AWADHI M. WILLIAM Tyr

AN ANALYSIS OF THE IMPACT OF TAXATION ON
ECONOMIC GROWTH IN TANZANIA: 1996-2019
AWADHI M. WILLIAM
A DISSERTATION REPORT SUBMITTED IN PARTIAL FULFILMENT OF
THE REQUIREMENTS FOR THE REQUIREMENTS OF THE DEGREE OF
MASTER OF SCIENCE IN ECONOMICS (MSc. ECON)
DEPARTMENT OF ECONOMICS
OF THE OPEN UNIVERSITY OF TANZANIA
2021
ii
CERTIFICATION
The undersigned certifies that, he has read and hereby recommends for acceptance
by the Open University of Tanzania a dissertation entitled: ‘The impact of taxation
on the economic growth of Tanzania between 1996 and 2019” in partial
fulfilment of the requirements for the degree of Master of Science in Economics of
Open University of Tanzania.
………………………………………
Dr. Felician Mutasa
(Supervisor)
………………………………..
Date
iii
COPYRIGHT
No part of this dissertation may be reproduced, stored in any retrieval system, or
transmitted in any form by any means, electronic, mechanical, photocopying,
recording or otherwise without prior written permission of the author or the Open
University of Tanzania in that behalf.

iv
DECLARATION
I, Awadhi M. William, declare that, the work presented in this dissertation is
original. It has not been presented to any other University or Institution. Where other
people’s works have been used, references have been provided. It is in this regard
that I declare this work as originally mine. It is hereby presented in partial fulfilment
of the requirement for the Degree of Master of Science in Economics (MSc-
Economics).
………………………………………
Signature
………………………………..
Date
v
DEDICATION
This dissertation is dedicated to my parents whose inspiration encouraged me to
complete this work.

vi
ACKNOWLEGEMENT
I would like to thank the Almighty God for his grace, in giving me the quality of life
and wellbeing to complete this work, since it is in my firm conviction that without
His endowments, I could never have accomplished it.
I owe much gratitude to my supervisor, Dr. Felician Mutasa who, in spite of being
occupied with different obligations, saved some time to oversee my work for this
study, empowering, and fundamentally adjusting the work, it would have been a
great hurdle finishing this study had it not been for his direction.
Much gratitude goes to my parents/mother/father/wife/fiancée/ for their
encouragement and support through all the season of this course.I additionally thank
all my lecturers at the Open University of Tanzania in the MSc (economics) program
for the knowledge they have invested in me. I therefore thank, Dr. T. Lyanga,
Professor. D. Ngaruko, Dr. F. Mutasa, and Mr. Abdul Kilimathe course coordinator
who was always there to provide the necessary assistance.
I also feel obliged to express gratitude towards all my course mates who imparted in
me immeasurable knowledge through their countless discussions.

vii
ABSTRACT
The study assessed the impact of total tax revenueon the economic growth of
Tanzania. The specific objectives were to examine the effect of total tax revenueon
the economic growth of Tanzania, and to examine the trends of different taxes such
as corporate tax, PAYE, taxes, excise duty, import duties, VAT and other domestic
taxes in Tanzania. The study used time series data of Tanzania for the period of
(1996 to 2019) collected from the Bank of Tanzania, National Bureau of Statistics
and the World Bank. Vector error correction model for estimation of the time series
was used since the data were only stationary in first difference I(1), and there was co
integration within the variables. The study also conducted stability test for VAR and
VECM models to ensure that the estimates were effective. The study revealed an
insignificant positive effect of total tax revenueon economic growth (GDPRate) of
Tanzania both in the short and long-run. Also found government spending has a
negative impact on economic growth rate both in the short and the long-run, contrary
to Wagner’s law of state (1876) which holds that for any country, public expenditure
rises constantly as income growth expands, however, many other empirical studies
have established that that government spending can have either positive or negative
effects on economic growth depending on how and to what sector the government is
spending on (Lee, Won and Jei, 2019). Capital accumulation which affected the
economic growth of Tanzania positively.It is the study’s recommendations that
government further institute systems which broaden the tax bases for taxes with
positive impact on the economic growth and review for possible reduction of
individual taxes that prove to have negative effects on the economic growth.
Keywords: Fiscal Policy, Economic Growth, Taxation, GDPRate, Total tax revenue.
viii
TABLE OF CONTENTS
CERTIFICATION.....................................................................................................ii
COPYRIGHT............................................................................................................iii
DECLARATION.......................................................................................................iv
DEDICATION............................................................................................................v
ACKNOWLEGEMENT...........................................................................................vi
ABSTRACT..............................................................................................................vii
TABLE OF CONTENTS.......................................................................................viii
LIST OF TABLES...................................................................................................xii
LIST OF FIGURES................................................................................................xiii
LIST OF ABBREVIATIONS.................................................................................xiv
CHAPTER ONE.........................................................................................................1
INTRODUCTION AND background to the study..................................................1
1.1 Background of the Study..................................................................................1
1.2 Statementofthe Problem...................................................................................4
1.3 Objective of the Study......................................................................................6
1.3.1 Specific Objectives of the Research.................................................................6
1.3.2 Hypothesis........................................................................................................6
1.4 Significance of the Study.................................................................................6
1.5 Scope of the Study............................................................................................7
CHAPTER TWO........................................................................................................8
LITERATURE REVIEW..........................................................................................8
2.1 Introduction......................................................................................................8
2.2 Definitions of Concepts....................................................................................8

ix
2.2.1 Economic Growth............................................................................................8
2.2.2 Fiscal Policy.....................................................................................................9
2.2.3 Government Spending......................................................................................9
2.2.4 Capital Accumulation.....................................................................................10
2.3 Theoretical Review........................................................................................11
2.3.1 Taxation Theories...........................................................................................11
2.3.2 Integration of Taxation into Economic Growth Theories..............................12
2.4 Empirical Review...........................................................................................13
2.5 Research Gap..................................................................................................17
2.6 Conceptual Framework..................................................................................18
2.6.1 Descriptions and Measurement of the Variables............................................19
CHAPTER THREE.................................................................................................21
RESEARCH METHODOLOGY............................................................................21
3.1 Introduction....................................................................................................21
3.2 Research Design.............................................................................................21
3.2.1 Research Approach........................................................................................21
3.3 Area of the Study............................................................................................21
3.4 Population of the Study..................................................................................22
3.4.1 Unit of Analysis.............................................................................................22
3.4.2 Sample Size....................................................................................................22
3.5 Types and Sources of Data.............................................................................22
3.5.1 Types of Data................................................................................................22
3.5.2 Sources of Data..............................................................................................22
3.5.3 Data Collections Methods..............................................................................23

x
3.6 Validity and Reliability..................................................................................23
3.6.1 Validity and Reliability..................................................................................23
3.7 Ethical Issues..................................................................................................24
3.8 Data Analysis.................................................................................................24
3.8.1 Theoretical Model..........................................................................................25
3.8.2 Estimation Model...........................................................................................26
3.8.3 Estimation Diagnostic Tests...........................................................................27
CHAPTER FOUR....................................................................................................33
PRESENTATION AND DISCUSSION OF RESULTS.......................................33
4.1 Introduction....................................................................................................33
4.2 Summary Descriptive Statistics.....................................................................33
4.2.2 Correlation Matrix..........................................................................................35
4.3 Summary of Diagnostic Tests’ Results..........................................................37
4.3.1 Linearity Test.................................................................................................37
4.3.2 Normality test.................................................................................................38
4.3.3 Autocorrelation Test.......................................................................................39
4.3.4 Heteroskedasticity Test..................................................................................40
4.3.5 Multicollinearitytest.......................................................................................42
4.3.6 Unit Root Test................................................................................................43
4.3.7 Cointegration Test..........................................................................................45
4.3.8 Causal Effects (Granger Causality test).........................................................46
4.3.9 The Causal Effect of Total Tax Revenue to Economic Growth of
Tanzania.........................................................................................................47
4.4 Vector Error Correction Model Estimation....................................................51

xi
4.4.1 The Effect of total tax revenue on the Economic Growth of Tanzania..........53
4.4.2 The effect of Capital Accumulation on the Economic Growth of
Tanzania.........................................................................................................55
4.4.3 The Effect of Government Spending on the Economic Growth of
Tanzania.........................................................................................................56
4.4.4 Stability Condition test of the VECM Estimates...........................................58
4.4.5 The Trends of Different Taxes such as Corporate Tax, PAYE, Taxes,
Excise Duty, Import Duties, VAT and other Domestic Taxesin
Tanzania.........................................................................................................60
CHAPTER FIVE......................................................................................................64
CONCLUSION.........................................................................................................64
5.1 Introduction....................................................................................................64
5.2 Summary of the Study....................................................................................64
5.3 Conclusion of the Study.................................................................................66
5.4 Recommendations and Policy Implications...................................................67
5.5 Area for Further Study...................................................................................69
REFERENCES.........................................................................................................70
APPENDICES..........................................................................................................76
xii
LIST OF TABLES
Table 4.1: Summary Statistics of Taxation Revenue, Capital Accumulation,
and Government spending and Economic Growth Rate in
Tanzania.................................................................................................33
Table 4.2: Correlation Matrix.....................................................................................36
Table 4.3: Normality Test (Jarque-Bera test).............................................................39
Table 4.4: Autocorrelation Test (Durbin-Watson Test).............................................40
Table 4.5: Heteroskedasticity (White’s test)..............................................................41
Table 4.6: Multicollinearitytest (VIF test).................................................................42
Table 4.7a: Unit Root Testat Level (Augmented Dickey Fuller test)........................43
Table 4.7b Unit Root Testat first Different (Augmented Dickey Fuller test)............43
Table 4.8: Cointegrating Test (Trace Test)................................................................45
Table 4.9: Causality Test (Granger Causality Wald TESTS)....................................47
Table 4.10: Long Run Estimation Result...................................................................52
Table 4.11: Short run Estimation Result....................................................................52
Table 4.12: VECM Stability Condition Test..............................................................59

xiii
LIST OF FIGURES
Figure 2.1: Conceptual Framework of the Study.......................................................19
Figure 4.2: VAR Stability Condition Test..................................................................49
Figure 4.3: Parameters’ Stability Test Result (sum of Recursive Residuals)............50
Figure 4.4: VECM Stability Test (eigenvalues of the Companion Matrix)...............59
Figure 4.5: Trends of PAYE, Corporation Tax and other Income Taxes..................60
Figure 4.6: Trends of VAT, Import Duty, Excise Duty, and other Domestic
Taxes and other Income Taxes.................................................................61

xiv
LIST OF ABBREVIATIONS
DEA Data Envelopment Analysis
GDP Gross Domestic Product
OECD Organization for Economic Cooperation and Development
OLS Ordinary Least Squares
SACU Southern African Customs Union
TRA Tanzania Revenue Authority
UEMOA West African Economic Monetary Union
USA United States of America
VAT Value Added Tax
VECM Vector Error Correction Model
WAEMU West African Economic Monetary Union
WDI World Development Indicator
WTO The World Trade Organization
NBS National Bureau of Statistics
TIC Tanzania Investment Centre

1
CHAPTER ONE
INTRODUCTION AND BACKGROUND TO THE STUDY
1.1Background of the Study
From the view of economists, a tax is an obligatory input of resources from the
private sector to the government. It is charged based on prearranged criteria that do
not consider any benefits that the taxpayer may receive, the tax burden is distributed
by the government among taxable individuals/businesses, aiming at influencing the
macroeconomic performance of the country’s economy through its fiscal policy,
sometimes the government imposes certain taxes so as to adjust consumption
patterns within the country’s economy (Goode, 1984).
It is broadly established that economic growth is the basis of improved prosperity,
and taxes are a significant tool for the economy (Myles, 2000). Although there are
many definitions of taxes, Appah (2010) defined tax as a compulsory levy imposed
on a subject or his/her property by the government to provide social amenities and
possible economic prosperity. Similarly, Chigbu and Njoku (2015) stress that
taxation is a key source of revenue for every economy and it’s commonly an
instrument used in reducing the gap between the rich and the poor.
Academically, there is a huge work of literature on taxes and their growth features,
as well as on widely varying methodologies and results (Gale, Krupkin, & Rueben,
2015). For example, McBride (2012) shares results of Congressional Research
Service, which has found support for the theory that taxes have no effect on
economic growth by relying on the U.S. experience from the time of the World War
II.
2
The degree to which tax revenue fuels economic performance in an economy
especially in developing countries continue to attract empirical debate, for example,
the Revenue Statistics in Africa, (2016) reported that in 2014, eight African
countries which include Cameroon, Côte d’Ivoire, Mauritius, Morocco, Rwanda,
Senegal, South Africa, and Tunisia reported tax revenues as a percentage of gross
domestic product (GDP) ranging from 16.1 to 31.3%. overall, these countries had a
rise in their taxation-to-GDP ratios in comparison to the average tax to GDP ratios of
Organization for Economic Co-operation and Development (OECD) countries, the
increase of 34.4% was only 0.2 percentage points higher in 2014 than in 2000.
The Revenue Statistics in Africa paper further reported that in that particular year
some African countries were significantly dependent on non-tax revenues, and more
specifically on grants such as foreign aid (Kenya), and resource rents from oil
(Nigeria and Angola) and Bauxite (Zambia); these countries’ economies tend to be
highly volatile that their finances could not be stabilized and predictive through tax
revenue. This leads to a question of whether the revenues from taxation in African
countries including Tanzania are growing as a proportion of national incomes.
It should be noted that non-tax revenues tend to vary more between years than tax
revenues, often due to changing commodity prices or the variability of grants. There
are several studies set in different timeframes and countries that have investigated
the relationship between taxes and economic output using cross data and established
that there is a negative relationship between taxes and economic growth, they
include Dowrick (1992), Easterly and Rebelo (1993), Badri and Allahyari (2013),
3
Zellner and Ngoie (2015), Seward (2008), Canicio and Zachary (1975), Lee and
Gordo (2005), Atems (2015), Ojede and Yamarik (2012), Dackehag and Hansson
(2012), Szarowska (2010) and Stoilova (2017).
Though most of the studies and theory underscores a negative relationship between
economic growth and tax, some studies have found a positive relationship (namely,
Orcan, 2009; Babatundel, Ibukun & Oveyemi 2017; and Tosun and Abizahed 2005).
Orcan (2009) investigated the impact of economic policy on economic growth in the
Republic of South Africa utilising the vector auto-regression (VAR) modeling. The
findings reveal that tax income have a positive association with economic growth.
However, tax alone takes a significantly long time to impact economic growth.
Babatundel, Ibukun & Oveyemi (2017) conducted a study to look at the link between
tax and economic growth in the Africa continent beginning in 2004 to 2013, the
findings for this study indicated that tax income has a positive link with GDP and
promotes economic growth in Africa perspective.
The tax trends in Tanzania have been changing over the years. The corporate
income-tax has been the biggest single source but has declined since the mid-1990s.
Levin (2001) analysed the sectorial development of corporate income taxes from
1996–2000. The main finding was that corporate income-tax payments in the
manufacturing sector had declined substantially over the period, from almost 41% of
total corporate tax payments in 1996 to just over 6% in 1998, since rising somewhat
but still low. The agriculture sector, which is usually assumed to be difficult to tax,
was actually contributing about a third of total corporate income taxes.

