Keywords

Electronic Fiscal Devices, EFD, Taxation, Small Business Owners.

Introduction

The sustainability of the economy of a nation depends on the adequacy of funds to run its operations (Chung & Trivedi, 2003). In order to generate funds, the government depends on various sources of income, one of which is taxation (Murphy, 2008). In Tanzania, taxation forms 12 percent of the total gross domestic product (GDP) (World Bank Africa Region, 2015). Some taxes are directly imposed on the citizens, such as income tax, while others are indirectly collected such as value added tax (Ngeni, 2016). Generally, the government has designed different means of collection which can cater for each taxation need, as applicable in the differing nature of businesses (Tehulu & Dinberu, 2014). For example, VAT is charged per benefits acquired, while taxes on small businesses are based on turnover (Marino & Zizza, 2012).

The Tanzanian government understands the value of integrating information systems to facilitate the monitoring of the tax compliance of its taxpayers. In ensuring that the government attains financial sustainability, one of its strategies is to ensure that businesses are using Electronic Fiscal Devices (EFDs) in their transactions (Nyamwananza, 2014). This is because the EFDs keep the record of all business transactions, which become evidence for determining tax liabilities (Eilu, 2018). However, the acceptance of EFDs to the taxpayers’ community is still a challenge. The study by Omweri, et al. (2011) conducted among VAT traders confirmed the increase of revenue collection after the deployment of EFDs in Tanzania.

Apparently, the deployment of a new element to anything comes with its challenges; in this case the introduction of Electronic Fiscal Devices is not exceptional. In Tanzania, some traders refuse to purchase a fiscal device because they feel it burdens them with unnecessary expenses (Lubua, 2014). On the other hand, some of those who possess EFDs refuse to issue receipts per every transaction (Geetha & Sekar, 2012). Other taxpayers collude with their customers and issue false receipts which do not represent the true value of the goods purchased (Murphy, 2008). At this point, the government is obliged to re-enforce compliance through its laws, with fines, imprisonment and penalties being the methods used for this (Tanzania Revenue Authority, 2018).

Although punishment is used in enforcing tax compliance, its success is influenced by several factors such as social, economic, cultural and physiological factors (Raskolnikov, 2006). Therefore, punishment needs to be strategized to get the best outcome from it (Chen et al., 2012). In the current study, and the Tanzanian context, the question remains whether punishments due to the non-use of EFDs have a significant impact on the rate of use. In practice, some tax administrations believe in severe punishments to ensure compliance, while some do not (Mohdali et al., 2014); the relevance of these positions in Tanzania depends on direct empirical evidence. With this background, the study responds to the question where the level of punishment impacts the rate of EFD use.

Study Problems

The tax administration of Tanzania adopted the Electronic Fiscal System with the aim of increasing its revenue collection from business owners. The system is thought to enhance the transparency of information between the revenue authority and the taxpayer (Chung & Trivedi, 2003). In the process of implementing the use of EFDs, different phases were used with the most recent phase requiring small business owners to use the system as well. Arguably, above 70 percent of people engaged in business fall under the category of small business owners; however, their contribution to the GDP is very low compared to their counterpart (Tanzania Revenue Authority, 2018). The contribution by small business owners toward tax collection is small as well. In this case, the introduction of EFDs is thought to boost tax collection among businesses of this category (Chan, 2012; Lubua, 2014). Unfortunately, this is met with stiff opposition from owners of businesses. Therefore, the Tanzania Revenue Authority uses different sessions to educate taxpayers on this responsibility. Moreover, it has introduced punishments for noncompliance (Tanzania Revenue Authority, 2018). However, the contribution of punishment to the rate of EFD use is not clear amid imposed punishments.

Study Importance

The importance of this study follows from the primary objective of the study which was to ultimately determine the impact of the taxpayer perception on EFD use. Given the significant number of small businesses in Tanzania and the importance of EFD uses in the collection of tax revenue in Tanzania, any significant relationship between perception and actual EFD use could be used in determining optimal tax policy in Tanzania.

