Factors Influencing Intention to Participate in Tax Evasion: Individual Perspectives

The study aims to examine the determinants that influence participation in tax evasion by individuals. This study uses religiosity and perception towards the government elected in the previous general election (GE14) as factors that may influence the individual’s behaviour to participate in tax evasion. This study collected data from primary sources using 500 questionnaires distributed manually as well as questionnaires distributed through a google form. The respondents of this study are individuals in Shah Alam, Malaysia. By using regression analysis, this study finds that religiosity and perception towards government have a significant influence on tax evasion. The implication of these results is that the government may design policies to prevent tax evasion from any religions or belief and provide a better solution to encourage the taxpayers to obey and support the government’s decisions.


Introduction
A tax is a compulsory financial charge or fees that are enforced by the government and collected from individuals or corporations. According to Tabandeh et al. [16], tax is an essential component of a government's revenue, and the government's expenditures are financed from the tax collection. There are a number of broad classifications of taxes, i.e., individual tax, company tax, stamp duty, real property gains tax and sales and service tax. However, the taxes that are most directly related to individuals in Malaysia are individual income tax and sales and service tax. In Malaysia, the effective year 2015, every individual who earns more than RM34,000 per annum after deducting the employees' provident fund is liable to register a tax file with the Inland Revenue Board of Malaysia (IRBM).
Besides, sales and service tax also has an indirect impact on citizens but not on their level of income. Sales and service tax, currently known as SST 2.0, was reintroduced on September 1, 2018, to replace the goods and service tax (GST). Under the new SST 2.0, goods are charged a sales tax of between five to ten per cent, while services are charged a service tax of six per cent. The SST2.0 is imposed on certain goods and services, for examples, live animals, food items, medicine, construction materials, general goods, selected vehicles, and most capital goods [8]. However, some issues have been raised because this initiative does not meet the citizens' expectations. Hence, they tend to make aggressive tax avoidance to reduce their individual's income tax.
The government has gone ahead with the public funding reform initiatives under the fiscal transformation programme to make the government fund stronger. Income tax is the most extensive part of the government's revenue, aiding over 50 per cent annually, to fund the nation's development facilities. As shown in Figure 1, government revenues had declined from January to July 2018. During  Figure 2, all taxes, i.e., SST/GST and direct taxes revenue declined from 2017 to 2019. On April 17, 2019, the New Straits Times reported that, barely 15 percent out of 15 million adults with occupation in Malaysia, pay income tax to the IRBM 1 . It also indicates that less than three per cent of those who are registered under IRBM are paying tax. It can be considered that most of the taxpayers are still attempting not to pay tax. In managing the fiscal policy, the government plays a vital role. By having an efficient well-designed tax system and the revenue collection, the government can give back to the citizens in the form of quality public services and infrastructures and people-centric programmes and projects 2 .
In November 2018, the Malaysian government had announced a voluntary disclosure programme to allow more taxpayers to register and declare their income to the IRBM. This voluntary disclosure can be made at any time commencing from November 3, 2018, until September 30 2019, which is expected to broaden the tax base and reduce the tax gap 3 .
Religiosity has become one of the debatable topics amongst past researchers [10,15,17,19]. According to Benk et al. [1], religiosity has a least possible, but significant, positive impact on taxpayers' attitudes. However, there are limited studies on the effect of religiosity on tax compliance, and they were mostly done in developed countries. Religiosity might have its power to affect the moralistic action amongst taxpayers to act in accordance with the tax system. According to Srivastava et al. [13], religion develops an individual's tolerable behaviour to abide by the principles and codes of manner drawn from religious scriptures such as the Bible and the al-Quran. Hence, it can be concluded that the more committed the believer is to his/her religion, the more responsible the individual is to comply with regulations.
History was made after 61 years since the independence of Malaysia. The government changed hands from the United Malay National Organisation (UMNO), led by Dato' Sri Mohammad Najib bin Abdul Razak, to Pakatan Harapan (Alliance of Hope) after the 14 th General Election on May 9 2018. Tun Dr. Mahathir bin Mohamad, the leader of Pakatan Harapan, was sworn in as the 7 th Prime Minister of Malaysia on May 10 2018. The biggest story that came out of the change in Malaysia's government in recent years was the corruption among the 'higher-ups'. This issue of corruption among politicians has been damaging to Malaysia's reputation [22]. Since the changes in government are very recent, there has been no study done The scenario of the change in the government makes things clearer in which the public trust is essential for the government to get supports. Citizens' optimistic attitudes on the government will intensify their determination to comply with tax regime [13]. However, many recent corruption cases in Malaysia happened, include VIPs. The tacit consequence of corruption is massive, and, at times, it also reduces the performance of public services, raises hardship among citizens, and creates hurdles in accomplishing societal welfare [14]. According to James et al. [7], a more corrupt society may allow more intention to participate in tax evasion as corrupt officials seek more income via bribes. As compared to past studies, this study focuses on two factors influencing individuals to participate in tax evasion, namely, religiosity and the people's perception of the new government of Malaysia.
The objective of this study is twofold. First, this study aims to explore more on the individuals' tax compliance by adding the new variables to examine the factors influencing tax evasion, which are religiosity and the perception of individual taxpayers towards the government. Thus, this study contributes by providing insights into the link between the factors influencing individuals to tax evasion. Second, this study aims to obtain opinions from the respondents about their perceptions of the government concerning tax evasion.
The remainder of this paper is structured as follows. First, the literature review and hypotheses development are discussed in Section 2. In a subsequent section, Section 3, the research design and method is described. The results and findings are presented and discussed in Section 4. The paper concludes with some concluding remarks, limitations and future avenue of research in Section 5.

