EXAMINING THE RELATIONSHIP BETWEEN ECONOMIC, DEMOGRAPHIC, LEGAL AND INSTITUTIONAL VARIABLES TOWARDS TAX EVASION : A CASE STUDY IN WEST KALIMANTAN

Tax evasion, one of the most common crimes in the world. There are so many cases of tax evasion that are not known by the public, in fact there are many cases of tax evasion that are closed and only revealed a few years later. This is also happened in Indonesia. Without exception, Indonesia is also one of the countries with the most tax evasion practices. To examine which factor, affect the tax evasion practice the most, this problem needs to be studied further. This research aims to examine and reveals how the economic variable, demographic variable, and legal and institutional variable affect tax evasion practice with West Kalimantan, Indonesia as the studied area. This research uses the multiple regression analysis as its research method. Questionnaire was used for the data collection where the questionnaire succeeded to obtain 120 respondents as research population, then 100 sample gained after screening process . The result of this research is that the economic variable and the legal and institutional variable has a positive effect on tax evasion practice and the demographic variable has a negative effect on tax evasion practice.


INTRODUCTION
Taxes are payments or fees imposed by local, state, and national governments on individuals or businesses to cover the cost of services, goods, and general government services.In Indonesia, people who are considered as taxpayer(wajib pajak) are people who are over 18 years old or the one who are already married.Tax can be collected based on the type of collection agency (central tax and local tax) and based on its nature (direct tax and indirect tax).All of these types of taxes will later be used by the government and used for the welfare of the people through infrastructure development, subsidies, and other activities.According to Direktorat Jendral Pajak (1983), the law that regulates about the taxation is the Law No.6 of 1983 where it has been changed to the newest one which is Law No.16 of 2009 that concerning about the General Provisions and Tax Procedures.
Although there are many tax administration regulations, many tax exemptions can still be found; such as tax avoidance, and tax evasion.Tax avoidance is a measure designed to reduce the amount of tax paid while maximizing the after-tax income.It is also a legal strategy to reduce your tax liability.Unlike tax evasion, tax evasion is classified as an illegal activity.Tax evasion is the practice of evading taxes or using illegal way to refuse paying taxes.Rahayu (2017) defines tax evasion as an active attempt made by taxpayers to illegally decrease or remove the tax burden since it breaches tax laws and regulations.The surge in tax evasion cases and the assumption of negative taxes has led in the general impression that tax evasion is an ethical activity to take.
A report by Tax Justice Network (2020), entitled "The State of Tax Justice" stated that countries in the world have lost more than 427 billion dollars annual taxes due to the abuse of global corporate tax and individual direct tax evasion.This is equivalent to almost $34 million in lost nurse's annual wages each year, or one nurse's annual wages per second.Each year, more taxes are lost in tax havens by multinational corporations than by private tax evaders where multinational corporations embezzle countries with $245 billion in taxes each year, while those moving their assets abroad embezzle their governments from $182 billion in less taxes each year.