Study Objectives

The first objective of the study was to determine the categorical relationship between demographic variables (age, gender and level of education) and the perceived fear of punishment. The second objective of the study was to determine the impact of the perceived fear of punishment (taxpayer perception) to the rate of using EFDs among small business owners in Tanzania.

Literature Review

Entrepreneurship Generally, choices by business owners are aimed at maximising their income (Muhrtala & Ogundeji, 2013; Gurama et al., 2015). In taxation, some identify loopholes in the tax system to evade tax for the sake of enhancing their income (Mohdali et al., 2014; Kiow et al., 2017). To address this challenge, the Revenue Authority of Tanzania enacted laws to punish tax evaders. The punishment is regarded as a deterrent (Ali, Yussof, & Isa, 2014). Arguably, punishments become useless if the ability to detect tax evaders is low. Therefore, strategies for detecting tax evasion are introduced, which include the use of EFDs in capturing the information from taxpayers for tax establishment (Tanzania Revenue Authority, 2018). Generally, it is a practice that if the business fails to comply with tax laws, it is subjected to punishment. The current study concentrated on the impact of punishment on enforcing the use of EFDs. This part of the literature addresses elements such as the relationship between the demographic factors and the fear of punishment. It also addresses the relationship between the fear of punishment and the rate of using the new technology. The technology in focus is the use of EFDs.

Theories of EFD use

This section discusses theories related to punishment and compliance in taxation and technology use. According to the Standard Model of tax evasion, there are two determinants of compliance in tax compliance: the probability of being detected and the level of punishment (Muhrtala & Ogundeji, 2013; Gurama et al., 2015). The assumption behind this theory is that taxpayers will always do things which increase their profit. If taxpayers find a loop hole to avoid tax, they are more likely to exploit it than comply (Sá, Martins, & Gomes, 2014). This statement is supported by studies such as those by Batrancea et al. (2012); Badara (2012), who confirmed that taxpayers would exploit the opportunity for non-compliance presented by their environment. Factors which increase non-compliance opportunities include corruption among taxpayers, inadequate staff to follow-up corruption activities, and even weak tax laws (Armborst, 2017). On the other hand, the variable known as the level of punishment identified by the Standard Model of tax evasion is discussed in studies such as those by Muhrtala and Ogundeji (2013); Gurama et al. (2015) on factors determining tax compliance. If the punishment is high, taxpayers will fear non-compliance, and vice versa. Based on the available literature, the impact of the fear of punishment on the rate of using EFDs is not adequately discussed, especially in the Tanzanian context (Muhrtala & Ogundeji, 2013; Gurama et al., 2015). Since the use of EFDs is an important aspect of tax compliance (refer Badara, 2012), the current study decided to determine whether the fear of punishment impacts the rate of punishment. This will add new information to existing theories.

According to the technology acceptance model of 2008, there is no direct relationship depicting the impact of the fear of punishment on the rate of using the new technology (Venkatesh & Bala, 2008). A close relationship to this is depicted by the impact voluntariness on the usefulness of the technology (Mandari et al., 2017). Based on the technology acceptance model, the voluntariness of using the technology enhances the perceived usefulness of the technology, which increases the readiness of using such a technology. This voluntariness happens when the user perceives direct benefits from the use of the technology (Selamat & Jaffar, 2011). In the study of using EFDs, it is difficult to draw a line between what the user perceives as benefit and what is considered as not beneficial. For example, EFD use is perceived as a good tool in keeping records, but users may equally consider it as a tool that will increase their tax dues. Therefore, it may be difficult to have businesses resolving to use the technology voluntarily, without government push (Dwivedi et al., 2019). In the context of the current study, a voluntary use of the technology is the opposite of the use of the technology without the fear of punishment. Since the government of Tanzania also uses punishment to ensure compliance, the current study decided to test its impact on the rate of using EFD.

Relating Demographic Variables with the Fear of Punishment

The literature suggests demographic variables that associate with different behaviours of people. For example, the study by Chan (2012); Lubua (2014) affirmed that in taxation, demographic factors are perceived to influence the behavioural decision taken by each taxpayer towards voluntary compliance. In technology adoption, this behaviour is also known as the rate of using the technology. While the next subsection discusses the behaviour of respondents (as the rate of EFD use) in relation to the fear of punishment, the current subsection focuses on demographic variables and the fear of punishment. The literature identified several demographic factors such as age, level of education and experience in business as the most discussed in the subject of taxation. For example, Hofmann et al. (2017) suggested that seniors are more fearful of punishment due to non-compliance. This argument is influenced by their own experience on the impact of non-compliance and the fear of ruining their reputation (Mohdali et al., 2014).

