Improving Ethics: Extending the Theory of Planned Behavior to Include

Moral Disengagement

Josh Cieslewicz, Greg Burton and Erv Black

Abstract

We extend the Theory of Planned Behavior (TPB) for ethics in the workplace. Using a path modeling methodology, we find evidence that, for ethics, moral disengagement is an antecedent to the TPB predictors of attitude, subjective norms, and perceived behavioral control

(PBC). We show that the TPB predictors mediate the influence moral disengagement has on ethical behavioral intentions. Thus, to improve ethical behavior, reducing moral disengagement is critical. We find support for including both types of PBC (self-efficacy and locus of control) when modeling ethical behavior. We use four ethics scenarios and international data to test this model. We also evaluate potential positive influences on ethics in the workplace and find that recency of ethics training interacts with religiosity and activates it to reduce moral disengagement. We also find that principles-based ethics training enhances professionals’ self- efficacy to behave ethically. Experience, including time as a member in a professional accounting organization, increases both locus of control and self-efficacy to behave ethically.

These variables—recency of ethical training, religiosity, principles-based professional ethics training, and experience—influence parts of the core TPB model, which in turn lead to improved ethical behavioral intentions.

KEY WORDS: moral disengagement, theory of planned behavior, ethics, principles-based ethics training 1. Introduction

The Theory of Planned Behavior (TPB) is a well-studied theory that has been used to measure the intention to behave ethically in many disciplines (see, for example, Chang, 1998;

Sutton, 1998; Armitage and Conner, 2001; and more recently, Alleyne et al. 2018; Seitz, 2020).

While the TPB incorporates specific antecedents to behavioral intention, we extend the model and show that prior TPB ethics studies have omitted an important precursor to ethical behavior: moral disengagement. We start with the TPB model proposed by Ajzen (1991) and later clarified by Ajzen (2002) and add measures of moral disengagement discussed by Bandura (1990a,

1990b, 1999).

At a theoretical level, we expect that moral disengagement will influence ethical intentions of practicing accountants through the attitudes, subjective norms, and perceived behavioral control constructs, which we hypothesize mediate ethical intentions in the TPB framework. Moral disengagement is the cognitive process individuals use to avoid the self- condemnation that ensues from contemplating or doing something that they know is wrong

(Moore et al. 2012). We show that there are antecedents that can reduce an individual’s propensity to morally disengage when facing ethical dilemmas. Using a path modeling methodology, we extend the TPB for ethics studies, and we also identify ways in which the core

TPB predictor variables (attitude, subjective norms, and perceived behavioral control) can be influenced to result in improved intention to be ethical and, ultimately, ethical behavior. We find that the TPB predictor variables mediate the relationship between moral disengagement and ethical intentions. In our expanded model, we also find that moral disengagement influences the

TPB variables which in turn influence ethical intentions. We also find that ethics training and religiosity reduce an individual’s propensity to morally disengage. Based on our findings, we

recommend that moral disengagement be combined with the TPB when the TPB is used to

model ethical behavior.

One of the contributions of our study is expanding the basic TPB theoretical model used

in studying ethical behavior. We combine the TPB, moral disengagement, and both PBC

constructs—PBC locus of control and PBC self-efficacy. We test the model across four different

ethics scenarios. We use international data that we gather from practicing accountants that, on

average, have a relatively high level of experience with ethics training and ethics expectations.

Furthermore, we contribute by identifying factors that indirectly influence ethical intentions by

influencing components of this expanded model.

We also expect other constructs in our model, namely the extent of professional work

experience and having formal professional ethics training, to reduce unethical intentions through

influencing the TPB predictor variable of perceived behavioral control (PBC). Unlike other TPB

ethics studies, we operationalize PBC with two variables, PBC locus of control and PBC self-

efficacy. We base our inclusion of the two PBC variables on work by Ajzen (2002). In that

study, Ajzen concluded that PBC has these two subcomponents, one characterized by self-

efficacy and one by locus of control. We find that the extent of professional experience

positively influences PBC locus of control and PBC self-efficacy. Likewise, we find some support that principles-based professional ethics training positively influences PBC self-efficacy.

Consequently, we suggest that this refinement in the application of the TPB model for ethics studies will more closely capture the intent to behave ethically. This increase in precision may reduce problems observed in some prior studies wherein PBC seemingly has no effect.

We expect that accountants with recent principles-based professional ethics training will

have lower moral disengagement1. We further postulate that beliefs and adherence to religious practices should generally lead a person to reduce moral disengagement. In considering these two variables together—religiosity and recent professional ethics training—we expect that recent professional ethics training will amplify the effect of religiosity on moral disengagement.

Accordingly, we investigate whether religiosity’s interaction with recent professional ethics training reduces moral disengagement. We find support for this expected relationship.

Understanding how recent professional ethics training improves ethical behavioral intentions is complex and indirect. We find that recent professional ethics training reduces an individual’s moral disengagement. Reducing moral disengagement leads to improved attitudes, subjective norms, and perceptions of control over behaving ethically. All these factors in turn influence ethical behavioral intentions. Furthermore, when there has been recent professional ethics training and the particular person experiencing it considers herself or himself religious, the recency effect is even stronger. Finally, experience, including time being a member of a professional accounting association such as the Association of Chartered Certified Accountants

(ACCA) or the Institute of Chartered Accountants in England and Wales (ICAEW), leads to an

improved sense that one has some say over being ethical (PBC locus of control), as well as a

belief that one can be ethical (PBC self-efficacy). Ultimately, the result is improved ethical intentions.2 These findings have bearing on decisions about requiring repetition of ethics

1 We refer to "principles-based professional ethics training” as “professional ethics training” in the remainder of this study. 2 While we would like to be able to study actual unethical behavior across a wide spectrum of real accounting situations, the details of most unethical behavior is not made public. This is a limitation of our study; however, our reliance on the TPB counters this weakness to some extent as explained further in the text.

training, the type of ethics training to invest in, and the value that should be placed on

participating in professional accounting organizations.

Accountants regularly work with confidential financial and nonfinancial personal and

company information. Stakeholders must have confidence that accountants will act in their

responsibilities with strict adherence to professional codes of conduct that include professional

values, attitudes, and ethics (IAESB, 2015; Jakubowsdki et al. 2002). Consequently, standards

for ethical conduct in the accounting professional are extensive (see, for example IESBA 2018).

The accounting profession relies on the accountant’s ability to act ethically, which increases the

public’s trust and safeguards the profession’s reputation. The accounting profession uses

education, training, rules, and examinations to engender ethical behavior. Ethical lapses occur through blatant acts of fraud and intentionally misstating financial statements or tax returns and

less observable actions such as those that violate client confidentiality, engage in collusive behavior, result in dishonest communications with clients, or pad expenses for personal gain. The focus of this paper is to create a model that explains the components of ethical behavior and evaluates the constructs that influence the commitment to act ethically.

While many accountants anticipate they will encounter ethical dilemmas during their

careers, most do not expect that they will have ethical lapses (Moore et al. 2006). However,

ethical lapses do occur. Detert et al. instructed that “if organizations knew more about whether

some individuals were more predisposed to moral disengagement than others, perhaps they could

target resources toward improving these individuals’ decision making processes” (p. 374).

Consequently, having a theory that predicts ethical behavior will allow for preemptive and

targeted interventions resulting in increased ethical behavior.

2. The Theory of Planned Behavior and Moral Disengagement

Theory of Planned Behavior

The Theory of Planned Behavior (TPB), formulated by Ajzen (1991), grew from the

Theory of Reasoned Action (TRA) proposed by Fishbein and Ajzen (1975). Both the TRA and

the TPB posit that behavioral intention is a determinant of actual behavior and that intention is

formed by the attitude towards the behavior (an individual’s favorable or unfavorable view

towards performing a behavior) and subjective norms (perceived social pressure or expectations

of others) (Fishbein and Ajzen, 1975; Ajzen, 1991). Ajzen’s 1991 theory added an additional

measure to the TRA, perceived behavioral control (PBC), that is also theorized to mediate

intention, to account for behaviors that individuals feel that they may not fully control (how

much control individuals perceive that they have concerning the ease of or difficulty in

performing the behavior). Taken together, the TPB combines an individual’s favorable or

unfavorable attitude towards performing the behavior, subjective norms, and perceived

behavioral control to form behavioral intentions which then lead to actual behavior (Ajzen,

1991; Ajzen, 2002).

The introduction of PBC to the TRA sparked significant research interest to understand

how the components of behavioral control influence intentions (for a summary, see Ajzen, 2001

and Armitage and Conner, 2001). At the heart of the research was the issue that a high degree of

perceived behavioral control should increase an individual’s intention and effort to perform the behavior (Ajzen, 2002). From this research, Ajzen (2002) determined that PBC consists of two factors that can be individually measured and assessed—perceived self-efficacy and perceived controllability. Perceived self-efficacy addresses the ease or difficulty an individual perceives in

performing a behavior while perceived controllability is a measure of the perceived extent that the individual controls the performance of the behavior (Ajzen, 2002).

The refined TPB has been shown to be robust in many settings. Recent studies include examining the intention to engage in pro-environmental behaviors (Yuriev et al. 2020), consumer intention to purchase or not purchase ethical products (Sun, 2019), sustainable usage of bike sharing (Si et al. 2020), drone food delivery services (Hwang et al. 2020), and parent-for-child behaviors (Hamilton et al. 2019). Of particular interest to our study is research using the TPB related to ethical behavior especially in the accounting domain, where ethics training and expectations to adhere to ethical behavioral principles and standards are high. Given that the

TPB is used to predict behavior from behavioral intentions, the TBC is particularly well suited to the study of ethics. The internal struggles individuals may have with their desires or commitment to ethical behavior are not normally observed, and even unethical behavior is difficult to identify given the many ways unethical conduct can be performed and unethical behavior’s secretive nature. Consequently, determining ways to evaluate and strengthen the intention to behave ethically is a good proxy for actual behavior.

One of the first studies to use the TPB to address ethical decision making in the accounting profession was conducted by Carpenter and Reimers (2005). Their study examined corporate managers’ decisions to engage in fraudulent financial reporting, which is an ethical choice given that fraudulent financial reporting violates explicit accounting ethics standards. The authors incorporated a series of statements into a questionnaire administered to 73 MBA students designed to measure intention, subjective norms, and PBC. Their analysis showed that attitude had significantly the strongest support for behavioral intention. They also found a positive and significant coefficient for subjective norm but not for PBC. They speculate that the

insignificance of their PBC finding could be due to the hypothetical situation described in the

scenario and that MBA students may not have enough work experience to have encountered

similar situations. The authors conclude that their study provided “strong evidence that the

theory of planned behavior can help explain ethical decision-making by business managers” (p.

124).

Another early study in accounting ethics was conducted by Buchan (2005). In addition to

intention, subjective norms, and PBC influencing attitude, Buchan added a separate measure for

ethical climate developed by Victor and Cullen (1987, 1988) rather than hypothesizing that

ethical climate influences intention through subjective norms. He also included a separate

measure for moral sensitivity using the Multidimensional Ethics Scale (MES) developed by

Reidenbach and Robin (1988, 1990). Buchan obtained responses from 95 professionals working in public accounting firms for four vignettes that described ethical dilemmas faced in public accounting practice. He found strong support for the attitude measure in support of ethical intentions but an insignificant affect from subjective norms. Additionally, he did not obtain an acceptable level of reliability with his measures of PBC which negated his attempt to speak to the relationship between PBC and intention. He attributed the failure to connect PBC and intention to confusion related to the wording of the questions. He also found no support for the hypothesized connection between moral sensitivity and ethical intention or instrumental climate and ethical intention. He noted that a study of ethical decision making in the U.S. metal finishing

industry by Flannery and May (2002) found a marginal effect of instrumental climate when the

construct was part of PBC rather than considered separately as Buchan had done.