4
The share of consumption taxes in GDP has fallen since the mid-1990s, while the
share of trade taxes has increased. Within the latter, the GDP-share of import tariffs
has declined, while the shares of excise duties and VAT on imports have risen. Tariff
rates have declined and the number of tariff bands has been reduced, but the
declining GDP-share of import tariffs has been attributed to generous exemption-
schemes and evasion. In a number of African countries, VAT has increased revenue
quite successfully. The Value Added Tax (VAT) was introduced in July 1998 (the
fiscal year 1998/99) replacing previous sales-taxes and parts of the stamp-duty and
entertainment taxes.
It was expected that VAT would lead to more revenue being generated in Tanzania
because of the significantly wider base that VAT enjoys as compared to the previous
sales-tax that was replaced by VAT. First, manufacturers previously excluded from
sales-tax are registered as taxpayers under the VAT-system. Second, a number of
items not previously liable to sales-tax are subjected to VAT, and the number of
goods excluded from VAT is limited. Third, VAT includes services such as
electricity, telephone, hotels, and restaurants, which previously were subjected to a
lower sales tax. Nevertheless, the introduction of VAT in Tanzanian has produced
less impact in comparison to a number of other countries.
1.2 Statementofthe Problem
Most developing nations endeavour to attain economic growth and development
through taxation, nevertheless the discussion on the effectiveness of taxes as an
instrument for promoting growth and development remains lacking since several
5
studies have indicated mixed impacts of the tax on economic growth. A positive
relationship between taxation and economic growth has been reported by several
studies (Dasalegn, 2014; Ugwunta and Ugwuanyi, 2015). Contrarily, the works of
Keho (2013), juniors and Tafirenyika (2010), Delessa and Daba (2014), Saima et al.
(2014) report a negative relationship between these variables. According to McBride
(2012) rising taxation diminishes investment and overall entrepreneurial activity due
to the fact that an unreasonably large portion of these activities are done by high-
income earners.
In contradiction two the above-conflicting opinions from different works of
literature, some studies have found no significant relationship between the variables
(N’Yilimon, 2014). Hungerford (2012) studying the United States of America (USA)
experience from 1975 to 2011 found support for the opinions that taxes have no
impact on the economic growth of the USA. Hungerford’s findings were also
supported by findings by Osundina and Olanrewaju (2013) that further report the
effect of taxation on national growth was insignificant and the overall tax burdens
have no effect on the growth of the economy.
Jens et al (2011) using 21 OECD countries between 1971 and 2004 finds that
corporate taxes have been most harmful to the economy, likewise taxes on personal
income, consumption, and property. Indeed, the progressivity of income taxation is
damaging to economic growth. When the top marginal rate on income increases,
productivity, and growth is adversely affected. It is because of these conflicting
findings on the relationship between taxation and economic growth and the overall
perceived importance of taxation to the countries’ development that this study was
6
necessary so as to further investigate both the short run and the long-run relationship
between these variables in the case of Tanzania covering a period between 1996 and
2019.
1.3 Objective of the Study
The main objective of this study was to analyse the impact of taxation on the
economic growth of Tanzania between 1996 and 2019
1.3.1 Specific Objectives of the Research
The study had the following specific objectives.
i. To examine the impact of total tax revenueon the economic growth in
Tanzania.
ii. To examine the trends of different taxes such as corporate tax, PAYE, other
income taxes, excise duty, import duties, VAT, and other domestic taxes in
Tanzania.
1.3.2 Hypothesis
The study was guided by the following hypothesis:
H1: There is a significant impact of total tax revenueon economic growth in
Tanzania.
1.4 Significance of the Study
The results of the research should prove useful to the Tanzanian government,
policymakers, legislature, and regulatory bodies such as the TRA in improving
taxation systems and therefore, tax policies in Tanzania. To the investors and
citizens, this study provides an insight into Tax rates and their impacts on economic
7
growth. Investors need to establish business strategies putting into consideration the
long-term effects (and consequences) of their decisions on the business and the
economy. The academicians and researchers may use the study as a foundation to
carry out any further research in this area.
1.5 Scope of the Study
According to Kothari (2006), the scope of the study clarifies the boundaries of the
research. As much as one might want to, they cannot study everyone, everywhere,
and everything (Miles and Huber man, 1994). Accordingly, this study focused on the
assessment of the impact of taxation on the economic growth of Tanzania between
1996 and 2019 mainly by using secondary data.

8
CHAPTER TWO
LITERATURE REVIEW
2.1 Introduction
A literature review can be referred to as the essential evaluation of section of
the published information through summarizing, classifying, and making assessment
prior to a new study; it entails an assessment of ideas and theories and associated
studies. According to (Fellows and Liu, 2003) literature review is the assessment
of documents from the research discovered in the literature that is associated to
the topic of interest. The literature overview will allow the researcher to have a clear
understanding of the topic in question, and be in a position to make
an integral assessment of the findings.
2.2 Definitions of Concepts
2.2.1 Economic Growth
Economic growth refers to the annual increase in the ability of the economic system
to produce goods and services. It can be measured in nominal terms (not inflated), or
in real terms (inflated). The growth of an economy involves both an increase in
productive capacity and an improvement in the lifestyle to the concerned
community. Gross Domestic Product (GDP) refers to the economic cost of all the
completed goods and services produced within a country's borders in a particular
time period. It consists of all of the non-public and public consumption, government
outlays, investments and exports fewer imports that take place within a described
territory and is measured annually. GDP is usually used as an economic indicator of
the standard health of an economy, (Lipsey and Chrystal, 2007).

9
2.2.2 Fiscal Policy
These involve taxation policies and how the government spends various resources at
its disposal in order to finance various development projects for the public.
However, to meet its development goals, the government must raise adequate
revenues; the main avenue for generating revenue in Tanzania is a number of taxes,
in addition to interest and repayments, licenses, fines and fees(Todaro and Smith,
2003).
2.2.2.1 Taxation
Revenue Statistics in Africa (2019) defines taxes as compulsory, unreciprocated
payments to the government. Taxes are said to be unreciprocated since the benefits
delivered by governments are generally not proportional to the taxes paid.
2.2.2.2 Total tax Revenue
According to (OECD, 2016) total tax revenue is defined as the revenues collected
from taxes on income and profits, social security contributions, taxes levied on
goods and services, payroll taxes, taxes on the ownership and transfer of property,
and other taxes. Total tax revenue as a percentage of GDP indicates the share of a
country's output that is collected by the government through taxes. It can be regarded
as one measure of the degree to which the government controls the economy's
resources. The tax burden is measured by taking the total tax revenues received as a
percentage of GDP. This indicator relates to government as a whole (all government
levels) and is measured in million USD and percentage of GDP.
2.2.3 Government Spending
According to Corporate Finance Institute [CFI], 2020, government spending refers to

10
money spent by the public sector on the acquisition of goods and provision of
services such as education, healthcare, social protection, and defence. In national
income accounting, when the government acquires goods and services for current
use to directly satisfy the individual or collective needs and requirements of the
community, it is classified as government final consumption spending. On the other
hand, when the government acquires goods and services for future use, it is classified
as government investment. This includes public consumption and public investment,
and transfer payments consisting of income transfers.
2.2.4 Capital Accumulation
In Karl Marx's economic theory, capital accumulation is the operation whereby
profits are reinvested into the economy, increasing the total quantity of capital.
Furthermore, Tuovila, Estevez, (2021) refers to Capital accumulation as an increase
in assets from investments or profits and is one of the building blocks of a capitalist
economy. The goal is to increase the value of an initial investment as a return on
investment, whether that may be through appreciation, rent, capital gains, or interest.
Capital accumulation primarily focuses on the growth of existing wealth through the
investment of earned profits and savings. This investment is focused in a variety of
ways throughout the economy. One method of growing capital is through the
purchase of tangible goods that drive production. This can include physical assets
such as machinery. Research and development can also drive production and is
known as human capital. Investment in financial assets, such as stocks and bonds, is
another means of capital accumulation if the value of those assets increases. Another
important factor of capital accumulation is appreciation. This is typically

11
investments in physical assets whose value grows over time, such as real estate.
2.3 Theoretical Review
2.3.1 Taxation Theories
The research identified theories that explain the research objective through negation
or in support. There is a number of theories that highlight the concept of taxation,
they include; the Benefit theory of taxation by Cooper (1994) which advocates that
the taxes levied on individuals should be in line with the benefit they receive from
the government. However, the implementation of this theory is difficult because it is
not easy to quantify all the benefits provided by the government, for instance the
state security protection provided to each individual taxpayer residing in the country.
Contrary to the benefit theory is the ‘Cost of service’ theory of taxation which posits
that the tax paid by the citizens should complement the cost of service rendered by
the government, thus a cost-benefit conjecture. This theory can be complimented by
the Ability to paytheory by Pigou (1920) which proposes that taxes should be paid in
consideration of one’s ability to pay taxes, while it seems a reasonable and just
proposition, it however, does not cater for how one’s ability to pay is defined .
The Ibn Khaldun’s (1332 to 1406) theory as championed by Islahi (2006) classifies
the arithmetic effect as well as the economic effect of tax rates on revenues. The
arithmetic effect posits that by lowering tax rates, the tax revenues decrease in equal
proportion and vice versa. On the other hand, the economic effect suggested that
lower tax rates have a positive impact on work, output and employment. This theory
is complimentary to the Laffer curve theory developed in 1979 by economist Arthur

12
Laffer; which position lower tax rates with an increase in the economy. It
reinforces supply-side economics; since Lower tax rates means more money into
taxpayers’ pockets, more many into the taxpayers’ hands creates more businesses,
more employment, and consequently a larger tax base which ultimately substitutes
any revenue lost from lowering the tax rates.
2.3.2 Integration of Taxation into Economic Growth Theories
The quest for ideal taxation whereby tax revenues are maximized for social welfare
and economic growth has been the essence of the variety of theories. Adam Smith
viewed taxation as a way of sustaining the government. Ricardo justified the
inclusion of capital tax which as part of factors of production (labour and capital) is
required (in part) to fund government activities. In its regulatory function, taxation
presents a mechanism to redistribute national income. In its catalytic role, taxation is
utilized to extend the price of effective demand, stimulate investment and prompt
economic development.
While assessing the effect of fiscal variables on economic growth, it is important to
understand that tax collection impacts economic growth exclusively through its
effect on individual growth variables which include the capital accumulation,
investment or human capital and technology advancement (Kotlán, Machová, and
Janíčková, 2011). Both Judd (1985) and Barro (1990) broadened neoclassical growth
model by national tax burden. Additionally, King and Rebello (1990), examined the
causes for inconsistency existing amid different countries in long-term economic
growth, they hypothesed that the solution to these inconsistencies lies in diverse tax
policies which incentives individuals to accumulate both physical and human

