Determinants of Tax Revenue Collection in Ethiopia: A Case of

Sululta Town Revenue Office

A Thesis Submitted to the School of Graduate Studies of Jimma

University in Partial Fulfilment of the Requirements for the Award of the
Degree of Masters of Science in Accounting and Finance (MSc)

BY:
ZEWUDE TARIKU BIKILA

JIMMA UNIVERSITY
COLLEGE OF BUSINESS AND ECONOMICS
DEPARTMENT OF ACCOUNTING AND FINANCE
MSC PROGRAM

JUNE, 2021
JIMMA, ETHIOPIA
Determinants of Tax Revenue Collection in Ethiopia: A Case of
Sululta Town Revenue Office

BY:
ZEWUDE TARIKU BIKILA

Under the guidance of:

Mr. Abel Worku (Assistant professor)
And
Mr. Hayatu Hairu (MSC)

A Thesis Submitted to the School of Graduate Studies of Jimma

University in Partial Fulfilment of the Requirements for the Award of the
Degree of Masters of Science in Accounting and Finance (MSC)

JIMMA UNIVERSITY
MSC PROGRAM

JUNE, 2021
JIMMA, ETHIOPIA
 CERTIFICATE
This is to certify that the thesis entitles “Determinants of Tax Revenue Collection in Ethiopia:
A Case Sululta Town Revenue Office”, submitted to Jimma University for the award of the
Degree of Masters of Science in Accounting and Finance (MSC) and is a record of bonafide
research work carried out by Mr. Zewude Tariku Bikila, under our guidance and supervision.

Therefore, we hereby declare that no part of this thesis has been submitted to any other
university or institutions for the award of any degree or diploma.
Main Adviser’s Name Date Signature
_______________ ___________ _________
Co-Advisor’s Name Date Signature
________________ ____________ _________
 DECLARATION
I hereby declare that this thesis entitled “Determinants of Tax Revenue Collection in
Ethiopia: A Case of Sululta Town Revenue Office”, has been carried out by me under the
guidance and supervision of Mr. Abel Worku (Assistant Professor) And Mr. Hayatu Hairu
(MSc).

The thesis is original and has not been submitted for the award of any degree or diploma to
any university or institutions.

Researcher‟s Name Date Signature

_________________ ___________ ________
 Abstract

The main objective of this study was to investigate the major determinants of tax revenue
collection. To achieve this objective a cross-sectional data was used. An explanatory and
descriptive research designs were applied to investigate the relationship that exists between
the level of tax revenue collection and its determinants. In this study, a mixed research
approach was applied. Both primary and secondary sources of data were utilized for this
study. A self-administrated structured questionnaire and interview were distributed to the
target population. As a result of Ethiopian business income taxpayer’s nature, a stratified
sampling method was used, and total of 279 sample data was collected from each category of
taxpayers, and an interview using purposive sampling method was applied to collect data
from employees of the tax revenue office. Due to the qualitative nature of the response
variable ordered logistic model was used with three categories. Descriptive and inferential
methods were used for data analysis. The results from the analysis suggest that illicit
financial out flows are found to be significant determinants of tax revenue collection in
Sululta town. The study has revealed technology and organizational strength of tax authority
have a positive and significant effect on tax revenue collection. The study further revealed an
increase in the level of political instability and tax delinquency decreases the probability of
collecting a high level of tax revenue. The study recommends the government should
strengthen tax authority by providing the latest technology that enables to collect tax
appropriately.

Key Words: Tax revenue collection, ordered logistic model, Tax delinquency

i
 Acknowledgments

Foremost, my ultimate thanks are to the almighty God for the gift of life, forgiveness, good
health, mercy and His grace of provision to achieve this success. I am indeed grateful to my
advisors Mr. Abel Worku (Assistant Professor) and Mr. Hayatu Hairu (MSc) who
professionally and skilfully supported and guided me in this research project.

I would like to warmly acknowledge the Sululta town revenue office for their support in
providing me all the required information and guidance. The cooperation of the sampled
taxpayers is also highly inspiring I would like to say thank you.

Lastly, but not least my special thanks go to my family, for their decent support along my
way. It gives me the joy to express my thanks to all my friends for their enormous support in
collecting data and moral support which ever inspires me to do my best in my endeavour.
May God bless you all!

ii
 Table of Contents
Abstract ....................................................................................................................................... i
Acknowledgments......................................................................................................................ii
List of figures .......................................................................................................................vii
LISTS OF ACRONYMS AND ABBREVIATIONS ........................................................ viii
CHAPTER ONE ................................................................................................................. 1
INTRODUCTION .............................................................................................................. 1
1.1 Background of the study ...................................................................................................... 1
1.2 Statement of the problems ................................................................................................... 3
1.3 Hypothesis to be tested ........................................................................................................ 6
1.4 Objective of the Study ......................................................................................................... 6
1.4.1 General objective of the study....................................................................................... 6
1.4.2 Specific objective of the study ...................................................................................... 6
1.5 Significance of the study...................................................................................................... 7
1.6 Scope and limitation of the Study ........................................................................................ 7
1.7 Structure of the Thesis ......................................................................................................... 8
CHAPTER TWO ................................................................................................................ 9
REVIEW OF RELATED LITERATURE .......................................................................... 9
2.1 Review of Theoretical Literatures ....................................................................................... 9
2.1.1 Definition of Taxation ................................................................................................... 9
2.1.2 The classification of taxes ............................................................................................. 9
2.1.3 Theories of taxation ..................................................................................................... 10
2.1.3.1 Theory of Tax Evasion ............................................................................................. 10
2.1.3.2 Theory of Technological Determinism .................................................................... 11
2.1.3.3 Institutional theory ................................................................................................... 12
2.1.3.4 The Expediency Theory of Taxation........................................................................ 13
2.1.3.5 Socio-Political Theory ............................................................................................. 13
2.1.3.6 Ability to Pay Theory ............................................................................................... 14
2.1.3.7 Benefit theory of taxation......................................................................................... 14
2.1.4 Characteristics of good taxation .................................................................................. 14
2.1.5 Functions of Taxation.................................................................................................. 16
2.1.6 Problems of tax collection ........................................................................................... 17
2.1.7 Categories of Business Income Tax Payers in Ethiopia .............................................. 17

iii
 2.1.8 Tax Declaration and Payment of Taxes ...................................................................... 18
2.1.9 Determinants of tax revenue collection ....................................................................... 19
2.2 Review of Empirical Literature ......................................................................................... 22
2.2.1 Evidence from other countries .................................................................................... 22
2.2.2 Evidence from Ethiopia ............................................................................................... 25
2.3 Summary and gap in the existing literature ....................................................................... 27
2.4 Conceptual Framework ...................................................................................................... 28
CHAPTER THREE .......................................................................................................... 29
RESEARCH METHODOLOGY ..................................................................................... 29
3.1 Research Design................................................................................................................. 29
3.2 Research Approach ............................................................................................................ 30
3.3 Study Area and Target Population ..................................................................................... 30
3.4 Types of Data ..................................................................................................................... 31
3.5 Data Collection Technique ................................................................................................ 32
3.6 Sampling Method ............................................................................................................... 32
3.7 Sample Size........................................................................................................................ 33
3.8 Variable Operation and Measurement ............................................................................... 34
3.8.1 Dependent Variable ..................................................................................................... 35
3.8.2 Independent Variables ................................................................................................. 35
3.9 Data analysis and interpretation methods .......................................................................... 39
3.10 Model Specification ......................................................................................................... 40
3.10.1 Ordered logistics model ............................................................................................ 41
3.10.2 Assumptions of Ordered logistics model .................................................................. 42
3.11 Validity and Reliability Test ............................................................................................ 43
3.11.1 Validity Test .............................................................................................................. 43
3.11.2 Reliability Test .......................................................................................................... 43
CHAPTER FOUR ............................................................................................................ 45
RESULT AND DISCUSSION ......................................................................................... 45
4.1 Descriptive Analysis .......................................................................................................... 45
4.1.1 Respondents Demographic Characteristics ................................................................. 45
4.1.2 Tax revenue collection ................................................................................................ 47
4.1.3 Tax revenue collection performance ........................................................................... 49
4.2 Econometric Analysis ........................................................................................................ 50

iv
 4.2.1 Chi-square Tests of Explanatory Variables against Tax Revenue collection ............. 50
4.2.2 Assumptions of the ordered logistic model ................................................................. 51
4.2.2.1 Multicollinearity ....................................................................................................... 51
4.2.2.2 Proportional odd assumption.................................................................................... 52
4.2.2.3 Test for Heteroskedasticity....................................................................................... 53
4.2.3 Ordered Logistics Model fitness ................................................................................. 53
4.2.4 Ordered Logistic Model Interpretations with the Coefficients ................................... 54
4.3 Results and Discussion ...................................................................................................... 57
CHAPTER FIVE .............................................................................................................. 65
SUMMARY, CONCLUSIONS AND RECOMMENDATIONS .................................... 65
5.1 Summary of major findings ............................................................................................... 65
5.2 Conclusions ........................................................................................................................ 65
5.3 Recommendations .............................................................................................................. 67
5.4 Future Research Direction ................................................................................................. 68
Reference ................................................................................................................................. 70
Appendix I ........................................................................................................................ 77
Appendix II ....................................................................................................................... 79
Appendix III ..................................................................................................................... 80

v
 List of Tables
Table 3.1 sample size for each taxpayer‟s category ............................................................... 34
Table 3.2 Summary of Variables and Measurement…………………………………………39
Table 3.3 Reliability of the instrument .................................................................................... 44
Table 4.1 Demographic characteristics respondents ................................................................ 46
Table 4.2 Respondents perception on the level of tax revenue collection…………………...47
Table 4.3 Respondents perception on the level of tax revenue collection…………………...47
Table 4.4 Descriptive Statistics for the explanatory variables of study ................................... 49
Table 4.5 Plan and performance of tax collection ................................................................... 50
Table 4.6 Chi-square Tests of Explanatory Variables against Tax Revenue collection .......... 50
Table 4.7 Correlation Matrix among the Variables ................................................................. 51
Table 4.8 Test of Parallel Lines ............................................................................................... 53
Table 4. 9Summary of parameter Estimates ............................................................................ 56
Table 4.10 Marginal effects on the probabilities of the level of tax revenue collection.......... 64

vi
 List of figures
Figure 2.1 Conceptual frameworks on determinants of tax revenue collection ...................... 28
Figure 3.1 Location map of Sululta town ................................................................................ 30

vii
 LISTS OF ACRONYMS AND ABBREVIATIONS
FDRE Federal Democratic Republic of Ethiopia

GDP Gross domestic product

ICTD International Centre for Tax and Development

IMF International monetary fund

OLM Ordered logistic model

OLS Ordinary least square

ONRSRA Oromia national regional state revenue authority

PO Proportional odd

SPSS Statistical Package for Social Science

TD Technological Determinism

TRC Tax revue collection

VAT Value added tax

UNDP United nation developmental program

WB World Bank Group

viii
 CHAPTER ONE
INTRODUCTION
1.1 Background of the study
The low level of tax collection is putting the economic development of poor countries at risk
(Akitoby, 2018). The general government imposes tax as a compulsory and unrequited
payment on citizens. A tax is not a voluntary contribution, but it is an involuntary
contribution imposed by the government (Ngugi, 2016). Tax is a financial charge levied by
governments in either directly paid by individuals or indirectly collected by intermediaries
and deposited to Government (Amin, et al., 2014).

In developing country increasing the amount of tax collected is crucial for development
programs. The tax has a broad area of benefits such as spurring infrastructure, strengthening
the social contract, and encouraging good governance (Basariya, et al., 2019). Similar to any
developing country, Ethiopia also needs to raise sources of revenue to finance at least three
functions such as defense, maintenance of law and order, and socio-economic development.
At the level of development, a government may take other steps like deficit financing, user
charging or foreign aid and debt but tax has limited cost and inefficiency than other measures
(Chaudhry & Munir , 2010). To avoid excessive foreign debt financing and strengthen
domestic revenue mobilization, developing countries are expected to broaden their tax bases
and improve their tax administration systems (Gupta, 2007).

In Ethiopia, the collections of taxes come into being with the emergence of state and
government, but there is no concrete documentary evidence as to when taxation was exactly
introduced. However as indicated in the Chamber of Commerce, (1996) the background of
tax in Ethiopia, traced back before Atse Zeryacob. The evidence is enacted regulation for the
collection of tax purpose by Atse Zeryacob (Chamber of Commerce, September,1996).

Tax collection improvement to some extent was occurring from one regime to another.
Further, the Dergue regime partially alleviated the tax assessment & collection problems that
existed during the imperial period (ibid).

After the Dergue regime, in the period of a transitional government with the establishment of
the regional governments, major changes took place in the taxation policy of the country.
They have tax and expenditure assignment powers to accomplish their duties and
responsibilities. Since that period, the administration of tax in Ethiopia has been based on

1
Federal and Regional levels of government. The 1995 constitution of the Federal Democratic
Republic of Ethiopia (FDRE), article 95 indicates that the federal and the state shall share
revenue. Furthermore, article 96-99 has separated the tax revenue shall be collected by the
federal government, state or regional government and jointly by the federal and regional
government (ibid).

Although the Ethiopian background of tax collection was tracked back to centuries, the tax
revenue of the country is still low to the ratio of GDP. In terms of the tax-to-GDP ratio,
Ethiopia‟s tax revenue performance is poor, at approximately 13%, and it is even lower than
the average for low-income countries (Yesegat & Fjeldstad, 2016). Further, as the finding of
Haque, (2009) many developing countries, like Ethiopia, had taken various reform programs
like the introduction of value-added tax (VAT) to increase the tax revenue of the countries,
though the level of taxation has still remained poor. The possible reasons for low levels of tax
revenue collection are inefficiency of the tax administration in implementing the tax rules and
regulations, and non-compliant behavior of the taxpayers (Yesegat & Fjeldstad, 2016).

Different studies were performed to identify factors that determine tax revenue collection in
other countries. The study conducted by (Mtasiwa, 2013) in Tanzania and Masarirambi
(2013) in Zimbabwe, revealed that tax evasion, social norms, tax avoidance, the complex
nature of tax rules and regulation were the determinants factors that accounted for inefficient
collection of tax revenues from taxpayers.

Some studies were also performed to identify factors that determine tax revenue collection in
Ethiopia, but still, the problems of tax revenue collection are not solved. The evidence by
Gobachew, et al., (2018) depicted that in Ethiopia the expenditure is higher than the revenue
collected. The main reason for this as indicated in the studies is a low level of tax collection.
Developing countries collect lower tax revenue comparing to developed countries, and
Ethiopia is not an exception of the developing country. In Ethiopia, there is a chronic gap
between budgeted and actual tax collection (World Bank, 2018. Although various studies
identified determinants of tax revenue, this study added new variables like tax delinquency.
Therefore, it is important to raise questions on what are the determinants of tax revenue
collection. The objective of this study is to answer that question. In Oromia's national
regional state, there are different tax administrators at each Zone and their respective towns.
Among these Sululta is the one. Thus to meet the objective of this study, it was conducted in
Sululta town.

2
1.2 Statement of the problems
As indicated, Tax revenue has vital importance for the sustainability of developed and
developing countries because of the following reasons. Firstly, taxation is the main source of
central government revenue, even in aid-dependent low-income countries. Secondly, tax aims
to meet social and public needs through providing public goods and services. Thirdly, it is
needed for the government to establish armed forces and judicial systems to ensure the secure
environment and justice of the society. However, the development program of many
developing countries bottlenecked by the low collection of tax revenue (Gobachew, et al.,
2018).

As indicated by Kibret & Mamuye (2016), in UNDP (2016), domestic resource mobilization
is critically important for sustainable financing of development plans. Nevertheless, domestic
resource mobilization in developing countries like Ethiopia is challenged by several factors
such as illicit financial outflows (tax evasion, tax avoidance and corruption), slow structural
transformation in the economy, substantial tax incentives like tax holidays and other
exemptions. In most developing countries, including Ethiopia, domestic resource
mobilization is low, and as a result, external debt is the main source to finance the resource
gap (Getinet & Ersumo, 2020).

For the reason that tax revenue is the main portion of the annual national budget, many
developed countries secure their national budget through collecting efficient tax revenues.
The World Bank (2018) stated that developed countries collecting more tax comparing with
developing countries than their GDP rate. Developing countries are unable to collect
sufficient tax revenue as a result they face an annual national budget deficit. As indicated in
IMF (2020) Ethiopia annual deficit is about 2.5 % which caused by the decline in tax revenue
collection as a result of a shift in GDP composition since 2014/15 to sectors that are
traditionally more prone to weak tax compliance (such as construction) and some temporary
impact of political unrest hindering revenue collection. The report further indicated tax
revenue collection was declined to 10% of GDP. Additionally, World Bank (2018) indicated
that to ensure basic service and a strong economy can be provided, the Ethiopian government
needs to collect at least 15% of GDP in taxes.

According to Akitoby (2018) “a typical developing country collects just 15% of GDP in
taxes, compared to the 40 % collected by a typical advanced country” (p.1). This shows the
collection of tax revenue by developing countries, including Ethiopia, is relatively low

3
comparing to developed countries. The Ethiopian domestic revenues collected from tax
sources have been significantly increasing in the past 10 years, compared to the preceding
decades. Even though tax revenue collection was growing at a reasonably high rate, when
analyzed with the growth in GDP, the growth has not been significant (Kibret & Mamuye,
2016).

To identify factors affecting tax revenue collection many studies were conducted. The study
conducted by Gituma (2017) in Kenya, stated that government policy, rules and regulations,
corruption, employee qualification, skills and training, technology and information systems
affected the optimal revenue collection. Amin, et al. (2014) revealed tax collection shrinks
due to increasing corruption and political instability. The study conducted by Yusnika &
Widyaningsih (2018) found tax revenue collections are affected by tax rates and good
governance.

There are also some Studies conducted in Ethiopia on the determinants of tax revenue
collection such study by Abate (2019) found the determinants of tax revenue collection are
lack of equity and fairness in presumptive tax assessment, the complexity of tax rules,
taxpayers poor perceptions of tax evasion and corrupt tax officials. Manaye, et al., (2019)
found that the effect of tax evasion and weak enforcement mechanism are reasons for the low
level of tax collection. As per Kibret & Mamuye (2016) suggested in UNDP (2016) the
reason for the low tax revenue of Ethiopia is the existence of untaxed income due to the poor
structure of the economy and administrative inefficiencies.

Particularly, as indicated by Oromia national regional state revenue authority (ONRSRA,

2018), tax revenue collection in Sululta town is low. Although several business activities in
the town are the source of business income tax, the collected tax is low. Specifically, the tax
authority stated that the tax revenue collection performance in Sululta town is below 80
percent. There may no study conducted in this town to reveal why the tax revenue collection
was low. Therefore, this study aimed to fill this gap.

Related studies were conducted to investigate the effects of macro-economic factors such as
inflation, foreign direct investment, foreign aid, urbanization and trade openness and
economic sectors which include agricultural, service and industrial sectors on tax revenue
Gobachew, et al., (2018); Workineh (2016); Bayu (2015); Kumari (2017) and Terefe &
Teera (2018). However, globally there are a few studies conducted on the effect of
organizational or institutional, administrative and demographic factors such as perceptions of

4
tax evasion and avoidance, corrupt tax officials and strength of the tax authority on tax
revenue collection. The studies conducted on the determinants of tax revenue collection in
Ethiopia cover some variables such as taxpayer‟s awareness, the capacity of the tax authority,
tax evasion, tax compliance, social norms, participatory tax system, equity and fairness of the
tax system (Abate (2019); Manaye, et al. (2019).

However, the big gap is those studies were geographically limited to specific towns and
cities, so generalizing to the country level as a whole leads to a wrong decision. Thus,
conducting a study at the country level or at each zone helps to know the overall determinants
of tax revenue collection in Ethiopia. Thus to fill this gap this study was conducted in Sululta
town, and this study may be the first to be conducted on the determinants of tax revenue
collection in Sululta town.