Cohen et al. (2010) combined the theory of the fraud triangle with the TPB to examine

managers’ unethical behavior in documented fraud cases. The theory of the fraud triangle posits

that fraud is a function of attitudes/rationalizations, incentives, and opportunities. The authors

suggested that the concept of attitudes in the theory of the fraud triangle is a broad concept that encompasses the three dimensions of the TPB as well the concept of moral obligation. They also stated that “the second and third components of the fraud triangle, the ‘incentives/pressures’ and

‘opportunities’ are not covered by the TPB because they represent external stimuli for the fraud behavior” (p. 276). The purpose of their study was to focus on behavior at the individual level and inform auditing standard setters to help auditors better look for fraud in audits. Cohen and his coauthors stated that incentives/pressures and opportunities are more easily observed than attitudes, which tends to be more internally focused. Consequently, their focus was on attitudes and they do not have separate measures for subjective norms or PBC. The authors examined press coverage of high-profile corporate frauds and categorized information into the three areas of the fraud triangle. The study found that incentives/pressures and opportunities were present in all cases of reported frauds while attitude/rationalization was present in some of the press- covered cases. They conclude that “auditors should better integrate the attitudes component when evaluating the potential for unethical behavior” (p. 287) and that “the TPB, is a useful framework for analyzing unethical behavior by managers that are associated with corporate fraud” (p. 272).

The TPB has also been used to examine whistle-blowing intentions concerning corporate and individual wrongdoings connected to financial fraud. The ethical implication is that a person observing unethical behavior has the choice to remain complicity silent or report the wrongdoing

(Brown et al. 2016). Brown et al. (2016) combined the theory of the fraud triangle and TPB to predict accountants’ whistleblowing intentions. In particular, they link the external factors of the fraud triangle, incentives with subjective norms and opportunities with PBC, as precursors to

behavioral intention. Incentives in this study are composed of extrinsic (economic and social norm) and intrinsic (moral) components while opportunities comprise environmental characteristics (information asymmetry, reporting channels, and regulations, policies, and procedures). They survey 284 professionals with significant corporate accounting experience and

find strong positive association with accountants’ attitudes and PBC with whistleblowing

intention but not for social norms.

Other whistleblowing studies have shown that external whistleblowing intentions of

public accountants in Barbados are influenced by attitudes, PBC, independence commitment,

personal responsibility for reporting, and personal cost of reporting, and that perceived high

organization support intensified intentions (Alleyne et al. 2018). Mustafida (2020) found that

attitude and subjective norms positively influenced whistleblowing intentions of Indonesian

professionals in fraudulent financial reporting but PBC is not significantly associated due to the

professional’s fear of retaliation.

Another study related to whistleblowing was conducted by Seitz et al. (2020). They

looked at the whistleblowing monetary incentives offered by the Dodd-Frank Wall Street Reform

and Consumer Protection Act. In addition to monetary incentives, the Dodd-Frank Act also

provides whistleblower protections. They posit that both monetary incentives and protections

should increase a whistleblower’s intention to report unethical financial reporting by mediating

attention and social norms (they did not tie features of the Dodd-Frank Act to PBC). Seitz et al.

provided financial professionals with one of two surveys designed to mimic the pre and post

Dodd-Frank Act environment. They found that the Dodd-Frank Act monetary incentives and

increased retaliatory protections offered by the Securities and Exchange Commission did not

change whistleblowing intention to report financial reporting frauds.

Cieslewicz (2016) used a field-based instrument to evaluate 91 professional accounting managers’ behavioral intentions to encourage subordinates to participate in collusive accounting information manipulation. In addition to TPB’s three factors theorized to affect behavior intentions, Cieslewicz added materialism as a factor that influences intention and three economic cultural variables: power distance, collectivism, and uncertainty avoidance. He found a significant association between attitude, PBC and materialism with a weaker but significant association with subjective norms and supervisors’ collusion intentions. Further, he found that power distance is positively associated with materialism and PBC, uncertainty avoidance influences subjective norms and materialism, and that collectivism directly impacts behavioral intention. Given the significance of the cultural variables, Cieslewicz concluded that considering the TPB as “a onesize-fits-all approach for all cultures around the world is not recommended” (p.

104).

Finally, a working paper by Barr-Pulliam (2017) used 265 corporate managers in an experiment to measure whether continuous internal auditing changes managers’ intent and behavior to manipulate earnings. The continuous internal auditing variable is hypothesized to mediate PBC and affect earnings management intention by increasing the degree of difficulty required to perform the earnings manipulation. Barr-Pulliam also attempted to measure the link between intention and behavior directly by having participants report both intent and actual behavior. He found that subjective norms and PBC are statistically and positively associated with intention to manipulate earnings. The link between attitude and intention was not statistically significant. He also found that managers who intend to manipulate earnings actually manipulate earnings, tying intention to behavior.

Many TPB studies of accounting ethics did not identify elements associated with PBC or did not find significance in the relationship between PBC and intention. Carpenter and Reimers

(2005) and Buchan (2005) both found insignificant association while Brown et al. (2016),

Cieslewicz (2016) and Barr-Pulliam (2017) found significant association. Other accounting related TPB studies did not hypothesize PBC relationships (Cohen et al., 2010; Seitz, 2020).

Consequently, a unifying theory derived from TPB accounting ethics research has not emerged.

Moral Disengagement

Albert Bandura proposed the theory of moral disengagement to explain why individuals perform inhumane acts without feeling moral distress (Bandura, 1990a, 1990b, 2002). Moral disengagement is defined as “an individual’s propensity to evoke cognitions which restructure one’s actions to appear less harmful, minimize one’s understanding of responsibility for one’s actions, or attenuate the perception of the distress one causes others” (Moore, 2008 p. 129).

Moral disengagement involves the various strategies (conscious or subconscious) individuals employ to rationalize engaging in unethical action in a way that they can still maintain their view of themselves as ethical persons. Cremer and Vandekerckove (2017) explained that “moral disengagement can thus be seen as a buffer that allows people to free themselves from feeling guilty and uneasy with the idea that they may have violated accepted ethical standards” (p. 442).

Moral disengagement is particularly well-suited to help explain unethical decision making. Moore (2008) stated that individuals who are more likely to morally disengage contribute to corruption in an organization. Detert et al. (2008) found empirical evidence that higher levels of moral disengagement are positively associated with unethical decision making.

Similarly, Johnson and Connelly (2016) learned that individuals with low levels of trait guilt have less moral disengagement which positively affects ethical decision making. Moore et al.

(2012) observed that an individual’s propensity to morally disengage explains a large range of undesirable workplace behaviors, including ethicality. Other studies have also documented that moral disengagement increased unethical behavior (Schweitzer and Hsee 2002; Shalvi et al.

2011; Duffy et al. 2012). This research led other researchers to conclude that moral

disengagement is an important mechanism to understand ethics (Treviño et al. 2006; Treviño et

al. 2014).

Integrating the Theory of Planned Behavior and Moral Disengagement

The TPB and moral disengagement connections proposed by our study focus on the

constructs that influence ethical behavioral intention. Azjen’s (1991) TPB framework specifies

that behavioral intentions are influenced by attitude, subjective norms, and perceived behavioral

control (PBC). In an ethics context, attitudes are how people view their responsibility towards

being ethical. They become manifest when a particular behavior is contemplated. Subjective

norms refer to what participants perceive others who are important to them think about a given

behavior. PBC refers to either perceptions of impediments to engaging in a behavior (locus of control) or a person’s sense as to whether she or he is capable of successfully engaging in a behavior (self-efficacy). In our study, locus of control refers to a participant’s sense of the extent

that external forces control whether they behave ethically (how much choice the participant has),

while self-efficacy refers to a participant’s belief in his or her capacity to behave ethically. We hold that moral disengagement influences attitude, subjective norms, and PBC, which in turn influence ethical intentions. In this way, we combine these concepts into one model. We reason

that moral disengagement will lead to negative attitudes, negative subjective norms, and poor

perceived behavioral control that then lead to unethical intentions. We expect that moral

disengagement does not have a direct relationship with ethical behavioral intentions; rather,

attitude, subjective norms, and perceived behavioral control mediate the effect of moral

disengagement on ethical behavioral intentions. Influences on an individual’s propensity to

morally disengage when faced with ethical dilemmas are thus important to understand as

precursors to ethical intentions.

3. Developing a Commitment to Ethical Behavior In professional accounting, ethical behavior is judged by adherence to a set of prescribed rules and codes of conduct, and individuals engaging in unethical behavior are sanctioned and

disciplined. Accountants can and do lose their licenses and certifications, serve jail terms, and

pay fines for making poor ethical choices. For instance, in the United States, the Sarbanes-Oxley

Act introduced monetary penalties and increased sentences for ethical violations. In a similar

vein, there are efforts in England to increase consequences for ethical breaches (Sikka et al.

2018).

Ethical behavior expectations, in an accounting context, are specific and the desired

ethical practices are uniform, widely accepted, understood, and promulgated in professional

standards. There is little room for ethical preference. The International Ethics Standards Board

for Accountants (IESBA 2018) specified five fundamental ethical principles for accountants:

Integrity, Objectivity, Professional Competence and Due Care, Confidentiality, and Professional

Behavior (p. 18). These principles are designed to shape attitudes concerning ethical conduct

from the beginning of an accountant’s training and carry through the remainder of the

accountant’s career. Fishbein and Ajzen (1975) indicate that there is a strong relationship

between attitude and behavior through the mediating role of intention. We suggest, since the

concept of accounting ethical conduct is continuously reinforced, is well defined with specific

rules and expectations, that accountants’ ethical behavioral attitudes will be highly correlated

with ethical intention. This is consistent with prior studies in accounting utilizing the TPB (see,

for example, Buchan, 2005; Cieslewicz, 2016).

Accountants who have knowledge of the profession’s ethics standards and requirements

understand what others inside and outside of the profession expect of them. This expectation is

summarized by the IESBA (2018) statement concerning professional behavior: “Professional

Behavior [means] to comply with relevant laws and regulations and avoid any conduct that the

professional accountant knows or should know might discredit the profession” (p. 18). This

ethics standard explicitly informs accountants what others expect and is directly related to subjective norms. The results of prior studies linking subjective norms to intention have been mixed, with findings of significant associations (Carpenter and Reimers, 2005; Barr-Pullium,

2017) weak associations (Cieslewicz 2016) and insignificant associations (Buchan, 2005; Brown et al. 2016). In an accounting ethics setting, subjective norms should be unambiguous and well understood. We expect subjective norms to be highly correlated with ethical intention.

A large component of accounting ethics codes of conduct and training focus on instilling in the minds of accountants that they are capable and empowered to act ethically. Thus, the PBC locus of control should be ingrained in the cognition of accountants. While the IESBA (2018) ethics principle Professional Behavior relates to both PBC locus of control and PBC self- efficacy, the IESBA’s Objectivity principle directly states that accountants should “not to compromise professional or business judgments because of bias, conflict of interest or undue influence of others” (p. 18; italics added). The Professional Competence and Due Care principle

addresses PBC self-efficacy by requiring accountants to “Attain and maintain professional

knowledge and skill at the level required to ensure that a client or employing organization

receives competent professional service, based on current technical and professional standards

and relevant legislation; and act diligently and in accordance with applicable technical and

professional standards” (IESBA 2018, p. 18). Accountants should feel impowered when they

have sufficient knowledge and capabilities to perform their responsibilities. We expect both PBC

locus of control and PBC self-efficacy to be highly correlated with ethical behavior.

One of the stated goals of International Education Standard (IES) 4 is for professional

accounting education to instill in accountants a commitment to ethical behavior (IAESB, 2014).

The standard explains that to foster professional values, ethics and attitudes should occur through

learning and development throughout an accountant’s career. Jacques (2003) indicated that ethics

education should focus on methods and means that combat systems that engender and encourage

ethical violations. The need is to provide ethics training that teaches accountants how to identify

ethical issues and appropriately respond to those issues. Accordingly, ethics training needs to be

more than rules-based memorization.