13
capitals.
Impact of Taxation on Accumulation of Capital: The impact of tax on
accumulation of capital and consequently economic growth has been discussed by
various authors Brebler (2012); Daniel and Jefferey (2013), and Feld and
Heckemeyer (2008) reports a negative association of increase in corporate tax on
foreign direct investment (FDI). They claim that a lower corporate tax stimulates
economic growth through encouraging foreign investment which in turn increases
capital accumulation in the country. Becker (2009) posits that when making
investment decisions, prospective investors are more concerned with higher taxes
such as corporate tax and labour taxes than other characteristics of domestic
economy such as infrastructure, market availability, and politic stability.
2.4 Empirical Review
The research has identified several empirical reviews across both developed and
developing countries regarding taxation and economic growth. Diverse empirical
studies have investigated the effects of taxes on economic growth. Results are far
from being conclusive, varying across countries, methodologies, and fiscal variables
involved. This study reviews previous empirical works from both the global and
local perspective. In a study by Padovano and Galli (2001)in the time period
between 1951-1990; consisting of 23 OECD countries found that effective marginal
personal income taxes are negatively correlated with GDP growth.
Likewise, a study by Tomljanovich (2004) in the time period between 1960-1990;in
the United States found higher tax rates negatively affect short-run growth, but not
14
long-run growth. Lee and Gordon (2005) in a study involving 70 countries in the
time period between 1980-1997 concludes that reducing corporate income tax by 1%
raises annual growth by 0.1% to 0.2%. Tosun and Abizadeh (2005) in a study
involving OECD countries in the time period between 1980-1999; found that shares
of personal and property taxes have responded positively to economic growth, while
the shares of labour taxes and goods and services taxes have reflected a relative
decline.
In a study by Alesina and Ardagna(2010) involving OECD countries in the time
period between 1970-2007; OECD found fiscal incentives based on tax cuts enhance
growth more than increased consumption. On the other hand, Gemmell, Kneller and
Sanz (2011) conducted a study in OECD countries in the time period between 1970-
2004; the study concluded that taxes on income and profit are most damaging to
economic growth over the long run.
The second in line are consumption taxes. Romer and Romer (2010) study in the
time period between 1945-2007 conducted in the United States found that tax
increase of 1% GDP leads to a fall in output of 3% after about two years, this is
supported by a study by Barro and Redlick (2011) conducted in the United states
between 1912-2016 which also found that cut in the average marginal tax rate of 1%
raises GDP per capita by around 0.5 in the next year.
Similarly, Ferede and Dahlby (2012) study in Canada in the time period between
1977-2006 found that reducing corporate income tax by 1% raises annual growth by
0.1 to 0.2%. In another study conducted in Nigeria by Onakoya and Afintinni (2016)
15
in the time period between 1980-2013 it found that there is a significant positive
relationship between petroleum profit tax, company income tax and economic
growth. Contrarily, there was insignificant relationship between customs, excises
and economic growth.
A study by N’Yilimon (2014) using unit root test on panel data suggests the absence
of a non-linear relationship between taxation and economic growth of the West
African Economic Monetary Union (WAEMU) countries. Anne (2014), in her own
study, adopted Ordinary Least Squares, Unit Root tests, Johanssen Cointegration
Test, Vector Error Correction Model (VECM), and finds a negative but insignificant
effect of income taxes on the Kenyan Economy.
Dasalegn (2014) discovers that VAT revenue plays an energetic positive role in the
national development of Ethiopia. Sekou (2015) in his study utilized the ordinary
least squares method and reports a positive and significant positive correlation
between the collection of taxes and growth in Mali. Chigbu. Akujuobi, & Appah,
(2012), Ogbona, and Appah (2012) and Confidence and Ebipanipre (2014) espoused
granger causality and Ordinary Least Squares (OLS) to reveal that tax revenue is
positively and significantly related to economic growth in Nigeria. A similar finding
which indicates that low personal income taxes boosted the economic growth was
reported by Bonu and Pedro (2009) who utilized descriptive technique in Botswana.
On the other hand, Keho (2010) reports that higher tax rates are strongly linked with
reduced economic growth in Cote d’Ivoire. Similarly, Saibu (2015) reports a
negative connection between tax burden and economic growth rate of South Africa
16
and Nigeria. Another study by Yaya (2013) adopted the Branson and Lovell (2001)
linear programming based methodology which is the Data Envelopment Analysis
(DEA) for New Zealand, the findings revealed that higher taxes are linked with poor
economic growth.
A research on the effect of tax incentives on economic growth in Kenya conducted
by Hilda (2014) utilized descriptive analysis, correlation analysis and regression
analysis and reports the similar inverse relationship between the variables. Lawrence
(2015) studied the effect of VAT on economic growth of Kenya and Edame and
Okoi (2014) using Ordinary Least Squares estimating technique, found a significant
and negative relationship between Value Added tax and economic growth.
Locally, most of the studies identified address other aspects of taxation rather than
the topic of interest. Makame (2015) reports that VAT and revenue collection has a
positive impact on GDP. Fjeldstad (2001) presents propositions about tax collection
by local authorities in Tanzania. Furthermore, another study by Fjeldstad and
Semboja (2002), examined factors influencing tax compliance behaviour in local
authorities in Tanzania.
Levin and Widell, (2014) estimate the amount of tax evasion in customs authorities
in both Kenya and Tanzania by calculating measurement errors in reported trade
flows between the two countries and correlating those errors with tax rates. The
study finds that the measurement error is correlated with the tax rates in Tanzania.
The study also introduced a third country into the analysis, the United Kingdom, and
tax evasion seemed to be more severe in trade flows between Kenya and Tanzania
17
compared to trade flows between the United Kingdom and Kenya/Tanzania. Finally,
the study also finds that the tax evasion coefficient is lower in the Kenya–United
Kingdom case compared to the Tanzanian–United Kingdom case which suggests
that tax evasion is more severe in the Tanzanian customs authority.
2.5 Research Gap
From the global perspective studies have come up with inconclusive findings with
regards to the impact of taxation on economic growth, while some studies have
found significant negative relationship, others have reported no significant effect, at
the same time other studies have reported significant effect only in some components
of taxation (Lee and Gordon, 2005; Widmalm, 2001; Dackehag and Hansson, 2012;
and Arnold, 2008; Santiago and Yoo, 2012; Mehrara, Masoumib, and Barkhi, 2014).
Studies conducted in Africa have also have left the same dilemma with some studies
indicating a negative impact between taxation and economic growth (Branson and
Lovell, 2001; Chiumia and Simwaka, 2012; Edame and Okoi, 2014;Lawrence, 2015;
Keho, 2013; Saibu, 2015). On the other hand, other studies in Africa have shown a
positive impact between the variables (Chigbu et al., 2011;Confidence and
Ebipanipre, 2014; Dasalegn, 2014; Enokela, 2010; Emmanuel, 2013; Ogbona and
Appah, 2012; Sekou, 2015; Wisdom, 2014).To add more conflicting results there are
studies that have shown no significant effect between these variables (N'Yilimon,
2014;Ugwunta and Ugwuanyi, 2015).
As evidenced from a review of literatures in this section from both developed and
developing countries there are mixed findings as far as the impact of taxation on
18
economic growth is concerned. Some studies have reported a positive impact, others
a negative impact, and others have reported no significant impact between the
variables. Furthermore, in Tanzania the studies identified addresses other aspects of
Taxations such as tax compliance Fjeldstad and Semboja (2002), and tax evasion
Levin, and Widell, (2014), making this study a necessity in order to bridge the
knowledge gap that exist with regard to the impact of taxation on economic growth
of Tanzania.
2.6 Conceptual Framework
To analyse the impact of taxation and economic growth of Tanzania, in consistency
with other growth literature this study will employ the real GDP per capita as a
proxy for economic growth and therefore the dependent variable and taxation
(TAX)as the independent variable (Katusiime, Agbola, &Shamsuddin, 2016).
However,when evaluating the impact of taxation on economic growth, it is necessary
to realize that taxation can be integrated into growth theories only through its impact
on individual growth variables such as capital accumulation and investment or
human capital and technology advancement that actually affect the economic growth
(Kotlán, 2010; Kotlán, Machová and Janíčková, 2011).
Since several other works of literature have been documented as having used these
individual growth variables as independent variables in studies of taxation and
economic growth (Kiyota & Urata, 2004; Nawaz et al, 2014, Macek, 2014), This
study too will use capital accumulation (CINV). Furthermore, this study will also
add government spending (GOV) to the control variables since taxes represent the
most important source of public budgets, it is necessary to incorporate government

19
spending into the analysis since those consequently represent the basic expenditure
part of public budgets(Macek, 2014).The Conceptual Framework Diagram is shown
in Fig. 2.1.
Independent Variables
Taxation Rate
Dependent Variables
Capital Economic
Accumulation Growth
Government
Spending
Figure 2.1: Conceptual Framework of the Study

Source: Researcher’s Construct, 2020
2.6.1 Descriptions and Measurement of the Variables
2.6.1.1 Dependent Variable
The dependent variable for this study is Economic growth rate; which is represented
by Growth domestic products rate (GDP). GDP is an indicator of the country’s
economic performance; on average, higher GDP indicates the country is increasing

20
the amount of production that is occurring in the economy and the residents have
higher pay and subsequently are spending more. A similar measure was used as a
representative of the economic growth rate in another study (Macek, 2014) on the
impact of taxation on economic growth: Case Study of OECD Countries.
2.6.1.2 Independent Variables
The independent variable for this study is Taxation (TAX) approximated by total tax
revenue; Total tax revenue presents the ratio of tax revenues to gross domestic
product (GDP) usually for the period of one calendar year. The taxes to be used in
the analysis are value added tax, income tax, corporate tax, and withholding tax. The
control variables are also independent variables for this study, they are;
Capital Accumulation (CINV) -measured by indicator of proportion of real
investment to GDP, expressed in purchasing power parity per one resident.
Government spending (GOV) - it is measured as government spending as a %
GDP.
21
CHAPTER THREE
RESEARCH METHODOLOGY
3.1 Introduction
In this chapter the research design was determined, area of the investigation
identified, data gathering and analysis instruments selected, the research approach in
gathering and analysing the data defined, as well as the estimation method
determined.
3.2 Research Design
This research embraced a causal research design; this is a research structure which
endeavours to uncover cause and effect connection between variables (Brüderl,
2015). Furthermore, this study’s primary goal was to examine the impact that
taxation has on the economic growth of Tanzania, henceforth, the causal relationship
existing between taxation variables and economic growth.
3.2.1 Research Approach
This research adopted a quantitative research approach since it aimed at quantifying
the impact of taxation on the economic growth of Tanzania. The quantitative
research approach insignificant since it empowers the estimation of the long-run
relationship of the numerical variables.
3.3 Area of the Study
The study area for this research is Tanzania, with the aim of determining the impact
of Taxation on the economic growth of Tanzania.

22
3.4 Population of the Study
This study collected data that are readily documented at various sources across
Tanzania and the world, so it did not involve dealing with a specific population.
3.4.1 Unit of Analysis
The units that was analysed in this study were real GDP per capita, capital
accumulation and investment, government spending, value added tax, personal
income tax, corporate tax, and excise duty in Tanzania.
3.4.2 Sample Size
The sample size for this study is based on the time series data on taxation and
economic growth of Tanzania collected between years (1996- 2019). This sample
size is based on the fact that while the study had intended to research on a much
longer observation, the study found out that data on Taxes were only well
documented after the formation of TRA.

3
3.5 Types and Sources of Data
3.5.1 Types of Data
This study collected annual time series data on GDP, Capital Accumulation, Human
Capital, Government spending, value added tax, personal income tax, corporate tax,
international trade taxes, and the withholding tax of Tanzania covering a period of
1996 - 2019. The choice of this period is based on the fact that the study could only
find well documented data on Tax after the years following the formation of TRA.
3.5.2 Sources of Data
The data for GDP were collected from the World Bank database; the World Trade
23
Organization (WTO) and the World Economic Forum database were consulted for
data comparison. Data on taxations were obtained from reviewing TRA documents;
governmentspending was collected from the ministry of finance, and capital
accumulation from Tanzania Investment Centre (TIC).
3.5.3 Data Collections Methods
3.5.3.1 Secondary Data
Secondary data specifically desk review method was used in collecting the data
needed to answer the research questions. The study used desk research because the
data needed for this study were readily documented and could be obtained from the
sources identified or through the internet. Because desk review can collect secondary
data from sources such as libraries, the internet, published reports and statistics and
reviewing relevant documents. It is a quick method of collecting information; this is
because some of the most consuming steps of normal researches for example survey
researches are significantly reduced as data is already gathered (Doolan and
Froelicher, 2009). This method is easier than field research since all the information
needed is already published, it is cost-effective and convenient since someone else
has already collected the data, and the researcher does not have to devote many
financial resources to the collection of data.
3.6 Validity and Reliability
3.6.1 Validity and Reliability
To measure the degree by which the researcher acquired the important variables to
clearly respond to the investigation questions, the study explored a few writings to
guarantee the correct variables were chosen just as the strategies used to find those
24
responses were the right one; likewise, the study gathered information over a long
enough period to guarantee the appropriate responses given are valid. To quantify
the reliability of secondary information for this exploration, the study guaranteed
that the data collected from dependable and reputable sources which are government
firms, for example TRA and national bureau of statistics in addition to the, the
World Bank. Sincereliability relates to the consistency of a measure, the study also
used the World Bank, the world trade organizations, world tax index and the like to
compare the data about taxation and economic growth. Furthermore, the study
conducted the appropriate diagnostic estimation tests such as autocorrelation,
normality, heteroscedasticity, multicollinearity, and unit root test all in the bid to
ensure the validity and reliability of the data for this study.
3.7 Ethical Issues
The main ethical issues with using secondary data are confidentiality and security.
However, since this study used desk research to collect data from the internet. For
the data that were obtained from the internet, they are covered by implied permission
(implied consent) as long as the original ownership of the data is acknowledged. By
implied permission, it means the authors by uploading their work online have
implied that other people have consent to use the documents (Nijhawan, 2013).
3.8 Data Analysis
The data analysis for this study was executed by Gretl, which enables its users to
execute all common econometric tests such as it is mentioned by Wooldridge (2009)
or Greene (2003) as the main econometric program.