The earlier studies such as the study conducted by Abate (2019) applied binary logistics by
classifying tax collection level as good or poor and Eja, et al. (2018) applied ordinary least
square which is other than the ordered logistic model; but ordered logistic model can
appropriately show the effect of each determinant on each level of tax revenue collection by
alluring more categories. Therefore, this study aimed to fill this gap by applying the ordered
logistics model.

Moreover, the study was motivated by the existence of an unequal income distribution of the
country, which might happen because of the economic the system country follows (mixed
and transition economy with the large public sector). This system created way only for a few
people to be advantageous leavening the country poor regarding infrastructure and stability
and based on the theory of Adam Smith this problem can be solved by collecting proper tax.
Therefore, identifying the determinants of tax collection is the basic issue.

The objective of this paper is to investigate determinants of tax revenue collection in Sululta
town, Oromia national regional state, and it added new variables to bridge the existing gap.
The outcome variable is tax revenue collection whereas the independent variables assessed
include tax evasion, corruption, employee competence, technology, organizational strength of
tax authority, political instability and tax delinquency. Further, demographic variables
(gender, work experience and educational level of taxpayers) were used as control variables.
The effect of employee competence on tax revenue collection is may new variable in Ethiopia
studies. Although study on the relationship between tax compliance and delinquency tax was

5
performed, tax delinquency may consider as a new variable and gap in the literature in this
area of study. Therefore, this study also tried to fill this gap.

1.3 Hypothesis to be tested

H1: Tax evasion has a negative and significant effect on tax revenue collection.

H1: Corruption has a negative and significant effect on tax revenue collection.

H1: Employee competence has a positive and significant effect on tax revenue collection.

H1: Technology has a positive and significant effect on tax revenue collection.

H1: Organizational strength of tax authority has a positive and significant effect on tax
revenue collection.

H1: Political instability has a negative and significant effect on tax revenue collection.

H1: Tax delinquency has a negative and significant effect on tax revenue collection.

1.4 Objective of the Study

The problems and research gaps that has been identified above need to be filled objectively;
this study addressed the identified gap by attaining the following general and specific
objectives.

1.4.1 General objective of the study

The general objective of this study was to examine the determinants of the tax revenue
collection system in Sululta town, Oromia national regional state, Ethiopia.

1.4.2 Specific objective of the study

 To examine the effect of tax evasion on tax revenue collection in Sululta town Revenue
Office;
 To evaluate the effect of corruption on tax revenue collection in Sululta town Revenue
Office;
 To investigate the effect of employee competence on tax revenue collection in Sululta
town Revenue Office;
 To examine the effect of technology on tax revenue collection in Sululta town Revenue
Office;
 To evaluate the effect of organizational strength of tax authority on tax revenue
collection in Sululta town Revenue Office;

6
  To investigate the effect of political instability on tax revenue collection in Sululta town
Revenue Office;
 To evaluate the effect of tax delinquency on tax revenue collection in Sululta town
Revenue Office;

1.5 Significance of the study

This paper may provide the following practical significance. Firstly, the findings of this paper
may help tax authorities and policymakers to know the determinants of tax revenue collection
and their effect, and thereby to design and implement appropriate tax collection systems.
Knowing the cause for low revenue collection is indispensable to take a reasonable response.
If tax is not collected properly, the government isn‟t able to provide public goods such as
infrastructure, health and education. The country may use debt financing which has an
interest payment, but domestic resource mobilization is critically important for sustainable
financing of development plans. Therefore by knowing where the problem is policymakers
can design an appropriate tax collection system.

Secondly, this paper has a contribution to the existing tax literature on determinants of tax
collection issues in Ethiopia and many other developing countries. The pieces of literature in
this area are scant, especially in Ethiopia. In this study new variable (tax delinquency) is
added and some of the existing variables are also used to prove their effect on tax revenue
collection. This can add something new to the existing literature.

Finally, it serves future researchers and academicians who may wish to study on tax revenue
collection factors related issues.

1.6 Scope and limitation of the Study

The study is focused on the determinants of tax revenue collection in the Sululta town tax
revenue office. Geographically the study was limited to Sululta town to examine the
determinants of revenue collection. For this study, the respondents are taxpayers and
employees of the Sululta town tax revenue office. The cross-sectional data analysis was
applied with Stata version 16 and Statistical Package for Social Science (SPSS) version 20.
The ordered logistic regression model was used to analyze the interrelationship between the
dependent variable which is tax revenue collection and the independent variable which are
limited to tax evasion, corruptions, employee competence, technology, organizational
strength of tax authority, political instability, tax delinquency and demographic variables
(gender, work experience and educational level of tax payers).

7
The study was conducted on the determinant of tax revenue collection in Sululta town which
may pose inappropriateness to generalize to the country as a whole. Further, the study
covered only tax revenue, not non-tax revenue. This may considered as limitation of this
study.

1.7 Structure of the Thesis

For limpidity, the thesis is organized logically under five interdependent chapters. The first
chapter contains the background of the study, statement of the problem, both general and
specific objective of the study, the significance of the study and confronted limitations. In the
second chapter relevant theoretical and empirical literature are included. In chapter three
researches methodology including research design, types of data and data collection
technique, sampling methods, model specification, data analysis and interpretation methods
are present. In chapter four results and discussion are included. Chapter five is the final
chapter and it presented the conclusion and recommendation of the study.

8
 CHAPTER TWO
REVIEW OF RELATED LITERATURE
In this chapter, related literature is stated. The broad areas covered by this section are
theoretical literature, empirical literature, the summary of the literature and the gap and
conceptual framework. Although there are various theories on taxation the most relevant for
this study are identified. The empirical study covers both external (from another country) and
internal (from Ethiopia) is presented to be proved or disprove the study. The vacuum in the
literature is also uncovered. The conceptual framework is developed based on a theoretical
and empirical study for visual view.

2.1 Review of Theoretical Literatures

Under this section, the review of literature on the theories is performed. A theory is a set of
concepts and also from which some of the concepts in the conceptual framework are also
used. Five theories of taxation with additional two theories are stated and they were served as
a benchmark for data analysis. Additionally, issues related to taxation are presented.

2.1.1 Definition of Taxation

The first duty of the sovereign is that of protecting the society from the violence and invasion
of other independent society can be performed only by means of military force (Smith, 1776,
p. 564). Those activities can be funded from the source of tax. Adams smith defines tax as
tax is a contribution from citizens for the support of the state. (Smith, 1776, p. 565)

A tax is a payment required (it is not voluntary) by a government (federal, state, or local) that
is unrelated to any specific benefit or service received (Not tied directly to the benefit
received by the taxpayer) from the government (Spilke, et al., 2019, p. 4).

The tax definition applicable in this study is the tax is a compulsory payment to the
government by the taxpayer without any expectation of some specified return.

2.1.2 The classification of taxes

Scholars classified tax in various methods. Many scholars have advanced various theories for
the classification of taxes into two main divisions direct and indirect tax. These terms can be
defined differently which sometimes leads to confusion. However most accepted logic behind
the direct tax is that tax is directly paid by the person on whom it is legally imposed. An
indirect tax burden is ultimately borne by final consumers (Tesfaye, 2011).

9
When both impact and incidence fall on the same person it is said to be a direct tax. A direct
tax cannot pass on to others, but it is borne by the person on whom it is imposed. The impact
of a tax is on the person who pays the tax in the first instance and the incidence of a tax
means the final or ultimate resting place of the burden of the tax payment. If the impact and
incidence fall on different persons it is an indirect tax. It is not borne by the person on whom
it is levied and can be passed on to others. The traditional theory of tax shifting stated a tax
may be imposed on some person and he may transfer it to another person. Ultimately, it may
be borne by the second person or the second person may again transfer it to others by whom
it is ultimately paid. That is, the person who bears the burden of tax in the last instance may
not necessarily be the person who originally pays the tax (Parameswaran, 2005).

According to FDRE income tax proclamation 979/2016, direct taxes are employment income
tax under schedule “A”, rental income tax under schedule “B”, business income tax under
schedule “C”, and other income under schedule “D”. Indirect taxes are taxes which borne by
someone other than the person responsible for paying them. The indirect taxes in income tax
proclamation 979/2016 are value-added tax VAT, Turn over tax (TOT), excise tax, stamp
duty, and import taxes.

2.1.3 Theories of taxation

This study utilized the theory of tax evasion, theory of technological determinism,
institutional theory, expediency theory of taxation, and socio-political theories as well as
common theories of taxation (ability to pay and benefit theory) were also used.

2.1.3.1 Theory of Tax Evasion

The formal theory of tax evasion was first developed by Michael Allingham and Agnar
Sandmo in 1972. The model answers two questions on tax evasion activities by an employer:
how much of his income should he report and how much should he evade? Thus according to
this theory, people have two options regarding tax payment. Firstly, they will evade if the
benefit of evading tax is higher than the probability of being caught and related cost or
penalty for evading tax. Secondly, they may be normal in declaring their actual income
(Sandmo, 2005).

Theory of Tax Evasion described tax evasion to have prevailed where tax evasion is not
detected by the tax authority, as well as when audit or punishments is uncertain. That is
people pay taxes primarily because of the fear of audit and punishment. Thus in turn implies
government can increase tax collection through raising audit activities and penalty rates.

10
Theory of Tax Evasion assumes that the probability of detection increases with the amount
that one evades. It is clearly stated in the theory of tax evasion article as following:

“In the A-S model with an endogenous probability of detection, it is assumed that the
probability of detection increases with the amount that one evades. But how does the
taxpayer form his views about this probability? One source of information about the
probability is the taxpayer‟s own observations of other taxpayers; he may have observed
neighbours or colleagues who work in the black economy and get away with it. This means
that we could write the taxpayer‟s subjective probability of detection as a function both of
his own evasion and his perceived amounts of evasion by others. Suppose now that his
perception of the amount evaded by others increases. His subjective probability of
detection falls, and he, therefore, decides to evade more. The others now perceive that he
evades more and therefore they may evade more themselves,” (Sandmo, 2005, p. 21).

Additionally, the model stated firms could evade indirect tax as they act as an agent on behalf
of the government. They serve as collectors of indirect taxes from consumers, and therefore,
the main stage of indirect tax evasion is at the final stage where transactions are made with
consumers (Sandmo, 2005). The theory of tax evasion is relevant in this study in that it
recognizes the importance to see how tax evasion affects revenue collection. According to
this theory, it is hypothesized that the existence of tax evasion has an impact on decreasing
the level of revenue collection.

2.1.3.2 Theory of Technological Determinism

Technological determinism (TD) is the theory that states a society's cultural values, social
structure, and history determined by technology, and it changes any mode of operation. It was
first proposed by Thorstein Veblen (1857–1929) American humanist and financial analyst,
but advancement to the theory was made over time for instance Karl Marx. It is a reductionist
theory which states technology is a social force that drives change. Hauer (2017) argued
social progress follows an inevitable course that is driven by technological innovation.
Technological determinism has two central concepts: firstly, technological development itself
follows a predictable, traceable path that is beyond any cultural or political influence; and
secondly, the technology, in turn, organizes society in a way to further develop itself
(Mardiana & Daniels, 2019).

As argued by Kelvin (2017) Theory of technological determinism depends on two

hypotheses; 1) belief that the technical base of a society is the fundamental condition

11
affecting all patterns of social existence and 2) belief that technological change is the single
most important source of change in society. Technological determinism is a moderate
approach emphasis on the autonomy of technological change and the technological shaping
of society (Acosta & Holbrook, 2008).

Technological determinism is the idea that technology has important effects on people‟s
lives. Technological changes have a major impact on the way people run their businesses.
The use of computers and the internet as one of the technologies has promoted the ability to
multitask in many responsibilities (Wood 2004). The theory of technological determinism is
relevant in this study in that it recognizes the importance of using technology to organizations
and society. According to this theory, it is hypothesized that the use of the latest technology
has an impact on increasing the level of revenue collection.

2.1.3.3 Institutional theory

The institutional theory asserts that the institutional environment can strongly influence the
development of formal structure in an organization. Early analyses of organizations by
sociologists used the term “institutional” to refer broadly to aspects of organizations involved
in mediating relations with external constituencies, and more specifically to securing
favorable perceptions of an organization by constituents as a way, ultimately, of ensuring
flows of necessary resources. This is exemplified in an early statement by a key proponent of
the structural-functionalist theory, Talcott Parsons (1956), who elaborated a “cultural-
institutional” view of organizations in the first issue of the Administrative Science Quarterly.

Parsons (1956) argued that, as components of a larger society, organizations need to

demonstrate “basic acceptance of the more generalized values of the superordinate system”
(p. 67). In this context, he distinguished three broad organizational levels, each with
distinctive functions: technical (production activities), managerial (coordination and control),
and institutional (managing external relations) (Parsons, 1960). In Parsons‟ view, the
institutional level was critical to articulating the connection between an organization‟s
espoused goals and the functioning of the larger society.

All societies have both formal institutions, which are the laws and regulations that define the
legal rules of the game, as well as informal institutions, which are the “socially shared rules,
usually unwritten, that are created, communicated and enforced outside of officially
sanctioned channels” (Helmke & Levitsky, 2004, p. 727). When these are fully aligned, tax
better tax collection will occur, because the values, norms and beliefs of taxpayers will be

12
fully aligned with the formal rules of the game. However, when the formal and informal
institutions/rules of the game are not aligned, which can be measured by the level of tax
morale, then engagement in tax non-compliance will become more prevalent (Williams,
2018)

Theory of Institutional is relevant in this study in that it recognizes the organizational

capacity in response to external (taxpayers). According to this theory, it is hypothesized that
institutional failure has an impact on decreasing the level of revenue collection

2.1.3.4 The Expediency Theory of Taxation

The expediency theory of taxation states that every tax revenue collection proposal must pass
the test of practicability, which must be the only consideration when the country government
is choosing a revenue collection proposal. The proposition is that the economic and social
objectives of the government should be treated as irrelevant since it is useless to have a tax
which cannot be levied and collected efficiently. However, there are pressures from
economic, social and political groups. Every group tries to protect and promote its own
interests and the county government is often forced to reshape tax structure to accommodate
these pressures (Bhartia, 2009).

2.1.3.5 Socio-Political Theory

The theory was first proposed by Wagner in 1883 in which theory main point is tax should
benefit social not a single individual. The idea of Wagner is not an individualist approach to a
problem, but he wanted that each economic problem should be looked at in its social and
political context and an appropriate solution found thereof. It obvious that society is a
collection of individuals, but it is more than the total sum of individuals. It had an existence
and entity of its own which needed preservation and taking care of society, but should be
used to cure the ills of society as a whole. Wagner, in other words, was advocating a modern
welfare approach in evolving and adopting a tax policy. Wagner was specifically in favor of
taxation as an instrument for reducing income inequalities (Chigbu, et al., 2012).

As this theory the central idea is not with the concern of individual benefits, it is a relevant
sense that activities like corruption and tax evasion offer benefits for individuals who commit
such activities. Thus this concept was taken as tax is to benefit sum of total individual
members than a single individual, and any activities that benefit individual reduce the level of
tax collection.

13
2.1.3.6 Ability to Pay Theory
The theory was developed by Pigou in 1920 and it is the most commonly developed principle
of equity. According to this theory tax liability, is considers as true form-compulsory
payment to the state without quid pro quo. A citizen is to pay taxes because they are able and
his or her relative share in the total tax burden is to be determined by his relative paying
capacity. The basic principle of this theory is that the burden of taxation should be shared by
the members of society on the principles of justice and equity and that these principles
necessitate that the tax burden is apportioned. The assumption is there is no commercial or
semi-commercial relationship between the citizens and government (Chigbu, et al., 2012).

2.1.3.7 Benefit theory of taxation

The theory was developed by Erik Lindahl in 1960. The central idea of this theory lay on
taxes is to be imposed on individuals according to the benefit conferred on them. In effect,
the more benefits a person derives from the activities of the State, the more he should pay to
the government. It is base on the assumption of exchange between government and citizens,
which are citizens, should pay tax because of goods and services they receive from the
government. However, in this quid pro quo setup, the issues of equitable distribution of
income and wealth are distorted. Rather, the benefits received are taken to represent the basis
for distributing the tax burden in a specific manner (Chigbu, et al., 2012).

Even though in this study no hypothesis was developed from the theory of ability to pay and
benefit theory, it is relevant in the sense that tax should be collected from anyone who has the
ability to pay and benefits from service provided by the government.

2.1.4 Characteristics of good taxation

Taxation is one of the major sources of Government revenue. At the same time, it has got a
special role to play in the redistribution of income and reduction in inequalities of income
prevailing in the society. To achieve both these objectives, the taxation system should be such
that it is acceptable to the general public, fair to all members of society, simple to operate and
capable of yielding the optimum revenue. Adam Smith, the famous eighteenth Century
British Economist, was the first to set four maxims in his book V named as a wealth of
nations published in 1776. He points out that the private revenue of individuals, arises,
ultimately from three different sources; rent, profit, and wages. Every tax must finally be paid
from someone or other of those three different sources of revenue, or from all of them
indifferently (Smith, 1776, p. 676).

14
The economist laid down about four basic canons of taxation on which a good tax system
should be based. The four canons laid down by Adam smith in his book V are:

1. Canon of Equity

The first one is the canon of equity which is also called as the canon of ability; all citizens of
the nation should contribute towards expenses of the Government “as nearly as possible in
proportion to their respective abilities.” In simple terms tax system should be such that the
citizens are liable to pay taxes as per individual capacity to pay. The riches are expected to
pay more while the poor are expected to pay less. The ability-to-pay taxes increase with the
rise in income, thus making the higher economic class to contribute more to the Government
revenue. This principle is based on the simple theory that the ability to pay taxes increases
more than proportionately to the increase in income because the utility of money gradually
diminishes with every increase in income. The basic principle underlying is that “equals
should be treated equally”. This principle makes a tax system “fair and just”. The equity level
determines now the tax burden is distributed among the taxpayers.

However, Modern economists did not agree to this opinion. Instead, they strongly put forth
„progressive‟ tax system which means increasing rates of taxes with the increase in income.
Considerations of fundamental fairness provide the main rationale for taxing income taxes,
which embrace both horizontal equity that is the need for equal treatment of persons with
comparable abilities to pay; and vertical equity: the need for appropriate differences in the tax
treatment of persons with different taxable capacities (Smith, 1776, p. 676).

2. Canon of Certainty

The second is the canon of certainty which is the tax which each individual is bound to pay,
ought to be certain and not arbitrary. This means the time of payment, the manner of
payment, the quantity to be paid, the place of payment, and the authority to whom the tax is
to be paid ought all to be clear and plain to the contributor, and to every other person. That is
to mean to protect the taxpayers from unnecessary harassment by the tax officials. The
uncertainty of taxation encourages the insolence and favors the corruption, of an order of men
who are naturally unpopular, even where they are neither insolent nor corrupt (Smith, 1776,
p. 676)

3. Canon Convenience

The third is the canon convenience which means every tax ought to be levied at the time, or
in the manner, in which it is most likely to be convenient for the contributor to pay it. The tax
15
should be levied in such a manner that its payment should cause the least hardship or
inconvenience to the taxpayer (Smith, 1776, p. 677)

4. Canon of Economy

Finally, the canon of the economy is every tax involves some revenue yield and a
corresponding cost of collection. Adam Smith noted, "Every tax to be so contrived as both to
take out and to keep out of the pockets of the people as little as possible over and above what
it brings into the public treasury of the government." The tax system should be economical to
operate and the tax should be such that the cost of its collection should be minimal. The
revenue from tax should be much more than the cost of its collection. Otherwise, the major
portion of the revenue will be taken away by the cost of its collection. This is particularly
important from the point of view that the costs spent on the collection and nothing to the
national output and resources, which are already in scarcity, should not be wasted. A heavy
tax burden also does not satisfy the canon of the economy. Thus, it is interpreted that if a tax
is quite heavy reducing savings of the public, reducing investments and, thus, the productive
capacity, the tax system seems to have failed to satisfy this canon (Smith, 1776, p. 677)

In addition to the above basic fundamentals of the tax system, given by Adam Smith, some
other writers like Bastable, Shirras, Mrs. Hicks have added a few more canons of taxation,
which are: Canon of Simplicity, Canon of Elasticity, Canon of Productivity, Canon of Co-
ordination, Canon of Variety and Canon of Expediency (Kulkami, 2001).