In response to the recognized need to provide more professional ethics education, the

International Accounting Education Standards Board (IAESB) commissioned a study to inform the standards setting process. As a result of the study, the IAESB published an Information Paper

(IP) in 2006 titled, “Approaches to the Development and Maintenance of Professional Values,

Ethics and Attitudes in Accounting Education Programs.” The IP introduced the Ethics

Education Framework (EEF) as a model for ethics education and learning. The EEF was developed from work reported by Cooper et al. (2008) that surveyed International Federation of

Accountants (IFAC) members and conducted 113 individual and focus group interviews. The IP

specifies that for a professional accountant to develop ethical commitment, ethics education

needs to progress through a four-stage continuum (IAESB, 2006):

Stage 1—Ethics Knowledge

Stage 2—Ethical Sensitivity

Stage 3—Ethical Judgment

Stage 4—Ethical Behavior

Stage 1, Enhancing Ethics Knowledge, recognizes that individuals entering the

accounting profession come with personal ethical perceptions and values developed through

family beliefs, religious affiliations, education, and interactions with peers and society. Because

the influences that shape ethical perceptions and attitudes are diverse, Stage 1 education

generally attempts to create ethical awareness in the accounting/business context and establish

expectations. It is during this stage in professional accounting education that professional codes

of conduct are introduced. As such, this level of education is predominantly rules-based.

Individuals usually receive this level of education in pre-certification programs such as

university courses. However, rules-based training is also the focus of many ethics continuing

professional development courses required to obtain or renew CPA licenses in the United

States.3

Stage 2 focuses on helping accounting professionals recognize ethical situations. This

stage builds on Stage 1 training by going beyond merely knowing and understanding the

parameters of ethics to being able to identify situations within those parameters and perceive the

3 For instance, the ethics course required by many US state societies, Professional Ethics: The American Institute of Certified Public Accountants' Comprehensive Course, focuses on rules-based learning objectives: • Apply the revised AICPA Code of Professional Conduct. • Recognize the importance and organization of the AICPA Code of Professional Conduct. • Apply AICPA ethics interpretations on integrity, objectivity, confidentiality, and advertising. • Recognize the SEC, GAO and other regulatory rules. Recall the obligations to exercise due care (AICPA 2019). 16

relevant details of the individuals, contexts, and events or practices involved. Most people

entering the accounting profession have been exposed to Stage 1 and Stage 2 training by the time

they begin their accounting careers.

A newer professional accountant’s efforts to raise concerns about ethical practices require

considerable confidence and all the skills and motivation that training can impart. Learning rules is necessary but not sufficient. Education in Stages 2, 3 and 4 should be the focus of continuing professional development and employer sponsored ethics training. This training should incorporate professional ethics education to accomplish the goal of committing accountants to choose the ethical path. The goal of this training is to help professionals use their understanding of ethics and professional codes of conduct in making the right choices, as well as to encourage commitment to ethical conduct in themselves as well as others.4

People progress through the stages of ethics education as they gain more experience and exposure to real-world issues. In theory, the EFF is a continuum that reinforces ethical knowledge, sensitivity, judgment, and commitment. As an individual progresses through and repeats the continuum, the process strengthens and enhances a person’s commitment to ethical behavior to a level that a person not only behaves ethically, but also advocates ethical behavior in others. As professionals advance into supervisory roles and become more distanced in time from their formal education, their decisions have increasingly stronger impacts on financial reports, their organizations, and also on those they mentor (Cieslewicz, 2016). It is therefore critical that their commitment to ethical behavior strengthens rather than degrades. Professional

4 The principles-based approach has been adopted by the ICAEW and the ACCA. For example, the ICAEW’s ethics course objectives are stated, “Ethics is more than just knowing the rules around confidentiality, integrity, objectivity and independence. As an ICAEW Business and Finance Professional you will be expected to be able to identify ethical dilemmas, understand the implications and behave appropriately.” The ICAEW further states that the course is designed to “help you get to grips with our ethical framework and practise how you would respond to an ethical situation (ICAEW 2019).” 17

ethics education and training strengthens a person’s resolve and confidence to act ethically,

which are two important constructs of the TPB that lead to ethical intentions. Ethics training

should also imbue in accountants a commitment to act ethically that will reduce the inclination to

morally disengage.

4. Hypotheses Development

Consistent with Ajzen’s (1991) theory that behavioral intentions are influenced by

attitudes, subjective norms, and perceived behavioral control, we hypothesize that attitudes

towards being ethical and subjective norms for being ethical are associated with intentions to

behave ethically. Perceived behavioral control is also one of the TPB predictors of behavioral

intentions. Ajzen (1991) observed, “Perceived behavioral control may not be particularly

realistic when a person has relatively little information about the behavior, when requirements or

available resources have changed, or when new and unfamiliar elements have entered into the

situation. Under those conditions, a measure of perceived behavioral control may add little to

accuracy of behavioral prediction” (pp 184-185). The TPB model becomes more robust when it

is adapted to specific behaviors (Ajzen, 2006; Cieslewicz, 2016). We posit that ethical expectations in the accounting profession are well known and have seen little change over time.

Therefore, PBC should a salient construct and also associated with intentions to behave ethically.

Ajzen (2002) explained that, depending on the type of behavior being studied, PBC could

be characterized more by an element of locus of control, or of self-efficacy. As ethical behaviors

are usually constrained by elements more outside of one’s control, locus of control should be

relevant. Likewise, as ethical behaviors depend on an individual feeling empowered, self-

efficacy should be relevant. Accordingly, both PBC locus of control and PBC self-efficacy

should influence intentions to be ethical. We first test our TPB model as we develop a unified

theory for evaluating ethics in accounting. In summary, with regards to the primary TPB

variables, we hypothesize the following:

HYPOTHESIS 1a. An attitude towards being ethical will positively influence intentions to be ethical.

HYPOTHESIS 1b. Subjective norms for being ethical will positively influence intentions to be ethical.

HYPOTHESIS 1c. PBC locus of control will positively influence intentions to be ethical.

HYPOTHESIS 1d. PBC self-efficacy will positively influence intentions to be ethical.

The TPB can be narrowly adapted to specific behaviors. In such cases, specific, context-

relevant variables are added to the model to explain the behavior. As the scope of the behavior

becomes less granular and more general, the variables used to model the behavior become less

specific and more general. This generalization continues until at the highest level, only the core,

generic model remains. However, the model can be applied at many different levels of

specificity. In our case, we are concerned about general ethical accounting behavior, but we

realize that the TPB is more accurate when it is applied to specific behaviors. To address this, we

test our higher level, theoretical model (see Figure 1) using four different specific ethics

scenarios. This allows us to gain some perspective on the usefulness of the theoretical model

across ethics scenarios. For ethical behaviors, we postulate that the core TPB variables are

influenced by moral disengagement. If someone disengages ethically, their ethical stance should be manifest in their attitudes towards behaving ethically, their perceptions of what others who are important to them think is acceptable behavior (subjective norms), their sense for whether or not they have control over behaving ethically, and their self-efficacy to be ethical. These in turn

should impact their ethical intentions. Accordingly, ethical engagement, or disengagement, is

critical to ethical behavior.

The construct of moral disengagement (Moore et al. 2012) captures this dimension. This

construct is more general in nature than the more specific TPB variables that are tailored to

specific behaviors. Moral disengagement as a construct is a gauge of the general moral state of a person that we expect will influence how a person will respond to specific ethical situations and inform their attitudes, subjective norms, and components of perceived behavioral control. Moral disengagement in an ethics context has direct applicability. Bandura’s (1986) theory suggested

that moral disengagement happens through eight different mechanisms which he classified into

three groups: 1) moral justification, euphemistic labeling, advantageous comparison; 2)

displacement of responsibility, diffusion of responsibility; and 3) disregarding or distorting the

consequences, dehumanization, and attribution of blame.

The first group involves mechanisms that lead to cognitive restructuring of reprehensible

behavior in a way that increases its moral acceptability (Bandura, 1986). Cognitive restructuring

repositions the behavior to seem somehow beneficial rather than despicable. These mechanisms

relate to the TPB attitude construct—an individual’s favorable or unfavorable view towards

acting ethically. Moral justification reframes an unethical act into an acceptable act by the

cognitive manipulation that the action is for the greater good of the company or even for the

individual’s private life. In accounting, this can occur by rationalizing that disclosing

confidential client information is necessary to enhance relationships with other clients and profits benefiting the company as well as helping the individual progress (Buchan, 2005). Euphemistic labeling changes the way an individual categorizes an act into a cognitive category that is

acceptable or even desirable. An accountant may self-justify fraudulent behavior to maintain the

20 accountant’s high standard of living or to protect the accountant’s reputation (Cohen et al. 2010).

Advantageous comparison pits the contemplated action against a more harmful activity to cast the action in a better light. An example is an accountant who understates expenses to meet income targets and therefore receive a promised bonus (Carpenter and Reimers, 2005, Barr-

Pulliam, 2017). Moral disengagement through cognitive restructuring will negatively impact the attitude construct of the TPB. We hypothesize the following:

HYPOTHESIS 2a. Moral disengagement will negatively influence attitudes towards being ethical.

The second grouping of Bandura’s mechanisms, displacement of responsibility and diffusion of responsibility are “cognitions which displace responsibility [that] tend to attribute the responsibility for one’s actions onto authority figures, who may have tacitly condoned or explicitly directed one’s behavior” (Moore, 2008 p. 130). Theses mechanisms relate to how much an individual perceives there is social pressure to perform an ethical behavior. This perceived social pressure is related to TPB’s subjective norms—perceptions that the individual has regarding how others who are important to them think about a given behavior. The individual can avoid feelings of social norm pressures through displacement of responsibility and diffusion of responsibility. For example, an accountant may displace responsibility for an action when remembering that a respected former business professor maked flippant comments during the student’s formative education about the need to strictly following Generally Accepted

Accounting Principles (GAAP) when faced with the decision to follow or violate GAAP, remembering that the trusted professor didn’t think it was important (Carpenter and Reimers,

2005). Diffusion of responsibility occurs when an accountant feels heavy influence to commit an

unethical act from others in leadership roles in the organization (Cohen et al. 2010). Both of

these mechanisms can be influenced by whether the accountant feels social pressure from

decision-makers in positions of power (Barr-Pulliam, 2007). Moral disengagement through

cognitive restructuring will negatively impact the subjective norm construct of TPB. We

hypothesize the following:

HYPOTHESIS 2b. Moral disengagement will negatively influence subjective norms for being ethical.

Ajzen (2002) explains that there are two components to PBC. Perceived self-efficacy

addresses the ease or difficulty an individual perceives in performing a behavior while perceived

controllability (locus of control) is a measure of the extent the individual controls the

performance of the behavior. The last group of Bandera’s mechanisms—disregarding or

distorting the consequences, dehumanization, and attribution of blame—suggests that individuals

morally disengage when they reframe the consequences of their actions to minimize their

perception of the distress their actions cause others (Moore, 2008). This moral disengagement

occurs when individuals separate the cognitive perceptions of their acts framed by locus of

control or self-efficacy. Accountants may disregard or distort the consequences of their action by cognitively shifting responsibility onto others because they have control (self-efficacy) over how the act is framed or feel they that they do not have control over the act (locus of control) because someone in authority is directing or expecting them to commit the act. For example, in the case of recording transactions that violate GAAP and will thus have negative consequences on those who rely on the financial statements, a manager may perceive that they have control over the choice or do not have control over the choice (Carpenter and Reimers, 2005). In this situation, both distortion of consequences and dehumanization facilitate moral disengagement

22 because it is easy for the manager to view the act as not harming anyone because violating

GAAP impacts companies who are faceless, have deep pockets, and as a collective group, nameless. Distortion can also arise when managers of positively publicized companies believe the accolades received in the press and will morally disengage to keep up the favorable image

(Cohen et al. 2010). Finally, attribution of blame is related to locus of control because it defers responsibility for the act. In whistleblowing situations, an individual who feels that reporting a wrongdoing will be ignored by the organization is not likely to become a whistleblower and can blame the lack of making a report on the inaction of management and attribute blame to management for the individual’s silent duplicity. Weak organizational polices, perceived hindrances, and company culture as well as fear of reprisals for whistleblowing all affect locus of control (Brown, 2016) as do established internal controls and monitoring of one’s work (Barr-

Pulliam, 2017). Moral disengagement through cognitive restructuring will negatively impact the

PBC construct of TPB. We hypothesize the following:

HYPOTHESIS 2c. Moral disengagement will negatively influence PBC locus of control.