25
3.8.1 Theoretical Model
The theoretical model for this study is based on the neoclassical growth model which
outlines how a steady economic growth rate results from a combination of three
driving forces: labour, capital, and technology (Solow and Swan, 1957). The theory
posits that the accumulation of capital within an economy, and how people use that
capital, is important for economic growth. The theory asserts that the relationship
between the capital and labour of an economy determines its output. Finally,
technology is thought to augment labour productivity and increase the output
capabilities of labour, based on this theory, the following production model has been
used as a measure for various sources of economic growth; -
Where, Y= total national product, K = the quantity of physical capital used, L = the
quantity of labour used, A = the state of technology.
However, based on the variables for our study the theoretical model in equation 1
can be modified to;
Whereby;
GDP is the economic growth rate measured in annual GDP rate, Taxation Revenue
(TAX)is the ratio of tax revenues to gross domestic product (GDP) usually for the
period of one calendar year, Capital accumulation (CINV)is measured by indicator
of proportion of real investment to GDP, expressed in purchasing power parity per
one resident, and Government spending (GOV) it is measured as government
spending as a % GDP.
3.8.2 Estimation Model

26
The study used time series estimation to estimate the effect of taxation on economic
growth of Tanzania. The analysis was done using Vector Error Correction (VEC)
Model to determine the short and long run relationship between taxation and
economic growth in Tanzania, the model’s selection is based on the fact that other
studies of similar nature have made use of the model, however, stationary test as
well as co integration test ascertained the fitness of the model.
From equation 2 the VEC model is formulated as;
Whereby,
is the coefficient of the ECT and the speed of adjustment, it measures the speed at
which GDP returns to equilibrium after changes in TAX, CINV, and GOV.ε = the
error term which denotes other variables that are not specified in the model, and
ECTt-1 is the error correction term which is a lagged OLS residual obtained from the
long-run co integrated equation expressed as follows;
i -1 = lag order
(Cointegration equation & long-run model)
The ECT explains that previous period’s deviation from equilibrium (which is the
error) influences short-run (SR) movement in the dependent variable.

27
3.8.3 Estimation Diagnostic Tests
These were some of the estimation diagnostic tests which have been performed;
3.8.3.1 Normality Test
The normality of the data for this study was measured by the jarque-bera test.
The Jarque-Bera (JB) statistic combines the two measures, and as follows:
In case of a large number of observations, if JB is higher than 6, it suggests that the
population distribution is unlikely to be a normal distribution.
3.8.3.2 Unit Root Test
The study executed a unit root test since it’s a significant test when managing time
series data in order to decide on the stationarityof the variables. Unit root test was
utilized to check the stationary nature of the variables utilized for this research. A
time series variable is said to be stationary if and only if its mean, variance and auto
covariance are constant (finite) and independent of time. There are different
techniques for testing for unit root, for example Classical economic hypothesis
variables that are to differenced to accomplish the estimation of stationary are known
as 1 (1) and those which are stationary are known as 1 (0) series.
The Augmented Dickey-Fuller (ADF) test is utilized in this examination since it
permits expansion of more lags to accomplish white noise error term, which is
needed for the distributional outcomes to be valid (no autocorrelation). The null
hypothesis tested under the Augmented Dickey Fuller (ADF); it is rejected when the
28
absolute value of the computed t-static is greater than the absolute of the critical
value (Gujarati, 2004). The Augmented Dickey Fuller (ADF) Test takes the
following formulation; with a drift and trend model specification representation;
Where
a Is a drift Component and bxt is trend component.
jis a measure of Lag length and is a measure of unit root.
The null hypothesis:
Ho: Y = 0 for existence of unit root while
H1: Y< 0 for no unit root.
3.8.3.3 Co-integration Test
When the time series variables integrated of order one that is their first differences
are stationary, a co-integration test is performed with the aim of ascertaining the
existence of the long run relationship between total tax revenueand economic
growth. Therefore, a co-integration test through Johansen’s co-integration was
executed; this test has the ability to estimate in excess of one co-integration
relationship of two or more time series.
The test is through the estimation of the following equation:
Where;
λ is the maximum Eigenvalue,
T is a sample size.
29
Trace test was utilized in defining the number of cointegrating vectors, the test
examines ‘r’ as a null hypothesis against ‘n’ as the alternative hypothesis of
cointegrating relations, where “n” refers to the total number of variables in the
system for . The tests’ computation follows the formula in
equation 8:
3.8.3.4 The Error Correction Model
The error correction model (ECM) is a vital model in determining the long-run
relationship among the dependent and independent variables, furthermore, the model
can yieldbetter short-run estimates that are economically meaningful when the
variables are cointegrated of order zero (Lesage, 1990). When the long-run model is
reformulated into an ECM,it is able to integrate both the short and long-run
dynamics of the models. Consequently the ECMtakes the form in equation 9:
Δℽt +α0 + i Δyt-1+ Δzt-1+µECTt-1+φt ………………………………..………9

i
Whereby;
ECTt-i is a one-period lag of the residual term (disequilibrium) from the long-run
relationship, α, ø, δ,φ and µ are parameters, and øis a white noise error term.
3.8.3.5 Autocorrelation Test
Autocorrelation occurs when the residuals are not independent from each other, or
simply when the value of y(x+1) is not independent from the value of y(x). The
study used the Durbin-Watson test to test for serial correlation. The Durbin-
Watson’s (d) tests the null hypothesis that the residuals are not linearly auto-
30
correlated. While d can assume values between 0 and 4, Values around 2 indicate no
autocorrelation. As a rule of thumb values of 1.5 < d < 2.5 show that there is no
auto-correlation in the data. The Durbin-Watson (dw) d statistic was given in the
following first-order process:
The d statistic is roughly: d = 2 – 2p, where p lies between +1 and -1. This statistic
lies between 0 and 4.
3.8.4.6 Multicollinearity Test
The study conducted a multicollinearitytest in order to determine the state of
collinearity within the independent variables. The variable inflation factor (VIF) was
used to test the presence of multicollinearity in the independent variable. The VIF
measures how highly correlated each independent variable is with the other
predictors in the model. With VIF > 5 there is an indication that multicollinearity
may be present; with VIF > 10 there is certainly multicollinearity among the
variables. When the values higher than 10 for predictor, this implies large inflation
of standard errors of regression coefficient as a result of this variable being in the
model (Gujarati, 2004).
Where 1- = Tolerance (T) which measures the influence of one independent
variable on all other independent variables is the coefficient of multiple

31
determination when is regressed on the k-1 remaining independent variables.
If multicollinearity is found in the data, centering the data (that is deducting the
mean of the variable from each score) might help to solve the problem. However, the
simplest way to address the problem is to remove independent variables with high
VIF values.
3.8.4.5 Heteroscedasticity Test
When running OLS regression it is assumed that residuals/errors have the same but
unknown variance (constant variance or homoscedasticity), the violation of this
assumption leads to a problem known as heteroscedasticity. The presence of
heteroskedasticity leads to inefficient OLS estimators, consequently, inefficient
regression predictions and invalid tests of hypotheses (t-test, and F-test).
White’s test was used to test for heteroskedasticitybecause of its implementation
simplicity and independence of normality assumptions. OLS assumes that;
For all j. That is, the variance of the error term is constant. (Homoscedasticity). If the
error terms do not have constant variance, they are said to be heteroskedastic. The
test can be performed as follows;
Consider the following Linear Regression Model (assume there are two independent
variable).
32
For given data, estimate the regression model and obtain the residuals ei’s.
By running the following regression model to obtain squared residuals from original
regression on the original set of independent variable, square value of independent
variables and the cross-product(s) of the independent variable(s) such as;
Find the R2 statistics from the auxiliary regression in equation 14.Test the statistical
significance of
n×R2∼χ2df,……………………………………………………..…………………... 15
Under the null hypothesis of homoscedasticity or no heteroscedasticity, where df is
number of regressors in equation 14.
If calculated chi-square value obtained in (15) is greater than the critical chi-square
value at chosen level of significance, reject the hypothesis of homoscedasticity in
favour of heteroscedasticity.
33
CHAPTER FOUR
PRESENTATION AND DISCUSSION OF RESULTS
4.1 Introduction
This chapter presents and discusses the analysis of the findings in regards to the
impact of taxation on economic growth of Tanzania between 1996 and 2019. The
findings include Summary descriptive statistics, and time series estimations.
4.2 Summary Descriptive Statistics
The summary statistics of the independent variables (taxation revenue, capital
accumulation, and government spending) and their impact on economic growth
inTanzania is presented in Table 4.1 and discussed thereafter.
Table 4.1: Summary Statistics of Taxation Revenue, Capital Accumulation, and

Government spending and Economic Growth Rate in Tanzania
Variable Mean S.D. Min Max
GDPRate 6.23 1.39 3.53 8.46
Capital accumulation (CINV) 28.4 8.04 14.9 37.6
Government spending (GOV) 9.90 1.60 7.54 13.1
Taxation Revenue (TAX) 11.4 1.06 9.50 13.3
Summary Statistics, using the observations 1996 – 2019, No. of observations =23
Source: Study’s Analysis (2020)
The results in table 4.1 showed that the mean value of Growth Domestic Product
(GDP) of Tanzania for the period of observation is 6.23%, this mean value is within
the acceptable level for developing countries (Amadeo, 2019). The standard
deviation for GDP rate is 1.39%; which indicates a slight dispersion in GDP from
one year to another, with the minimum GDP being 3.53% and the maximum being
8.46%.Some of the reason for the slightly higher GDP rate is Tanzania as a
34
developing country, the slightly higher GDP rate is attributed to the fact that
developing countries are starting from a lower base income level than their
developed country counterparts.
Then there is also the presence of Official Development Assistance (ODA) from
developed to developing countries; higher population growth rates in developing
countries; and the general shift of labour from agriculture to faster growing service
and industries that's taking place in developing countries. Foreign Direct Investments
(FDI) is also growing faster into developing countries than into developed
economies. General trade liberalization is also helping developing countries achieve
a higher GDP in comparison to developed countries.
The mean value of Capital accumulation (CINV) of Tanzania for the period of
observation is 28.4%, the standard deviation for CINV is 8.04%; this shows a big
deviation from the CINV mean from one year to another, the minimum and
maximum mean of CINV in the time of observation was 14.9%, and 37.6%
respectively. The average capital accumulation for Tanzania for the period is slightly
higher than the average capital accumulation for the world which for the last ten (10)
years is 24.4%, meaning Tanzania has done well in the aspect of capital
accumulation which is an important aspect in economic development of the country
since increased capital accumulation means more disposable income for investment.
The mean for government spending (GOV) measured as a percentage of GDP for the
period of observation is 9.90%, the standard deviation for GOV is 1.60%; this shows
only a slight deviation from the GOV mean in the years of observation, the
35
minimum and maximum mean of GOV in the time of observation was 7.54%, and
13.1% respectively. The level of government spending in Tanzania, is consistent
with levels of government spending for most developing nations, since on average
the level of government spending for less developed countries is 10%, while for high
income countries, the level is much higher (Mauro et al, 2015; OECD, 2020). While
a lower government spend may seem as prudential, it could however mean that the
government is unable to provide important services to the public.
The mean value of Taxation Revenue (TAX) is 11.4%, with the standard deviation
for TAX being 1.06%, this shows taxation revenue in the years of observation was
not far off the mean value whereby there was a minimum of 9.50%, and the
maximum TAX of 13.3%.The average tax revenue per GDP ratio of 11.4% is on the
low end of many other developing countries since typically, a developing economy
collects about 15 percent of GDP in taxes, compared with the 40 percent collected by
a typical advanced economy (Akitoby, 2018). So the country still has much room for
improvement as much as tax collection is concerned.
4.2.2 Correlation Matrix
Statistically, correlation is used to denote association between two quantitative
variables, the association is assumed to be linear, and that is one variable increases
or decreases a fixed amount for a unit increase or decrease in the other. Correlation
coefficient measures the degree of association between the respective variables;
thecoefficients are measured on a scale that varies from +1 through 0 to -1. When
one variable increases as the other increases the correlation is positive, when one
36
decreases as the other increases it is negative. The result of correlation analysis in
order to measure the degree of association between the variables for this study is
shown in table 4.2.
Table 4.2: Correlation Matrix

GDPRate CINV GOV TAX
GDPRate 1.0000
CINV 0.4543 1.0000
GOV 0.2984 0.3197 1.0000
TAX -0.3922 -0.4296 -0.3937 1.0000
Statistically, correlation is used to denote association between two quantitative
variables, the association is assumed to be linear, and that is one variable increases
or decreases a fixed amount for a unit increase or decrease in the other. Correlation
coefficient measures the degree of association between the respective variables; the
coefficients are measured on a scale that varies from +1 through 0 to -1. When one
variable increases as the other increases the correlation is positive, when one
decreases as the other increases it is negative.
Apart from showing the association between independent and dependent variables,
correlation analysis can also indicate whether the independent variables are too
strongly correlated which could be a sign of multicollinearity problem within the
independent variables of interest. Fortunately, the magnitude of the correlation
coefficients in this case is less than 0.80, suggesting that there is no multicollinearity
within the variables.

