The Influence of Religiosity on Tax Morale in Indonesia

(1)

The Influence of Religiosity on

Tax Morale in Indonesia

Master Thesis Public Administration

Universiteit Leiden

Faculty of Governance and Global Affairs

Harry Novalino

S1783068

13 March 2017

(2)

1

ABSTRACT

The main objective of this study is to examine the influence of individuals’ intrinsic value, namely religiosity, as well as external values, namely perception of government, interaction with tax authority, social interaction, and threat of punishment on the tax morale of taxpayers in Indonesia. The finding of this study is expected to contribute to the current literature of tax compliance discussion, especially in developing countries.

The data is collected through online self administered survey that involves 300 respondents from various provinces in Indonesia. Majority of respondents are salaried taxpayers and muslim; a reflection of population in Indonesia. The findings of this study show that taxpayers’ perception of government, interaction with tax authority, religion, and social interaction significantly and statistically influence tax morale. While social interaction is found to have negative association with tax morale, three other variables is shown to have positive association. Threat of punishment, however, does not influence tax morale significantly.

The results put the importance of the taxpayers’ perception of government and tax authority in encouraging their compliance behavior. The Indonesian government then needs to address this aspect seriously and try to improve its integrity to increase tax morale. Fighting corruption may be the best feasible action in the mean time. The significance influence of religiosity on tax morale is also apparent based on the result. It means that the policy to foster the religiosity level is proven to positively contribute to compliance behavior. Maintaining and proving proper environment that allows people to practice their religion belief and teaching is one of the examples.

(3)

2

CHAPTER 1 INTRODUCTION

1.1 Background

Tax compliance without doubt is one of many factors that influence global economy. This issue has been a concern and essential question for government worldwide, including not only developing but also developed countries. Particularly in Indonesia, tax compliance remains a huge challenge for the government. In 2014, the compliance of tax return filing in 2014 was recorded only at 58.87% (DGT, 2015). It becomes more important as the mission of Indonesian Directorate General of Taxes (DGT) is to ensure the implementation of a sovereign and

autonomous state with revenue collection based on high compliance on tax voluntary and fair law enforcement.

There are enormous and rich researches in the tax compliance field. They are ranging from the ones that focus more on the level of individual to the ones that emphasize on the company; ones that dig more about non-compliance to ones that put the importance on the compliance; and ones that try to gain insight from economic aspect to the ones who try to solve the puzzle from non-economic factors. However, all of them share the same objective which is to help government, especially tax authority, to gain more revenue from taxation in order to pay the expenditure.

An earlier common theme in tax compliance studies is that the traditional economics-of-crime approach. It puts the attention to the economic factors such as tax audits, tax penalty, and tax rate. Economics-of-crime approach, while indeed containing many insights, is not enough to

(6)

5 fully understand the reason behind compliance behavior of taxpayers. This approach views taxpayers as individuals who try to maximize their utilities using the evasion gamble; weighing the benefit and success of probability of the evasion against the risk of being detected and punished is made to determine the decision to pay taxes or to evade. Therefore, our endeavor to more fully comprehend tax compliance is important. It can be done by introducing social science aspect into the field of tax, such as behavior of motivation. This study aims to provide additional attempt to review this new approach in order to move beyond economics-of-crime perspective and provide more understanding of taxpayers’ compliance decisions.

The classical tax evasion model assumes that deterrence policies are the answer to increase tax compliance because audit probability and tax penalty is positively correlated with tax compliance. The idea behind this framework is that individuals pay taxes because they afraid of being caught and penalized. Therefore, for rational taxpayers, maximizing utility can be done by matching the benefit of successful tax evasion against the probability of being detected and punished. In other words, tax evasion is a rational action, according to the economic deterrence approach, as long as the reward for evading taxes is bigger that the cost of being caught. Consequently, tax compliance as the expected outcome is only achieved because the fear of taxpayers of economic consequences of detection and punishment (Alm, Sanchez, and De Juan 1995, p. 5; Torgler 2011, p. 3). Consequently, to increase the compliance of taxpayers, tax authority has to increase the rate of penalty and the tax audit.

At first, solving the puzzle of tax compliance using economic model was favored because of the simplicity and distinct policy prediction that it provided. The purely economic analysis of tax evasion then would imply that rational taxpayers will be very likely to evade tax because the probability for them to be caught and penalized is very small. However, the level of compliance

(7)

6 in the real life does not fall in the level predicted by the theory. The simplicity that consider individuals pays taxes because – and only because – economic deterrence effect has failed to explain why the level of tax compliance is higher than the prediction, especially in countries that have high level of compliance and low economic deterrence level (low tax audit and penalty rate). For example, in 2014, tax revenue from audit process only contributes to 2.48% of total tax income in Indonesia. It shows that taxpayers’ attitude (in Indonesia, in this case) is not greatly determined by threat of punishment, in this case tax audit. In line with this example, Torgler (2011, p. 3-4) points out that widespread noncompliance behavior is rarely happen even though economic deterrence effect (tax audit and tax penalty) is quite small in most countries. These phenomena might indicate that the taxpayers’ compliance decision must be affected in ways that are not fully captured by basic economic approach.

As a result of this, variety of factors that may influence tax compliance beside economic aspect are started to be analyzed. The more recent studies in tax compliance then put the

emphasis on the economic factors. Instead of finding the answer behind taxpayers’ non-compliance, researchers started to find the answers of relatively high compliance. The existence of intrinsic motivation of individuals to behave honestly as well as provide information and pay taxes properly and correctly is suggested. Researchers used the concept of social norm of tax compliance or tax morale to call this suggestion.

Schmolders (1959, p. 341) suggests that taxation, as a form of public finance application, is suggested to develop its general principle from the root of genuine knowledge of human behavior. The factors that positively encourage taxpayers’ willingness to comply may be the reflection of integrity, good faith, and fairness shown in the tax system through principle of publicity, completeness, specialization and truthfulness. Another important point suggested is

(8)

7 that governments need to introduce moral persuasion as a means to attract voluntary compliance of taxpayers because taxation is a government program so it cannot be separated from the functioning of public relations of government.

Study on the determinants of tax morale might help to answer the ever-growing question that why people pay taxes, rather than why people evade taxes. Introducing the concept of tax morale means that governments and tax authorities may have additional possible policy choices beside coercion using economic deterrence factors to increase tax compliance. Then, in order to develop better policy recommendation, the study on the determinants of tax morale becomes important because further understanding of tax morale could lead to a variety of methods available for government to increase voluntary tax compliance.

Recent studies of tax compliance analyze non-economic factors from two perspectives. The first perspective is the internal values from taxpayers themselves, such as the value in their self or family, culture embedded in themselves from society, and religion teaching and beliefs. The second perspective is the external values, for example: the influence of society, government actions, as well as the treatment of tax authority towards taxpayers.