HYPOTHESIS 2d. Moral disengagement will negatively influence PBC self-efficacy.

Disengaging morally is likely to erode attitudes, subjective norms, and PBC through

Bandera’s eight mechanisms, which in turn will lead to intentions to behave unethically.

Likewise, not disengaging morally should lead to improved attitudes, subjective norms, and perceptions of behavioral control, which in turn will lead to intentions to behave ethically. Moral disengagement can change people’s perceptions of not only what they should and can do

(attitude), but also beliefs about what others think is appropriate (subjective norms). This occurs in part through social projection, which suggests that people tend to assume others are similar to

23 themselves (Robbins and Krueger, 2005). The effect of moral disengagement on intentions to be ethical is expected to happen through the effect moral disengagement has on the traditional TPB predictor variables. This mediated relationship is hypothesized as follows:

HYPOTHESIS 3. The influence of moral disengagement on intentions to be ethical will be mediated. Moral disengagement will influence (a) attitudes towards being ethical, (b) subjective norms for being ethical, (c) PBC locus of control, and (d), PBC self- efficacy, which will all in turn influence intentions to be ethical.

If moral disengagement can be limited, the ultimate impact on intentions to be ethical should improve. We are, therefore, interested in factors related to ethical behavior that reduce moral disengagement. Many studies have looked at factors that lead to moral disengagement including, goal setting (Barsky, 2011), trait guilt and shame (Johnson and Connelly, 2016), anger and hostility (Shen et al. 2019), abusive leadership (Valle et al. 2019), and creativity (Zheng et al. 2019). These trait factors can fit within the general definitions of the antecedents to moral disengagement described by Detert et al. (2008); empathy, trait cynicism, chance locus of control, and moral identity. An important consideration of antecedents to moral disengagement is that they are “malleable propert[ies] of individuals, a result of the continued reciprocal influences of the individual on his or her environment and vice versa” (Moore 2008, p. 131). Consequently, an effective way to reduce moral disengagement is to change an individual’s propensity to morally disengage.

Detert et al. (2008) applied moral disengagement specifically to ethical decision making.

They found that individual differences in empathy, trait cynicism, locus of control, and moral identity are antecedents to moral disengagement and that moral disengagement predicts unethical

decision making. Additionally, they found that moral disengagement mediates the relationships

between unethical decisions and these individual differences.

Antecedents to moral disengagement that moderate ethical behavior will be related to

empathy, trait cynicism, chance locus of control, and moral identity. Empathy is the way

individuals notice and have feelings of concern for others (Detert et al. 2008). Ethics training is

designed to promote concern for others. In accounting ethics training, a fundamental principle is safeguarding the profession’s reputation by being ethical in fact and in appearance. Stanga and

Turpen (1991) asserted that “accountants have special obligations to behave ethically because of the considerable trust that investors and creditors place in them” (p. 740). Consequently, professional accounting ethics training puts ethical conduct in the frame of not hurting others through unethical conduct. This leads to empathetic development. McPhail (2001) emphasized

the need for training in empathy with this warning: “accountancy has become dangerously dehumanized and that one of the most important objectives for any business ethics education must be to develop an empathy with ‘the other’” (p. 279). Developing empathy is a goal of EEF

Stage 2—Ethical Sensitivity training and applying empathy is part of Stage 3—Ethical Judgment

and Stage-4—Ethical Behavior.

Trait cynicism manifests itself when an individual feels disillusioned and lacks trust in

other people, social conventions, ideologies, and organizations (Detert et al. 2008). One way to

overcome trait cynicism is through moral meaningfulness. May et al. (2014) define moral

meaningfulness in terms of ethical behavior as “the value of ethics in one’s work life” (p. 70).

Moral meaningfulness becomes salient when individuals understand ethical conduct and

reasoning sufficiently that behaving ethically is meaningful and fulfilling (May et al. 2014).

Professional ethics training strengthens the view that accountants belong to a trustworthy and

moral profession, concerned for the public well-being and good. It adds meaningfulness to ethical behavior and reduces trait cynicism. Understanding the need and benefit of ethical

behavior is part of EEF Stage 1—Ethics Knowledge and becomes internalized in Stage 4—

Ethical Behavior.

Locus of control orientations as antecedents to moral disengagement are somewhat

different than PBC locus of control orientations. Locus of control as related to moral

disengagement is defined as how individuals consider and view the events in their lives (Detert

et al. 2008). At the heart of locus of control is whether individuals believe that they control

outcomes or whether outcomes happen by chance or are controlled by others (Detert et al. 2008).

Higher internal locus of control gives people the belief that their behavior and outcomes are

connected. One of the purposes of ethics training is to help people connect their actions to ethical

outcomes. Instilling the ethical belief that an accountant is in control of ethical outcomes is part

of Stage 1—Ethics Knowledge training and is a constant theme in Stages 2-4.

Moral identity is a well-studied antecedent to ethical behavior. It is a factor that relates to

the way individuals think about themselves (Detert et al. 2008). Aquino et al. (2007) suggested

that individuals with high moral identity will be concerned with the well-being of others—even

those external to an organization—and be less likely to morally disengage. Ethics training helps

individuals believe that they can be ethical and embrace ethical values. At the Stage 1—Ethics

Knowledge level, accountants are taught about professional ethics standards, codes of conduct,

and expectations. The goal is for accountants to gain a fundamental understanding of

professional values, attitudes, and ethics sufficient to equip them “with the intellectual

underpinning to appreciate the requirements and expectations in different situations, professional

and otherwise” (IAESB, 2006 p. 97). As accountants progress through the four-stage continuum

26 and learn about professional ethics standards and expectations (Stage 1), become equipped to recognize ethical issues (Stage 2), apply the proper responses to ethical issues (Stage 3) and finally act on principles rather than just believe in them (Stage 4), ethical behavior becomes ingrained and part of their moral identity.

Moore (2008) suggested that moral disengagement is malleable and is developed over a period of time and “would be amenable to intervention or learning” (p. 131). This means that the propensity to morally disengage can change over time from social norms, interactions with others, experiences, and training. Additionally, Moore (2008) stated that moral disengagement is context-dependent, meaning that individuals “apply different set[s] of ethical standards in different contexts (p. 131).

The accounting ethics continuum relies on continual training to reinforce the commitment to and expectation of ethical behavior. The International Code of Ethics for Professional

Accountants by International Ethics and Standards Board of Accountants (IESBA) instructs that accountants should comply with five fundamental principles of professional ethics: integrity, objectivity, professional competence and due care, confidentiality and professional behavior

(IESBA 2019). In a professional context, a means to this end is making sure that regular ethics training occurs. Without this effort, people forget, get distracted, or get caught up in what is easier or seemingly more advantageous. With this reasoning, we expect that recency of professional ethical training will reduce moral disengagement.

HYPOTHESIS 4a. Recency of professional ethics training will be associated with lower moral disengagement.

We expect that recent professional ethics training and religiosity will be positively associated with low moral disengagement, which will help strengthen ethical behavioral

intentions. The salience of ethics training can grow stale over time, necessitating regular training.

Religiosity is broadly defined as having a religious orientation and involvement. Kohlberg

(1981) argues that religion addresses the Why be moral? question and is positively associated

with moral development. Religiosity has been shown to have a strong impact on work values

(Parboteeah et al., 2009), which affect attitudes and behavior (Ramasamy et al., 2010). Graafland

(2017) found a positive relationship between religiosity, attitude, and subjective norms when

consumers considered socially responsible products. We posit that the interaction of recent

professional ethics training and religiosity will directly reduce moral disengagement which in

turn will directly influences attitude, subjective norms, and PBC, leading to improved ethical

intentions.

In theory, professionals who consider themselves to be religious, regardless of any

religion they may belong to, should generally exhibit moral behavior. Religiosity is broadly

defined as having a religious orientation and involvement. Kohlberg (1981) argued that religion

addresses the Why be moral? question and is positively associated with moral development.

Religiosity has been shown to have a strong impact on work values (Parboteeah et al., 2009),

which affect attitudes and behavior (Ramasamy et al., 2010). A study by Bloodgood et al. (2008)

found that students who frequently attended worship services were less likely to cheat than those who rarely attended religious services. However, focusing on the students who attended religious

services less frequently, those who also had classroom instruction on ethics exhibited a lower

likelihood of cheating. The authors suggest that the observed change from the less religious

group is because they had more “room to grow.” Omer et al. (2018) found that auditors working

in offices located in highly religious U.S. Metropolitan Statistical Areas issue more going

concern opinions (demonstrating adherence to professional standards and more skepticism about

the client) than those located in lower religious areas. Nelson (2009) argues that applying

professional skepticism is an indication of ethical predisposition.

In our case, consistent with the religiosity construct, we focus on whether professionals

consider themselves to be religious, without regard to actual attendance at religious services. We

ask, “How important is religion in your life?” and “Regardless of whether you belong to a

particular religion, how religious would you say you are?” Our expectation is that higher

religiosity will reduce moral disengagement.

HYPOTHESIS 4b: Higher religiosity will be associated with lower moral disengagement.

Elements of religiosity are also imbedded in empathy, trait cynicism, chance locus of

control, and moral identity. Religiosity fosters empathy, teaches people to be concerned for the

welfare of their neighbors, instills the idea that individuals are responsible for and control their

own behavior and that they are offspring of deity and should view themselves as capable of doing good. Additionally, religiosity establishes the expectation that a person should treat others, and themselves, fairly.

As with higher religiosity, we expect that recency of professional ethics training will be associated with lower moral disengagement. This is consistent with a COSO Internal Control –

Integrated Framework premise that regular communication to employees that ethical behavior is

expected will improve the control environment (COSO, 2013). We also expect an interaction

between religiosity and recency of professional ethics training because they are both antecedents

to ethical behavior encompass empathy, trait cynicism, chance locus of control, and moral identity. We reason that recency of professional ethics training should activate religiosity,

whereas the lack of recency of professional ethics training will not activate religiosity.

HYPOTHESIS 4c: Recency of professional ethics training will interact with higher religiosity and result in lower moral disengagement.

Anecdotally, professional ethics training may not change all people’s attitudes about

behaving ethically. It also may not change their perceptions about what others think about ethical behavior. However, receiving professional ethics training should give people a sense of empowerment to be ethical because it communicates what is expected. Professional ethics training should influence professionals’ perceptions of their self-efficacy to behave ethically.

This is hypothesized as follows:

HYPOTHESIS 5. Professional ethics training will increase Perceived Behavioral Control (PBC) self-efficacy.

Prior research indicates that an individual’s sense of control influences their ethical

intentions and behavior (Cieslewicz 2016; Webb and Sheeran 2006). To improve ethics, it is

helpful to understand what influences PBC locus of control. We previously hypothesized that

moral disengagement will influence PBC locus of control. Experience should also influence PBC

locus of control. Dane and Sonenshein (2015) reviewed the contradictory results prior research

has found on experience’s relation to ethical decision making. They found that “experience is not

a unitary construct with straightforward effects” (p. 86). They recommended that researchers not

reduce experience to a single relationship with ethical decision making but recognize that

experience takes on different forms and can be complex in the moral domain. Following this

guidance, we suggest that experience, measured by the time spent in the accounting profession

and exposure to codes of conduct and ethical situations, will enhance PBC locus of control rather

than impact intentions directly.