2.1.5 Functions of Taxation

Tax revenue is treated as one of the most important may be of raising the funds for
government activities. Apart from the raising of funds, the tax policies are being framed to
achieve several economic and social objectives. The primary function of most taxes is to
divert economic potential from the private sector to the government for its use (Kulkami,
2001). The objectives of taxation in developing countries are virtually the same.

The contemporary theory of taxation shows taxes has three main functions such as a fiscal
function, regulatory functions and stimulating function (Eckhardt, 2016).

Fiscal Function: the first and important function of tax is fiscal function. It is a function of
raising tax revenue to finance the government budget and thereby to enables the country in
maintaining their institutions and performs various duties. This function is derived from the
definition of taxes in which taxes are levied primarily to raise revenue for government
expenditures. This revenue finance government expenditure which is either the provision of

16
goods and services which members of the public cannot provide such as defense law and
order to the provision of goods and services which the federal and state governments feel are
better provided by itself such as health services and education (Eckhardt, 2016).

Regulatory functions: the second purpose or function of taxation is to act as a regulatory

function. This function of taxes is aimed at molding the income and capital of taxpayers. This
means tax can act as a redistribution of income. Taxes regulate the price through taxing the
most spending area by the public and private sectors which in turn reduce inflationary
(Eckhardt, 2016).

Stimulating function: the third function of taxation is stimulating. Not all activities performed
by citizens are acceptable and it also harms the state. Therefore, the government needs to
discourage such activities through taxes. In contrast, there are activities that need
encouragement by the government. Thus, if tax is to act as encouraging or discouraging
certain activities, it is said to be a stimulating function of tax (Eckhardt, 2016).

2.1.6 Problems of tax collection

According to Chinyere (2002) cited Lawal (1982) the following stated problems mainly
affect tax collection.

Inadequate staff or manpower to carry out the assignment efficiently and thus has contributed
to the low revenue generated for the state. The other problem is mismanagement of tax
collected. The collected taxes were not been utilized for the purpose for which it was
collected thus makes taxpayers not give out their wealth for the state. Furthermore, bribery
and corruption are the main problems. In this day, tax collector personal interest has
overridden their official interest in the performance of their duties consequently affects
revenue generation for the state (Chinyere, 2002).

Additionally, lack of voluntary compliance from taxpayers these attitudes of taxpayer cause
tax avoidance evasion and delinquency. Poor accounting records are the other problem
related to tax collection. Most business traders professional do not keep proper records of
their income and expenditure. An inadequate facility for employees is also another problem
(Chinyere, 2002).

2.1.7 Categories of Business Income Tax Payers in Ethiopia

According to income tax proclamation No. 979/2016, for the purpose of assessment and
collection business tax, taxpayers are categorized as category “A”, “B” and “C”. as per this

17
proclamation Category “A” taxpayers shall include any company incorporated under the laws
of Ethiopia or a foreign, and any other business having a gross income of birr one 1,000,000
(one million Br) or more as per aforementioned regulation article 3(1/a). category “A”
taxpayers liable for business income tax shall keep books of account prepared in accordance
with the financial accounting reporting standards and in particular shall keep a record of the
business assets and liabilities of the taxpayers, including acquisition date, the cost of
acquisition, any costs of improvement in relation to assets, and the current net book value of
the assets. Additionally, they shall keep the record of all daily income and expenditures
related to taxpayers' business and record of all purchases and sales of trading stock, and
service provided and received and received by the taxpayer (income tax proclamation,
979/2016).

Category “B” taxpayers include being a person other than a body, having gross income of
500,000 or more but less than 1,000, 0000 (five hundred thousand or more than but less than
one million birrs) as per aforementioned regulation article 3(1/b). Category “C” taxpayers
liable for business income tax shall keep the record of daily income and expenditure, all
purchases and sales of trading stock, salary and wages register and any other document
relevant in determining the tax liability of taxpayers. Category “C” taxpayers include being a
person other than a body, having gross income of less than 500,000 (five hundred birr) as per
aforementioned regulation article 3(1/c). Category “C” may keep a record of gross income
and shall keep such other records as may be specified in the regulation, but not required
(Income tax proclamation , No.979/2016).

2.1.8 Tax Declaration and Payment of Taxes

As per tax proclamation taxpayers under schedule C income have to follow the procedure
stated in the tax for declaring their taxable income. The period of declaration of file for each
category is differing. Category “A” taxpayers are required to declare their taxable income
within four months from the end of the tax year. The tax declaration filed by category “A”
taxpayer shall be accompanied by the taxpayer‟s profit or loss statement and balance sheet for
the year. Category “B” taxpayers are required to declare their taxable income within two
months from the end of the year. The tax declaration filed by category “B” taxpayer‟s shall
be accompanied by the taxpayer‟s profit or loss statement for the year. Category” C”
taxpayers shall pay tax income together with the annual turnover and the amount derived
from the sources other than the main operation within 30 days from the end of a fiscal period,
6th of August every year (Income tax proclamation , No.979/2016).

18
2.1.9 Determinants of tax revenue collection
In theory, different factors affect tax revenue collection. Accordingly, tax evasion and
avoidance are among the first. The theory of tax evasion is indicated in the above section in
detail. Tax evasion is considered as a violation of the law. Taxpayers engage in illegal
activity through refrains from reporting income from labour or capital which is in principle
taxable. On the other hand, tax avoidance is reducing tax payment within the legal framework
of the tax law. Tax can be avoided through exploiting loopholes in the tax law in order to
reduce one‟s tax liability; converting labour income into capital income that is taxed at a
lower rate provides one class of examples of tax avoidance. When the taxpayer is involved in
tax avoidance, he/ she has no reason to worry about possible detection; it is often imperative
that taxpayers make a detailed statement about his transactions so that he/she can get the tax
reduction that he/she desired. Tax evasion and tax avoidance reduce the amount of tax to be
paid by the taxpayer (Sandmo, 2005).

Corruption provides benefit for an individual who commits which against the socio-political
theory. According to Transparency International (2004), corruption is the misuse of entrusted
public power for private gain. One of the greatest challenges of the contemporary world is
corruption. Corruption minimizes good government, basically distorts public policy, harms
the private sector, leads to the misallocation of resources and private sector development and
particularly hurts the poor.

Concepts of equity and fairness are the main issue of tax policy. Horizontal, vertical, and
individual equity are the minimum three distinct dimensions in this concept. Vertical equity is
most commonly manifested in progressive tax and expenditure structures. Horizontal equity
is a people with the same income or wealth brackets should pay the same amount of taxes.
The demand of individual equity is the one in which individuals freely engage in transactions
of their own choosing. Moreover, citizens are keenly sensitive to arguments about fairness in
almost every policy debate. The most obvious requirement of equity or fairness is to treat
equal people in equal circumstances in an equal way (Eugene , 1975).

One of the organizational or institutional factors that affect tax collection is the skill or
competence of the employee. Employee competence enables them to perform their job well
and achieve organizational excellence, which could be the organizational duty to improve
their employee‟s performance. Employee competence has no single definition. To be specific,
the employee competencies employees need to develop in the job of tax, accounting and

19
auditing include general business knowledge, knowledge of accounting, skills in problem-
solving, interpersonal skills, communication skills and personal attitude (Udin, et al., 2012, p.
430).

The vrooms expectancy theory developed by Victor Vroom in 1964, employee‟s motivation
is an outcome of how much an individual wants a reward, the valuation that the likelihood
that the effort will lead to projected performance and the belief that the performance will lead
to remuneration. The model further indicated that employee performance is based on
individual issues such as persona, skills, knowledge, practice and abilities. Expectancy is the
confidence that better efforts will effect in better performance. It is predisposed by factors
such as ownership of suitable skills for the execution of jobs, accessibility of right resources,
accessibility of vital information and getting the required backing for carrying out the job
(Adams, 2012).

Technology or mode of tax collection is another issue in tax collection. As indicated by

United Nations (2007) technology in taxation is the use of e-taxation which is a process
where tax documents or tax returns are submitted through the internet, usually without the
need to submit any paper return and it encompasses the use of internet technology. The
technology acceptance model developed by Davis in 1989, stated that when users are offered
a new technology, a number of aspects influence their choice about how and when they will
use it, remarkably, seeming used and apparent usefulness. The use of technology would users'
work performance (Davis, 1989).

Digitalization may improve tax enforcement technology through collecting more and more
reliable information on the economic outcomes of taxpayers. Further, it improves the equity-
efficiency trade-off by implementing more complex tax systems to better target income
redistribution. In doing so, digitalization potentially enables the government to collect the
same amount of revenue or to redistribute the same amount of income as in current tax
systems by lowering tax rates. (Jacob, 2017, p. 25).

Tax compliance and the educational level of taxpayers can also affect the amount of revenue
collected. Tax compliance is the taxpayer‟s decision to comply with tax regulations and laws
through paying accurate tax on time (Younde, 2019). Mainly two factors affect the
improvement of tax collection. These are lack of tax compliance and difficulty in finding
effective instruments for improving compliance given institutional constraints (Ahmed, et al.,
2012).

20
ICTD showed it is believed that citizens with better tax education are able to act and reacting
on complex tax systems, and can avoid overpayment due to intricate reporting requirements.
Citizens without tax knowledge may create uncertainty and untrust between government and
taxpayers, this reduces tax collection. Tax education can be given to citizens in different form
such as in school, national taxpayer days and tax edutainment (radio, TV and social networks
are the main channels for bringing tax education into people‟s homes) this can improve their
tax compliance and payment which in turn increase tax collection (Mascagni & Santoro,
2018).

Adequate tax education enhances tax collection. Tax education increases the knowledge level
of taxpayers in terms of technical skills needed to determine their tax liability and awareness
of tax regulations. Several ways can upgrade tax revenue mobilization such as the knowledge
level and through tax education of individual taxpayers. Further, individual tax payer‟s
technical skills in tax accountability, awareness of available tax regulations, and Sanctions
(penalties fines prosecutions) are factors for better tax revenue mobilization. Tax education
improves tax collection by reducing the level of non-compliance, increasing tax
accountability and improving the knowledge level on tax regulations (Eugene , et al., 2020).

The organizational strength of tax authority was already known as a core factor for effective
tax compliance and collection. OCED (2001) indicated that tax revenue authority has a role
to ensure compliance with tax laws. However, the effectiveness of tax authority is based on
some of the external factors such as the willingness of taxpayers to comply with tax rules,
public support for the priorities of the government and the state of the economy. “Good
revenue authorities are strategically focussed and responsive to changes in their environment
and that of their taxpayers” (p.3).

Using the theory of agency, tax revenue is produced through combining the tax base, tax rate,
and the tax administrator‟s collection effort. The politicians first determine tax rates and
bases by passing tax legislation, but tax revenue is not automatically generated. The process
requires a tax collection effort. To realize tax revenue, tax administrators implement tax
legislation by inputting a tax collection effort. Without the collection effort, the government
would have tax receivables with an unknown amount of tax revenue. Thus, the level of tax
revenue depends on the level of the tax collector‟s collection effort, holding other factors
constant (Jensen & Meckling, 1976).

21
Political instability has a probability to shorten policymakers‟ horizons leading to suboptimal
short-term macroeconomic policies. Additionally, frequent switch of policies, creating
volatility is a may result from political instability and thus, negatively affecting
macroeconomic performance. Considering its damaging repercussions on economic
performance the extent at which political instability is pervasive across countries and time is
quite surprising (Aisen & Veiga, 2011).

Tax delinquencies are considered as the debts owed to the tax government by taxpayers. Even
though they have been considered with other aspects of tax collection, such as tax evasion
and tax avoidance, tax delinquencies independently have an important impact in the tax
collection process. Financial penalties and income garnishment are some of the most
traditional tools used to collect tax delinquencies. Additionally, tax delinquencies are
collected through shaming penalties, by which the identities of tax debtors are publicly
revealed (Truglia & Troiano, 2015). Likewise in Ethiopia, the mechanism applied to
delinquent tax is tax lien in which it charges the asset of delinquent taxpayers until the tax
already due is paid (Yilma, 2013).

2.2 Review of Empirical Literature

Tax is a main source of revenue for the government. Different empirical studies were
undertaken to identify factors affecting tax collection. Thus, in this section studies performed
on the area of determinists of tax collection were reviewed.

2.2.1 Evidence from other countries

Various studies have been conducted on the determinants of tax revenue collection from
which the most related study reviewed as follows.

Amin, et al. (2014) explore the “Factors affecting tax collection in Pakistan” using time series
data from 1980-2010, with Autoregressive Distributed Lagged (ARDL) model was applied to
find out the long-run relationship between variables. Since the Stationary of data in time
series analysis is very important the authors performed the stationary test using to check the
stationary of all variables using Augmented Dickey Fuller (ADF). In the study, only a
secondary source of data was used.

The authors applied macro-economic variables as determinants of tax revenue collection.

They found that inflation has a positive but insignificant relationship with tax collection.
Further, the result suggested that real per capita income, corruption and political instability

22
have a negative and significant effect on tax collection. They also found trade openness has a
positive and significant relationship with tax revenue.

Ngicuru (2017) conduct a study on “An empirical review of factors affecting Revenue
collection in Nairobi country, Kenya”. The author used some specific objectives of the study
which were to establish the effect of revenue diversification on revenue collection, establish
the effects of administration on revenue collection, identified how tax structure affects
revenue collection and indicated how various forms of revenues affect revenue collection.
The author used a descriptive research design. The study was done using primary data
collected from tax employees using questionnaires.

The finding shows the diversification of revenue sources can increase tax revenue. Moreover,
tax revenue collection can be affected by tax administration practices, competent staff and
adoption of the latest technology. Additionally, the author found there is a significant
relationship between tax structures and tax revenue collection.

Eja, et al. (2018) conduct a study on “Inhibiting factors to tax revenue generation in cross
river state, Nigeria” using a cross-sectional survey design and the ordinary least squared
multiple regression model for data analysis. The objectives of the study were to examine the
impact of lack of infrastructural facilities, tax evasion and avoidance and experienced
personnel on tax revenue generation in Cross River State. The author's used a primary source
of data and a structured questionnaire was distributed to sampled staff of Cross River State
board of internal revenue using random sampling techniques.

The founding of the study revealed there is a negative and insignificant relationship between
the lack of infrastructural facilities and government tax revenue. The authors also found that
there is a negative and significant relationship between tax evasion and avoidance and
government tax revenue. Additionally, they found there is a positive and significant
relationship between experienced personnel and government tax revenue. The authors also
suggested the government should encourage tax payer‟s morale through the provision of
quality infrastructural facilities and other incentives to enhance its revenue generation.

Wayau (2017) studied “The effect of environmental factors on revenue collection in Kitui
County” using primary data collected from ministries involved in revenue collection. The
data were analyzed using descriptive statistics such as mean and standard deviation. The
study suggested that political interference, staff competence, internal control and staff
motivation have an impact on tax revenue collection. An improved staff competence,

23
internal control and staff motivation have a positive impact on revenue collection. On the
other hand, political interference was found to affect revenue collection negatively.

Mugambi, (2018) conducted a study in Kenya on “Factors affecting the implementation of

revenue Collection systems” The methodology used was a cross-sectional mixed method and
the questionnaires were distributed to the sampled employees. The study was performed
using both qualitative and quantitative approaches. To present the data result regression
analysis, ANOVA, correlation and descriptive statistics were used.

The finding indicated technology or the ability to make payments using mobile money
platforms can increase revenue collection. He also found out that employee‟s skills including
experience, knowledge, skills & competence will improve service delivery and revenue
collection. According to this study technology and employee, skills have a positive and
significant impact on tax collection.

Adeniyi and Folayan (2018) conducted a study to identify the effects of tax evasion on
government revenue generation. The authors undertook a study using primary data collected
from taxpayers and employees with the support of secondary data. The ordinary least square
was used for data analysis. The authors found that tax evasion has a significant effect on the
revenue generation of the government. Further, the corruption actives performed by
government officials provoke taxpayers to evade more.

Ataro, et al. (2016) studied on “Factors Affecting Revenue Collection Efficiency in County
Governments in Kenya: A Case Study of Trans-Nzoia County” by using primary data
collected from employees from the various departments, and stratified sampling technique to
select them. The author's findings suggest that staff competencies, internal control and
compliance have a positive effect on improving revenue collection practice. Revenue
collection efficiency can be achieved by employing strong audit systems in managing the
revenue collection to reduce the loss of revenue and using an automation system in tax
collection.

Mohammed and Muturi (2018) conducted a study on “factors affecting revenue collection
efficiency in county Governments of Kenya” using primary data collected from employee
work on tax-related issues and secondary data obtained from financial reports. The authors
applied descriptive statistics and inferential statistics (correlation analysis and regression
analysis) to analyze data. The result suggested that tax collection efficiency is affected by
staff competence, taxpayer‟s education (taxpayers‟ knowledge of tax laws) and tax

24
compliance. Staff competence has a positive and insignificant effect on tax revenue collection
efficiency while taxpayer‟s education and compliance have a positive and significant effect
on revenue collection efficiency. Authors further indicated the effect of technology which
shows an improvement in technology leads to collect tax efficiently.

Kelvin (2017) conducted a study on “the impact of system automation on revenue collection
in Kenya revenue authority” using descriptive study design and multiple regression analysis
to investigate the effect of technology on revenue collection. To meet the objective and show
how Simba technology changed the amount of revenue collected secondary data was utilized.
The result suggested that amount of revenue collected was significantly increased as a result
of the technology used. There is a positive and significant relationship between technology
and revenue collection.

2.2.2 Evidence from Ethiopia

Under this sub-section, some of the previous empirical studies conducted in Ethiopia are
stated.

Gobachew, et al. (2018) conducted on the “Determinants of Tax Revenue in Ethiopia” to

identify the determinants of tax revenue in Ethiopia. They used secondary time series data set
for the years 1999/00 to 2015/16 and multiple variables regression model with an OLS
estimator of a variable. They employed a quantitative research method with both descriptive
statistics and econometric tools to analyze and present the data. By using multiple variables
regression model using OLS method they fund that industry sector share to GDP, per capita
income and trade openness as measured by the share of export and import to GDP have a
significant positive effect on tax revenue whereas agriculture sector share to GDP and the
annual rate of inflation has a significant and negative effect on tax revenue as measured by
the share of tax revenue to GDP.

Manaye, et al., (2019) conducted a study on “Determinants of revenue collection efficiency

and administration: the case of local government authorities in Wolaita Zone, Ethiopia” to
identify the determinants of revenue collection efficiency and administration. They used
descriptive types of research design and quantitative research approach. To meet the
proposed objective authors used primary type of data and which is collected from employees
through the questionnaire. They used the systematic random sampling technique of
probability sampling. Further to analyze the effect of independent variables on dependent
variables, they applied a multiple regressions model.

25
The study finding reveals that administration system and enforcement mechanisms have a
positive and significant effect on revenue collection efficiency while tax evasion has a
negative and insignificant effect on revenue collection efficiency. Their overall finding shows
the existence of some weakness in the revenue collection efficiency process in local
government authorities.

Abate (2019) conducted a study on “Factors affecting presumptive tax collection in Ethiopia:
evidence from category “C” taxpayers in Bahir-Dar city” by employing a cross-sectional
survey design and the quantitative research approach. The structured questionnaire was
applied to collect primary data from category “C” taxpayers and only primary data was used.
A simple random sampling technique was used to select a sample from the total population.
The study used a binary logistic regression model to analyze the relationship between
dependent and independent variables.

The finding of the study shows that taxpayers‟ knowledge about tax rules, perception of
equity and fairness, participatory tax system, taxpayers‟ attitudes toward the government and
organizational strength authority have a positive and significant effect on tax collection. The
study further shows corrupted behavior of tax officials has a negative and significant effect
on tax collection. However, taxpayers‟ perceptions of tax evasion and Mode of tax collection
have a negative and insignificant effect on tax collection while social norms have a positive
and significant effect on tax collection.