Religious value is suggested to have influence on shaping individuals attitude and

behavior. Therefore, it is expected to encourage positive behavior and prevent negative behavior in their daily life. Consequently, religiosity is presumed to positively influence voluntary

compliance towards tax laws while at the same time negatively affect deviant acts such as tax avoidance. In other words, religiosity is possibly to pose as one of possible explanations behind the relatively high level of positive compliance attitudes as shown in existing researches. The abovementioned discussion leads to my research question:

(9)

8

RQ. What is the influence of religiosity, as internal factors, and other external factors on taxpayers’ compliance?

The objective of this study is to empirically assess the relationship between religiosity as internal value of taxpayers on the tax morale of taxpayers in Indonesia. This study tries to address the influence of religiosity in the formula of taxpayers’ compliance, while including other relevant non-economic factors; perception of government, interaction with tax authority, social interaction, and threat of punishment.

1.2 Justification

The trend of tax compliance studies has shifted from emphasizing on the economic factors to deter non-compliance behavior of taxpayers, such as tax evasion, to stressing more on the non-economic factors that may influence the willingness of taxpayers to voluntarily comply with tax laws, such as tax morale. However, there is limited evidence particularly from the perspective of religiosity to be the aspect that influences compliance. Riahi-Belkaoui (2004, p. 141-142) recommends that social and religious norms should be included in future research to provide a clear picture of tax morale in building a complete theory of tax compliance and contribute to a better public policy in this field.

Religiosity is specially chosen as dependent variable due to several reasons. This study is intended to explore tax morale in Indonesia which is known as the country with the biggest muslim population. Based on the national survey of Indonesian Statistic Agency, in 2010 the

(10)

9 proportion of muslim people in Indonesia is 87.18 percent1. The census also shows that more than 99 percent of Indonesian populations are adhered to a religion. Therefore, considering the fact that religiosity is have not received much attention in tax morale research and the rate religion adherence in Indonesia, acknowledging its influence in tax morale formulation is preferred

This study then not only adds to a literature on the influence of religiosity on tax compliance, but also expands the study of limited research in Indonesia regarding tax compliance, such as Iskandar, Bhaduri, and Wunscher (2016). While they try to find out the influence of economic factors to a specific tax law namely environmental tax, this study examines the influence of non-economic factors, religiosity, to tax morale of Indonesian taxpayers.

Furthermore, this thesis also contributes to the field of tax research by using real

religiosity commitment that might influence taxpayers’ compliance. This study is similar to the previous study from Mohdali (2012) that examines the effects of religiosity on taxpayers’ compliance in Malaysia and Benk et al. (2016) that studies Turkey’s taxpayers’ compliance attitudes. The former studies about religiosity and tax compliance includes one from Leonardo (2011) that uses ideology rather than religiosity, and Grasmick, Bursik, and Cochran (1991), and Eiya, Ilaboya, and Okoye (2016) that do not specifically group religiosity item to intra and interpersonal value.

This study uses survey to Indonesian taxpayers regarding their individual religiosity and compliance attitude. This method is suggested by Richardson (2008, p. 76) that recommends future research to investigate the relationship between culture and tax evasion in countries

individually using appropriate survey methodologies. Furthermore, McKerchar and Evans (2009,

(11)

10 p. 182) emphasize the need of research and practical experience to foster voluntary tax

compliance and enhance taxpayers’ morale in developing countries.

The findings from this study are expected to give better understanding to Indonesian tax authority (DGT) regarding key variables that might be significantly associated with compliance. I also hope that DGT could design and implement appropriate strategies in tax compliance program using the evidence from this thesis. Specifically, this study may give assistance to DGT because it is in line with their mission to ensure the implementation of a sovereign and

(12)

11

CHAPTER 2 LITERATURE REVIEW

2.1 Neoclassical Model of Tax Compliance Study

The classical view of tax compliance can be seen in the definition by Roth, Scholz, and Witte (1989, p. 2) which define tax compliance as a condition where taxpayers file all their taxation responsibility at a proper time with correct calculation of their tax liability according to rules, regulation, and court decision applicable at the time when they file the tax return. The aforementioned definition only treats compliance in regard with the tax return obligation; there are other studies that have different views regarding compliance. James and Alley (2000, p. 29) endeavor to describe compliance not only from the tax gap approach, but also focus on the behavior of taxpayers. In line with this, the definition from The Internal Revenue Service (IRS) also includes the intrinsic motivation from taxpayers themselves. IRS defines tax compliance as “a system of compliance that relies on individual citizens to report their income freely and voluntarily, calculate their tax liability correctly, and file a tax return on time”.

If tax compliance can be considered as positive attitudes of tax commitment, then tax avoidance and tax evasion could be seen as deviant acts of the commitment. The distinction of those two action stems from the legality of taxpayers’ action. Tax evasion, according to Sandmo (2005, p. 645) is a violation of law by, for example, refraining from reporting taxable income from labor or capital or by claiming ineligible deduction. On the other hand, tax avoidance is associated with reducing tax liability within the legal limitation of tax regulation (Sandmo 2005, p. 645) or by understating the liability due to the misunderstanding caused by complexity or

(13)

12 ambiguity of tax laws (Slemrod 2007, p. 26). From this, we can see that tax evasion could be done through two ways; some manipulate the ambiguity of tax laws to reduce the tax burden while others might interpret the laws according to their understanding.

Moving further into a broader perspective, the definition of compliance from James and Alley (2000, p. 29) extends the definition of compliance from traditional approach which initially focuses on economic aspect of tax gap to the wider approach which includes taxpayers’ behavior. Added to the definition from IRS, it shows that the traditional approach is started to blend with socio-economic and psychological factors. This blending of economic and non-economic factors reflects the change in focus of studies from the concern of taxpayers assumed to evade tax to a focus on the willingness and voluntary compliance of taxpayers. Furthermore, the blend might be needed to completely understand the unique behaviors of taxpayers.

The starting point of neoclassical model for the literature on tax evasion is developed from the model by Allingham and Sandmo (1972). The model is an extension from the

economics-of-crime framework developed by Becker (1968) to the field of taxation. According to the models, taxpayers will determine their compliance decision based on the cost-benefit calculation. They will weigh the benefits of evading taxes versus the cost of being caught. Included in the economic deterrence approach are tax rate, tax penalties, and prospect of being audited. They are initially viewed as effective instruments to threaten people not to avoid taxes as evident in a number of studies (for example see Becker, Büchner, & Sleeking, 1987; Beck, Jon, & Jung, 1991). The underlying assumption of this approach is that higher tax penalty and audit probability are associated with less tax evasion. Another foundation in this model is suggested by Alm and McKee (1998, p.260); tax compliance decision is made strictly due to the fear of the possibility of being caught and penalized. Thus, the predictions in this model are

(14)

13 straightforward; tax evasion could be minimized by increasing tax penalties and audit probability (control).