There are several reasons experience should enhance PBC locus of control. Stated in

terms of the International Accounting Education Standards Board’s EEF discussed earlier,

individuals progress through a continuum that reinforces ethical knowledge, sensitivity,

judgment, and commitment. With additional experience comes anecdotal confirmation that

people get caught when engaging in unethical behavior. Additional experience should lead to a

sense that unethical behavior is being watched for and may be hard to get away with, as well as

the realization that it is possible to work through and resolve ethical dilemmas without resorting

to unethical behavior. Our measure of experience combines age and time as a member of the

ACCA or ICAEW. As the length of time increases in which one has been a member of the

ACCA or ICAEW, a sense of pride in professionalism should develop. Experience should

increase one’s perception that he or she can succeed in being ethical. Likewise, experience

should increase the perception that behaving ethically is very much within one’s control.

Experience should thus impact both the self-efficacy and locus of control elements of PBC, as follows:

HYPOTHESIS 6a. Additional experience will increase PBC locus of control.

HYPOTHESIS 6b. Additional experience will increase PBC self-efficacy.

We expect that we can incorporate the above hypotheses into an integrated model that can explain many aspects of ethical behavior. Testing the integrated model with a culture-rich, international dataset validates the model and provides a way to understand ethical behavior and effects of ethics training internationally. We incorporate the above hypothesized relationships into such a theoretical model (see Figure 1). The theoretical model seeks to advance our

understanding of ethical behavior, linking moral disengagement with the TPB. It also illustrates

what ethics training should be expected to accomplish. Professional ethics training should

enhance professionals’ PBC self-efficacy that they can behave ethically, while recency of professional ethics training should act to reduce moral disengagement. Recency of Professional

Ethics Training should also interact with religiosity and activate it to reduce moral

disengagement. Experience, which is a composite measure incorporating both age and time as a

member in professional organizations, should increase both PBC locus of control and PBC self-

efficacy to behave ethically.

Insert Figure 1

5. Research Method

Sample

The survey data were gathered from accounting professionals who were undergoing or

who had finished their accounting certification/qualifying process. Two professional accounting

organizations sent invitations on our behalf to participate in our study to approximately 105,000

individuals. Not everyone who started the surveys (3,143) completed the surveys (1,692).

Typically, we find that surveys that require just a few minutes of time are more likely to be

completed, whereas this survey required closer to 20 minutes of time. We were not able to

identify any response bias other than the survey took too much time. While the sample size is

adequate for statistical analysis and is inclusive of many parts of the world, the usable response

rate was 1.61% of the individuals invited to participate (105,000), with the final sample being n

= 1692 (mean age = 33; 44 percent women)5. The survey was designed to include a number of

5 The organizations sending the invitation to participate in our study indicated that they typically have less than a 2% response rate to similar survey requests.

different instruments related to ethics training. These instruments were randomly assigned to

participants so that only a subset of the 1,692 participants received both the TPB and the moral

disengagement instruments. The individuals who were randomly assigned both the TPB and moral disengagement instruments and who completed the survey was n = 270 (see Table 1 for descriptive statistics). The information reported in this study is derived from this subset of 270 responses.

Table 1 shows the descriptive statistics for our subset sample (n = 270). The majority of respondents (95.2 percent) had completed a university degree or professional qualification and

49.6 percent reported that their degree was specifically in accounting.6 More than 69 percent

were currently employed in a full-time accounting position. In our sample, 38 percent reported

having a professional ethics course as part of their accounting certification/qualifying process; 76

percent had ethics training in some other way; and 38 percent had ethics training within 18

months of the survey. Our survey participants were from the following geographic areas

identified in the Global Leadership and Organizational Behavior Effectiveness (GLOBE) work

(House et al., 2004): 14 percent African, 39.5 percent Anglo, 9.7 percent Confucian, 8.8 percent

Eastern European, 10.3 percent Middle Eastern, and 9 percent Southeast Asian.

Insert Table 1

Procedures Each person who consented to complete a survey answered questions about previous ethics training and whether they had completed an ethics module as part of the certification/qualifying process, as well as when the ethics module was completed, if applicable.

6 In many parts of the world an accounting degree is not necessary to become a certified or chartered accountant, e.g., United Kingdom. In these countries an apprenticeship model is the norm.

The subjects then answered the TPB survey and the moral disengagement survey (the order they

received the surveys was randomly determined). The two surveys were designed to be completed

in approximately 20 minutes.

Measures

Professional ethics training is defined based on whether a person received ethics training

as part of his or her testing and preparation to become either a chartered accountant or a certified chartered accountant. It is measured as 1 if the person has received the ethics training and 0 otherwise. In addition, we asked each person that received ethics training from the professional certifying body to give us the length of time since the ethics training had occurred: within the last

18 months (coded 1), or longer (coded 0). Experience is the latent construct resulting from two manifest variables (or items), age and length of time as a member of the ACCA or ICAEW.

Moral disengagement was assessed using the eight-item reduced version (Moore et al.,

2012) of Bandura’s measure of Mechanisms of Moral Disengagement (Bandura et al., 1996).

Moral disengagement occurs when people abandon their moral responsibilities by convincing

themselves that ethics standards do not apply in a particular situation. When people morally

disengage, they can continue to perceive themselves as ethical and just while engaging in

unethical actions.

As will be discussed later, each hypothesis was tested using four different ethics

scenarios. After each scenario, items related to the core TPB variables of attitude, subjective

norms, perceived behavioral control, and ethical intentions were presented. Each of the TPB

variables was measured for each ethics scenario and is unique to that particular behavior. This

allows us to see whether the theoretical model fits different ethics situations.

Items were measured using 7-point Likert scales. Attitude was measured by items such as: “In your opinion, was the behavior (good/bad).” Subjective norms— what participants think others who are important to them think about a given behavior—was measured by items such as,

“Most accounting professionals who are important to me will look down on me if I perform the behavior described in the scenario (unlikely/likely).” PBC items refer to either perceptions of impediments to engaging in a behavior (locus of control), or a person’s sense as to whether or not he or she is capable of successfully engaging in a behavior (self-efficacy) (Ajzen, 2002). An example of an item measuring PBC locus of control is, “How much control do you have over the behavior described in the scenario? (complete control/absolutely no control).” An example of an item measuring PBC self-efficacy is, “For me to perform the behaviour described in the scenario is: (easy/difficult).” Intentions to be ethical was measured using items such as “I may perform the behavior described in the scenario in the future.”

Ethics Scenarios

We tested our hypotheses using four different accounting-relevant ethical behaviors associated with four different scenarios. The scenarios were adapted from earlier studies including Buchan (2005), Claypool et al. (1990), Cohen et al. (1995), and Armstrong (1985).

1 Maintaining Confidentiality

• Scenario: John, a Chartered Accountant, serves as the auditor of Shrinking

Company, a privately held firm. Shrinking's market share has declined drastically,

and John knows that Shrinking will soon be bankrupt. Another of John’s audit

clients is Solid Company. While auditing Solid’s accounts receivable, John finds

that Shrinking Company owes Solid $200,000. This represents 10% of Solid’s

receivables.

• Action: John warns the client, Solid Company, about Shrinking's impending

bankruptcy.

2 Resisting Collusive Supervision

• Scenario: A public accounting firm has recently acquired a new client through a

very low bid. The partner suggests the audit hour budget for inventory-related

items will be 100 hours. The senior’s experience with similar clients suggests that

in order to have reasonable assurances of no material errors or irregularities, the

audit will take a minimum of 150 hours. Performance evaluation is based, in part,

on efficiency.

• Action: The senior accepts the budget as is. Subsequently, in cases where

judgement is required to determine the number of audit procedures, the senior

performs fewer procedures.

3 Being Truthful with Clients

• Scenario: A partner is developing a bid for a new client. The partner deliberately

sets the bid significantly below cost. The partner knows that the audit will lose

money in the first few years. However, the expectation is that the firm will be able

to raise the audit fee a few years down the road to generate a profit.

• Action: In response to a question from the prospective client, the partner indicates

that fees should not be expected to rise significantly in the foreseeable future.

4 Exercising Integrity with Travel Expenses

• Scenario: A supervisor, the mother of two small children, has been promoted and

assigned to an engagement which requires travel away from home for the firm on

a regular basis. Because these trips are frequent and inconvenience her family life,

she is contemplating charging some small personal expenses while traveling for

the firm. She has heard that this is common practise in the firm.

• Action: The supervisor charges the firm $50 for family gifts.

Evaluation of the Survey Instrument

A recognized strength of our methodology in which we survey professionals is the enhancement of external validity. This strength, however, requires that adequate care be taken when creating the survey instrument. We carefully compiled our survey by adapting various components of our instrument from prior studies. For instance, we utilized an existing, tested scale for moral disengagement and used well-tested TPB scenarios that have been used by several published studies. Post hoc, we also evaluated the validity of the survey instrument.

A weakness of survey instruments, or of any research instrument for that matter, is that participants may move through the instrument too quickly. For instance, they may click “7,” “7,”

“7” without carefully reading the related questions. To guard against this, we reverse coded some of the items and then, after the surveys were completed, we mathematically reversed the responses so that all the scales align in the same direction. Our factor analyses indicate that related items load together, and did not load on other constructs, suggesting that participants did

37 take adequate care in completing the survey, and also that the discriminant validity of the constructs is acceptable.

Common method bias may occur with survey instruments through association of predictor variables with criterion variables. Consistent with Podsakoff et al. (2003), Harman's single-factor test was used to evaluate whether a single factor accounted for the majority of the total variation. A single factor did not account for more than 50% of the variance, with all factors accounting for less than 20% of the variance.

6. Results

PLS Path Modeling

We evaluate the hypothesized relationships with PLS path modeling. PLS path modeling permits evaluating theoretical and measurement models simultaneously and has been used extensively outside of accounting as well as in more recent behavioral accounting research. For this study, the primary reason for using PLS path modeling includes its ability to evaluate a complex series of relationships simultaneously, including latent variables, their related constructs, mediation, and path analysis. PLS path modeling is also suitable for smaller sample sizes, as explained by Chin and Newsted (1999).

The composite reliability of the latent variables (e.g., perceived behavioral control) is measured with Dillon-Goldstein’s Rho as recommended for PLS path modeling (Vinzi et al.,

2010). The Dillon-Goldstein’s Rho for each latent variable is at least 0.75, while a cut-off of 0.70 is recommended (Vinzi et al., 2010). The loadings of the items for each latent variable (e.g., loadings of the items measuring attitude on the attitude construct) are all satisfactory and exceed the cutoff of 0.60 (Hair et al., 2010), except for those related to moral disengagement. As moral

disengagement is an established construct, there are several items (8), and the loadings are all

above 0.45 with no cross loading on other constructs, we did not drop any of the items. For

moral disengagement, this also leads to a lower mean communality, but improving it through

dropping items with lower loadings did not significantly change the path coefficients. For the reasons stated above, we did not remove any of the items. For the other constructs, the mean communalities (AVE) are all above 0.5. Based on these measures, the reliability and convergent validity of the latent variables are acceptable. The discriminant validity of the constructs was also evaluated and found to be acceptable as discussed earlier.

We are interested in both the hypothesized relationships related to each specific ethical behavior, as well as how consistent the hypothesized relationships are across the four ethical behaviors. That is, we are interested in whether the relationships are true across different ethics scenarios and may be suggestive of a model with usefulness beyond the specific ethical behaviors in this study. First, we present each model and briefly discuss them in turn. We then tabulate the combined results across all four models for comparative purposes to assess the overall theoretical model (see Table 2 later). These models present two-tailed results, although the hypotheses are directional.

