“The effect of morality and Machiavellianism on government fraud: The moderating role of education”

The high number of fraud cases in governments has increasingly become a concern among academics and practitioners. Many approaches can be used to minimize fraud cases. However, the role of education in the influence of personality on fraud in the Indonesian government has not been considered. Thus, this study aims to analyze the influence of morality and Machiavellianism as personality factors leading to fraud. Education was tested as a moderating variable. This quantitative study involved a sample of Pekanbaru City government officials. To collect data, questionnaires were sent directly to the respondents. Through the purposive sampling technique, 251 samples were collected. A structural equation with partial least squares was used to test the hypotheses. The test results show that all the proposed hypotheses are accepted. Morality affects fraud with a p-value < 0.001, less than 0.05. Machiavellianism affects fraud with a p-value < 0.001. Education moderates the relationship between morality and fraud with a p-value of 0.048, less than 0.05. Moreover, education moderates the relationship between Machiavellianism and fraud with a p-value < 0.001, less than 0.05. In conclusion, this study contributes to shaping government officials’ behavior to prevent fraud. Character building and high education standards for officials must be integrated to reduce the risk of abuse in Indonesian government.


INTRODUCTION
The Indonesia Corruption Watch (ICW) shows that in 2021, the Attorney General's Office had to deal with 371 corruption cases involving 814 suspects. These are the highest numbers of cases and defendants in the past five years (Rizaty, 2022). According to the Corruption Eradication Commission, Riau Province is one of Indonesia's ten most corrupt areas (Kamis, 2020). Twenty-eight civil servants in Riau have been caught in corruption cases. Several regional heads, ranging from governors and regents to low-level government officials, have yet to be spared from corruption cases (Merdeka, 2019).
Fraud is an unlawful activity carried out intentionally to obtain something using deception. Meanwhile, corruption is one type of fraud that occurs frequently in Indonesia. There is no longer a stigma associated with corruption; people of all ages can engage in it. Corruption is a form of public sector fraud that is pervasive. Typical actions include manipulating records and removing documents as well as markups that harm the nation's economy.
The significant prevalence of corruption cases in Riau Province suggests that government operations continue to suffer from inadequate oversight. Therefore, it is imperative to investigate these fraudulent acts in order to determine the underlying causes.
Personality can play a role in determining whether someone commits fraud. A lack of morality and a tendency toward Machiavellianism can motivate someone to cheat. However, the level of education can control a person's behavior. A person's education provides knowledge for individuals and, more than that, provides an understanding of ethical behavior (Raharjo, 2015).

LITERATURE REVIEW AND HYPOTHESES
Every business faces the risk of fraud, as the practice itself has a long history. Companies such as Enron, WorldCom, and Satyam, among others, have all fallen victim to this detrimental phenomenon. Deceptive behaviors are exhibited when individuals aim to deceive or mislead others. Misrepresentation can be classified into two primary categories: intentional and unintentional, as identified by Ramamoorti (2008).
Fraud is a purposeful misapplication or abuse of an organization's assets to obtain a personal benefit (ACFE, 2014). The Association of Certified Theft Examiners states that work theft has some characteristics: (1) covertness; (2) theft from the company by an employee who has a fiduciary duty to the company; (3) promises of substantial financial benefits to the employee; and (4) relatedness to the resources, earnings, or holdings of the company that employs the person (ACFE, 2014). It distinguishes among three significant types of fraudulent activities: theft of assets, fraudulent financial statement preparation, and corrupt government officials. Individual morality is one of the personality elements that encourage fraud committed by individuals. Morality is the conspicuous actual dimension of every human group (Walker, 2007 Kohlberg (1971) and Sosler (2019) propose that when individuals reach a higher level of morality, they are more likely to prioritize the community's interests over their own personal or organizational concerns. Thus, they try to avoid committing unethical actions that harm many people. Morality significantly contributes to fraud (Setiawan et al., 2020). Fraud can also be avoided by always being open to each other within agency secrecy. If morality is well maintained, then cheating can be avoided. In government, individual morality becomes an important thing to prevent fraud. Therefore, this study also suspects that individual morality will also affect fraud that occurs in the government.
The next element of personality is Machiavellian traits. Machiavellianism is a form of manipulative character. When someone with this personality has a goal, he/she will think of and do various ways to achieve it. The gone theory by Bologna (1993) reveals that greed is one of the driving factors for fraud (Talib & Kusumawati, 2020). It relates to Machiavellian nature because it will use any means to achieve its goals. People that have a high Machiavellian trait prevalence tend to lie more frequently (Triantoro et al., 2020). They have a personality that lacks concern for human relationships because they disregard conventional morality and have poor ideological commitment (Dammak et al., 2022). It can be shown by the fact that they ignore traditional morality. Theoretically, individuals with high Machiavellian tendencies are not concerned with judging morally ambiguous behaviors. They are more inclined to engage in ethical or unethical ways to attain their aims (Zirman & Basri, 2014 A person's level of education provides more than just knowledge for individuals. Someone highly educated provides an understanding of behaving more ethically (Head, 2020). Individuals with higher education are also considered to have high ethics and moral sense (Rottweiler & Gill, 2022). Thus, education and morality can prevent someone from committing immoral acts such as fraud. Triantoro et al. (2020) found that those who score highly on the Machiavellian trait of government officials are also more likely to lie. The correlation between Machiavellian characteristics and deceit can be attenuated or amplified by the degree of formal education. When a person has high Machiavellian traits and a high level of education, he will be controlled by his knowledge and understanding, preventing the individual from committing unethical actions (Kurniawan & Anjarwati, 2020). Thus, the interaction between Machiavellian traits can reduce the level of fraud. Based on previous discussions, this study suspects that the level of education can influence the effect of morality and Machiavellianism on cheating among government personnel in Indonesia.
Based on the literature review, this study aims to analyze how morality and Machiavellianism affect fraud in the local government of Pekanbaru City, Indonesia. This study also investigates how educational level moderates the relationship between morality and Machiavellianism. Therefore, the following hypotheses are developed: H1: Morality has a negative effect on fraud.
H2: Machiavellianism has a positive effect on fraud.
H3: The level of education moderates the effect of morality on fraud. H4: The level of education moderates the effect of Machiavellianism on fraud.

