THE IMPACT OF LOYALTY AND RETALIATION ON ACCOUNTANTS’ PROPENSITY TO BLOW THE WHISTLE: AN EXPERIMENT

This study experimentally examines the effect of loyalty level and retaliation strength, on accountants’ propensity to blow the whistle (PBW) when faced with a serious wrongdoing. It is argued that individuals with higher levels of loyalty are more likely to blow the whistle than are individuals with lower levels of loyalty. Besides, it is expected that individuals facing a weak retaliation are more likely to blow the whistle than are individuals facing a strong retaliation. Seventy working students enrolled in postgraduate of accounting program in two Indonesia universities participated in the study. Participants responded to three hypothetical whistle-blowing scenarios on one of two conditions –i.e., strong or weak retaliation for whistleblowing, and completed an instrument that measured loyalty. This study examines the arguments using the 2x2 between subjects design. Unconsistent with the philosophical theory of Larmer (1992), Corvino (2002), and Vandekerckhove and Commers (2004), this study empirically finds that the level of loyalty does not affect their PBW. In other words, this study fails to obtain empirical evidence about the effect of loyalty level on PBW. Besides, unlike results reported by Arnold and Ponemon (1991) and Liyanarachchi and Newdick (2009), a significant effect of retaliation strength on participants’ PBW is not found.


INTRODUCTION
Transparency in financial reporting is the mantra of the day, but the means to that transparency remains elusive. One element of transparency is the discovery of unethical situations and reporting the wrongs in the corporate environment. That's why whistleblowing has become an important issue for accountants today. Whistleblowing is the disclosure by organization members (former or current) of illegal, immoral, or illegitimate practices under the control of their employers, to persons or organizations that may be able to effect action (Near and Miceli, 1985). Some writers had tried to explore the reasons enabling accountants to choose if they will blow the whistle. Accounting professionals are exposed to ethical dilemmas on a regular basis and have the opportunity to uncover wrongdoing, possibly even before the effects are so disastrous that a company's viability is in question (Shawver and Clements, 2008).
Some studies had tried to analyze whistleblowing from a variety of views. Taylor and Curtis (2010) found that moral intensity relates to the likelihood of reporting and perseverance in reporting. Shawver and Clements (2008) found that accounting professionals are able to recognize unethical actions and that they are more likely to blow the whistle on situations involving higher materiality levels when they are guaranteed their job and assured of anonymity. Shawver and Clements (2008) also found that job satisfaction, organizational commitment, perceptions of ethical values, and perceptions of an enforced ethics code are not significant factors in the whistleblowing intentions for the sample of practicing management accountants. Elias (2008) found that auditing students generally felt that whistleblowing was necessary in cases of fraud, although they were slightly less likely to do it themselves, probably due to the high costs involved such as retaliation and the difficulty of finding a future job in the same profession. Also, more professionally committed auditing students were more likely to blow the whistle compared to other students. Similar relationships were found regarding students socialized early in the accounting profession.
Although popular interest in whistle blowing continues to increase, little is known about why some employees who observe wrong doing report it, while others do not. Many studies have tried to grasp the significant variables to account for the decision making process of whistle blowing. However, they have largely explored certain factors in the specific situation or environment and this has led to a lack of theoretical development in this area (Ponnu et al., 2008). Miceli and Near (1988) pointed out that there was no general or comprehensive theory on why some employees intend to report illegal or unethical behaviour in an organisation, while others do not. Until the present time, this observation holds true. For successful implementation of an ethical or legal system to protect whistleblowers and prompt employees to report wrongdoing, a comprehensive theoretical framework that provides a systematic explanation of whistle blowing intention is necessary. In addition, one of the major interests in whistle blowing research has been the gap between attitude and intention as well as intention and action. Even if many employees find whistle blowing to be morally right and necessary, only a few intend to blow the whistle in an organisation and far fewer take action of blowing the whistle (Ponnu et al., 2008).
Prior explications of whistleblowing behavior have suggested a number of important factors, including prosocial orientation (Dozier and Miceli, 1985), social status (Miceli and Near, 1984), and power and justice (Near et al., 1993). Gundlach et al. (2003) suggested that numerous reports of whistleblowing incidents appeared in the media, capturing the interest of management practitioners and society in general. This interest is mirrored by the attention academicians have given whistleblowing.
Academicians had tried to develop and analyze many research models relating to whistleblowing. Conceptual and theoretical research concerning the relation of employee loyalty with whistleblowing had been much discussed by researchers in philosophical and theoretical debate context. But still there remains a gap, those models did not give much attention to empirical testing concerning the relationship between employee loyalty and the propensity to blow the whistle.