There are a number of theoretical and empirical evidence to support the view that higher penalties and audit probability encourage compliance. One of the initial researches in the field is conducted by Allingham and Sandmo (1972) which suggests that deterrence policy could be used to ensure that people pay their taxes. The usage of threat of punishment, including audit probability, severity of penalty, and increasing of tax rate, could positively increase compliance.

However, there also other studies that point towards other direction. Witte and Woodburry (1985, p. 9), for example, suggest that incremental penalty have no significant impact on tax compliance and tax audits have an insignificant impact on salaried taxpayers. It might indicate that threat of punishment through tax penalty and audit may be effective only for certain group of taxpayers. Therefore, it might not be appropriate to rely heavily on the economic deterrence approach to solve the tax compliance problems.

In accordance with the idea, Slemrod (2007, p. 45) also suggests that deterrence as an element of compliance decision-making process by taxpayers, but we should also recognize the importance of moral values in each individual that might shape taxpayers attitude and behavior. In addition, Alm et al. (1995 p. 15) also describe that in order to full explain tax compliance we require various factors that may motivate taxpayers behavior that go further than standard economic approach to include, for example, theories of social science. In another study, James, Murphy, and Reinhart (2005, p. 188) explain that it is better to explore other determinants to fully understand non-economic factors that may influence taxpayers’ behavior rather than relying heavily on economic approach to encourage compliance. The willingness of taxpayers to comply with tax regulation, for example, could be better explored by using social psychological

(15)

14 approach. To sum up, these studies indicate that deterrence model cannot be used solely to explain the reason why many people comply with tax laws voluntarily.

It has also been argued and realized that, given that the level of taxpayers’ compliance is relatively high compared to the actual rate of tax audits and penalties, the real challenge in the puzzle of tax compliance is not to answer to question why people evade tax but why people comply (Alm, Sanchez and De Juan, 1995, p. 3). This reason shifts the focus of tax compliance studies from economic deterrence approach to social psychological approach in order to gain a better understanding of taxpayers’ behavior by considering non-economic factors. While

neoclassical framework believes that economic factors are the main determinants of compliance decision of taxpayers, social psychologist theorists have different view. Most of the studies that fall on the social psychological framework focus on the non-economic elements that might influence taxpayers’ decision making process, such as peer influences and individual personal attitudes (for example Kaplan & Reckers 1985).

2.2 Social Psychological Model of Tax Compliance Study

Social psychological approach is a result of a shift in the focus of the attempts to explore tax compliance issue from different angles other than deterrence economic approach. James and Alley (2000, p. 33) suggest that individuals should not be viewed only as independent and selfish beings who always try to maximize their utilities; they also interact each other according to differing attitudes, beliefs, norms, and roles.

The studies of tax compliance from non-economic framework examine not only the external but also internal values of taxpayers that may shape their attitudes and behavior in order to understand the way taxpayers react to tax laws (for example Torgler, 2006; Bobek, Roberts

(16)

15 and Sweeney, 2007). These non-economic factors then are explored in recent research into tax morale (Feld and Frey, 2002; Torgler, 2003; Torgler, 2006). Even though the valid determinants of tax morale is remain uncertain, Frey (2003, p. 391) suggests that tax morale is determined by the combination of extrinsic motivation from rewards and punishments derived from external environment and intrinsic motivation from internal values of taxpayers themselves which then will guide their behaviors and attitudes, including tax compliance decisions.

External factors in the tax compliance research that may influence tax morale,

particularly, include taxpayers’ perception of government and treatment from tax authority. The perception of government includes, for example, the opinion of trust, equity, fairness, and types of government. Another extrinsic element is the impact of social interaction, such as peer influence, on taxpayers’ compliance attitudes. Regarding the intrinsic values, Mohdali (2012) suggests that the main factor from internal value of taxpayers that could determine their attitudes is morality. She proposes that morality could be viewed from two perspectives; individuals’ own moral value, and their adherence to other particular institutions, such as beliefs and religiosity.

Reasoned action theory by Ajzen and Fishbein (1980) is one of the examples of social psychological approach that has been applied to the field of tax compliance. The theory assumes that individuals’ behavior is essentially determined by their behavioral intention which is created from the combination of social influences and individuals’ attitude. In other words, the intentions of certain behaviors of taxpayers are based on their personal judgments and others’ opinions. This corresponds to the suggestion from social psychological approach that individual morality and beliefs and peer influences are the main factors that may affect taxpayers’ compliance decisions (Grasmick and Bursik, 1990, p. 857).

(17)

16 There are also similar findings found in other studies (for example Smith and Kinsey, 1987). The idea of conformity is introduced to explain that individuals are likely to follow other people’s ideas and actions. The conformity has purposes such as for gaining social acceptance, avoiding rejection, and also acquiring information that may be needed to make right decision. In the field of tax compliance, perception of society may have strong positive or negative influence on taxpayers’ attitude. Once tax evasion becomes widely accepted in society, for instance, it becomes justifiable for some taxpayers to evade the tax (for example Sandmo, 2005).

2.3 Dependent Variable - Tax Morale

The common definition of tax morale in tax literatures is the taxpayers’ intrinsic motivation to pay taxes. Torgler and Murphy (2004, p. 301) define tax morale as “moral principles or values held by individuals about paying their taxes”. It is important to understand the determinants of tax morale in order to explain it as one of the factors that influences tax compliance. Another possible definition of tax morale is suggested by Graetz and Wilde (1985, p. 358). They define tax morale as the commitment of individuals as a form of responsibility and respect for the law. Aforementioned definitions seem to have a very strong connection with the definition of voluntary tax compliance. Voluntary tax compliance is described as taxpayers’ attitude that comes from their inner and natural willingness to cooperate, originated from their moral obligation towards the benefit of public (Kirchler and Wahl, 2010, p. 343). Based on these definitions, moral values as well as sense of responsibilities to other people could be recognized as main determinants of taxpayers’ attitude in their compliance decision according to tax laws.

Another aspect that may support tax morale, according to prior researches, relates to social norms which focuses on the personal moral beliefs and society influence. According to

(18)

17 Bobek, Roberts, and Sweeney (2007, p. 61), individuals’ morality has more influence on tax compliance than others’ influence. The importance of this moral belief is evident as a number of studies, with slightly different term, have tried to examine it in the field of tax compliance. For example, while Slemrod (1998) uses civic duty to elucidate morality, Torgler et al. (2008) use the term of willingness to obey.

To complete the study of tax morale, Riahi-Belkaoui (2004, pp. 141-142) suggests to add social and religious norms as determinants of tax compliance in order to complete the puzzle in the theory of tax compliance, which deals not only with economic factors but also the

institutional and moral aspects. In other words, morale aspects could act as a complement for economic approach or, perhaps, the missing link of tax compliance study from non-economic perspective. The non-economic aspects that may influence tax morale are discussed in the following sections.

