ADAMA SCIENCE AND TECHNOLGY UNIVERSITY

FACULITY OF BUSSINES AND ECONOMICS

DEPARTMENT OF ECONOMICS

IMPACT OF FOREGIN DIRECTINVESTMENT ON ECONOMIC

GROWTH IN ETHIOPIA

BY: - MANDEFRO GEDEFAW ALEMU

DECEMBER, 2016

ADAMA, ETHIOPIA
 IMPACT OF FOREGIN DIRECT INVESTMENT ON ECONOMIC
GROWTH IN ETHIOPIA

A THESIS SUBMITTED TO THE SCHOOL OF GRADUATE STUDIES OF ADAMA

SCIENCE AND TECHNOLOGY UNIVERSITY IN PARTIAL FULFILLMENT OF THE
REQIRMENT FOR THE DEGREE OF MASTER OF SCIENCE IN DEVELOPMENT
ECONOMICS

BY: - MANDEFRO GEDEFAW ALEMU

ADVISOR

Dr. ADEM KEDIR (PhD)

DECEMBER, 2016

ADAMA, ETHIOPIA
 APPROVAL OF BOARD OF EXAMINATIONS
COLLAGE OF BUSSINES AND ECONOMIC
ADAMA SCIENCE AND TECHNOLOGY
UNIVERISTY
As thesis advisor, I hereby certify that I have read and evaluated this research work contribution,
under my guidance, by Mandefro Gedefaw Alemu entitled “Impact of Foreign Direct
investment on Economic Growth in Ethiopia”. I recommend that it has to be submitted as
fulfilling the thesis requirement.

Dr Adem Kedir (PhD) ______________ ______________

Major advisor Signature Date

As member of board of examiners of the M.Sc. Thesis open defense examination, we

certify that we have read and evaluated the thesis prepared by Mandefro Gedefaw
Alemu entitled “Impact of Foreign Direct investment on Economic Growth in
Ethiopia” and examined the candidate. We recommended that it has to be accepted as
fulfilling the thesis requirements for the degree of masters of Science in Economics major
with Development Economics.

_________________ _______________ ________________

Name of Chairperson Signature Date

_________________ ________________ _________________

Name of Internal Examiner Signature Date

_________________ _________________________________
Name of External Examiner Signature Date

i
 ACKNOWLEDGMENT

Writing a master thesis is a challenge. First and foremost I would like to express my sincere
gratitude to God the almighty for his protection, guidance and abundant blessings throughout the
whole research process, with him everything is possible. What is certain is that this thesis would
not exist without the inspiration, knowledge, help and support from a large group of people. I
want o thank everyone who have motivated and contributed to this process.

Foremost, I would like to express my deepest gratitude to my advisor Dr. Adem Kedir for the
continuous support of my research, for his patience, motivation, enthusiasm, and immense
knowledge. His guidance helped me in all the time of research and writing of this thesis.

Besides my advisor I would also like to think my parents and friends for supporting me and
bearing with me throughout this process, especially Kalkidan Yenalem and Adugna Gedefaw
were always willing to help and give their best suggestions. My research would not have been
possible without their helps.

ii
 Table of Contents
ACKNOWLEDGMENT................................................................................................................. ii
LIST OF TABLE ........................................................................................................................... vi
LIST OF FIGURE......................................................................................................................... vii
LIST OF APPENDIXES.............................................................................................................. viii
LIST OF ACRONYMS ................................................................................................................. ix
ABSTRACT .................................................................................................................................... x
CHAPTER ONE ............................................................................................................................. 1
1. INTRODUCTION ............................................................................................................ 1
1.1 Background of the study .................................................................................................. 1
1.2 Statement of the problem ................................................................................................. 4
1.3 Hypothesis of the study .................................................................................................... 6
1.4 Objectives of the study ..................................................................................................... 6
1.4.1 General Objective ..................................................................................................... 6
1.4.2 Specific Objectives ................................................................................................... 6
1.5 Research questions ........................................................................................................... 7
1.6 Significance of the study .................................................................................................. 7
1.7 Scope of the study ............................................................................................................ 8
1.8 Limitation of the study ..................................................................................................... 8
1.9 Organization of the research paper................................................................................... 8
CHAPTER TWO ............................................................................................................................ 9
2 LITERATURE REVIEW ................................................................................................. 9
2.1 Introduction ...................................................................................................................... 9
2.2 Concepts and definitions of FDI ...................................................................................... 9
2.3 Theoretical Review ........................................................................................................ 10
2.3.1 Solow Growth Theory............................................................................................. 10
2.3.2 Neoclassical growth theory ..................................................................................... 11
2.3.3 Endogenous Growth Theory ................................................................................... 11
2.3.4 Economic Geography Theory ................................................................................. 12
2.3.5 Industrial Organization Theory ............................................................................... 12
2.3.6 Internalization Theory............................................................................................. 14
2.3.7 Product life Cycle Theory ....................................................................................... 14

iii
 2.3.8 Caves Theory .......................................................................................................... 15
2.3.9 Location Theory ...................................................................................................... 16
2.3.10 An Eclectic Theory ................................................................................................. 17
2.4 Views on the Impact of FDI ........................................................................................... 19
2.5 Role of FDI in Growth ................................................................................................... 20
2.6 Empirical Evidence on Impact of FDI ........................................................................... 20
2.7 Conceptual Frame Work ................................................................................................ 24
CHAPTER THREE ...................................................................................................................... 26
3. METHODOLOGY ................................................................................................................ 26
3.1 Econometric Model ........................................................................................................ 26
3.2 VAR Time Series Analysis ............................................................................................ 26
3.2.1 The choice of the variables ..................................................................................... 27
3.2.2 Testable hypotheses ................................................................................................ 27
3.2.3 Testing the stationary of time series ....................................................................... 28
3.2.3.1 Augmented Dickey-Fuller ...................................................................................... 28
3.2.3.2 Phillips Perron test .................................................................................................. 29
3.2.4 Model identification ................................................................................................ 29
3.2.4.1 Lag length selection ................................................................................................ 29
3.3 VEC Model .................................................................................................................... 30
3.3.1 Johansen Co-integration Test .................................................................................. 30
3.3.2 VEC Estimation ...................................................................................................... 30
3.3.3 Model Diagnostics .................................................................................................. 30
3.3.3.1 Test of Residual Autocorrelation (Breusch-Godfrey Serial Correlation) ............... 30
3.3.3.2 Normality of the Residuals ..................................................................................... 31
3.3.4 Granger Causality Test ........................................................................................... 32
CHAPTER FOUR ......................................................................................................................... 33
4. RESULT AND DISCUTION ................................................................................................ 33
4.1 Data Description ............................................................................................................. 33
4.2 Descriptive statistics ....................................................................................................... 33
4.3 VAR Modeling ............................................................................................................... 34
4.3.1 Testing the Stationarity ........................................................................................... 34
4.4 Model identification ....................................................................................................... 37

iv
 4.4.1 The VAR Order Selection Analysis ........................................................................ 37
4.5 Co-integration Analysis.................................................................................................. 38
4.6 Vector Error Correction Model ...................................................................................... 40
4.6.1 Model Estimation .................................................................................................... 40
4.7 Model Diagnostics.......................................................................................................... 43
4.7.1 Test of Residual Autocorrelation /serial correlation ............................................... 43
4.7.2 Normality Testing of Residual ................................................................................ 43
4.7.3 Lag Exclusion Test ................................................................................................. 44
4.8 Granger Causality test results......................................................................................... 45
CHAPTER FIVE .......................................................................................................................... 47
5. SUMMURY, CONCLUSION AND RECOMMENDATIONS ........................................... 47
5.1 Summary ........................................................................................................................ 47
5.2 Conclusion...................................................................................................................... 48
5.3 Recommendation ............................................................................................................ 50
REFERENCE ................................................................................................................................ 51
APPENDIX ................................................................................................................................... 59

v
 LIST OF TABLE
Table 1:-Summary of Main Findings of the Empirical Literature Reviewed………………25
Table 4.1:- Descriptive Statistics for Variables Used In the Study…………………………33
Table 4.2: - Unit Root Test Result (At Level)………………………………………………35
Table 4.3: - ADF and PP Stationary Test First Difference of RGDP, FDI, X and FCF ……35
Table 4.4: - ADF and PP Stationary Test At Second Difference Integration……………….36
Table 4.5:- VAR Model Lag Order Selection Criteria ……………………………………..37
Table 4.6: - Unrestricted Co-Integration Rank Test (Trace and Max- Eigen) value………..38
Table 4.7:- Wald Coefficient Test…………………………………………………………...42
Table 4.8: - Lag Exclusion Test …………………………………………………………….45
Table 4.9:- Granger Causality Test ………………………………………………………....45

vi
 LIST OF FIGURE

Figure 4.1:- Graphical Approaches for Unit Root Test Before Stationary Of
RGDP, FDI, X and FCF ………………………………………………………..34
Figure 4.2:- Graphical Approaches for Unit Root Test After Stationary Of
RGDP, FDI, X and FCF……………………………………………………….37
Figure 4.3:- Co-Integration Relation Graph ………………………………………………...39
Figure 4.4:- Inverse Root of AR Characteristics…………………………………………. ...44

vii
 LIST OF APPENDIXES

Appendix table A1:-GDP, FDI FCF and X Data ………………………………………..59

Appendix table A2:-Vector Auto Regression Estimation in Unrestricted VAR………….60
Appendix table A3:-Vector Error Correction Estimation...………………………………61
Appendix table A4:- Model Estimation of RGDP ……………………..…………………63
Appendix table A5:- Model Estimation of FDI……..…………………………………….63
Appendix table A6:- Vector Error Correction Residual Correlation LM...……………….64
Appendix table A7:-Vector Error Correction Residual Normality test………………..….65
Appendix table A8:- Roots of Characteristics Polynomial ……………………….. ……..65

viii
 LIST OF ACRONYMS

ADF ………………… Augmented Dickey Fuller

DCs………………......Development Countries

DI…………………..….Domestic Investment

EEA ………………..…Ethiopian Economic association

FDI………………....…Foreign Direct Investment

GDP…………………...Gross Domestic Product

IMF……………….…...International Monetary Fund

IV …………………….Instrumental Variable

LDCs ………………… Less Development Countries

MNC……………..……Multinational Corporations

MNE …………………..Multinational Enterprises

NBE………………...….National Bank of Ethiopia

OLI ……………….…....Ownership, Location, and Internationalization

PP ……………………..Phillies Perron

TNC………………….....Trans National Corporations

UNCTAD…………...… United Nations Conference on Trade and Development

EU………………………European Union

U.S. ……………….…....United State

US$ ……………….…....United states Dollar

WIR ……………….…...World Investment Report

OLS………………….… Ordinary Least Square

UNESCO……………......United Nations Educational, Scientific and Cultural Organization

WB…………………...…World Bank

ix
 ABSTRACT
This study examines the impact of foreign direct investment on economic growth of Ethiopia
using annually collected time series data for 1980/81-2014/15. Economic growth is
representations by real gross domestic product and foreign direct investment proxies by the
inflow of foreign direct investment. Other control variables such as fixed capital formation and
export have been incorporated. A unit root test was used to determine whether or not the data
was stationary. The Johansen co-integration test was then used to test for co-integration.
Additionally the Granger causality procedure was used to test the direction of causality between
foreign direct investment and economic growth. In order to fully account for feedbacks, a vector
autoregressive model is utilized. It was exposed that all the variables were non-stationary in
level and first difference form but were all found to be stationary after second differencing.
Further, the variables were integrated of order two, I (2). The results show that there is a stable,
long-run relationship between foreign direct investment and economic growth in Ethiopia from
the period 1980/81-2014/15.becuase that, a unit increase in the FDI encourages on average,
leads to an increase of about 4.157028 units in GDP. The pair-wise Granger causality result
shows that there is a unidirectional causality that runs from FDI to economic growth of
Ethiopia. The results indicate that foreign direct investment has highly significant positive effect
on economic growth. Hence, the researcher therefore recommend that, FDI facilitate
economic growth, so the government has to apply much effort in order to attract more FDI
into the country.

Keywords: Ethiopia, Foreign Direct Investment, Economic Growth, Stationarity, Co-

integration, Vector Error Correction Model, Granger Causality.

x
 CHAPTER ONE

1. INTRODUCTION

1.1 Background of the study

Foreign direct investment (FDI) is gradually significant network for resource flows between the
developed and developing countries. It affects economic growth of developing countries
positively through transfer of technology, capital, improved managerial skills, and know-how,
productivity, job creations, entrepreneurial ability and access to markets (Li and Liu (2005).most
countries seek to attract FDI through their attractive frameworks. In this sense the attraction has
become an obvious objective of economic policy both in developed and developing countries.
Generally foreign direct investment is a phenomenon resulting from globalization, which
involves the integration of the domestic economic system with global market. It is accomplished
through opening up of the local economic sector as well as domestic capital for foreign investors
to establish business, within the economy.

Most countries attempt to attract FDI, because of its positive implication as a tool of economic
development. That means the attitude of many developing countries towards the importance of
FDI has changed remarkably and action has been taken by these countries to ease restriction on
their inflow. The policy designed by developing countries towards FDI is based on the
assumption that increases the country‟s output, productivity and produce externalities and
technology transfer (Damooei and Tavakoli, 2006). FDI has been directed by the multinational
enterprise (MNE) into the economies of the developing world. Theoretically, FDI in the slow
growth model promotes economic growth by increasing the volume of investment, raise
economic growth by generating technological diffusion from developed countries to developing
countries, because the absence of appropriate technology and financial resource is obstructing for
development (Borensztein, Gregorio & Lee, 1998).

Globally, beginning from this mid- 1980s, FDI has grown much faster than either trade or
income. For instance, between 1985 and 1997 while worldwide nominal Gross Domestic
Product (GDP) increased at a rate of 7.2 per cent per year, worldwide imports at 9.2 per cent, and
the worldwide nominal inflows of FDI increased at 17.6 percent (Friedman, 2000). Considering
the benefits of FDI for growth and development, most African countries have undertaken varies
policy reforms to create conducive investment environment in order to attract a considerable

1
amount of FDI. According to the United Nations Conference on Trade and Development
(UNCTAD) of 2005 world investment report, the average annual FDI flows to Africa doubled
during the 1980s to 2.2 billion US$ compared to the 1970s, but increased significantly to 6.2US$
Billion and $13.8 Billion respectively during the 1990s and 2000–2003. However, it is low
compared to other developing regions. FDI inflows to the continent amounts to 36 Billion US$ in
2006, which was 20% higher than the previous record of $30 Billion in 2005 and twice the 2004
value of $18 Billion and rose to a historic value of $53 Billion in 2007 (UNCTAD,2008).

Moreover, recently (2009–2010) the trends of FDI flow to developing countries has been the
main source of capital inflows with greater stability and productive investment. The total FDI
inflow to these developing countries was 579 Billion US$ in 2005, this figure has increased to
1095 Billion US$ in the 2010, however, African share was ten percent and the flow is
characterized by uneven distribution among countries in the region (UNCTAD, 2011).
Commonly, FDI it assumed to be benefiting for a poor country economic growth, like Ethiopia.
The total FDI inflows in to Ethiopia have increased continuously from 135 million US$ in 2000
up to 545 US$ in 2004.

According to World Investment Report (WIR) of 2012, Ethiopia is one of the developing
countries receiving diversified FDI. It has been categorized among countries receiving annual
FDI ranging between 100 and 499 Million US$. An academic and policy area, the issue related
to FDI is not only attracting a significant amount of foreign capital but also its likely impact on
host economy. There has been a long debate in the literature on effect of foreign direct
investment on host economy like Ethiopia. The expected impact of FDI on the receivers
countries remains more argumentative in empirical than the theoretically leaving blind spots on
FDI as cure for development problem of developing countries. As result, in the recent literature
attentions are diverted to the impact of FDI on host country.