Scenario 1: Maintaining Confidentiality

Figure 2 models the accounting ethical behavior of maintaining confidentiality. We begin at the right side of the model with intentions to maintain confidentiality and then work upstream to the left. Intentions to maintain confidentiality is an instance of behavioral intentions, the dependent variable in the TPB framework. Intentions to maintain confidentiality is influenced directly by attitude towards maintaining confidentiality (0.494, p = 0.000), subjective norms for

maintaining confidentiality (0.319, p = 0.000), PBC locus of control (0.093, p = 0.016), and

PBC self-efficacy (0.066, p = 0.090). These variables that influence intentions to maintain confidentiality are themselves further explained by considering factors that impact them.

Insert Figure 2

The next principal set of relationships in the model involves moral disengagement. In the theoretical model, moral disengagement functions in part as an exogenous or independent variable that has a negative relationship with each of the TPB predictors of behavioral intentions.

When moral disengagement is low, attitude, subjective norms, and perceived behavioral control

(both locus of control and self-efficacy) related to behaving ethically are all high, which further leads to higher intentions to behave ethically. More specifically, attitude towards maintaining confidentiality (-0.254, p = 0.000), subjective norms for maintaining confidentiality (-0.204, p =

0.001), PBC locus of control (-0.108, p = 0.079), and PBC self-efficacy (-0.213, p = 0.000) are all inversely related to moral disengagement. In summary, moral disengagement has a negative relationship with each of the TPB predictors of intentions to maintain confidentiality. This situation is important to understand as it then becomes imperative from an ethics training point of view to understand what can be done to minimize moral disengagement.

Moral disengagement is less likely when there has been recent professional ethics training and religiosity is high. Recent professional ethics training (abbreviated in the figure as

Recent Ethics Training) is negatively related with moral disengagement (-0.335, p = 0.018).

Religiosity is not significantly related to moral disengagement, but it interacts with recent ethics

training. The interaction of religiosity and recent ethics training has a negative relationship with

moral disengagement (-0.291, p = 0.048). Moral disengagement is lowest when someone has

both a high sense of being religious and has recently had ethics training.

Experience is positively related to both PBC locus of control (0.139, p = 0.024) and PBC self-efficacy (0.160, p = 0.012). With more experience (comprised of age and time as a member of a professional accounting association), comes both a greater sense that it is within one’s power to be ethical and that one is capable of being ethical. As mentioned above, PBC locus of control and PBC self-efficacy are both further positively related with intentions to maintain confidentiality.

Professional ethics training has a positive relationship with PBC self-efficacy (0.138, p =

0.027). When a practicing accountant receives ethics training, he or she has a greater belief in her ability to behave ethically. It seems that such training is not just instructive, but empowering.

Taken as a whole, the results across the models are similar, suggesting that the model is

appropriate for international use. Yet there are noteworthy differences to discuss (see Table 2,

presented later, for a summary across all four ethical behaviors). Having walked through each

relationship in the model for maintaining confidentiality, we will now focus on instances within

each of the remaining three models for ethical behaviors that vary materially from the results for

the model for maintaining confidentiality. In doing so, we refer to the respective figures.

Scenario 2: Resisting Collusive Supervision

Figure 3 models the ethical behavior of resisting collusive supervision. This ethics

scenario involves pressure from a supervisor to enter into collusion. While nearly all the

relationships in the model for resisting collusive supervision are significant, two relationships are

not significant that are significant in the model for maintaining confidentiality. First, professional

ethics training does not have a statistically significant relationship with PBC self-efficacy (0.065,

p = 0.283). Experience also does not have a statistically significant relationship with PBC locus

of control (-0.001, p = 0.982). As explained below, we believe this is due to the nature of the

specific ethical behavior that is focused on – resisting collusive supervision.

Insert Figure 3

We obtain an international sample to allow us to evaluate whether our model may be

appropriate for international use. At an international level, however, there are differences that are

difficult to completely accommodate with a one-size-fits-all model (Cieslewicz, 2014).

Considering this particular behavior—resisting collusive supervision, Cieslewicz (2016) suggests

that direct pressure from supervisors activates cultural effects that other ethical circumstances

may not. For instance, in a highly collectivistic society, if it has been deemed that an action is in

the best interest of the collective, it may be presumed that an employee would concur with a

supervisor and support the collective. In a highly individualistic society, an employee might be

more likely to speak out based solely on his or her own prerogative of wishing to comply with professional standards. Both may be motivated by positive aspects of different cultures, yet they could lead to different ethical behaviors. Power distance could also be influential given a subordinate/supervisor relationship is present. Resisting collusive supervision is always daunting because subordinates are often dependent on supervisors for promotions, raises, and the continuance of employment. Additionally, in a high power distance society, there is also an assumption that authority should not be challenged. We draw from this that in preparing ethics training for different areas around the world, it may be helpful to augment or reinforce them for such regional tendencies. We also note that while some regional tendencies make ethical behavior in specific ethical situations more difficult and less likely, other regional tendencies may make ethical behavior more likely. Such advantages should be identified and capitalized on in ethics training.

Our study does not focus on such cultural and other regional differences. We only mention them here to suggest that these differences may explain why certain relationships for certain ethics scenarios are not significant. In our study, we are interested in a model that is as internationally useful as possible. Importantly, for this scenario, while experience and professional ethics training do not have statistically significant relationships with PBC locus of control and PBC self-efficacy, respectively, and thus do not influence intentions to resist collusive supervision, there is another statistically significant pathway within the model that may compensate. The interaction of religiosity and recency of ethics training influences moral disengagement, which further influences PBC locus of control and PBC self-efficacy, which further influence intentions to resist collusive supervision.

Scenario 3: Being Truthful with Clients

Figure 4 models the ethical behavior of being truthful with clients. Most of the relationships are significant in the model. We focus again on those that are not significant. In this model, experience does not have a significant relationship with PBC self-efficacy (-0.026, p

=0.683), whereas this relationship is significant in the other three models. Moral disengagement also does not have a significant relationship with PBC locus of control -0.089, p =0.140). Given one of our goals is to test the overall model in several different scenarios, and this relationship is at least consistent in direction with the others, we do not view this last relationship as a substantial problem with this component of the theoretical model.

Insert Figure 4

Scenario 4: Exercising Integrity with Travel Expenses

As with the model for resisting collusive supervision, two relationships – the same

relationships – are not significant (see Figure 5). First, professional ethics training does not

influence PBC self-efficacy (0.061, p =0.341). Experience also does not influence PBC locus of

control (0.073, p = 0.228). In this case again, the specific behavior described in the scenario may invite cultural influences that outweigh the influence of experience and professional ethics training. In a highly collective society, family is very important. For some, it may be viewed as

unethical to not bring a gift home for the daughter. As with the collusive supervision scenario,

the other pathways (identified earlier) to influence ethical intentions consists of statistically

significant relationships.

Insert Figure 5

Summary of Model Results Across All Four Ethics Scenarios

The results across the four ethics scenarios are summarized in Table 2. In Table 2 (as

well as in Table 3), the various measures of specific ethical behavioral intentions (e.g., intentions

to maintain confidentiality and intentions to resist collusive supervision) are referred to simply as

“Ethical Behavioral Intentions.” While the usefulness of measures of goodness of fit for PLS

path models is questioned (Mahama and Cheng 2013; Vinzi et al. 2010), the absolute goodness

of fit index for the model in Figure 2 is .302, Figure 3 is .295, Figure 4 is .300, and Figure 5 is

.293. Also, although the hypotheses are directional, we report two-tailed results.

Insert Table 2

To summarize, there is substantial support for all of the relationships in the model, except

for the relationships between 1) professional ethics training and PBC self-efficacy, and 2)

experience and PBC locus of control. These two relationships are statistically significant for two

of the four scenarios, indicating partial support for this part of the model. The remainder of the

model is more robust across the ethics scenarios. We now proceed to explain another aspect of

the model focused on how moral disengagement influences ethical intentions.

Mediation

Hypothesis 3 indicates that the influence of moral disengagement on intentions to be

ethical will be mediated. That is, moral disengagement will influence (a) attitudes towards being

ethical, (b) subjective norms for being ethical, (c) PBC locus of control, and (d), PBC self-

efficacy, which will all in turn influence intentions to be ethical. To the extent it is supported,

this is an extension of the TPB insofar as its use with ethical accounting behaviors is concerned.

Baron and Kenny (1986) explain conditions that must be satisfied in testing for

mediation. We illustrate these conditions using the ethical behavior of exercising integrity with

travel expenses. First, the direct relationship between the two variables must be statistically

significant if the relationship is to be mediated. In Table 3, under integrity with travel expenses,

moral disengagement has a statistically significant relationship with ethical behavioral intentions

(-0.281, p = 0.000). The first condition establishing a direct relationship is satisfied. In reference

to Table 3, to be more concise, we have used shortened variable names (e.g., “attitude” instead of

“attitude towards exercising integrity with travel expenses”).

Second, moral disengagement must have a statistically significant relationship with the

mediating variable(s), and the mediating variable(s) must in turn then influence the dependent

variable, ethical behavioral intentions. In Table 3, moral disengagement is related to attitude (-

0.260, p = 0.000), subjective norms (-0.165, p =0.006), and PBC self-efficacy (-0.229, p =

0.000). These three mediating variables then in turn impact ethical behavioral intentions. That is,

attitude (0.317, p = 0.000), subjective norms (0.246, p = 0.000), and PBC self-efficacy (0.328, p

= 0.000) impact ethical behavioral intentions. Each link in these three mediating relationships is statistically significant. The other predicted mediating variable of PBC locus of control has one link in the mediating relationship that is not statistically significant (0.067, p = 0.109). In total for step 2, 7 of the 8 required relationships are significant. This suggests that the effect moral disengagement has on ethical behavioral intentions occurs through the mediation of the TPB predictor variables of attitude, subjective norms, and PBC. A third condition, however, must also be met.

The third condition, that the original direct relationship between moral disengagement and ethical behavioral intentions (-0.281, p =0.000), loses strength when the mediating relationships are introduced into the model, also occurs. The direct relationship ceases to be significant (-0.043, p = 0.316) when the mediating relationships are entered into the model. This

is a case of full mediation as the direct relationship does not just lose strength, it becomes

insignificant.

Insert Table 3

We present results of similar mediation analyses for each of the four ethics scenarios (see

Table 3). In total across the four ethics scenarios, all four of the direct relationships that are

expected to be mediated (i.e., step 1) are significant. All four of these relationships also either

lose strength or become insignificant when the mediating pathways are introduced into the model

(i.e., step 3). For step 2, 29 of 32 relationships involving the hypothesized mediating variables

(e.g., attitude, subjective norms, PBC locus of control, and PBC self-efficacy) are statistically

significant. For two of the ethics scenarios (exercising integrity with travel expenses and maintaining confidentiality), there is support for full mediation for most of the hypothesized

mediating relationships. For the other two ethics scenarios (e.g., resisting collusive supervision

and being truthful with clients) there is support for partial mediation for most of the hypothesized

mediating relationships. Partial mediation occurs when the original direct relationship (e.g.,

between moral disengagement and ethical behavioral intentions) is weakened but does not

disappear when the mediating variables are introduced into the model. Overall, there is a trend

across all the behaviors indicative of mediation. This lends support for the addition of moral

disengagement in the TPB model explaining ethical accounting behavior.

7. Discussion

We build on the premise that the TPB is appropriate for modeling ethical behavior. We

extend the model and link it to another important ethical concept, that of moral disengagement.

We also do not just focus on ethical problems in accounting, but to solutions to ethical problems

that are associated with professional ethics training. We link the TPB and moral disengagement to professional ethics training, the recency of professional ethics training, religiosity, and experience. We also present a model illustrating how these latter four variables counter unethical accounting behavior.

Moral disengagement involves the strategies that individuals employ to rationalize unethical action in a way that they can still maintain their view of themselves as ethical persons.