37
Furthermore, to interpret the strength and direction of association between the
independent and dependent variables of this study; it should be noted that when the
coefficient has a value of -1 or 1 it means there is a perfectly negative or positive
relationship between independent and dependent variable respectively. Likewise a
value of -1 to -0.5/1 to 0.5; means stronger negative relationship/stronger positive
relationship, -0.5 to -0.1/0.5 to 0.1; means a weaker negative relationship/weaker
positive relationship, and a value of -0.1 to 0.1; means little to no relationship
between the variables.
The finding in table 4.2 show a weaker positive relationship between capital
accumulation and economic growth of Tanzania at 0.45, meaning an increase of
capital accumulation leads to a weak increase in economic growth of Tanzania.
There is also a weaker positive relationship between government spending and
economic growth of Tanzania at 0.3, meaning an increase in government spending
leads to a small increase in the economic growth of Tanzania. The result also shows
a weaker negative relationship between annual tax increase rate and economic
growth of Tanzania at 0.39, meaning an increase in tax leads to a small decrease in
the economic growth of Tanzania.
4.3 Summary of Diagnostic Tests’ Results
4.3.1 Linearity Test
One of the crucial assumptions when performing OLS regression is the relationships
between the independent and the dependent variable should be linear; to test for the
linearity assumption, the study used scatterplots as shown in Fig. 4.1;

38
GDPRate
40
30
CINV
20
10
14
12
GOV
10
8
14
12
TAX
10
4 6 8 10 20 30 40 8 10 12 14
Figure 4.1: Linearity Test (Scatter Plots)
Figure 4.1 indicates that the relationship between the independent variables (TAX,
GOV, and CINV) and the dependent variable (GDPRate) is nearly linearsince no
data point is predominantly that far from the rest of the data set.
4.3.2 Normality test
In line with multiple regression assumptions, residuals of the data have to be
normally distributed, in order to test this assumption the study used the JB (Jacque-
Bera) Test of normality; the result of the test is shown in table 4.3; the hypothesis for
the normality test was;
H0:Normally Distributed
H1:Not Normally Distributed

39
Table 4.3: Normality Test (Jarque-Bera test)

Test for normality of uhat1:
Jarque-Bera test = 1.29558, with p-value 0.523201
In line with multiple regression assumptions, residuals of the data have to be
normally distributed, while data does not need to be perfectly normally distributed
for the tests to be reliable; however normal distribution should peak in the middle
and is symmetrical about the mean. To test for this multiple regression assumptions
of normality for this study, the JB (Jacque-Bera) Test of normality was used through
the Gretl software, the decision criteria when using the JB (Jacque-Bera)
Test of Normality is to reject the null hypothesis of normality when the test statistics
has a p-value larger than the optimal p-value which is 0.05. the result in table 4.3
revealed that the test statistics for the Jarque-Bera test is 1.29558, with p-value
0.523201, since the p-value is larger than 0.05 then the test has failed to reject the
null hypothesis that the residuals in the model are normally distributed, concluding
that the residuals are normally distributed. This conclusion indicate that the data for
the variables used for this study which includes economic growth (GDPRate),
Taxation, government spending and capital accumulation can be used for various
appropriate parametric tests to ensures effective multiple regression estimates.
4.3.3 Autocorrelation Test
OLS assumes that the error terms between various observations are uncorrelated.
nevertheless, when dealing with time series it is not uncommon for the error terms
at a particular period to correlate with the error terms in preceding periods, hence
40
autocorrelation problem, to test for this OLS assumption, this study used the Durbin-
Watson test for autocorrelation under the hypothesis;
H0: no autocorrelation, and
H1: There is serial autocorrelation.
Table 4.4: Autocorrelation Test (Durbin-Watson Test)

Durbin Watson d- statistical(4, 24) =1.666693
OLS assumes that the error terms between various observations are uncorrelated.
nevertheless, when dealing with time series it is not uncommon forthe error terms at
a particular period to correlate with the error terms in preceding periods, hence
autocorrelation problem, when the disturbance term exhibits serial correlation, the
value of the standard error of the parameter estimates is affected and the predictions
based on ordinary least square estimates will be inefficient. To test for this OLS
assumption, this study used the Durbin-Watson test for autocorrelation, under the
Durbin-Watson test, the rule of thumb is values between 1.5 < d < 2.5 show that
there is no autocorrelation in the data. The test result in table 4.2.3 show that the
Durbin-Watson’s test statistics value is 1.667, since the value lies between 1.5 < d <
2.5, it means the test failed to reject the null hypothesis that there is no
autocorrelation, concluding that there is no autocorrelation problem within the
variables, hence values of the estimated parameters are correct and unbiased.
4.3.4 Heteroskedasticity Test
Under OLS regression, it is assumed that residuals/errors have the same but
unknown variance otherwise known as homoscedasticity, the violation of this

41
assumption leads to the problem of heteroscedasticity. To test for this assumption,
the study utilized White’s test for heteroskedasticity, under the following the
hypothesis;
H0:homoscedasticity
H1: heteroskedasticity
Table 4.5: Heteroskedasticity (White’s test)

Test statistic for the heteroskedasticity P-Value
12.77 0.1733
Under OLS regression, it is assumed that residuals/errors have the same but
unknown variance otherwise known as homoscedasticity, otherwise it is
heteroskedasticity. With heteroskedasticity present it means the model coefficients
estimated using ordinary least squares (OLS) are biased and error terms are inflated,
therefore the estimation of their variance is not reliable. In addition, since the
standard errors are biased when heteroskedasticity is present, this in turn leads to
bias in test statistics and confidence intervals. It was therefore necessary to perform a
test for heteroscedasticity.
The result in Table 4.5 revealed that the test statistic for the heteroskedasticity test is
12.7698 which have a p-value of 0.173303; this means that the test failed to reject
the null hypothesis of homoscedasticity because the tests’ p-value of 0.17 is greater
to 0.05, concluding that there is no problem of heteroskedasticity within the

42
variables.
4.3.5 Multicollinearitytest
Multiple regressions assume that the independent variables are not highly correlated
with each other, to test this assumption; the study used the variable inflation factor
(VIF). The VIF measures how highly correlated each independent variable is with
the other predictors in the model. With VIF > 5 there is an indication that
multicollinearity may be present; with VIF > 10 there is certainly multicollinearity
among the variables. The result of the test is shown in table 4.6.
Table 4.6: Multicollinearitytest (VIF test)

. estat vif
Variable VIF 1/VIF
TAX 1.35 0.742219

CINV 1.27 0.788624
GOV 1.22 0.817170
Mean VIF 1.28
Multiple regressions assume that the independent variables are not highly correlated
with each other, since the high correlation of the explanatory variables can lead to
imprecise estimation of the regression and slight fluctuations in correlation may lead
to large differences in regression coefficients, to test the multiple regression
assumption on multicollinearity, the study used the variable inflation factor (VIF).
The VIF measures how highly correlated each independent variable is with the other
predictors in the model. When VIF greater than 5 there is a possibility of
multicollinearity; but when VIF is greater than 10 it is no longer a possibility but

43
rather certainty that there is multicollinearity among the variables. As a rule of
thumb, for standardized data, mean VIF>10 indicate harmful collinearity. However,
the result from Table 4.6 shows that mean VIF was 1.28, meaning that the
independent variables are not highly correlated.
4.3.6 Unit Root Test
To avoid spurious estimation results this study conducted a unit root test through
Augmented Dickey Fuller test (AIC) criterion; the hypothesis for a unit root test are;
H0:non-stationary data
H1:stationary data
Table 4.7a: Unit Root Testat Level (Augmented Dickey Fuller test)
Variable Test Statistic Mackinnon P value (zt)
GDPRate -3.395 0.0111
TAX -2.216 0.2006
CINV -1.112 0.7104
GOV -1.762 0.3993
Table 4.7b Unit Root Testat first Different (Augmented Dickey Fuller test)
Variable Test Statistic Mackinnon P value (zt)
GDPRate - -
TAX -3.721 0.0038
CINV -4.700 0.0001
GOV -5.925 0.0000
Stationary of a time series is crucial for the application of various econometric
techniques, if a time series is non-stationary, one can study its behaviour only for the
time period under consideration. Non-stationary data implies that trend is not
deterministic, hence forecast tend to be unreliable. Augmented Dickey-Fuller test

44
was effectively used to test for stationary for each variable, simply because ADF test
can be used with serial correlation and it can handle large complex set of time series
models. (Epapha, 2017).
The result in table 4.7a shows that the p-value for GDPRate is 0.0111, this p-value is
less than 0.05; which means the test rejected the null hypothesis of non-stationary.
On the other hand, the result also shows the p-values for the rest of the variable are
superior to 0.05, which means the test failed to reject the null hypotheses in the case
of CINV, GOV and TAX, thus necessitating the need for testing their stationary in
first difference. By introducing the first difference to all the variables the result in
table 4.7b shows that the p-values for CINV, GOV and TAX are 0.0001, 0.0000, and
0.0038 respectively, these p-values are all less than 0.05; which means by testing in
the first difference, the ADF test rejected the null hypothesis of non-stationary for all
the three variables. Overall, the time series data for GDPRate is stationary at level,
implying that the time series data on Gross Domestic Product growth rate is
integrated of order zero (0) while the annual time series data on GOV, CINV, and
TAX are all stationary at first difference, implying that they are integrated of order
one (1).
The finding with respect to government spending, capital accumulation and total tax
revenue substantiates the theoretical assertion that most economic time series are
usually not stationary at level, but they attain stationarity after first differencing. The
fact that these variables were not stationary in level but rather into their first
difference is indicative of possible long – run relationship between the variables; this
necessitated the determination of the long-run correlation between these variables

45
using a cointegration test.
4.3.7 Cointegration Test
When variables are not stationary in their level, they can be integrated by order one,
and when variables are intergratedof order one it is suggestive of the existence of the
long-run relationship between the variables. To clarify the existence of long-run
equilibrium relationship between two variables a cointegration test was performed
through a trace test, the result of the test is shown in table 4.8. The hypothesis for the
tests was;
For r=0;
H0; No cointegrating equation
H1: There is at most onecointegrating equation
For r=1;
H1: There is at most two cointegrating equation
For r=2;
H1: There is at most 3 cointegrating equation
For r=3;
H1: There is is at most 4 cointegrating equation
Table 4.8: Cointegrating Test (Trace Test)

Rank Trace test statistics P-value
0 71.379 0.0000***
1 32.384 0.0239***
2 11.349 0.1937
3 3.655 0.0560**
Johansen test, Lag order up to 2, *** , ** shows the statistical significance at 1% and 5% level of
46
significance respectively. No. Of observation =22

When variables are not stationary in their level, they can be stationary after first
difference, and when variables are intergratedafter first differenceit is suggestive of
the existence of the long-run relationship between the variables. To clarify the
existence of long-run equilibrium relationship between two variables a cointegration
test was performed through a trace test.
When trace statistics is used to determine the number of cointegration equations, the
decision criteria is a null hypothesis of ‘r’ cointegrating relations is rejected against
the alternative of ‘n’ cointegrating relations, if the tests’ p-value is less that the
significance p-value of 0.05. From table 4.8 the test’s p-value is less than 0.05 in two
occasions, first when r=0 and when r=1, this means the null hypothesis of no
cointegrating relation at r=0 is rejected in favour of at most one cointegrating
equation, furthermore, the null hypothesis of no cointegrating relation at r=1 is
rejected in favour of at most two cointegrating equations. Therefore, the trace test
confirmed the existence of at most two long-run equations, hence, the existence of
long run relationship.
4.3.8 Causal Effects (Granger Causality test)
According to Sprites, Glymour, and Scheines, (2000), in statistics, when the value of
one event, or variable, increases or decreases as a result of other events, it is said
there is causation. To test for causality between total tax revenue (TAX) and
economic growth of Tanzania the granger causality test was used; the hypothesis for
the test is;

47
(H0)1: the lagged values of tax (TAX) do not cause GDPRate.
(H1)1the lagged values of tax (TAX) do cause GDPRate.
(H0)2: the lagged values of GDPRate do not causeTAX.
((H2)2: the lagged values GDPRate do causeTAX
Table 4.9: Causality Test (Granger Causality Wald TESTS)

Dependent variable: GDPRate
Excluded Chi-sq. df P-value
CINV 1.1834 2 0.553
GOV 4.2679 2 0.118
TAX 0.09993 2 0.951
ALL 5.2215 6 0.516
Dependent variable: TAX

GDPRate .73276 2 0.693
CINV .08319 2 0.959
GOV 9.143 2 0.010 ***
ALL 11.829 6 0.066
Dependent variable: CINV
GDPRate 4.4623 2 0.107

TAX 12.306 2 0.119
GOV 4.2608 2 0.540
ALL 11.829 6 0.099
Dependent variable: GOV

GDPRate 24.895 2 0.000***
CINV 27.591 2 0.000***
TAX 3.3524 2 0.187
ALL 47.149 6 0.000***
Vargranger, separator(1), Granger causality Wald tests, Lag order up to 2, *** shows the
statistical significance at 1% level of significance
4.3.9 The Causal Effect of Total Tax Revenue to Economic Growth of Tanzania
According to Heine, H., (1970); Causality is a genetic connection of phenomena
through which one thing (the cause) under certain conditions gives rise to, causes
something else (the effect). Causality is an active relationship, a relationship which
brings to life something new, which turns possibility into actuality. A cause is an
48
active and primary thing in relation to the effect. But "after this" does not always
mean "because of this". In statistics, when the value of one event, or variable,
increases or decreases as a result of other events, it is said there is causation. To test
for causality between total tax revenue (TAX) and economic growth of Tanzania the
granger causality test was used.
Result in Table 4.9, it is shows that the p-value for TAX is 0.951; this value is
greater than 0.05 which means the test failed to reject the null hypothesis that tax
(TAX) does not cause GDPRate. Meaning there is no causal effect from the tatal tax
revenue to economic growth. On the other hand, the finding also indicates that the p-
value of GDPRate is 0.693, since this value is greater than the 0.05 value, it means
the test also fails to reject the null hypothesis that GDPRate, do cause total tax
revenue. So, concluding that there is no causal relationship between total tax revenue
and economic growth of Tanzania in the period of observation of this study.
This result is inconsistent with Szarowská (2013) report which posits that there is
bidirectional causality between tax and GDP growth, meaning not only does tax
granger cause GDP growth but also GDP growth also granger causes tax, and also a
report by McBride, W (2012) who posits a unidirectional causality from GDP
growth to changes in taxes. The fact that this study has not found evidence of
causality between tax and economic growth in Tanzania, could be a result of
working with total tax revenue rather than individual taxes from which most of the
other studies looking for causality have worked with, so more studies in the topic,
especially on individual taxes, are highly encouraged.