Shenkute (2018) conducted a study on “Factors affecting tax revenue collection in the case of
Addis Ketema sub-city tax office”. The author used primary data surveyed from category „A‟
and „B‟ taxpayers and supported with secondary data. The study was conducted using
descriptive statistics and inferential analysis. The findings show tax revenue collection is
affected by tax evasion, tax payer‟s awareness, and tax compliance. Tax evasion has a
negative and significant effect on tax revenue collection while tax compliance and tax payer‟s
awareness have a positive and significant impact on tax collection.

Tumoro, (2020) explored “Challenges of business income tax administration with special
emphasis on collection process in Ethiopia: the case of Hadiya zone, Hossana town” using a
descriptive research design with a qualitative approach. He applied both primary and
secondary sources of data. To achieve the objective of identifying the challenges of business
income tax administration and collection process in Ethiopia, he collected data from
taxpayers and employees in the tax authority bureau using questionnaires and structured

26
interviews. The author used a proportionally stratified sampling technique to select
respondents from each category (Category „A‟, „B‟ and „C‟) of taxpayers.

The finding revealed that challenges associated with business income tax collection
procedures are taxpayer‟s attitude towards tax system, lack of good tax administration and
implementation, tax collectors unethical practice during tax collection, availability of
unregistered traders or illegal traders, tax evasion and avoidance.

Ergo (2018) investigated performance determinants of business income tax collection in the
Boditi Town Revenue Collecting Branch Office, Wolaita Zone; SNNPR, Ethiopia using
primary data collected through five Likert scale questionnaires and interview and secondary
data. Descriptive research design with quantitative and qualitative techniques and regression
analysis were used. The result suggested that fairness and equity, culture and social norm,
taxpayer‟s education, attitude towards the government and competency of tax experts have a
positive and significant impact on tax collection performance. Further, the finding indicated
organizational strength of tax authorities has a positive and insignificant effect on tax
collection.

2.3 Summary and gap in the existing literature

Tax revenue collection is the generation of income for the government from legal tax sources.
There are different theories on tax which consist of factors that affect tax collection such as
the ability to pay, benefit theory, theory of tax evasion, institutional theory and technological
determinisms. These theories indicated factors affecting tax revenue collection such as tax
evasion, corruption, tax avoidance, tax rate, penalties and strong organizational activities like
conducting a frequent audit.

Literature shows several empirical studies have identified the effects of macro-economic
variables on tax revenue. These studies identified the impact of GDP, trade openness,
inflation, foreign aid, foreign direct investment, urbanization, agricultural sector, industrial
sector, service sector and mining sector on tax revenue generation Workineh (2016);
Gobachew, et al. (2018); Bayu, (2015); Gupta (2007); Jeza, et al. (2016); Terefe & Teera
(2018) and Kumari (2017). On contrary, even though scant study, some studies identified the
effect of institutional or organizational, administrative factors such as taxpayer‟s awareness,
the capacity of the tax authority, tax evasion, mode of tax collection, political instability,
employee qualification, participatory tax system, equity and fairness of the tax system,

27
 technology, social norms and tax Compliance on tax revenue collection Abate (2019); Amin,
et al. (2014); Eja, et al. (2018) Mugambi (2018); Kelvin (2017) and Manaye, et al. (2019).

However, in these studies, the effect some variable like tax delinquency, political instability
and employee qualification were not included in the Ethiopia study, and Even though study
on the effect of tax delinquency on tax compliance is conducted, a study on the effect of tax
delinquency on tax revenue collection is scant globally. Therefore, this study tried to bridge
this gap.

2.4 Conceptual Framework

From the theoretical and empirical reviewed literature in above; the developed the following
schematic representation of the conceptual framework.

Independent variables
 Tax Evasion
 Corruption
 Employee competence
 Technology
 Organizational strength
of tax authority Dependent Variable
 Political instability
 Tax delinquency
Tax Revenue
Collection
Control Variables
 Gender
 Work experience
 Educational status

Figure 2.1 Conceptual frameworks on determinants of tax revenue collection

28
 CHAPTER THREE
RESEARCH METHODOLOGY
The objective of this study is to investigate determinants of tax revenue collection in the
Sululta town revenue office. To achieve the objective properly, this chapter discussed the
methodology of how the study was done scientifically. It presented a description of the
research design; that is, sampling frame, sample size, study area, the tools and procedures
used in the data collection and analysis. The sequential patterns from the nature of data,
sampling procedures, data analysis and interpretations have been aligned to attain the purpose
of the study.

3.1 Research Design

Descriptive and explanatory research design was used. Research design is the overall plan of
how this study was done. It is a blueprint for the collection, measurement, analysis of
collected data; thus it shows how the study is implemented from problem definition to
conclusion of the study. It is through research design that one can answer various questions
such as “what is the objective of the project? What data is needed? How data will be
collected and analyzed?” (Kothari, C, 2004). Because the objective of this study was to
investigate the determinants of tax revenue collection, descriptive and explanatory designs
were used to evaluate the relationship between dependent and independent variables.
Additionally, because descriptive and explanatory research design involves both quantitative
and qualitative data; the mix of both approaches was used for this study.

In cross-sectional research design analyses, data collected from the population at a specific
point in time (Saunders, et al., 2009). Thus a cross-sectional survey design was applied to
gather data about the determinants of tax revenue collection in respect of category “A”, “B”
and “C” taxpayers in Sululta town. A cross-sectional study as the data was collected on one
occasion and represents a snapshot of the respondents‟ responses at that specific point in
time. This design was used because the aim is to find out the determinants of tax revenue
collection, by taking a cross-section of taxpayers in Sululta town. Moreover, the
characteristics of the variables would not change due to the shorter period spent on data
collection.

29
3.2 Research Approach
To accomplish this study, a research approach was selected based on the objective and the
research hypothesis addressed. Research approaches consist of qualitative, quantitative and
mixed approaches (Cooper & Schindler, 2014). A quantitative approach is consists of
numerical data which is subjected to quantitative analysis. On the other hand, the qualitative
approach involves data in textual form that is concerned with the subjective valuation of
attitudes, thoughts and behavior.

In this study, a mixed research approach was applied. That is combination of both
quantitative and qualitative approaches. Mixed methods approach was used because it offer
to draw from the strengths and minimize the weaknesses of the quantitative and qualitative
research approaches thereby putting in more accurate result (Johnson & Onwuegbuzie, 2004).

3.3 Study Area and Target Population

The study was conducted to investigate the major determinants of tax revenue collection in
the Sululta town revenue Office. Sululta is a town of Sululta special zone surrounding
Finfinne (Addis Ababa) to the northern part, which is about 25km from the capital of the
country. Sululta special zone is among zones of Oromia National Regional state founded in
1937. Several business activities are operating in Sululta town including manufacturing,
merchandise, service, Small and medium enterprise and agriculture which are a source of
business income tax for the government.

Figure 3.1 Location map of Sululta town

Source: Sululta town municipality

30
The factors that affect tax collection can either be within the tax revenue office or taxpayers.
Therefore, collecting data from both taxpayers and employees (manager, auditor and tax
collectors) of the tax revenue office was considered reasonable to reach an appropriate
conclusion. The specifically targeted taxpayer are those included under business income tax
during the study, which includes category “A”, “B” and “C” taxpayers.

3.4 Types of Data

A primary and secondary type of data was used for this study. Kothari (2004) states there are
two types of data depending on the sources and techniques one uses for gathering data. Types
of data that are collected afresh and for the first time is and happened to be original in
character is primary data. In this study, primary data was used, which was collected from
taxpayers and employees of the tax revenue office. The data was collected properly; this may
enable the study to be representative. On the other hand, if data have already been collected
by someone else and have already been passed through the statistical process is said to be
secondary data. Secondary data was also used to check for the trends and gaps in tax
collection.

As indicated by Hox (2004), in social science one can use primary or secondary data.
Although primary data collection is costly and time-consuming, the operationalization of the
theoretical constructs, the research design and the data collection strategy can be tailored to
the research question, which ensures that the study is coherent and the information collected
indeed helps to solve the problem. On the other hand, although secondary data collection is
lower costs and faster access to relevant information, the data were originally collected for a
different purpose; therefore it may not be optimal for the research problem under
consideration or, in qualitative data may not easy to interpret without explicit information on
the informant.

Since both types of data have their own weakness, this study applied both of them
simultaneously to the best so that the research hypothesis and problem identified can be
solved. Therefore, both primary and secondary types of data were applied for this study.

31
3.5 Data Collection Technique
There are several methods of collecting primary data, particularly in surveys and descriptive
researches, such as observation method, interview method, questionnaires, schedules, and
other methods which include warranty cards, and distributor audits.

In this study, a self- administrated questionnaire guided by five points Likert scale for
taxpayers and semi-structured interview for employees were used. The questionnaire was
adopted from other studies such as Abate (2019) and Gituma (2017), but major improvements
were made to questionnaires. The self-administrated questionnaire was distributed to
respondents, and it was managed closely. Likert scaling is a summated rating scale that
consists of a series of statements concerning an attitude object. Scores of individual items
were summated to produce a total score for the respondent. Likert scale was used because it
enables to collect of more information than merely yes or no answers. Additionally, the
Likert scale was applied, because the Likert scale can make data collection easy and improve
the accuracy of data. Moreover, the respondents may not be bored in writing a long sentence
and they can fill with enthusiasm.

With regard to this study‟s independent variables, respondents were asked to indicate their
level of agreement on five-point Likert scale, as follows: 1 = Strongly disagree; 2 = Disagree;
3 = Neutral; 4 = Agree; 5 = Strongly agree. The data collection instrument or questionnaire
was first prepared in the Afan Oromo and Amharic language considering taxpayers and then
translated into the English language for data analysis.

Secondary data was collected from the tax collection records and unpublished working papers
which show the trends of tax collection and budget and actual tax collected in the Sululta tax
revenue office.

3.6 Sampling Method

To meet the objective of the study appropriate sampling method was tried to be designed. The
target respondents were taxpayers which are about 4,115 in total and employees which are
about 25 in total. Although the census method covers all the members by the inquiry no
element of chance is left and the highest accuracy is obtained, the census method is costly in
terms of time, money and energy and further, there is no way of checking the element of bias
or its extent except through a resurvey or use of sample checks. Therefore sampling methods
is appropriate in such a case. In the case of sampling, it is possible to get much more accurate
information than is possible by collecting data from the representative body and if

32
homogeneity exists in target populations. This study also targets the population, when
observed as they are, they seem somehow heterogeneous, but it is possible to make them
homogenous by using appropriate sampling methods. This study was applied a probability
sampling method for taxpayers while the non probability of purposive sampling method was
used to select from total employees.

According to Kothari (2004), based on the representation basis sampling methods can be
classified as probability and non-probability sampling methods. The technique of probability
sampling is based on the concept of random selection, whereas non-probability sampling is
„non-random‟ sampling. Random sampling, systematic sampling, stratified sampling, cluster
sampling and multi-stage sampling are considered as probability sampling. Purposive
sampling, quota Sampling and convenience sampling are considered as non-probability
sampling. The selection of sampling methods based on the homogeneity and representation of
the sample.

In this study the target population was heterogeneous, consisting of category “A”, category
“B” and category “C” taxpayers. This fits the characteristic of stratified probability sampling.
Hence the stratified probability sampling method was used for taxpayers.

The basic procedures followed for the stratified sampling method were as follows; firstly, the
strata were formed based on common characteristics of the group to be put in each stratum.
That is first the stratum category “A”, the second stratum category “B”, and the third stratum
category “C”. Secondly, the sample size was calculated based on the Cochran formula.
Thirdly, the disproportionate allocation method was used to identify the specific sample size
of respondents from each stratum. Finally, from each stratum, the determined sizes of
respondents were selected using a simple random sampling method.

3.7 Sample Size

For determining sample size there are three approaches those are using a census for small
Populations, using a sample size of a similar study which may lead to repeat the same
problem and using formulas to calculate the sample size. In this study, the formula was to
calculate sample size

Cochran's (1963) formula was used to calculate the sample size due to the large number of
populations the study.

33
Where n is a sample size, Z2 is the abscissa of the normal curve cut off an area α at the tails (1
- α equals desired confidence level is 95%), e is desired level of precision, p is the estimated
proportion of attribute and q is 1-p.

( )( )
( )

The total number of taxpayers is 4,115 from which Category “A” taxpayers total number is
157, Category “B” taxpayers is 239 and Category “C” taxpayers is 3,719.

( )

The next step was to determine the size of the sample to be sampled from each stratum by
using a disproportionate stratified sampling method, which is as shown in table 3.1.

Table 3.1 sample size for each taxpayer‟s category

Stratum Total Proportionate Percent Disproportionate Percent

population Sample size sample size
Category “A” 157 13 3.7% 42 12%
Category “B” 239 21 5.98% 63 18%
Category “C” 3,719 317 90.32% 246 70%
Total 4,115 351 100% 351 100%

Source: own construction

A disproportionate stratified sampling method was used for allocating sample size because
with proportionate stratified sampling method the sample size from category “A” and “B”
taxpayer is too small to represent population. Thus to make accurate inference from the
sample the disproportionate stratified allocation method was used.

3.8 Variable Operation and Measurement

The dependent variable used in this study is tax revenue collection and the independent
variable applied are tax evasion, corruption, employee competence, technology,
organizational strength of tax authority, political instability, tax delinquency and controlling
variables include gender, work experience of taxpayers and educational level of taxpayers.
For simplification purposes based on OECD (2013) tax evasion, corruption and tax
delinquency were considered as administrative factors, because these variables occurred with
tax administrative and effacement capacity. In analyzing tax compliance Palil (2010)

34
indicated that institutional factors consist role of the tax authority capacity, simplicity of tax
system and probability of performing an audit. Accordingly, in this study institutional factors
(tax office inherent factors) include employee competence, technology and organizational
strength of tax authority. Nut & Nut (2020) also considered political instability as an
institutional factor; thus it is considered as an institutional factor. Additionally, Demographic
variables (controlling variables) include gender, work experience of taxpayers and
educational level of taxpayers.

3.8.1 Dependent Variable

Tax revenue collection is a process of government billing the public and members of the
public for taxes on income and profits, social security contribution, the tax levied on goods
and services, payroll taxes, taxes on ownership and transfer of property and other tax. The
increment in tax revenue collection shows that an increase in tax revenue of the government.
This study was focused on tax revenue collection from business income tax.

The dependent variable (tax revenue collection) has ordinal nature. The previous studies like
Abate (2019) used only two categories for outcome variable (i.e. poor or good), but this study
expanded it by using three categories which coded as low tax revenue collection „0‟, medium
tax revenue collection „1‟ and high tax revenue collection „2‟. Due to the ordinal nature of the
data, the dependent variable was measured on an ordinal scale.

3.8.2 Independent Variables

The independent variables used in this study as explained here under their explanations.
Before explaining variables showing their measurement of scale is important. There is
considerable debate and inconsistency in literature on the issue of whether Likert scale data
should be analyzed as an ordinal or interval measurement scale. This debate created two main
schools of thought. One school considers the Likert scale as ordinal and the other treats it as
an interval scale. Both thoughts agreed on the question: “whether points on items are
equivalent and equidistant?” Points on the scale are not close enough to consider them equal
which means strongly agree is definitely away from agree and agree is away from neutral,
and they should be considered as a non-equivalent entity. The conflict arises on asking
another question: “if the points on scale are non –equivalent, are they equidistant which mean
is „neutral‟ of same distance from „agree‟ as „agree‟ from „strongly „agree‟? (Joshi, et al.,
2015).

35
 A. Likert scale as ordinal (Categorical)

The first school of researchers and statisticians consider the Likert scale as an ordinal scale.
They argue that choices or responses are arranged in some ranking order. However, as this
scale doesn‟t show the relative magnitude and distance between two responses quantitatively,
it can‟t be treated as an interval scale. Therefore, if Likert scale is measured as an ordinal
scale, measurement of central tendency such a mode and median, and measurement of
dispersion such as range and quartile can be used (Joshi, et al., 2015)

B. Likert scale as Interval scale

On the other hand, the second school interprets this dilemma from a different perspective,
stating that when the aim of the researcher is to „combine‟ all the items in order to generate a
„composite‟ score for an individual rather than separate analysis of a single item responded
by all individuals, then this individualistic summative score (for all the items) of a participant
shows a sensible realistic distance from the individual summative score of another individual;
hence, can be labeled as „interval estimates‟. Therefore, if the Likert scale is measured as an
interval scale from the measurement of central tendency a mean and standard deviation from
the measurement of dispersion can be used (Joshi, et al., 2015).

In this study the aim is not to separately analyse a single item responded by all individuals,
rather to combine every single item to obtain a composite score. Thus, the Likert scale was
measured as an interval scale. Other studies such as Abate (2019); Gituma (2017) and
Shenkute (2018) also analyzed the Likert scale as an interval scale.

Tax evasion

Tax evasion is the efforts by individuals, firms, trusts and other entities to evade the payment
of taxes by breaking the law. Tax evasion usually implies taxpayers‟ activities of intentional
misrepresenting or concealing the true state of their affairs to the tax authorities in order to
minimize their tax liability, and includes, in particular, dishonest tax reporting such as
understating their actual income, profits or gains; or in reverse overstating deductions. As
indicate by Adeniyi & Folayan (2018) tax evasion is one of the main factors that reduced tax
revenues. Additionally, Mtasiwa (2013) and Masarirambi (2013) state the negative
relationship between tax revenue collection and tax evasion.

Corruption

36
In this study, corruption can be defined as unlawful or unauthorized action performed by
public tax officials using their position to receive a bribe in exchange for making a benefit
available to a member of a specific public group like taxpayers. Different empirical pieces of
evidence show that corruption reduces tax revenues Amin, et al. (2014) and Adeniyi &
Folayan (2018). Additionally, Abate (2019) indicated the corrupt behavior of tax officials
resulted in a low level of tax collection.

Employee’s competence

Employees are staffs of an organization who work toward achieving the organization's
objectives and mission. Employee competence is progress that arises for the purpose of
developing staff capacity and potential within corporate framework Harrison (2007). The tax
administration personal has the power to influence the business community through tax, and
employee's fault in tax revenue collection causes a poor level of tax collection. Employee
qualifications improve tax collection performance and it has a positive relationship with tax
collection (Mugambi, 2018) and (Eja, et al., 2018).

Technology

Technology has influenced the daily life of human beings and the ways they interact with
others. It also affected the way tax systems are designed and administered in developing
countries. Technology in taxation enables to build visualizations using provision data and
easily facilitate multi-period comparisons and trend analyses. Moreover, it also offers a
chance for proper tax accounting processes. Additionally, policymakers need to focus on the
effect of dynamic in technology on the design of tax systems and tax instruments in raising
revenue. One study in Kenya proved that the amount of tax revenue collection is increased
after the implementation of a new automated tax collection system which is the Simba system
(Kelvin, 2017). Another study also indicated the positive effect of technology on tax
collection (Ngicuru, 2017) and (Mohammed & Muturi, 2018). However, Abate (2019)
revealed that there is an insignificant relationship between tax revenue collection and mode
of tax payment.

Organizational strength of the tax authority

The organizational strength of the tax authority can be in terms of having adequate
technological and human resources, delivering quality tax services to taxpayers, having
strong law enforcement potential, and transparent complaints resolution procedures are ideal
for increasing tax revenue. It also involves the way how tax authority performs tax audits

37
frequently. Organizational strength of tax authority is a factor for the low level of business
income tax collection (Ergo, 2018). According to Manaye, et al. (2019) there is a significant
relationship between tax administration and tax revenue collection. Similarly, Abate (2019)
stated that the organizational strength of tax authority is positively affected tax revenue
collection.

Political instability: political instability is may relate to violent conflicts of low intensity.
Political instability arises due to types of government coalitions with paramilitary groups in
rural areas, displacing people and expropriates land from its owners. According to Estrada
(2010) levels of violence vary from massacres against the opposition political groups to
assassinations of presidential candidates. Amin, et al., (2014), that one main reason for a low
level of tax revenue collection is political instability. During the period of political instability,
all government bodies including tax authority think about bringing stability than collecting
tax; therefore, tax revenue collection falls from the budgeted amount. Based on Wayau,
(2017) also political instability has a negative impact on tax collection, and it blocks the
overall economic growth.