2.4 Independent Variable

2.4.1 Religiosity

To understand how religiosity might influence tax compliance, it is better to start from the theory of modernization. The first conception in the modernization theory is that traditional cultures values are replaced with modern values, while the second is that there is transition process and during the process traditional values remain. Inglehart and Baker (2000, p. 49) propose that even though there are extensive changes brought by economic development, there is a cultural heritage with enduring effects that historically shaped by Protestantism or

(19)

18 persistence of traditional values, this approach also views religion as one of the central elements in society’s culture.

The relationship of religion and economy is believed to interact in two directions. First, the extent to which religiosity responds to economic development, or how economic

development might affect religiosity level. Barro and McCleary (2003a, p. 34) find that as per capita Gross Domestic Product (GDP) increases, religiosity measurements, such as religious attendance and religious beliefs, decline. However, they also find that per capita GDP does not influence religiosity when particular economic development developments such as education, urbanization and life expectancy are held constant. Second, the extent to which religiosity might influence economy. To support this view, Barro and McCleary (2003b, p. 779) suggest that religious beliefs could encourage certain behaviors; for example, honesty and work ethic that supposedly may increase productivity. It then can be said that economic growth responds positively to religious beliefs.

According to Lehrer (2004, p. 721), economic and demographic behavior can also be determined by religion. He emphasizes that association with religious affiliation and the state of being religious individuals relate to well-being and healthy individuals. It is due to the fact that individuals’ religion beliefs have self-enforcement moral rules that ask people to benefit not only to themselves but also larger society. However, the study of the impact of religion on economic development has been neglected since many scholars believe that the effect of religiosity decreases with higher economic development and improved technology, even though a number of researchers have disagreed.

Religion, defined by the Oxford Dictionary (2017), is the belief in and worship or

(20)

19 describes by Heill et al. (1998) in Worthington et al. (2003, p. 84) which views religion as “the feeling, thought, experiences, and behaviors that arise from a search for the sacred”. Allport (1950, p. 26) proposes that description of religion is very complex because of two reasons; there are numerous underlying foundation of religion, and its influence in the lives of individuals is so varied. Forming rational interpretation of religion is also difficult, thus uniformity of product is impossible. However, the interesting part of religion is that it could act as core intrinsic and comprehensive motivation, as value that underlies all things (Allport, 1961, p. 301). It then is argued to go beyond the basic desires such as to gain rewards or to avoid punishment.

Religiosity, on the other hand, is defined as the condition of being religious. Johnson et al. (2001, p. 25) define religiosity or religious commitment as the quality of individuals’ commitment to their religion and its teachings that will be reflected on their attitudes and behavior. This definition proposes that religiosity has effects on individuals’ behavior and it has been studied by several researchers. The result shows that religiosity is expected to influence certain behaviors. Greeley (1963, p. 21) describes the religion and behavior relationship based on three different models; Weberian, need disposition, and predictor models. The Weberian model considers that society is influenced by religion through the values in each individual. The

influence emerges as a result of a religious network not only between individuals themselves, but also between individuals and society as an institution. The second model, need disposition, assumes that religion is a creation of social system. The social system would produce the need disposition that in the end will lead to certain kinds of beliefs and religious activities. The distinction of the first and second model lies on the position of religion whether as independent or dependent variable. The third model, on the other hand, views religion not as dependent or independent variable, but rather as a predictor of behavior. This model assumes that the

(21)

20 relationship between culture, personality, and society are so complex. It is because that religion could affect the organization which in turn may influence individuals’ role in the system. On the other hand, the role could then shape the value they support and in turn have an effect on their religious beliefs. In other words, it says that individuals rely on religious values as a source of their morality to shape their attitudes and behavior.

Religiosity, in the field of tax compliance, has not been widely recognized as one

potential element that might explain the puzzle of tax compliance, until several researchers such as Torgler (2003b) and Welch et al. (2005) emphasized its importance. In tax compliance study, religiosity can be related to the suggestion made by Margolis (1997, p. 247) that religious faith, in any religion, can act as moral constraint to make a distinction between good and bad behavior. In addition, religious beliefs are likely to prevent deviant behavior by imposing self-guilt,

particularly in tax evasion case (Grasmick et al., 1991, p. 263).

Tittle and Welch (1983, p. 676) in their attempt to study the relationship between religiosity and deviant acts – one of the deviants acts is tax evasion – find that the effect of individuals’ religiosity on future deviant behaviors is determined by the characteristics of their religious communities. In other study, Welch, Tittle and Petee (1991, p. 169) use data collected from Catholic parishioners and find that individuals’ religiosity is negatively correlated with tax evasion. Several studies also support the view that religiosity may positively affect tax morale. Torgler (2003b, p. 297), using data from World Values Survey (WVS) for 1990 in Canada, reveals that tax morale is positively associated with religiosity. Similar opinion also suggested by Stack and Kposowa (2006, p. 349) who conclude that people with no religious affiliation were more likely to view tax fraud as acceptable.

(22)

21 To sum up, there is an indication of positive influence of religiosity on tax morale. Individuals are likely to use religion as a way out of their problems (Keister 2003, p. 178) and it indicates a new aspect in dealing with tax morale issues. This domain of research could offer an opportunity for investigation because religiosity seems to be inadequately examined in earlier tax morale studies. Thus:

H1: More religious people tend to have higher tax morale.

2.4.2 Perception of Government

Psychological tax contract as the complicated interaction between taxpayers and governments has been used as an element that determines tax morale (Feld and Frey, 2007, p. 103). This tax contract is apparent in various number of tax compliance studies as trust in government and legal system may significantly influence tax compliance (for example: Torgler, 2003b, p. 297; 2004, p. 36; Torgler and Murphy, 2004, p. 334; Richardson, 2008, p. 75). Another similar opinion is suggested by several studies that the perception of taxpayers whether tax system is fair or unfair, as well as taxpayers’ positive attitudes towards government based on trust will encourage tax compliance behavior which in turn persuade taxpayers to pay their tax (Murphy, 2003, p. 234; Richardson, 2006, p. 164; McKerchar et al., 2013, p. 18).

Beside the trust in the government as well as tax system, the provision of public services also holds important key in tax morale. According to Alm et al.(1992, p.36) and Torgler (2003a, p. 137), taxpayers are more likely to comply with tax regulation when they consider public service from government as appropriate and equitable. Feld and Frey (2007), in addition, in their research find that even though taxpayers do not receive public goods comparable to the paid taxes, they are willing to pay taxes if they perceive the political process as fair and lawful.

(23)

22 Another determinants regarding perception of government that might influence tax morale is the involvement of taxpayers in political process. Their involvement is shown to have positive effect on the compliance, and taxpayers with direct democratic rights are also found to be more compliant (Torgler, 2005, p. 530; Torgler and Schneider, 2007, p. 465). Overall, the perception of government, either from the performance, tax system, or political system, can be one of non-economic factors that are crucial to maintain or increase taxpayers’ voluntary compliance. Thus:

H2: Taxpayers with better perception of government tend to have higher tax morale.