The main idea underlying the FDI liberalization policies of many developing countries and FDI
promotion efforts of international donors such as WB and IMF is the view that FDI inflows
foster economic growth. Theoretical arguments assign a key role for FDI in economic growth.
While these theoretical arguments are quit straightforward and widely accepted, the empirical
evidence is much more ambiguous, or as De Mello (1997) puts it: “whether FDI can be believed
to be a catalyst for output growth, capital accumulation and technological progress, appears to be

2
a less controversial hypothesis in theory than in practice” The empirical macro-economic
literature shows a clear link between FDI and GDP growth but the direction of causality is not
always clear (Carkovic and Levine, 2002; Nunnenkamp, 2003). Also when the heterogeneity of
the host economies is recognized in empirical studies, the link between FDI inflows and growth
becomes ambiguous (Nunnenkamp, 2003). Economic theory predicts FDI to create growth
multiplier effects through vertical and horizontal spillover effects; including the transfer of
technology and know-how to domestic firms, the formation of human capital, etc. In addition to
these advantages; some points which supports the concept that FDI promotes growth are
explained by, Agrawal and Khan (2011)

1. FDI acts as a means of transportation for transfer of advanced manufacturing

technologies from DCs to the LDCs,

2. FDI increase competition in the host country‟s markets.

3. FDI helps the host countries improve their foreign exchange reserves (or balance-of
payments position) by increasing exports,

4. FDI brings along with it the management know-how needed to runs the facilities,

5. FDI provides the financial resources needed by the host country,

6. FDI enhances the training and employment opportunities for the people of LDCs,

7. FDI reduces the burden of imports on the host countries through import substitution and

8. FDI acts as catalyst for increasing domestic savings and investment.

As a result of these benefits, many developing countries are now actively seeking for promoting
FDI by trying to create a favorable environment for it. So Ethiopia is one of them in order to
offers attractive investment opportunities for foreign companies and has adopted a number of
policies to attract foreign direct investment into the country. Some of the measures taken include
economic and political reforms aiming at macroeconomic and political stability, investment in
infrastructure and human capital and liberalization of trade (Haile and Assefa 2006).

Therefore, this paper seeks to analyze FDI inflows into Ethiopia and to investigate its effect on
economic growth empirically. This gives scope for necessary policy initiative in terms of
attracting more FDI to sector which the country has a competitive advantage. So this paper

3
focuses on the FD-led growth hypothesis in the case of Ethiopia. The study is based on time
series data from 1980/81-2014/15.

1.2 Statement of the problem

From the international viewpoint, the relationship between FDI and economic growth, and the
stability of this growth, is essential reflections of the host countries evaluate the trade-offs
associated with foreign access. This has been considered in the context of longer term
performance, stemming from the argument by Romer (1993) that an idea gap has held back
growth in emerging markets. If an idea gap has obstructed growth, FDI can bring a catch-up on
or examine the process. Investment in general is realized as one of the host important variables in
driving economic growth and development. Hence, FDI is believed to help as a strong instrument
for the reinforcement and spread of business opportunities throughout developing and
industrialized economies thereby enhancing economic development. It has been argued in
several studies that FDI contributes positively to economic development in host economies.

Generally Foreign Direct Investment (FDI) is considered as an option for financing development
both in developed and developing countries. This is mostly true where FDI brings invisible
financial resources and fills the gap between desired investment and domestically mobilized
savings, when it facilitates technology transfer and entry into export markets, as well as
strengthens the export capabilities of the host country resulting in productivity gains (Caves
2007; Ayanwale, 2007; Borensztein et al, 1998).

Conversely, some Studies conducted in developing countries in particular are found to be

contradicting or opposing with the above facts. Many researchers argue that FDI has an adverse
effect on development. They argue increased FDI does not always contribute to upgrading but
sometimes may even act to reduce the host country‟s long run potential, leading to a crowding-
out affect whereby domestic firms are displaced or outcompeted by foreign-owned and hence
affecting economic development negatively (Mwlima, 2003, Athukorala, 2003). As a result, the
role of FDI in promoting growth and curative development problems became controversial.

The agreement in the literature appears to be that FDI increase growth through productivity and
efficiency gains by local firms. The empirical evidence is not common; however, available
evidence for developed countries appears to support the idea that the productivity of domestic
firms is positively related to the existence of foreign firms (Globeram, 1979). The results for

4
developing countries are not so clear, with some finding positive spillovers (Blomstrom, 2000)
reporting limited evidence. Still others find no evidence of positive short run spillover from
foreign firms. Some of the reasons offered for these mixed results are that the functional forward
and backward linkages may not necessarily be there (Aitken et.al.1999). Further, the role of FDI
in export promotion remains controversial and depends crucially on the motive for such
investment (World Bank, 1998).

In general, the above conflicting evidence is not exception to Ethiopian economy. Because, the
vast body of literature states that FDI helps to promote economic growth through various
channels. Exactly, three main channels can be detected through which FDI affected growth:
Increase capital accumulation, Raises level of knowledge or skill, and Increase competition of
firms in the host country. With these assumption, developing countries are selecting for polices
to attracted FDI to block their development financing gap (Razafimahefa and Hamori, 2007).
Majority of the literature used foreign direct inflows as a proxy variable for foreign direct
investment and real per capita as a measure of economic growth to test the hypothesis of positive
relationship between foreign direct investment and economic growth. Among the more important
time series studies, the following studies may be mentioned: P.P.Awasantha (2003), Soltani
Hassen and Ochianis (2012), Kyuntae Kim and Hokyung Bang (2008), Dejene G. (2014) and
Getinet and Hirut (200).

Although, Ethiopia has meaningfully improved the functioning of its market economy; on the
other hand significant steps towards macroeconomic stability and structural reform are also
enhancing the attractiveness of foreign investments. Nowadays Ethiopia has become attractive
investment destination. According to www.ethiopianvestor.com the major reasons are: Political
and social Stability, Macroeconomic stability and Growing economy, adequate guarantees and
Protection, Transparent laws and efficient procedures, Abundant investment Opportunities,
Abundant labor force, Wide domestic, Regional and International market opportunity,
Competitive investment incentive packages, Welcoming attitude of the people to FDI, Amusing
climate and Fertile soils and Low production cost.

As a result from the above attractive investment opportunity, the current Ethiopian government
investment Proclamation No. 691/2010 and Article 39 of the Investment Proclamation
No.769/2012 allowed foreign investors to invest in all economic sectors, except those currently

5
reserved for domestic private investors, state investment or joint investment with government.
This study is recognize the growing confirmations from cross-country and country specific
studies that the association between foreign direct investment and economic growth. Because
Successful and sustainable economic growth requires continued improvement in investment and
productivity. Generally the research gap was identified from Chapter two in conceptual frame
work parts; this paper tries to explain whether FDI has a positive and significant impact on
economic growth in Ethiopia. So it links the gap between the various empirical literatures and
examining the developmental effects of FDI in Ethiopia economic growth by using a time series
data over the period 1980/81-2014/15. Therefore the expected outcome from this study could
also be useful in improving policy design, institutional setup, implementation, monitoring and
evaluation of FDI.

1.3 Hypothesis of the study

The relationship between FDI and economic growth are hypothesized based on past grounds of
theories and empirical findings. Therefore the following hypothesizes are being set to guide the
study: hypothesis 1: FDI will positively relate to economic growth and significant affect
economic growth measured in real gross domestic product. Hypothesis2: There will a causal
relationship between FDI and real gross domestic product and the direction of causality will from
FDI to real gross domestic product.

1.4 Objectives of the study

1.4.1 General Objective

The general objective of this study is examining the impact of foreign direct investment on
economic growth in Ethiopia from the period 1980/81-2014/15.

1.4.2 Specific Objectives

The specific objectives of the study are:

 To identify the relationship between FDI and economic growth in Ethiopia, and

 To determine the direction of relationship between FDI and economic growth.

6
1.5 Research questions

The paper attempt to answer the following basic questions:

 Is there a causal relationship between foreign direct investment and economic

growth in Ethiopia?

 Does foreign direct investment promote economic growth in Ethiopia?

1.6 Significance of the study

Foreign direct investment is important to most countries, because no country is self-sufficient,

i.e. there is mutual interdependence among countries. And can be particularity energetic for
developing countries like Ethiopia. In many instances, developing countries have both the
demand for a good and service, and the labor and natural resources to supply it, but they lack the
access to capital or financial resource necessary to begin producing. In the United States, most
businesses start when an entrepreneur goes to a bank and takes out a loan or finance. Larger
enterprises may go to an investment bank to sell stocks or bonds, to get their businesses going.
But in many developing countries, like Ethiopia either bank does not exist in adequate numbers
or they do not have enough capital to lend the people. Therefore, FDI provides essential capital
or financial resource to help catalyst the creation of productive enterprises, But

There are a few studies analyzing the empirical relationship between FDI and economic growth
in Ethiopia. Therefore, I believe that this thesis will provide more light on the benefits and costs
of the existing FDI, Help in filling knowledge gap in such area, Be important in elaborating the
significance of FDI to economic growth in Ethiopia, Help experts to re-evaluate foreign relations
as well as take measures that will help gain the positive benefits of FDI, Important because FDI
stimulates economic growth and spurs economic development as a whole, Be a vital tool for
policy makers in developing countries as th evidence, Provide guidelines on how to improve
their economic policies so as to attract more FDI in their countries, Help as a good tool for
academicians as well as investors to evaluate the importance of FDI on economic growth in
Ethiopia, Assist as a reference tool for researchers seeking to find out the effect of FDI on
economic growth, Come up with mechanisms to enhance the positive impacts of FDI on
Economic Growth through such paths as change of policy, Search to provide better
understanding characteristics and trends of FDI activates in Ethiopia.

7
Generally, the expected outcome from this study could also be useful in improving policy design,
institutional setup, implementation, monitoring and evaluation of FDI. As well as, it can suggest
further study in the area more precisely. So, the result of this study on the impact of FDI on
growth and development is important not only to researchers interested in economic
development but also to people responsible for formulating development policy. Hence, the
findings and empirical results of this study are provide empirical evidence and try to contribute
to the policy debate on the linkage FDI for policy.

1.7 Scope of the study

This paper are limited to the quantitative analysis of the relationship between the selected
macroeconomic variables, namely economic growth measured in real gross domestic product (as
dependent variable) , foreign direct investment, export, capital formation, (as independent
variable). So this study is rotate on the orbits of the impact of FDI on Ethiopian economic growth
based on time-series data over the period 1980/81-2014/15.

1.8 Limitation of the study

The limitations of the study are concerned with the problems of time constraints, money
constraints and lack of some requirements (unavailable or insufficient data especially foreign
direct investment for in-depth research investigation about the study. So the study was consider
only some key factors that determine Economic growth including FDI inflows and also considers
only some macroeconomic determinants of economic growth.

1.9 Organization of the research paper

This paper is organized into five chapters. Chapter one is the introductory parts of the paper; it
includes the background of the study, statement of the problem , hypothesis of the study,
objectives of the study, research question, significance of the study, scope of the study,
limitation of the study and organization of the research paper. Chapter two is review of related
literature and it has two parts: theoretical literature and empirical literature. Chapter three is
research methodology and data description. Chapter four is empirical analysis and discussion of
findings. Chapter five is conclusion of major summary findings, and recommendations.

8
 CHAPTER TWO

2 LITERATURE REVIEW

2.1 Introduction

This section presents the analysis of theoretical and empirical literature on the effect of foreign
direct investment on economic growth. So there is a large body of theoretical and empirical
literature on the impact of FDI on economic growth. But empirical evidence on the relationship
between FDI and economic growth is still questionable or unclear. However, FDI is assumed to
have a number of positive effects on the economy of the host country, (such as transferring
resource, increasing employment opportunities, improving the balance of payment, transferring
technology, productivity gains, introduction of new processes, managerial skills and know-how,
employee training, building capital, human formation, increase competition) and in general it is a
significant factor in modernizing then host country‟s economy and encouraging its growth.

2.2 Concepts and definitions of FDI

In history FDI is believed to begin in the 19th centuries. But now today there is no
straightforward and clear-cut definition of FDI, as different organizations use slightly different
definitions. FDI is defined as an investment involving a long-term relationship and control of a
resident entity in one economy. FDI implies that the investor applies a significant degree of
influence and control over the management of the enterprise resident in the other economy.
International Monitory Fund (IMF) defines FDI as “investment that is made to acquire a
permanent interest in an enterprise operating in an economy other than that of the investor, the
investor‟s purpose being to have an effective voice in the management of the enterprise” ( IMF,
1977). Similarly, UNCTAD (1999) defines FDI as “an investment involving a long term
relationship and reflecting a permanent interest and control of a resident entity in one economy
(foreign direct investor or parent enterprise) in an enterprise resident in an economy other than
that of foreign direct investor”.

FDI represents the primary means of transfer of private capital (physical or financial),
technology, personnel and access to brand names and marketing advantage (Makola, 2003). He
defines FDI as a form of international economic integration that brings gains to both parties
according to the principle of comparative advantage. FDI inflows are combined by main three

9
capital components. First is equity capital, which is the foreign direct investor‟s purchase of
shares of an enterprise in a country other than its own. Second, reinvested earning which is the
earning of the direct investor‟s share. Third, is an intra-company loan or an intra- company debt
transaction which involves borrowing and lending of funds between direct investors and
association enterprises (Jones, J and wren, C 2006).

Similarly, Dunning (1993), also recognized FDI by the following three ways; First, market
seeking FDI that refers to the purpose of helping local and regional markets, host countries
characteristics that can attract this kind of FDI are the sizes of per capital income, GDP growth
and the growth potential of the market. Second, resource seeking FDI refers to FDI for acquiring
resource that is not available in the home country. Like natural resource, raw material, and
availability of skilled and unskilled labor. Third also, Efficiency seeking FDI, this type of FDI
happens when a firm can gain from the common governance of the geographically dispersed
actions, especially in the existence of economic scale and scope, and diversification of risks.

2.3 Theoretical Review

2.3.1 Solow Growth Theory

The role of foreign direct investment (FDI) in stimulating economic growth is one of the host
issues in the development literature. This theory was developed independently by Solow and
Swan in 1956 through the Solow-Swan model. In the standard slow type growth model, FDI
enables host countries to achieve investment that exceeds their own domestic saving and
enhances capital formation. The Solow model emphasizes capital accumulation and exogenous
rates of change in population and technological progress. This model predicts that all market-
based economies will eventually reach the same constant growth rate if they have the same rate
of technological progress and population growth. Moreover, the model assumes that the long-run
rate of growth is out of the reach of policymakers (Solow, 1956). The key idea of this framework
is that growth is caused by capital accumulation and autonomous technological change. Solow
views the world as one in which output, Y, is generated by the population function Y=F (K, L),
where K is the capital sock and L is the labor force. He postulated that the production function
displays constant return to scale, so that doubling all inputs would double output. However,
holding one input constant- say, labor and doubling capital will yield less than double the amount
of output. Referred to as the law of diminishing marginal returns, this is one of the distinguishing

10
elements of the Solow model. In the Solow growth model, where technological progress is
exogenous, income will rise with the level of physical or human capital (accumulated human
knowledge

2.3.2 Neoclassical growth theory

The neoclassical theory was developed by neoclassical economists, initially introduced by
Veblen in 1900. The theory was further developed by various neoclassical theorists who consider
production to be directly related to reproduction. Neoclassical theory stipulates that international
capital flows with differentiated rates of return across countries lead to capital arbitrage, where
capital seeks the rate of return.

The neoclassical growth theory in contrast, which dates back to Solow (1956) and Rostow
(1956), assign an important role to FDI as a growth-enhancing factor to developing countries. In
the growth model by Rostow (1956), FDI is realized as a way of meeting the capital and
technological transfers required for economic transformation. Solow (1956) emphasizes on the
increased FDI and progress in technology as important variables in output growth, hence
development. Moreover, according to this theory, economic growth comes from two sources,
factor accumulation, and total factor productivity growth. As a result, FDI plays a double
function by contributing to capital accumulation and by increasing total factor productivity
(Lucas, 1988). Likewise, in the neoclassical growth models FDI promotes economic growth by
increasing the volume of investment and/or efficiency.