In a sense, it is a more sophisticated and more developed construct for that of rationalization found in the fraud triangle commonly included in accounting education. This is more especially so when we link moral disengagement to intentions to be ethical. We find that moral disengagement does not have a simple direct relationship with intentions to be ethical; rather, attitude towards being ethical, subjective norms for being ethical, and perceived behavioral

control (both PBC locus of control and PBC self-efficacy) mediate its effect on intentions to be ethical. In evaluating four different ethics scenarios, we find that moral disengagement is a general precursor that leads to more specific attitudes, subjective norms, and perceptions of

behavioral control that lead to intentions to behave unethically.

We find that professional accounting ethics training does have an effect on ethical

intentions. Our findings support the view that professional ethics training empowers people to

believe they can act ethically—it enhances their PBC self-efficacy—and PBC self-efficacy

influences ethical intentions. We also find that ethics training has a shelf life; that is, it can

become stale and needs to be refreshed in order to strengthen ethical intention. We conclude that

a significant one-time training is not sufficient.

We find that professional ethics training does have an effect on intentions to be ethical.

However, the effect is not simple and direct. Professional ethics training empowers people to believe they can act ethically; that is, it enhances their PBC self-efficacy. Self-efficacy in turn influences intentions to be ethical. We find that helping accountants establish a foundation in ethical behavior is not enough. Regular ethical maintenance, or continued ethics training, helps to reduce moral disengagement. When a person having experienced recent professional ethics

training considers himself or herself to be religious, the effect of these ethics tune-ups is even stronger. We also find that experience, including time as a member of a professional accounting association such as the ACCA or ICAEW, leads to an improved sense that one has some say over being ethical (PBC locus of control), as well as a belief that one can be ethical (PBC self- efficacy). These in turn lead to improved intentions to be ethical. Thus, when an accountant faces

an ethical dilemma, the commitment to act ethically is engrained and ethical actions naturally

follow.

8. Limitations and Future Research

Because the relationships between ethics training, ethical intentions, and ethical behavior are not necessarily simple and direct, further research is needed to enhance understanding of these relationships. For instance, earlier we discussed the link between ethical intentions and ethical behavior. Factors that strengthen or weaken this relationship should be researched. For instance, actual opportunity—arising because of weak internal controls—could mediate this

relationship.

Factors that affect moral disengagement in an accounting environment are important

avenues of future research. Moral disengagement occurs when people decide that particular

ethics standards do not apply to themselves. It involves the various strategies (whether conscious

or not) individuals employ to rationalize engaging in unethical action in a way that they can still

maintain their view of themselves as ethical persons. Further research should consider what

causes moral disengagement. For instance, what might management or boards do that would set

off moral disengagement in employees, and what might they do that would reduce moral

disengagement? Certain acts might increase trust in management, which could be associated with

reductions in moral disengagement. Other acts might lead to reduced trust and heightened moral

disengagement. Such additional influences on the core model we introduce could lead to clearer

expectations of what may lead to ethical failures.

References

Ajzen, I. (1991). The Theory of Planned Behavior. Organizational Behavior and Human

Decision Processes, 50(2), 179-211.

Ajzen, I. (2001). Attitudes. Annual Review of Psychology, 52, 27-58.

Ajzen, I. (2002). Perceived behavioral control, Self-Efficacy, Locus of Control, and the theory of

planned behavior. Journal of Applied Social Psychology, 32(4), 665-683.

Ajzen, I. (2006). Constructing a TPB Questionnaire: Conceptual and Methodological

Considerations. Working paper. Available at:

https://pdfs.semanticscholar.org/0574/b20bd58130dd5a961f1a2db10fd1fcbae95d.pdf

Ajzen, I., and M. Fishbein. (1980). Understanding attitudes and predicting social behavior.

Prentice-Hall, Englewood Cliffs, NJ.

American Institute of Certified Public Accountants. (2002). Consideration of Fraud in a

Financial Statement Audit, Statement on Auditing Standards No. 99. Technical report,

AICPA, New York, NY.

American Institute of Certified Public Accountants. (2019). Retrieved from

https://www.aicpastore.com/AST/Main/CPA2BIZ_Primary/YellowBookSingleAudit/PR

DOVR~PC-732302/PC-732302.jsp?cm_vc=HOMEZ2 October 8, 2019.

Aquino, K., Reed, A., Thau, S., & Freeman, D. (2007). A grotesque and dark beauty: How moral

identity and mechanisms of moral disengagement influence cognitive and emotional

reactions to war. Journal of Experimental Social Psychology, 43, 385–392.

Armitage, C.J., and M. Conner. (2001). Efficacy of the Theory of Planned Behaviour: A meta-

analytic review. British Journal of Social Psychology, 40, 471-499.

Armstrong, M.B. (1985). Internationalization of the Professional Ethic by Certified Public

Accounting: A Multidimensional Approach. Unpublished doctoral dissertation,

University of Southern California, Los Angeles, CA.

Bandura, A. (1990a). Mechanisms of Moral Disengagement, in W. Reich (ed.), Origins of

Terrorism: Psychologies, Ideologies, States of Mind, Cambridge University Press, New

York), pp. 161–191.

Bandura, A. (1990b). Selective Activation and Disengagement of Moral Control, Journal of

Social Issues, 46, 27–46.

Bandura, A. (1999) Moral Disengagement in the Perpetuation of Inhumanities, Personality and

Social Psychology Review, 3, 193–209.

Bandura, A. (2002) Selective Moral Disengagement in the Exercise of Moral Agency, Journal of

Moral Education, 31, 101–119.

Bandura, A., C. Barbaranelli, G.V. Caprara, and C. Pastorelli. (1996). Mechanisms of moral

disengagement in the exercise of moral agency. Journal of Personality and Social

Psychology, 71, 364-374.

Baron, R.M. and D.A. Kenny. (1986). The Moderator-Mediator Variable Distinction in Social

Psychological Research: Conceptual, Strategic, and Statistical Considerations. Journal of

Personality and Social Psychology, 51(6), 1173-1182.

Bloodgood, J.M., W. H. Turnley, and P. Mudrack. (2008). The Influence of Ethics Instruction,

Religiosity, and Intelligence on Cheating Behavior. Journal of Business Ethics, 82, 557–

571.

Brown, J. O., J. Hays, and M. T. Stuebs, Jr. (2016). Modeling Accountant Whistleblowing

Intentions: Applying the Theory of Planned Behavior and the Fraud Triangle. Accounting

and the Public Interest American, 16(1), 28–56.

Buchan, H.F. (2005). Ethical decision making in the public accounting profession: An extension

of Ajzen's theory of planned behavior. Journal of Business Ethics, 61(2), 165-181.

Carpenter, T.D., and J.L. Reimers. (2005). Unethical and Fraudulent Financial Reporting:

Applying the Theory of Planned Behavior. Journal of Business Ethics, 60: 115-129.

Carrington, M. J, B. A. Neville, and G. J. Whitwell. (2010). Why Ethical Consumers Don’t Walk

Their Talk: Towards a Framework for Understanding the Gap Between the Ethical

Purchase Intentions and Actual Buying Behaviour of Ethically Minded Consumers.

Journal of Business Ethics, 97, 139–158.

Chin, W.W., and P.R. Newsted. (1999). Structural equation modeling analysis with small

samples using partial least squares. In R. H. Hoyle (Eds.), Statistical Strategies for Small

Sample Research, (pp. 307–341). Thousand Oaks, CA: Sage Publications.

Cieslewicz, J.K. (2014). Relationships between national economic culture, institutions, and

accounting: Implications for IFRS. Critical Perspectives on Accounting, 25(6), 511-528.

Cieslewicz, J.K. (2016). Collusive accounting supervision and economic culture. Journal of

International Accounting Research, 15(1), 89-108.

Cohen, J.R., L.W. Pant, and D.J. Sharp. (1995). An exploratory examination of international

differences in auditors’ ethical perceptions. Behavioral Research in Accounting, 7, 37-61.

Claypool, G.A., D.F. Fetyko, M.A. Pearson. (1990). Reactions to Ethical Dilemmas: A study

pertaining to certified public accountants. Journal of Business Ethics, 9(9), 699-706.

Cooper, B. J, P. Leung, S. Dellaportas, B. Jackling, and G. Wong. (2008). Ethics Education for

Accounting Students—a Toolkit Approach. Accounting Education: an International

Journal, 17(4), 405-430.

Committee of Sponsoring Organizations of the Treadway Commission. (2013). Internal Control

– Integrated Framework. Technical report, COSO, New York, NY.

Cressey, D. R. (1953). Other People’s Money: A Study of the Social Psychology of

Embezzlement. Free Press, Glencoe, IL.

Dane, E. and S. Sonenshein. (2015). On the role of experience in ethical decision making at

work: An ethical expertise perspective. Organizational Psychology Review, 5(1), 74–96.

Duffy, M. K., Scott, K. L., Shaw, J. D., Tepper, B. J., & Aquino, K. (2012). A social context

model of envy and social undermining. Academy of Management Journal, 55, 643–666.

Fishbein, M. and I. Ajzen. (1975). Belief, Attitude, Intention and Behavior: An Introduction to

Theory and Research. Addison-Wesley, Reading, MA.

Gibson, A.M. and A.H. Frakes. (1997). Truth or Consequences: A Study of Critical Issues and

Decision Making in Accounting. Journal of Business Ethics, 16(2), 161-171.

Graafland, J. (2017). Religiosity, Attitude and the Demand for Socially Responsible Products.

Journal of Business Ethics, 144(1), 121-138.

Hair, J. F., W. C. Black, B. J. Babin, and R. E. Anderson. (2010). Multivariate Data Analysis: A

Global Perspective, 7th edition. Pearson, Upper Saddle River, NJ.

Hamilton, K., A. van Dongen, and M.S. Hagger. (2019). An extended theory of planned behavior

for parent-for-child health behaviors: A meta-analysis. Health Psychology.

https://doi.org/10.1037/hea0000940

Hassan, L.M., E. Shiu, and D. Shaw. (2016). Who says there is an intention-behavior gap?

Assessing the empirical evidence of an intention-behavior gap in ethical consumption.

Journal of Business Ethics, 136, 219-236.

House, R.J. et al. (eds.): (2004). Culture, Leadership, and Organizations: The GLOBE Study of

62 Societies. In R.J. House et al. (Eds.), Culture, Leadership, and Organizations: The

GLOBE Study of 62 Societies. Thousand Oaks, CA: Sage Publications.

Hwang, J., I. Kim, and M.A. Gulzar (2020). Understanding the Eco-Friendly Role of Drone Food

Delivery Services: Deepening the Theory of Planned Behavior. Sustainability, 12, 1-12.

International Accounting Education Standards Board (IAESB). (2006). Approaches to the

Development and Maintenance of Professional Values, Ethics and Attitudes in

Accounting Education Programs. Technical report, International Federation of

Accountants, New York, NY.

International Accounting Education Standards Board (IAESB). (2014). Initial Professional

Development – Professional Values, Ethics, and Attitudes (Revised). Technical report,

International Federation of Accountants, New York, NY.

International Accounting Education Standards Board (IAESB). (2015). Education Standards for

Professional Accountants and Aspiring Professional Accountants. Technical report.

International Federation of Accountants, New York, NY.

International Auditing and Assurance Standards Board (IAASB). (2009). The auditor’s

responsibilities relating to fraud in an audit of financial statements. International

Standard on Auditing 240. Technical report, IAASB, New York, NY.

International Ethics and Standards Board of Accountants (IESBA). (2018). 2018 Handbook of

the international code of ethics for professional accountants. Retrieved from

https://www.ethicsboard.org/international-code-ethics-professional-accountants.

Institute of Chartered Accountant in England and Wales. (2019). Ethics Learning Programme.

Retrieved from https://www.icaew.com/groups-and-networks/become-a-bfp/ethics-

learning-programme, October 10, 2019.