Delinquency Tax

Tax delinquency is essentially taxes owed to the tax revenue office that is not paid within the
due date by taxpayers. Tax is considered as delinquent once taxpayers miss the filing or
payment. Likewise, tax evasion, the government‟s fiscal health in terms of revenue and
equity can be damaged by tax delinquency. Tax delinquencies can increase in loss of revenue
and costs of collection would also be high. The government is compelled to implement
enforcement collection processes that are costlier to both the government and the taxpayer.
Tax delinquency can be seen as a way that honest taxpayers can shift toward non-compliance,
leading to a violation of the equity principle in taxation. Thus a reduced compliance rates
decrease tax revenue and increase tax collection costs in the long run Jang & Eger (2019).

38
 Table 3.2 Summary of Variables and Measurement
Variables specification Symbols Description Measurement of Expected
scale Sign
Tax Revenue collection TRC The level of revenue collected from Ordinal scale DV
tax sources. (0=low,
1=medium, &
2=High)
Tax evasion Teva The level of taxpayer evades from tax. likert scale (-)
Corruption Corr The effect of corrupted tax officials on likert scale (-)
tax collection.
Employees competence Epl The skill of employees in the tax office likert scale (+)
on collecting tax properly.

Technology Tech The convenience of technology in the likert scale (+)

collection of tax.
Organizational strength of Tautho The capacity of tax administration on likert scale (+)
tax authority collection tax revenue on time.
Political instability PIs The level to which political instability likert scale (-)
affects tax revenue collection.
Tax delinquency Tdel The level of Tax does not collect likert scale (-)
before the due date.
Gender Gender Dummy variable (1=male, 2=female) -
Work experience of WExp Dummy variable (1= 5 and below, -
taxpayers 2=6-10, 3=11-15, 4=16 and above),
this shows how long taxpayers have
been paying tax in the current
business.
Educational level Edu A dummy variable (1= G12 and -
Below, 2=Diploma, 3= first degree and
above
Source: own construction

3.9 Data analysis and interpretation methods

The objective of this paper is to identify determinates of tax revenue collection, to meet these
objectives data were collected as indicated in the above section and both descriptive and
econometric methods were applied for analyses. The data analysis was followed the
following procedure.

The collected data was processed so to get ready for analysis. Initially from the collected
questionnaire, data were edited to detect errors and omissions. Then coding activity was
performed which was the process of assigning numerals or other symbols to each sample so
that responses can be put into a limited number of categories or classes. The next step was

39
classifying data on the basis of common characteristics. The data were classified according to
the data related independent variables and other data. Finally, the data was entered into the
software.

The next step was data analysis. Analysis is the computation of certain measures to search for
patterns of relationship that may exist among the data groups. It consists of computing the
values of unknown parameters of the population and drawing inferences from testing
hypotheses. Both descriptive analysis and inferential analysis were used. In descriptive
statistics what is or what the data shows were described using descriptive statistics like mean
and standard deviation. With inferential statistics, the conclusions were drawn from the
sample data.

In this study, a computer software package was used in performing data analysis. The
software package is the most advanced technology in the research area to process big data
with higher accuracy and speed. Techniques involving trial and error process are quite
frequently employed in research methodology which involves several calculations and work
of repetitive nature. However, such techniques can be best performed by the computer, thus
reducing the tediousness of researchers on the one hand and producing the final result rapidly
on the other. Computation of means, standard deviations, correlation coefficients, t„ tests,
chi2 test, analysis of variance, analysis of covariance, factor analysis and various
nonparametric analyses are just a few of the programs and subprograms that are available at
almost all computer software. There are about five software packaged used in social sciences
including SPSS, STATA, R Foundation for Statistical Computing, SAS and E-views. For this
study Stata version, 16 and SPSS version, 20 were used.

The proportional odd results were interpreted using marginal effect and coefficient. Mood
(2010) argued that the odds ratio has some drawbacks; thus alternative probability‐based
measures, including marginal effects (partial derivatives) have advantageous over odd ratio.
Therefore, instead of odd ratio ordered logistics standard coefficient interpretation and
marginal effect applied.

3.10 Model Specification

In this study, the aim is to identify the determinants of tax revenue collection which have
qualitative nature. When the dependent variable is qualitative in nature, using the OLS model
for estimation pose several problems. There are appropriate alternatives for such categorical
variables. Those include logistic model, probit model, Tobit model, Poisson regression

40
model, multinomial logit, multinomial probit, ordered logit and ordered probit model. The
first four models are commonly used for binary variables meaning when the dependent
variable has only two categories. The last four models are appropriate if the outcome variable
has three or more categories (Gujarat, 2004).

The multinomial logit and multinomial probit are used when outcome variable categories
have no nature order. The ordered logit and ordered probit model are used when outcome
variable categories have a nature of the order. In this study dependent variable was designed
with an ordering nature, which is from low to high. There is no significant difference between
ordered logit and probit models. Therefore an ordered logit model was used in this study.

3.10.1 Ordered logistics model

As outcome variable has an ordering nature an ordered logistic regression model was used in
order to investigate whether direct relationships exist between the determinant factors
proposed as independent variables and the tax revenue collection (TRC). Ordinal logistic
models (OLM) are used to predict the level of tax revenue collection. The most used model
for ordinal regression is the Proportional Odd (PO) model which assumes that the effect of
each predictor remains the same for each category of the response variable. Hence, the
Proportional Odd model was used in an ordered logistic regression model.

In an ordered logistic model only the ordering can be interpreted and not the actual numerical
values, therefore to estimate parameters instead of ordinary least square (OLS) a maximum
likelihood (ML technique was applied.

Dependent variable (TRC) has three categories 0= low, 1= medium and 2= high. In the
ordered logistic model, there is an observed ordinal variable, which is tax revenue collection
(Y). This dependent variable (Y,) in turn, is a function of another variable, which is latent
variable (yi*), that is not measured and whose values determine what the observed ordinal
variable Y equals.

Y has j possible outcomes Y=j for j=1… M. Natural ordering (i. e, j + 1 are in some sense
‟‟better‟‟ than j) that is:

, * +- (1)

The latent variable model is given by:

For i=1…N (2)

41
Where, y*= Latent (unobserved) measure of tax revenue collection, Xi = a vector of
explanatory variables, α, βi= coefficients estimate, and ɛ= a random error term.

The observed ordinal variable (Y) is determined based on the latent outcome variable model
as follows:

(3)

Where is a set of threshold or cut point of the level of tax revenue collection estimated with
parameter β.

To estimate the in (odds) of being at or below the jth category, the proportional odds model
can be estimated as follows:

, (
( | )
. ( |
/ (4)
)

Different software packages may estimate parameters differently and the ordered logistic
regression model can be expressed in different forms (Liu, 2009). The above developed
model was used in Stata.

3.10.2 Assumptions of Ordered logistics model

The ordered logistics model (OLM) has commonly four main assumptions to be checked
before running the model. These assumptions are:

1. The first assumption is dependent variable should be measured at the ordinal level.
2. Independent variables can be continuous or ordinal or categorical (including
dichotomous variables).
3. There must be no multicollinearity problem. Multicollinearity problem occurs when two
or more independent variables are highly correlated with each other. If the problem of
multicollinearity exists, it would be difficult in identifying as which variable contributes
to the explanation of the dependent variable and technical issues in calculating an
ordinal regression. Accordingly, the VIF test was used, and it showed there is no
Multicollinearity problem.
4. Proportional odds assumption or parallel line test is another assumption of OLM which
is a fundamental assumption of in ordinal logistics model. The assumption of

42
 proportional odds shows that each independent variable has an identical effect at each
cumulative split of the ordinal dependent variable. To test PO Brant test was used with
Stata and a parallel line test with SPSS indicated the assumption is met.

3.11 Validity and Reliability Test

Before using the data collected through a data collection instrument which was a
questionnaire, the validity and reliability were tested.

3.11.1 Validity Test

Validity tests the extent to which differences may found with a measuring instrument reflect
true differences among those being tested. It measures the degree of agreement of the result
obtained from the research questionnaire with the real world. Validity shows the degree to
which an instrument measures what it is supposed to measure. There are about three types of
validity such as content validity, face validity, criterion-related validity and construct validity.

For this study content validity was established. Content validity was used to assess whether a
test is representative of all aspects of the construct. Thus, the questions in the questionnaire
were checked by advisors who understand well this topic and irrelevant questions were
dropped and the questionnaire was reconstructed for data collection.

3.11.2 Reliability Test

Reliability is the degree to which the data collection instrument and analyzing procedures
yield consistent findings (Saunders, et al., 2009). Reliability analysis measures the internal
consistency of a group of items, which used in questionnaire construction. It is the extent to
which the same questionnaire would produce the same result if the study is to be conducted
again on the same topic. It is used to describe the consistency of the questionnaire.

To determine the consistency of the questionnaire Cronbach‟s alpha coefficient was used. As
indicated by Nunnally (1978) an alpha score of higher than 0.70 is generally considered
acceptable. Therefore, in this study, seven explanatory variables are included in the model to
investigate their effect on tax revenue collection. The reliability test shows seven independent
variables have an alpha value of greater than 0.7. The highest alpha value is 0.823
(Corruption) while the lowest alpha value is 0.703 (employee competence).

43
Table 3.3 Reliability of the instrument

Variables observation Cronbach’s alpha

Tax evasion 279 .704
Corruption 279 .823
Employee competence 279 .703
Technology 279 .711
The organizational strength of tax 279 .712
authority
Political instability 279 .732
Tax delinquency 279 .711

Source: Computed from own survey data, 2021

44
 CHAPTER FOUR
RESULT AND DISCUSSION
Investigating determinants of tax revenue collection and their effect is a key and basic issue
to design and implement a tax collection system. To better indicate factors affecting tax
revenue collection the survey data were collected from Sululta town which is one of the
special zones surrounding Addis Ababa in the Oromia National Regional State.

Summarizing the determinants of a single category of taxpayers to all other categories may
not result in a valid conclusion. Thus, sample respondents were selected from each category
(category „A‟, „B‟ and „C‟) of taxpayers.

In this chapter collected data were analyzed using descriptive statistical tools and
econometrics tools. Data analysis was performed with the help of SPSS version 20 and Stata
version 16. The chapter is organized into three main sections. Accordingly, in the first
section, descriptive analysis is presented. The second section presented econometrics
analysis, and the final section deals with results and discussion.

4.1 Descriptive Analysis

Descriptive analysis was used to describe characteristics of respondents, relevant aspects of
tax revenue collection and provide detailed information about each variable. Descriptive
statistical tools used were the mean and standard deviation to describe respondent‟s responses
on related variables in the study. However, before directly deep in through descriptive
analysis, the reliability of the instrument test is presented.

4.1.1 Respondents Demographic Characteristics

To meet the objective of the study 351 sample of questionnaires were distributed for
taxpayers, and 279 (79%) questionnaires were returned which is rationally satisfactory.
According to Nulty (2008) the response rate is acceptable as it surpassed 70% of total sample
size. Additionally, an interview was made with 3 employees which were 2 males and 1
female (manager, auditor and tax collector). The following table (Table 4.1) shows the
gender, work experience and educational status of respondents. As indicated 60.1% of the
respondents were male while 39.9% of the respondents were female. This implies that the
majority of respondents were male, and as the proportion of employees in this percentage is
small, the table may show the majority of the taxpayers are male.

45
Work experience shows taxpayers tax paying experience within the current business.
Regarding the work experience of respondents, about 39.78% of the respondents responded
to below 5 years while 6.09% of respondents were above 16 years. This implies that
responded taxpayers have a high proportion of below 5 years tax paying experience within
the current business.

Concerning the educational status of respondents, 81.36% were G12 and below. On the
contrary, 5.38% of respondents were the holder of the second degree and above. From this
result, the educational status of the majority of the taxpayer was G12 and below. This
indicates most educated people have not participated in business activities and the likelihood
of having less awareness about tax could be high among taxpayers which in turn affect tax
revenue collection. However, taxpayers with adequate education status have a probability to
enhance ways for tax collection.

Table 4.1 Demographic characteristics respondents

Gender of respondents Frequency Percent

Male 172 60.1

Female 110 39.9

Total 282 100

Work experience of respondents Frequency Percent

Below 5 111 39.78

6-10 109 39.07

11-15 42 15.05

Above 16 17 6.09

Total 279 100

Educational status of respondents Frequency Percent

G12 and below 227 81.36

Diploma 37 13.26

First degree and above 15 5.38

Total 279 100

Source: Computed from own survey data, 2021

46
As indicated below table (table 4.2) data were collected from each category (Category „A‟,
„B‟ and „C‟) of taxpayers. Accordingly, the respondents from category „A‟ were 13.98%,
category „B‟ taxpayers were 18.64% and category „C‟ taxpayers were 67.38%. This implies
all categories of taxpayers were considered in this study to make an inference that reduces
intuitive generalization from one category to all categories of taxpayers. Therefore, it is
reasonable to say factors included in this study are the determinants of tax revenue collection
for the three categories (Category „A‟, „B‟ and „C‟) of taxpayers.

Table 4.2 Proportion of respondents from each category of taxpayers

Category of taxpayers Frequency Percentage

Category “A” 39 13.98
Category “B” 52 18.64
Category “C” 188 67.38
Source: Computed from own survey data, 2021

4.1.2 Tax revenue collection

Under this subsection, the level of tax revenue collection in Sululta town and results from the
respondent‟s response on the determinant of tax revenue collection is described. Table 4.3
shows the respondent's perceptions on the level of tax revenue collection. Accordingly, 43%
of respondents evaluated the tax collection level of Sululta town by tax revenue authority as a
medium which is followed by 31% of respondents who responded to a low level. However,
only 26% of respondents responded to the high tax collection level. This result indicates that
the tax revenue authority needs to improve its tax revenue collection.

Taxpayers have a probability of knowing who pays tax and who evade tax as well as source
informal economy that is not taxed. Thus based on respondents response, the tax authority is
not collecting the amount of tax expected to be collected.

Table 4.3 Respondents perception on the level of tax revenue collection

Level of tax revenue collection Frequency Percent

Low 84 31
Medium 117 43
High 78 26
Source: Computed from own survey data, 2021

The questionnaires were determined on a five-point Likert scale representing 1 strongly

disagree to 5 representing strongly agree. The mean value of the respondent‟s responses to
variables was used to show the extent of their agreement or disagreement on the issues.
Moreover to make interpretation easy all questions in the instrument were developed

47
positively, and the scales were reassigned as follows 1 - 1.8= Strongly Disagree, 1.81 – 2.6 =
Disagree, 2.61 – 3.4= Neutral, 3.41 – 4.20= Agree and 4.21 – 5 = Strongly Agree (Dhaba,
2017). Further standard deviation result was used to show the variability of respondent‟s
response on the issue under study. The larger standard deviation (Standard deviation>1)
shows the existence of variability of the responses.

Tax evasion and corruption are illicit financial flows. Based on the above classification of
agreement or disagreement as shown in table 4.4 the mean value of tax evasion is 3.61 and its
standard deviation is 0.95. This indicates taxpayers were evading tax, and the respondents'
responses deviation is low. Since there is an agreement with the existence of tax evasion and
low variation among respondents, there could be a probability of collecting low tax revenue.
Corruption is the explanatory variable with the highest mean value (3.85) and the standard
deviation is (1.06). The mean value implies that taxpayers agreed with the existence of
corruption during tax assessments and collection. Thus, corruption is one of the factors that
can reduce tax revenue collection. Further, the standard deviation value indicates that the
respondents' responses are deviated in between the level of agreement on the scales of
measurements.

The other explanatory variables are employee competence, technology and organization
strength of tax authority. The mean value of Employee competence is 2.91 and its standard
deviation is 0.94. The result implies respondents‟ responses were neutral about employee
competence. The mean value and standard deviation of technology are 2.83 and 1.27
respectively. The response is with high deviation (standard deviation>1); this indicates the
inconsistency of response, which in turn it implies the existence of a problem with
technology (mode) of tax collection. The mean value and standard deviation of organizational
strength of tax authority are 2.39 and 1.01 respectively. This implies that the disagreement on
the organizational strength of tax authority in terms of technological and human resources as
well as adequacy on enforcement mechanisms. Further, it shows tax authority has a low
capacity to levy and collect tax from all categories (category A, B, C) of taxpayers. Thus, this
may cause some activities not to be taxed and less amount of tax to be collected.

Political instability and tax delinquency are other explanatory variable. From the survey
result, the mean value of political instability is 3.11 and the standard deviation is 1.01. This
implies respondent‟s response is neutral but with a variation on the issue. The respondent‟s
response on tax delinquency shows the mean value and the standard deviation are 3.68 and

48
0.97. From this, it can be said there is tax delinquency which in turn be a factor for a low
level of tax collection.

Table 4.4 Descriptive Statistics for the explanatory variables of the study

Variables Obs. Mean Standard Maximum Minimum

deviation
Teva 276 3.61 0.95 5 1
Corr 279 3.85 1.06 5 1
Epl 279 2.91 0.94 5 1
Tech 279 2.83 1.27 5 1
Tautho 279 2.39 1.01 5 1
PIs 279 3.11 1.01 5 1
Tdel 279 3.68 0.97 5 1
Source: Computed from own survey data, 2021

4.1.3 Tax revenue collection performance

Tax authority uses the annual tax collection budget and compares it with the collected tax
revenue to evaluate its performance. Accordingly, to support data collected from taxpayers
and employees, documented data was obtained from the Sululta tax revenue office. Sululta
revenue office plans to the action of tax collection by broadening source of income that is
base for socio-economic development.

Table 4.5 shows the plan and performance of tax revenue collection from all categories
(category A, B, C) of taxpayers. The performance percentage of tax revenue collection from
the year 2011 to 2020 shows great fluctuation. The highest revenue collection performance
was in the year 2011 (157.3) indicated above plan. In contrast, the lowest revenue collection
performance was in the year 2016 (68.5%). Tax authority was underperformed particularly
during the years 2012, 2013, 2016, 2018, 2019 and 2020. This implies, from data for ten
years, tax collection performance was below plan for six years. The average tax collection
performance was 87.5%, and the overall trends of tax revenue collection were unstable. Thus,
it can be said there is a probability of collecting low tax revenue.

49
Table 4.5 Plan and performance of tax collection

Year Plan (ETB) Performance (ETB) Performance in

percent
2011 9,668,494.00 15,212,686.58 157.3
2012 19,266,677.00 17,379,985.95 90.2
2013 28,014,299.00 25,633,647.65 91.5
2014 31,365,769.00 43,022,755.14 137.3
2015 55,000,000.00 62,482,736.34 113.6
2016 92,621,649.00 63,766,991.19 68.5
2017 72,374,931.00 76,921,048.38 106.28
2018 177,239,407.59 141,791,026.07 79.9
2019 150,116,993.26 124,506,920.92 82.2
2020 192,137,058.30 153,780,319.44 80
Total 827,825,278.15 724,498,117.66 87.5
Source: Sululta tax revenue office, 2020

4.2 Econometric Analysis

The ordered logistics model (OLM) is applied using maximum likelihood estimation to
estimate the parameters of the OLM equation. Before running regression, the assumptions of
OLM and model fitting information were tested.

4.2.1 Chi-square Tests of Explanatory Variables against Tax Revenue collection

Before analyzing the ordered logistic model, the association of each explanatory variable
with the outcome variable was checked using the Pearson chi-square test. Table 4.6 shows tax
evasion, corruption, technology, organizational strength of tax authority, political instability
and tax delinquency variables are significantly associated with tax revenue collection at
p<0.05.

Table 4.6 Chi-square Tests of Explanatory Variables against Tax Revenue collection

Variables Chi-squire p-value

Tax evasion 87.0667 0.000
Corruption 87.9285 0.000
Employee competence 45.5154 0.057
Technology 42.4667 0.011
Organizational strength tax authority 42.4704 0.005
Political instability 50.5668 0.001
Tax delinquency 74.3474 0.000
Gender 4.7369 0.094
Work experience of taxpayers 10.6168 0.101
Educational level of taxpayers 7.3582 0.118
Source: Computed from own survey data, 2021
50
 4.2.2 Assumptions of the ordered logistic model
The four main assumptions such as dependent variable should be measured at ordinal level,
explanatory variables should be continuous, ordinal or categorical, absence of
multicollinearity and proportional odd were tested. Accordingly, the first and second
assumptions are directly observable and met which is dependent variable was measured at the
ordinal scale and explanatory variables were measured at the interval scale (continuous). This
implies OLM can be used.