2.3.3 Endogenous Growth Theory

On the other hand, endogenous growth model (e.g. Romer (1993), Lucas (1988), and Barro and
Sala-i-Martin (1997) that highlight the importance of improvement in technology, efficiency,
and productivity suggest that FDI can positively influence the growth rate in so far as it generates
increasing returns in production through externalities and production spillovers. On further
theoretical arguments why developing countries may not gain from FDI; Krugman (1998),
argues that the transfer of control from domestic to foreign firms may not always be beneficial to
the host countries because of the adverse selection problem.

FDI undertaken within a crisis situation under “fire sale” may transfer ownership of firms from
domestic to foreign firms that are less efficient. This concern is particularly important to the
developing countries including the sub Saharan African countries, where, as part of privatization,

11
state owned enterprises, are sold to foreign firms simply because foreign firms have more
available funds than domestic ones. As pointed out by Salz (1992) and Agosin (2005) FDI may
also “crowd out” domestic firms through unfair competition. There is also a concern that the
reserve nature of many foreign owned firms and their minimal linkage to the rest of the economy
could reduce the potential spillover contribution to the national economy. Moreover, the
potential subsequent outflow of foreign firms' lower earnings to their parent companies could
also cause deterioration in the balance of payments. It is also argued that foreign corporations
tend to produce unsuitable goods that are handmade to satisfy the wealthy portion of the host
country‟s consumers, thereby increasing inequality and engaging in transfer pricing.

2.3.4 Economic Geography Theory

Yarbrough & Yarbrough (2002) discuss recent theoretical models of economic geography that
attempt to explain the spatial location of FDI. They assume that the decision of a Trans National
Corporation (TNC) on which area to locate investment depends on a set of characteristics of the
host area affecting firm‟s revenue or costs such as factor endowments, market size, per capital
income, skilled labor and availability of public infrastructure, among others. Aiello et al. (2009)
argue that other things being equal, a change in infrastructure expenditure influences the cost
faced by the firm in adjusting its current capital stock to the target level. They argue that this is a
reasonable assumption, given that the adjustment costs depend not only on the firm‟s internal
characteristics, but also on external factors, such as the provision of public infrastructure. In the
current global structure, geographical proximity and cultural & linguistic attractions are
becoming one important determinant of FDI (Jinayu, 1997)

2.3.5 Industrial Organization Theory

Hymer was one of the innovators who established a systematic framework in the study of FDI.
He was supported by Kindle Berger, (1969), (Caves, (1971), and Dunning (1979). among others
(Hymer, 1976). Homer‟s theory posits that firms operating abroad have to compete with
domestic firms that are in an advantageous position in terms of culture, language, legal system
and consumer‟s preference. Furthermore, foreign firms are also exposed to foreign exchange
risk. These disadvantages must be offset by some form of market power in order to make
international investment profitable. The sources of market power– the firm-specific advantage in
Hymer terms or monopolistic advantage in Kindle Berger‟s terms are in the form of patent-

12
protected superior technology, brand names, marketing and management skills, economies of
scale and cheaper sources of finance. According to Hymer, technological superiority is the most
important advantage as it facilitates the introduction of new products with new features.
Moreover the possession of knowledge helps in developing other skills such as marketing and
improved production process. A significant feature of this theory is that it articulates the point
that the advantages are transmitted effectively from one unit of a firm to another unit of that firm,
irrespective of the fact that they are either located in one country or in more than one country
(Caves, 1971). The Hymer-Kindleberger (HK) theory addressed the question of why foreign
multinational company is able to compete with indigenous /original firms in the host economy,
given the various advantages of indigenous firms. The indigenous firm has knowledge of
domestic market, consumer tastes, the legal and institutional framework of local business and
customs. In contrary, foreign firms face costs of operating business such as the scarcity of the
host country‟s information, difficulty in communication, fluctuation in exchange rates and
sometimes discrimination from political instability when they enter a new business environment
(Jones, J and wren, C 2006).

The theory states that the foreign firms must possess some rewarding advantages which allow
them to compete on equal terms with domestic firms. Foreign firms have advantages specific to
their ownership in order to compete with domestic firms. These potential advantages include
innovative, ownership of a brand name, the possession of special marketing skills, and access to
patented or generally unavailable technology, favored access to sources of finance, team-specific
managerial skills, plant economies of scale and economies of vertical integration. Thus,
international production/ FDI/ arise due to the fact that it is difficult to sell these advantages.
They cannot be sold because they are inherent in firm managerial experiences and organizational
capabilities (Moran, 1978). However, the theory could not explain certain aspect of FDI flows.
While the existence of some firm-specific advantages explains why a foreign firm can compete
successfully in the domestic market, such advantages do not explain why such competition must
take the form of FDI. It over-emphasizes the role of structural market failure and ignores the
transaction-cost side of market failure.

13
2.3.6 Internalization Theory

The Internalization theory was founded conceptualized by (Buckley and Casson, 1976) which
explains how multinational companies developed and became so strong and how they manage
their goals in Foreign direct investment, this theory asserts that MNCs are consolidating their
inside accomplishments so as to develop specific advantages, which then to be exploited. To
increase profitability, some transactions should be carried out within a firm rather than between
firms and this is one of the reasons why multinational companies exist. This theory may answer
the question why production is carried out by the same firm in different locations. Mostly,
technologies or knowhow can be sold and licensed. However, sometimes, there are technologies
that are embodied in the mind of a group of individuals and not possible to write or sale to other
parties. This difficulty of marketing and pricing know how forces multinational companies to
open a subsidiary in a foreign country instead of selling the technology. In addition, a number of
problems may arise if an output of a firm is an input to other firm in other country. For instance,
if each has a monopoly position, they may get into a conflict as the buyer of the input tries to
hold the price down while the firm that produces input tries to raise it. Nevertheless, these
problems can be avoided by integrating various activities within a firm rather than
subcontracting the activities (Krugman and Obstfeld, 2003). (Casson, 1976), developed the
theory by focusing on two kinds of integration which are vertical and horizontal integration. In
this theory, multinational companies work globally having foreign operations and transactions
with other firms located abroad by means of a governance structure with contracting.

It occurs in two different cases. First, when there isn‟t any market that can provide the basic
products the multinational companies need and the second reason is that the external or foreign
market which can supply such products isn‟t effective and fails to supply the goods needed by
such multinational companies. Global competitive advantages can be developed by means of
internalization by forming international economies of scale and scope.

2.3.7 Product life Cycle Theory

The Product Life Cycle Theory was developed by Vernon in 1966 based on the experience of the
U.S market. This theory was used to explain certain types of FDI made by U.S companies in
Western Europe after World War II (WW2). This theory explains how trade patterns change over
time. According to Vernon, a product has a life cycle that has four main stages innovation,

14
growth, maturity and decline. In the first stage, the U.S companies create new innovative
products for local consumption and export the surplus for the use of foreign markets. This stage
characterized U.S„s advantage of technology on international competitors as they were able to
manufacture the needed goods. The second stage was characterized by the MNC shifting the
production to the developing country. This would help the MNC cut on export costs, gain more
profits and also import some of the goods back to the home country. The third stage saw the
developing country‟s competitor export the goods to the MNC‟s home country. This marked
clear competition between the host country and the MNC home country, thus labeled the
maturity stage because the developing country‟s competitors have stabilized enough to export to
the MNC‟s country. The last stage was characterized by the developing country‟s markets
remaining viable target markets for other MNCs due to the stabilized economy. The MNC home
country market is seen to diminish. This theory explains both trade and FDI by outlining the
various phases of a MNC‟s Foreign Direct Investment and the changes over time. New products
are introduced to meet the local needs. These new products are first exported to similar countries
with similar needs, preferences and incomes. Increase in sale of the new product attracts
competitors and an increase of demand through exports to advanced countries occurs. Further
innovation in production, cost reduction and market process takes place and manufacturing is
shifted to foreign countries. Worldwide production through exports declines as a result of large
scale production and manufacturing is shifted to developing countries, making technology
standard. Market for the product concentrates in the less developed countries and demand shifts
to the developed countries, making the original innovator the importer.

2.3.8 Caves Theory

Caves developed a theory that distinguishes types of FDI in 1971. According to his theory, there
are two types of firms that engage in FDI: horizontal and vertical FDIs. Horizontal FDI takes
place when a firm enters into its own product market within a foreign country, whereas vertical
FDI occurs when a firm enters into the product market at a different stage of production
(Danning, 1993). Horizontal FDI firms will undertake if it either possesses a unique asset which
others do not have or because of the adverse effects of tariffs on its exports. Both reasons are
likely to result in FDI occurring in market structures characterized by oligopoly and product
differentiation abroad. First, the asset must be a public good within the firm so that once
provided the sunk cost has occurred and the firm‟s advantage can be used in other national

15
markets, for instance, the possession of superior knowledge. This allows the firm to offset any
informational disadvantages that compared with foreign local firms that will have accumulated
knowledge on the social, economic and cultural factors in that market. Second, profits made in
the host country must depend upon production in that country, as this ensures that the firm has to
locate abroad if it is going to be successful in production. The theory states that both
characteristics will be found in a market with product differentiation so that the firm can move
into these markets at minimum cost. In general, horizontal FDI is a feature of oligopolistic
markets where products are differentiated (Caves, 2007). Vertical FDI occurs when firms seek to
avoid strategic uncertainty and put up entry barriers to prevent foreign firms from entering the
market. It is argued by Caves that vertical FDI is more likely if profits in the foreign market are
dependent on long-term prices and large investments size. These together ensure market
structure that is characterized by a few suppliers. But, FDI is unlikely to occur when there is no
technological complementarily between stages of production and competitive market.

2.3.9 Location Theory

The theory was first developed by Heinrich (1826). He notes that because transportation costs
and, of course, economic rents vary across goods, different land uses and use intensities will
result with increased distance from the market place. The Location theory is generally concerned
with location specific advantages of production. So theory in itself is concerned with the
geographic location of an economic activity. It explains FDI in the context of the location
specific factor differentials. Location theory explains about supply (cost factors) and demand
(market factor) variables that affect the distribution processes of firms. i.e. It comprises “supply
oriented location theory” and “demand oriented location theory”. The comparative advantage,
the availability of raw materials, and transportation cost are main determinants in this theory.
FDI exists because of immobility of these factors of production (Claudia, Kleinert, Lipponer,
Toubal, Midelfart, 2005 and Markusen, 1995). The theory‟s explanation for FDI can be
discussed more by the following factors. First, the availability and cost of inputs can explain the
existence of FDI. A firm considers the source of input and cost of production in order to choose
the location. Thus, a firm investing abroad may be attracted by the availability of some inputs in
another country, which are scarce at home, or by the lower cost of inputs abroad such as cheap
labor cost. The lower labor costs can be the main reasons for FDI in developing countries (Jones,
J and wren, C (2006).

16
Second, marketing factors are the main driving force that stimulates foreign firms to invest
abroad. A firm can get many advantages by locating a production plant near the market. Firms
can conduct business smoothly because of locating the firm abroad and hence can better exploit
the local market. Furthermore, the production through the setting up of subsidiaries in a host
country may be more accepted by the local people than direct exporting. Finally, FDI is
stimulated by the existence of trade barriers. Subsidiaries of foreign firms are often set up in
another country that is not yet subject to trade restrictions. Then, the products are exported to
those markets that have imposed restrictions on the exports of the investing country.

2.3.10 An Eclectic Theory

Electric Paradigm theory was developed by Dunning (1988). This theory integrates/ constitutes
from three different theories of FDI: such as the industrial organization theory, the internalization
theory and the location theory (Dunning, 2000). According to this theory, there are three
conditions that must be satisfied if firms to engage in FDI. First, the firm must have some
ownership advantages with respect to other firms. These advantages usually arise from the
possession of firm-specific intangible assets. Second, it must be more beneficial for the firms to
use these advantages rather than to sell or lease them to other independent firms. Finally, it must
be more profitable to use these advantages in combination with at least some factor inputs
located abroad. Thus, if FDI to take place, the firms must have ownership and internalization
advantages, and a foreign country must have location advantages over the firms home country.
Dunning further divides these advantages into three groups. They are: (1) Ownership advantages,
(2) Location advantages, and (3) Internalization advantages. These three advantages create the
famous OLI model. These advantages are: benefits the firm can achieve from its size, monopoly
power and better resource capacity and usages; and benefits derived from the enterprise's ability
of operation and management such as know-how, organizational and marketing systems.

Location advantages are key determinants of host countries during the MNC‟s screening process
when looking for a suitable location to invest in. The specific advantages to the country here are
economic benefits comprising of both quantitative and qualitative factors of production, transport
costs, existence of raw materials, technological advancement and the market itself, political
benefits such as market structure, government legislation and policies, legal, and special taxes
that affect FDI inflows and social advantages such as cultural diversity and behavioral changes

17
such as attitudes towards foreigners (Dunning, 1979). So there are two types of location
advantages. The first type gained from attractions of special location advantages provided by the
host country, such as cheaper labor forces market for the product and the government's better
policies. The second one is generated from the limitations of the home. The investors are forced
to decide on direct investment abroad because they suffer from disadvantages in their own
countries such as a small market for their products, lack of raw materials and higher production
costs.in short, Location advantage (L) of firms is determined by natural resource endowments;
market size; assets (skilled labor, technology & infrastructure); and the sub region‟s regulatory
and policy position.

Internalization advantages refer to the benefits that the firms can secure by using its ownership
advantages internally between the parent company and its subsidiaries. According to this theory,
the importance and role played by O, L and I are different which determines the firms' choice of
international trade or direct production abroad. So internalization advantage shows flexibility
level and capacity of the firm in producing and marketing its own subsidiaries (Dunning, 1979).
In short, Internalization advantage (I) assesses how firms create and use their essential
competencies.

Ownership advantage is a firm specific advantage that gives power to firms over their
competitors. So the use of these benefits in a foreign market results in higher marginal
profitability or lower marginal costs than other competitors in the foreign market (Dunning,
1988). This includes advantage in technology and information, in management techniques
(entrepreneurial skills and organizational systems), easy access to finance, economies of scale
and capacity to coordinate activities, natural endowments, marketing, manpower, and capital. Of
these three advantages, ownership advantages are essential (Dunning, 1979).i.e. Ownership
advantage (O) asserts how foreign firms enjoy competitive advantage relative to domestic firms
Therefore there is no enterprise that can engage in FDI without any ownership advantages.
However, if the firm has only ownership advantages without the other two advantages, it will
benefit from licensing instead from FDI. If the firm has advantages of ownership and
internalization but not location advantages, it will prefer to sell its products by exporting. In
conclusion, FDI occurs only when a firm has all these three types of advantages. In other word,
all these factors are energetic for foreign direct investment as they show the level and pattern of
foreign direct investment.

18
Generally, one of the features of the eclectic theory is that it makes it to clear the difference
between structural and multinational market failure. The eclectic theory offers a better and clear
understanding of foreign direct investment when it is compared to the other theories as it defines
all the three variables of (OLI) all together creating an explanation from which everyone can
easily understand how the variables work and form foreign direct investment

2.4 Views on the Impact of FDI

There are two views on the impact of FDI on host economy: the benevolent model that argues
for FDI and the malign model that argue against. These two alternative conceptualizations guide
the understanding of the impact of FDI and its potential contribution to the economic
development for the host economy. The benevolent model assumes that FDI is more useful to the
economies which are fixed in the vicious circle of under development. If the potential host
economy is hindered in poverty loaded equilibrium with vicious circle of poverty, FDI can break
this by complementing local saving and supplying more effective management, marketing and
technology to improve productivity. The gain in national income depends on the size of the
capital flows and the elasticity of the demand for capital. Furthermore, technological and
managerial inputs, transfers and spillovers to local firms may reason the nation‟s production
function to shift upward. Therefore, under competitive condition (which is presence of foreign
firms and FDI may enhance); FDI should raise efficiency, expand output and lead to higher
economic growth in the host economy. The emphasis on the new resources that the foreign
investors bring to remove the bottlenecks that discourages the development process is a common
theme among international business groups and multilateral agencies that need greater
acceptance of FDI in the developing countries (Moran, 1993). Whereas as in malign model also
there is a long history of criticism of Multinational Corporation, much of it centers of the
possibility that foreign investors will prevent the way of laws that constrain socially undesirable
practices, such as pollution regulation or healthy safety and minimum wage requirements or
ignore laws already enacted. In the previous stage, a few studies showed that foreign capital had
a negative impact on the growth of the developing economies. The foreign firms made negative
impact on the host economy because they operated in industries where there were
substantial/extensive barriers to entry and increasing market concentration (Moran, 1993).