Not only investigated variables expected as potential motives of whistleblowing, some studies also investigated variables expected as hindering factor of whistleblowing. One of the most popular and of the most investigated variables is retaliation. According to Rehg et al. (2008), retaliation against whistleblowers represents an outcome of a conflict between an organization and its employee, in which members of the organization attempt to control the employee by threatening to take, or actually taking, an action that is detrimental to the well-being of the employee, in response to the employee's reporting, through internal or external channels, a perceived wrongful action. Keenan's definition (2002) explicitly included both "taking an undesirable action against an employee or not taking a desirable action." Despite its importance, very little attention in accounting research has been paid to whistleblowing, possibly because it is a hidden activity, which makes it a difficult topic to research. Also, given the clearness of accountants role in whistleblowing, still there remains a question if the accountants themselves have intention to react correctly? As such, researchers can only examine behavioral intentions rather than the actual behavior of respondents. Besides, still there remains a gap in literature concerning the relationship between loyalty and whistleblowing because it is discussed in theoretical and philosophical area but not empirical.
This research attempts to contribute empirical study in accounting field concerning the relationship between employee loyalty and whistleblowing. The writer chooses this topic to contribute empirical evidence for literature relates to whistleblowing and employee loyalty which has not been investigated by former studies. Also, this research investigates the effect of retaliation on whistleblowing. So, this study specifically attempts to investigate the impact of loyalty and retaliation on accountants' propensity to blow the whistle (PBW). Miceli and Near (1984) define whistleblowing as actions taken by current or former organizational members to report illegal, unethical, or illegitimate activities, which are under the control of management, to persons who are willing and able to correct such misconduct. According to Rocha and Kleiner (2005), the term whistleblowing refers to allegations made by someone inside an organization, and hence, this term is used to distinguish allegations made by insiders from those made by outsiders.

Whistleblowing
Accountants in industry are often keys to the planning and control processes of their organizations. These accountants may find themselves faced with the dilemma of whether or not to blow the whistle on an organizational wrongdoing. Some people, both inside and outside the profession, would view the accountants' decision to blow the whistle as morally and professionally justifiable; while others would see it as harmful to the employing organization and would advise against doing so (Chiasson et al., 1995).
The topic of whistleblowing is important because it contributes to improvements in internal control systems (Patel, 2003). Moreover, whistleblowing can improve longterm organizational effectiveness because whistleblowers may suggest solutions to organizational problems (Brief & Motowidlo, 1986). Additionally, organizational members, investors, creditors, and society in general can benefit from the cessation of organizational wrongdoings revealed by whistleblowers (Miceli et al., 1991).
Recognizing its importance, the Chartered Institute of Public Finance and Accountancy in the U.K. has called for effective whistleblowing procedures to be established in order to restore public confidence in accounting standards. In the U.S., the Committee of Sponsoring Organizations of the Treadway Commission (COSO 1992) has emphasized the role of communication in internal control. In particular, COSO advocated the importance of whistleblowing. In Australia, a study by the Australian Society of Certified Practicing Accountants and the Institute of Chartered Accountants in Australia published by The Middleton Report in 1994 also emphasized the importance of whistleblowing as an issue for the profession. Further importance of this topic is shown in Knapp's collection in 1993 of well-known audit cases and the importance placed on whistleblowers' assistance in detecting fraud (Patel, 2003).
The year 2002 will be remembered as the year that corporations failed many of their constituencies. Stories of corporate greed and wrongdoings created headlines. The importance of whistleblowing as an internal control mechanism is most clearly illustrated by Time Magazine in 2002, selecting three whistleblowers at Enron, WorldCom, and the FBI as Persons of the Year for 2002. This, and other incidents in the U.S., have triggered increasing attention on the topic of whistleblowing as an internal control mechanisms in other industrialized countries such as Australia, the U.K., Canada, and France. In an economic climate where corporations and governments cannot be trusted to be ethical, whistleblowing may be one way of revealing wrongdoings (Patel, 2003).
Researchers began studying whistleblowing in the late 1960s or early 1970s; since then a substantial literature has developed. More than three decades ago, Bowman et al. (1984) in Brewer and Selden (1998) reviewed this literature and found approximately 1,400 contributions. Most items consisted of anecdotal accounts and single incident case studies. A few scholars were exploring the behavioral and organizational implications of whistleblowing, but the state of knowledge on these topics was highly speculative. During the past decade, this void largely has been filled as researchers have studied samples of whistle blowers intensely, using sound descriptive and empirical research methods. Their findings converge to provide an emerging picture of whistleblowers that is remarkably consistent and somewhat surprising (Brewer dan Selden, 1998).
The position of whistleblowers has improved considerably in recent years. Where they were once regarded as individuals who betrayed their employers, in the "post Sarbanes-Oxley age" (following the US whistleblowing legislation that was adopted in 2002 and sought to reform corporate governance in the wake of a number of highprofile corporate financing scandals) they are pictured as courageous people with strong ethical principles, who have the right to anonymity and protection. This change in perception can be found in the text of several whistleblowing policies. The chemical company DSM, for example, "encourages its employees who have concerns about suspected serious misconduct to come forward and express these concerns without fear of punishment or unfair treatment" (Europian Industrial Relations Review, 2006).