2.4.3 Interaction with the Tax Authority

Feld and Frey (2007, p. 115) propose that personal relationship between taxpayers and tax authority is a part of psychological tax contract and is considered to be a function of tax morale. The interaction between them might result in either positive or negative experiences. These experiences are the one that will influence individuals’ decision making process, either to comply or not to comply. This cooperation between taxpayers and tax authority to willingly build a successful community partnership to enhance voluntary tax compliance is identified as synergistic tax climate by Kirchler et al. (2008, p. 211).

Feld and Frey (2002b, p. 97) in their study find that positive treatment from tax authority to taxpayers has a positive effect on tax morale. Other studies confirm this finding, for example friendly treatment by tax authority during tax audit (Feld and Frey 2007, p. 115), as well as gentle reminder letter to taxpayers that deliver the impression fairness principle (Wenzel 2006, p. 358) is found to have a significant positive effect on taxpayers’ compliance attitude. Another

(24)

23 finding is found by Wallschutzky (1984, p. 383) who suggests, based on taxpayers survey in Australia, that the way taxpayers are treated can influence the future compliance of taxpayers. However, it is important to note that the impact of treatment on taxpayers’ compliance might be varies in every country based on the culture of the society and tax authority.

Another important point is that psychological contract between taxpayers and tax authority is based on a relationship of trust and disrespectful treatment may deteriorate the psychological contract (Feld and Frey 2002, p. 97), which in turn can adversely affect compliance attitude (Murphy 2004, p. 203). Furthermore, from the cooperation between

taxpayers and tax authority, there will be a significant amount of information that is beneficial to provide taxpayers needs and supposedly will motivate voluntary tax compliance attitudes. Thus:

H3: Taxpayers who experience better interaction with tax authority tend to have higher

tax morale.

2.4.4 Social Interaction

Values and beliefs in society may be incorporated in each individual within the society. By observing and internalizing the values from the society, the members are likely to adopt and adapt these values as their code of conduct and behavior in life. How societal perception of tax compliance may influence taxpayers as member of society can be explained using pro-social behavior theory. The assumption behind this theory is that people take into account of others’ behavior when act. Reciprocity theory, as one of pro-social behavior theories, believes in positive attitudes as a respond of other positive attitudes, and negative attitudes as a respond to other negative attitudes (for example see Fehr and Falk, 2002). In tax compliance context,

(25)

24 according to reciprocity theory, there will be compliance behavior of one or more taxpayers may positively influence others behavior. In other words, there will be feeling of uncomfortable for not paying taxes if taxpayers know that others are paying their taxes or vice versa (Mohdali and Pope 2012, p. 126).

There are a growing number of studies which try to find the effect of social interactions and peer influences on tax morale. Scholz and Lubell (1998, p. 411) discover that the willingness of taxpayers to pay taxes increases as they believe if other people are bound by same social contract, or as Wallschutzky (1984, p. 383) suggests, if tax evasion is apparent in society, taxpayers are likely to evade. Other studies also find positive correlation between social interaction and tax morale. A study by Molero and Pujol (2012, p. 158-159) using survey to students as respondents indicates that tax morale can be improved if they identify that only small group of people involved in tax non-compliance. Torgler (2004a, p. 36) also reveals that

taxpayers are influenced by information regarding how other people manage their tax affairs. For instance, if taxpayers are aware that tax evasion is evident, they are more likely to evade taxes. Furthermore, Wenzel (2007, p. 42) also finds that tax compliance is also influenced by the identification of groups that relate to a compliance pattern. From this, it can be inferred that the perception of taxpayers of other taxpayers’ actions might influence their future compliance decision. In line with the intrinsic and extrinsic motivation to pay taxes voluntarily, Bobek et al. (2007, p. 61) propose that combination of taxpayers’ inner values and values of others in the society is the most influential factor in influencing taxpayers’ compliance attitudes. To sum up, integrating right values in society might generate positive results. Thus:

H4: Taxpayers who are more influenced positively by social interaction tend to have

(26)

25

2.4.5 Threat of Punishment

The threat of punishment as economic factors in the determinant of tax morale consists of tax audits and tax penalties. Witte and Woodbury (1985, p. 9) propose that audit probabilities has different impact on different group of taxpayers. They argue that audit probabilities may have significant influences on self-employed taxpayers but only small effect on salaried taxpayers. Hence, it is beneficial for tax authority to have a policy that targets specific groups of taxpayers rather than randomly audit taxpayers. In line with this idea, Trivedi et al. (2003, p. 190) also suggests that audit probabilities is found to influence only comparatively risk averse taxpayers or taxpayers who have intention to evade.

With regards to the tax penalties, the influence on tax morale is also varied. Some researchers indeed find that penalty is positively associated with tax compliance, for example Park and Hyun (2003) and Beck, Jon, and Jung (1991). However, there are also different

opinions which show that greater penalty is ineffective to enhance compliance, for example Alm, Jackson, and McKee (1992). Furthermore, increasing tax penalty is argued to have an inherent risk (Skinner and Slemrod, 1985, p. 348-349). The first risk is that tax penalty may have an issue of inequality between tax evaders who are caught and those who are escape. The second risk is that government may penalize evaders who evade intentionally and other who do that

unintentionally uniformly. Overall, even though there are contrasting opinions regarding threat of punishment, this variable is still an important part to understand tax morale. It may be more beneficial to include an economic factor in the formulation of tax morale to shed more light in the puzzle of tax morale.

(27)

26 However, in the bigger scale, Gneezy and Rustichini (2000, p. 2) propose that in the field of legal studies, punishment is justified to reduce the number of future crimes; yet, when the discussion shifts to economic analysis, the effect of punishment is not direct and it requires a set of market formula. Further, they find that the introducing of penalty increases the number of behavior being punished due to the change of people’s perception when they anticipate the penalties of their behavior (p. 3). The penalty is then considered as a cost that justifies the violation; as long as ones afford to pay the costs, the violation is deemed to be allowed (p. 14). While Gneezy and Rustichini study (2000) conclude that punishment can be considered as a price for violating the regulation, a study by Kreps (1997) finds that extrinsic motivation could lead to diminution of intrinsic motivation. One of possible explanation is that when individuals do something without the encouragement of extrinsic incentives, they will consider their efforts as reflecting their enjoyment. When the extrinsic incentive is introduced, the efforts may be attributed towards the incentive, developing distaste for the required efforts (Kreps, 1997, p. 362). The punishment, in this case tax penalties and tax audits, is considered as an external motivation for taxpayers to pay taxes. On the other hand, tax morale is considered as intrinsic motivation of taxpayers, thus:

H5: Taxpayers who are more deterred by punishment tend to have lower tax morale.

2.4.6 Chapter Summary

Despite the numerous studies on individual taxpayers that have been undertaken, the tax morale puzzle still remains. Non-economic framework probably complements neoclassical model of economic deterrence approach as an attempt to learn more the complex attitude of tax

(28)

27 morale. The blend of both factors in tax morale research is then expected to provide richer insight in understanding the attitude of taxpayers.

