Analysis of Tone at the Top Principle’s Implementation in Fraud Prevention on Public University in Kupang City

Fraud prevention can be done by embedding fraud awareness to an entity’s employees and this actively demonstrates that management needs to create a valid control environment by creating a tone at the top to erode the hypocritical behavior of each leader at all levels. This research was conducted on public universities in Kupang City namely Universitas Nusa Cendana, Politeknik Negeri Kupang, Politeknik Kesehatan Kementerian Kesehatan Kupang, and Politeknik Pertanian Kupang. The goal of this study was to discover the impact of tone at the top principle on fraud prevention on public universities in Kupang. The sampling technique used was purposive sampling with total amount of sample was 46. All data was analyzed with SmartPLS 3.0 as the analysis tool the technique used was simple linear regression. The outcomes of this study partially show that the principle of tone at the top has a notable effect on the fraud prevention in public universities in Kupang. Meanwhile, the results of the study concurrently perform that the principle of tone at the top 43,8% affected fraud prevention while 56,2% was influenced by other factors.


INTRODUCTION
University as one of the economic entities that manage funds sourced from individuals, communities and or the government are obliged to submit accountability reports to stakeholders or shareholders. The groundwork of laws and regulations are one of the criteria used in financial management and accountability which include planning and budgeting, implementation and administration, accountability and reporting, and auditing.
Education sector is one of the important sectors in Indonesia, because this sector is one of the indicators of the country's progress. Quoted from the Merdeka.com website (1/9/2020), in 2021 the Government has budgeted 20% of the APBN for the education sector, which is Rp. 549.5 T and Rp. 5.30 T of the budget is budgeted for Universities. However, ironically, based on the monitoring of Indonesian Corruption Watch (ICW) quoted from the Kompas.com website (2020) during 2015 to 2019, there were 202 corruption cases involving 465 people. From the 202 corruption cases that occurred in the field of education, there were at least 20 cases or around 10% that took place in the world of high schools or colleges with state losses approaching Rp. 81.9 M. Indonesian Corruption Watch (ICW) quoted from the Kompas.com website (2020) also noted the motives of corruption that often occur in the education sector, namely price inflation, embezzlement of college goods and assets, misuse of budgets for research funds and scholarships, illegal levies, fictitious activities/projects, abuse of authority, and others.
It must be admitted that fraud prevention efforts do not always come to light, due to the many motivations of the perpetrators and also the fraud method. There are many theories and methods of prevention that can be done, one of which is by building a quality corporate or organizational culture. This is in line with the opinion of several researchers which through their research elaborated that if an organization does not instill organizational values and ethical rules as a culture, it is will give slight chance for the employees to implement honesty values. The ethical culture of this organization is usually impacted by the leadership style of the top brass of the organization. For that being said, the important part of this case is to balanced perceptions between people who will influence behavior with people whose behavior will be influenced. With the power possessed, a leader has the means or tools to influence the behavior of followers and subordinates with the power they have. With regard to the correlation between power and leadership, leaders should need to assess their own behavior before judging or influencing the behavior of others. In another perspective, power can be considered as the potential for influence.
In simple terms, tone at the top is a statement, expression, speech, will, and color that comes from the top management. Tone at the Top is an example that must be built in an organization or company to stay away from acts that violate. But unfortunately, many organizational leaders are not fully aware of this. A good leadership style that can show consistency between what is said and what is done by the company's leadership. Thus, fraud prevention should be started by eradicating hypocritical behavior. This means that fraud prevention must be started and proven by each leader at all levels.

LITERATURE REVIEW AND HYPO-THESIS Attribution Theory
According to Fritz Heider in Luthans (2005:20) the founder of attribution theory, proposed that attribution theory basically explains about a person's behavior, how we determine the causes and motives of a person's behavior. Attribution theory depicts how a person's reaction to events around them, by knowing their reasons for the events they experience. This theory elaborates that we can predict a person's behavior in dealing with certain situations regarding to their certain attitudes and characteristics that can be observed.