49
While the other microeconomic variables used in this study were not of priority
when examining the causality relationship, however, the finding indicates that the p-
value of GDPRate and capital accumulation (CINV) on government spending
(GOV) are 0.000 and 0.000 respectively, both p-values are less than the 0.05,
meaning in this case the null hypotheses are rejected and hence confirms that both
economic growth (GDPRate) and capital accumulation causes government spending.
Furthermore, government spending has a p-value of 0.010 on total tax revenue, since
the p-value is less than 0.05; it means the null hypothesis of GOV not causing TAX
is rejected hence confirming that government spending does cause total tax revenue.
4.3.9.1 Stability Test of the VAR Model
To ensure that the relevant estimation results are effective, it is required that the
model is stable, and the condition of the model stability is the reciprocal of the
modulus of all the roots are in the unit circle. If the reciprocal of the roots exists
outside the unit circle, then the model may not be stable, and the model’s estimates
may be ineffective. The unit root diagram of the model is shown in Figure 4.2.
50
Roots of the companion matrix
1
0.261
.5
0.380
Im aginary
0 0.094
0.639 0.244
0.094
-.5
0.380
0.261
-1
-1 -.5 0 .5 1
Real
Points labeled with their distances from the unit circle
Figure 4.2: VAR Stability Condition Test

4.3.9.2 Parameters Stability Test (Recursive Residuals)
Time series regression/estimation assumes that the coefficients estimates are stable
over time. To test that assumption the study used the sum of recursive residuals
parameters’ stability test, which basically test whether there are structural breaks in
the residuals, the result of the test is shown in figure 4.3, the hypothesis for the test
is;
H0:no parameter instability
H1: parameter instability

51
CUSUM lower
upper
CUSUM
0 0
2001 2019
Year
Figure 4.3: Parameters’ Stability Test Result (sum of Recursive Residuals)

4.3.9.3 Stability test of the VAR Model
To ensure that the relevant estimation results by the VAR model are effective, it is
required that the model is stable, and the condition of the model stability is that the
reciprocal of the modulus of all the roots are in the unit circle. If the reciprocal of the
roots exists outside the unit circle, then the model may not be stable, and the model’s
estimates may be ineffective. According to the results in Figure 4.2, the reciprocal of
the model’s roots is in the unit circle, so the VAR model is stable and therefore the
model estimates are effective.
4.3.9.4 Parameters Stability Test (Recursive Residuals)
Time series regression/estimation assumes that the coefficient estimates are stable
over time. This study used the square roots of the recursive residuals to check the
52
model assumptions of normality and homoscedasticity, and other aspects of model
misfits such as change of regime, outliers, and omitted predictors, in place of plots
based on ordinary residuals. The parameters stability test basically tests whether
there are structural breaks in the residuals, according to the result in figure 4.3 the
cumulative sum square line is within the 95% confidence bands around the null; the
test fails to reject the null hypothesis of no parameters instability at 5% level,
elaborating that the test fails to reject the null hypothesis of no parameter instability,
meaning the VAR parameters are stable and consequently effective.
4.4 Vector Error Correction (VEC) Model Estimation
Vector Error Correction (VEC) Model was used to determine the short and long run
relationship between taxation and economic growth in Tanzania in line with
Balcazar (2017) who posits that VEC models are the best suited for analysing
multivariate time series. Furthermore, the variables were co-integrated of order one,
the cointegration test revealed that there exist at most two (2) cointegrating equations
between the variables for this study making the selection of the VECM even more
compelling. The Vector Error Correction (VEC) Estimation results are shown in
4.10 and 4.11;
Table 4.10: Long Run Estimation Result

Variable Coefficient Std.Error T-Statistic P-value
d-GOV -0.5298055 0.118154 -4.48 0.000***
d-TAX 0.1439203 0.1125362 1.28 0.201
d-CINV 0.2037802 0.0416661 4.89 0.000***
cons -0.3153978 - - -
Dependent Variable: GDPRate, Observations 1998-2019 (21)
*** = Significant at 1%
53
Table 4.11: Short run Estimation Result

Vector error-correction model
Sample: 1999 - 2019 No. of obs = 21

AIC = 14.54708
Log likelihood = -125.7443 HQIC = 14.83853
Det(Sigma_ml) = 1.866639 SBIC = 15.89003
Equation Parms RMSE R-sq chi2 P>chi2
D_dGDPRate 6 1.42162 0.8291 72.77052 0.0000

D_dGOV 6 .766532 0.7608 47.70915 0.0000
D_dTAX 6 1.30621 0.1016 1.696735 0.9454
D_dCINV 6 3.05231 0.4616 12.86038 0.0453
Coef. Std. Err. z P>|z| [95% Conf. Interval]
D_dGDPRate
_ce1
L1. -2.082348 .4158788 -5.01 0.000 -2.897456 -1.267241
dGDPRate
LD. .2350031 .2291632 1.03 0.305 -.2141485 .6841548
dGOV
LD. -.3658746 .2482439 -1.47 0.141 -.8524237 .1206745
dTAX
LD. .0784663 .289924 0.27 0.787 -.4897743 .646707
dCINV
LD. .265587 .1052925 2.52 0.012 .0592176 .4719564
_cons .0470768 .3126106 0.15 0.880 -.5656286 .6597822
From the VECM result in Table 4.10, it is indicated that economic growth rate
(GDPRate) is positioned as the dependent variableand in interpreting the results, the
signs of the coefficients from the cointegration equation are reversed in the long-run.
4.4.1 The Effect of total tax revenue on the Economic Growth of Tanzania
The result in table 4.10 shows that in the long-run, taxation has a positive impact on
the economic growth rate at 0.144 that is in the long run each unit-point increase in
taxation will also cause an increase of 0.144 unit-points in the economic growth rate
54
of Tanzania, on average ceteris paribus. Likewise, the result in table 4.11 indicates
that the coefficient estimate of taxation on GDPRate is 0.08 unit-points; meaning in
the short – run taxation has a 0.08 unit-points statistically insignificant positive effect
on GDPRate in Tanzania on average ceteris paribus. Accordingly, the result also
indicates that taxation positively affects the economic growth of Tanzania in both the
short and long run. This finding is not in support of the finding by Chiumia and
Simwaka, 2012; Edame and Okoi, 2014; Lawrence, 2015 to name a few who all
reports that taxation negatively affects the economic growth in the respective
countries of study.
On a different note, the study is in support of many other authors who have reported
a positive relationship between taxation and economic growth (Emmanuel, 2013;
Sekou, 2015; Wisdom, 2014). Overall, the purpose of government among others is to
provide basic amenities, protect the lives and property of the citizens, and create an
enabling environment for individuals and corporate organizations to strive. However,
for the government to carry out this responsibility, it needs to mobilize revenue
through taxation of the citizens and corporate organization. Thus the whole essence
of tax is to generate revenue which could be used to advance the welfare of the
citizen and to regulate the economy (fiscal policy). While taxation plays a significant
role in income redistribution, protection of the weak and infant industry, the revenue
generated through it plays a crucial role in promoting economic growth and
development.
The level of economic growth and development of any economy depends on the
55
amount of revenue generated and channelled towards the development of the
country; one of the sure ways of generating revenue is through tax. A country’s tax
system is a major determinant of the macroeconomic indexes for developed and
developing economies, however, if the tax is poorly administered the revenue
generated will not contribute adequately to economic growth and development,
perhaps this is the reason studies in the relationship between tax and economic
growth have revealed a number of conflicting findings.
Assessing this study’s result from both neoclassical and Keynesian theoretical
models perspectives, it is predicted that higher taxes reduce economic activity, even
though there is less agreement on the exact mechanisms that generate this result. On
the other hand, taxes may be a benefit for the economy because the taxes are the
basic source for financing public goods and services, and in this way can increase the
living standards and wealth of the whole society. If collected taxes are used
efficiently, providing public services can increase the productivity of human and
fixed capital in the private sector, and promote long-term economic growth.
Economic theory mostly links taxation to growth through its influence on the
decisions of economic agents that is taxation at least theoretically changes economic
decisions and can thereby affect economic growth. If a simple production function is
considered, it is evident that taxcan affect GDP and economic growth through its
impact on capital accumulation of both physical and human capital, and its effect on
the total factor productivity (TFP).
However, the Neo-classical view is that income and wealth must first be produced
56
and then consumed, meaning that taxes on the factors of production, such as capital
and labour, are particularly disruptive of wealth creation. Corporate taxes reduce the
incentive to invest and to build capital. When there is less investment, it means fewer
employment and lower wages, furthermore, when higher income is taxed at higher
rates, it negatively affect the returns to education, because it is expected that high
levels of education leads to higher income so when that is not the case the incentive
to build human capital through education is reduced.
4.4.2 The effect of Capital Accumulation on the Economic Growth of Tanzania
The result in table 4.10 indicates that other things constant in the long-run, capital
accumulation has a positive effect on the economic growth rate at 0.204 that is in a
long run each unit-point increase in capital accumulation will also cause an increase
of 0.204 unit-points in economic growth rate in Tanzania. On the other hand, the
result in table 4.11 also shows that the coefficient estimate of capital accumulation
on GDPRate is 0.27; meaning in the short – run capital accumulation has a 0.27 unit-
point statistically significant at 1% level positive effect on GDPRate in Tanzania on
average ceteris paribus. In unanimity, the results indicate that capital accumulation
has positive effects on the economic growth of Tanzania both in the short and the
long run.
Overall, according to Tuovila (2020), capital accumulation refers to an increase in
assets from investments or profits and is one of the building blocks of an economy; it
is the process of acquiring additional capital stock that is used in the productive
process. Capital accumulation can involve investment in physical fixed capital (e.g.
factories, machines), portfolio investment such as the purchase of bonds, shares, and
57
crypto currencies, investment in assets, such as housing. So a country that creates a
conducive environment for the prosperity of capital accumulation has a great chance
of positively affecting its economy since capital accumulation can occur through
reinvesting profit from the business, foreign direct investment (important for
developing economies with low capital basis), technological innovation which
increases the productivity of capital, increase in human capital such as better
educated workforce enables an increase in production possibility frontier,
discovering new sources of raw materials, such as oil and gas reserves, increased
level of savings, all this are important activities towards economic growth of a
country. In the Harod-Domar model of economic growth, a higher proportion of
income that is saved enables more investment and higher rates of economic growth.
(Though Keynesians note that higher savings are not always invested but can be
saved without investment).
4.4.3 The Effect of Government Spending on the Economic Growth of Tanzania
The result in table 4.10 shows that in the long-run, government spending has a
negative impact on the economic growth rate at 0.0.53units that is in a long run each
unit-point increase in government spending will also cause a decrease of 0.53 unit-
points in economic growth rate in Tanzania, on average ceteris paribus. On the other
hand, the result in table 4.11 also reveals that the coefficient estimate of government
spending on GDPRate is -0.37; meaning in the short-run government spending has a
0.37 unit-points statistically insignificant negative effect on GDPRate in Tanzania on
average ceteris paribus.
The finding of this study on the effect of government spending on the economic
58
growth of Tanzania are contrary to the Wagner’s law of state (1876) which holds
that for any country, public expenditure rises constantly as income growth expands.
The theory predicts that the development of an industrial economy will be
accompanied by an increase share of public expenditure in GDP. Furthermore, the
findings also both supports and conflicts with many other studies (Fölster and
Henrekson, 2001; Jin-young, and Ho, 2015; Saez et al, 2017), a classic example is a
study by Jong Chan Lee, Joong Won and Young Jei (2019) who when studying the
relationship between Government Expenditures and Economic Growth for China and
Korea revealed that government spending has both positive and negative effects on
economic growth depending on how and to what sector the government is spending
on. Overall, increased government spending is likely to cause a rise in aggregate
demand (AD). This can cause higher growth within the short-run. It may also
probably cause inflation.
Higher government outlay will have an effect on the supply-side of the economy
based on which area of government expenditure is increased. If spending is targeted
on raising the standard of infrastructure, this might improve productivity and growth
within the long-term aggregate supply. If spending is targeted on welfare benefits or
pensions, it may reduce inequality, but it could crowd out more productive private
sector investment. Spending on welfare benefits helps to reduce levels of inequality.
For example, benefits to the unemployed enable them to maintain a minimum
income and avoid absolute poverty. There is a potential higher welfare benefit which
could reduce incentives to work, but on the other hand, welfare benefits can also
help the labour market to function more efficiently.