19
In this case, the foreign firms are creating to lower the domestic savings and investment by
extracting rent. According to this model, foreign firms have a potential to drive out the local
producers from business and substitute imported inputs. In such a situation, the foreign firms
might not link the gap between domestic investment and foreign exchange. In addition, the
repatriation of profit by these foreign firms drains out the capital from the host country.

2.5 Role of FDI in Growth

FDI can be seen a means of transportation for industrial development and technological
progress.it increases productivity and technological progress in a host country, FDI might
therefore have positive impacts on economic growth. In developing countries like Ethiopia, a
combination of advanced management skills and new technologies is likely to increase the
efficiency of economy. Thus, FDI may be the main channel through which advanced technology
is transferred to developing countries and hence affect economic growth positively (Neuhaus,
2005). Contrary to the above argument for positive effect of FDI on economic growth, there is
also debate on the possible adverse effects of FDI on economic growth. The debate has focused
on the economic condition of the recipient economy. Human capital and the financial market
development in the host country may influence the FDI effects. Borensztein, et al (1998)
describes the importance of human capital in the host country. These authors suggest that the
FDI effects on economic growth depend on the level of human capital available. Their empirical
results indicate that the higher productivity of FDI holds only when the host country has a
minimum threshold stock of human capital. If they are correct, FDI can only be expected to
contribute to economic growth when the host economy has a sufficient capability to absorb
advanced technologies. According to Alfaro, Chanda, Kalemli, and Sayek (2004) local financial
market development matters for contribution of FDI to host country‟s economic growth.

2.6 Empirical Evidence on Impact of FDI

At macro level empirical findings on the effect of FDI inflow on economic growth are mixed.
However, the positive contribution of FDI to growth is argued for by many researchers.
Therefore to summarize, there have been various empirical evidences that investigated the
impact of FDI on economic growth. For instance Bengoa and Sanchez-Robes (2003) found that
FDI has a significant positive effect on economic growth of developing countries. But the size of

20
its impact may vary from country to country depending on human capital, domestic investment,
infrastructure, macroeconomic stability, investment policy and liberalized capital market.

Similarly, De Mello (1999) and Borensztein et.al, (1998) found that there is a relationship
between FDI and economic growth, though this tend to be so because of the host country
characteristics such as human capital. Thus, the favorable growth enhance impact of FDI is
dependent on the conditions and characteristics of a host country enabling environment.
Investment recipient countries with better endowment of human capital and strong institutional
capacity are supposed to benefit more from FDI induced technology transfer and thereby
productivity gain. Studies by Blomstrom, Lipsey and Zejan (1994) found that FDI promotes
growth only in higher income developing countries. The study was a cross country analysis of 78
developing countries and they found no positive effect for lower income developing countries.
Balasubramanyam et.al (1996) investigated how foreign direct investment impacted economic
growth in developing countries using cross-sectional data and the Ordinary Least Square (OLS)
regression method. They found that FDI has a positive impact on economic growth only in
countries that have export promoting strategy.

The sector impact FDI on economic growth has not been studied broadly; the earliest research to
consider this matter was Alfaro (2003). The study was based on 47 countries in the primary,
secondary and service sectors for the period from 1981-1999 in cross country regression. In his
finding, positive growth effect of FDI was found on manufacturing sector, negative in the
primary sector and the service sector was found to be ambiguous. Nunnenkamp et al. (2006)
also found that the growth impact of FDI varies across sectors. The study was based on panel co-
integration framework for 15 industrial in the primary, secondary and service sector in India.
They found no causal relationship in the primary sector, temporary growth effect of FDI in the
service, and FDI stock and growth were mutually supporting in the manufacturing sector.
Similarly, a study conducted by IMF(2011) on Eastern Central and South Eastern Europe
countries indicate that the impact of FDI on trade depends on whether the sector are tradable
and non-tradable. The paper defined manufacturing, agriculture, mining, retail, hotels, and
restaurants as tradable sectors and electricity, transport, communication, real estate and financial
intermediation as non-tradable sectors. They study was conducted by applying both a cross
sectional and time series econometrics method. According to this research finding, FDI in
tradable sector is positively associated with higher export implying that there is a positive

21
correlation between stock of FDI going to these sectors and export performance of countries. On
the other hand, their empirical finding shows that FDI in non-tradable sector is positively
associated with import.

A few scholars have also emphasized on the way in which the growth effects of FDI depends on
the financial market conditions of the recipient country. Alfaro et al. (2004), emphasize that
growth effect of FDI depends on sound financial markets of the host country. He used cross-
country data for the period of 1975-1995 and found that FDI alone plays a vague role in
promoting economic growth, however, when several financial development measures are
included positive effects are found. Olofsdotter (1998) also argues that the beneficial effect of
FDI is stronger in those countries with higher level of institutional capacity. Similarly, Chang
(2009) applied panel co-integration and panel error correction model for 37 countries using
annual data from 1970-2002, found positive relationship between FDI and growth when financial
development measurement are included. When a country has a solid financial system as its
foundation, it follows that it is in a better position to more effectively earn the benefits form FDI
inflows.

Likewise, Li and Liu (2005) find that FDI has positive impact on economic growth of both
developed and developing countries. i.e. in both case FDI significantly and positively affect
economic growth; however, the interaction of FDI with technology gap of the countries behave
differently implying the difference of technology and hence absorptive ability of nations. In sum,
they concluded that there is a strong complementary connection between FDI and economic
growth both in developing and developed countries. The empirical analysis of Seetanah and
Khadaroo for 39 Sub-Saharan African countries over 1980-2000 exposed that FDI is an
important element in explaining economic performance of Sub-Saharan African countries.

Haile & Assefa in (2006) tried to examine the nature and factors that attracting FDI in Ethiopia
as a case study of their empirical research concentrating on theoretical relation between
economic growth and foreign direct investment in a edition to policy regimes, results of this
study concluded implicated that the growth rate of real gross domestic products bedsides free
trade with exports promotions have positive effects on attracting foreign capitals rather than non-
stable macro level regard with lack of infrastructure capabilities seems to have negative impacts
on attracting foreign direct investment in Ethiopia.

22
Tang and Selvanathan (2008) investigated the causal link between FDI, domestic investment and
economic growth in china for the period 1988-2003. The empirical result shows that there is
single-directional causality from FDI to domestic investment and economic growth. Similarly, an
empirical works of Magnus and Fosu (2008) for Ghanan economy indicated that the null
hypothesis that FDI does not granger cause GDP were not rejected, that is there is a one way
causal relationship between FDI and GDP growth in which the direction of causality is from FDI
to GDP growth. On the other hand, some research work clime that the contribution of FDI to
growth is not positive. For instance, Carkovic and Liven (2002) assert that FDI does not have a
healthy independent influence on growth based on their studies conducted for 75 countries.

Similarly, the research work done by Mwlima (2003) does not support the important of FDI in
economic growth. He concluded that there is no real evidence that FDI brings development.
Similarly, for Sri Lanka, Athukorala (2003) tested the FDI- led growth hypothesis using time
series data from 1959 to 2002. The result did not support the link between FDI and economic
growth. In the same way, Nunnenkamp and Spatz (2003) claim that conclusive evidence to
support the view that developing countries should draw on FDI to promote economic
development hard to come up. They contend the result on the growth impact of FDI is
ambiguous because of highly aggregated FDI data. Asiedu (2002) conducted a study on 32 sub-
Saharan African countries and 39 non sub Saharan African countries over a period of 10 years
(1988-1987). She argues that FDI inflows in to sub Saharan African countries are for market
seeking. Asiedu (2004) argues that natural resource and market size are the chief determinates
of FDI. Asiedu (2002) finding indicates that FDI in Africa is not solely determined by
availability of natural resource and that can play an important role in directing FDI through trade
reform. In addition to the above, she argues FDI policy instruments used to attract foreign
investors. But as explained in Asiedu (2004) the investment incentive by itself cannot be enough.
The host country should increase other determinants like infrastructure and market size,
macroeconomic and political stability, efficient institution and improvement in infrastructure.

Likewise, Meskerm Daniel (2014), Selamawit Berhie (2015), Birhe Esthete and Thomas Gebrie
(2012), Mulat Chanie (2012) and Dejene Gizew (2014) findings also indicates that mostly FDI in
Ethiopia has positive and significant effect on economic growth. But Meskerm Daniel findings
show the effect comes after two year lags.

23
Generally, FDI may have deep effects; those there are of various kinds such as growth, export,
technology, know-how transfer etc. however, these effects are found to vary across countries and
depend on several circumstances/factors such as institutional development, human capital
development, government policies, sector, etc. therefore government took a number of
measurements to improve its investment climate and attractiveness to potential investors. So this
paper is attempted to examine the effects of foreign direct investment on economic development
in Ethiopia by measured through real GDP growth, and spillovers effects.

2.7 Conceptual Frame Work

The conceptual frame works were designed from theoretical basis, practical observations and
empirical evidences. Most of the theoretical models imply that FDI is beneficial for the host
country‟s economic growth and also there is a widespread belief among policymakers that FDI
generates positive productivity for host countries. So among those different theories, Solow‟s
growth model has clearly stated the importance FDI in economic growth process. Because the
key idea of this theory is that economic growth is caused by capital accumulation and
technological change. Meaning that, the role of technological progress and capital accumulation
has been clearly stated as an important driver of economic growth in the model. Due to these
facts, Solow‟s growth model is my theoretical framework/base for this thesis. But different
empirical evidence fails to confirm this belief. Especially, In the case of developing countries
many empirical literatures finds on the effect of FDI on host countries economic growth is still a
debatable one. Some empirical studies indicate positive link between FDI and economic growth.
Because it is an alternative source of capital, technology and skill gain, increase production and
trade networks; enhance socioeconomic development. (Haile and Assefa, 2006; Meskerm Dan
(2014), Selamawit Berhie (2015), Birhe Esthete and Thomas Gebrie (2012), Mulat Chanie
(2012), Dejene Gizew (2014) Demollo, 1997; Ayanwal,2002; Agrawal,2005; Sala and
Trivin,2014; Sukar and Hassan,2011; Kabundi andloots,2012; Zakari et.al,2012; Lenka and
sharma,2015; Njoupouognigni and Ndambendia,2010; and Li and Liu, 2015). Contrary to this,
some other scholars (Carkovic and liven, 2007; Alege & Ogundipe, 2013; Mwlima, 2003;
Athukorala, 2003) argue that FDI retards economic growth through crowding out of domestic
infant industries , exploiting local resources, repatriating profits to their home countries, opening
door for corruption by some public officials. But, my finding were confirm with some empirical
studies about the positive effect of FDI on economic growth in Ethiopia. Some empirical studies

24
identify FDI has positive impact on economic growth depends on various aspect of the host
countries. Like economic sector, effect human capital, export promoting countries, higher
income developing countries, tradable and non-tradable sectors, natural resource etc.

Table 1:- Summary of Main Findings of the Empirical Literature Reviewed

Author Types of data Countries and time Result
period
Alfro(2003) Cross section 47 developing Sector effects of FDI and negative
and panel data countries 1981-1999 effect on primary sector, positive in
manufacturing sector and ambiguous
effect on service sector
Bengoa and Panel data and 18 Latin American FDI has positive effect on growth but
Sanchez-Robels time series countries 1970-1995 it depends on human capital,
(2003) economic stability and liberalized
capital market. .
Borensztein et al. Cross section 69 developing FDI has positive effect but its
(1998) countries 1970-1989 magnitude depends on human capital
in host country.
Balasubramanyam cross section 46 developing FDI has positive effect only for
et.al. (1996) countries 1970-1985 export promoting countries.
Blomstrom et al. Cross section 78 developing FDI has positive effect on growth
(1994) and panel data countries 1960-198 only in higher income developing
countries.

IMF (2011) Cross section Eastern Central and The study found positive correlation
and time series South Eastern between export performance and FDI
Europe countries stock in tradable sectors

Asiedu(2002) Cross section 32 Sub-Saharan and FDI has positive effect on growth
and time series 39 Non-Sub Saharan based on natural resource and market
African countries seeking

But, this paper was investigated basically the constructive effects of FDI on economic growth in
Ethiopia based on the whole economic activates rather than disaggregated and explain whether
FDI has a positive and negative impact on economic growth in Ethiopia in overall from the
period of 1980/81-2014/15.

25
 CHAPTER THREE

3. METHODOLOGY

There are many types of multivariate time series models to choose from First of all there is a
decision about linear and non-linear models and then the specific type of model within these two
categories. Many non-linear models have been specially designed and tailored for the problem
area they are applied to, especially within the financial domain, the choice of which model to use
can be very difficult. The data set under consideration within this thesis will be modeled by one
of the multivariate linear time series models, the main linear models are the Vector
Autoregressive (VAR) process.

3.1 Econometric Model

The basis of the research model is the augmented Cobb-Douglas production function with FDI
incorporated as one of the factor inputs. Theoretical model of first approach is formulated as
follows:

( ) ( )
The model hypothesizes that GDP is a function of FDI. Although GDP might also be affected by
some other factors such as Fixed Capital Formation (FCF) and Export (X) they are included in
this model additionally. But we would like to examine primarily the causal relationship between
economic growth (GDP) and foreign direct investment (FDI). Therefore other variables are
excluded in the economic model of this study.

The functional relationship between the variables and proxies can be expressed as:-

( )

3.2 VAR Time Series Analysis

The vector Autoregressive (VAR) model, proposed by Smith (1980), is one of the most
successful, flexible, and easy to use models for analysis of multivariate time series. It is applied
to grasp the mutual influence among multiple time series. VAR models extended the Univariate
Autoregressive (AR) model to dynamic Multivariate time series by allowing for more than one
involving variables. All variables in a VAR model are treated symmetrically in a structure sense;

26
each variable has an equation explaining its explaining its evolution based on its own lags and
lags of the variables.

Let ( ) t denotes a ( ) vector of time series variables. A VAR model

with p lags can then be expressed as follows;

Where the i- period back observation is called the i- lag of Y, is ( ) coefficient

matrix, c is a ( ) vector of constants (intercepts) and is an ( ) unobservable zero mean
white noise vector process to satisfying the following properties.

 The error term is normally distributed with mean zero [E ( ) = 0]

 The error term has to be constant variance (Homoscedasticity) or time invariant. That is
( )
 The error has to be independent white noise process with time invariant( ) = 0 for
any non-zero k, meaning that there is no correlation among the errors across time , in
particular, no serial correlation in individual error terms.

3.2.1 The choice of the variables

We have to determine a list of variables which can be assumed to affect each other inter
temporally. When we choose the variables, it is necessary to take three aspects into
consideration;

 The chosen variables should be related to the research problem.

 The choice of the variables should be in accordance with the theoretical hypothesis.
 Data used for fitting the model must be available and of good quality.

In this study, the VAR model set up used as GDP and FDI as inter-affect variables to generate
the VAR model.

3.2.2 Testable hypotheses

Based on the above framework, I test the hypothesis that FDI has an effect on economic growth
in Ethiopia. The null hypothesis is that FDI does not have an effect on growth in Ethiopia, while
the alternative hypothesis is that FDI Positive effect on growth in Ethiopia.