Fraud
Fraud is a common term among accountants, auditors, investors and the public engaged in economics and accounting. Fraud has a very broad and varied scope of understanding. Fraud is unlawful acts carried out intentionally for certain purposes (manipulation or giving false reports to other parties) carried out by people from inside or outside the organization to gain personal or group benefits directly or indirectly harming other parties (ACFE ,2016).
Fraud prevention can be done by eliminating the factors that cause fraud. The concept of fraud prevention can be explained as an effort to instill awareness about the existence of fraud (fraud awareness) and an effort to evaluate the risk of fraud (fraud risk assessment). Due to its intangible nature, risk awareness cannot be applied directly by management decisions, but management can take steps to increase risk awareness by creating an appropriate control environment. In other words, fraud prevention efforts can start from internal control.
The control environment creates an atmosphere of control in an organization and influences personnel awareness of control. According to Murtanto (2005:32) the control environment shows the atmosphere or atmosphere (sets the tone) in an entity/company that affects the control consciousness of the people in the organization.

Tone at The Top Principle
Tone at the Top is an example that must be built in an organization or company to stay away from acts that violate. Business ethicists believe that having a good ethical climate created by leaders can help prevent fraud and other unethical practices. Thus, to create a control environment that is aware of the risk of fraud, top management must create a tone at the top related to risk management through communication and expectations.

Leadership Style
Leadership concerns the process of social influence that is intentionally carried out by one person on another to structure activity and influence within a group or organization (Robbins, 2013: 83).
Mas 'ud (2004:34) says that leadership style consists of five dimensions of leadership style, namely: a. Participative style, which is a leadership style in which the leader expects suggestions and ideas from subordinates before making a decision.
There are numerous advantages of this leadership style, including: the job satisfaction from the team members will increase and tend to be productive because they feel involved; improve employee talent development; and team members feel as part of a larger and meaningful system and are motivated to achieve more than just financial satisfaction. b. Nurturant Style, which is a leadership style in which the leader pays attention to subordinates in career advancement, provides guidance, direction, assistance and behaves well and appreciates subordinates who work on time. c. Authoritarian Style, namely a leadership style that does not require main ideas from subordinates and prioritizes power and prestige so that a leader has high confidence in decision making.
Authoritarian leaders have full power over their members, so team members have little opportunity to express opinions. d. Bureaucratic Style, namely a leadership style that follows the rules strictly and convinces his subordinates that they also follow the same rules. This system is a suitable system for jobs that involve hazardous work risks or where large sums of funding are involved. Bureaucratic leadership is also very useful in organizations where employees work in routines. e. Task Oriented Style, namely a leadership style in which a leader requires subordinates to be disciplined in terms of work or tasks. This leadership style requires team members to complete existing work or tasks according to predetermined rules.

METHODS
This research includes quantitative research using primary data. This research was carried out in February -April 2021 at 4 State Universities in Kupang City, namely Universitas Nusa Cendana, Politeknik Negeri Kupang, Politeknik Kesehatan Kementrian Kesehatan Kupang and Politeknik Pertanian Kupang. The independent variable used is the principle of tone at the top which is proxied to 4 leadership styles, namely participative style, nurturing style, bureaucratic style, and task-oriented style. While the dependent variable used is fraud prevention which is proxied into 7 indicators of the control environment, which are integrity and ethical values, commitment to competence, management philosophy and leadership style, organizational structure, delegation of authority and respsonsibility, human resource management policies and practices, and management attention and direction. A set questionnaire was used in this study to collect data which measured by 4 ordinal scales where the answers to each instrument item have gradations ranging from Strongly Agree (SA), Agree (A), Disagree (D), and Strongly Disagree (SD). This study uses four alternative answers by eliminating the alternative answer "Less Agree (LS)". This is done because "Less Disagree (DS)" indicates the subject is not sure of the answer given (Azwar, 2007). Elimination of alternative answer "Less Disagree (LS)" was carried out by researchers as an effort so that the subject only gave answers that were believed by the subject.
In this study, the data then tabulated and analyzed with SmartPLS 3.0 analysis tool. The analysis of data was carried out in two stages, namely (1) measurement (outer) model test and (2) structural (inner) model test.

RESULTS AND DISCUSSION Model (Outer) Measurement Model Convergent Validity
The value of convergent validity is the value of the loading factor on the latent variable with its indicators. The value expectation is > 0.7, or the limit of 0.6 is often used as the minimum limit of the loading factor value. However, in this study, the loading factor value used is > 0.7. The indicators that meet these criteria are PS2, NS1, NS2, NS3, NS4, BS1, INE 1, KTK1, KTK2, KTK3, FMGK1, FMGK2, FMGK3, SO1, SO2, PWTJ1, KSDM1, and KSDM2.