In addition, females (in Africa) are considered as more fearful of punishment, which drives many of them to tax compliance (Chung & Trivedi, 2003; Hofmann et al., 2017). Other studies put forth the aggressiveness and physical tolerance of men as a factor toward their fearlessness attitude (Lane, 2012; Curiel & Bishop, 2018). In this regard, the two groups would need different approaches in enforcing their compliance (Gwaro et al., 2016). In addition, the study by Curiel and Bishop (2018) cited education to be an important determinant of compliance, where those with college education are more inclined to comply, than their counterparts. With the information on demographic variables, the study developed hypotheses to test their categorical relationship to the fear of punishment.

The Rate of using EFDs

The introduction of EFDs to business owners was not well received and thus implementation was not smooth (Gemmell & Ratto, 2017). At first, business owners rejected acquiring EFDs, arguing that buying the device was costly (Tanzania Revenue Authority, 2018). Following dialogues with the government, a consensus to implement was arrived at and the implementation proceeded. Studies conducted in other settings of Africa show that users of EFDs may accept the use, while in practice they do not issue tax invoices to their clients (Deyganto, 2018; Tehulu & Dinberu, 2014). This is the same assumption which obliged the Tanzania Revenue Authority to enact laws which allow punishment to those who fail to comply with this government directive (Tanzania Revenue Authority, 2018). Examples of such punishment include fines, penalties and even imprisonment. With this government position, the question remains whether enforced punishment significantly impacts the rate of using EFDs. Bearing this in mind, the next hypothesis was developed.

During this study, data on the impact of punishment on the rate of using EFDs were limited. However, studies exist that present information on how the punishment impacts the general tax compliance. Since the rate of EFD compliance is expected to correlate with tax compliance (Chan, 2012; Doran, 2009), the same information is used to understand how the perceived punishment would affect the rate of EFD use. According to the study by Batrancea, Nichita and Batrancea (2012), penalties, fines and the interest on accrued amount are important to enforce compliance if not abused. These penalties are abused in a case where corruption thrives or even in a case where employees lack the required competency to compute the outstanding tax liability and their implication on penalties (Gberegbe & Umoren, 2017). Another aspect which can lead to the abuse of penalties and fines occurs where the tax officer is required to use professional judgment to determine the appropriate level of punishment for the taxpayer; this nature of subjectivity can affect the level of fairness and lead to complaints. Since punishment is proven to impact tax compliance in general, the current study was interested to determine whether the same affects the attitude of taxpayers toward using EFDs among small business owners. To accomplish this objective, the next null hypothesis was formulated.

Study Hypotheses

The following null hypotheses were relevant to the study:

H₁: There is no categorical relationship between gender and the fear of punishment among small business owners.

H₂: There is no categorical relationship between age and the perceived fear of punishment among small business owners.

H₃: There is not categorical relationship between the level of education and the fear of punishment among business taxpayers.

H₄: The fear of punishment does not significantly impact the rate of using EFDs among small business owners of Tanzania.

Methodology

This was a quantitative study, with the survey research design at its core. The knowledge was acquired objectively, through hypotheses testing (Collins, 2010). This approach allows other researchers to verify the relevance of findings obtained (Makombe, 2017). Furthermore, the approach enables the study to make generalisations to other settings with characteristics similar to those of the population of the study (Tuli, 2010; Collins, 2010). In operationalising the study, the Arusha tax region of Tanzania was selected as the research area because it is among the top regions in tax contribution, but with mixed payers (Tanzania Revenue Authority, 2018). In addition, the study focused on the small business owners of the Arusha tax region as the population for study because they form the largest category of taxpayers in the country, with above 2/3 of the whole population (Tanzania Revenue Authority, 2018).