4.2.2.1 Multicollinearity
The absence of multicollinearity is the third assumption of ordered logistics regression. When
two or more independent variables are highly correlated with each other it can be said
multicollinearity exists. The variance inflation factor (VIF) and correlation coefficients were
used to detect the existence of multicollinearity among seven independent and three
controlling variables. The mean value of VIF is 1.09 which is less than 10. This shows the
absence of multicollinearity problems.

Table 4.7 Correlation Matrix among the Variables

Teva Corr Epl Tech Tautho PIs Tdel Gender WExp Edu

Teva 1.0000
Corr 0.2019 1.0000
Epl -0.0015 0.0326 1.0000
Tech -0.2249 0.0359 0.0809 1.0000
Tautho -0.0609 -0.0568 0.0534 0.1299 1.0000
PIs 0.1659 0.0856 -0.0056 -0.0743 -0.0329 1.0000
Tdel 0.2530 0.3185 0.0541 -0.1054 -0.1161 0.1821 1.0000
Gender 0.0559 0.0020 -0.0510 -0.0183 0.0152 0.0261 -0.0345 1.0000
WExp 0.1066 0.0754 0.0607 0.0373 0.0169 0.1271 0.1205 0.0056 1.0000
Edu 0.0394 0.0268 -0.0698 0.0699 0.1189 0.0882 -0.0116 -0.0433 0.1239 1.0000

Source: Source: Computed from own survey data, 2021

Multicollinearity is said to exist when there is a high correlation among independent variables
(r=0.9 and above) (Pallet, 2005). As indicated in above table 4.7 the correlation coefficient
among variables is less than the cut point (0.9) this shows there is no multicollinearity

51
problem. Thus seven independent variables (tax evasion, corruption, employee competence,
organizational strength of tax authority, political instability, technology and tax delinquency)
and three controlling variables (gender, work experience of taxpayers and educational status)
were included in the final model.

The correlation matrix between explanatory variables shows a positive correlation between
tax evasion, corruption, political instability and tax delinquency. On the other hand, there is a
positive relationship between employee competence, organizational strength of tax authority
and technology. However, tax evasion, corruption, political instability and tax delinquency
have a negative relationship with and employee competence, organizational strength of tax
authority and technology.

4.2.2.2 Proportional odd assumption

The proportional odd assumption is fundamental for the ordered logistics model which
indicates the identical effect of each independent variable at each cumulative split of the
ordinal dependent variable. It is known the assumption should be tested if the ordinal model
has to be used properly. PO assumption is frequently rejected when the number of
explanatory variables is large (Brant, 1990). Where the PO assumption is justified ordinal
regression models can be a powerful means of summarizing relationships that utilize all the
information present in the ordinal outcome.

In this study, the proportional odd test was performed to check whether the assumption is
upheld or violated. Brant test was used from Stata V. 16.0 software, and the result suggested
chi-square 20.04 with a degree of freedom 13(p-value 0.094). The null hypothesis of PO
states the location parameters are the same across response categories. A decision rule is to
accept the null hypothesis (PO assumption fit) if the p-value is insignificant at 0.05% and
otherwise reject it. Since the p-value is insignificant, the null hypothesis is failed to reject,
and the PO assumption is upheld.

Additionally, SPSS 20 result for PO indicated in table 4.8, the value of chi-square is 10.826
with a degree of freedom of 10(p-value 0.371) which is insignificant at 0.05. This indicates
PO assumption is upheld, and it was appropriate to analyze the level of tax revenue collection
using OLM.

52
Table 4.8 Test of Parallel Lines
Model -2 Log Likelihood Chi-Square df Sig.
Null Hypothesis 408.181
General 397.355 10.826 10 .371
The null hypothesis states that the location parameters (slope coefficients) are the
same across response categories. Link function: Logit

4.2.2.3 Test for Heteroskedasticity

To make sure standard errors are not distorted and any inferences made could not be
misleading and for the efficiency of the model, the problem of heteroskedasticity has been
detected. The null hypothesis of the test states that the errors are homoskedastic and
independent of the regressors and that there is no problem of misspecification.
Heteroskedasticity problem was tasted using Breusch-Pagan test with the decision rule states
that, if the p-value of the test is lower than any of the chosen significance levels 1%, and 5%
it indicates a possible problem of heteroskedasticity; whereas if the p-value of the test is
greater than any of the chosen significance levels 1%, and 5% indicate the possible absence
heteroskedasticity problem.

The Breusch-Pagan Test result shows chi-square 1.69 with 1 degree of freedom at the chi-
square p-value 0.194. Thus, shows that the chi-square p-value is more than 5%, this implies
the null hypothesis that the residuals are homoskedastic at 1%, and 5% significance level is
failed to reject. Therefore, there is no problem of heteroskedasticity with the residuals of the
model.
4.2.3 Ordered Logistics Model fitness

As stated earlier the OLM was run to determine the relationship between TRC and
explanatory variables. The first result shows the iteration Log which is a listing of the log-
likelihoods for each of the iteration. The first iteration for the iteration of zero is the log-
likelihood of the null or empty model; that is, a model with no predictors. Next to the zero
iteration, the predictors are included in the model. The final iteration was converged to -
202.53195.

The model summary information shows the number of observations of 279 which is the
number of observations undertaken in the ordered logistic regression model. The Likelihood
Ratio (LR) Chi-Square test shows that at least one of the predictors‟ regression coefficients is
not equal to zero in the model. The LR Chi-square is 198.78 which was computed as -2(null

53
model – fitted model)) = - ((-301.92184) - (-202.53195)). Thus, the LR Chi-square (13)
shows that the model fits the data well as compared to the null hypothesis.

The Pro > Chi2 indicates the probability of obtaining these Chi-square statistics (198.78) if
there is, in fact, no effect of the predictor variables. This p-value is compared to a specified
alpha level, this is the willingness to accept type I error, which is typically set at 0.05. The
small p-value from the LR test, <0.0001, would lead to conclude that at least one of the
regression coefficients in the model is not equal to zero. The model chi-square is highly
significant and indicates that explanatory variables have a significant effect on the level of tax
revenue collection.

Additionally, the pseudo-R-square value was 0.3292; this implies the null hypothesis which
states that all the coefficients in the model are equal to zero is rejected. According to
Hensher, et al. (2005) values of Pseudo-R-square between 0.2 and 0.4 are considered
to be good fits; therefore, the overall fit of the ordered logistic model was found to be worthy.
4.2.4 Ordered Logistic Model Interpretations with the Coefficients
When the assumption of proportional odd holds, the coefficients of explanatory variables in
an ordinal logistic regression model are interpreted in terms of the logarithm of the ratio of
the odd of a particular category to the reference category. Interpretation of the parameters
corresponding to the significant variables is presented in this section.

The result of each variable was interpreted using the coefficient interpretation method.
Coefficients were interpreted according to the standard interpretation of the ordered logistic
model coefficient interpretation, which is “as one-unit increase in the predictor, the response
variable level is expected to change by its respective regression coefficients in the ordered
log-odds scale while the other variables in the model held constant”. As indicated table
below (Table 4.9) six explanatory variables (tax evasion, corruption, technology,
organizational strength of tax authority, political instability and tax delinquency) and one
controlling variable (educational level of taxpayers) are statistically significant.

Tax evasion (Teva): This is the ordered log-odds estimate for an increase in tax evasion
activities by taxpayers on the expected level of tax revenue collection given the other
variables are held constant in the model. As indicated in table 4.9 an increase in tax evasion
activities of taxpayers would result in a 0.908 unit decreases in the ordered log-odds of being
in a higher tax revenue collection level while the other variables in the model are held

54
constant. This implies tax evasion has a negative effect on tax revenue collection; which
shows tax collection decrease as more taxpayers evade tax.

Corruption (Corr): it shows, given the other variables are held constant in the model, the
ordered log-odds effect of increases in corruption activities on the expected level of tax
revenue collection. The econometric result suggested an increase in corruption activities of
tax officials would result in a 1.077 unit decreases in the ordered log-odds of being in a
higher tax revenue collection level while the other variables in the model are held constant.
This indicates an increase of corrupted tax officials could lower the level of tax revenue
collection.

Technology (Tech): this is the ordered log-odds effect of an increase in using the latest
technology on the expected tax revenue collection level assuming the other variables are held
constant in the model. The result indicated an increase in using the latest technology would
result in a 0.338 unit increase in the ordered log-odds of being in a higher tax revenue
collection level while the other variables in the model are held constant. This shows the use
of the latest technology increases the level of tax revenue collection.

The organizational strength of tax authority (Tautho): assuming the other variables are
held constant in the model, it is the ordered log-odds estimate for an increase in
organizational strength of tax authority on the expected tax revenue collection level. The
econometric result revealed an increase in organizational strength of tax authority would
result in a 0.344 unit increase in the ordered log-odds of being in a higher tax revenue
collection level while the other variables in the model are held constant. This implies the tax
collection would be improved when the tax authority sustained strength in all dimensions.

Political instability (PIs): it shows, given the other variables are held constant in the model,
the ordered log-odds effect of increases in political instability on the expected level of tax
revenue collection. From table 4.9 an increase in level political instability would result in a
0.683 unit decreases in the ordered log-odds of being in a higher tax revenue collection level
while the other variables in the model are held constant. From this, it can be said political
instability has negative implications on tax collection, and as instability increases in the
country, the probability of collecting the tax would be low.

Tax delinquency (Tdel): it is the ordered log-odds estimate for an increase in tax delinquent
activities of taxpayers on the expected level of tax revenue collection given the other
variables are held constant in the model. The result shows an increase in tax delinquent

55
activities of taxpayers would result in a 0.655 unit decreases in the ordered log-odds of being
in a higher tax revenue collection level while the other variables in the model are held
constant. This implies as taxpayers delinquent frequently, the probability of collecting low
tax revenue would increase.

Educational level of taxpayers (Edu): Educational level of taxpayers is a controlling

variable that affects tax collection. As a one-unit increase in the educational level of
taxpayers (from diploma to First degree and above), tax revenue collection level is expected
to change to a higher level by 1.78 units in the ordered log-odds scale while the other
variables in the model held constant. This implies as the taxpayer‟s educational level
increases, the probability of collecting the tax would increase.

The marginal effects (Appendix III) indicated that the probabilities of tax revenue collection
to be becoming one of the three-level: low (0), medium (1) or high (2) would be indicated as
.17326681, .64677145 and .17996174 respectively. This implies the most likely the tax
revenue collection was at a medium level. Thus, as a result of variables included in the
model, the probability of collecting a medium level of tax revenue has occurred.

Table 4.9 Summary of parameter Estimates

TRC Coef. Std. Err. z P>|z| [95% Conf. Interval]

Teva -.908847 .1762669 -5.16 0.000 -1.254324 -.5633702

Corr -1.077311 .1591089 -6.77 0.000 -1.389159 -.7654634
Epl .1292307 .1465243 0.88 0.378 -.1579517 .4164131
Tech .3379584 .1093983 3.09 0.002 .1235416 .5523752
Tautho .344311 .1364022 2.52 0.012 .0769676 .6116545
PIs -.6834582 .144968 -4.71 0.000 -.9675902 -.3993262
Tdel -.6547667 .1616769 -4.05 0.000 -.9716477 -.3378858

Gender
Male -.0319387 .2729987 -0.12 0.907 -.5670063 .5031289

WExp
6-10 -.276087 .3009228 -0.92 0.359 -.8658849 .3137109
11-15 .0474589 .4166884 0.11 0.909 -.7692354 .8641533
Above 16 .527114 .5402355 0.98 0.329 -.5317281 1.585956

Edu
Diploma .3780533 .3835455 0.99 0.324 -.3736822 1.129789
First degree and above 1.779961 .6714558 2.65 0.008 .4639316 3.09599

LR chi2(13) = 198.78 Pseudo R2 = 0.3292

Prob. > chi2 = 0.0000 Likelihood Ratio = -202.53195

Source: Computed from own survey data, 2021

56
4.3 Results and Discussion
Determinants of tax revenue collection (TRC) were analyzed using ordered logistics
regression in the above section. The assumptions of OLM are fitted to investigate
determinants of tax revenue collection. This section presents results and discusses the effects
of explanatory and controlling variables on TRC, and hypotheses were tested in
interconnectedness with the econometric result.

Explanatory variables

Tax evasion is one of the determinants of tax revenue collection covered in this study.
Accordingly, the econometric result shows tax evasion has a negative and significant (p-
value<0.01) effect on tax revenue collection. Thus this result is consistent with the hypothesis
of the study.

H1: Tax evasion has a negative and significant effect on tax revenue collection.

Therefore, the null hypothesis is rejected, because its p-value is in the rejection region. This
implies tax evasion has a statistically significant impact on tax revenue collection. The result
is consistent with the finding of Adeniyi and Folayan (2018); Eja, et al. (2018); and Tumoro,
(2020) who concluded the final effect of tax evasion on revenue generation is to results in
revenue loss. Thus, from this, it can be said the expected number of tax revenue may no be
collected from both formal and informal economies due to the weak enforcement
mechanisms on tax evasion. However, the finding of Abate (2019) and Manaye, et al.,
(2019) showed tax evasion has a negative and insignificant effect on tax collection.

The marginal effect (table 4.10) shows an increased level of tax evasion activities by
taxpayers, being other factors held constant, the likelihood of tax revenue collection occurs in
low tax collection level increases by about 12% while it decreases the probability collecting
medium and high tax level by about 2% and 9% respectively. This denotes that if taxpayers
evade more tax, the probability of collecting lower tax revenue is anticipated.

As Tax evasion is a violation of the law, it affects the economy of the country and law
negatively. Taxpayers were evading tax by hiding their actual income while filling tax returns
and undermining their business income while assessed. Additionally, there is less probability
that tax evaded taxpayers were punished, and this may open more ways or loopholes for high
tax evasion. This is a clue in that tax fines and penalties are discouragement for potential tax
evasion. Generally, this study finding is consistent with tax evasion theory, which indicated
tax evasion reduces the tax to be collected.

57
An interview result also indicates the existence of tax evasion activities. Although it is illegal
to evade tax, in oppose to the socio-political theory which states sum is better than individual,
evaders see individual advantage only. The summary of an interview on tax evasion indicated
here under:

Tax evaders use two main ways to refrain from paying the required amount of tax. Firstly,
they undermine the actual income earned while preparing statements if they are category „A‟
and „B‟ taxpayers. It can be accomplished through undermine statements: hiding the sale‟s
invoice and show exaggerated purchase invoice so that they can not be coughed by the
auditor. Category „C‟ taxpayers also evade tax by providing false information on their daily
income while they are assessed. Secondly, even though they are required to do, they perform
selling without using invoice.

Corruption is unlawful or unauthorized action performed by public tax officials using their
position to receive a bribe in exchange for making a benefit available to a member of a
specific public group particularly taxpayers. The hypothesis of the study was stated as
follows.

H1: Corruption has a negative and significant effect on tax revenue collection.

The econometric result indicated that corruption has a negative and significant (p-value<0.01)
effect on tax revenue collection. This finding supports the study hypothesis. Thus, the null
hypothesis is rejected, because its p-value is in the rejection region. This implies corruption
has a statistically significant impact on tax revenue collection. This result is in line with the
finding of Amin, et al. (2014); Abate (2019) and Adeniyi and Folayan (2018) as corrupted tax
officials increase, there is a probability that tax collection decrease. Additionally, corruption
has multiple effects on tax collection: firstly, it directly reduces tax collection. Secondly,
since, it is positively correlated with tax evasion, it creates loopholes for tax evasion.

The level of tax revenue collection decline may because of corrupted tax officials. Tax
officials offer a reduction of tax payment (tax liability) for taxpayers by receiving bribes. Tax
officials commit corruption, it is not surprising the probability that they are punished is low
which may be derived from institutional weakness. It is obvious corruption gives benefit to
the people who commit it, but it affects TRC negatively and thereby the economy of the
country as a whole.

The marginal effect shows that being other factors held constant, an increase in the level of
corruption activities by tax officials, the probability of collecting low level of tax revenue

58
increase by about 14.2% whereas the likelihood of collecting medium and high level of tax
revenue collection decrease by about 2.7% and 11.5% respectively. The implication is that if
corruption activity mainly by corrupted tax officials increases, the TRC probability of being
at a low level is realized.

An interview result also confirms the existence of corruption and its impact on tax collection.
In the tax collection corruption is mainly engaged by tax assessors. As tax assessors
independently assess the tax of category „C‟ taxpayers, there would be a probability in which
they receive a bribe instead of collecting full tax. This has happened because it is difficult to
control them unless they are accountable.

Another important determinant of tax revenue collection is employee competence.

Employee‟s competence is the skill of tax staff on mobilizing revenue from appropriate
sources the appropriate amount of tax revenue.

H1: Employee competence has a positive and significant effect on tax revenue collection.

This study found that employee qualification has a positive and insignificant (p>0.05) effect
on tax revenue collection. This result is consistent with the finding of Mugambi (2018) and
Eja, et al. (2018). Therefore, the null hypothesis is failed to rejected, because its p-value is
not in the rejection region. This implies the employee competence has a statistically
insignificant impact on tax revenue collection.

The summary of an interviewee on employee competence is indicated as next: Tax

authority provides training to improve employee competence. The kinds of training include
audit and assessment, cash handling and tax collecting. However, employees have a
problem regarding the application of competence. For instance, as there is no protection for
auditors, they may not do what they are expected to and can do. This implies even though
employees have competence, they are not applying in tax collection; thus, the effect of
employee competence is insignificant in tax collection.

Technology is another determinant of tax revenue collection. In taxation, technology is the

use of e-taxation which reduces a paper movement rather advanced with internet technology.

H1: Technology has a positive and significant effect on tax revenue collection.

The econometric result shows the technology has a positive and significant effect (p<0.01) on
tax revenue collection. The result was consistent with the study hypothesis. Thus, the null
hypothesis is rejected, because its p-value is in the rejection region. This implies technology

59
has a statistically significant impact on tax revenue collection. The finding is consistent with
Mugambi, (2018); Mohammed and Muturi, (2018) and Ngicuru (2017) who stated the usage
of the latest technology is positively related to TRC. TRC appeared to be low as a result of
the manual system in providing service for taxpayers. Although tax collection practice was
started in Ethiopia long years ago, tax revenue collection is still low. This may happen
because of working with the manual system. Therefore, the advanced mode of tax collection
affects TRC positively in which this result was different from the finding of Abate (2019) on
the mode of tax collection, who stated mode of tax collection has a negative and insignificant
effect on tax collection.

In terms of marginal effect as shown in table 4.10, other factors held constant, in a change in
the use of the latest technology, the probability of tax revenue collection occurs in the low
level decreases by 4.4% while the probability of medium and high level of tax revenue
collection increases by 0.8% and 3.6% respectively. This connotes that if tax revenue were
collected using the latest technology, the probability of collecting high tax revenue is
upsurged.

An interview on the effect of technology on tax revenue collection is indicated as the

following: The current mode of tax collection or technology is not suitable for effective tax
collection. The tax collection system is manual. Whenever a taxpayer makes tax payment
tax he/she must present physically which may induce a negative attitude in taxpayers on
tax. This can reduce the amount of tax to be collected and require a long time which causes
tax collection not completed on time.

The organizational strength of tax authority is another determinant of tax revenue collection,
which indicates tax authority capacity in terms of competent staff, technology, and
enforcement mechanisms in respect of tax collection.

H1: Organizational strength of tax authority has a positive and significant effect on tax
revenue collection.

In line with Abate (2019) and Ngicuru (2017,) the econometric result implied there is a
positive and significant (p<0.05) relationship between the organizational strength of tax
authority and tax revenue collection. The finding supported the hypothesis of the study. Thus,
the null hypothesis is rejected, because its p-value is in the rejection region. This implies
organizational strength of tax authority has a statistically significant effect on tax revenue
collection.

60
Possessing organizational strength, in terms of adequate human and technological resources,
delivering quality tax services to taxpayers and possessing strong law enforcement potential
were among the factors for organizational strength of tax authority which in turn affects tax
collection.