59
Spending on pensions, an aging population requires higher government spending,
pensions, and health care spending. But pension spending has no impact on boosting
productivity. Spending on education and training if successfully targeted on
improving skills and education, government spending can increase labour
productivity and enable higher long-term economic growth. By spending on
infrastructure, it means higher spending on roads and railways can help remove
supply bottlenecks and enable greater efficiency, this can also boost long-term
economic growth. Spending on higher debt interest payments, since the government
has higher debt and higher bond yields, and then it can cause increased costs of
borrowing. This spending will go to investors and have no benefit for the economy.
4.4.4 Stability Condition test of the VECM Estimates
In order to check if the VEC Model is correctly specified, the study used the
vecstable test which checks whether the number of cointegrating equations has been
correctly specified. The result of the test is shown in table 4.12

60
Table 4.12: VECM Stability Condition Test

. vecstable
Eigenvalue stability condition
Eigenvalue Modulus
1 1
1 1
1 1
-.3947924 + .709172i .811656
-.3947924 - .709172i .811656
-.5199543 + .1783358i .549687
-.5199543 - .1783358i .549687
-.2620188 .262019
The VECM specification imposes 3 unit moduli.

Roots of the companion matrix

1
.5
Im ag in ary
0-.5
-1
-1 -.5 0 .5 1
Real
The VECM specification imposes 3 unit moduli
Figure 4.4: VECM Stability Test (eigenvalues of the Companion Matrix)
In order to check if the VEC Model is correctly specified, the study used the
vegetable test which checks whether the number of cointegrating equations has been
correctly specified. The companion matrix of a VECM with k endogenous variables
and r cointegrating equations has K-r unit eigenvalues. If the process is stable, the
61
moduli of the remaining r eigenvalues are strictly less than one. Because there is no
general distribution theory for ascertaining whether the moduli are too close to one
can be difficult. The result of the test in table 4.12 indicates that the model is stable
since the moduli of the remaining eigenvalues are strictly less than one; also because
the graph option was specified the vecstable plotted the eigenvalues of the
companion matrix. The test result in figure 4.4 indicates that none of the remaining
eigenvalues appears close to the unit circle. Consequently, the stability check does
not indicate that the model is misspecified meaning the estimation coefficients from
the VECM are effective.
4.4.5 The trends of Different Taxes such as Corporate Tax, PAYE, Taxes,
Excise Duty, Import Duties, VAT and other Domestic Taxes in Tanzania
Figure 4.5: Trends of PAYE, Corporation Tax and other Income Taxes

62
Figure 4.6: Trends of VAT, Import duty, excise duty, and other domestic taxes
and other Income Taxes
Apart from investigating the effect of total tax revenue on economic growth, the
study also wanted to examine the trend statistics for various taxes in Tanzania. By
capturing the trends the study was close to uncovering what events may have led to
differing trends in taxes throughout the observation period. The trend for PAYE
beginning in 1996 kept on a consistent increase until 2017 when it reached the all-
time high of 17.4% of the overall tax revenue. Moreover, from 2018 to 2019 there
was a sharp drop of PAYE taxes from 16.46% to 10.59% which was the all-time low
for almost ten years (2001-2019). This sharp decline may be attributed to the passing
of the 2016/17 Finance Act which saw some taxes sharply increase their contribution
to the overall revenue while others decreasing.
On corporate tax, in the first year of observation, its contribution to the overall tax
revenue was 10.79%, however, by the year 2001, it had become one of the least
63
contributing taxes at only 5.4%, perhaps because of the increase in revenue collected
from other taxes such as other income taxes which was having one of its best
performing years in the same year. After 2001, other income taxes had an
inconsistent upward movement up to 2016 when it dropped from 15% in 2014 to
10.54%, increased again to 15.12% in 2017 only to have the all-time drop to 1.2% in
2019. This huge drop can be attributed to tax reforms following the passing of the
2016/17 Finance Act. These reforms saw an abrupt and sharp increase in some of the
taxes such as excise duty, income taxes, and other domestic taxes perhaps to the
extent of belittling the contribution of corporate taxes.
With the exception of year 2001 when other income taxes increased up to 11.78%
for all the other years these taxes have been contributing for an average of 7% up
until 2018 when it increased from 9.5% to 15.3%, proving to be one of the taxes that
have increased as a result of the tax reforms following the passing of the 2016/17
Finance Act. Another tax that seems to have benefited from the tax reforms in the
2016/17 Finance Act is the excise duty, this tax had fared around 17% thorough out
the time of observation up until 2018 when it increased from 16.17% to 27.1%.
This huge increase is attributed to the rationalizations of the tax base include a
widening of the tax base of Excise Duty on financial services relating to the payment
of funds. Banks and Mobile Financial Providers for example M-Pesa, TigoPesa,
Selcom, etc. now impose excise duties and VAT on commissions charged on
financial transactions, meaning this reform has actually managed to increase the tax
collected through excise duty to within the same level as the tax collected through
VAT.
64
The introduction of VAT saw it become the leading tax category in contribution to
the total revenue collected at 23.9%, the tax kept increasing up to 2005 when it
reached the all-time high at 43.1%. Afterward, it has gradually lowered to an average
of 28% throughout the period of observation, and it still is the highest revenue
collection tax followed closely by excise duty. Other domestic taxes started as high
as 15.17% but eventually dropped to the 2% level in 2001, the tax remained at that
level up until 2018 when it rose from 2.5% to 10.2% in 2019, like the income taxes,
this sharp rise is also attributed to the tax reforms in 2016/17 Finance Act. Import
duty on the other hand has no much trend to report since it consistently operated on
the same level throughout the observation period.

65
CHAPTER FIVE
CONCLUSION
5.1 Introduction
This last chapter presents the summary of the findings, conclusions which were
reached, recommendations for stakeholders and their policy implications as well as
areas for further study.
5.2 Summary of the Study
This research intended to examine the impact of taxation on the economic growth of
Tanzania. The research includes six chapters; the first chapter gave the background
of taxation and the economic growth of Tanzania. Developed the gap to be bridged
and why it was significant for the gap to be filled, it indicated the scope of the study
and lastly developed three (3) specific objectives to be realized by the researches’
findings. These objectives were; to examine the effect of total tax revenueon the
economic growth of Tanzania, to examine the causal effect of total tax revenueon the
economic growth of Tanzania, and to examine the trends of different taxes such as
corporate tax, PAYE, taxes, excise duty, import duties, VAT and other domestic
taxes in Tanzania
To better understand the topic, the second chapter involved the review of several
studies on the topic of this research; the review included theoretical and empirical
studies from different authors, while there were studies indicating the existence of a
significant relationship between taxation and economic growth (Wisdom,
2014).Others reported no relationship between taxation and economic growth
(N'Yilimon, 2014; Ugwunta and Ugwuanyi, 2015). Furthermore, in order to better

66
understand the effect of taxation and economic growth the study reviewed several
theories which included Cost of servicetheory of taxation, Ibn Khaldun’s theory and
Neoclassical Growth Theory - (The Solow Growth Model).
The essence of the various theories has been the quest for the optimum taxation
where tax revenues are maximized for social welfare and economic growth. Adam
Smith regarded taxation as a means of sustaining the government. Ricardo provided
justification for capital tax which as part of factors of production (labour and capital)
is required (in part) to fund government activities. In its regulatory function, taxation
provides a mechanism to redistribute national income. In its catalytic role, taxation is
applied to increase the value of effective demand, stimulate investment and engender
Chapter three developed the research methodology for this study by adopting a
quantitative research approach, and causal research design; this is a research design
which attempts to reveal a cause and effect relationship between variables (Brüderl,
2015). The study area which is Tanzania was introduced in this chapter; it introduced
the data types and sources as secondary data collected from different reputable
sources. It is in this chapter where measurement of the variables was introduced as
well as the estimation model for estimating the time series data developed.
In chapter four, the appropriate diagnostic tests which includes the normality tests,
unit root tests, heteroskedasticity and autocorrelation test were carried out
successfully, the study presentedthe findings analysed quantitatively using statistical
softwares, which were Stata and Gretl; the study presented the descriptive statistics
67
summary, as well as estimation of the time series data whereby the findings
indicated that on average over the period of investigation the GDPratefor Tanzania
was 6.23%, capital accumulation (CINV)was 28.4%, government spending (GOV)
was 9.9%, and taxation revenue (TAX) was 11.4%.
Furthermore, the study found that in the long-run, capital accumulation and taxation
have positive impacts on economic growth rate while government spending has a
negative impact on economic growth rate, on average ceteris paribus. Furthermore,
the study revealed that in the short-run capital accumulation and government
spending have statistically significant positive impacts on economic growth while
taxation has an insignificant negative impact on economic growth of Tanzania. The
fifth chapter discussed the findings presented in chapter four.
5.3 Conclusion of the Study
This research had intended to examine the impact of taxation on the economic
growth of Tanzania. Based on annual data from 1996 to 2019 the effect as well as
the causal effect of total tax revenueand economic growth of Tanzania is analysed
using Vector Error Correction Model and the granger causality test; the following
main conclusions are obtained: Firstly, there is an insignificant positive effect of
total tax revenueon economic growth (GDPRate) of Tanzania both in the short and
long-run. Therefore we reject null hypothesis that there is a significant relationship
between taxation and economic growth.
Secondly, the granger causality test of variables in the VAR model shows that there
exists no causality between tax and economic growth of Tanzania; hence the
68
rejection of the hypothesis that there exist a significant causal relationship between
total tax revenue and economic growth. Thirdly, government spending has a negative
impact on economic growth rate both in the short and the long-run. Regardless of
this study’s finding, government spending may have both negative and positive
effect on the economic growth of the country depending on how and to what sector
the government is spending on. Fourthly, capital accumulation has positive effects
on the economic growth of Tanzania both in the short and the long run, capital
accumulation is an important aspect in economic development of the country since
increased capital accumulation means more disposable income for investment.
5.4 Recommendations and Policy Implications
Based on the findings on the effect of total tax revenueon economic growth of
Tanzania, the causal effect of total tax revenueon economic growth of Tanzania, and
the trends of different individual taxes, the following main recommendations are
given; Firstly, The finding shows that taxation has a positive impact on economic
growth rate of Tanzania both in the short and long-run, however the average tax
revenue per GDP ratio of 11.4% is on the low end of many other developing
countries.
A developing economy collects about 15 percent of GDP in taxes, compared with
the 40 percent collected by a typical advanced economy (Akitoby, 2018). So the
country still has much room for improvement as much as tax collection is concerned.
That being the case the government should institute appropriate systems which
broaden the tax bases for taxes with positive impact on the economic growth at the
69
same time reviewing for possible reduction of individual taxes that prove to have
negative effects on the economic growth.
Secondly, Fourthly, recent trends in taxes have shown a sharp increase in excise
duty, income taxes and other domestic taxes, meaning the tax reforms recently
introduced have served the purpose of increasing these taxes, however, the trends
have also shown a huge decrease in some of the taxes, specifically PAYE and
Corporate taxes, this should be alarming to the government since its indicative of
poor performance of the labour market as well as the private sector. The poor
performance of the corporate tax could be an indicator of the private sector not
making profit which can be detrimental to the short and long term economy of the
country, as a poor performing private sector means less employment and less tax to
the government. So the government should harmonize policies that may be
weakening the private sector and encourage private public partnership (PPP) for
better overall economic growth of the country.
Thirdly, the finding in this study shows that government spending has a negative
impact on economic growth rate of Tanzania both in the short and the long-run,
however, past studies such as Jin-young, and Ho, (2015) and Saez et al, (2017) have
indicated that government spending can have both positive and negative effects on
the economy of the country depending on how and to what sector the government is
spending on, so the fact that this study revealed a negative impact means that the
government may have focused more spending on sectors that do not have any
positive impact to the economy, that being the case, the government should pay
special attention to sectors that have huge potential of positively affecting the
70
economy of the country because with increased economy the government will be in a
much better position to serve other sectors that do not have direct contribution to the
economy of the country.
Fourthly, since capital accumulation matters greatly for a country’s standard of
living and growth, there is need for an increased interest in the mobilization of
domestic resource to finance savings, investment and consequently economic
growth. If savings can be improved, thereby leading to increased investment then,
then the country’s economic growth can be positively affected. Therefore, the
government should specifically pay particular attention to economic shocks, as well
as the investment climate in Tanzania so as to ensure macroeconomic stability and
5.5 Area for Further Study
The mixed findings revealed by the past studies, in addition to the findings of this
study, and the fact that this study found no causality between tax and economic
growth calls for future research directed towards individual taxes since some of the
studies that found causality were conducted on individual taxes.