Literature Review and Hypotheses Development
Taxation is the primary source for a country to generate revenue for the government, who, in turn, uses this money to fund its operations to serve the public better. Malaysia's tax system is the government's primary source of revenue. The tax revenue constitutes 73.9 per cent of the total revenue, or 12.2 per cent of Malaysia's gross domestic product (GDP), while non-tax revenue represents 26.1 per cent of the total revenue or 4.3 per cent of the GDP 4 . According to the latest figures, Malaysia's tax revenue, including sales tax revenue, stands at USD 10.833 billion. The amount of Malaysia's tax money is used to improve public welfare such as fund education, home security, public facilities and defence.
Income tax in Malaysia is imposed on income accruing in or derived from Malaysia except for resident companies carrying on business in air/sea transport, banking, or insurance that is assessable on the world income scope 5 . Tax rates charged on the taxable income will increase with an increase in the individual's taxable income for that year of assessment. Therefore, the higher the chargeable income, the higher the final amount of tax payable is.
Attitude and action from the individuals are derived from their belief-religiosity as the interpersonal attribute, which affects their action. A more religious person tends to trust others, human, organization and government [4]. This also might affect other people's attitudes. From a spiritual point of view, tax evasion gives different perceptions among different religions which have not always been justified, and it can be said that there is no absolute opposition to tax evasion, or there are still situations where one can justify it up to a certain extent.
According to Srivastava et al. [13], religiosity and attitude toward the government influence the customers' indirect participation in tax evasion. These two factors have important implications for the ethical decision-making process of a consumer. Hence, both factors seem interrelated each other, which the intention to perform an attitude is from their belief and trust towards others.

Religiosity and Tax Evasion
The relationship between religiosity and ethics is sometimes unclear, which religiosity may be changed [6,24]. According to Benk et al. [1], the religious values intrinsic in Turkish Muslim individuals have a significant favourable influence to shape their attitudes in tax-paying, either voluntarily or mandatorily enforced by the tax authority. Swimberghe et al. [15] found that a person with high religiosity has a tendency to adhere to the rules and teachings and obey the laws and conducts of his/her religion more diligently. This tendency would explain that religiosity positively influences tax-paying behaviour.
Guiso et al. [4] found a positive relationship between religiosity and trust in peoples and institutions. Guiso et al.'s [4] finding was contradictory to that of Wadim and Inna [23], who found that, in the Czech Republic, belief and the importance of God in one's life has no significant effect on the tax morale. It means that religiosity does not impact the tax morale among individuals significantly. While Walker et al. [24] and Clark and Dawson [2] found a negative association, Wong [25] found a weak relationship between religiosity and tax evasion. Religiosity refers to the state of being religious. Individuals with higher religiosity seem to be more intellectual, have higher moral dimensions, and tend to follow the rules.
Therefore, based on these arguments, this study proposed the following hypothesis: 5 Retrieved from PWC website on 29 th October 2019 at https://www.pwc.com/gx/en/services/tax/publications/paying-taxes-2020. html H1: There is a significant relationship between religiosity and individuals' intention to participate in tax evasion.