The marginal effect indicated that being other factors held constant, an increase in
organizational strength of tax authority decreases the likelihood of low level of tax revenue
collection on average by about 4.5% while it increases the probability of collecting medium
and high level of tax revenue on average by 0.8% and 3.6% respectively. This is consistent
with the study of Abate (2019) stated organizational strength of tax authority affects tax
collection positively.

Regarding the organizational strength of tax authority, the interview result shows tax
authority has a human resource that enables to collect tax, but not technology. Even though
the human resource is not more than ample, the main problem with the tax authority is not a
human resource. The problem is assigning the wrong person on assessing tax amount which
especially works for category „C‟ taxpayers. Moreover, due to the existence of corruption and
inappropriate positioning, human resource is not utilized in collecting revenue. Tax authority
use enforcement mechanisms such as closing delinquent taxpayers' businesses until they
made tax payments and performing the audit of statements.

Another determinant of tax revenue collection is political instability which may relate to
violent conflicts of low stability.

H1: Political instability has a negative and significant effect on tax revenue collection.

Econometrics result suggests that political instability has a negative and significant impact
(p<0.01) on tax revenue collection. It supported the study hypothesis. Therefore, the null
hypothesis is rejected, because its p-value is in the rejection region. This implies political
instability has a statistically significant effect on tax revenue collection. The result is
consistent with the study of Wayau (2017) and Amin, et al., (2014); in that political
instability reduces tax revenue collection. During political instability, taxpayers have to stop
business oppression. This in turn reduces their income and thereby reducing tax revenue
collection.

To be specific within each level of TRC, marginal effect indicated that an increase in political
instability increases the probability of collecting a low level of tax revenue on average by 9%
while it decreases the likelihood of collecting medium and high level of tax revenue

61
collection by 1.7% and 7.3% respectively. This implies if the government would works on
reducing political instability and maintaining stability, the probability of collecting higher tax
revenue is achieved.

An interview on the effect of political instability shows: political instability is disastrous for
tax collection. There is a possibility of losing a high source of tax as business canter may
damage. Further, during political instability, illegal activities like tax evasion would
increases. It is almost difficult to tax category „C‟ as their taxation is based on assessing daily
income in which political instability creates a chance of inappropriately assessing their actual
income.

The final main explanatory variable is tax delinquency, and it is a tax not paid within the due
date.

H1: Tax delinquency has a negative and significant effect on tax revenue collection.

The study found tax delinquency has a negative and significant effect (p<0.01) on tax
revenue collection. This finding supported the hypothesis of the study. Thus, the null
hypothesis is rejected, because its p-value is in the rejection region. This implies tax
delinquency has a statistically significant effect on tax revenue collection. Up to this study is
conducted no study considered tax delinquency variable as determinants of tax revenue
collection. However, the study of Jang & Eger (2019) revealed tax delinquency can be seen
as a factor that shifts honest taxpayers toward non-compliance. Further another study
revealed tax compliance enhances revenue collection Mohammed & Muturi (2018). This
indicates that if taxpayers are delinquent, the probability of tax compliance is low, which in
turn it reduces the level of tax revenue collection. Therefore, using the finding of Mohammed
& Muturi (2018) as a bridge the result of this study supported that tax delinquency reduces
tax revenue collection significantly.

Based on items included in the data collection instrument some of the factors for tax
delinquency were low penalty if delinquent and less information on tax due date. Taxpayers
may not pay tax before the due date because of the low penalty. This implies if the penalty
were strong, they would pay the tax instead of being delinquent.

The marginal effect indicated that being other factors held constant, an increase in tax
delinquency increases the likelihood of tax revenue collection being in the low level on
average by about 8.6% while it decreases the probability of tax revenue collection being in
the medium and high level of TRC on average by 1.6% and 7% respectively.

62
An interview regarding to tax delinquency also shows as it affects TRC as next. Some
taxpayers do not pay tax on time. They fail to pay tax before the due date because of less
awareness of the tax due date and carelessness. It affects tax planned to be collected which
creates underperformance in tax collection. Particularly tax delinquency is high in the area of
rent and lease as they do not pay on time.

Control variables

As stated earlier in addition to the explanatory variables discussed, three control variables
such as gender, work experience of taxpayers and educational status were also included in
this study. Although the objective of this study was not to detect the effect of these
controlling variables on TRC, it was used to enhance the validity of the study by limiting the
influence of explanatory variables.

Regarding the educational status of taxpayers econometric results suggest if a taxpayer is the
holder of the first degree and above, it has a positive and significant (p<0.01) impact on tax
revenue collection. The marginal effect shows that being other factors held constant, in a one-
level change in the education of taxpayers that is from diploma to first degree and above, the
probability of collecting a low level tax revenue decreases by 20% while the probability of
collecting high TRC increases by 22.6%. This implies that the higher the education level
taxpayers have, the higher the probability of tax revenue collection being in higher tax
revenue collection level. Thus, as taxpayers educational level increase tax authority can
collect more tax revenue.

63
Table 4.10 Marginal effects on the probabilities of the level of tax revenue collection

Level of tax revenue collection

Low Medium High
Variables Dydx Std. Err Dydx Std. Err dydx Std. Err
Teva .120* .021 -.023* .008 -.097* .017
Corr .142* .018 -.027* .009 -.115* .014
Epl -.017 .019 .003 .004 .014 .015
Tech -.044* .013 .008** .003 .036* .011
Tautho -.045* .017 .008** .004 .036** .014
PIs .090* .017 -.017* .005 -.073* .015
Tdel .086* .021 -.016** .007 -.070* .016
Gender
1=Male .004 .036 -.0008 .0068 -.0034 .029

WExp
2=6-10 .037 .040 -.008 .009 -.029 .031
3=11-15 -.006 .054 .001 .008 .005 .046
4=Above 16 -.065 .064 .003 .006 .062 .066
Edu
2=Diploma -.051 .050 .0098 .008 .041 .042
3=First degree -.200* .056 -.026 .042 .226** .095
and above
Source: Computed from own survey data, 2021
Note: * and ** represent significant at 1% and 5% of significance level respectively

64
 CHAPTER FIVE
SUMMARY, CONCLUSIONS AND RECOMMENDATIONS
In this chapter, the conclusions derived from the finding of this study on determinants of tax
revenue collection in Sululta town are described. The conclusions were based on the purpose
and hypothesis of the study. Based on the implications of these conclusions the resultant
recommendations are also presented.

5.1 Summary of major findings

The determinants of tax revenue collection were examined using the ordered logistic model
and an interview was also used. Tax evasion has a negative effect on tax revenue collection
which causes a reduction in tax collection. Further, corruption has a double effect on tax
collection: in one way it directly reduces the amount of tax expected to be collected, and in a
second way it creates a loophole for tax evasion. On the other hand, the use of the latest
technology increase the amount of tax expected to be collected. Likewise, technology
organizational strength of tax authority also increases tax revenue collection. Political
instability has a negative effect on tax revenue collection. Tax is not collected during political
instability as it causes taxpayers to stop their business operations. Tax delinquency decreases
tax revenue expected to be collected. Additionally, the educational status of taxpayers has a
positive effect on tax revenue collection. An increase in the educational status of taxpayers
opens ways for the tax authority to collect more tax revenue.

5.2 Conclusions
In line with most other developing countries, tax revenue collection remains a major problem
of resource mobilization in Ethiopia. In Sululta town, which is one of Oromia's special zone
surrounding Finfinne (Addis Ababa), several business activities are performing which are the
main sources of business income tax for the government and country growth. However, this
study found that tax revenue collection is found at a medium level as various factors hinder
the collection of the required amount of tax revenue.

Ethiopia meets goals regarding infrastructure, defense and maintains rules using both external
debt and internally mobilized resources. Nevertheless, as external debt is costly, domestic
resource mobilization is critically important for sustainable financing of development plans.
Thus, various studies were performed to identify determinants of tax revenue, but most
studies were focused on the effect of microeconomic variables such as GDP, inflation, trade

65
openness, foreign direct investment, and economic sectors such as agricultural, service, and
industry sector on tax revenue. However, there are few studies on the effect of social,
administrative, institutional or organizational and taxpayer‟s demographic characteristics on
tax revenue collection. In particular, up to this study, there is no comprehensive determinant
of tax revenue collection study on the business income tax in Sululta town.

The main purpose of this study is to investigate the determinant of tax revenue collection in
Sululta town. To meet this objective, primary data was collected from taxpayers through
questionnaires, and from employees such as managers, auditors and tax collectors through
interviews.

Drawing on data from Sululta town in Oromia national regional state and assuming all
categories of taxpayers (category „A,‟ „B‟, „C‟), this study has investigated the impact of
administrative, organizational or institutional and demographic determinants of tax revenue
collection using ordered logistic model. Tax revenue collection was categorized into three
levels which were low, medium and high.

The result of the study suggested the level of tax evasion activities by taxpayers, activities of
corrupted tax officials, the inconveniences of mode or technologies used to collect tax, the
capacity or organizational strength of tax authority on collecting tax, the existence of political
instability and tax-delinquent are potentially detrimental to the level of tax revenue
collection. As a result of these potential determinants, the Sululta town tax revenue office
collects a medium level of tax revenue.

The major conclusions of the study are:

 Tax evasion is found to increase the probability of collecting a low level of tax revenue
collection. However, it decreases the likelihood of collecting a high level of tax revenue.
Taxpayers evade tax by performing a transaction without an invoice and providing
misstated information for the tax authority.
 Corruption has an important effect on the probability of decreasing the high tax revenue
collection level, but it has the effect of increasing the probability of collecting a low level
of tax revenue. That is, it decreases tax revenue collection.
 The use of the latest technology in collecting tax revenue is found to increase the
likelihood of collecting a high level of tax revenue, but it has the effect of decreasing the
probability of collecting a low level of tax revenue.

66
 The organizational strength of tax authority is found to have a significant impact on
increasing the probability of collecting high tax revenue. However, it significantly
decreases the probability of collecting a low level of tax revenue. It is also found that the
main factors for organizational strength tax authority are possession of technological and
human resources and ways of its enforcement mechanisms.
 Political instability is also another determinant of tax revue collection. As the level of
political instability increases the probability of collecting a low level of tax revenue
would increase. On contrary, political instability decreases the likelihood of collecting a
high level of tax revenue.
 Tax delinquency is found to be another factor that determines tax collection negatively
and significantly. As taxpayers being delinquent increases, the probability of collecting
tax revenue occurs at a low level is anticipated. However, if tax delinquency decrease, the
probability of collecting a high level tax revenue is would occur.

5.3 Recommendations
Reducing or eliminating the determinant that negatively affects tax revenue collection
requires addressing its root of sources like the weak perception of taxpayers on tax evasion
and delinquency, manual mode of the tax collecting system, the weak tax authority in terms
of technology and frequent occurring of political instability. Based on the major finding, the
following policy implication recommendations have been provided.

Firstly, the Ethiopian government, particularly, ministry of education, must focus on

providing education for citizens on tax knowledge from the elementary class. This has two-
dimensional advantages. Firstly, if citizens are well aware of tax law, they may not receive
goods and service without legal receipt provided by the tax authority to taxpayers, which
facilitate tax audit and reduce tax evasion. Secondly, educated taxpayers feel a sense of civic
responsibility and pay taxes properly. Thus, creating awareness for taxpayers can increase the
probability of having a positive attitude toward paying tax so that tax evasion and tax
delinquency may avoid. Additionally, educated citizen feels confident about the amount of
tax they pay and feel a positive sense of fulfilling their social responsibility instead of giving
bribe for reducing tax liability.

Secondly, tax authority must employ serious enforcement mechanisms such as audits and
strong penalties, which require tax authority to develop the latest technology and employee
competence. Penalizing tax evaders has likely to create a negative sense of tax evasion in

67
others taxpayers, so they will not evade tax anymore. However, to change the perceptions of
taxpayers, the penalty should be strong enough. This is work especially for big business
owners who may prefer paying penalties by evading more tax. Thus, introducing a strong
penalty or improving penalty is not the responsibility to be left for tax authority alone, but it
requires Ethiopian Revenue and Custom Authority (ERCA) involvement.

Thirdly, the government should strengthen tax authority by providing the latest technology
that enables them to collect tax efficiently and effectively. This can be accomplished by
adopting an automated tax collection system from developed or any other developing
countries. Alternatively, the government can develop its own automated tax revenue
collection system domestically. Further, tax authority should also have a clear organizational
structure and avoid inappropriate assignment of individual without his/her position qualified
for. This can enable to improve organizational strength of tax authority.

Fourthly, the government should work on bringing and maintaining stability so that tax
collection will not be affected by political instability. This can be accomplished by creating
citizens with full social responsibility and citizens who love their country. Additionally,
people's knowledge about the purpose of tax toward the country enables them to be
independent of the political party, thereby opening ways for the tax authority to collect more
tax revenue. Therefore, to create patriot citizens the government should be accountable for
tax revenue collected and on how it is used, and education on tax should be given to citizens
both formally (in school) and informally (radio, television and programs on awareness
creation).

Finally, the tax authority must provide adequate information for taxpayers about the due date
of tax payment to reduce delinquent tax. Taxpayers should know about the due date of tax
payment and its penalty if they fail to pay on time. Strong penalty for delinquent taxpayers
and timely available information on due date has a probability of reducing tax delinquent.
Thus, tax authority should use various means such as pamphlets, on television and radio for
giving information about tax payment due date.

5.4 Future Research Direction

Even though various aspects of tax revenue collection are covered in this study, several other
issues should be studied further. To improve the literature in the area of tax revenue
collection, future research should be conducted on determinants of tax revenue collection on
each category of taxpayers („A‟, „B‟ and „C‟) separately. Undertaking a study on each

68
category of taxpayers enables us to investigate the significant effect of each variable on each
taxpayer separately. This is important because not all variables necessary have an equal effect
on categories of taxpayers.

As this study was limited to Sululta town; it is also important to further consider the study to
the country as a whole or at each region and zone. Concluding a study from one or a few
towns to the country may lead to hasty generalization. Thus, it adds more literature if the
study is extended to other areas.

There are several factors that affect tax revenue collection. In this study, seven explanatory
and three control variables were included in the model, but tax revenue collection is not
determined by these factors only. It is important to consider the effect of other social,
administrative, organizational and institutional factors and demographic factors such as
location of tax authority, source of tax, age of taxpayers, religion and level of income.

Even though it is costly to collect data by modifying data collection techniques such as an
interview with taxpayers, it can help to identify determinates of revenue collection. An
interview with taxpayers has the power to reveal or identify new variables that determine tax
revenue collection.

69
 Reference
Abate, A. A., 2019. Factors affecting presumptive tax collection in Ethiopia: evidence from
category “C” taxpayers in Bahir Dar city. Journal of Tax Administration, 5(2), pp.
74-96.

Acosta, M. S. & Holbrook, A., 2008. Some notes on theories of technology, society and
innovation systems for science and technology policy studies, Vancouver BC: s.n.

Adams, M., 2012. Agency Theory and the International Audit. Managerial Auditing Journal,
9(8), pp. 8-12.

Adeniyi , A. G. & Folayan , D. O., 2018. Effects of Tax Evasion on Government Revenue
Generation in Oyo State, Nigeria. European Journal of Accounting, Auditing and
Finance Research, 6(1), pp. 76-89.

Agarwal, R. C., 2003. Public Finance, “Theory & practice” without of date. s.l.:Published by
Lakshmi Narain Agarwal.

Ahmed, E. & Stren, N., 1989. Handbook of Development Economics. In: H. C. a. T.

Srinivasan, ed. Taxation for developing countries. London: Elsevier Science
Publishers B.V, pp. 1006-1087.

Ahmed, N. et al., 2012. Improving Tax Compliance in Developing Economies: Evidence from
Bangladesh, s.l.: s.n.

Aisen, A. & Veiga, F. J., 2011. How Does Political Instability Affect Economic Growth?.
IMF Working Paper .

Akitoby, B., 2018. Raising Revenue. Finance and development, 55(1).

Amin, A., Nadeem, A. & Majeed , . P., 2014. Factors Affecting Tax Collection in Pakistan:
An Empirical Investigation. Journal of Finance and Economic, 2(5), pp. 149-155.

Ataro, P. . O., Muturi, W. & Wandera, R. W., 2016. Factors Affecting Revenue Collection
Efficiency in County Governments in Kenya: A Case Study of Trans-Nzoia County.
International Journal of Recent Research in Commerce Economics and Management
(IJRRCEM), 3(2), pp. 85-90.

Basariya, R., Arab, H. R. & Ahmed, R. R., 2019. A STUDY ON THE USES OF
TAXATION IN DEVELOPING COUNTRIES. Journal of critical reviews, 6(6), pp.
368-375.

Bayu, T., 2015. Analysis of Tax Buoyancy and Its Determinants in Ethiopia (Cointegration
Approach). Journal of Economics and Sustainable Development, 6(3).

Bhartia, H. L., 2009. Public Finance. 14th ed. New Delhi: Vikas Publishing House PVT Ltd..

70
Chamber of Commerce, A. A., September,1996. History of tax in Ethiopia. Addis Ababa,
unpublished Handout.

Chaudhry, I. S. & Munir , F., 2010. Determinants of Low Tax Revenue in Pakistan. Pakistan
Journal of Social Sciences (PJSS), 30(2), pp. 439-452.

Chigbu, E., Eze, A. L. & Ebimobowei, A., 2012. An Empirical Study on the Causality
between Economic Growth and Taxation in Nigeria. Current Research Journal of
Economic Theory, 4(2), pp. 29-38.

Chinyere, C. J., 2002. The Use of Consultants in Tax collection Analysis of the Accelerated
Revenue Generation Programme in River State: Nsukka.

Cochran, W., 1963. Sampling Techniques. 2nd Ed. ed. New York: John Wiley and Sons, : s.n.

Cooper, D. R. & Schindler, P. S., 2014. Business Research methods. 12th ed. New York:
McGraw-Hill/Irwin.

Cooper, S. G., 1994. The benefit theory of taxation. Australia: s.n.

Davis, J., 1989. Mainstreaming social budgeting into the budgetary process in Kenya
KIPPRA Discussion Paper. s.l., s.n., p. No 57.

Dhaba, G., 2017. Factors Affecting Rental Income Tax Payers Compliance with Tax System :
In Case of Hawassa City Administration , SNNRS , Ethiopia. Research Journal Of
Finance And Accounting, 8(9), p. 75–83.

Due & Fitzgerald, J., 1999. Government Finance: Economics of the Public Sector(Irwin
Series in Economics). s.l.:Richard d Irwin.

Eckhardt, P., 2016. Adam Smith‟s View on the Functions of Taxation. 7(29).

Eja, B. R., Indika, S. E. & Ogar, A., 2018. Inhibiting factors to tax revenue generation in
cross river state, nigeria. Global journal of social sciences,, 17(1), pp. 11-19.

Ergo, A. E., 2018. Performance Determinants of Business Income Tax Collection: A Study
Conducted in the Boditi Town Revenue Collecting Branch Office, Wolaita Zone;
SNNPR,Ethipia. International Journal of Business and Management Invention
(IJBMI), 7(8), pp. 35-46.

Ergo, A. E. E. E., 2018. Performance Determinants of Business Income Tax Collection: A

Study Conducted in the Boditi Town Revenue Collecting Branch Office, Wolaita
Zone; SNNPR,Ethipia. International Journal of Business and Management Invention
(IJBMI), 7(8), pp. 35-46.

Eugene , S. A., Abiire, M. A., Atindaana, P. A. & Ayimpoya, R. N., 2020. Tax Education
And Tax Compliance In Ghana. Journal of Accounting, 8(1), p. 2020.

Eugene , S. C., 1975. “And Equal (Tax) Justice for All?”. s.l.: Urban Institute Press.

71
Fairchild, F. R., 1934. The Problem of Tax Delinquency. The American Economic Review,
24(1), pp. 140-150.

Garner, B. A., 1999. Black's Law Dictionary. 7th ed. s.l.:West Group.

Getinet, B. & Ersumo, F., 2020. The Impact of Public External Debt on Economic Growth in
Ethiopia: The ARDL Approach to Co-integration. Journal of Economics and
Sustainable Development , 11(11), pp. 25-39.