In order to obtain the sample, the study further defined its population to only include small business owners who visited the Arusha revenue office, in the month of January 2019. To qualify the respondents also had to possess the Electronic Fiscal Device. An estimation of 1000 small business owners, qualified for these criteria. Following Bartlett, Kotrlik & Higgins (2001), at the standard error of 0.05, at least 213 units of the sample can embody 1000 units of the population. Therefore, the sample for this study comprised 279 units. The systematic sampling method was employed by utilising names of respondents documented to the register in January 2019. Systematic sampling supports the quantitative research approach (Wilson, 2010).

The study used a closed-ended questionnaire to obtain data. This type of questionnaire is popular because it allows the collection of a large amount of data, and it supports the quantitative analysis adopted by the current study (Geetha & Sekar, 2012). According to Table 1, the questionnaire had five variables, and gender was the only binary variable. Other variables of the questionnaire were ordinal in nature; therefore, the Likert scale was used. Before using the questionnaire, the researchers ensured that it received ethical clearance from the University of KwaZulu-Natal to ensure that the study is compliant to ethical standards. The data collection tool was a closed-ended questionnaire; and the researcher personally collected data from small business owners who visited the Revenue Authority of Tanzania.

Table 1: Key Variables Of The Study
Variable	Information measured	Scale
Gender	The gender category	Categorical data
Age	The age group of the respondent	Likert scale
Level of education	The acquired level of education	Likert scale
Fear of punishment	The taxpayer’s perception of the severity of the punishment due to non-compliance	Likert scale
Rate of using EFDs	The perceived rate of using EFDs in percent	Likert scale

In addition, the study used two approaches to attain the validity of data collected. The validity of data is important because users want to know whether the solution provided to the problem is relevant or not (Doran, 2009). The first approach of assessing the validity of the questionnaire is based on face validity. The face validity ensured that different contents of the questionnaire provided the right measurement for the study. This was accomplished through two academic experts, in this research field, who reviewed the questionnaire and ascertained that it met the requirement for capturing data for testing hypotheses. In addition, the study used the exploratory factor analysis (EFA) to test the coherence of items of the questionnaire. This approach is supported by Williams (2007) in testing validity of items in the questionnaire. The study observed the Kaiser-Meyer-Olkin measure of sampling adequacy as 0.801. The approved minimum threshold is 0.6; therefore, the sample met the validity requirement (Ball, 2019; Taherdoost, 2018). Furthermore, in the component matrix, each factor had a construct with at least 0.3 correlation level, which is the minimum threshold. Furthermore, the study used the Cronbach alpha to test the reliability of the questionnaire. In this case, the study observed that all relationships used in testing the inter item consistence had the minimum acceptable alpha ratio. According to Taber (2018), the minimum acceptable ratio is 0.6. The results in Table 2 show the acceptable level of reliability.

Table 2: Reliability Testing
Reliability test	Cronbach alpha	0.813
	Cronbach Alpha Based on Standardised Items0	0.783
	Number of Items	9
Exploratory factor analysis (EFA)	Kaiser-Meyer-Olkin measure of sampling adequacy	0.801
	Component matrix (Principal component analysis)	All factors had above 0.3
	Number of items in the questionnaire	10

Subsequently, the study used the Social Science Statistical Package (SPSS) in analysing the data collected. In analysing data, certain statistical models were employed: First, descriptive statistics were applied to know the frequency of data in every variable. Secondly, the one-way analysis of variance (ANOVA) was used to test the categorical relationships between variables. The one-way ANOVA is relevant in testing categorical relationships where variables are in a Likert scale (Murphy, 2008). Furthermore, the study used the ordinal regression model to test the impact of independent variables to respective dependent variables. The ordinal regression model is relevant in testing the influence of one variable to another, where the dependent variable is in a Likert (ordinal) scale (Collins, 2010).

Empirical Results

Data Analysis and Results

This section presents results of the research governed by two objectives. The first objective determined the categorical relationship between demographic variables (the age, gender and level of education) and the perceived fear of punishment. The second objective determined the impact of the perceived fear of punishment on the rate of using EFDs. The study targeted small business owners who use EFDs in Arusha, Tanzania. Before addressing the objectives of the study, a preliminary analysis was conducted to determine descriptive information of demographic variables of the study.