Average Variance Extracted (AVE)
AVE is used to evaluate every construct or latent variable's discriminant validity. The model has better discriminant validity if the resulting value is greater than 0.5. In this study, the AVE value of each construct was above 0.5. Hence, there is no issue with convergent validity in the model being tested or it can be said that this research model has fit the evaluation for discriminant validity.

Discriminant Validity Test
This type of test is used to guarantee that every concept of each latent variable used in this study is distinctive from other variables. The table below shows the results of the research's discriminant validity model by assessing the crossloading values.
From the results of the estimated cross loading in Table 2, it shows that the loading value of each indicator item on the construct is greater than the cross-loading value. Thus, it can be concluded that all constructs or latent variables already have good discriminant validity.

Composite Reliability Test
In addition to assess convergent validity and discriminant validity, composite reliability was carried through by looking at the looking at construct reliability or latent variables which are measured by looking at composite reliability and Cronbach's alpha values. To determine a reliable model, the composite reliability and Cronbach's alpha values must be greater than 0.7. Table 3, the model shows that the composite reliability and Cronbach's alpha values for all constructs are above 0.7. Therefore, it can be decided that all constructs have good reliability.

Structural (Inner) Model Test
The assessment of the inner model in this study can be carry out by analyzing the R-square for the dependent latent variable.

Predictive Relevance
Ghozali (2014), the results of the calculation of predictive relevance or Q2 show how well the observed values are produced by the model and also the estimated parameters. The value of Q2>0 indicates that the model has predictive relevance. In this research model, the Q2 value is greater than 0 (zero) so that the predictions made by the model are considered relevant.

Quality Index
Quality index is used to show how good our research model is. Quality index can be measured in several ways, one of which is using the Normed-fit Index (NFI). Table  6 shows the NFI value of this research model.

Path Coefficients
The path coefficient calculations' results in this study presented by Table 4.9 show that the path has a coefficient value of 0.662. The path coefficient value that is in the range -1 to 0 means that the hypothesis has a negative influence and conversely the path coefficient value that is in the range 0.1 to 1 means that the hypothesis has a positive influence. Therefore, it can be interpreted that the Tone at The Top Principle has an effect on fraud prevention or the hypothesis is accepted.

Boot-strapping Result
In PLS, the relationship between each variable was tested by utilizing a simulation using the bootstrapping method on the sample with the purpose to minimize the problem of abnormal research data and test the significance of the independent variable on the dependent variable. The test results using the bootstrapping method from the PLS SEM analysis shown as follows.