Whistleblowing has a checkered reputation, at best (Taylor and Curtiz, 2010). First, it directly implies ethical failure and involves one person judging the ethical behavior of another. Second, whistleblowing is often anonymous, depriving the reported-on individual the right to face his or her accuser. Third, whistleblowing often entails reporting outside of the established lines of communication and authority. Finally, whistleblowing requires trust in those at the top of the organization to take appropriate actions when they learn of misdeeds by their employees. Unfortunately, even when wrongdoing is detrimental to a large number of people external to the organization (e.g. fraudulent financial reporting resulting in artificially high stock prices), those internal to the organization often view the whistleblower's report (rather than the initial wrongdoing) as the cause of their losses. Evidence of widespread retaliation and cost to the whistleblower himself is well documented (Jos et al., 1989). However uncomfortable we are with the notion of reporting on the behavior of others, whistleblowing is an important organizational control. Indeed, industry surveys and academic research (Miceli and Near, 2002) support the contention that reporting mechanisms aid in the prevention and detection of unethical behavior. Whistleblowing, the disclosure by an organization member of perceived organizational wrongdoing to authority that is able to take action (Near and Miceli, 1985), gives dilemma to the managers. Whistleblowers identify and sometimes offer solutions for organization problems. Yet in the same time, they also threaten organization authority structure or workgroup functionalization (Greenberger et al., 1987;Jensen, 1987). Possibly, when this threat is considered to get bigger, the whistleblower will face retaliations (Parmerlee et al., 1982).
Similar to the statement above, Liyanarachchi and Newdick (2009) suggest that given the intimate knowledge insiders or employees have about their organization, relying on employees to blow the whistle is sensible. But it can also be problematic. For example, Appelbaum et al. (2006) state that a study conducted in the United States by the Ethics Resource Center and reported in the January 2005 edition of Strategic Finance pointed out that 44% of all non-management employees do not report misconduct they observe. Similarly, addressing the fear of retaliation, Rocha and Kleiner (2005) state that the dilemma the potential whistleblower has is a struggle between doing what is right and suffering the consequences, or just being quiet, pretending it does not exist.

Loyalty and Whistleblowing
Loyalty has a very long and interesting discussion among the researchers in their writings. Over the last thirty years, a variety of definitions for loyalty have appeared in the organizational literature. Some descriptions can be traced to earlier work on the relationship between firms and their employees, which emphasized the devotion of workers to their organizations as reflected in their compliance with instructions from supervisors. Other definitions have emerged more recently from research on organizational commitment (Meyer and Allen, 1991) and related variables (Bhappu, 2000;Werhane, 1999), in which loyalty has sometimes been used as a synonym for one or more forms of commitment.
Today, definitions of loyalty range from specific to broad, and capture attitudes and behaviors involving a variety of foci. As the set of definitions continues to expand, it becomes increasingly difficult to determine exactly what is meant by ''loyalty" and how it should be measured. This leads to contradictory findings about the presence or absence of loyalty in organizations and makes it more difficult to identify loyalty's antecedents and outcomes (Coughlan, 2005).
About thirty years ago, loyalty was viewed broadly as an employee's feeling of attachment to an organization (Buchanan, 1974). This attachment was later connected to more specific behaviors such as supporting an organization and the individuals within it (Niehoff et al, 2001) and practicing good citizenship (Rusbult et al, 1988) . Such attachment was also thought to be evident when an employee passed on an attractive position with another organization (Logan, 1984) or simply remained with one organization for some length of time (Cunha, 2002).
In recent years, definitions of loyalty have centered on employees not harming their colleagues or the firm that employs them (Coughlan, 2005). Such conceptualizations hint at the importance of ethical behavior and trust in discussions of loyalty. Coughlan (2005) suggested that among the earliest to address the issue of loyalty in organizations is Allport in 1933 who suggested it connoted "a selection of values, an adherence to some principle of conduct we consider good". Support for their inclusion in descriptions of loyalty can be found in a variety of sources, including writings by philosophers. For example, Oldenquist (1982) suggests "our loyalties define the domains in which we are willing to universalize moral judgments". Similarly, Pettit (1988) argues that the reason for a loyal individual's action must be universalizable within a given community; otherwise the rationale will not have the same motivational hold on all members. According to the principle of universalizability, in order to determine whether a potential act is morally appropriate, individuals must first identify the moral principle that guides the behavior. They must ask themselves whether that principle could be conceived of as a universal law. If it cannot be, the action must be foregone or another principle must be utilized as the moral basis for action (Coughlan, 2005).
The conceptualization of loyalty advanced by these philosophers (Oldenquist, 1982;Pettit, 1988) acknowledges many of the dimensions found in Allport's definition, including notions of voluntary choice (of domains or communities) and morality. Coughlan (2005), by drawing on major elements from the definitions discussed above, proposes a new conceptualization of loyalty that is based in morality and distinguishable from commitment. Accordingly, he propose "loyalty is reflected in behavior that can be tied to an implicit promise, voluntarily made by an individual operating in a community of interdependent others, to adhere to universalizable moral principles in pursuit of individual and collective goals". Coughlan (2005) also examines former instruments attempting to measure loyalty and corrects by distinguishing between loyalty and commitment, until a new instrument is more precisely established to measure loyalty.