Perception towards Government and Tax Evasion
According to Torgler [20], attitude and trust towards the government are the critical factors of tax-paying behaviour. Trust is one of the fundamental elements in the understanding of tax evasion. Rothstein and Uslaner [11] justified the type of trusts, such as social and generalized trusts, that was important and became the reasons why people tend to pay tax. This finding explained that people's faith is dependent on how others react to a circumstance. It is found that individuals are willing to obey tax compliance if they trust that most other citizens would pay the same. According to Wadim and Inna [23], trust in the government and attitude towards the government seems to have no significant effect on tax morale. It means that belief in the government appears to have no impact on the tax attitude of individuals. On the other hands, other past empirical findings suggest that trust towards the government, authorities, and the legal system has a significant positive effect on tax morale [e.g., 12,17,18]. Torgler [17] also found that trust in the government had a highly significant positive effect on tax morale.
Crowe [3] found that the most common reasons why people justify tax evasion were unfairness of the tax system, government corruption, wasting of tax funds, not receiving any benefit, and the inability to pay taxes.
Therefore, based on these arguments, this study proposed the following hypothesis: H2: There is a significant relationship between individuals' perception towards government and their intention to participate in tax evasion.

Research Method and Design
This study uses a simple random sampling method to select the sample. Five hundred questionnaires containing twenty-five questions were distributed to the individuals' resident in Shah Alam, Malaysia. From the five hundred questionnaires distributed, this study expected about seventy-five per cent rate of return. According to Salkind [9], to perform a quantitative method, the minimum sample size should be thirty, and the maximum is five hundred.
The samples were collected among taxpayers registered in Klang Valley, Shah Alam, Malaysia. Department of Statistics reported that in 2018, Selangor has the highest growth in manufacturing industries with 7.3 per cent. Shah Alam is estimated to have a population of 650,000 and 29,026 industry units all over the city. These figures make Shah Alam one of the most significant tax contributors in Selangor 6 . On average, information that was received from the officer in the Inland Revenue Board Shah Alam Branch, 356,000 taxpayers are staying in Shah Alam. As the capital city in Selangor, possibly respondents in Shah Alam will provide data intended to answer the objective of this study.
The questionnaires were divided into four sections. Section A includes information on the independent variable, religiosity. Section B provides information to measure perception towards the government, the other independent variable. Section C contains information about the dependent variable (tax evasion), and Section D includes personal information about the respondents (demographic information). The response to 24 of the 25 items in Sections A, B, and C of the questionnaire was measured based on a 5-point Likert-scale (1 = strongly disagree to 5 = strongly agree). One of the items in Section B is an open-ended question.
Data were collected in two ways, namely, via Google form (online) and manual distribution of questionnaires to taxpayers in selected companies around Shah Alam. More than four weeks (from May 10, 2019, until June 12 2019) were needed to gather all the information to achieve the minimum requirement for the analysis. The follow-ups were done weekly on every Wednesday of each week. The two forms of questionnaires distribution were carried out simultaneously to minimize the time consumed for data collection. Overall, 284 questionnaires were returned online, and 118 out of 500 questionnaires were returned manually. The total of the respondents is 408. However, only 402 are valid for data analysis. The data were processed manually in the Microsoft spreadsheet before they were transferred to the SPSS software. Table 1 shows the demographic profile of the 402 respondents. The background of the respondents is comprised of gender, age, marital status, families, annual income, ethnicity, and qualification. The personal backgrounds of the respondents are essential to indicate their personal commitment, age, and annual income that would be eligible for income tax. The respondents' academic qualification would demonstrate their personal knowledge about tax compliance.