27
3.2.3 Testing the stationary of time series

This is the first step in any time series econometric analysis. Much of the theory for both the
linear univariate and multivariate time series analysis has been concentrated on a particular
family of times series referred to as stationary time series. In short, a time series is stationary if
its first and second moments (mean and variance) are time invariant. One interpretation of the
Wald decomposition theorem [Lutkepohl1993] is that a VAR process can be used to represent
stationary multivariate time series. With multivariate time series, there is another family called
stable multivariate time series; Smith (1980) suggests that non-stationary time series are still
feasible in VAR modeling. But in practice, using the non-stationary time series in VAR
modeling is problematic with regards to statistical inference since the standard statistical tests
used for inference are based on the condition that all of the series used must be stationary. That
means, stationary is important for time series analysis for some statistical evidence. In order to
do that, ADF and Phillips–Peron unit root tests are used. The hypothesis in unit root test is as
following:

If the test statistics of a variable is less than the critical in absolute terms, then the null hypothesis
cannot be rejected. Therefore, the first difference of a variable should be tested. If test statistics
of a variable is more than critical value in absolute terms, then the null hypothesis can be
rejected, which indicates the stationary condition of a variable.

3.2.3.1 Augmented Dickey-Fuller

In conducted the DF test, it was assumed that the error term was uncorrelated. But in case
is correlated, Dickey and Fuller have developed a test known as the Augmented Dickey-Fuller
(ADF) test. This ADF test is a modification of the DF test and involves Augmenting the Dickey-
Fuller equation by lagged values of the dependent variable. This done is to ensure that the error
process in the estimating equation is residually uncorrelated and also captures the possibility
that is characterized by a higher order autoregressive process. Thus the inclusion this lagged
value of the dependent variable were clean up any serial correlation in . The lag length is
dictated by the frequency of the data as well as the sample size and for annual data one or two

28
lags usually suffices (Wooldridge, 2009). In this ADF we still test whether δ = 0 and the test
follows the same asymptotic distribution as the DF statistics, so the same critical values can be
used.

3.2.3.2 Phillips Perron test

Phillips-Peron test (named after Peter C.B. Phillips and Pierre Perron) is a unit root test. That is,
it is used in time series analysis to test the null hypothesis that a time series is integrated of order
2. Like the Augmented Dickey-Fuller test, the Phillips-Perron test addresses the issue that the
process generating data for might have a higher order of autocorrelation than is admitted in the
test equation-making endogenous and thus invalidating the Dickey-Fuller t-test. Though the
Augmented Dickey –Fuller test addresses this issue by introducing lags of as regressor in
the test equation, the Phillips- Perron test makes a non- parametric correction to the t-test
statistic. The test is highly with respect to unspecified autocorrelation and heteroscedasticity in
the disturbance process of the test equation. A Then unit root test at level and at first difference
has been designed to test Stationary and non-stationary of the data series at both ADF and PP test
statistics.

3.2.4 Model identification

3.2.4.1 Lag length selection

Lag length selection is one of the most important steps after stationary test and before co-
integration test. Hence, we employed the information criteria methods like Akaike‟s information
criterion (AIC), Schwarz‟s Bayesian information criterion (SBIC), Hannan-Quinn information
criterion (HQIC), Likelihood ratio (LR), and Final prediction error (FPE), show the optimal lag
length where the information criterion is smallest. The number of lags for unit root test is
determined by referring to Akaike‟s Information Criterion (AIC) and Likelihood-Ratio test. It is
known that the more lags there are, the less the degrees of freedom are. When we determine the
number of lags, we choose the one with the minimum AIC and LR. Other test like SBIC, FPF
and HQIC can also be used. If the AIC value are not minimized using the same model, we
instead apply a likelihood-ratio (LR) test (Johannes 1995). The LR-statistic can be expressed as
follows.

( ( ) ( )) ( )

29
Where k is the lag order, L is the maximized likelihood of the model and n is the number of
variables.

If , we do not reject the null hypothesis that all the estimates in the coefficient matrix are
zero. Then we can reduce the lag order until the null hypothesis is rejected.

3.3 VEC Model

3.3.1 Johansen Co-integration Test

Before VEC model estimation, number of co-integrating equations should be defined. In order to
do that, Johansen co-integration test is run by using lagged values of variables. The number of
lags is defined based on the criteria that used in VAR model. By running this test, it is checked
whether the variables are co-integrated or not, which indicates whether there is a long-run
relationship between them. The hypotheses for this test are:

3.3.2 VEC Estimation

After running Johansen co-integration test and finding co-integrating vector, VEC estimation can
be done. The long run relationship between variables can be analyzed through looking at co-
integration equations. It also enables one to analyze long-run relationship with co-integrating
error which implies break points or disturbances in the long run equilibrium, Moreover, short-run
relationship can be observed based on changes of the lagged values in VEC estimation which
gives details of the relationship regarding past and current values of variables and explain
causality among them.

3.3.3 Model Diagnostics

A wide range of procedures is available for checking the adequacy of VAR and VECMs. They
should be applied before a model is used for specific purpose to ensure that it represents the data
adequately.

3.3.3.1 Test of Residual Autocorrelation (Breusch-Godfrey Serial Correlation)

In Ordinary Least Squares (OLS) regression, time series residuals are often found to be serially
correlated with their own lagged values. Serial correlation means (a) OLS is no longer an

30
efficient linear estimator, (b) standard errors are incorrect and generally overstated, and (c)OLS
estimates are biased and inconsistent if, as in this paper, a lagged dependent variable is used as a
regressor.

This test is an alternative to the Q-Statistic for testing for serial correlation. It is available for
residuals from OLS, and the original regression may include autoregressive (AR) terms. Unlike
the Durbin-Watson Test, the Breusch-Godfrey Test may be used to test for serial correlation
beyond the first order, and is valid in the presence of lagged dependent variables. The hypothesis
of the Breusch-Godfrey Test given by;

The Breusch-Godfrey Test regresses the residuals on the original regressor and lagged residuals
up to the specified lag order. The number of observations multiplied by is the Breusch-
Godfrey Test statistic.

3.3.3.2 Normality of the Residuals

The Jarque-Bera normality tests for univariate and multivariate series are implemented and
applied to the residuals of a VAR (p) as well as separate tests for multivariate Skewness and
kurtosis (Bera and Jarque 1980, 1981; Jarque and Bera 1987; Lutkepohl 2006). A multivariate
version of this test can be computed by using the residuals that are standardized by a Choleski
decomposition of the variance-covariance matrix for the centered residuals. Please note, that in
this case the test result is dependent upon the ordering of the variables. The hypothesis to test
normality of residual is given by:

The test statistics ( ) and the multivariate Skewness, , and Kurtosis test,
aredistributed as ( ).

31
3.3.4 Granger Causality Test

Based on the VAR estimation, Granger causality test is run in order to observe the causality and
its direction among variables or Granger causality test is implemented to identify how much the
one factor significant in forecasting the other one (granger, 1987). This test simply checks
whether or not the past values of one variable would explain or imply a change in present values
of the other variables. In that respect, a change in the past values of one variable would enable
one to predict present values of the other variables. In principle, if this is the case that changes in
variable is observed and then changes in variable is happened, then it can be said that X
Granger cause . In other words, if past values of variable increases the prediction or
forecasting of Y variable, then it is said that Granger cause . The hypotheses of Granger
causality test are the following:

Where and are random time series variables.

32
 CHAPTER FOUR

4. RESULT AND DISCUTION

4.1 Data Description

The study is based on the annual time series data observed from 1980/81-2014/15 where used to
appropriate model. The secondary data where obtained from Ethiopian Ministry of Finance and
Economics Development and National Bank of Ethiopia. The number of observation is 35, in
fact the number of observation for some of the variables are less than 35. The discussion of this
thesis begins with describing the data set and then extended to the results from the model
selection procedure. Finally results will be interpreted and discussed. The data analysis was
performed by using Eviews 6.

4.2 Descriptive statistics

In the empirical analysis, four aggregate series namely, Real Gross Domestic Product (RGDP),
Foreign Direct Investment (FDI), Fixed Capital Formation (FCF), and Export (X) were used
(table 4.1). Some descriptive statistics including the number of observation, sum of the
observation, mean, standard deviation, minimum and maximum values of the series under study
are presented in Table 4.1. The results show that the values of summary statistics are more or
less similar in magnitude.

Table 4.1 Descriptive statistics for the variables used in the analysis
Variables Observation Sum Mean Maximum Minimum Std. Dev.
name
RGDP 23 7,528,876 327,342 748,021 145,798 184,285

FDI 23 89,400 3,887 13,726 57.28 4,138

FCF 23 2,176,828 94,645 305,056 29,027 75,009

X 23 937,223 40,749 85,955 9,783 23,495

Source: Author’s Estimation using Eviews

33
Figure 4.1: Graphical approach for unit root test before stationary of RGDP, FDI, X and FCF

800,000

700,000

600,000

500,000

400,000

300,000

200,000

100,000

0
1975 1980 1985 1990 1995 2000 2005

RGDP FDI FCF X

Source: Author’s Estimation using Eviews 6

The line plot of the is presented in Figure 4.1. Clearly, the GDP, FDI, FCF and X have an
increasing trend showing that it is a non-stationary process.

4.3 VAR Modeling

4.3.1 Testing the Stationarity

Before starting any econometric analysis, unit root tests of related series must be made in order
to be careful of “artificial regression” problem i.e. first it must be investigates the time series
properties of RGDP, FDI, FCF, and X. Because the standard econometric theory of time series
data requires that variable should be stationary if an inference is to be adequate. That means,
non-stationary in the data is often considered a problem in empirical analysis and as such it needs
to be checked in order to prevent ending up with misleading results. Therefore, it is necessary to
test for stationary of time series variables before running any sort of time series analysis. This
study uses the ADF and PP test to detect the presence of unit roots in the time series. As shown
in Table 4.2, the same conclusion is obtained regarding the assumption of stationary on GDP,

34
FDI, FCF and X like we expected from the line plot. i.e. the test were done both at level and with
intercept and trend; The entire variables were not found stationary at level. This can be seen by
comparing the observed values in absolute terms of both the ADF and PP test statistics with the
critical values also in absolute terms of the test statistics at the 5% level of significance. Hence,
the result shows insignificant at 5% levels of significance; which fail to assume stationary or
there is a unit root in the data series.

Table 4.2: Unit root test result (At level)

Variabl Level with intercept Level with intercept and trends
es T- statistics Prob* T-statistics Prob*
Name
ADF PP ADF PP ADF PP ADF PP
RGDP 12.17446 13.07481 1.0000 1.0000 5.011176 5.522789 1.0000 1.0000
FDI -1.073889 -1.748008 0.7060 0.3947 -3.551400 -3.625263 0.0583 0.0507
FCF 7.407993 18.32294 1.0000 1.0000 6.131144 20.25537 1.0000 1.0000
X 0.225000 0.269798 0.9703 0.9731 -2.024636 -2.011999 0.5675 0.5741
Source: Author’s Estimation using Eviews 6

The result of ADF and PP unit root tests are presented in the above table (Table 4.2). As it is
seen, the null hypothesis is not rejected at the levels of variables. Therefore, we need to take first
differences of variables (RGDP, FDI, FCF and X) in order to see whether they are stationary or
not at first difference. The results that are presented in the Table 4.3
st
Table 4.3 ADF and PP Stationarity test at 1 difference of GDP, FDI, X and
FCF
Variables Level with intercept Level with intercept and trends
Name T- statistics Prob* T-statistics Prob*

ADF PP ADF PP ADF PP ADF PP

RGDP 1.020879 -0.762175 0.9958 0.8166 -0.894715 -2.859551 0.9440 0.1878
FDI -8.757003 -8.757003 0.0000 0.0000 -8.561540 -8.561540 0.0000 0.0000
FCF 1.798554 -2.085821 0.9996 0.2513 0.366083 -4.132587 0.9982 0.0136
X -4.638240 4.540056 0.0007 0.00010 -4.872566 -4.550866 0.0023 0.0049
Source: Author’s Estimation using Eviews 6

All the variables were differenced once and both the ADF and PP test were conducted on them
as shown in Table 4.3. The coefficients compared with the critical values ( 5% level of

35
significance) tells that all the variables were also found non-stationary at first difference except
FDI and X at level with intercept and level with intercept and trend in ADF and PP test. But on
the basis of this, the null hypothesis of non-Stationarity is accepted. This implies that the
variables are not integrated of order one. Therefore, we need to take second differences of
variables (RGDP, FDI, FCF and X) in order to see whether they are stationary or not at second
difference level. The result is arranged as shown below.

Table 4.4 ADF and PP Stationarity Test at Second Difference Integration

Variables Level with intercept Level with intercept and trends

Name T- statistics Prob* T-statistics Prob*

ADF PP ADF PP ADF PP ADF PP

RGDP -8.203104 -8.538261 0.0000 0.0000 -8.644938 -23.85840 0.0000 0.0000
FDI -8.396323 -40.61428 0.0000 0.0000 -8.578200 -42.27040 0.0000 0.0000
FCF -14.24464 -14.02471 0.0000 0.0000 -15.31543 -16..29964 0.0000 0.0000
X -7.335431 -12.61706 0.0000 0.0000 -7.263912 -14.11759 0.0000 0.0000

The results that are presented in both Table 4.4 and Figure 4.2 indicate that all series are
stationary. This means that the Second differences of GDP, FCF, X and FDI variables are co-
integrated of order two. Because, the coefficient compared with the critical value (5% level of
significance) tells that all the variables in ADF and PP tests were found stationary. Meaning that,
we can reject null hypothesis of non Stationarity rather we accept alternative hypothesis.
Therefore, we decided all variables are stationary in PP and ADF test and it is safe to conclude
that the variables are stationary and integrated of order two.

36
Figure 4.2: Graphical approach for unit root test after stationary of GDP, FCF, FDI and FDI

40,000

30,000

20,000

10,000

-10,000

-20,000

-30,000

-40,000
1975 1980 1985 1990 1995 2000 2005

DRGDP2 DFDI2
DFCF2 DX2

Source: Author’s Estimation using Eviews 6

4.4 Model identification

4.4.1 The VAR Order Selection Analysis

Information criteria method is the optimal ways to identify the proper lag length for VAR model.
To determine appropriate lag length for the VAR model; LR, FPE, AIC, SBIC, HQIC were used.
So the lowest/smallest information criteria value and number of star are the essential ways to
determine the optimal lag length. Among each information criterion AIC value is better to select
lag order of the model. Since our sample size is small; when applying the VAR model, the
number of lags should not have to be large given the minimum Information criteria rules. So in
order to avoid the loss of information, we determine that the optimal number of lags is 2 as
shown in Table 4.5.

Table 4.5: VAR model lag order selection criteria

Lag LogL LR FPE AIC SC HQ

0 -376.2478 NA 6.13e+15 42.02754 42.12647 42.04118

1 -366.8231 15.70794 3.38e+15 41.42479 41.72158 41.46571
2 -357.4430 13.54894* 1.90e+15* 40.82700* 41.32165* 40.89521*
3 -355.5928 2.261415 2.55e+15 41.06587 41.75838 41.16135

Endogenous: DRGDP2 DFDI2 * Indicates lag order selected by the criterion

Source: Author‟s Estimation using Eviews

37
So from the above Table 4.5result, all lag section criteria choose optimal lags two at 5% level of
significance. i.e. all information criteria were supported at lag length two and lowest information
criterion value was found here. We can conclude that there is a possibility of existence of co-
integration by applying the Johansson (1991) Trace and Maximum- Eigen value co-integrating
tests. Therefore the estimated VAR model can be expressed as VAR model-two.

