler. 1999. ‘Cutoff Criteria for Fit Indexes in Covariance Structure Analysis: Conventional Criteria versus New Alternatives’. Structural Equation Modeling: A Multidisciplinary Journal 6(1):1–55. Independent Broad-Based Anti-Corruption Commission (IBAC). 2017. Perceptions of Corruption: Survey of Victorian State Government Employees. Melbourne: IBAC. Independent Commission Against Corruption (ICAC) New South Wales. 1994. Unravelling Corruption: A Public Sector Perspective. Survey of NSW Public Sector Employees’ Understanding of Corruption and their Willingness to Take Action. Sydney: ICAC. Independent Commission Against Corruption (ICAC) New South Wales. 2018. Corruption and Integrity in the NSW Public Sector: An Assessment of Current Trends and Events. Sydney: ICAC. Independent Commissioner Against Corruption (ICAC) South Australia and Office for Public Integrity (OPI). 2018. ICAC Public Integrity Survey 2018 South Australia. Adelaide: ICAC. King, G., III. 1997. ‘The Effects of Interpersonal Closeness and Issue Seriousness on Blowing the Whistle’. Journal of Business Communication 34(4):419–436. Lavena, C. F. 2016. ‘Whistle-Blowing: Individual and Organizational Determinants of the Decision to Report Wrongdoing in the Federal Government’. American Review of Public Administration 46(1):113–136. Lasthuizen, K., L. Huberts, and L. Heres. 2011. ‘How to Measure Integrity Violations’. Public Management Review 13(3): 383-408. Lee, J.-Y., S.G. Heilmann, and J.P. Near. 2004. ‘Blowing the Whistle on Sexual Harassment: Test of a Model of Predictors and Outcomes’. Human Relations 57(3): 297–322. Lewis, D., A.J. Brown, and R. Moberly. 2014. ‘Whistleblowing, Its Importance and the State of the Research’. In International Handbook on Whistleblowing Research. Cheltenham: Edward Elgar, 1-33. Lind, E. A. 2001. ‘Fairness Heuristic Theory: Justice Judgments as Pivotal Cognitions in Organizational Relations.’ In Advances in Organizational Justice, eds J. S. Greenberg and R. Cropanzano. Stanford, Stanford University Press, 56–88. MacNab, B. R. and R. Worthley. 2008. ‘Self-Efficacy as an Intrapersonal Predictor for Internal Whistleblowing: A US and Canada Examination’. Journal of Business Ethics 79(4):407–421. Mesmer-Magnus, J. R. and C. Viswesvaran. 2005. ‘Whistleblowing in Organizations: An Examination of Correlates of Whistleblowing Intentions, Actions, and Retaliation’. Journal of Business Ethics 62(3):277–297. Miceli, M.P. and J.P. Near. 1985. ‘Characteristics of Organizational Climate and Perceived Wrongdoing Associated With Whistle-Blowing Decisions’. Personnel Psychology 38(3): 525–544. Miceli, M.P. and J.P. Near. 1988. ‘Individual and Situational Correlates of Whistle-blowing’. Personnel Psychology 41(2): 267-281. Miceli, M. P. and J. P. Near. 1992. Blowing the Whistle: The Organizational and Legal Implications for Companies and Employees. New York: Lexington Press. Miceli, M.P. and J.P. Near. 2005. ‘Standing Up or Standing By: What Predicts Blowing the Whistle on Organizational Wrongdoing?’ In Research in Personnel and Human Resources Management, ed. J. Martocchio, Greenwich, CT, JAI/Elsevier, Vol. 24, 95-136. Miceli, M. P., J. P. Near, M. T. Rehg, and J. R. Van Scotter. 2012. ‘Predicting Employee Reactions to Perceived Organizational Wrongdoing: Demoralization, Justice, Proactive Personality, and Whistle-Blowing’. Human Relations 65(8):923–954. Miethe, T. D. and J. Rothschild. 1994. ‘Whistleblowing and the Control of Organizational Misconduct’. Sociological Inquiry 64(3):322–347. Near, J. P. and M. P. Miceli. 