According to Table 3, there were more male (68%) than female (38%) respondents in the sample. Since data were obtained through systematic probability sampling, it is safe to suggest that the studied category of business owners was dominated by men, as opposed to women. This position is supported by Aikaeli & Mkenda (2014); Muhanga (2017), who confirmed more men in microbusinesses (in Tanzania) than women. Another part of the analysis revealed that small businesses are more dominated by young individuals as 69% of all respondents were 40 years and below. This age category comprises the active working class in the society. This finding is supported by the Ministry of Trade and Industry (2012) and Isaga (2015) who acknowledged the same category to feature extensively in the Tanzanian economic activities. The other studied variable is the level of education, and findings listed in Table 3 indicate that most of small business owners had secondary education and below (60%). This is possibly because people with a high level of education tend to opt for employment rather than self-employment (Chen, Harold, Little, Mark, & Zhao, 2012).

Table 3: Demographic Characteristics Of The Participants
Characteristic	Category	Frequency	Percent
Gender	Male	106	38
	Female	173	62
	Total	279	100
Age	18-30 years	98	35
	31-40 years	96	34
	41-50 years	68	24
	Above 51 years	17	6
	Total	279	100
Level of Education	Primary education	79	28
	Secondary education	90	32
	Certificate or Diploma	88	32
	Degree and above	22	8
	Total	279	100

Demographic Variables and the Fear of the Punishment

In addition, the study used different models to test the categorical relationship between demographic variables and the perceived fear of punishment. Table 4 presents the results of the chi-square and the one-way ANOVA model as observed in the analysis.

Table 4: Categorical Relationship Testing
	Output variable	P-value	Analytical model
Gender	Fear of punishment	0.462	Chi Square test
Age	Fear of punishment	0.365	One Way ANOVA
Level of education	Fear of Punishment	0.841	One Way ANOVA
Experience in business	Fear of punishment	0.994	One Way ANOVA

The first analysis tested the null hypothesis which suggested that there is an insignificant categorical relationship between the gender of respondents and their fear of punishment due to the none use of EFDs. The analysis used the chi-squared model to test the relationship because the gender of respondents is a dichotomous variable. According to results summarised in Table 4, the observed p-value was 0.462. With these results, it is confirmed that there is no significant categorical relationship between the gender of respondents and the fear of punishment. Therefore, the fear of punishment between male and female is almost the same. This is possibly due to the taxpayer’s education constantly shared by the revenue authority to all business owners (Tanzania Revenue Authority, 2018). This position is in contrast to the prominent belief among Africans that females are more responsive to fear than males (Hofmann et al., 2017).

Furthermore, the study used the one-way ANOVA to test the null hypothesis suggesting that there is no significant categorical relationship between the age of respondents and the perceived fear of punishment. According to results presented in Table 4, the p-value is 0.365, which is greater than the threshold p-value of 0.05. This indicates an insignificant categorical relationship between the age of respondents and the fear of punishment. Therefore, the age of the respondent is not a determinant of the fear of punishment due to EFD non-compliance. However, this observation is against the findings of Hofmann et al. (2017); Deyganto (2018), who observed senior citizens to be more fearful of punishment than the younger generation. The reason could be the similarity on the level of awareness across age groups.

In addition, Table 4 presents results for the analysis of the null hypothesis suggesting that the level of education does not have a significant categorical relationship with the fear of punishment. The one-way ANOVA results indicate that the p-value is 0.994, which is greater than the threshold. This information suggests no significant categorical relationship between the level of education and the fear of punishment. It must be understood that regardless of the level of education, the most important knowledge is how to comply with tax laws. It is possible that differences in understanding tax issues were addressed through ongoing training issued by the revenue authority (Tanzania Revenue Authority, 2018); therefore, the knowledge of respondents was equivalent throughout. Considering this situation, it was difficult for the level of education to have a significant difference in its relationship with the fear of punishment.

The Impact of the Fear of Punishment on the Rate of Using EFD

The second objective of this study determined the impact of the fear of punishment on the rate of using EFD among small business owners of Arusha, Tanzania. This was motivated by the fact that the Tanzania Revenue Authority instituted different punitive measures to those who fail to adopt the use of electronic fiscal devices.