Results and Discussion
From Table 1, it shows that from the 402 respondents, 50% of them are female, and another 50% are male. Most of the respondents (44.8%) are aged between 26-35 years old, followed by 29.9% within the range of 36-45 years old. The majority (76.4%) of the respondents are married. As referring to the number of children, the majority (51.7%) of the respondents are having between 1 to 3 children. The frequency for ethnicity indicates that the respondents come from a variety of ethnic; namely, Malay, Chinese, Indian and Others with the majority (58.8%) are Malays.
According to the IRBM, a person with an annual income of RM34,001 is liable to pay tax. Hence, the results show that 97 per cent of the respondents would be liable to pay tax. The majority (70.4%) of the respondents hold a degree qualification, while 16.2% and 0.7% of the respondents have higher academic qualifications, namely, Master degree and PhD, respectively. Information on income and educational qualification together indicates that the respondents are knowledgeable about tax compliance and aggressive tax reduction. Table 2 shows the descriptive analysis to measure the central tendency and variability of the dependent and independent variables. It presents the mean, standard deviation, minimum and maximum scores of the data set. It also shows the kurtosis and skewness of the data to indicate the normality of the data collected. Table 2 shows the average mean for the variables ranges from 3.44 to 4.18. This suggests that all of the respondents agree with the statements for all the variables measured.
The result for the dependent variable (tax evasion) presented in Table 2 shows the minimum score and maximum score of 2.40 and 5.00, respectively. The mean value of 3.79 represents the central tendency of the data. As for normality, the table shows that religiosity and tax evasion are negatively skewed, where the scores are clustered at the higher end and the tails point towards the lower or more negative scores. Meanwhile, the perception towards the government is positively skewed with scores clustered at the lower end and the tails point towards the higher or more positives scores. Since the skewness values of the data are between 0 and 1.0, the data could be considered normal. The kurtosis values are within the range of ±1, showing that the data are normal. 108 Factors Influencing Intention to Participate in Tax Evasion: Individual Perspectives

RELIABILITY ANALYSIS
The reliability statistics are used to assure that the variables are free from errors and, therefore, yield consistent results. The construct reliability test is conducted using Cronbach's alpha method to assess the reliability of the variables measured using the likert-scale. The results are shown in Table 3.  Table 3 shows the reliability statistics for the variables. The Cronbach's Alpha is 0.771, which is acceptable, and it implies that the 24 scaled items are internally consistent.
The summary items statistics in Table 4 show that the mean score of the items is 3.918. This result indicates that the respondents are moderately inclined with the statements in the questionnaires. The minimum score for the items is 2.80, and the maximum is 4.72.

NORMALITY TEST
This section explains further the normality test for the dependent variable, tax evasion. It comprises the minimum and maximum bound, mean, standard deviation, variance, skewness, and kurtosis for the 402 respondents to assess the likelihood of a random variable underlying the data set to be normally distributed. Table 5 replicates the results of the descriptive statistics and normality test of the dependent variable shown in Table 2 for easier reference.
This study evaluates the data normality based on the skewness and kurtosis. Normal distribution data exists if each of the skewness value is between -2.0 and +2.0, whilst the kurtosis should fall between -7.0 to +7.0 [5]. Table 5 shows that the dependent variable (tax evasion) has a leptokurtic kurtosis with a score of 0.083 and skewness of -0.282, which these values are within the range of the threshold points. Therefore, the data in this study are normally distributed.