4.5 Co-integration Analysis

Once we know VAR lag order, we can apply the Johansson maximum Likelihood method of co-
integration to obtain the number of co-integration vectors. The theory of co-integration proposed
by Engel and Granger (1987) is considered as one of the most important new concept in the field
of econometrics and time series analysis. The co-integration test clearly identifies the real long-
term relationship between the variables in the model. Most financial variables are not stationary,
this implies that the statistical estimation may sound good but in reality it is incorrect. Co-
integration, therefore allows estimating the long-term relationship between stationary variables
integrated of same order. Furthermore, the Johansen (1988) test was preferred to determine the
existence and number of co-integration between variables in the model by using the following
two tests. Namely, the trace statistics test and the maximum Eigen value tests were be used to
determine the number of co-integrating vectors are present. So the hypotheses for these tests are:
- there is no co- integration (null hypothesis) and there is at most one co-integration (alternative
hypothesis). Since we discovered through ADF and PP tests that GDP and FDI are integrated of
order two, so, we can check for co-integration, since all variables are stationary of order two.

Table 4.6: Unrestricted Co-integration Rank Test (Trace and Maximum Eigen Value)

Hypothesized Eigen Trace 0.05critical Prob** Max-Eigen 0.05critical Prob**

No. of CE(s) value statistics value statistic value

None* 0.605555 30.93850 15.49471 0.0001 16.74494 14.26460 0.0199

At most 1* 0.545488 14.19355 3.841466 0.0002 14.19355 3.841466 0.0002

Trace and Max-eigenvalue statistic test indicates 2 co-integrating eqn(s) at the 0.05 level
* denotes rejection of the hypothesis at the 0.05 level **MacKinnon-Haug-Michelis (1999) p-values
Source: Author‟s Estimation using Eviews

38
Normalized co-integrating coefficients (standard error in parentheses)
DRGDP2 DFDI2
1.000000 -4.157028
(1.52290)

Table 4.6indicates, the null hypothesis of no co-integration equation is rejected by both tests.
Because, the values of trace statistics were (30.94, 14.2) greater than critical value (15.5, 3.8)
respectively, as well as their respective probability value for both test is small and less than 5%
level of significance. Hence, the null hypothesis was rejected rather we accept the alternative
hypothesis. In another hand the Max-Eigen value of co-integration test was (16.8, 14.2) greater
than critical value (14.3, 3.8) respectively and probability value is less than 5%. So there is co-
integration or long-run relationship among them.

Consequently, the co-integration vector is given by: ( )

The values correspond to the co-integration coefficients of GDP (normalized to one), and FDI,
respectively. The co-integration equation is given by.

GDP = 2467.387+4.157028FDI

To support the statistical test of co-integration presented above, we represented the co-
integration graph generated from the estimated VAR model at the selected order. The co-
integration graph shown in Figure 4.3 below, confirmed the existence of co-integration among
the variable used in the study.

Figure 4.3 Co-integration relation graphs

80,000

60,000

40,000

20,000

-20,000

-40,000

-60,000
90 92 94 96 98 00 02 04 06

Cointegrating relation 1

Source: Author‟s Estimation using Eviews

39
4.6 Vector Error Correction Model

After determined variables in the Johansen co-integration test; the result indicates that the two
variables are co-integrated; we keep estimating the short run behavior and the adjustment to the
long run relationship, which is represented by VECM.

4.6.1 Model Estimation

Coefficient estimates of the VEC model are presented in (table A3). This table consists of two
parts; the first part covers the detail of the co-integration vector which is derived by normalizing
the GDP. The result shows that, the long run coefficients of GDP have a positive long run
relationship with FDI as expected in the theory. The long run equation is specified as follows:

GDP = 2467.387+4.157028FDI

The value 4.157028 indicates that a unit increase in the FDI encourages on average, leads to an
increase of about 4.157028 units in GDP.

The second part of the (table A3), indicates adjustment coefficients of the error correction term
(cointEq1) which identifies the fraction of the long-term gap that is closed in each period (Year).
The first equation GDP equation (1) below; the coefficient have a negative sign and significant
with large t-values ( -1.59067); which shows the remaining long term GDP departs by about 1.40
times in each period, while the gap in the FDI equation close by about 0.509 times in each year
(Eqn (2). The remaining long-run for FDI closes by 50.99% in each period (year).

Finally, using the error correction term as another independent variable in the unrestricted VAR
model we can estimate the following Vector Error Correction Model.

Model of real gross domestic product (RGDP2):-

 D(DRGDP2) = C(1)*( DRGDP2(-1) - 4.15702758784*DFDI2(-1) - 2467.38691713 )

+C (2)*D (DRGDP2 (-1)) + C (3)*D (DRGDP2 (-2)) + C (4)*D (DFDI2 (-1))
+ C (5)*D (DFDI2 (-2)) + C (6) + C (7)*FCF + C (8)*X……….Eqn (1)
 ∆(DRGDP2) = -1.39680*( DRGDP2(-1) - 4.15702758784*DFDI2(-1) - 2467.38691713 )
+ 0.27799601*∆ (DRGDP2 (-1)) - 0.177808*∆ (DRGDP2 (-2))
- 4.0242398*∆ (DFDI2 (-1)) - 1.8543982*∆ (DFDI2 (-2)) - 6498.4701
- 5.23542928472e-05*FCF + 0.1606656*X………………. Eqn (1)

40
Model of foreign direct investment (FDI):-

 D(DFDI2) = C(9)*( DRGDP2(-1) - 4.15702758784*DFDI2(-1) - 2467.38691713 )

+ C (10)*D (DRGDP2 (-1)) + C (11)*D (DRGDP2 (-2)) + C (12)*D (DFDI2 (-1))

+ C (13)*D (DFDI2 (-2)) + C (14) + C (15)*FCF + C (16)*X …………Eqn (2)

 ∆ (DFDI2) = 0.50988*(DRGDP2 (-1) - 4.15702758784*DFDI2 (-1) - 2467.38691713

- 0.26932*∆ (DRGDP2 (-1)) - 0.17773*∆ (DRGDP2 (-2)) - 0.02095*∆ (DFDI2 (-1))

- 0.4395299*∆ (DFDI2 (-2)) + 1783.14892 - 0.01556*FCF - 0.0103280*X …….Eqn (2)

Where: ∆ stands for second difference (D), the value in the bracket is the error correction term
and the coefficients of error correction term are called adjustment coefficient.

Generally each independent variable is associated with each coefficient. (Table A4and A5), So
to identify the significant relationship between the dependent variable and the independent
variables of the above equation (1) and (2) individually, we state the p-values. Hence the general
guide line is if the p-value is greater than 5%, such independent variables is not significant to
explain the dependent variable (RGDP2) and (FDI2), while, if the p-value is less than 5%, such
independent variables is significant to explain the dependent variables (RGDP2 ) and (FDI2).
Therefore from the above equation, Eqn (1); C (1)*[RGDP2 (-1), FDI2(-1)], C (2)*D (RGDP2)
(-1)), C (3)*D (RGDP2 (-2)), C (4)*D (FDI2 (-1)), C (5)*D (FDI2 (-2)), C (7)*FCF, and C (8)*X
independent variables are not significant variables to explain the dependent variable (RGDP2),
because, p-value is greater than 5%.Similarly in Eqn (2); C (9)*[(RGDP2) (-1), FDI2 (-1)]
independent variables are significant variables to explain the dependent variable (FDI2),
because, p-value is less than 5%. Meaning that, 0.03 is less than 5%.But, the rest independent
variables are not significant variables to explain the dependent variable (FDI2), such as C (10)*D
(RGDP2 (-1)),(11)*D (RGDP2 (-2)), C (12)*D (FDI2 (-1)), C (13)*D (FDI2 (-2)), C (15)*FCF,
and C (16)*X; in equation (Eqn2). Since p-value is greater than 5%.

However, due to the above two equation results especially in Eqn (1) no one independent
variable individually explains the dependent variable (RGDP).Therefore, it is better to see the
significance of independent variables jointly to explain the dependent variables (RGDP and FDI)
based on Wald coefficient test. So the general guide line is, if the coefficients of the independent
variables are jointly equal to zero, and if p-value is greater than 5%,we can‟t reject the null

41
hypothesis like C(1)=C(2)=0).That means the independent variables are not jointly significant
variables to influence the dependent variables (RGDP2 and FDI2).But if the coefficients of the
independent variables are not jointly equal to zero and if p-value is less than 5%, we reject the
null hypothesis like C (1) =C (2) ≠0), i.e. The independent variables are jointly significant
variables to explain the dependent variables (RGDP2 and FDI2). The result is arranged as shown
below table 4.7

Table 4.7 Wald coefficient test

Equation one (RGDP) Equation two (FDI)

Coefficient of the Independent variables Probability Coefficient of Independent variables Probability

independent the independent
variables variables

C(2)=C(3)=0 RGDP(-1)& RGDP(-2) 0.2393 C(10)=C(11)=0 RGDP(-1)& RGDP(-2) 0.0846

C(4)=C(5)=0 FDI(-1)&FDI(-2) 0.3619 C(12)=C(13)=0 FDI(-1) & FDI(-2) 0.0015***

C(7)=C(8)=0 FCF & X 0.7045 C(15)=C(16)=0 FCF & X 0.3416

C(2)=C(4)=0 RGDP(-1)&FDI(-1) 0.0230*** C(10)=C(12)=0 RGDP(-1)&FDI(-1) 0.0000***

C(3)=C(5)=0 RGDP(-2) & FDI(-2) 0.0478*** C(11)=C(13)=0 RGDP(-2) & FDI(-2) 0.0000***

C(2)=C(7)=0 RGDP(-1) & FCF 0.9096 C(10)=C(15)=0 RGDP(-1) & FCF 0.1403

C(2)=C(8)=0 RGDP(-1) & X 0.8605 C(10)=C(16)=0 RGDP(-1) & X 0.1690

C(3)=C(7)=0 RGDP(-2) & FCF 0.8780 C(11)=C(15)=0 RGDP(-2) & FCF 0.0581

C(3)=C(8)=0 RGDP(-2) & X 0.7315 C(11)=C(16)=0 RGDP(-2) & X 0.0740

C(4)=C(7)=0 FDI(-1) &FCF 0.3649 C(12)=C(15)=0 FDI(-1) &FCF 0.8036

C(4)=C(8)=0 FDI(-1) & X 0.3668 C(12)=C(16)=0 FDI(-1) & X 0.9891

C(5)=C(7)=0 FDI(-2) &FCF 0.3657 C(13)=C(15)=0 FDI(-2) &FCF 0.3452

C(5)=C(8)=0 FDI(-2) & X 0.3814 C(13)=C(16)=0 FDI(-2) &X 0.2501

42
Therefore in Eqn (1); C (2) with C (4), and C (3) with C (5) are jointly significant variables to
determine the dependent variable (RGDP), because, 0.0230 and 0.0478 are less than
0.05.However in Eqn (2); C (12) with C (13), C (10) with C (12) and C (11) with C (13) are
jointly significant variables to explain the dependent variable (FDI). Since, 0.0015 0.0000 and
0.0000 respectively are less than 0.05.but the rest independent variables still are not significant
variables jointly to explain the dependent variables (RGDP and FDI) in both equation. That
means all probability of the independent variables jointly is greater than 0.05.

4.7 Model Diagnostics

We test normality and autocorrelation of residuals for the VEC model. We see that residuals are
normally distributed and there is no autocorrelation in this model.

4.7.1 Test of Residual Autocorrelation /serial correlation

The test for residual serial correlation was performed based on Breusch Godfrey serial
correlation LM test. Therefore the test is determined by p- value, i.e. If p-value is greater than
5%, there is no serial correlation in the residual of the model and the null hypothesis are not
reject rather we accept. The null hypothesis is no serial correlation in the residual of this model.
So the (Table A6) presents the residuals of the Breusch Godfrey test for VEC model residual
serial correlation. The test is used to test for the overall significance of the residual
autocorrelations up to lag 2. The result is insignificant at 5% level of significant which failed to
reject the null hypothesis, i.e. there is no autocorrelation or error terms are not serially correlated.

4.7.2 Normality Testing of Residual

Multivariate version of the Jarque Bera tests is used to test the normality of the residuals. It uses
Skewness and kurtosis to determine whether the error terms are multivariate normally or not. In
addition to this we look the p- value, i.e. if p – value is greater than 5%, we can‟t reject the null
hypothesis. Because null hypothesis refers that residual are normally distributed. The results in
(Table A7) indicate that the null hypothesis has failed to be rejected. Meaning that, the error
term is normally distributed at 5% level of significance.

In addition to the above test, we conducted stability check test for VEC (Table A8)and found to
impose two unit roots indicating that the VECM is stable and there is two potentially co-

43
integration equation in the system. Similarly, the roots of characteristic Polynomial as clearly
shown below figure 4.4 complement the existence of the two co-integration equations.

Figure 4.4 inverse roots of AR characteristic Polynomial

Inverse Roots of AR Characteristic Polynomial

1.5

1.0

0.5

0.0

-0.5

-1.0

-1.5
-1.5 -1.0 -0.5 0.0 0.5 1.0 1.5

Source: Author‟s Estimation using Eviews

Inverse root of AR characteristic Polynomial shows no point resides outside the circle. This
again confirms the VECM is stable and there is potentially co-integrated equation in the system.
The result showed each error terms is normally distributed this confirmed the existence of co-
integration among the variables. From a statistical point of view, co-integration of two economic
variables implies that they move together over time. So that deviation from their long-term trend
was adjusted over time (Engle and Granger, 1987). Normally the result shows the null for no-
integration is rejected at 5% level of significance by the Trace and Max- Eigen statistics. Here, it
can be conclude that there is a long-run relationship among the variables which is explained by a
linear combination of I (2).

4.7.3 Lag Exclusion Test

To check whether the chosen lag is optimal, Wald lag exclusion test is used. Given that VAR
modeling requires uniform lag length for each variable, the result in Table 4.8shows that second
lag is significant for all variables at 5 percent level of significance. That is, the value in the
square brackets indicates probability value for the corresponding chi-square statistics. Therefore;
the VAR is found suitable for the data set and hence could be adopted.

44
 Table 4.8: Lag exclusion test

VEC Lag Exclusion Wald Tests

Date: 12/21/16 Time: 14:29
Sample: 1973 2007
Included observations: 18

Chi-squared test statistics for lag exclusion:

Numbers in [ ] are p-values

D(DRGDP2) D(DFDI2) Joint

DLag 1 7.541001 25.25501 27.70864

[ 0.023041] [ 3.28e-06] [ 1.43e-05]

DLag 2 6.082757 22.20406 22.92343

[ 0.047769] [ 1.51e-05] [ 0.000131]

Df 2 2 4

4.8 Granger Causality test results

In order to determine whether real GDP and FDI affect each other over time, a pairwise granger
causality test is performed. The result of pairwise granger causality Wald test is shown in the
table 4.9.

Table 4.9 VEC Granger causality test

VEC Granger Causality/Block Exogeneity Wald Tests

Dependent variable: D(DRGDP2)

Excluded Chi-sq df Prob.
D(DFDI2) 6.435499 2 0.0400
All 6.435499 2 0.0400

Dependent variable: D(DFDI2)

Excluded Chi-sq df Prob.

D(DRGDP2) 1.492401 2 0.4742

All 1.492401 2 0.4742

45
Since, we understand two models, (such as RGDP2 and FDI2). To determine the hypothesis we
depend on p-value. Means, if p-value is greater than 5%, we can‟t reject null hypothesis and no
granger causality between them. But if p- value is less than 5%, we can also reject null
hypothesis and has granger causality them. Therefore from the above Table4.9, the result
explains the existence of a unidirectional causality between the variables that runs from foreign
direct investment to real gross domestic product. Implying that FDI does influence economic
growth (RGDP), or it implies that past values of FDI has significantly contribute to the
prediction of economic growth and we can‟t accept H0, Because (0.04 < 0.05).But real gross
domestic product (RGDP) doesn‟t affect foreign direct investment. Means p-value is greater than
5%, because (0.47 > 0.05) and we can‟t reject H0 rather we accept. This result was similar to by
Dejene G, 2014 in Ethiopia. But his study investigated the impact of foreign direct investment
(FDI) on economic growth of Ethiopia in agreement of four other core macroeconomic variables
that include domestic investment, Inflation, government consumption and trade. Generally the
above result indicates the granger causality or association between RGDP and FDI in Ethiopia
for the period 1980/81-2014/15.