In Indonesia, according to a Tax Justice Network report by Rustandi & Luib (2020), tax evasion in Indonesia costs the government $4.86 billion or IDR 68.7 trillion annually.The report titled "The State of Tax Equity 2020: Tax Equity in the Age of Covid-19" details $4.78 billion or IDR 67.7 trillion as a result of corporate evasion.taxes in Indonesia.The remaining $78.83 million or IDR 1.1 trillion was lost due to tax evasion by individuals or private taxpayers.
Companies often hide their profits in countries considered tax havens to avoid disclosing the actual profits earned in the country where the company operates.This strategy allows companies to pay much less tax than otherwise.On the basis of individual taxpayers, wealthy people often hide their assets and declare their assets abroad, which is against the law.This trend is driven by corporate tax abuse, with low-income countries losing the equivalent of 5.5% of tax revenue collected and high-income countries by 1.3%.To get a clearer picture, the Indonesian Ministry of Finance is targeting tax revenue of IDR 1,198.82trillion in 2020, equivalent to 5.7% of the total reported tax evasion.Compared to 2019, when Indonesia had just reached the target of tax loss of IDR 1.332 billion, equivalent to 5.16% of the total reported tax loss.
Research about tax evasion has been conducted by Marriott (2017) where she found that age, income level and income source have impact towards the attitude of tax evasion however Taing & Chang (2021) found that demographic variable has no relations on tax evasion and 'Aini Haron et al., (2018) has not found relationship between income level and tax evasion.Rashid (2020) studies found tax complexity, have positive effect towards tax evasion attitude but Komang et al. (2017) result stated that tax system/complexity has negative effect on tax evasion.On the other hand, the study by Cullen et al. (2018) showed that a positive outlook on government lowers tax evasion and the result also differs from Taing & Chang (2021), where it's believed that attidue toward the government in term of trust especially has no effect on tax evasion.
Differentiation of previous researchs result had led to further examination which needs to be conducted to examine the correlation between economic, demographic and legal & institutional variables towards the tax evasion practices.Using the 7 indicators namely; income level, income source, age, gender, education, trust in the governance, and tax complexity.Data collection process will be carried out using questionnaire and conducted using Likert Scale method.And the research will be conducted in West Kalimantan due to the studies on tax evasion in West Kalimantan is very limited.it was also supported by the information from Badan Pusat Statistik (2023) the GDP of West Kalimantan grew by 4% in the second quarter of 2023 compared to the second quarter of 2022, where the increase of GDP will also increase the amount of tax that need to be payed for each individual.The research aims were to analyze and discover the effect of economic, demographic and legal & institutional variable toward tax evasion practice with West Kalimantan, Indonesia as studied area.Wallschutzky (1984) was the one that presented the first review of the literature on the factors related to tax evasion and emphasized by Jackson and Milliron (1986).According to Wallschutzky (1984), tax evasion means payment of less tax than the law requires on one's true taxable income.He explains that the possible cause of tax evasion can be categorized by 3 categories, namely; exchange indicator, social orientation and administrative control.