The influence of non-economic factors on tax compliance has been studied broadly since 1990s, such as taxpayers’ perception of government, tax authority, as well as society. However, the effect of religiosity as internal value of taxpayers had been overlooked. Only more recently then religiosity is discussed as one potential determinant of compliance attitude. Therefore, it is obvious that further theoretical and empirical contributions are necessary to expand the currently limited religiosity studies in the field of tax compliance.

(29)

28

CHAPTER 3 RESEARCH DESIGN AND METHODOLOGY

3.1 Introduction

This chapter focuses on the research design and the methodology used in this thesis. It begins with the explanation and the reason behind the usage of survey as a method to gather the data. It then describes the explanation of the sampling method as well as the measurement and operationalization of variable used. Finally the brief description of analysis is discussed.

3.2. Survey

Survey research, today, is a frequently used method of research in the social science. Survey is used mainly in a research with individual people as the units of analysis (Babbie 2013, p. 229). He also suggests that survey is the most appropriate technique for researchers who interested to study extremely large population, as well as to measure attitudes and orientation in the population. This study tries to examine the effect of religiosity to Indonesian taxpayers’ morale, therefore survey can be considered to be the most effective method. To collect the attitude and responds from respondents, this study uses self-administered survey method. Self-administered questionnaire is an instrument of survey that is used to gather information form respondents who complete the instrument themselves (Bourque and Fielder, 1995, p. 2). Self-administered questionnaire must stand alone, which means that all the information needed to complete the questionnaire must be provided on the form itself, because there is no interviewer who will provide instruction or clarify information (Bourque and Fielder, 1995, p. 17).

(30)

29 Furthermore, variables used in this study are sensitive topics. The main explanatory variable is religiosity level, while the dependent variable is tax morale. Other variables are also employed as control variables, namely social interaction, perception towards government, and threat of punishment. Due to the sensitive topics asked, self administered survey is argued to be the method that can yield most responses, as explained by Lee (1993, p. 75) and Bourque and Fielder (1995, p. 13). It is because respondents are likely to provide answers anonymously when they are asked about their attitude regarding sensitive issues. People are also more likely to present complete and truthful information regarding sensitive topics if self-administered questionnaire is used rather than in an interview. Lee (1993, p. 75) states that self administered survey is preferred because respondents usually try to manage good impression of themselves in the eyes of interviewer. They tend to under-report the behaviors that are normally regarded as threatening or sensitive in the presence of interviewers. It comes from the argumentation that when an interviewer is absence, respondents are less likely to feel threatened by sensitive topic questions (Lee, 1993, p. 98).

The other consideration is because the limitation of researcher. Firstly, this research is conducted as a requirement of master study. The research is conducted in the Netherlands while the data collection is completed in Indonesia. Secondly, there is also budget limitation. Bourque and Fielder (1995, p. 9) suggest that the lower cost of application is one of the advantage of survey because it only involves minimal cost, especially when it is compared to other methods such as in-person and telephone interview. Hence, self administered survey is preferred.

More specifically, the self-administered survey used is online survey. There are several reasons why online survey is chosen. Van Selm and Jankowski (2006) briefly describe the

(31)

30 others. Online survey is supposed to be cheaper as related to conventional paper-and-pencil surveys. Manfreda and Vehovar (2008, p. 282) explain that compared to mail survey, online survey has lower or even no costs for paper, printing, envelopes, postage, and related

administrative work. It also has an advantage when compared to paper-and-pencil questionnaires because it spares the costs of data entry and editing procedures. Compared to

interviewer-administered-survey, the costs reduction happens from the absence of interviewers. Second, online survey also provides the possibility of anonymity for respondents and it can help

researchers to facilitate the sharing of respondents’ experience and opinions regarding sensitive issues thus make it possible for researchers to reach the respondents who usually are difficult to be reached. Finally, online survey provides researchers with easier data collection because the usage of internet, and helps respondents with a simple and efficient way to answer and submit their responds. In addition, Manfreda and Vehovar (2008, p. 272) also mention that internet survey could decrease the time intervals between contacts of interviewer and interviewees.

While some researchers agree that online surveys may be beneficial and increase response rate, others also find that this type of survey has its own drawback. For example, Babbie (2013, p. 259) illustrates the representativeness problem of online survey. It is because not everyone of interest in the study can be reached via internet. However, he also explains that the problem will decline as more and more people have gained access to internet. The other consideration, for this study especially, is also because this mode of surveys is considered to be the most inexpensive for researcher and the one that can produce the most response in short period of time.

(32)

31

3.3 Questionnaire Design

The questionnaire used in this study is divided into three sections. The first section is designed to obtain personal details of respondents, such as age, gender, religion, education level, employment status, and income. Tax statements are positioned in the next section which includes respondents’ perception of tax morale, social interaction, threat of punishment, and perception towards government. The last section gathers the level of religiosity based on the religiosity score. The data are measured using several measurement scales (see Mike Procter 2001, p. 254-255). Nominal scale is used for data that do not need any order or structure, such as gender, religion, and employment status. This type of scale is used for identification purposes. Other items are measured using ordinal scale, including education level and total income. The main scale employed in this study is interval scale that is used to measure respondents’ attitudes towards certain issues. An interval scale of five-point Likert scale is used to measure respondents’ attitude regarding tax and religiosity dimensions.

Online-self-administered survey is used to collect the data. Several endeavors are adopted to make it easy for the respondent to complete the questionnaire and to increase the response. Plain design in the user interface is adopted in the survey to provide easy visual due to the reason that internet users tend to read and scan quickly rather than read the text thoroughly (Manfreda and Vehovar, 2008, p. 272, 276). The survey introductory page is also used to persuade

respondents to take time to complete the questionnaire, to emphasize the ease of participating, to instruct respondents, and the last but not least to explain the incentive given. The virtual

incentive is used to incentivize respondents, because generally incentive will increase response rates, web survey included.

(33)

32 The next objective of questionnaire design is to reduce or to avoid measurement error (Simmons, 2008). Firstly, poor questionnaire layout might cause questions to be unnoticed. On the other hand, respondent-friendly questionnaire is more attractive and may encourage

respondents to complete the questionnaires. Due to this reason, this study uses grid (table or matrix) question type. This mode is proven to have several advantages. Manfreda and Vehovar (2008, p. 277) describe that grid question makes questionnaires look short, eliminate redundancy, and also reduce the effort for respondents with keyboard and mouse action in a computer.