FACTOR ANALYSIS
This section discusses the results of the factor analysis. This technique is used to reduce many variables to a smaller set of underlying factors that summarise the essential information contained in the variables. Before doing the factor analysis, Kaiser-Meyer-Olkin (KMO) and Bartlett's tests were conducted to assess the adequacy of the data for factor analysis. Table 4.6 shows the results for Kaiser-Meyer-Olkin (KMO) and Bartlett's tests.
The KMO value for this study is 0.784, which falls into the range of "good". Therefore, this study is convinced that the sample size is adequate for factor analysis. In addition, Bartlett's test of sphericity is significant (p-value < 0.001). Factor analysis was conducted using the Rotated Component Matrix to maximize the loading of a variable on one factor while minimizing its loading on all other factors. Table 7 shows the results of the rotated component matrix.   Table 7 shows that all the questions in the variables being separated in 7 Factors in which, they are differently ranked. The highest rank which represents religiosity is REL7: I rarely join the activities of my religious organization. Hence, this reverse question shows that religiosity has a relationship with tax evasion. To answer the next independent variable, all the items that load highly on the ranging between 0.834 -(-.0784). The highest rank question for perception towards government is PTG2: In my opinion, the way the government spent the taxpayers' money is reasonable and appropriate. The items show the highest answer, which indicates that tax compliance has positive with perception towards the government. The dependent variables are ranging from 0.457 -0.761. The highest question rank is TE3: I am already informed that tax evasion is an illegal action. Hence, this shows that all the respondents have knowledge about the tax system and tax evasion. The results of the factor analysis indicate that the items in the questionnaire fit the variables for which they were intended to measure (i.e., religiosity, perception towards the government, and tax evasion).

CORRELATION ANALYSIS
The correlation coefficient among the variables in this study was measured using Pearson correlation. For the correlation analysis, the data were run through a bivariate correlation that helps to answer the research questions. The results are shown in Table 8.
The results in Table 8 show that religiosity is positively correlated to tax evasion with a Pearson correlation coefficient (r) of 0.400, and the significance value is less than 0.001 (p-value < 0.001). These values indicate that there is a significant positive association between religiosity and tax evasion. The results also show that perception towards the government has a significant positive association with tax evasion (r = 0.175, p-value < 0.001). Table 9 presents a summary of the model for this study. The results show 17.3 per cent of the tax evasion variation is explained by the variation in religiosity and perception towards the government (R 2 = 0.173). Meanwhile, 16.9 per cent of the variation in tax evasion is explained by religiosity and perception towards the government after considering the sample size and number of independent variables (Adjusted R 2 = 0.169). The analysis of variance, ANOVA, was also conducted to test further the existence of relationships between the dependent and independent variables. Based on Table 10, the F-value for two independent variables (religiosity and perception towards government) is 41.744, which is greater than F critical (F (2,399) = 3.01). This F-value indicates that there is a significant linear relationship between tax evasion and the two independent variables of religiosity and perception towards the government.  The results as per Table 11  The coefficient indicates that for one percentage point change in the tax evasion (TE), the religiosity (BREL) will increase by an average of 36.1 per cent, and perception towards the government (BPTG) also increases by an average of 12.1 per cent. These values denote again that religiosity and perception towards the government have a positive relationship with tax evasion. As shown in Table 4.11, the t-value for religiosity is 8.289, which is greater than t (399) of 1.968, and the p-value is less than 0.000 (p < 0.001). These values indicate that there is a significant linear relationship between religiosity and tax evasion. H1 is accepted since there is evidence that religiosity could affect the tax evasion.

REGRESSION ANALYSIS
Concerning the perception towards the government, the t-value is 2.496, which is greater than t (399) of 1.968, while p-value is less than 0.001. These values indicate that there is a significant linear relationship between perception towards the government and tax evasion; hence H2 is accepted. Besides, the evidence that perception towards the government could affect tax evasion, the hypothesis. Based on the 5 per cent level of significance, and the 95 per cent level of confidence, religiosity has 0.000 level of significance and perception towards the government has 0.013 level of significance. Hence, it can be concluded that religiosity and perception towards the government have a significant relationship with tax evasion (p < 0.05).
In order to support the second objective, the analysis of the opinions by respondents shows mixed perception towards the government, which may be related to their response of the overall questionnaire. From 32 respondents, those who highly agree with the statements given in items 1 to 9 of Section B gave a positive perception compared to those who disagree. Among the opinions were "I support the government and believe that government will improve from the previous crisis" and "New government tell us directly regarding expenses than the previous one." Some of the respondents were neutral on their perception but still have trust in government, e.g., "The government can only expect taxpayers to be sincerely accountable if they are sincerely accountable with their spending." Therefore, it was concluded that the respondents who trust the government are willing to support whatever decisions made by the government.
However, some respondents have a lack of trust towards government. The statement is as follows: " These types of opinions indicate that trust towards the government has little impact on the attitude and beliefs of the citizens. Some specific individuals disagree with the government but are, somehow, still committed to paying tax.
"The current government has not been transparent and seems to be worse compared to the previous government in spending the tax money. However, I still believe that it is my duty as a citizen to pay my taxes; however much I distrust the current government." This outcome is supported by Chan and Mo (2000), which stated that people comply with the tax law because of their perceived moral obligations to obey the tax laws.
Also, some opinions indicate that religiosity is also interrelated to their perception towards the government.
"New government totally discriminate people based on racial. There is no concept of multiracial in our country if the leaders also implement this kind of behaviour." "….religious problem getting worse day by day…I have no trust in the government." These respondents believed that racism might reduce their trust in the government. The issues regarding religion may also influence their faith in the government. These opinions support the relationships between both the independent variables (religiosity and perception towards the government) and intention to participate in tax evasion.