LITERATURE REVIEW Tax Evasion
Tax evasion practice always related with the behavior of individual.That's where the theory of planned behavior takes part.Ajzen (1991) explains that the theory of planned behavior is used to predict whether a person will do or not perform a behavior and he believes that intentions of an individual can be explained by 3 determinants, namely; attitudes towards the behavior, subjective norms, and perceived behavior control.And the theory has been used by lots of researcher, and one of the researchers Alleyne and Harris (2017) which conducted research with the aim to predict the intentions of an individuals to conduct a tax evasion practice and it's found that tax evasion practice has been significantly impacted by the attitude towards perceived behavioral control, moral obligations, and attitudes toward the behavior.

Economic Variable (Income Level, Income Source)
An Economic variable is a variable collected by using economic determination as its source.Several indicators that can be used to describe economic aspects, is nations, a government, a corporate or an individual.Economic variables have several types; (1) quantity (Gross Domestic Product (GDP) and unemployment), ( 2) quality (productivity, utility) and ( 3) distribution (Gini coefficient).This research uses the quantity economic types, where income level will be considered as part of Gross Domestic Product (GDP), and the income sources as the part of unemployment factors.For economic variable, this research used 2 indicators namely, income level and income source.
Income in economics can be defined as the monetary outcome or result obtained from the use of goods or services by people.To classify income, there is something called income level or level of income.Income level is the amount of money that people can earn over a given period of time; such as week, month, and year.Income levels can be classified into 4 levels; (1) low income, (2) lower-middle income, (3) upper-middle income, and (4) high income.This research uses the income level as the part of the Gross Domestic Product (GDP), because income level explains how to categorize an income earned by each individual.With the existence of an income level, it is possible to simplify the calculation of Gross Domestic Product (GDP).
Beside income level, this research also uses the income source as the part of the unemployment factors.Income source can be defined as an entity that provides income or something that provides a regular supply of money, such as employment, investments, pensions and others.Income sources can be classified into five types; (1) wages, (2) personal income (business & etc.), (3) government transfer, (4) investment income and (5) other income.This research uses the income source as the part of unemployment factors because income source collected the data about where the income is obtained.Thus, it can be seen whether a person actually gets income from working or not.Gërxhani & Schram (2006) has conducted research about tax evasion where the results of the studies shows that lower level of complexity and the higher level of education, services income source, fairness and tax morale will impact on lower tax evasion practice happening across the countries and Marriott (2017) also support the result.So, the suggested hypothesis was: H1: Economic variable has a relationship or effect towards tax evasion practice Demographic Variable (Age, Gender, Education) A demographic variable is a variable collected by researchers with the source from demographics aspect to describe the shape and distribution of a sample for use in random statistics.It can be categorized as age, genre, education, occupation, and religion as its measurement.In this research, age, gender, and education variables are preferable more than other variables and used as the indicator's variable for demographic variable.
Age can be defined simply as how long a person has lived or a thing has existed, where this research uses the age of the taxpayer especially in Indonesia.In Indonesia, to become a taxpayer, there are 2 conditions that must be fulfilled; objective and subjective.The objective condition is regarding whether the income of the person has exceeded the specified limit or not.Subjective requirement is a condition of compliance with the provisions related to taxable objects in the income tax law.In Indonesia, a person considered a taxpayer is a person who is over 18 years old or who was born and resided in Indonesia for more than 183 days in 12 months.
Gender is a set of characteristics that are intertwined and distinguish between masculinity and femininity.These characteristics can then include gender, be it male, female, or intersex.As a social construct, eventually gender varies from society to society and can change from time to time.
Education is a situation, thing, activity, behavior, or process of changing the character and behavior of a person or group for the purpose of human maturity.Education is carried out through teaching and training with the objective to prepare an individual for a better future.There are many levels of education in Indonesia; from kindergarten to postgraduate.This research uses final educational background as a parameter to know if a person understood enough about taxes.
Research conducted by Pardisi and Mcgee (2023) revealing the three are significant differences in the attitude toward tax evasion between females and males in the youngest age group which also supported by Marriott (2017).So, the suggested hypothesis was: H2: Demographic variable has a relationship or effect towards tax evasion practice Legal and Institutional Determinants (Trust in the Governance, Tax Complexity) Legal and institutional determinants are simply regulating the legality; such as law and the institutional in a country.There are some determinants variables that is include in the legal and institutional determinants, such as; (1) legal system, (2) tax complexity, (3) democracy, (4) trust in the governance, (5) enforcement, (6) probability of detection, and (7) tax authority contract.In this research, trust in the governance and tax complexity variable is given priority over other indicators.
When talking about trust in the governance, it can be defined as how the people put their trust into the governance of a country.This means, they fully trust the government to regulate, guide and enforce the law.In Indonesia, there are many fields entrusted by the community to the government to manage, for example taxes.As mentioned earlier, tax is a mandatory payment that is either paid by a person or a business.Tax in Indonesia is used to cover the cost for the activities that are done by the governance; mostly for the infrastructure.The trust in the governance is important because this will affect how the community pays their taxes.For example, it can be said that those who place great trust in governance will certainly pay taxes regularly.But for those who don't, they won't want to pay taxes.
Tax complexity is a study that studies the complexity of the tax system in a country.Some problems that can be found in the tax complexity of a country is; (1) tax ambiguity and uncertainty, (2) frequently changes on tax law, (3) several rules, and (4) confusing forms of taxes.Tax complexity can be detrimental to a country's economy and to every tax codecompliant individual.The more complex tax regulation of a country, it can lead to some tax problem; like tax evasion.
Research about tax evasion which include trust in governance as it variable have been conducted by Hasanah and Widiyati (2021) and Palil et al. (2016) which the result shows that the variable of trust in the government affects tax evasion.Based on some research above, the hypothesis for this research is: H3: Economic variable has a relationship or effect towards tax evasion practice