The study began with an analysis to test the null hypothesis suggesting that there is not a significant categorical relationship between the fear of punishment and the rate of EFD use. In this case, the study used the one-way ANOVA model. According to results presented in Table 5, the observed p-value is 0.178. The p-value is greater than the threshold. Therefore, there is no significant categorical relationship between the fear of punishment and the rate of EFD use.

In addition, the study tested the impact of the fear of punishment on the rate of using EFD. In this process, the study used the ordinal regression to conduct the analysis. According to results in Table 6, the two variables fit to the ordinal regression model because the model fitting p-value is 0.037. However, none of the parameter estimates of the perceived fear of punishment showed a significant difference from the reference value. All of the p-values for parameter estimates are above 0.05. Furthermore, the observed Nagelkerke Pseudo r-square is 0.025, which is equally low. In this case, the fear of punishment is not a drive toward understanding the rate of EFD use expressed by an individual. This suggests that the re-enforcement of punishment by the revenue authority and the developing fear do not impact the rate of compliance to EFD uses (Tanzania Revenue Authority, 2018). People would not fear the punishment if the regime is corrupt to the extent that it provides a lot of avenue for avoidance (Chung & Trivedi, 2003; Lubua & Pretorius, 2019). Also, people would be less fearful if the magnitude of the punishment is easy to carry because it causes low harm to evaders (Cuccia, 2013).

Table 5: Fear Of Punishment And Rate Of Efd Use -One-Way Anova Test
	Sum of squares	df	Mean Square	F	Sig.
Between groups	9.135	4	2.284	1.588	0.178
Within groups	393.926	274	1.438
Total	403.061	278

Table 6: Fear Of Punishment The Rate Of EFD Use-Ordinal Regression Test
Element of measurement		Value
Model fitting information		P-value = 0.037
Nagelkerke Pseudo r-square		r-square = 0.025 or 2.5%
Parameter estimates for the input variable (level of punishment)	High	p-value = 0.701
	Moderate	p-value = 0.296
	Low	Reference value

Conclusion and Recommendations

Conclusion

The major aim of this study was to determine the impact of the fear of punishment on the rate of using electronic fiscal devices among small business taxpayers. This was guided by two specific objectives. The first objective determined whether there is a significant categorical relationship between demographic characteristics and the fear of punishment. The following demographic variables were included in the analysis: Age, gender, and the level of education. In addressing this objective, the study found no categorical relationship between any of the demographic variables and the fear of punishment. The second objective of the study determined the impact of the fear of punishment on the rate of EFD uses. In this analysis, neither a categorical nor causal relationship was found between the fear of punishment and the rate of using EFDs in business.

Recommendations

The main implication of this study is its ability to provide relevant information to the Tanzania Revenue Authority that demographic variables such as gender, age and education are not useful in understanding the response of small business owners toward punishment. Furthermore, punishment alone is not adequate in influencing the use of EFDs among small business owners.

Prior to considering the recommendations of the study it is important to acknowledge the limitations to the current study. The first limitation was that the study was conducted among small business owners who use EFDs. Results could be otherwise if the target population was different. Other limitations include the fact that the study was conducted in Arusha in Tanzania, and the fact that the study was quantitative in nature.

The study recommends the following:

1. The Revenue Authority of Tanzania should not use age, gender, and level of education as factors for determining the effectiveness of punishment among small business taxpayers.

2. The revenue authority should not use punishment to create fear as the tool to influence compliance with EFD use among small business owners. This is because the level of fear does not impact the rate of EFD use.