Furthermore, cognitive effort needed is reduced because same response categories are used for all items. Those all lead to the decrease of respondents’ burden; hence, increasing the response rate and data quality. Babbie (2013, p. 239) adds that grid questions allow efficient use of spaces, help respondents to complete the questionnaire faster, help respondents to review their previous answer thus increase the comparability of responses given to different questions. Despite the aforementioned advantage, there are several disadvantage of grid question. Babbie (2013, p. 239-240) and Manfreda and Vehovar (2008, p. 277) list several disadvantages; it may foster a

response-set among several respondents, which means that they tend to answer question uniformly to decrease the efforts, or to develop a pattern of, for example, agreeing or not agreeing with all the statements in the questionnaire. To counter this problem, there are several trap questions employed in the questionnaire. The questions ask respondent to choose a specific answer, for example, disagree. The aim is to differentiate respondents who answer in a pattern or uniformly so they can be omitted in further data analysis.

(34)

33

3.4 Respondents

This study involves individual taxpayers in Indonesia as respondents. As it tries to examine the tax morale – the willingness to comply with tax voluntarily – the samples are not limited to only working people, but also people who are not working. However, the respondents are limited to the level of individual because obviously the level of religiosity can be measured only from a person perspective, not from a company or other type of institutions.

Due to the abovementioned limitations, data are collected through web survey using volunteer opt-in panel. Opt-in panel is used mainly in market research. In this mode, lists of respondents are acquired through participants’ self inclusion to a research or survey institution. The lists then are maintained by a professional market research institution, and are usually a large database of volunteer respondents (Manfreda and Vehovar, 2008, p. 266).

The data collection in this study is conducted by PT Nusaresearch in Indonesia based on the principal agreement no. NS/AGR/012017. PT Nusaresearch is a branch of World Wide System Corporation (W&S Co) in Japan. W&S Co. is a subsidiary of W&S Group with head office located in Japan and branches in several Asian countries, such as Vietnam, Indonesia, and Thailand. The services provided by W&S Co. include market research, website design, and internet marketing.

Respondents are collected using simple random sampling from the database maintained by Nusaresearch. To sustain the quality of respondent database, panel quality management employs two screening methods; Algorithm logic (Al) check and human check. Al check consists of automatic system check for multiple accounts identified by respondent identities (such as email, phone number, citizen identity card, etc), check for impersonator accounts, check for panelists who have been inactive for an extended period, identify invalid respondents via trap

(35)

34 questionnaires, and regular checking for invalid responses via regular survey. The human check, moreover, consists of check by panel executive officer everyday to check new register and approve correct personal data, email respondents who have not completed information form or respondents with unsuitable data, confirm validity of suspicious information personal data by calling the respondents, and call respondents who requested during exchange point reward period.

Sample size in this study is limited to 300 respondents due to the limitations. However, this number is deemed to be sufficient according to Fowler (2009, p. 44). According to Fowler, a sample of 150 people will describe a population of 15,000 or 15 million without any virtually difference in the degree of accuracy; increasing the number of respondents will only give small impact to the objective of representing the population (2009, p. 44-45). Furthermore, Babbie (2008, p. 142-143) also argues that the accuracy of estimates is not determined by the size of sample and population; in other words the size of population is almost always irrelevant. It is due to the fact that the calculation of sampling error always assumes that population are infinitely large. Thus, the sample size would equal 0 percent of the whole; sufficiently small to approach the situation in infinitely large population. The proportion may be considered as relevant if the sample size, say, represents five percent or more of the population. On this basis, the size of sample at 300 respondents is considered to be adequate.

3.5 Measurement

Several constructs are developed to represent the variables in this study. Dependent variable is represented by tax morale. The independent variables are religiosity as internal value of taxpayers, perception of government, interaction with tax authority, and social interaction as

(36)

35 external value of taxpayers, and threat of punishment as economic factor that may affect tax morale.

Respondents’ attitude towards tax is measured in the second section of questionnaire, including tax morale and their perception of external dimension, namely government, tax authority, society, and threat of punishment. The constructs used are adapted from existing literature. The tax statements are measured on five-point Likert scale (1 = strongly disagree, 5 = strongly agree). The items are shown in the table 3.1 with the list of the referred literature.

The next section of questionnaire measures the religiosity level of respondents.

Religiosity questions in this study are adapted from Worthington et al. (2003). They create the 10-item measurement of religious commitment (RCI-10) which can be used research,

counseling, and health psychology. RCI-10 is argued as the most efficient and psychometrically sound instrument, and is also expected to be able to distinguish between people with high

religious commitment and those that are not as highly committed. The statements are listed in the table 3.2. Religiosity construct in this study is also measured by five-point Likert scale (1 = not at all true of me, 5 = totally true of me). The statements used are more general so they can cover all religions and avoid bias against certain religions. Worthington et al. (2003) conclude that a full-scale RCI-10 score of 38 or higher would justify considering a person to be highly religious because it is considered to be one standard deviation higher than the mean.

(37)

36

Table 3.1 Tax Items and the reference

No. Tax Morale Reference

1. I pay taxes as required by the regulations because it is my

responsibility and something that I have to do

Kirchler and Wahl 2010

2. I pay taxes as required by regulations to support the country and

other citizens

3. I pay taxes as required by regulations because I want to

contribute to everyone's good

4. I pay taxes as required by regulations because it is the normal

and right thing to do

5. I pay taxes as required by regulations because it is my duty as a

citizen

No. Perception towards Government Reference

6. The government spends taxpayers' money in transparent manner

Mohdali and Pope (2012)

7. Government is spending public money prudently

8. The services provided by government are comparable to the

amount of taxes paid

9. Public funds are allocated equitably

10. The taxes paid is fair because the tax base has included allowable

deductions Niemirowski and Wearing

(2006)

11. Tax laws has ensured that everyone pays fair amount of taxes

No. Social Interaction Statements Reference

12. My surroundings influence me in my decision to pay taxes

Mohdali and Pope (2012)

13. My decisions to pay taxes are mostly affected by the experiences

and suggestions of my friends

No. Interaction with Tax Authority Statements Reference

14. DGT staffs are helpful with my enquiries about tax return

Niemirowski and Wearing (2006)

15. DGT is professional in the way it treats me as a taxpayer

16. DGT staffs treat me fairly

17. DGT is fair and considerate with those who get audited

No. Threat of Punishment Statements Reference

18. The probability of being detected and punished when evading tax is high so it is likely for taxpayers to be discovered and punished.

Verboon and van Dijke (2007)

19. I pay taxes as required by the regulations because I know the probability of being audited is very high.

20. The DGT almost always finds out if someone has overly understated their income or overstated their deductions.

(38)

37

Table 3.2 Items for religiosity

No. Item

1. I often read books and magazines about my faith.

2. I make financial contributions to my religious organization. 3. I spend time trying to grow in understanding of my faith.

4. Religion is especially important to me because it answers many questions about the meaning of life.

5. My religious beliefs lie behind my whole approach to life. 6. I enjoy spending time with others of my religious affiliation. 7. Religious beliefs influence all my dealings in life.