METHOD
The population data used to carry out this research is Indonesian people in West Kalimantan Province.The samples taken by using purposive sampling method.The samples are the individuals who already qualified in term of age to pay taxes; which is more than 18 years old who were in the West Kalimantan.The data were collected through questionnaires distributed to individuals in West Kalimantan.To make it easier for reaching the respondents, the questionnaires that have been distributed electronically via google form.The question in the questionnaires consisting 4 sections, namely for the tax evasion, economic, demographic, and legal and institutional section.For the answer option, the questionnaires are using a Likert Scale; which the respondent will have to choose only 1 option over 5 option.The type of the questionnaires is a closed questionnaire; which help the respondents to make the decision quickly by choosing between several answer option.To ensure that the questionnaire is filled out properly by the respondent, the questionnaires is also equipped with some classification data or private information of a respondent, or demographic and economic question that contain the information on respondent's name, age, gender, last education, income level, and income source.For the purpose of the study, the number of minimum samples needed is where the sample size should be at least 10 times the number of variables studied.Therefore, at least 30 samples should be taken for this test because its uses in total 3 variables, namely economic, demographic, and legal and institutional variable, however this study use 100 sample to increase its result reliability.
In this research, the variable is measured by using the questionnaires.For the dependent variable; tax evasion variable is measured by using the Likert Scale with some conditions namely, answer option with point 1 will indicating strongly disagree and the answer option with point 5 will indicating strongly agree.For tax evasion section, it is consisting in total 5 question and 5 answer option each.For the independent variable; the economic, demographic, and legal & institutional variable is also measured by using the Likert Scale with the same conditions as the dependent or tax evasion section namely: answer option with point 1 will indicating strongly disagree and the answer option with point 5 will indicating strongly agree.For the economic variable, the total questions are 5; which 3 questions is for the income level variable and 2 questions for the income source variable.The demographic variable is also containing 5 questions; which 2 questions for age variable, 2 questions for gender, and 1 question for education level.Lastly, the legal and institutional variable is containing 5 questions; which 3 questions is for trust in the governance variable and 2 questions for the tax complexity.
The questionnaire that has been distributed managed to get a total of 120 respondents.However, this study only took a total of 100 respondents due to eliminating several answers that did not match the research criteria mentioned before and also multiple answers by the same respondent.In order to get the correct conclusion, collected data must be sort by running some test to ensure the data was valid and reliable to be used in this research.According to Field (2013) validity means "measure what is intended to be measured".In this case, Validity test used to determine whether the data collected was suitable to be used as a measurement to reach the purpose of the research.This study used Pearson Correlation Coefficient as its technique to test the data qualify/valid or not.If the result of shows that Pearson's Coefficient value > tabulated value, the indicator considered as valid.Meanwhile, reliability is concerned with the extent to which the evaluation of an object provides stable and consistent results.In this study, we use Cronbach's coefficient alpha as an instrument to test how reliable our data was.Multiple regression analysis is used as the analysis model in this research.Using the SPSS application, all of the hypothesis between the economic variable (income level and income source), demographic variable (age, gender, and education), and legal and institutional variable (trust in the governance and tax complexity) were tested in order to know the effect on tax evasion practice.The research model is described on Figure 1.The regression equation for this research is:

RESULT AND DISCUSSION Test Result
The result of testing in this research is conducted using the SPSS, with 100 samples of data and using the 5% of significance.All test that has been carried out are presented as follow: Demographic traits of sample/respondent used in this research was illustrated on Table 1.Where it can be seen that the sample was dominated by 18-25 years old respondent at 79%, followed by 26-35 years old at 16%, and the rest are above 35 years old.Female respondent was higher on the sample at 58% than male respondent which at 42%.Sample also dominated by respondent from the Pontianak as the Capital of West Kalimantan at 73%, followed by Singkawang at 9% and hence on.63% of Last education of respondent was Senior High School, 35% was Diploma -Bachelor and the rest was Master-Doctor.Mostly respondent income range was below Rp4.500.000 which take 57% of the total respondent, followed by Rp4.500.000-Rp5.000.000 at 19% and hence on.Lastly, the source of income on the first place was from Parents at 61%, second place was from Work at 18%, third place from Business at 9% and hence on.Based on Table 2, there were 100 valid data (n) from 100 respondents.For every item on the research questionnaire, the minimum value that participants can provide is 1, and the greatest value that participants can provide is 5.The average value of 9 of indicator is greater than 3 only, and 6 indicators has the average below average of 3 namely EVA4, EVA5, ECMC5, LGIN3, LGIN4, and LGIN5.Most indicator is averaging higher than 3 than the one that averaging below 3 where this could indicate that the majority of respondents concur with the provided statement.Furthermore, responses from respondents doesn't indicate a neutral answer, where it can be proven an average score that is less than or more than 3.And the standard deviation of each indicator is lower than the mean of each indicator, which indicates outlier data doesn't occur on the sample data.LGIN 1 0,696 Qualified 17 LGIN 2 0,831 Qualified 18 LGIN 3 0,817 Qualified 19 LGIN 4 0,813 Qualified 20 LGIN 5 0,594 Qualified Source: SPSS v26, (2023) Kurnianingsih and Dwi Atmoko (2022) stated that the reliability test result can be measured by using Cronbach coefficient alpha value, where the data will be considered as reliable if alpha value is greater than 0,7.The result of test that provided in Table 4 shows that each of variables consist in the study has passed reliability test.One Sample Kolmogorov Smirnov test was used to examine and determine whether the data distributed normally or not.The result of this test is presented on Table 5.The result shows 0,200, where 0,195 > 0,05, It means the data distribution was normally distributed.In this research, a test was run to examine the possibility of multicollinearity in data.Dian and Manurung (2022) believed that if the tolerance value is > 10% and the VIF value is < 10, it can be concluded that there is no multicollinearity between the independent variables in the regression model and vice versa.Table 6 is showing the result of Multicollinearity Test.The tolerance for each model is greater than 0,1 and the VIF of is lower than 10.So, Multicollinearity does not occur in the data.This research used Glejser test to check whether the heteroscedasticity occurred or not.If the significance result of each variable is higher than 0,05, the heteroscedasticity does not occur in the data.The result of heteroscedasticity is presented on Table 7 and it's found that the significance results of each variable are greater than 0,05 except the DMGR which result on 0,042.To solve this problem, researcher transform the dependent variable into natural logarithm form (LN) and reprocess the data by using the transformed dependent variable for the calculation where the result shows each variable value > 0,05.It means that the heteroscedasticity does not occur in the data, so the data can be used to test the model in this research.Coefficient of Determination need to be examined to determine the percentage of contribution/effect of independent variables towards dependent variable.Due to the independent variable is more than 2, this research used Adjusted R Square to determine the coefficient of determination.The coefficient of determination can be examined on the Table 8.The value of Adj R Square is 0,787.It means that the percentage of contribution effect of independent variables towards dependent variable is 78,7%.F-Test was used in this research to test the significance of the regression models.For the 100 sample of data, the F value must be higher than the value from the F-table.The value of Ftable for 100 sample of data was 2,7 and the significance value must be under 0,05 to consider the regression models is significant.The F-Test result was presented on the Table 9.The F value of this study is 122.621 and the significance is 0.000b.The F value is greater than 2,7 and the significance was lower than 0,05 which means the study pass the F-test.T-test was used in this study to determine the partial effect of the independent variable on the dependent variable.It also used to test the independent variable whether it affect the dependent variable or not and to determine how significance the independent variable affect dependent variable.The significance level of each independent variable must be lower than 0,05 to be considered have an effect on dependent variable.The result of T-test is presented on Table 10.It can be seen on the table; the significance of each independent variable is lower than 0,05.And for the t value, for 100 sample data, the positive t value must be larger than 1.98498 and for the negative t value must be lower than -1.98498.From the table, it's shown that ECMC and LGIN variable have a positive t value and greater than 1.98498, and for the DMRG variable have a negative t value significance -12.677 < -1.98498, so DMGR also passed the T-Test.

Economic variable and tax evasion
The conduct of tax evasion was significantly influenced by economic factors.The study found a statistically significant and positive correlation between economic factors and tax evasion.This is consistent with Wallschutzky (1984) belief in exchange indicators, according to which people may evade taxes if they believe there is a financial benefit to underreporting their taxable income, such as in the event of a high tax burden or unstable economic conditions.These economic factors affect perceived behavioral control in the context of theory of planned behavior by Ajzen (1991) because they make people feel like they have less control over adhering to tax laws when they are facing financial difficulties, which increases their propensity to commit tax evasion.For example, those with high tax loads or unstable economies due to undependable income source may be more inclined to underreport their income in order to get financial relief, which increases the likelihood of tax evasion.
The result of economic variables which can be interpreted as positive and significant effect toward tax evasion was in line with the research conducted by Gërxhani &Schram (2006), andMarriott (2017), where they also found that economic variables had an impact toward tax evasion.However, the result was contradicted with founding of Aini Haron et al., (2018),where they found income level which is the representative indicator for economic variables has no effect toward tax evasion.