References

1. Al-Mamun, A., Entebang, H., Mansor, S.A., &amli; Yasser, Q.R. (2014). The Imliact of Demogralihic Factors on Tax Comliliance Attitude and Behavior in Malaysia. Journal of Finance, Accounting and Management, 5(1), 109-124.
2. Armborst, A. (2017). How Fear of Crime Affects liunitive Attitudes. Euroliean Journal on Criminal liolicy and Research, 23(3), 461-481.
3. Chan, C.T. (2012). An Exlianded Model of Taxliayer Comliliance: Emliirical Evidence from the United States and Hong Kong. Journal of International Accounting, Auditing and Taxation, 9(2), 83-103.
4. Chen, Y., Harold, T., Little, J., Mark, T., &amli; Zhao, Q. (2012). Factors Motivating the Adolition of e-Learning Technologies. Journal of e-Learning &amli; Higher Education, 1-17.
5. Chung, J., &amli; Trivedi, V.U. (2003). The Effect of Friendly liersuasion and Gender on Tax Comliliance Behavior. Journal of Business Ethics, 47(2), 133-145.
6. Collins, H. (2010). Creative Research: The Theory and liractice of Research for the Creative Industries. New York: AVA liublications.
7. Cuccia, A. (2013). The economics of tax comliliance: what do we know and where do we go? Journal of accounting literature, 13, 81-116.
8. Curiel, R.li., &amli; Bisholi, S.R. (2018). Fear of Crime: The Imliact of Different Distributions of Victimisation. lialgrave Communications, 4(46), 6-12.
9. Deyganto, K.O. (2018). Factors Influencing Taxliayers’ Voluntary Comliliance Attitude with Tax System: Evidence from Gedeo Zone of Southern Ethioliia. Universal Journal of Accounting and Finance, 6(3), 92-107.
10. Al-Mamun, A., Entebang, H., Mansor, S.A., &amli; Yasser, Q.R. (2014). The Imliact of Demogralihic Factors on Tax Comliliance Attitude and Behavior in Malaysia. Journal of Finance, Accounting and Management, 5(1), 109-124.
11. Armborst, A. (2017). How Fear of Crime Affects liunitive Attitudes. Euroliean Journal on Criminal liolicy and Research, 23(3), 461-481.
12. Chan, C.T. (2012). An Exlianded Model of Taxliayer Comliliance: Emliirical Evidence from the United States and Hong Kong. Journal of International Accounting, Auditing and Taxation, 9(2), 83-103.
13. Chen, Y., Harold, T., Little, J., Mark, T., &amli; Zhao, Q. (2012). Factors Motivating the Adolition of e-Learning Technologies. Journal of e-Learning &amli; Higher Education, 1-17.
14. Chung, J., &amli; Trivedi, V.U. (2003). The Effect of Friendly liersuasion and Gender on Tax Comliliance Behavior. Journal of Business Ethics, 47(2), 133-145.
15. Collins, H. (2010). Creative Research: The Theory and liractice of Research for the Creative Industries. New York: AVA liublications.
16. Cuccia, A. (2013). The economics of tax comliliance: what do we know and where do we go? Journal of accounting literature, 13, 81-116.
17. Curiel, R.li., &amli; Bisholi, S.R. (2018). Fear of Crime: The Imliact of Different Distributions of Victimisation. lialgrave Communications , 4(46), 6-12.
18. Deyganto, K.O. (2018). Factors Influencing Taxliayers’ Voluntary Comliliance Attitude with Tax System: Evidence from Gedeo Zone of Southern Ethioliia. Universal Journal of Accounting and Finance, 6(3), 92-107.
19. Geetha, R., &amli; Sekar, M. (2012). E-Filing of Income Tax: Awareness and Satisfaction Level of Individual Taxliayers in Coimbatore City, India. Research Journal of Management Sciences, 1(4), 6-11.
20. Gemmell, N., &amli; Ratto, A.M. (2017). The Effects of lienalty Information on Tax Comliliance.
21. Gray, D.E. (2014). Theoretical liersliectives and Research Methodologies. In D. E. Gray, Doing Research in the Real World. Greenwich: SAGE. 15-38.
22. Gwaro, O.T., Maina, K., &amli; Kwasira, J. (2016). Influence of Online Tax Filing on Tax Comliliance among Small and Medium Enterlirises in Nakuru Town, Kenya. IOSR Journal of Business and Management , 18(10), 82-92.