Most loyalty discussion in literatures expects moral dilemma between employee loyalty and whistleblowing. The standard view of ethicists is that employees possess prima facie duties of loyalty and confidentiality to their employers and that whistleblowing cannot be justified except on the basis of a higher duty to the public good (Larmer, 1992). Generally, based on this standard view, there are three schools concerning the relationship between employee loyalty and whistleblowing.
The first view suggested by Duska in 1985 is against the standard view. Duska follows John Ladd in defining loyalty as a "wholehearted devotion to an object of some kind". That object may be a person or a group of persons; thus, the fact that companies are not individual persons is not, by itself, a bar to loyalty's being appropriate toward them. According to Duska, however, not every group is a proper object of loyalty (Corvino, 2002). Ronald Duska has argued that employees do not have even a prima facie duty of loyalty to their employers and that whistleblowing needs, therefore, no moral justification (Larmer, 1992).
The standard view believes that employees have a prima facie duty of loyalty to their companies, one that sometimes conflicts with other duties, such as the duty to blow the whistle in response to dangerous or unethical practices (Corvino, 2002). Duska's view is more radical and denies the existence of any such duty. According to Duska, "one does not have an obligation of loyalty to a company, even a prima facie one, because companies are not the kind of things that are proper objects of loyalty." Duska is not claiming, of course, that one may treat one's colleagues callously or deceitfully, or that employees have no obligations to their companies. Rather, he is suggesting that whatever duties an employee has, loyalty is not among them, since loyalty to companies involves a kind of category mistake (Corvino, 2002).
The second view is in contrast to Duska, suggesting that loyalty to one's employer is appropriate. According to it, one has an obligation to be loyal to one's employer and, consequently, a prima facie duty to protect the employer's interests. Whistleblowing constitutes, therefore, a violation of duty to one's employer and needs strong justification if it is to be appropriate (Larmer, 1992).
Behind this view lies the assumption that to be loyal to someone is to act in a way that accords with what that person believes to be in his or her best interests. To be loyal to an employer, therefore, is to act in a way which the employer deems to be in his or her best interests. Since employers very rarely approve of whistleblowing and generally feel that it is not in their best interests, it follows that whistleblowing is an act of betrayal on the part of the employee, albeit a betrayal made in the interests of the public good (Larmer, 1992).
The third view argues those two views before, is that employee loyalty and whistleblowing are compatible and have no moral dilemma between them. Some writers have argued the view of Duska and Bok above, like Larmer (1992), Corvino (2002), and Vandekerckove and Commers (2004). Larmer (1992) suggests that Duska's view does not seem adequate, however. First, it is not true that loyalty must be quite so reciprocal as Duska demands. Ideally, of course, one expects that if one is loyal to another person that person will reciprocate in kind. There are, however, many cases where loyalty is not entirely reciprocated, but where we do not feel that it is misplaced. So it is not nonsense to suppose that loyalty may be appropriate even though it is not reciprocated. Inasmuch as he ignores this possibility, Duska's account of loyalty is flawed. Second, even if Duska is correct in holding that loyalty is only appropriate between moral agents and that a company is not genuinely a moral agent, the question may still be raised whether an employee owes loyalty to fellow employees or the shareholders of the company. Third, it seems wrong to suggest that simply because the primary motive of the employer is economic, considerations of loyalty are irrelevant. An employee's primary motive in working for an employer is generally economic, but no one on that account would argue that it is impossible for him or her to demonstrate loyalty to the employer, even if it turns out to be misplaced. All that is required is that his or her primary economic motive be in some degree qualified by considerations of the employer's welfare. Similarly, the fact that an employer's primary motive is economic does not imply that it is not qualified by considerations of the employee's welfare. Given the possibility of mutual qualification of admittedly primary economic motives, it is fallacious to argue that employee loyalty is never appropriate. Beside criticisms from Larmer (1002), Corvino (2002) and Vandekerckove and Commers (2004) also criticize Duska's view with long criticisms.
Criticisms against Bok's view relating to loyalty, Larmer (1992) suggests that the concept is mistaken. It ignores the fact revealed by Near and Miceli (1985)  By ignoring the possibility that blowing the whistle may demonstrate greater loyalty than not blowing the whistle, it fails to do justice to the many instances where loyalty to someone constrains us to act in defiance of what that person believes to be in her best interests. I am not, for example, being disloyal to a friend if I refuse to loan her money for an investment I am sure will bring her financial ruin; even if she bitterly reproaches me for denying her what is so obviously a golden opportunity to make a fortune (Larmer, 1992).
Some criticisms by Larmer (1992), Corvino (2002), and Vandekerckove and Commers (2004) against the view of Duska and Bok take us to a more adequate definition of being loyal to someone is that loyalty involves acting in accordance with what one has good reason to believe to be in that person's best interests. To the degree that an action is genuinely immoral, it is impossible that it is in the agent's best interests. Loyalty does not imply that we have a duty to refrain from reporting the immoral actions of those to whom we are loyal. An employer who is acting immorally is not acting in her own best interests and an employee is not acting disloyally in blowing the whistle. Indeed, the argument can be made that the employee who blows the whistle may be demonstrating greater loyalty than the employee who simply ignores the immoral conduct, inasmuch as she is attempting to prevent her employer from engaging in self-destructive behaviour (Larmer, 1992).