METHODOLOGY
All government officials working for the Pekanbaru City Government were included in this study. In this analysis, purposive sampling was employed to choose the research samples. The selected sample criteria are government officials with positions as head, secretary, and deputy head of the division.
Questionnaires were sent to government officials in all municipal governments to collect the research data. A total of 300 questionnaires were sent, and 251 questionnaires were collected and then processed.
This study used the dependent variable -fraud. Morality and Machiavellian characteristics are the independent variables. The education level is the moderating variable. Variable measurements are presented in Table 1.
The paper employed a structural equation model in Warp-PLS to analyze the collected data. Descriptive and inferential statistical methods are combined in PLS, which is a component-based or variance-based (SEM) structural equation model (Ghozali & Latan, 2015). SEM testing with PLS comprises a two-stage process. The outer model testing phase involves checking the validity and reliability of the model. Convergent and discriminant validity are the two main types of validity tests. When a questionnaire predicts some other variable or construction, it can be subjected to a reliability test to see how well it predicts (Ghozali, 2014). The questionnaire is considered reliable if one's responses to a statement are constant or steady throughout time. An indicator Table 1. Definition and measurement of variables Fraud is defined as providing false statements in the shape of significant facts, and the truth cannot be believed. As a result, it influences or induces others to act or to act in a negative manner (ACFE, 2014). The standard of education is consistent across the board and is passed on without interruption from one generation to the next in every region of the world. The acquisition of knowledge is one of the most vitally essential tools for the evolution of the human species, and these endeavors are carried out in a manner that considers each community's worldview and historical context (Tirtarahardja, 2005

RESULTS
The study collected 380 questionnaires, of which 251 were deemed valid and analyzed. The return rate is 66.05%. The questionnaire uses a 5-point Likert scale. The government officials were grouped based on criteria including length of service, gender, status, age, religion, and education. Table 2 presents the demographics of the sample. The descriptive statistics are illustrated through the minimum, maximum, average value (mean), and standard deviation (Table 3).  Table 4 indicate convergence in validity measures.
In this case, a loading factor value of > 0.5 was obtained after deleting multiple false positives. This study confirms that the indicator of convergent validity has been attained. Convergent validity can be tested in various ways besides factor loading, for as by inspecting the mean-variance of the extracted data (AVE). Convergent validity is satisfied when the AVE is more significant than 0.50 (Hair et al., 2010). Based on Table 4, the AVE is consistently more than 0.5.
Furthermore, the reliability test measures the consistency of measuring instruments. It is a measure that indicates the extent to which a questionnaire can consistently measure a variable. Cronbach's alpha and the composite reliability should be more than 0.70 for the reliability to be considered adequate (Hair Jr et al., 2021).  Table 4, the R-Square value corresponding to the fraud variable is 0.625. Individual morality and Machiavellianism control 62.5% of the fraud variable, moderated by education level. The remaining 37.5%, on the other hand, is almost certainly affected by additional factors that were not investigated in this study. The relative change in R-squared shows that an independent hidden variable influences the dependent variable under study (Ghozali & Latan, 2015). When the R-square number is higher, it indicates that the prediction model of the suggested research method is more accurate.
The next step is a test of discriminant validity. Discriminant validity is considered high when there is a substantial distinction between the traits the measurement instruments and the theoretical frameworks for those variables should not measure. Knowing the cross-loading value of each latent variable is necessary for testing the reflective measurement model for discriminant validity. For example, the indicator's structure-measurement item connection may be more robust than any other structure. It suggests that the fundamental structure can forecast the indication more accurately than the other structures (Fornell & Larcker, 1981). Table 5 shows discriminant validity testing the correlation of the latent variables.   Table 6 presents the results of validating the model fit, demonstrating that all metrics meet the expectations of testing quality indexes like APC, ARS, and AVIF. APC and ARS are acceptable with p-value < 0.05 and AVIF < 5 (Sholihin & Ratmono, 2013).
Analysis of the path coefficient and testing of hypotheses are the next steps. Figure 1 depicts the interdependencies between the variables, and Table 7 lists the path coefficients and significance levels.
The path coefficient of the relationship between morality and fraud is -0.368, and the p-value is <0.000, less than the significance level (0.05). These results indicate that morality has a negative impact on fraud (H1 is accepted). The path coefficient value of the relationship between Machiavellianism and fraud is 0.432, with a p-value of 0.001, less than the significance level (0.05). These results prove that Machiavellianism has a beneficial influence on fraud (H2 is accepted). The path coefficient value of the relationship between education and  Note: * p < .05;** p < .01; *** p < .001.