46
 CHAPTER FIVE

5. SUMMURY, CONCLUSION AND RECOMMENDATIONS

5.1 Summary

This paper studied the impact of Foreign Direct Investment (FDI) on economic growth of
Ethiopia from the period 1980/81-2014/15. The paper totally organized by five chapters. The
first chapter explores the background of this study and statement of the problem by elaborating
on what FDI needs, its impacts both positive and negative and hypothesis of the study, Objective
of the study, significance of the study, limitation of the study, organization of the paper, and
research question. The second chapter goes into various definitions, and theories and empirical
paradigm regarding on FDI and economic growth. So the vast body of literature states that FDI
helps to prompt economic growth through transfers of technology, raises level of knowledge and
skill and capital accumulation. Therefore Solow‟s growth model is my theoretical framework of
the finding. Because, the theory clearly stated that capital accumulation and technological
progress is an important driver for host country like Ethiopian economic growth. Some empirical
studies also approve the finding (Demollo, 1997; Ayanwale, 2007; Agrawal, 2005). Chapter
three also refers with the research method to be applied in this study. The next part explains the
research design. Based on the research questions, the methodology uses a quantitative research
design that helps to identify the numerical characteristics of the effects of FDI on economic
growth in Ethiopia. The next chapter offers a detailed report of the findings in the case study.
Findings from chapter four show that there is a significant positive relationship between foreign
direct investment and economic growth in Ethiopia. This positive relationship means that there is
a direct proportionate relationship between foreign direct investment and economic growth. This
means that more foreign direct investment leads to higher levels capital accumulation and
technological progress. Based on the above, we need to enhance more foreign direct investment
in order to promote economic growth. Policy implications of these findings are that FDI is a
prerequisite for economic growth in Ethiopia. The last chapter was also shows summery,
conclusion and policy recommendation.

47
5.2 Conclusion

This study examines the impact of foreign direct investment on economic growth in Ethiopia
using vector autoregressive methodology. Fixed capital formation and export have been included
as independent variables. The econometric approach used was co-integration and Vector Error
Correction Model (VECM). The data are used time series covering the period 1980/81-2014/15.
The statistical properties of the series was tested especially unit root test for Stationarity and
Johansen co-integration test was employed to discover the long relationship among variables.
While Granger Causality test was also employed to identify the direction of causality and Vector
Error Correction Model (VECM) was adopted to estimate long run and short run effect of FDI on
economic growth. Econometric and statistical methods have been applied in order to answer the
research question.

Generally, the result of stationary test indicates that all the series (variables) were integrated of
order two (two differences) that is I (2) by using Augmented Dickey Fuller (ADF) and Phillips
Perron (PP) and therefore conform to stationary at second difference. The Johannes co-
integration result shows that the variables are co-integrated. The Trace and Max-Eigen statistic
test show co-integration equations at 5% level of significance. A result of co-integration test
indicates that there is long run relationship among variables; meaning that all variables are
moving together in the long run (long run association ship). That means, the first basic research
question (i.e. is there a causal relationship between FDI and economic growth in Ethiopia?) were
answered. In addition to this, this question was answered by pairwise Granger Causality test.
Because, the tests showed that there is a unidirectional causality between foreign direct
investment and economic growth that run from foreign direct investment to economic growth.
Implying that FDI does influence economic growth of Ethiopia, but RGDP was does not granger
cause of FDI. This result was confirmed /approved by Dejene G, 2014; Tang and Selvanathan,
(2008), Magnus and Fosu, (2008). So it is one side feedback and it‟s completely consist with
theory proposed that more FDI inflows increase stocks of capital. In other word, FDI can
supplement the current stocks of awareness and technology in the host economy through labor
preparations systems, skill acquirement and diffusion an new managerial performs and
organizational appointments.

48
The second research question of this paper is also (does foreign direct investment promotes
economic growth in Ethiopia?). So Vector Error Correction Model estimation results provide
sufficient answer. The result estimated method (VECM) under VAR system shows that FDI are
positive and statistically significant determinants of economic growth with the test at 5% level of
significance in the long run. Short run estimation results suggested that both coefficient of lagged
variables RGDP (-1) and FDI (-1), RGDP (-2) and FDI (-2) have positive effects on economic
growth and were statistically significant jointly. In other words they are jointly influencing the
dependent variable in the short run. Generally based on empirical finding of this paper we can
say that FDI has plays important role to economic growth and promote economic growth in short
run. The coefficient of error correction term (speed of adjustment) toward the long run
equilibrium was large high it was about (0.5099). Sense that if there was deviation from the
equilibrium level only (50.99%) is corrected in one year as variables moving towards
equilibrium in the long run. This is most probably due to the fact that there are an other
variables that affect the economic growth which are not included to the empirical model such as
labor, human capital, saving and government expenditure. FDI can be seen from this findings as
an essential part of an open and operative worldwide economic system which organizes a major
promoter to growth.

49
5.3 Recommendation

Foreign direct investment is an important indicator to boost the economic growth of Ethiopia as a
medium in order to acquire skills, stock capital, management know-how and access to new
markets knowledge, technologies and at the same time to reduce debts. FDI does influence and
energetic for Ethiopian economic growth, meaning that, it‟s completely consisting with the result
finding and theory proposed that more FDI inflows increase economic growth. Therefore, the
results show that there is a stable, long-run relationship between foreign direct investment and
economic growth in Ethiopia from the period 1980/81-2014/15. Because that, a unit increase in
the FDI encourages on average, leads to an increase of about 4.16units in GDP. Based on the
findings of this study plus theoretical basis, and empirical evidences, the following policy
recommendations are recommended to attract and sustain more FDI in Ethiopia. Due to the
positive effects of FDI on the Ethiopian economy, the government should continue to keep its
open door policy to FDI and MNCs in the future. The government also needs to go a step further
and actively seek to attract FDI by advertising our economy and eventually set up national
investment promotion agencies regarding on investment promotion policies. Ethiopia should
adopt a positive approach towards FDI promotion, and clearly look for ways to increase its
benefits by attracting more bilateral and multilateral trade agreements, improving the quality of
infrastructure by way of channeling more resources to its development. Under these types of
policies, foreign investors are targeted at the industry level in order to meet Ethiopia‟s specific
needs that fit in with its developmental priorities.

In addition to this especially based on the practical observations the following recommendations
are suggested. Resource utilization by investor including land and water should get better
attention and needs strong regulation to keep FDI trend healthy, Investment proclamation of
Ethiopia should consider time based situation and also there must be amendment to protect the
country economy from the foreign investor, The investment climate of Ethiopia needs at most
care by all stakeholders and must back up by laws and regulations as well and Degree of
openness (trade liberalization), Economic and political stability are so essential to achieving
sustainable capital inflow, to achieve this, approving an investment friendly environment by
enhancing foreign investor legal protection, restructuring procedure for businesses. If the above
remarks are look over by the concerned bodies the current FDI development from European
Union d other major trade partners will attract a win-win project in Ethiopia.

50
 REFERENCE
Agrawal, G (2015). Foreign Direct Investment and Economic Growth in BRICS Economies:
A panel Data Analysis, Journal of Economics, business and management, 3(4), April
2015,
Allege P, and Organdie, A (2013). Sustaining Economic Development of West
African Countries: A system GMM Panel Approach. MPRA Paper 51702,
Nigeria: Covenant University, Ota, Ogun State.
Agrawal, G. and Khan, M.A. (2011). Impact of FDI on GDP: A Comparative study of china and
India, International journal of business and management, 6 (10), p. 71-79.

Agosin, M.R. and Machado, R. (2005). FDI in Developing Countries: does it crowd in domestic
investment? Oxford studies, 33(2), p 149-162

Agosin, M.R. and R. Mayer (2000), “foreign Investment in Developing Countries”, UNCTAD
Discussion papers, No. 146

Aitken, B.J. and Harrison Ann .E (1999), “DO domestic firms benefit from direct foreign
investment? Evidence from Venezuela”, American Economic Review, vil.89, pp.605-618

Aiello, F., Lona, A. and Leonida, L. (2009), Regional Infrastructure and Firm Investment: theory
and Empirical Evidence for Italy, Working paper, No. 639 London: Queen Mary
University.

Alfaro, L. (2003). Foreign Direct Investment and Growth: Does the sector Matter? UAS:
Harvard business school.

Alfaro, L., Chanda, A. Kalemli- Ozcan, S. and S. Sayek, (2004), “FDI and Economic Growth:
the Role of Local Financial Markets,” Journal of International Economics 64(1), p89-112

Asiedu, E. (2002). On the Determinates of Foreign Direct Investment to Developing Countries:

is Africa Different? World Development, 30(1), 107-119.

Asiedu, E. (2004). Policy Reform and Foreign Direct Investment in Africa: Absolute Progress
but Relative Decline. Development Policy Review, 22(1), 41-48.

51
Athukorala, W.P. (2003). The Impact of Foreign Direct Investment for Economic Growth:
A Case Study in Sri Lanka: paper submitted for the 9th International Conference on Sri
Lanka Studies, University of Peradeniya, and Sri Lanka.

Adelegan J.O (2000) Foreign Direct Investment and Economic Growth in Nigeria: a seemingly
unrelated model; African Review of Money, Finance and Banking, Supplementary issue
saving and Development, Milan, Italy, 5-25.

Ayanwale, A.B. (2007): FDI and Economic Growth: Evidence from Nigeria, African Economic
Research Consortium (AERC), and Research paper 165, Nairobi.

Balasubramanyam, V.N. (1996), M. Slaisu and D. Dapsoford Foreign Direct Investment in EP

and is countries, Economic Journal, 106: 92-105.

Barro, R. and X. Sala-i-Martin (1997), “Technology Diffusion, Convergence, and Growth”,

journal of Economic Growth, 1-26

Bengoa, M. and Sanchez- Robles, 2003, “FDI, Economic Freedom, and Growth: New evidence
from Latin America, European Journal of political Economy, 19, 529- 545.

Biruhe Eshete and Thomas Gebre (2012), FDI development between the European Union and

Less development countries; Business opportunity in Ethiopia, HAAGA-HELIA

University of applied science

Blomstrom, M., S. Globeram and A. Kokako (2000), “The Determinants of Host Country
spillovers from FDI”, center for Economic policy Research Discussion paper No 235.

Blomstrom, M. Lipsey. R. E., and Zejan, M. (1994). What explains developing country growth?
National Bureau of Economic Research (working paper No 4132).

Borensztein, E., De Gregorio, J. And Lee, J-W. (1998). HOW Does Foreign Direct Investment
Africa Economic Growth? Journal of International Economics, 45(1), 115-135.

Buckley, P., and Casson, M. (1976), the Future of the Multinational Enterprises, London:
Macmillan.

52
Carkovic, M. And Levine, R., (2002). Does Foreign Direct Investment Accelerate economic
growth? Retrieved January 25, 2012 Multinational Enterprise and Economic Analysis.
New York: Cambridge University press.

Caves, R.E. (2007): Multinational Enterprise and Economic Analysis. New York: Cambridge
university press.

Caves, R.E. (1971) “International Corporations: The Industrial Economics of FDI”, 7: P.1-27.

Chakraborty, C. And Nunnenkamp, P. (2006): Economic reforms, FDI and its economic
effect in India, Kiel Institute for the World Economy. (Working paper No.1272).

Dejene. G. (2014), the impact of FDI on Economic Growth, Journal of Poverty, Investment and

Development, Vol 15, 2015

De Mello, L.R. Jr. (1997), “FDI in Developing Countries and Growth: A Selective Survey”. The
Journal of Development studies, 34(1), 115-135

De Mello, L.R. (1999). FDI-led growth: Evidence form time series and panel data. Oxford
Economic Papers, 51, 133-151.

Damooei, J and Tavakoli A., (2006): The Effects of FDI and Imports on Economic Growth
a Comparative analysis of Thailand and Philippines (1970-1998), The Journal of
Developing Areas Volume 39, No. 2, 79-100

Dunning, J.H. (1993): Multinational Enterprises and the Global Economy, reading:
Addison Wesley.

Dunning, J.H. (1979): International Production and the Multinational Enterprise London: Allen
and Unwin: 2-26.

Dunning, J.H. (1988): Explaining International Production. London: Uniwin Hyman: 26-38.

Friedman, J. (2000), „What Attracts Foreign Multinational Corporations? Evidence from Branch
Plant Location in the United States‟, Journal of Regional Science, 32, 403–18
Globerman, S., (1979): “FDI and spillover efficiency benefit in Canadian manufacturing
industries”. Canadian Journal of Economics, 12: 42-56.

Gujarati, (2004), basic Econometric; Fourth Edition

53
Haile Getinet and Hirut Assefa, (2006), “Determinants of FDI in Ethiopia: A Time –series
Analysis.” Paper prepared for the 4th international conference on the Ethiopian economy.
Addis Ababa, June 10.University of Westminster, London .

Helmut Lutkepohl, New Introduction to Multiple Time Series Analysis Helmut Lutkepohl

And Markus Krsizig, Applied Time Series Econometrics. University, Avichai Sin. (1993)

Hymer, S. (1960), “Multinational Corporations and Foreign Direct Investment” London:

Rutledge for the United Nations.

Hymer, S. H. (1976), “The International Operations of National Firms. Cambridge, MA:

The MIT Press: 49-116.

IMF Growth in Sub-Africa: Performance, implements and policy requirements World Economic

Outlook, chapter VI.2011, Washington, DC: IMF

IMF, (1977): Balance of Payments Manual. 4th ed, Washington. DC.

Jeffrey M. Wooldridge, 2009. A text book of Econometrics, Michigan State University P.474
Jinayu, O. (1997), FDI in China and its impact on manufacturing growth. ISS working
paper, series No. 237.the Hague Netherlands

Jones, J and Wren, C (2006), FDI & the regional economy, Regional Science, vol.47.No 4, 2007

Johansson, S. (1988) “Statistical Analysis of Co-integration Vectors” Journal of Economic

Dynamic And Control, 12:pp.231-254.

Johansson, S. (1995), Likelihood-based Inference in Co-integrated Vector

Autoregressive Models, Oxford University press

Johansson, S. (1991) “estimation and hypothesis of co-integration vectors” in Gaussian vector

Autoregressive model, econometric, 1551-1580

Kabundi A. and Loots, E (2012), Foreign Direct Investment to Africa: trends, dynamics

&challenges, SAJEMSNS, 15(2)

Kindleberger C.P. (1969), American Business Abroad. The International Executive 11:11-1

Krugman, P. (1998), Fire sale FDI”. Working Paper, Massachusetts Institute of technology

54
Krugman, Paul R. and Obstfeld Maurice, (2003), International economics: Theory and policy
(6th edition). USA: Pearson Education, Inc.

Kyuntae Kim and Hokyung Bang, 2008 .The Impact of Foreign Direct Investment on

Economic growth: A case study of Ireland.

Lenka S. K. & Sharma P. (2015), FDI as a Main Determinant of Economic Growth:

A Panel Data Analysis, Annual Research Journal of Symbiosis Centre for

Management Studies, 1, Januray 2013-January 2014, 84–9

Li, X. and Liu, X (2005), FDI and Economic Growth: An Increasingly Endogenous Relationship.
World Development, 33(3), 393-407.

Lucas, R. (1988), “On the Mechanics of Economic Deployment”, Journal of economic

development, 22(3), 3-42.

Magnus, F.J. and Fosu, O.A. (2008): Bivariate Causality Analysis between FDI Inflows and
Economic Growth in Ghana. International Research Journal of Finance and Economics,
15, 104-112.

Makola M. (2003), the Attraction of the FDI by African Countries. Biennial ESSA conference:
some rest, west: Cape Town 17-19 September 2003. Retrieved on February 20, 2012
form http:/www.essa.org.za/download/2003 conference.