8. It is important to me to spend periods of time in private religious thought and reflection.

9. I enjoy working in the activities of my religious organization.

10. I keep well informed about my local religious group and have some influence in its decisions.

3.6 Data Analysis

The data collected from the questionnaire was analyzed using Statistical Packages for Social Science (SPSS). The data gathered through online survey can be directly entered into an electronic file compatible with SPSS so the manual data coding is not needed. The data then is analyzed using main descriptive statistics, correlation and regression analysis.

The descriptive analysis describes the background of respondents using frequency, mean, and standard deviation. It is intended to portray the demographic and socio-economic status of respondents. Descriptive statistics is also used to present the snap shot of respondents’ attitude towards tax issues and their religiosity level.

(39)

38 Relationships between variables are tested using Pearson correlation coefficient (r) analysis and p value. All the independent variables, namely religiosity, perception of government, interaction with tax authority, social interaction, and threat of punishment, are tested against the dependent variable, namely tax morale. The next step is performed to predict the determinants of tax morale variable. From this analysis, b value is obtained to measure the strength of each independent variable in determining tax morale. Adjusted R2 is used to calculate the success of model provided.

(40)

39

CHAPTER 4 DATA ANALYSIS AND FINDINGS

4.1 Survey

One of online survey disadvantages is the difficulty to obtain response rate. As explained by de Vaus (2002, p. 128), calculating the respondents who refuse to respond in an online survey is not possible because researchers do not have any control over the respondents. It is particularly true when third parties, for example the usage of company representatives or opt-in panel in research company, were employed in the process of distributing questionnaires in the study. The screening process is conducted through the usage of trap questions in the survey. Respondents are asked to directly choose one option, for example “Please choose disagree”. It is intended to eliminate respondents who answer all the statements uniformly. Respondents who have wrong answer in the trap questions then are removed from the lists and will not be used in the data analysis. For this study, total respondents used are 300 people. Therefore, the survey was distributed for a period of time until the quotas of 300 usable responses are obtained.

Reliability tests are performed for each of the variables. The Cronbach alpha coefficient is presented in table 4.1. The measurement employed in this study is considered to be reliable and internally consistent for all of the variables because the value of Cronbach alpha exceeds 0.6 for all components.

(41)

40

Table 4.1 Reliability Coefficient of Components

Components Number of items Cronbach alpha

Tax Morale 5 0.899

Perception of Government 6 0.927

Interaction with Tax Authority 4 0.912

Social Interaction 2 0.828

Threat of Punishment 4 0.664

Religiosity 10 0.906

The responses received in this study are expected to reflect the proportion of Indonesian citizen as taxpayers as a whole. The only data available that can be used is obtained through Indonesian Central Agency of Statistics from the national population census in 20102. There is no more recent data available. The comparison between response received and the population of Indonesia by gender, religion, and age is shown in the table 4.2, while the comparison by region is presented in table 4.3. The comparison in the tables shows that the surveys can be considered to reflect Indonesian population to an acceptable level for gender and religion, even though the sample may not be a perfect representation of the population. Regarding the age, the proportion of groups of age in the sample may not be a reflection of population especially in the group of people with the age between 30 and 39, and group of people with age above 50. Firstly, it can be caused by the difference of period when the census was performed. Secondly, it maybe because the sample only consists of people who use and are exposed to internet while the actual number of people who use internet in the population might be lower. The ranking of each group in the sample is similar to the one in the population. The proportion of male and female is roughly 50% for both groups. The three religions, namely Islam, Protestant, and Catholic still dominate

percentage of population and sample with Islam compromise approximately 80% of the total. The sample and population also reflected in the composition of respondents’ regions which are

(42)

41 dominated from provinces in Java Island, which are DKI Jakarta, Jawa Barat, Jawa Tengah, and Jawa Timur. The slight difference proportion between sample and population may be caused by the difference of survey and census conducted.

Table 4.2 Comparison of Respondents with Indonesian Population by Gender and Religion3

Reference Percentage of Response Percentage of Population

Gender Male 53.00 50.34 Female 47.00 49.66 Religion Islam 79.00 87.18 Protestant 11.00 6.96 Catholic 4.67 2.91 Hinduism 1.67 1.69 Buddhism 3.00 0.72 Confucius 0.33 0.05 Atheism 0.33 - Age 20-29 35.67 31.67* 30-39 44.67 29.46* 40-49 16.00 23.49* 50-59 3.67 15.38*

*Because the respondents’ age only ranges from 20 to 59, the percentage of population is also from the number of people from the age of 20 to 59

Regarding educational background, more than half of the respondents belong to the highly educated with 60.67 percent at bachelor (undergraduate) and postgraduate levels. The rest of respondents are at high school level (20.00 percent), diploma level (19.00 percent), and below high school level (0.33 percent). Approximately half of the respondents are working as private company staff (54.33 percent). Self-employed respondents and professional – doctors, teacher, etc. – are 10.00 percent and 10.33 percent respectively. Civil servants seize 9.33 percent of the total and others employment status make up for the rest with 16.00 percent. The summary is presented in table 4.4.

(43)

42

Table 4.3 Comparison of Respondents with Indonesian Population by Province4

No. Province Percentage of Response Percentage of Population

1. _Aceh 3.33 _1.89 2. Sumatera Utara 6.67 5.46 3. _{Sumatera Barat} - _2.04 4. _Riau 1.00 _2.33 5. _Jambi 0.67 _1.30 6. _{Sumatera Selatan} 3.00 _3.14 7. _Bengkulu 0.33 _0.72 8. _Lampung 1.33 _3.20

9. _{Kep. Bangka Belitung} 0.33 _0.51

10. _{Kepulauan Riau} 0.67 _0.71 11. DKI Jakarta 11.00 4.04 12. _{Jawa Barat} 15.33 _18.12 13. _{Jawa Tengah} 12.67 _13.63 14. DI Yogyakarta 4.67 1.45 15. _{Jawa Timur} 15.00 _15.77 16. Banten 5.00 4.47 17. Bali 4.33 1.64

18. Nusa Tenggara Barat 2.00 1.89

19. Nusa Tenggara Timur 0.67 1.97

20. Kalimantan Barat 3.33 1.85 21. Kalimantan Tengah 0.67 0.93 22. Kalimantan Selatan 1.67 1.53 23. Kalimantan Timur 1.00 1.50 24. Sulawesi Utara 0.33 0.96 25. Sulawesi Tengah 0.33 1.11 26. Sulawesi Selatan 3.00 3.38 27. Sulawesi Tenggara 1.33 0.94 28. Gorontalo - 0.44 29. Sulawesi Barat - 0.49 30. Maluku - 0.65 31. Maluku Utara - 0.44 32. Papua Barat 0.33 0.32 33. Papua - 1.19 4.2 Main Findings

The findings of this study regarding taxpayers’ perception towards tax and other issues as well as the level of their religiosity are discussed in this section. The relationship between these components and their impact on tax morale will also be described.

The Influence of Religiosity on Tax Morale in Indonesia