Jacques, E. (2003). Ethics for Management. Management Communication Quarterly, 17(1), 136–

142.

Jakubowsdki, S. T., P. Chao, S. K. Huh, and S. Maheshwari. (2002). A cross-country comparison

of the codes of professional conduct of certified/chartered accountants. Journal of

Business Ethics, 35(2), 111–129.

Kohlberg, L. (1981). The Philosophy of Moral Development: Moral Stages and the Idea of

Justice. Harper & Row, San Francisco, CA.

Mahama, H., and M. Cheng. (2013). The effect of managers’ enabling perceptions on costing

system use, psychological empowerment, and task performance. Behavioral Research in

Accounting, 25(1), 89–114.

Mayhew, B.W., and P. R. Murphy. (2014). The Impact of Authority on Reporting Behavior,

Rationalization and Affect. Contemporary Accounting Research, 31(2), 420–443.

Moore, C., Detert, J. R., Treviño, L. K., Baker, V. L., Mayer, D. M. (2012). Why employees do

bad things: Moral disengagement and unethical organizational behavior. Personnel

Psychology, 65, 1–48.

Moore, D.A., Tetlock, P.E., Tanlu, L., and Bazerman, M. (2006). Conflicts of interest and the

case of auditor independence: Moral seduction and strategic issue cycling. Academy of

Management Review, 31(1), 10-29.

Nelson, M. W. (2009). A model and literature review of professional skepticism in auditing.

Auditing: A Journal of Practice & Theory, 28(2), 1–34.

Parboteeah, K. P., M. Hoegl, and J. Cullen. (2009). Religious groups and work values: A focus

on Buddhism, Christianity, Hinduism, and Islam. International Journal of Cross Cultural

Management, 9(1), 51–67.

Podsakoff, P. M., MacKenzie, S. B., Lee, J. Y., and Podsakoff, N. P. (2003). Common method

biases in behavioral research: A critical review of the literature and recommended

remedies. Journal of Applied Psychology, 88(5), 879-903.

Public Company Accounting Oversight Board (PCAOB). (2005). Consideration of fraud in a

financial statement audit. AU Section 316. Technical report. PCAOB, Washington, D.C.

Ramasamy, B., M. C. H. Yeung, and A. K. M. Au. (2010). Consumer support for corporate

social responsibility (CSR): The role of religion and values. Journal of Business Ethics,

91(1), 61–72.

Robbins, J. M., and J. I. Krueger. (2005). Social projection to ingroups and outgroups: A review

and meta-analysis. Personality and Social Psychology Review, 9(1), 32–47.

Si, H., J. Shi, D. Tang, G. Wu, and J. Lan. (2020). Understanding intention and behavior toward

sustainable usage of bike sharing by extending the theory of planned behavior.

Resources, Convervation & Recycling, 152, 1-10.

Sikka, P., C. Haslam, C. Cooper, J. Haslam, J. Christensen, D.G. Driver, T. Hadden, P. Ireland,

M. Parker, G. Pearson, A. Pettifor, S. Picciotto, J. Veldman, and H. Willmott. (2018).

Reforming the Auditing Industry. Policy paper available at:

http://visar.csustan.edu/aaba/LabourPolicymaking-AuditingReformsDec2018.pdf

Stanga, K. G. and R. A. Turpen. (1991). Ethical Judgements on Selected Accounting Issues: An

Empirical Study. Journal of Business Ethics, 10 (10), 739–747.

Sun, W. (2019). Toward a theory of ethical consumer intention formation: re-extending the

theory of planned behavior. AMS Review https://doi.org/10.1007/s13162-019-00156-6

Sutton, S. (1998). Predicting and explaining intentions and behavior: How well are we doing?

Journal of Applied Social Psychology, 28, 1317–1338.

Schweitzer, M. E., & Hsee, C. K. (2002). Stretching the truth: Elastic justification and motivated

communication of uncertain information. Journal of Risk and Uncertainty, 25, 185–201.

Shalvi, S., Dana, J., Handgraaf, M. J., & De Dreu, C. K. (2011). Justified ethicality: Observing

desired counterfactuals modifies ethical perceptions and behavior. Organizational

Behavior and Human Decision Processes, 115, 181–190.

Treviño, L.K., den Nieuwenboer, N.A., & Kish-Gephart, J.J. (2014). Unethical behavior in

organizations. Annual Review of Psychology, 65, 635–660.

Treviño, L.K., Weaver, G.R., & Reynolds, S. (2006). Behavioral ethics in organizations: A

review. Journal of Management, 32, 951–990.

Vinzi, V. E., L. Trinchera, and S. Amato. (2010). PLS path modeling: From foundations to

recent developments and open issues for model assessment and improvement. In V. E.

Vinzi, W. W. Chin, J. Henseler, and H. Wang (Eds.), Handbook of Partial Least

Squares: Concepts, Methods, and Applications, (pp. 47-82). Dusseldorf, Germany:

Springer-Verlag.

Webb, T. L. and P. Sheeran. (2006). Does Changing Behavioral Intentions Engender Behavior

Change? A Meta-Analysis of the Experimental Evidence. Psychological Bulletin, 132(2),

249–268.

Yuriev, A., M. Dahmen, P. Paillé, O. Boiral, and L. Guillaumie. (2020). Pro-environmental

Behaviors through the Lens of the Theory of Planned Behavior: a Scoping Review. Resources,

Conservation & Recycling, 155, 1-12.

Figure 1 Theoretical Model of Ethics Training and Ethical Intentions

Figure 2 Model for Ethics Training and Maintaining Confidentiality

Note: Hypotheses are directional, but two-tailed results are presented.

Figure 3 Model for Ethics Training and Resisting Collusive Supervision

Note: Hypotheses are directional, but two-tailed results are presented.

Figure 4 Model for Ethics Training and Being Truthful with Clients

Note: Hypotheses are directional, but two-tailed results are presented.

Figure 5 Model for Ethics Training and Exercising Integrity with Travel Expenses

Note: Hypotheses are directional, but two-tailed results are presented.

TABLE 1 Descriptive Statistics (N = 270)

Panel A: Demographics

Gender (% female) 44.6% Mean Age 31.9 How religious are you? (Mean: 1 very, 7 not at all religious) 4.0 How important is religion? (Mean: 1 very, 7 not at all important) 4.3 Highest level of education: University Degree or Professional Qualification 95.2% Degree Subject (percent in each): Accounting 49.6% Business 8.4% Economics 6.5% Finance 9.8% Science 9.5% Social Science 3.8% Other 12.4% Most important influence on values (Mean: rank 1-6) employer 4.12 family 2.02 friends 3.49 professional body 3.69 religion 3.71 schooling/university education 3.68 General ethics as part of degree course 31.2% Business ethics training as part of degree course 21.1% Accounting ethics training as part of degree course 31.3% Employer sponsored ethics training 36.1%

Panel B: Descriptive Statistics

Variable Mean Std Dev Minimum Maximum Professional Ethics 0.38 0.49 0 1.00 Other Ethics 0.76 0.43 0 1.00 Recent Ethics 0.38 0.49 0 1.00 African 0.14 0.35 0 1.00 Anglo 0.39 0.49 0 1.00 Confucian 0.10 0.30 0 1.00 Eastern Europe 0.09 0.28 0 1.00 Middle Eastern 0.10 0.30 0 1.00 Southeast Asia 0.09 0.29 0 1.00 MD (n= 482) 5.87 0.79 2.25 7.00 TPB (n= 517) 4.57 0.49 1.00 6.00

TABLE 2

Summary of Model Results Across All Four Ethics Scenarios

This table summarizes the results illustrated in the figures. There is support overall for the model as discussed in the text. Construct names have been shortened in the table. Hypotheses are directional, but two-tailed results are presented.

Hypothesized Relationship: Ethical Behavior: Collusive Being Truthful Integrity with Confidentiality Supervision with Clients Travel Expenses

Exogenous Variable Endogenous Variable Regression Regression Regression Regression Hyp. (Independent Variable) (Dependent Variable) Coefficient p value Coefficient p value Coefficient p value Coefficient p value

H1a Attitude Intentions to be Ethical .494 .000 .204 .000 .293 .000 .280 .000

H1b Subjective Norms Intentions to be Ethical .319 .000 .416 .000 .417 .000 .275 .000

H1c PBC Locus of Control Intentions to be Ethical .093 .016 .117 .006 .094 .019 .075 .079

H1d PBC Self-Efficacy Intentions to be Ethical .066 .090 .295 .000 .208 .000 .345 .000

H2a Moral Disengagement Attitude -.254 .000 -.270 .000 -.241 .000 -.259 .000

H2b Moral Disengagement Subjective Norms -.204 .001 -.227 .000 -.166 .006 -.162 .007

H2c Moral Disengagement PBC Locus of Control -.108 .079 -.216 .000 -.089 .140 -.197 .001

H2d Moral Disengagement PBC Self-Efficacy -.213 .000 -.211 .000 -.188 .002 -.204 .001 Professional Ethics H5 Training PBC Self-Efficacy .138 .027 .065 .283 .124 .048 .061 .341

H6a Experience PBC Locus of Control .139 .024 -.001 .982 .235 .000 .073 .228

H6b Experience PBC Self-Efficacy .160 .012 .194 .002 -.026 .683 .152 .019

H4a Recent Ethics Training Moral Disengagement -.335 .018 -.326 .024 -.295 .021 -.279 .016

H4b Religiosity Moral Disengagement -.117 .085 -.111 .102 -.111 .093 -.132 .041 Religiosity X Recent Ethics H4c Training Moral Disengagement -.291 .048 -.282 .061 -.252 .058 -.232 .053

TABLE 3

Mediation Analyses for Four Ethical Behaviors

We extend the Theory of Planned Behavior for ethical behaviors by adding moral disengagement. The TPB holds that attitude, subjective norms, PBC locus of control, and PBC self-efficacy influence behavioral intentions, which influence behavior. For four ethics scenarios, we find that moral disengagement influences intentions to be ethical through the mediating variables of attitude, subjective norms, PBC locus of control, and PBC self-efficacy. This table contains the results of mediation tests following Baron and Kenny (1986).

Hypothesized Relationships Ethical Behaviors

Mediating Variables are Bolded below Resisting Being Truthful Integrity with The Direct Relationships are Italicized below Confidentiality Collusive Supervision with Clients Travel Expenses

Regression Regression Regression Regression Independent Variable Dependent Variable Coefficient p value Coefficient p value Coefficient p value Coefficient p value

Step 1: Is the Direct Relationship that is Hypothesized to be Mediated Statistically Significant ?

Moral Disengagement Intentions to be Ethical -.278 .000 -.335 .000 -.285 .000 -.281 .000

Step 2: Are the Relationships that Comprise the Hypothesized Mediating Pathways Statistically Significant?

Moral Disengagement Attitude -.292 .000 -.262 .000 -.244 .000 -.260 .000

Attitude Intentions to be Ethical .544 .000 .226 .000 .317 .000 .317 .000

Moral Disengagement Subjective Norms -.200 .001 -.230 .000 -.168 .006 -.165 .006

Subjective Norms Intentions to be Ethical .279 .000 .384 .000 .386 .000 .246 .000

Moral Disengagement PBC Locus of Control -.198 .001 -.231 .000 -.147 .015 -.230 .000

PBC Locus of Control Intentions to be Ethical .051 .183 .089 .036 .086 .031 .067 .109

Moral Disengagement PBC Self-Efficacy -.245 .000 -.274 .000 -.196 .001 -.229 .000

PBC Self-Efficacy Intentions to be Ethical .031 .426 .265 .000 .175 .000 .328 .000

Step 3: With the Introduction into the Model of the Mediating Pathways, Does the Strength of the Original Direct Relationship Decrease (Partial Mediation) or Become Insignificant (Full Mediation)?

Moral Disengagement Intentions to be Ethical -.035 .379 -.091 .040 -.093 .021 -.043 .316