Loyalty implies a desire that the person to whom one is loyal take no moral stumbles, but that if moral stumbles have occurred that the person be restored and not simply punished. A loyal friend is not only someone who sticks by you in times of trouble, but someone who tries to help you avoid trouble. This suggests that a loyal employee will have a desire to point out problems and potential prob lems long before the drastic measures associated with whistleblowing become necessary, but that if whistleblowing does become necessary there remains a desire to help the employer. The conclusion is that to whistleblow for reasons of morality is to act in one's employer's best interests and involves, therefore, no disloyalty, more than that is evidence of higher loyalty level of employee (Larmer, 1992).
Explanation above leads us to the first hypothesis: H 1 : Individuals with higher levels of loyalty are more likely to blow the whistle than are individuals with lower levels of loyalty.

Retaliation and Whistleblowing
Studies identify that potential retaliation against whistle-blowing is important in explaining one's decision to engage in whistle-blowing (Liyanarachchi and Newdick, 2009). For example: The changes in cultural values and events have made it easier for employees to stand up for what is right. However, whistleblowing has many consequences and retaliation seems to be present in almost all the cases. (Rocha and Kleiner, 2005).
There are many forms of retaliation. Bok in 1980 listed downgrading, giving work without responsibility, giving more work, and work that requires new skills or qualifications as forms of retaliation against whistleblowers. Another more severe form of retaliation is to order whistleblowers to take psychiatric fitness-for-duty examinations (Liyanarachchi and Newdick, 2009). Retaliation may also come in the form of coercion to silence the whistle-blower or termination of employment (Mesmer-Magnus and Viswesvaran, 2005;Parmerlee et al., 1982). These retaliatory measures vary from one another in terms of their severity or strength, and potential whistleblowers may evaluate not only the possible retaliation but also the strength of such measures when deciding to blow the whistle.
Some countries had long ago established act and law attempting to protect whistleblowers in their countries. For example, The Whistleblower Protection Act of 1989 is a United States federal law that protects federal whistleblowers who work for the government and report agency misconduct. A federal agency violates the Whistleblower Protection Act if agency authorities take (or threaten to take) retaliatory personnel action against any employee or applicant because of disclosure of information by that employee or applicant. Whistleblowers may file complaints that they believe reasonably evidences a violation of a law, rule or regulation; gross mismanagement; gross waste of funds; an abuse of authority; or a substantial and specific danger to public health or safety. The Whistleblower Protection Enhancement Act was introduced in 2009 by Senator Daniel Akaka to amend federal personnel law relating to whistleblower protections.
Indonesia has established The Law of The Republic of Indonesia number 13 of 2006 concernning Protection of Witness and Victim. The law establishes Witness and Victim Protection Agency, hereinafter referred to as LPSK, which is the LPSK tasked and authorized to provide protection and rights of others to witness and/or victim as set forth in this law. The witness is a person who can provide information in the interest of the inquiry, investigation, prosecution, and examination in court on a criminal case he heard himself, he seen himself, and/or he has experienced himself. Victim is a person who suffered physical, mental, and/or economic loss caused by a criminal offense. The law defines the threat as any action that will produce, either directly or indirectly, which resulted in the witness and/or the victim to fear and/or forced to do or not do something regarding to giving testimony in a criminal justice process. Protection defined by the law is all effort of right fulfillment and providing assistance to provide sense of security to witness and/or victim that must be performed by the LPSK or other LPSK in accordance with the provisions of this law. Till this research writing, revision of the law is still proposed and discussed by concerning agencies.
However, research indicates that the organization is more likely to retaliate against the whistleblower (Near and Miceli, 1985). Miceli and Near in 1994 noted that organizations retaliate against whistleblowers in order to prevent a full public knowledge of the complaint, discredit the whistleblower and discourage other potential whistleblowers (Elias, 2008). Retaliation can take several forms such as isolation of the whistleblowers, character defamation and these actions may sometimes be imposed by lower levels managers with/without the knowledge of corporate executives (Parmerlee et al., 1982). In a survey of corporate executives, Keenan and Krueger (1992) found that only 53 percent of surveyed executives believed their companies protected whistleblowers from reprisal. A survey conducted by the Ethics Resource Center found that 44 percent of all non-management employees did not report misconduct they observed out of fear of retaliation (Verschoor, 2005). Miceli and Near in 1992 showed that employees reporting more serious or entrenched wrongdoing will face more retaliation (Elias, 2008). In fact, Glazer and Glazer in 1989 found that 89 percent of whistleblowers will have difficulty finding employment in the public sector.