23. Hofmann, E., Voracek, M., Bock, C., &amli; Kirchler, E. (2017). Tax comliliance across sociodemogralihic categories: Meta-analyses of survey studies in 111 countries. Journal of Economic lisychology, 62, 63-71.
24. Kelliher, F. (2005). Interliretivism and the liursuit of Research Legitimisation: An Integrated Aliliroach to Single Case Design. The Electronic Journal of Business Research Methodology, 3(2), 123-132.
25. Krauss, S. (2005). Research liaradigms and Meaning Making: A lirimer. Qualitative Reliort, 10(4), 758-770.
26. Lane, J. (2012). Theoretical Exlilanations for Gender Differences in Fear of Crime. In J. Lane, Routledge International Handbook of Crime and Gender Studies. London: Routledge, 55-72.
27. Lubua, E.W. (2014). Influencing Tax Comliliance in SMEs through the Use of ICTs. International Journal of Learning, Teaching and Educational Research, 2(1), 80-90.
28. Lubua, E.W., &amli; liretorius, li.D. (2019). Cyber-security liolicy Framework and lirocedural Comliliance in liublic Organisations. liroceedings of the International Conference on Industrial Engineering and Olierations Management (lili. 1-13). liilsen: IEOM Society International.
29. Makombe, G. (2017). An Exliose of the Relationshili between liaradigm, Method and Design in Research. Gordon Institute of Business Science-University of liretoria. liretoria: The Qualitative Reliort.
30. Mannion, R., Blenkinsolili, J., &amli; liowell, M. (2018). Understanding the Knowledge Galis in Whistleblowing and Slieaking uli in Health Care: Narrative Reviews of the Research Literature and Formal Inquiries, a Legal Analysis and Stakeholder Interviews. Southamliton (UK): NIHR Journals Library.
31. Marino, M.R., &amli; Zizza, R. (2012). liersonal Income Tax Evasion in Italy: An Estimate by Taxliayer Tylie. Tax Evasion and the Shadow Economy, 1(1), 33-60.
32. Mohdali, R., Isa, K., &amli; Yusoff, H.S. (2014). The Imliact of Threat of liunishment on Tax Comliliance and Non-Comliliance Attitudes in Malaysia. International Conference on Accounting Studies 2014 (lili. 291-297). Kuala Lumliar: Elsevier.
33. Murlihy, K. (2008). Enforcing Tax Comliliance: To liunish or liersuade? Economic Analysis and liolicy, 38(1), 113-135.
34. Ngeni, F.B. (2016). Tax Administration in Tanzania: An Assessment of Factors Affecting Tax Morale and Voluntary Tax Comliliance towards Effective Tax Administration. International Journal of Finance and Accounting, 5(2), 90-97.
35. Nyamwananza, T.M. (2014). An analysis of SMEs attitude and liractices towards tax comliliance in Zimbabwe. Sage olien.
36. Omweri, M.L., Obongo, B.M., Obara, li.M., &amli; Onsongo, C.O. (2011). The effectiveness of electronic tax registers in lirocessing of value added tax returns: liersliectives from registered VAT taxliayers in Kisii Town, Kenya. African Journal of Business &amli; Management, 1, 44-54.
37. Raskolnikov, A. (2006). Crime and liunishment in Taxation: Deceit, Deterrence, and the Self-Adjusting lienalty. Columbia Law Review, 106(3), 569-642.
38. Tanzania Revenue Authority. (2018). Effect of usage of EFD Mashines in lirivate sector. Dodoma: Tanzania lirinting documents.
39. Tehulu, T.A., &amli; Dinberu, Y.D. (2014). Determinants of Tax Comliliance Behavior in Ethioliia: The Case of Bahir Dar City Taxliayers. Journal of Economics and Sustainable Develoliment, 5(15), 268-273.
40. Tuli, F. (2010). The Basis of Distinction Between Qualitative and Quantitative Research in Social Science: Reflection on Ontological, Eliistemological and Methodological liersliectives. Ethioliian Journal of Education &amli; Science, 6(1), 98-108.
41. Wilson, J. (2010). Essentials of Business Research: A Guide to Doing Your Research liroject. Calfornia: SAGE liublications.
42. World Bank Africa Region. (2015). Why should Tanzanian liay taxes. [Internet: &nbsli;httli://www.worldbank.org/tanzania/economiculidate retrieved on June 7, 2020].