Conclusion
This study highlights the two variables, religiosity and perception towards government and participation in tax evasion. First, this study provides an insight into the current issues of the government. The results in this study would raise the awareness of the government bodies on the probability of the citizens accepting the government. They would also assist the government in understanding the citizens' behaviour towards tax compliance so that they would be encouraged to pay the taxes owed willingly. Understanding the citizens' behaviour is also necessary to increase the government's accountability and integrity. The increase in responsibility and integrity is essential to raise the citizens' trust in the government's performance.
Most of the past research on tax evasion looked at external factors such as economic factors that impact the government fund. Nevertheless, this study sought to divulge the internal factors that may significantly cause tax evasion. The findings of this study could help the government to assess its current governance structure for enhancements on more effective and efficient governance. Consequently, preventing the possible risk of the individuals participating in aggressive tax avoidance. In addition, this study would assist the policymakers, especially the tax authority handling the tax revenue, to identify the factors of individuals' participation in tax evasion, which would enable them to enhance the effectiveness of the tax system. By identifying the factors influencing tax evasion, the tax authority would also be able to identify the necessary corrective actions required to gain the trust of the individual taxpayer.
This study also added to the empirical research on the accountability of the government in employing government revenue. Since the government acts as an agent to the citizens of one country, it has to maximize citizens' welfare and to spend the money wisely. Thus, this study may help the government to ascertain remedy needed to improve trust by the public, which, implicitly, would give rise to the tax revenue.
There are a few limitations of this study. Firstly, the mechanism of identifying the number of respondents suitable as the study sample. Although there are thousands of individual taxpayers registered in Selangor, only 500 taxpayers in Shah Alam were selected as the sample. Furthermore, the number of taxpayers registered in Shah Alam is only an estimated number. Thus, the representativeness of the sample might be compromised, which might also deteriorate the accuracy of the overall result. Besides, the process of sample selection unavoidably involves compromise since the taxpayers are individuals from various races and religions. In order to meet the main objective of religiosity, this study had to monitor the number of sampling units to ensure a fair number of responses were received from the respondents of different religions. The gathering of data from the IRBM Shah Alam branch also made it more difficult due to the lack of data available on its website. It should be noted the different sample sizes may yield different results, although the research objectives may be the same.
The results of this study are promising for future research. First of all, there are rooms for the upcoming studies to compare the results two to five years later to see how far the government may have increased the taxpayers' and other citizens' trust in the government via its performance. As 2018 was the first year of change in the new government, future studies may expand the assessment of the perception towards the government over a more extended period. Second of all, future studies may also examine the changes in tax compliance and tax evasion from November 2018, when the government announced the voluntary disclosure campaign that encourages the taxpayers to, not only declare the actual amount of the chargeable tax imposed to them but also to promote the eligible taxpayers to register with IRBM.
Next, future research may also use qualitative method to extend the research on the variables used in this study since the quantitative method may limit the real feedback from the respondents on their perception towards the government. Currently, the theory that is widely used to explain individuals' behaviour is the theory of reasoned action. Thus, future research might want to focus on different approaches to analyze the effect of the behaviour of individuals on aggressive tax reduction. Furthermore, researchers might consider extending the period of data collection and widening the scope of the research to get more complete results.