Studies on the effects of retaliation on whistleblowing show mixed results (Miceli and Near, 1988;Near and Miceli, 1995;Near et al., 2004). Therefore, although an individual's decision to blow the whistle is likely to be affected by retaliation against such an act, this impact may be complex making its prediction difficult. Liyanarachchi and Newdick (2009) make Keenan in 1995 as an example who shows that managers' fear of retaliation has a strong influence on feeling obliged to blow the whistle. Mesmer-Magnus and Viswesvaran (2005) report that while threat of retaliation negatively affects the whistle-blowing intentions, threat of retaliation is not related to actual whistle-blowing behavior. Arnold and Ponemon (1991) and Liyanarachchi and Newdick (2009) found that retaliation strength has a significant effect on propensity to blow the whistle. Despite the mixed nature of empirical results, the general pattern is that the harsher the retaliation, the less likely an individual is to blow the whistle. Accordingly, the strength of potential retaliation is likely to influence an individual's decision to blow the whistle.
This leads to the second hypothesis: H 2 : Individuals facing a weak retaliation are more likely to blow the whistle than are individuals facing a strong retaliation.

Research Design
The experimental method was used. Independent variable examined is loyalty level, which is measured to be high and low loyalty. The treatment in this experiment is retaliation, included 2 levels of retaliation (strong or weak); thus the experiment confirmed to a 2x2 between subject factorial design. Two versions of the experimental task were created, one involving the strong retaliation (penalty) and the other involving the weak retaliation (affiliation) condition. Participants were randomly assigned into one of the four conditions. (Arnold and Ponemon, 1991;Greenberger et al., 1987).
In this research, loyalty is measured by using the instrument developed by Coughlan (2005) who offers an alternative construct to measure loyalty in a way more precisely and completely. Loyalty, according to his research, has several dimensions. Those dimensions are normative commitment, attitudinal loyalty, and applied loyalty.
The dimension of normative commitment represents the feelings of obligation to remain with one's organization. Though this dimension is not explicitly defined as loyalty, it seems to suggest that it means remaining with one's firm, perhaps because of a specific investment that the organization makes in the employee or because of a perceived psychological contract. On the whole, this dimension is measured by six items focusing on one specific decision of an employee, whether or not to leave the organization. The dimension of attitudinal loyalty focuses on the role of moral values in guiding the behavior of employees. This dimension consists of three items which are meant to reflect loyalty's three unique components: its voluntariness, its emphasis on adherence and its moral basis. The dimension of applied loyalty consists of three items relating to the application of moral principles hold together in dilemma cases at workplace generally. For each item, subjects were asked to indicate their level of agreement on a seven-point Likert-type scale, ranging from 1 (Completely Disagree) to 7 (Completely Agree).
In this research, retaliation is measured by using instrument designed by Arnold and Ponemon (1991) which is used and developed by Liyanarachchi and Newdick (2009). They classify the strength of retaliation into either strong (penalty) or weak (affiliation) retaliation.
Penalty or strong retaliation refers to the disciplinary consequences in the form of threats to person or property, lawsuits, job termination, or imprisonments and, thus, is a stronger form of retaliation. Affiliation refers to the form of relationships with others inside or outside the organization, i.e., other individuals are subject to retaliation rather than the whistle-blower. Thus affiliation is a weaker form of retaliation when compared to the penalty-type retaliation for the whistleblower (Arnold and Ponemon, 1991). For example, if an individual lost his or her job due to blowing the whistle, then this would be retaliation in the form of a penalty. On the other hand, if someone else lost their job due to an individual blowing the whistle, then this is retaliation in the form of affiliation (or weak retaliation).
The variable of whistleblowing is measured by using a research instrument designed by Arnold and Ponemon (1991) which is also used and developed by Liyanarachchi and Newdick (2009). This instrument involved three whistleblowing scenarios and each participant was asked to indicate the likelihood that the individual in each of the scenario would blow the whistle. The three scenarios allowed for sufficient variability in the dependent variable, thus allowing for more consistent and reliable observations on PBW. Writer modifies the instrument in accordance with the condition of Indonesia, yet the whistleblowing scenarios are presented in a third-person context as the original instrument. This approach is adopted in studies in this area of research (Arnold and Ponemon, 1991) and is recommended by Rest (1986) when using a hypothetical ethical scenario. For each hypothetical case, participants were asked to assess the likelihood of the individual described in the scenario blowing the whistle. Participants indicated the PBW on a sevenpoint Likert-type scale for each scenario (1 = "very unlikely" to 7 = "very likely").

Experimental Procedure and Protocol
All participants began the experiment by reading the experiment instruction and fill assent sheet to follow the experiment. Instruction and assent sheet are adopted from Tayler (2010) then adjusted by this research requirement. Hereafter participant listened to instruction of word-of-mouth of researcher containing general instruction of experiment activity. While participants were reading instruction sheet and fills participation assent sheet, experiment materials were distributed to participant consisting of cases and research questionnaires. Next, participants were asked to open instrument materials concurrently and complete experiment cases about whistleblowing. After completing cases tasks of whistleblowing, participants were requested to answers one question of manipulation check. Next, participants were requested to answer questionnaires which measures to their loyalty level, whether it is high or low. Finally, participants were requested to complete demography data.