Research Discussion
Based on testing using the Partial Least Square (PLS) analysis technique, the results showed that Tone at The Top Principle contributed 43.8% of the influence on Fraud Prevention at State Universities in Kupang City. In testing the hypothesis, Tone at The Top (X) Principle on Fraud Prevention (Y) shows a path coefficient value of 0.662 with a t value of 11.573. The t value is greater than t table (1.96) or the p-values are less than 0.05 (α=5%). These results mean that the application of the Tone at the Top principle has a significant direct effect on the Prevention of Fraud at State Universities in Kupang City.
The tone at the top principle has a significant effect because based on theoretical studies it explains that the leader has the influence to determine the policies and nuances that exist within the organization in accordance with the leadership style they have. Leadership style is a behavioral etiquette used by a person when that person tries to impact the behavior of others. Therefore, it becomes important for leaders to assess their own behavior based on ethics before judging or influencing the behavior of others. Another reason is that the effectiveness of internal control comes from within the people who design and implement it. In the concept of the tone at the top principle, the leadership style that reflects this principle is the participative leadership style, nurturant leadership style, bureaucratic leadership style, and task-oriented leadership style, while the authoritarian style leadership style is not included in this context because it tends to use command and control that is no longer suitable for today's work environment, which is more based on trustbased relationships. This is in accordance with the concept in attribution theory in order to change someone's behavior they need to be given or they need to be perceived internal attributes more than external attributes.
Fraud prevention can be done by creating a conducive control environment. According to Murtanto (2005) the control environment (control consciousness) shows the atmosphere or atmosphere (sets the tone) in an entity/company that affects the conscious control of the people in the organization. To create a control environment that is conducive and aware of the risk of fraud, top management or in this case the leadership must create a tone at the top related to risk management through communication and expectations. As stated in the research of Braumann, Grabner & Porch (2020) that "tone from the top in risk management does not work in isolation but rather in a system with other, more formal, and with control practices". This statement confirms the research results that the principle of tone at the top has a significant effect, but also has a direct effect on fraud prevention. It is said to have a direct effect because this principle is applied in the agency's internal control system, namely as part of the control environment. Hence, from a systems perspective, the control environment is indirectly designed to investigate or examine the role of top management in creating fraud risk awareness. Thus, hypocritical behavior can be avoided and in itself can prevent fraud.
The loading factor value is used to assess how much the indicator contributes to the variable. In variable Y, the indicator with the biggest loading factor value is SO1. This shows us hint that the organizational structure makes a major contribution to fraud prevention. However, based on descriptive statistics, the mean value of SO1 indicates that management should concentrate more or develop the concept of this indicator to achieve the objectives of fraud prevention. Meanwhile, the indicator with the lowest loading factor value is INE1. This shows that leaders who act according to what is said make little contribution in terms of preventing fraud. However, the high mean value of this indicator indicates that there is a need to increase awareness of these principles and concepts in order to be able to take fraud prevention measures. Meanwhile, there are also index that have a loading factor value and a small mean value, namely the FMGK1 indicator. This means that the leaders in 4 State Universities in Kupang City have not always emphasized the importance of achieving internal control objectives to employees. Thus, the control system and the people who run it must be able to emphasize the importance of internal control to employees.
The results of the R Square calculation show that there is a 43.8% contribution to the Tone at The Top Principle on Fraud Prevention. According to the relationship between the loading factor value and the mean value, it can also be explained what factors are the contribution. The indicators that represent the Tone at The Top Principle variable have a balanced loading factor value and mean value or in other words the loading factor value is high and the mean value is also high. So, it can be said that respondents or leaders in 4 PTNs in Kupang City agree with the statements related to indicators and have implemented statements as a reflection of the tone at the top principle in preventing fraud in PTNs in Kupang City.
The indicators in question are nurturant leadership style, participative leadership style, and bureaucratic style leadership style. Nurturant style shows that the leader provides guidance, schedules work and respects employees who do the work on time, does not "golden child" certain employees and the leadership also mingles with employees. Then, the Bureaucratican be style indicator is in the statement that the leader places company regulations as an orientation in carrying out tasks. The last indicator is Participative Style in the statement that the leader is involved in communication to determine the role of employees.
Based on the value of loading factors after being eliminated, which can be located in Figure 4.6, it shows that the measuring indicators used in assessing the effect of variable X on variable Y are indicators BS1, NS1, NS2, NS3, NS4 and PS2. Of these indicators, the indicator that gives the largest contribution is the NS3 indicator. This shows that the leadership is aware that risk awareness will not arise if there is a special behavior given to certain employees. Meanwhile, the indicator with the lowest loading factor value is NS4, meaning that the leadership has not fully mingled with the employees.
The BS1 indicator indicates that the leadership places organizational regulations as an orientation in carrying out tasks. This shows the leader is aware of every activity and behavior. existing and related to the agency must be based on organizational regulations. How effective the implementation of the regulation depends on how the leader supervises the organization's systems and activities and also how aware employees are of the importance of the regulation as a limit to behavior.
Indicators NS1, NS2, NS3, and NS4 indicate that leaders provide guidance, schedule work and reward employees who do work on time, do not "golden children" for certain employees and leaders also mingle with employees. The leadership needs to pay attention to the fulfillment of the element of compassion that is natural and is in the system.
The PS2 indicator indicates that the leader is involved in communication to determine the role that employees should play. Building communication and trust is part of the organizational culture that can prevent the risk of fraud. Thus, the leadership also takes a management approach to open communication and escalation of risk management.

CONCLUSION
Based on the outcomes the analysis of the effect of the tone at the top principle in preventing fraud at State Universities in Kupang City, it can be concluded that the tone at the top principle has a crucial and direct effect on fraud prevention at State Universities in Kupang City. Tone at the top principle has a direct and significant effect because this principle is applied in the agency's internal control system so that it is an inseparable part between top management, employees and the control system in creating awareness of fraud risk.