DISCUSSION
The analysis indicates that morality has a negative impact on individuals' propensity to engage in fraudulent activities. This study reveals that fraud tends to decrease as the moral character of government officials in Pekanbaru City increases. Conversely, fraud is more likely to occur when the moral character of government officials in Pekanbaru City is lower. The data show that the morality of government officials in the Pekanbaru city government is quite good. This result confirms Kohlberg's theory of moral growth, which describes each person's moral development. A person with weak moral reasoning will likely act in ways that a person with solid moral sense would not. Higher-order moral reason makes people less likely to behave dishonestly (Wilopo, 2006).
On the other hand, someone who has low morale will commit fraud. Individuals with insufficient moral reasoning tend to do something that will cause legal sanctions (Septiningsih, 2021 The third hypothesis test results indicate that one's educational background can influence the correlation between morality and fraud. The average level of education of respondents is in higher education. However, the path coefficient value shows a positive direction. This study indicates that fraud tends to occur for someone with a high intellectual understanding. This supports the findings of ACFE (2014) that officials with higher education commit fraud in government.
In line with Kabuye et al. (2018), the success or failure of fraud is determined by the capability or ability of the perpetrator to commit fraud. The perpetrator must own two powers to succeed in his fraudulent actions, including position and intelligence. The high status of a person in an organization will make it easier for him/her to commit fraud, and the risk of being detected will also be smaller. A high position is synonymous with a high level of education. For instance, government officials adept at analyzing internal control systems can benefit from this since it simplifies identifying vulnerable points in the system. The results of the fourth hypothesis testing demonstrate that education level can attenuate the association between Machiavellianism and fraud.
The study shows that education has a positive moderating effect on Machiavellianism. The results indicate that the respondents, on average, have a higher education degree. It suggests that a person's education level is not necessarily correlated with any reduction in his Machiavellian tendencies. This result is in line with Sabău (2013), that, in general, someone who is socially educated tends to be honest, but at the same time, individuals have different needs. Therefore, they will be motivated to lie, steal, and commit fraud. If it is supported by high greed, then high education will increase the incidence of fraud. According to the gone theory, the factors that cause corruption are greed, opportunity, need, and disclosure (Restya & Amalia, 2019 This study has contributed to the literature emphasizing the importance of character-building to reduce the occurrence of fraud. The findings have implications for the government that morality and Machiavellianism can lead to fraud. Education also plays a role in encouraging cheating. However, the results of this study can only be generalized in a limited area. The research findings show that there are still many opportunities for future researchers to investigate the causes of government fraud, e.g., from the point of view of hexagon theory. Further research can also develop a broader scope of government or analyze other areas of the public sector.

CONCLUSION
This investigation aims to examine the effect of morality and Machiavellianism on fraud among government personnel in Indonesia. This study also analyzes how educational level influences the link between morality and Machiavellianism. According to the findings, high morality in individuals will reduce fraud. Machiavellianism is beneficial to corrupt government officials in that it helps them avoid being exposed to fraud. The higher the nature of Machiavellianism, the higher the level of fraud that will occur. This finding proves that the level of education can moderate the relationship between morality and Machiavellian traits. According to the study results, high education makes people more likely to commit fraud.
This study has certain limitations, primarily from its focus on the government officials in Pekanbaru City. To enhance the generalizability of the findings, further research should encompass a broader range of contexts and geographic areas. Additionally, this study solely explores the impact of morality and Machiavellianism on fraud occurrence, neglecting other aspects of the fraud hexagon theory. Future studies can delve into a more comprehensive examination of the various characteristics associated with the fraud theory.
This study suggests that individual morality should be strengthened, and it is essential to emphasize how important it is to maintain ethical behavior following the ethics code. Governments can provide ethics-related training, improve oversight and regulation, and enhance individual morale.