71
REFERENCES
Alesina, A. & Ardagna, S. (2010). Large changes in fiscal policy: taxes versus
spending. Tax Policy and the Economy, 24, 35-68.
Amadeo. K. (2019). What is the Ideal GDP Growth Rate? How Fast Should the
Economy Grow?, Retrieved on 9th July, 2020 from: https://www.thebalance.
com/what-is-the-ideal-gdp-growth-rate-3306017.
Angahar, P. A. & Alfred, S. I. (2012). Personal income tax administration in Nigeria:
challenges and prospects for increased revenue generation from self-
employed persons in the society. Global Business and Economics Research
Journal, 1(1), 1-11.
Appah, E. & Oyandonghan, J.K. (2011). The Challenges of Tax Mobilization and
Management in the Nigerian Economy. Journal of Business Administration
and Management, 6(2), 128-136.
Bania, N., Gray, J. & Stone, J. (2007). Growth, taxes and government expenditures:
growth hills for U.S. states. National Tax Journal, 60(2), 193-204.
Barro, R. & Redlick, C. (2011). Macroeconomic Effects of Government Purchases
and Taxes. Quarterly Journal of Economics, 15369, 51-102.
Bhartia, H. L. (2009). Public Finance. New Delhi: Vikas Publishing House PVT
Ltd.
Chigbu, E. E. & Njoku, C. O. (2015). Taxation and the Nigerian Economy: (1994-
2012). Management Studies and Economic Systems, 2(2), 111-128.
Chigbu. E. E, Akujuobi, L. E. & Appah, E. (2012). An Empirical Study on the
Causality between Economic Growth and Taxation in Nigeria. Current
Research Journal of Economic Theory, 4(2), 29-38.

72
Dackehag, M., Hansson, A. (2012). Taxation of Income and Economic Growth: An
Empirical Analysis of 25 Rich OECD Countries. Lund University
Department of Economics. Working Paper No. 2012:6.
Ebipanipre, G. M & Confidence, J. I. (2014). Taxation as an instrument of economic
growth (The Nigerian perspective). Info. Knowl. Manage. 12(4), 4-53.
Ferede, E. & Dahlby, B. (2012). The impact of tax cuts on economic growth:
Evidence from the Canadian provinces. National Tax Journal, 65(3), 563-
594.
Folster, S. & Henrekson, M. (2001). Growth effects of government expenditure and
taxation in rich countries. Eur. Econ. Rev, 45, 1501–1520.
Gale, W., Krupkin, A. & Rueben, K. (2015). The Relationship between Taxes and
Growth at the State Level: New Evidence. National Tax Journal, 68(4), 919-
942.
Gemmell, N., Kneller, R., &Sanz, I. (2011). The Timing and Persistence of Fiscal
Policy Impacts on Growth: Evidence from OECD Countries. The Economic
Journal, 121(550), 33-58.
Heine, H. (1970) Werke, Briefwechsel, Lebenszeugnisse. Band 5. Reisebilder I.
1824-1828. Akademie-Verlag, Berlin. Editions du CNRS, Paris, S. 29.
Helms, L. J. (1985). The Effect of State and Local Taxes on Economic Growth: A
Time Series Cross-Section Approach. Review of Economics and Statistics,
67(4), 574-582.
Ibadin, P. O. & Oladipupo, A. O. (2015). Indirect taxes and economic growth in
Nigeria. Ekonomskamisaoipraksa, 2, 345-364.
Jin-young, H.; Ho, L.S., (2015); The Effectiveness of Government Expenditure on

73
Economic Growth and Volatility. Korean Assoc. Public Policy, 33, 85–105.
Judd, K. (1985). Redistributive Taxation in a Simple Perfect Foresight Model.
Journal of Public Economics. 28 (1), pp. 59-83.
Lee, Y., & Gordon, R. (2005). Tax Structure and Economic Growth. Journal of
Public Economics, 89(5-6), 10271043. DOI: 10.1016/j.jpubeco.2004.07.002
Macek, R (2014). The Impact of Taxation on Economic Growth: Case Study of
OECD Countries. Review of Economic Perspectives , 14(4), 2014, 309–328,
Machová, Z., Kotlán, I. (2013). World Tax Index: New Methodology for OECD
Countries 2000-2010. DANUBE: Law and Economics Review, 2013 (2),
165-179.
Maina, A. W. (2014). Income taxes and economic performance in Kenya. A research
paper submitted to the University of Nairobi, school of economics,
unpublished masters dissertation, Nairobi University, Kenya.
Mankiw, G., Romer, D., Weil, D. (1992). A Contribution to the Empirics of
Economic Growth. The Quarterly Journal of Economics. 107 (2), 407-437.
Mankiw, G., Weinzierl, M. &Yagan, D. (2009). Optimal Taxation in Theory and
Practice. Journal of Economic Perspectives, 23(4), 147-174.
Mauro, P., Romeu, R., Binder, A., & Zaman, A. (2015). A modern history of fiscal
prudence and profligacy. Journal of Monetary Economics, 76, 55-70.
McBride, W. (2012). What is the Evidence on Taxes and Growth? Tax Foundation,
Special Report 207.
Mitra, P., & Stern, N. (2003). Tax systems in transition. World Bank Policy
Research Working Paper 2947. Washington, DC: The World Bank.
Myles, G. (2000). Taxation and Economic Growth. Fiscal Studies, 21(1), 141-168.
74
Nerre, B., Dragojlović, A., Ranđelović, S., &Đenić, M. (2014). Tax reform in
Serbia: Experiences and perspectives. In Tax Reforms: Experiences and
Perspectives: Conference Proceedings, 20 Jun 2014 (pp. 79-96). Zagreb:
Institute of Public Finance.
OECD, (2016). Total tax revenue, in OECD Factbook 2015-2016: Economic,
Environmental and Social Statistics, OECD Publishing.
OECD, (2020), General government spending (indicator). doi: 10.1787/a31cbf4d-en
(Accessed on 17 November 2020)
Onakoya, A. B. & Afintinni, O. I. (2016). Taxation and economic growth in Nigeria.
Asian Journal of Economic Modelling, 4(4), 199-210.
Padovano, F. & Galli, E. (2001). Tax rates and economic growth in the OECD
countries (1950-1990). Economic Inquiry, 39(1), 44-57.
Pere, E. & Hashorva, A. (2011). Tax Systems in West Balkans Countries. Romanian
Economic and Business Review, 6(2), 80-94.
Reed, R. (2008). The robust relationship between taxes and U.S. state income
growth. National Tax Journal, 61(1), 57-80.
Romer, C. & Romer, D. (2010). The Macroeconomic Effects of Tax Changes:
Estimated Based on a New Measure of Fiscal Shocks. American Economic
Review, 100(3), 763-801.
Sáez, M. P., Álvarez-García, S. & Rodríguez, D. C., (2017) Government expenditure
and economic growth in the European Union countries: New evidence. Bull.
Geogr. -Socio-Econ. Ser., 36, 127–133.
Saima, S. A., Tariq, F. R, Muhammad, A. & Sofia, A. A. (2014). Taxation effects
on economic activity in Pakistan. J. Fin. Econ., 6(2):215-219.

75
Salami, G. O., Apelogun, K. H., Omidiya, O. M. & Ojoye, O. F. (2015). Taxation
and Nigerian economic growth. Research Journal of Finance and
Accounting, 10(6), 93-101.
Sekou, M. (2015). The impact of tax collection in achieving revenue targets: The
Directorate General of taxes of Mali case study. Theoretical Economics
Letters, 5(3), 403-409.
Solow, R. (1956). A Contribution to the Theory of Economic Growth. The Quarterly
Journal of Economics. 70 (1), 65-94.
Sprites, P., Glymour, C. & Scheines, R. (2000): Causation, Prediction and Search,
2nd edition. New York, SpringerVerlag.
Swan, T. (1956). Economic Growth and Capital Accumulation. The Economic
Record. 32(2), 334-361.
Szarowská, I. (2013). Effects of taxation by economic functions on economic growth
in theEuropean Union. In Jircikova, E., Knapkova, A., Pastuszkova, E.
(Eds.). Proceedings of the 6th International Scientific Conference: Finance
and the performance of Firms in Science, Education and Practice. 746-758,
Zlin: Tomas Bata University.
Tomljanovich, M. (2004). The role of state fiscal policy in state economic growth.
Contemporary Economic Policy, 22(3), 318-330.
Tosun, M. S. & Abizadeh, S. (2005). Economic growth and tax components: An
analysis of Tax changes in OECD. Applied Economics, 37(19), 2251-2263.
Ugwunta, O. D. & Ugwuanyi, U. B. (2015). Effect of distortionary and no
distortionary taxes on economic growth: Evidence from Sub-Saharan African
countries. J. Account. Taxation. 7(6), 106-112.

76
Wisdom, T. (2014). Tax Revenue and Economic Growth in Ghana: A Cointegration
Approach. MPRA, pp. 58532. Retrieved on 20 th March, 2020 from;
http://mpra.ub.uni-muenchen.de/58532/.
Yaya, K (2013). The structure of taxes and economic growth in Cote d’ivoire: An
econometric investigation. J. Res. Econ. Int. Fin. (JREIF) 3(2), 39-48.

77
APPENDICES
Appendix 1: Multiple Graphs Showing Trend Statistics of Different taxes in
Tanzania between 1996-2019
GDPRate PAYE CorporationTax

8.5 18 16
8 17 14
7.5 16
7 15 12
6.5 14 10
6 13 8
12
5.5 11 6
5 10 4
4.5 9
4 8 2
3.5 7 0
19972000200320062009201220152018 19972000200320062009201220152018 19972000200320062009201220152018
OtherIncomeTaxes Excisesduty VAT

16 28 44
15 26 42
14 24 40
13 38
12 22 36
11 20 34
10 18 32
9 16 30
8 28
7 14 26
6 12 24
19972000200320062009201220152018 19972000200320062009201220152018 19972000200320062009201220152018
OtherDomesticTaxes Importduty
16 18
14 17
12 16
10 15
8 14
6 13
4 12
2 11
0 10
19972000200320062009201220152018 19972000200320062009201220152018

AWADHI M. WILLIAM Tyr

Uploaded by

Copyright:

Available Formats

AWADHI M. WILLIAM Tyr

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

AWADHI M. WILLIAM Tyr

Uploaded by

Copyright:

Available Formats

AN ANALYSIS OF THE IMPACT OF TAXATION ON

ECONOMIC GROWTH IN TANZANIA: 1996-2019

A DISSERTATION REPORT SUBMITTED IN PARTIAL FULFILMENT OF

THE REQUIREMENTS FOR THE REQUIREMENTS OF THE DEGREE OF

MASTER OF SCIENCE IN ECONOMICS (MSc. ECON)

OF THE OPEN UNIVERSITY OF TANZANIA

by the Open University of Tanzania a dissertation entitled: ‘The impact of taxation

on the economic growth of Tanzania between 1996 and 2019” in partial

fulfilment of the requirements for the degree of Master of Science in Economics of

Open University of Tanzania.

No part of this dissertation may be reproduced, stored in any retrieval system, or

transmitted in any form by any means, electronic, mechanical, photocopying,

University of Tanzania in that behalf.

I, Awadhi M. William, declare that, the work presented in this dissertation is

of the requirement for the Degree of Master of Science in Economics (MSc-

This dissertation is dedicated to my parents whose inspiration encouraged me to

complete this work.

His endowments, I could never have accomplished it.

Much gratitude goes to my parents/mother/father/wife/fiancée/ for their

who was always there to provide the necessary assistance.

me immeasurable knowledge through their countless discussions.

insignificant positive effect of total tax revenueon economic growth (GDPRate) of

economic growth of Tanzania positively.It is the study’s recommendations that

INTRODUCTION AND background to the study..................................................1

1.1 Background of the Study..................................................................................1

1.2 Statementofthe Problem...................................................................................4

1.3 Objective of the Study......................................................................................6

1.3.1 Specific Objectives of the Research.................................................................6

1.4 Significance of the Study.................................................................................6

1.5 Scope of the Study............................................................................................7

2.2 Definitions of Concepts....................................................................................8

2.2.1 Economic Growth............................................................................................8

2.2.2 Fiscal Policy.....................................................................................................9

2.2.3 Government Spending......................................................................................9

2.2.4 Capital Accumulation.....................................................................................10

2.3 Theoretical Review........................................................................................11

2.3.1 Taxation Theories...........................................................................................11

2.3.2 Integration of Taxation into Economic Growth Theories..............................12

2.4 Empirical Review...........................................................................................13

2.5 Research Gap..................................................................................................17

2.6 Conceptual Framework..................................................................................18

2.6.1 Descriptions and Measurement of the Variables............................................19

3.2 Research Design.............................................................................................21

3.2.1 Research Approach........................................................................................21

3.3 Area of the Study............................................................................................21

3.4 Population of the Study..................................................................................22

3.4.1 Unit of Analysis.............................................................................................22

3.4.2 Sample Size....................................................................................................22

3.5 Types and Sources of Data.............................................................................22

3.5.1 Types of Data................................................................................................22

3.5.2 Sources of Data..............................................................................................22

3.5.3 Data Collections Methods..............................................................................23

3.6 Validity and Reliability..................................................................................23

3.6.1 Validity and Reliability..................................................................................23

3.7 Ethical Issues..................................................................................................24

3.8 Data Analysis.................................................................................................24

3.8.1 Theoretical Model..........................................................................................25

3.8.2 Estimation Model...........................................................................................26

3.8.3 Estimation Diagnostic Tests...........................................................................27