Participants
Participants in this research are students who already finished their undergraduate study in accounting majors, and are carrying on postgraduate study of accounting whether Magister Science of Accounting, Master of Accounting, Magister of Accounting Knowledge, or Accountant Profession Education Program. The students selected as respondents just only who are working at a particular organization or institution, and have completed college subjects concerning accounting at undergraduate in advance. The experiments were executed six times on regular lecturing term and participation was voluntarily. Researcher selects accounting students as respondents with these criteria because of several motives (Liyanarachchi and Newdick, 2009). Firstly, students in this level had achieved a good understanding of accounting and business ethics issues due to their previous learning experiences at university. Secondly, whistleblowing obviously has more cost for a new recruited accounting employee and relatively just finished his or her study. Thirdly, students who are working at a particular organization or institution are of course more relevant for examination to their loyalty level.

Manipulation Check
Researcher uses the manipulation check to determine if participants understand the condition -i.e. treatmentexperienced by the whistleblowers in the experiment cases by one question. That question asks if participants know who would be threatened if the wrongdoings are revealed. Participants are requested to choose one of two available answer options, which are the whistleblowers or others (not the whistleblowers). The result of manipulation check analysis points out that from 101 participants, just as much 70 participants (69.31%) who pass the manipulation check, 29 participants (28.71%) drop off manipulation check, and 2 participants (1.98%) do not answer completely.

Statistical Analysis
Internal validity is measured by undertaking factor analysis (Cooper and Schindler, 2011). If the loading factor is above 0.50 therefore the item is valid and appropriate (Hair et al., 2010). The result points out that all item has one loading factor upon 0.5 except item L6. Based on that result, all question from loyalty questionnaire is valid and can be used for further analysis, but item question number 6 is not valid so is discarded and not used for further analysis.
In this research, measurement of item homogeneity of questionnaire refers to item-total correlation and Cronbach's alpha value (Cooper and Schindler, 2011). The result of reliability examination using Pearson Correlation Formula points out correlation values which ranging from 0.347 until 0.784. Those values are greater than one presupposed which is 0.31 (Popham, 2002). Besides, by seeing Cronbach's Alpha value, if alpha > 0.60 therefore the questionnaire can be relied on (Hair et al., 2010). Questionnaire in this research gets Cronbach's Alpha value as big as 0.833. Remembering the item number 6 is discarded based on previous validity test before, therefore the Cronbach's Alpha value -if item number 6 is discarded-is 0.813. This two reliability examinations point out that those items except item number 6 can be used for further analysis.
The result of One Sample Kolmogorof-Smirnov Test points out that residual value is normally distributed with significance level even greater than 0.05, which is 0.232. Besides, significance level of One Sample Kolmogorof-Smirnov Test for the dependent variable (PBW) is greater than 0.05, which is 0.147. This result points out that normality test accomplishes normality assumption for ANOVA, so ANOVA can be used for data analysis in this research.
The measurement result of homogeneity of variance using Levene's Test of Equality of Error Variances points out that significance level is upon 0.05, which is 0.897, therefore H a is refused. It points out that the data variance is homogeneous or no significant difference among sample groups. This result supports the homogeneity of variance assumption of ANOVA, so ANOVA can be used for data analysis.

Hypothesis Results
The dependent variable examined in this research is propensity to blow the whistle (PBW), meanwhile independent variables are loyalty level (high and low) and retaliation strength (strong and weak). Table 1 presents the descriptive statistic to see the difference between treatments and also among groups, by seeing the means and standard deviations of dependent variable, which is propensity to blow the whistle (PBW). As expected, table 1 shows that mean PBW increases as the threat of retaliation becomes weaker (5.0351> 5.0000). A similar pattern exists for high loyalty, that is, mean PBW is greater at weak retaliation than at strong retaliation (5.1167> 4.8939). The same pattern happens on weak retaliation, which is, mean PBW is greater once loyalty gets higher (5.1167> 4.9444). In contrast to hypothesis, mean PBW is just greater as loyalty is low (5.0476> 5.0000). The same pattern happens on low loyalty, which is, mean PBW is greater as the threat of retaliation is strong (5.2333> 4.9444).
And so is it on strong retaliation, mean PBW is greater as loyalty is low (5.2333> 4.8939). Table 2 shows the two-way ANOVA results of this research. The first hypothesis states that individuals with higher levels of loyalty are more likely to blow the whistle than are individuals with lower levels of loyalty. But, the result of ANOVA above shows that respondents' loyalty levels have no significant effect on their PBW at level 0.05 (F=0.073; p=0.789). Thus, the first hypothesis is not empirically supported in this research.
The second hypothesis states that individuals facing a weak retaliation are more likely to blow the whistle than are individuals facing a strong retaliation. The result of ANOVA above shows that retaliation strength has no significant effect on their PBW at level 0.05 (F=0.011; p=0.915). Thus, the second hypothesis is not empirically supported in this research either.