Demographic variable and tax evasion
It was discovered that there is a statistically significant negative correlation between tax evasion and demographic factors such age, income, and education.Wallschutzky (1984) social orientation category and theory of planned behavior, which place an emphasis on attitudes and subjective standards, are in line with this.People who, as demonstrated by their demographic traits, have more positive views about tax compliance and a greater sense of social responsibility are less likely to engage in tax evasion.In theory of planned behavior by Ajzen (1991), the roles of attitudes and subjective norms reflect the influence of these demographic variables, since those who adhere to higher social norms are less likely to engage in tax evasion.As an example, people who are older and high educated person showed a decreased propensity to evade taxes where older people become wiser and educated person gain knowledge to not break a law, indicating the impact of demographic traits.
For the demographic variables, the result shows that demographic variable do have an impact toward tax evasion even though the result was negative and significant effect, the result was supported by previous researcher, where Pardisi and Mcgee (2023), Marriot (2017) assert that demographic variable has an impact toward tax evasion practice.In contrast, it was not in line with the research conducted by Taing & Chang (2021), where it was stated that demographic variable has no impact toward tax evasion.

Legal & Institutional variable and tax evasion
The research revealed that institutional and legal factors significantly positively influenced tax evasion behavior.Wallschutzky (1984) approach, which emphasizes the influence of institutional and legal procedures on tax evasion, is consistent with this assessment.People are more inclined to evade taxes when they believe that the law is unjust or insufficiently enforced.These factors are related to perceived behavioral control in the theory of planned behavior framework by Ajzen (1991), which affects people's perceptions of their capacity to adhere to tax laws.A thorough grasp of the complex dynamics present in the field of tax evasion is offered by this integrated viewpoint.For example, a culture where tax evasion is perceived as a more accessible option due to lack enforcement of tax regulations and trust toward the government may increase the likelihood of engaging in tax evasion.Similarly, tax complexity may result in individuals with lower levels of education choosing not to pay taxes because they are unable to comprehend the tax system.
Lastly, the result of legal & institutional variable shows that legal & institutional variable also has an impact toward tax evasion which is positive and significant.The result was similar with Rashid (2020) where he found that tax complexity has positive impact toward tax evasion.Cullen et al. (2018), Hasanah and Widiyati (2021) and Palil et al. (2016) also found that trust in government affect tax evasion.On contrary, the result was not similar with Komang et al. (2017), where he asserts that tax system/complexity as negative effect toward tax evasion, and Taing & Chang (2021) result also contradictive with the result, where he has found that attitude toward the government in term of trust has no effect on tax evasion.

CLOSING Conclusion
Based on the statistical analysis that has been done, it can be concluded that every hypothesis is accepted, and H0 was rejected.Economic variable and legal & institutional variable have an impact toward tax evasion, specifically have positive and significant impact towards tax evasion practice.On the other hand, the demographic variable also has an impact toward tax evasion practice, where it has a negative and significant effect towards tax evasion.The coefficient determination asserts that the percentage of contribution or effect of independent variables towards dependent variable is 78,7%, where it means economic, demographic and legal & instructional variable on this study are able to explain the tax evasion with the contribution of 78.7%.

Suggestion
This study has several limitations, the sample of this study was limited.So, the next suggestion is to take larger sample where it may give a different result.The study only targets an individual who lived in West Kalimantan, So the data may not so vary, the next suggestion was to extent and widen the data collection to make the data more vary.

Table 3 ,
described the result of the validity test.To run the validity test, this research uses Pearson Correlation.By using 100 sample and 5% level of significance, data is considered valid if obtained a value that is greater than the critical value from the r table.In this case, for 100 sample, the 0.1966 is the critical value and the result illustrate that each of indicator's value > 0.1966.