Meskerm .D (2014), the impact of FDI on economic growth of Ethiopia, university Oslo

Mwlima N. (2003) “FDI in Africa “Labor resource and research institute (LaRRI) south Africa

Moran, T. (1978): Multinational Corporation and Dependency: A dialogue for dependentistas

and non-dependentistas, international Organization, 32, 79-100.

Mulat Chanie (2012), the impact of FDI on Economic growth, Arba Minch University

Njoupouognigni M. & Ndambendia H. (2010), Foreign Aid, Foreign Direct Investment

And Economic Growth in Sub-Saharan Africa: Evidence from Pooled Mean Group

Estimator (PMG), International Journal of Economics and Finance, 2(3)

55
Nunnenkamp, p. & Spatz, J. (2003), “FDI and Economic Growth in Developing Countries”:
Kiel working paper No. 1176, Kiel institute for world Economy.

Razafimahefa and Hamori, (2007), International trade and FDI competitiveness

P.P.A Wasantha Athukorala (2003), the Impact of Foreign Direct Investment for

Economic Growth: A Case Study in SriLanka, Matara, SriLanka.

R.F. Engel and C.W.J. Granger, Co-integration and error correction: Representation,
Estimation, and testing, Econometrical, 55(2), 1987, 251-276
R.E. Lucas, why doesn‟t capital flow from rich to poor countries? American Economic Review,
papers and proceeding, 80 (1990), 92-96

Romer, P, (1993): “India gap and object gaps in economic development”, journal of Monetary
Economics vol 32 No 3.

Rostow, WW (1956), the Take-off into Self- Sustained Growth, The economic Journal, Vol.66
No.261, p.25 48.

Olofsdotter, K. (1998). “FDI, country capabilities &economic growth”, Weltwirtschaftliches

archive, 134(3): 534-47

Sala H. & Trivin P. (2014), Openness, Investment and Growth in Sub-Saharan Africa, Journal

of African Economies, 1–33. http://dx. http://dx.doi.org/ 10.1093/jae/ejt027.

Sukar, A., & Hassan, S. (2011). The Effects of Foreign Direct Investment on Economic

Growth: The Case of Sub-Sahara Africa. South western Economic Review.

Seetanah, B. & Khadaroo, A.J.(2007), “FDI and growth: new evidence from sub-Saharan Africa
countries”http:/www.csae.ox.ac.uk/conference/2007-EDiA-LAWBiDC/paper /169

Selamawit Berhe (2015), impact of FDI on economic growth Copenhagen business school

P. Romper, Endogenous technological change, Journal of political economy, 98, (1990), 71-102.

Phillips, P.C.B and P. Perron (1988), Testing for Unit Root in Time Series Regression,

Biometrical 75:pg. 346-355

56
Saltz, I.S. (1992), “The Negative Correlation between FDI & Economic Growth in the Third
World: theory & evidence”, RI vista International Di Science Economic the commercial,
19(7), 617-633.

Smith, S. C. (1980). Economic Development, (11th ed.). New York San Francisco Upper
Saddle River: Addison-Wesley
Solow, R. (1956). “A Contribution to the theory of economic growth”, Quarterly journal of
Economics, 70(1), 65-94

Soltani Hassen and Ochi Anis (2012).Foreign Direct Investment (FDI) and Economic Growth:

An approach in terms of co-integration for the case of Tunisia, Journal of Applied Finance

& Banking, vol.2, no.4, 2012

Tang, S. & Selvanathan, S. (2008). FDI, Domestic investment, and Economic Growth in China:
A Time Series Analysis. United Nation University, World institute for Development
Economic research

UNCTAD (1999), FDI in Africa: Performance and potential, United Nation, Geneva, UNCTAD/
IIT/Misc.15.

UNCTAD, (2005): World Investment report: Trans National Corporations and the
Internalization of R&D. United Nations New York and Geneva.

UNCTAD, (2008): World Investment Directory, Volume X Africa, United Nation, New York
and Geneva.

UNCTAD, (2011): FDI in LDCs: Lessons Learned from the decade 2001-2010 and the Way
Forward, United Nation, New York and Geneva.

Vernon, R. (1966), “International Investment and International Trade in the Product Cycle”
Quarterly Journal of Economics, 80,190-207.

Yarbrough, B. V. and Yarbrough, R.M. (2002), The World Economy: Trade and Finance
South Western college publishers.

World Bank, (1998), World Development Report, published by Oxford University press the
World Bank.

57
Wooldridge, J. M. (2005). Introductory Econometrics: A Modern Approach.
WIR, 2012: Methodological Note World Investment Report 2012: Towards a New Generation

Of Investment Policies, WIR

Zakari A. Mohammed, H. and Adamu Y. (2012), Does FDI Cause Economic Growth?

Evidence from Selected Countries in Africa and Asia, African Journal of

Social Sciences, 2(4)

58
 APPENDIX
Table:-A1 RGDP, FDI, FCF and X Data, (in million birr)

G.Y EFY RGDP FCF FDI X

1980/81 1973 115224.11238 21880.29572 9872.16215

NA

1981/82 1974 115110.57556 22075.35370 NA 8778.05781

1982/83 1975 126706.98715 21632.01668 NA 9227.03059

1983/84 1976 118729.13523 28002.45495 NA 10136.31517

1984/85 1977 107221.23738 16066.81869 NA 7006.56011

1985/86 1978 117837.32505 27052.04092 NA 8889.59963

1986/87 1979 134380.23714 29349.56101 NA 8924.53030

1987/88 1980 134308.81324 38447.71901 NA 8708.37999

1988/89 1981 134767.01362 27202.43476 NA 9808.74250

1989/90 1982 140247.62402 24516.09897 NA 8692.68909

1990/91 1983 135164.66433 19684.29879 NA 6023.20078

1991/92 1984 130176.97769 16754.46178 NA 4726.70569

1992/93 1985 145798.54619 29026.60967 153.8761 9782.59615

1993/94 1986 148275.62095 31468.85113 87.6576 13584.68938

1994/95 1987 156247.19547 35957.63415 309.3994 18187.86640

1995/96 1988 172839.41230 40856.44050 57.276 18232.76220

1996/97 1989 180910.93085 43065.04580 406.4509 23647.50028

1997/98 1990 178301.48708 42820.90386 931.2201 22789.07441

1998/99 1991 188990.35793 44833.86910 870.014 21448.79442

1999/00 1992 199102.15332 44107.90940 449.1135 24120.19302

2000/01 1993 215630.47924 50746.40604 923.5166 26047.13116

2001/02 1994 218896.66898 57704.48814 1478.89 27828.23635

2002/03 1995 214166.13178 51974.93255 530.099 28766.84930

2003/04 1996 243233.53479 70589.84040 1413.216 36606.85097

59
2004/05 1997 271980.92362 70671.13390 3588.377 41452.14248

2005/06 1998 301449.05038 83052.59233 2803.819 42056.52502

2006/07 1999 335983.09935 81226.60027 10465.21 43057.66963

2007/08 2000 372230.81754 90924.08089 6259.699 42799.87617

2008/09 2001 404996.38646 100416.44481 6095.143 42853.68962

2009/10 2002 455825.63572 122967.06777 11470.56 62643.81663

2010/11 2003 515078.54100 140903.95012 7793.471 85954.96957

2011/12 2004 559621.56400 181854.38055 13725.83 76581.74383

2012/13 2005 618842.22900 200121.62335 6844.169 76774.84648

2013/14 2006 682454.17800 256481.01772 7219.121 78953.12895

2014/15 2007 748021.08600 305056.11001 5523.52 73051.79310

GDP by Expenditure Approach at Constant Prices 2010/11=100

Source:-NBE, MoFED

Table: - A2
Vector Auto regression Estimates
Date: 12/20/16 Time: 21:03
Sample (adjusted): 1989 2007
Included observations: 19 after adjustments
Standard errors in ( ) & t-statistics in [ ]

DRGDP2 DFDI2

DRGDP2(-1) -0.338279 0.069200

(0.22388) (0.06414)
[-1.51100] [ 1.07893]

DRGDP2(-2) -0.516397 0.004173

(0.21803) (0.06246)
[-2.36848] [ 0.06681]

DFDI2(-1) 0.184294 -1.361092

(0.68044) (0.19494)
[ 0.27084] [-6.98223]

DFDI2(-2) -0.463389 -0.785095

(0.72730) (0.20836)
[-0.63713] [-3.76796]

60
 C 4942.762 -561.7270
(2668.69) (764.538)
[ 1.85213] [-0.73473]

R-squared 0.381835 0.807207

Adj. R-squared 0.205216 0.752123
Sum sq. resids 1.58E+09 1.30E+08
S.E. equation 10620.30 3042.555
F-statistic 2.161919 14.65418
Log likelihood -200.1986 -176.4473
Akaike AIC 21.59986 19.09972
Schwarz SC 21.84839 19.34825
Mean dependent 2577.615 -75.97245
S.D. dependent 11912.76 6111.116

Determinant resid covariance (dof adj.) 1.04E+15

Determinant resid covariance 5.67E+14
Log likelihood -376.6455
Akaike information criterion 40.69952
Schwarz criterion 41.19660

Table: - A3

Vector Error Correction Estimates

Date: 12/20/16 Time: 21:13
Sample (adjusted): 1990 2007
Included observations: 18 after adjustments
Standard errors in ( ) & t-statistics in [ ]

Co-integrating Eqn: CointEq1

DRGDP2(-1) 1.000000

DFDI2(-1) -4.157028
(1.52290)
[-2.72969]

C -2467.387

Error Correction: D(DRGDP2) D(DFDI2)

CointEq1 -1.396802 0.509883

(0.87812) (0.20069)
[-1.59067] [ 2.54063]

D(DRGDP2(-1)) 0.277996 -0.269330

(0.63905) (0.14605)
[ 0.43501] [-1.84406]

61
 D(DRGDP2(-2)) -0.177808 -0.177727
(0.35004) (0.08000)
[-0.50796] [-2.22157]

D(DFDI2(-1)) -4.024240 -0.020952

(2.84453) (0.65011)
[-1.41473] [-0.03223]

D(DFDI2(-2)) -1.854398 -0.439530

(1.35033) (0.30861)
[-1.37329] [-1.42421]

C -6498.470 1783.149
(10162.4) (2322.58)
[-0.63946] [ 0.76774]

FCF -5.24E-05 -0.015555

(0.10299) (0.02354)
[-0.00051] [-0.66085]

X 0.160666 -0.010328
(0.36558) (0.08355)
[ 0.43948] [-0.12361]

R-squared 0.660697 0.953462

Adj. R-squared 0.423184 0.920885
Sum sq. resids 2.13E+09 1.11E+08
S.E. equation 14581.88 3332.625
F-statistic 2.781735 29.26829
Log likelihood -192.8264 -166.2581
Akaike AIC 22.31405 19.36201
Schwarz SC 22.70977 19.75773
Mean dependent 581.9810 -148.4361
S.D. dependent 19199.72 11848.35

Determinant resid covariance (dof adj.) 1.92E+15

Determinant resid covariance 5.93E+14
Log likelihood -357.2209
Akaike information criterion 41.69121
Schwarz criterion 42.58158

62
Table: - A4

Dependent Variable: D(DRGDP2)

Method: Least Squares
Date: 12/20/16 Time: 21:28
Sample (adjusted): 1990 2007
Included observations: 18 after adjustments
D(DRGDP2) = C(1)*( DRGDP2(-1) - 4.15702758784*DFDI2(-1) -
2467.38691713 ) + C(2)*D(DRGDP2(-1)) + C(3)*D(DRGDP2(-2)+
C(4)*D(DFDI2(-1)) + C(5)*D(DFDI2(-2)) + C(6) + C(7)*FCF + C(8)*X

Coefficient Std. Error t-Statistic Prob.

C(1) -1.396802 0.878124 -1.590666 0.1428

C(2) 0.277996 0.639054 0.435012 0.6728
C(3) -0.177808 0.350043 -0.507961 0.6225
C(4) -4.024240 2.844533 -1.414728 0.1875
C(5) -1.854398 1.350332 -1.373290 0.1997
C(6) -6498.470 10162.45 -0.639459 0.5369
C(7) -5.24E-05 0.102988 -0.000508 0.9996
C(8) 0.160666 0.365579 0.439483 0.6697

R-squared 0.660697 Mean dependent var 581.9810

Adjusted R-squared 0.423184 S.D. dependent var 19199.72
S.E. of regression 14581.88 Akaike info criterion 22.31405
Sum squared resid 2.13E+09 Schwarz criterion 22.70977
Log likelihood -192.8264 Hannan-Quinn criter. 22.36861
F-statistic 2.781735 Durbin-Watson stat 2.365272
Prob(F-statistic) 0.069534

Table:-A5

Dependent Variable: D(DFDI2)

Method: Least Squares
Date: 12/20/16 Time: 21:50
Sample (adjusted): 1990 2007
Included observations: 18 after adjustments
D(DFDI2) = C(9)*( DRGDP2(-1) - 4.15702758784*DFDI2(-1) -
2467.38691713 ) + C(10)*D(DRGDP2(-1)) + C(11)*D(DRGDP2(-2)) +
C(12)*D(DFDI2(-1)) + C(13)*D(DFDI2(-2)) + C(14) + C(15)*FCF + C(16) *X

Coefficient Std. Error t-Statistic Prob.

C(9) 0.509883 0.200691 2.540632 0.0293

C(10) -0.269330 0.146053 -1.844055 0.0950

63
 C(11) -0.177727 0.080001 -2.221573 0.0506
C(12) -0.020952 0.650106 -0.032229 0.9749
C(13) -0.439530 0.308613 -1.424213 0.1848
C(14) 1783.149 2322.583 0.767744 0.4604
C(15) -0.015555 0.023538 -0.660848 0.5236
C(16) -0.010328 0.083551 -0.123613 0.9041

R-squared 0.953462 Mean dependent var -148.4361

Adjusted R-squared 0.920885 S.D. dependent var 11848.35
S.E. of regression 3332.625 Akaike info criterion 19.36201
Sum squared resid 1.11E+08 Schwarz criterion 19.75773
Log likelihood -166.2581 Hannan-Quinn criter. 19.41658
F-statistic 29.26829 Durbin-Watson stat 1.952593
Prob(F-statistic) 0.000007

Table: - A6

VEC Residual Serial Correlation LM

Tests
Null Hypothesis: no serial correlation
at lag order h
Date: 12/21/16 Time: 14:15
Sample: 1973 2007
Included observations: 18

Lags LM-Stat Prob

1 9.047837 0.0599
2 7.060831 0.1327

Probs from chi-square with 4 df.

64
Table:-A7
VEC Residual Normality Tests
Orthogonalization: Chloesky (Lutkepohl)
Null Hypothesis: residuals are multivariate normal
Date: 12/21/16 Time: 14:18
Sample: 1973 2007
Included observations: 18

Component Skewness Chi-sq Df Prob.

1 0.125446 0.047210 1 0.8280

2 -0.301419 0.272561 1 0.6016

Joint 0.319771 2 0.8522

Component Kurtosis Chi-sq Df Prob.

1 0.754321 3.782306 1 0.0518

2 0.835728 3.513055 1 0.0609

Joint 7.295362 2 0.0261

Component Jarque-Bera df Prob.

1 3.829516 2 0.1474
2 3.785616 2 0.1506

Joint 7.615132 4 0.1067

Table: - A8
Roots of Characteristic Polynomial
Endogenous variables: DRGDP2 DFDI2
Exogenous variables: FCF X
Lag specification: 1 2
Date: 12/21/16 Time: 14:21

Root Modulus

1.000000 1.000000
-0.716877 - 0.621290i 0.948638
-0.716877 + 0.621290i 0.948638
-0.176091 - 0.747756i 0.768211
-0.176091 + 0.747756i 0.768211
-0.473421 0.473421

VEC specification imposes 1 unit root(s).

65