1996. ‘Whistle-Blowing: Myth and Reality’. Journal of Management 22(3):507–526. Near, J. P., M. P. Miceli, and T. M. Dworkin. 2013. Whistle-Blowing in Organizations. New York: Taylor and Francis. Near, J. P., M. T. Rehg, J. R. Van Scotter, and M. P. Miceli. 2004. ‘Does Type of Wrongdoing Affect the Whistle-Blowing Process?’ Business Ethics Quarterly 14(2):219–242. Podsakoff, P. M., S. B. MacKenzie, J. Y. Lee, and N. P. Podsakoff. 2003. ‘Common Method Biases in Behavioral Research: A Critical Review of the Literature and Recommended Remedies’. Journal of Applied Psychology 88(5):879–903. PricewaterhouseCoopers (PwC). 2016. Fighting Fraud in the Public Sector IV. Australia: PwC. Roberts, P., A.J. Brown, and J. Olsen. 2011. Whistling While They Work: A Good-Practice Guide for Managing Internal Reporting of Wrongdoing in Public Sector Organisations. Canberra: ANU E Press. Robinson, S.N., J.C. Robertson, and M.B. Curtis. 2012. ‘The Effects of Contextual and Wrongdoing Attributes on Organizational Employees’ Whistleblowing Intentions Following Fraud’. Journal of Business Ethics 106(2): 213–227. Rothwell, G. R. and J. N. Baldwin. 2006. ‘Ethical Climates and Contextual Predictors of Whistle-Blowing’. Review of Public Personnel Administration 26(3):216–244. Sims, R. L. and J. P. Keenan. 1998. ‘Predictors of External Whistleblowing: Organizational and Intrapersonal Variables’. Journal of Business Ethics 17(4):411–421. Smith, R. 2010. ‘The Role of Whistle-Blowing in Governing Well: Evidence From the Australian Public Sector’. American Review of Public Administration 40(6):704–721. Somers, M. and J. C. Casal. 2011. ‘Type of Wrongdoing and Whistle-Blowing: Further Evidence that Type of Wrongdoing Affects the Whistle-Blowing Process’. Public Personnel Management 40(2):151–163. Søreide, T. 2014. Drivers of Corruption: A Brief Review. Washington DC: The World Bank. Taylor, J. 2018. ‘Internal Whistleblowing in the Public Service: A Matter of Trust’. Public Administration Review 78(5): 717-726. Trevino, L. K. and S. A. Youngblood. 1990. ‘Bad Apples in Bad Barrels: A Causal Analysis of Ethical Decision-Making Behavior’. Joumal of Applied Psychology 75(4):378–385. van der Wal, Z., A. Graycar, and K. Kelly. 2016. ‘See No Evil, Hear No Evil? Assessing Corruption Risk Perceptions and Strategies of Victorian Public Bodies’. Australian Journal of Public Administration 75(1):3–17. Victor, B. and J. B. Cullen. 1988. ‘The Organizational Bases of Ethical Work Climates’. Administrative Science Quarterly 33(1):101–125. Wortley, R., P. Cassematis, and M. Donkin. 2008. ‘Who Blows the Whistle, Who Doesn’t and Why?’ In Whistleblowing in the Australian Public Sector: Enhancing the Theory and Practice of Internal Witness Management in Public Sector Organisations, ed. A. J. Brown. Canberra: ANU E Press, 53–82.  Figure 1. Typology of corruption witnessed in the APS, 2015-2016 Sources: APS Employee Census 2015, 2016  Table 1. Descriptive statistics and correlations Variable M (SD) 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 1. Whistleblowing .33 (.47) 2. Fraud .14 (.35) .25 3. Conflict of interest .21 (.41) .09 -.04 4. Unlawful disclosure of information .04 (.20) .10 .04 .16 5. Perverting the course of justice .03 (.17) .08 .05 .10 .11 6. Nepotism .25 (.43) -.08 -.09 .11 .01 .00 7. Cronyism .64 (.48) -.29 -.26 .05 -.01 .01 .07 8. Bribery .03 (.16) .00 .02 .05 .06 .08 .01 -.05 9. Theft of official assets .07 (.25) .09 .13 .04 .07 .07 .00 -.11 -.08 10. Blackmail .01 (.10) .01 .00 .07 .07 .10 .05 -.00 .13 .04 11. Conspiring with criminals .01 (.11) .03 .05 .08 .05 .00 .01 -.01 .12 .04 .07 12. Adhocracy culture 2.56 (.94) .12 .04 -.06 -.07 -.04 -.02 -.15 -.03 -.03 -.01 .01 13. Hierarchy culture 3.38 (.94) .00 -.07 -.04 -.04 -.05 -.04 -.02 -.05 -.05 .00 -.02 .42 14. Clan culture 2.55 (.93) .14 .05 -.06 -.04 -.06 -.05 -.17 -.02 -.02 -.03 .00 .78 -.47 15. Market culture 3.35 (.96) .05 -.03 -.04 -.05 -.05 -.05 -.05 -.06 -.07 -.01 -.02 .58 -.71 .57 16. Satisfaction 2.90 (.88) .13 .06 -.07 -.06 -.05 -.03 -.14 -.04 -.02 -.05 -.01 .57 -.34 .66 .41 17. Commitment 3.05 (1.09) .13 .04 -.02 -.05 -.04 -.00 -.12 -.05 -.04 .03 .01 .49 -.29 .51 .37 .51 18. Job performance 6.91 (1.93) .06 .03 .00 .01 -.00 -.03 -.04 -.05 -.01 .02 .04 .12 -.02 .11 .07 .11 .19 19. Self-efficacy 3.50 (.85) .31 .08 .01 .04 .01 -.03 -.14 -.04 .03 .01 .05 .21 -.16 .21 .18 .20 .21 .14 20. Reporting of workplace bullying .49 (.50) .43 .09 .03 .05 .04 -.02 -.08 -.01 .03 .03 .05 .11 -.00 -.09 .06 .12 .13 .11 .21 ∣r∣ > .03 are significant at .05 % level of significance Table 2. Comparative trends by agency type: Analysis of variance results Variable Specialist Regulatory Policy Operational F   (N=201) (N=95) (N=567) (N=192)   Internal whistleblowing .22 (.42) .26(.44) .26(.44) .35(.48) 9.88*** Conflict of interest .27(.44) .15(.36) .20(.40) .21(.41) 2.25† Cronyism .71(.45) .67(.47) .71(.45) .62(.49) 7.03*** Nepotism .26(.44) .08(.28) .19 (.39) .27(.44) 10.39*** Mean values, followed by standard deviation values in parentheses. The highest mean value for each variable is in bold. Level of significance, p: † = p < .10; * = p < .05; ** = p < .01; *** = p < .001.  Table 3. Logistic regression results: Internal whistleblowing in the APS Variable Model I   Model II     B(SE) Odds ratio 95% C.I. B(SE) Odds ratio 95% C.I. Gender (1=male) -.07(.06) .93 .83-1.04 -.07(.07) 0.93 .82-1.06 Age (1=45 yrs and above) .10(.06)† 1.10 .99-1.23 .03(.07) 1.03 .91-1.17 Education (1=university) -08(.06) .92 .82-1.03 .02(.07) 1.02 .90-1.16 Level (1=junior) -.12(.06)* .89 .79-1.00 -.05(.07) .96 .84-1.09 Employment (1=fulltime) .14(.06)* 1.15 1.02-1.29 .08(.07) 1.09 .95-1.24 Organisational tenure (1=at least 5 yrs) -.15(.06)** .85 .76-.95 -.14(.07)* .87 .77-.99 Organisational size (1= at least 1001 employees) -.03(.06) .97 .83-1.08 .05(.07) 1.05 .92-1.20 Organisational type (1=operational)a .23(.06)*** 1.25 1.05-1.40 .18(.07)*. 1.20 1.04-1.38 Corruption Fraud .34(.05)*** 1.40 1.26-1.56 .35(.06)*** 1.42 1.26-1.60 Conflict of interest .17(.05)** 1.18 1.07-1.31 .20(.06)** 1.22 1.08-1.37 Unlawful disclosure of information .13(.05)** 1.13 1.03-1.24 .10(.05)† 1.10 .99-1.22 Perverting the course of justice .11(.05)* 1.12 1.02-1.23 .11(.06)† 1.11 .99-1.22 Nepotism -.12(.06)* .89 .79-.99 -.11(.06)† .89 .79-1.01 Cronyism -.47(.06)*** .62 .56-.70 -.50(.06)*** .61 .54-.69 Theft of official assets .07(.05) 1.07 .97-1.19 .05(.06) 1.05 .94-1.18 Organisational culture Adhocracy .01(.09) 1.01 .84-1.21 -.14(.11) .87 .71-1.08 Hierarchy -.18(.08)* .84 .72-.98 -.22(.09)* .81 .67-.96 Clan .16(.09)† 1.18 .98-1.41 .13(.12) 1.14 .91-1.43 Market .14(.09) 1.15 .97-1.37 .07(.10) 1.07 .88-1.31 Other Job performance -.10(.06) .91 .80-1.03 Satisfaction .02(.09) 1.02 .86-1.21 Organisational commitment .13(.08)† 1.13 .98-1.31 Self-efficacy .62(.07)*** 1.85 1.60-2.13 Reporting workplace harassment .99(.07)*** 2.69 2.36-3.07 Pseudo R2 (Nagelkerke) .19 .42 Model χ2 255.17***     641.19***     Level of significance, p: † = p < .10; * = p < .05; ** = p < .01; *** = p < .001. Confidence interval, C.I. a 1=operational agency, 0=specialist, regulatory and policy agency 34