DISCUSSION AND CONCLUSION
This research attempts to contribute empirical research in accounting area concerning the relationship between employee loyalty and whistleblowing. This research intends to contribute empirical evidence for literature concerning the relationship between whistleblowing and employee loyalty that hasn't is examined by former studies. Besides, this research also examines retaliation effect on whistleblowing. So, specifically this study attempts to examine the impact of loyalty and retaliation on accountants' propensity to blow the whistle (PBW).
The first hypothesis states that individuals with higher levels of loyalty are more likely to blow the whistle than are individuals with lower levels of loyalty. This research result failed to give empirical evidence to theory proposed by Larmer (1992), Corvino (2002), and Vandekerckove and Commers (2004). They suggest that employee loyalty and whistleblowing are compatible and have no moral dilemma between them, even the higher one's loyalty level, the higher his propensity to blow the whistle, and an employee who blows the whistle may be demonstrating greater loyalty than the employee who simply ignores the immoral conduct, inasmuch as he or she is attempting to prevent his or her employer from engaging in self-destructive behaviour. The conclusion of their theory is that to whistleblow for reasons of morality is to act in one's employer's best interests and involves, therefore, no disloyalty, more than that is evidence of higher loyalty level of employee (Larmer, 1992). Yet, the result of this research failed to give empirical support for this theory. The result shows that loyalty levels have no significant effect on one's propensity to blow the whistle.
This research result supports criticism by Varelius (2009) who suggests that this new approach of loyalty to dealing with the moral problem of whistle-blowing offered by Corvino (2002), Larmer (1992), and Vandekerckhove and Commers (2004) is not acceptable. The focus of that argument is that being loyal to one's employer is not incompatible with blowing the whistle about their wrongdoing, because employee loyalty and whistleblowing serve the same goal, the moral good of the employer, and even blowing the whistle demonstrates a higher loyalty level. According to Varelius (2009), the arguments of Corvino, Larmer, and Vandekerckhove and Commers to support their conception of loyalty are not sufficient to be acceptable. According to him, the proponents of the new conception of loyalty present to support their view do not amount into a satisfactory explanation for why we have been mistaken in accepting our original conception of loyalty and why it ought to be replaced with the new conception of loyalty they propose.
This result of research -on one hand-is in accordance with theory revealed by Duska that loyalty and whistleblowing don't have any conflict absolutely and no relationship each other, although -on the other hand-this research result still points out that respondents tend to consider that loyalty is a variable which exists in their relationship with organization. It is shown by mean of all of respondents' answer to loyalty instrument which is greater than 4 (median of scale of 1 to 7), which is as big as 5.027.
The second hypothesis states that individuals facing a weak retaliation are more likely to blow the whistle than are individuals facing a strong retaliation. The research of Lianarachchi and Newdick (2009) concludes that individuals facing a weak retaliation are more likely to blow the whistle than are individuals facing a strong retaliation. The result of this research is in contrast to previous research concerning the effect of retaliation strength that can reduce one's propensity to blow the whistle.
Causal factor why the second hypothesis is not supported is possibly because respondents consider that cases they found constitute serious wrongdoings, so they tend to blow the whistle whatever the level of retaliation faced. Miceli et al. (1991) suggest that possibility of why retaliation level has no effect to whistleblowing can caused by perceived seriousness of a malpractice. If a wrongdoing is serious enough or intolerable, someone will tend to blow the whistle whatever the consequence. In other words, in a wrongdoing case considered serious one, the threat of retaliation is not so counted and considered. So, fear effect of retaliation is more likely to be found in malpractice case that is not serious.
Pretty much other possibility of causal factor of difference of this result with previous research is because of the difference of its respondent used in this research. Previous research used undergraduate students who had not worked yet (Liyanarachchi and Newdick, 2009), while this research use students of some postgraduate programs who are working. This research has several limitations: First, this research uses experimental design to examine the effect of loyalty and retaliation on accountants' propensity to blow the whistle. Pretty much information needed by respondents to make decision was not ranged in cases given, since in the real world, situations faced by accountants are more complex to decide whether or not to blow the whistle. Second, one out of six times experiment is done at the moment respondent would follow semester examination, so some respondents did not fully participate or participated in a state of far from concentration. Third, the researcher limitation to measure how far respondents really internalized the manipulations given on the cases.
Limitations in this research make opportunities for future research. Research concerning whistleblowing in Indonesia particularly in accounting area is still not much done so it is expected that future research would deepen the study about it. Several suggestions of researcher for future research are: First, future research can insert any other variable that is able to explain the relationship of variables in this research, e.g. perceived seriousness of a malpractice, religiosity, etc. Second, future research can also examine specific segment of accountant as management accountant or auditor, either public or non-public. Third, future research is suggested to have more perfect experimental design to be able to figure reality condition. Fourth, future experiment should be executed not on busy time so respondents could concentrate to follow the experiment. Fifth, future research would better have a design enabling the researcher to examine more precisely on how far respondents are able to internalize the manipulation given. Sixth, future research gets opportunity to investigate this topic by using other method besides experiment, which is survey for instance.