Gituma, H. K., 2017. Determinants of effective revenue collection by Embu county Kenya.
Thesis paper.

Gobachew, N., Debela, K. L. & Shibiru, W., 2018. Determinants of Tax Revenue in Ethiopia.
Economics, 6(1), pp. 58-64.

Gupta, A. S., 2007. Determinants of Tax Revenue Efforts in developing country. IMF
Working paper.

Harrison, K., 2007. Corruption and tax compliance: challenges to tax policy and
administration. Sofia, Bulgaria: Centre for the Study of Democracy, s.n.

Hauer, T., 2017. Technological determinism and new media. International Journal of
English, Literature and Social Science, 2(2), pp. 1-4.

Helmke, G. & Levitsky, S., 2004. Informal institutions and comparative politics: a research
agenda.. Perspectives on Politics, 2(4), pp. 725-740.

Hensher, D., Rose, J. & Greene, W., 2005. Applied choice analysis: a primer. New York,
USA: Cambridge university press.

Imam , P. A. & Jacobs, D. F., 2007. Effect of Corruption on Tax Revenues in the Middle
East. IMF Working Paper, pp. 1-34.

Income tax proclamation , T. E., No.979/2016. s.l., s.n.

International Monetary Fund, 2020. 2019 article IV consultation and requests for three-year
arrangement under the extended credit facility and an arrangement under the
extended fund facility—press release and staff report, Washington, D.C.: s.n.

Jacob, B., 2017. Digital revolutions in public finance. Washington, DC: s.n.

Jang, S. & Eger, R. J., 2019. The Effects of State Delinquent Tax Collection Outsourcing on
Administrative Effectiveness, Efficiency, and Procedural Fairness. American Review
of Public Administration, 49(2), p. 236–251.

Jensen, M. C. & Meckling, W. H., 1976. Theory of the firm managerial behavior, agency
costs and ownership structure. Journal of Financial Economics, Volume 3, pp. 305-
360.

72
Jeza, M. T., Hassen, A. A. & Ramakrishna, G., 2016. The Relationship between FDI flows
and Tax Revenues in Ethiopia: an Evidence based on ARDL Model with a Structural
Break. Paper presented in International Conference on Ethiopian Economy,
organized by EEA during 21-23 July 2016, Addis Ababa, Ethiopia, s.n.

Johnson, R. & Onwuegbuzie, A., 2004. Mixed methods research: A research paradigm whose
time has come. Educational Researcher, 33(7), pp. 14-26.

Joshi, A., Kale, S., Chandel, S. & Pal, D. K., 2015. Likert Scale: Explored and Explained.
British Journal of Applied Science & Technology, 7(4), pp. 396-403.

Kelvin, G., 2017. The Impact Of System Automation On Revenue Collection in Kenya
Revenue Authority. (A Case Study of SIMBA). [Online]
Available at: https://mpra.ub.uni-muenchen.de/80343/
[Accessed December 2020].

Kibret, H. & Mamuye, R., 2016. Performance and Prospects of Tax Collection in Ethiopia.
Working Paper.

Kothari, C, 2004. Research Methodology Methods and Techniques. 2nd ed. New Delhi: New
Age International publisher.

Kulkami, M., 2001. Business Taxation and Auditing. s.l.:s.n.

Kumari, D. T. H., 2017. The determinants of tax revenue in Ethiopia. Journal of Marketing
Strategy (JMS), 5(3), pp. 198-209.

Liu, X., 2009. Ordinal regression analysis: Fitting the proportional odds model using Stata,
SAS and SPSS. Journal of Modern Applied Statistical Methods, 8(2), p. 632–645.

Manaye, M. K., Patnaik, B. C. M. & Satpa, I., 2019. Determnants of Revenue Collection
Efficiency and Administration: The Case of Local Government Authorties in Wolaita
Zone, Ethiopia.. International Journal of Engineering and Advanced Technology
(IJEAT), 8(6), p. n.

Mardiana, H. & Daniels, H. K., 2019. Technological Determinism, New Literacies and
Learning Process And The Impact Towards Future Learning. Journal of Educational
Science and Technology, 5(3), pp. 219-229.

Masarirambi, C., 2013. An Investigation into Factors Associated with Tax evasion in
Zimbabwe Informal Sector: A survey of MbareMagaba Informal Traders. PhD
Dissertation.

Mascagni , G. & Santoro, F., 2018. Why is tax education important for compliance?
Summary of African Tax Administration Working Paper 1 ICTD. Issue 23.

Mikesell, J. L., 1976. Property market dynamics, local economies, and tax delinquency. State
and Local Government Review, 8(2), pp. 41-45.

73
Mohammed, D. & Muturi, W., 2018. Factors Affecting Revenue Collection Efficiency in
count Government, Kenya. International Journal of Social Sciences and Information
Technology, 4(9), pp. 196-205.

Mood , C., 2010. Logistic Regression: Why We Cannot Do What We Think We Can Do, and
What We Can Do About It. European Sociological Review, Volume 26, pp. 67-82.

Mtasiwa, A. M. J., 2013. Factors causing inefficiency on tax revenue collection in Tanzania.
Masters thesis.

Mugambi, F. K., 2018. Factors affecting implementation of revenue collection systems in

county governments in Kenya, a case of Meru county. International Journal For
Research In Business, Management And Accounting, 4(8), pp. 9-30.

Ngicuru, P. N., 2017. An empirical review of factors affecting revenue collection in Nairobi
county, Kenya. International Journal of Economics, Commerce and Management,
5(8), pp. 324-359.

Ngugi, J. N., 2016. Factors Influencing Optimal Revenue Collection in County Governments
in Kenya:A Case of Kiambu County Government. International Journal of
Innovations, Business and Management (IJIBM) , 10(4), p. n.

Nulty, K., 2008. Research Methodology: Methods and Techniques. London, UK:INTRAC:
s.n.

Nunnally, J., 1978. Psychometric theory. 2nd ed. ed. New York: McGraw-Hill.

Nut, . A. C. & Nut, . F. . M., 2020. Modelling the Influences of Economic, Demographic, and
Institutional Factors on Fiscal Pressure. [Online]
Available at: doi:10.3390/su12041681 www.mdpi.com/journal/sustainability

OECD Committee of Fiscal Affairs Forum on Strategic Management, 2001. Principles of

Good Tax Administration, s.l.: s.n.

OECD, 2013. Policy framework for investment user’s toolkit. [Online]

Available at: www.oecd.org/investment/pfitoolkit

Old , M. O. & Bird, R. M., 1967. Taxation in developing countries. Revised Edition ed. USA:
The John Hopkins press.

Oromia national regional state revenue authority , 2018. [Online]

Available at: http://www.oromiyaa.gov.et

Palil, M., 2010. ax Knowledge and Tax Compliance Determinants in Self Assessment System
in Malaysia. thesis submitted to The University of Birmingham for the degree of
Doctor of Philosophy.

Pallet, J., 2005. SPSS survival manual: A step by step guide to data analysis using SPSS for
Windows (Version 10). Australia: s.n.

74
Parameswaran, P., 2005. Public Finance and Taxation. Addis Ababa:
R.PARAMESWARAN.

Parsons, T., 1960. Some ingredients of a formal theory of organization Structure and process
in modern society. n T. Parsons (Ed.) ed. s.l.:Glencoe, IL: Free Press.

Pigou , A. C., 1920. The economics of welfare. london: : Macmillan: s.n.

Richardson, G., 2008. The relationship between culture and Tax Evasion across countries.
Journal of International Accounting, Auditing and Taxation, 17(2), pp. 67-78.

Sandmo, A., 2005. The theory of tax evasion: A retrospective view. National Tax journal,
58(4), pp. 643-663.

Saunders, M., Lewis, P. & Thornhill, A., 2009. Research Methods for Business Students. In:
New York: Pearson.

Saunders, Thornhill & Lewis, 2009. Research Methods for Business Students. 5th Edition ed.
s.l.:Prentice Hall.

Shenkute, Y., 2018. factors affecting tax revenue collection in the case of Addis Ketema sub
city tax office. Masters thesis.

Slade , K., 1939. The Ability-to-Pay Theory of Taxation. The American Economic Review,
29(1), pp. 92-101.

Smith, A., 1776. Wealth of nations. Book V ed. England: Published by Bibliomania.com.ltd.

Spilke, C. B. et al., 2019. Taxation of Individuals and Business Entities. 2019 Edition ed.
New Yor: Published by McGraw-Hill Education.

Terefe, K. D. & Teera, J., 2018. Determinants of tax revenue in East African countries: An
application of multivariate panel data cointegration analysis. Journal of Economics
and International Finance, 10(11), pp. 134-155.

Tesfaye, . M. A., 2011. Ethiopian Tax Accounting: Principles and practice. 1st ed. Addis
Ababa: Publisher not identified.

Truglia, R. P. & Troiano, U., 2015. Shaming tax delinquents. Massachusetts, s.n.

Tumoro, D. T., 2020. Challenges of business income tax administration with special
emphasis on collection process in Ethiopia: The case of Hadiya zone, Hossana town.
Journal of Accounting and Taxation, 12(3), pp. 85-98.

Udin, N., Manaf, N., Ishak, Z. & Rusniza, A., 2012. Tax Professionals‟ Views on Continuing
Professional Development (CPD) Programs in Malaysia In 6th Knowledge
Management International Conference. Malaysia, s.n.

Wagner , A., 1883. Finanzwissenschaft, Leipzig. Partly reprinted in Classics in the Theory of
Public Finance. 3rd ed. London: Macmillan: s.n.
75
Wayau, N. D., 2017. The effect of environmental factors on revenue collection in Kitui
County.

Williams, C., 2018. Entrepreneurship in the Informal Sector: An Institutional Perspective.

London: Routledge: s.n.

Workineh, A., 2016. Determinants of tax revenue in Ethiopia (johansen co-integration

approach). International Journal of Business, Economics and Management, 3(6), pp.
69-84.

World Bank, 2018. Annual report (investing in opportunity and Ending Poverty), Washington
D.C: s.n.

Yesegat, A. W. & Fjeldstad, H. O., 2016. Business people’s views of paying taxes in Ethiopia
(ICTD Working Paper 43). Brighton, UK, Brighton, UK: International Centre for
Tax and Development, Institute of Developmental Studies.

Yilma, K. M., 2013. Notes on the salient features of tax liens under Ethiopian law. [Online]
Available at: http://dx.doi.org/10.4314/mlr.v7i1.9

Younde, S. L. S., 2019. Tax compliance. in Farazmand A. (eds0 Global public

administration, public policy, and Governance, springer, cham., s.l.: s.n.

Yusnika, F. & Widyaningsih, A., 2018. Factors that Influence Tax Revenue and Government
Expenditure in the Asia Pacific Region. s.l., s.n.

76
 Appendix I
Jimma University
College of Business and Economics
Research Questionnaires
Questionnaire for the data collection on the determinant of tax revenue collection from
taxpayers
General Instruction: This questionnaire is designed to identify the determinants of tax
revenue collection in Sululta town. You are kindly asked to participate in the study by
providing your opinion for each part. The information you provide will be highly
confidential.
Thank you for your cooperation in advance!
Part I. BACKGROUND INFORMATION
Please tick mark (√) on appropriately.
1. Please indicate your gender Male ( ) Female ( )
2. How long have you been paying tax in this business?
0-5 years ( ) 5-10 years ( ) 10-15 years ( ) Over 15 years ( )
3. To date, what has been your highest formal qualification?
Secondary School Level ( ) Diploma ( )
First degree and above ( )
4. To which taxpayer category do you belong to?
Category A ( ) Category B ( ) Category C ( )
Part II: Revenue Collection
How do you evaluate the tax collection efficiency of Sululta tax revenue office?
Low ( ) Medium ( ) High ( )
Part III: Indicate your level of agreement to the following items that are related to
determinants of tax revenue collection. Express your opinion by encircling on the following
number according to your level of agreement.
1= strongly disagree, 2= disagree, 3 = Neutral, 4 = Agree, 5 = strongly agree
Tax evasion
No. Attribute Level of agreement
1 There is high degree to which I aware of people who evade 1 2 3 4 5
their taxes.
2 I think people evade tax due to the fact that the probability of 1 2 3 4 5
being caught is low.
3 People evade tax because of low amount of penalty imposed 1 2 3 4 5
on evaders by tax authority.

77
Corruption
No. Attribute Level of agreement
1 Tax officials demand bribes when they assess tax. 1 2 3 4 5
2 There is high degree to which I believe tax officials are 1 2 3 4 5
corrupted
3 Tax officials accept bribes when they offer to reduce tax 1 2 3 4 5
liability

Employment competence
No. Attribute Level of agreement
1 Sululta tax revenue office employees have tax administration 1 2 3 4 5
skills.
2 I think Sululta tax revenue office employees have tax 1 2 3 4 5
assessment skills.
3 I think employees are open to serve tax payers transparently. 1 2 3 4 5
4 I think qualification of tax collectors (officials) and their 1 2 3 4 5
experience is good.
Technology
No. Attribute Level of agreement
1 The tax collection is manual. 1 2 3 4 5
3 The mode of payment is time consuming and tedious. 1 2 3 4 5
3 The payment process is not advanced 1 2 3 4 5

The strength of tax authority

No. Attribute Level of agreement
1 The authority has adequate human and technological resources. 1 2 3 4 5
2 The law enforcement potential of the authority is strong. 1 2 3 4 5
3 The authority delivers quality tax services to taxpayers. 1 2 3 4 5

Political instability
No. Attribute Level of agreement
1 As a taxpayer I think, I don‟t have to pay tax because there is 1 2 3 4 5
no guaranty for my business during crises.
2 When political instability exists, it affects our income report is 1 2 3 4 5
checked and assessed by government body.
3 Political instability caused to stop our business operation. 1 2 3 4 5

Tax delinquency
No. Attribute Level of agreement
1 As taxpayer I know what tax delinquency means and it penalty. 1 2 3 4 5
2 There is a case when we didn‟t file our tax return before due 1 2 3 4 5
date.
3 The process and penalty of delinquency tax is easy. 1 2 3 4 5

Thank you for your participation!

78
 Appendix II
Jimma University
College of Business and Economics
Research Questionnaires
Interview for the data collection on the determinant of tax revenue collection from
employee
This interview is designed to identify the determinants of tax revenue collection in Sululta
town. You are kindly asked to participate in the study by providing your opinion for each
question. The information you provide will be highly confidential.
Thank you for your cooperation in advance!
1. How do you observe taxpayer's activities on tax evasion?
2. What mechanisms do they use to evade?
3. How well do you control tax collectors and assessors?
4. Do you think some tax officials are corrupted?
a. How and when do they commit corruption
5. How do you improve the competence of your employee on tax collection and
assessment?
6. Is there training for employees those perform tax collecting and assessing task or for
any other employees?
a. What change have you observed on tax collection capability after training?
7. What kind of technology do you use to collect tax?
a. Do you think current technology has improved tax collection?
8. What is the current status of Sululta tax revenue authority in terms of technology and
human resource, enforcement mechanisms and others if any?
a. Side of it strength
b. Side of its weakness
9. How well Sululta tax revenue office does use it a technological and human resource?
10. How do you evaluate the effect of tax collection during political instability?
11. Do taxpayers always pay tax on time?
a. If not how it affects tax collection?
b. Why taxpayers do you think not to pay tax on time?

Thank you for your cooperation!

79
 Appendix III
1. The ordered logistics regression result

2. Test of proportional odd assumption (test of parallel line)

80
3. Test of Multicollinearity

4. Correlation matrix for explanatory variables

5. Test for heteroskedasticity

81
6. Reliability of instrument

82
7. Marginal effect

83
84
 8. Percentage of respondents responses to each questions

Statements Response Frequency Percent

There is high degree to which I aware of S. disagree 25 8.9
people who evade their taxes. Disagree 23 8.2
Neutral 53 18.9
Agree 109 39.3
S. agree 69 24.7
Total 279 100
I think people evade tax due to the fact S. disagree 25 8.9
that the probability of being caught is Disagree 54 19.2
low. Neutral 39 13.9
Agree 99 35.2
S. agree 62 22.8
Total 279 100
People evade tax because of low amount S. disagree 18 6.4
of penalty imposed on evaders by tax Disagree 20 7.8

85
authority. Neutral 39 13.9
Agree 128 45.6
S. agree 74 26.3
Total 279 100
Statements Response Frequency Percent
Tax officials demand bribes when they S. disagree 20 7.1
assess tax. Disagree 18 6.4
Neutral 58 20.6
Agree 107 38.1
S. agree 76 27.0
Total 279 100
There is high degree to which I believe S. disagree 23 8.2
tax officials are corrupted Disagree 19 6.8
Neutral 30 10.7
Agree 88 31.3
S. agree 119 42.3
Total 279 100
Tax officials accept bribes when they S. disagree 27 9.6
offer to reduce tax liability Disagree 23 8.2
Neutral 25 8.9
Agree 81 28.8
S. agree 123 43.8
Total 279 100
Statements Response Frequency Percent
Sululta tax revenue office employees S. disagree 60 21.4
have tax administration skills. Disagree 34 12.1
Neutral 88 31.3
Agree 60 21.4
S. agree 37 13.2
Total 279 100
I think Sululta tax revenue office S. disagree 67 23.8
employees have tax assessment skills. Disagree 24 8.5
Neutral 87 31.0
Agree 85 30.2
S. agree 16 5.7
Total 279 100
I think employees are open to serve tax S. disagree 75 26.7
payers transparently. Disagree 34 12.1
Neutral 55 19.6
Agree 75 26.7
S. agree 40 14.2
Total 279 100
Statements Response Frequency Percent
The manual tax collection system is S. disagree 88 31.3
suitable. Disagree 58 20.6
Neutral 27 9.6
Agree 50 17.8
S. agree 56 19.9
Total 279 100
The mode of payment is time consuming S. disagree 92 32.7
and tedious. Disagree 38 13.5
Neutral 13 4.6

86
 Agree 71 25.3
S. agree 65 23.1
Total 279 100
The payment process enables me to pay S. disagree 87 31.0
tax on time. Disagree 54 19.2
Neutral 22 7.8
Agree 48 17.1
S. agree 68 24.2
Total 279 100
Statements Response Frequency Percent
The authority has adequate human and S. disagree 76 27.0
technological resources. Disagree 105 37.4
Neutral 37 13.2
Agree 43 15.3
S. agree 18 6.4
Total 279 100
The law enforcement potential of the S. disagree 86 30.6
authority is strong. Disagree 92 32.7
Neutral 40 14.2
Agree 45 16.0
S. agree 16 5.7
Total 279 100
The authority delivers quality tax S. disagree 84 29.9
services to taxpayers. Disagree 83 29.5
Neutral 32 11.4
Agree 49 17.4
S. agree 31 11.0
Total 279 100
Statements Response Frequency Percent
As a taxpayer I think, I don‟t have to pay S. disagree 38 13.5
tax because there is no guaranty for my Disagree 54 19.2
business during crises. Neutral 53 18.9
Agree 81 28.8
S. agree 53 18.9
Total 279 100
When political instability exists, it S. disagree 37 13.2
affects our income report is checked and Disagree 70 24.9
assessed by government body. Neutral 100 35.6
Agree 35 12.5
S. agree 37 13.2
Total 279 100
Political instability caused to stop our S. disagree 32 11.4
business operation. Disagree 57 20.3
Neutral 44 15.7
Agree 103 36.7
S. agree 43 15.3
Total 279 100
Statements Response Frequency Percent
As taxpayer I know what tax S. disagree 27 9.6
delinquency means and it penalty. Disagree 31 11.0
Neutral 135 48.0
Agree 52 18.5

87
 S. agree 34 12.1
Total 279 100
There is a case when we did not file our S. disagree 26 9.3
tax return before due date. Disagree 12 4.3
Neutral 41 14.6
Agree 80 28.5
S. agree 120 42.7
Total 279 100
The process and penalty of delinquency S. disagree 24 8.5
tax is easy. Disagree 28 10.0
Neutral 8 2.8
Agree 78 27.8
S. agree 141 50.2
Total 279 100

88