University of Ghana http://ugspace.ug.edu.gh UNIVERSITY OF GHANA THE EFFECT OF EXTERNAL DEBT ON ECONOMIC GROWTH IN GHANA: AN ARDL APPROACH BY ANGELA ACKON (10701538) THIS LONG ESSAY IS SUBMITTED TO THE UNIVERSITY OF GHANA, LEGON IN PARTIAL FULFILMENT OF THE REQUIREMENT FOR THE AWARD OF MASTER OF SCIENCE IN DEVELOPMENT FINANCE DEGREE JULY, 2019 University of Ghana http://ugspace.ug.edu.gh DECLARATION I, ANGELA ACKON, do hereby declare that this work is the result of my own research and has not been presented by anyone for any academic award in this or any other university. All references used in this work have been fully acknowledged. I bear sole responsibility for any shortcomings. …………………………………… …………………………… ANGELA ACKON DATE (10701538) i University of Ghana http://ugspace.ug.edu.gh CERTIFICATION I hereby certify that this dissertation was supervised in accordance with procedures laid down by the University. …………………………………… …………………………… DR. LORDINA AMOAH DATE (SUPERVISOR) ii University of Ghana http://ugspace.ug.edu.gh DEDICATION This dissertation is first and foremost dedicated to the almighty God without whom this work would not have been a success. I also dedicate this work to my lovely Mother Esther Agbenowosi for her support. iii University of Ghana http://ugspace.ug.edu.gh ACKNOWLEDGEMENT I would like to express my profound appreciation and praise to the almighty God for the gift of life and good health, and offering me yet another opportunity to pursue my second degree and for His divine grace and protection throughout the entire period of my study. I wish to thank my supervisor, Dr. Lordina Amoah for her constructive criticism and guidance throughout the preparation of this dissertation. In spite of her busy schedules, she still made time to go through every word, phrase and sentence in the work. May the almighty God richly bless you! I also thank my immediate family (Alex, Adrian and Anna) for their encouragements during my low moments. I also want to say a Special thank you to Abraham Mensah-Belley for his support and guidance throughout the entire program. I cannot thank you enough. iv University of Ghana http://ugspace.ug.edu.gh TABLE OF CONTENTS DECLARATION ........................................................................................................................ i CERTIFICATION ..................................................................................................................... ii DEDICATION .......................................................................................................................... iii ACKNOWLEDGEMENT ......................................................................................................... iv TABLE OF CONTENTS ............................................................................................................v LIST OF FIGURES ................................................................................................................ viii LIST OF TABLES .................................................................................................................... ix ABSTRACT ...............................................................................................................................x CHAPTER ONE .........................................................................................................................1 INTRODUCTION ...................................................................................................................1 1.1 Background of the Study ....................................................................................................1 1.2 Statement of the Problem ...................................................................................................3 1.3 Objectives of the Study ......................................................................................................5 1.4 Research Questions ............................................................................................................5 1.5 Significance of the Study ...................................................................................................5 1.6 Research Scope and Limitations ........................................................................................6 1.7 Organization of the Study ..................................................................................................6 CHAPTER TWO ........................................................................................................................7 LITERATURE REVIEW ........................................................................................................7 2.1 Introduction .......................................................................................................................7 2.2 Theoretical Review ............................................................................................................7 2.2.1 Neoclassical Growth Theories -Solow Growth Model .................................................8 v University of Ghana http://ugspace.ug.edu.gh 2.2.2 The Dual Gap Theory ................................................................................................ 10 2.3 Empirical Literature Review ............................................................................................ 11 CHAPTER THREE ................................................................................................................... 17 RESEARCH METHODOLOGY ........................................................................................... 17 3.1 Introduction ..................................................................................................................... 17 3.2 Research Design .............................................................................................................. 17 3.3 Data Sources .................................................................................................................... 17 3.4 Empirical Methodology ................................................................................................... 18 3.4.1 Model Specification .................................................................................................. 18 3.4.2 Estimation Methodology ........................................................................................... 19 3.4.2.1 Unit Root Test .................................................................................................... 19 3.4.2.2 The Autoregressive Distributed Lagged (ARDL) approach ................................. 19 3.4.2.3 Diagnostic and Stability test ................................................................................ 22 3.4.3 Economic Growth (GDP) .......................................................................................... 22 3.4.4 External Debt (EDEBT) ............................................................................................ 22 3.4.5 External Debt Servicing (EDEBTSR) ........................................................................ 23 3.4.6 Exchange Rate (EXC) ............................................................................................... 23 3.4.7 Financial Development (FINDVT) ............................................................................ 24 3.4.8 Government Spending (GOVSPD) ............................................................................ 24 3.4.9 Gross Domestic Investment (GRDINVT) .................................................................. 25 3.4.10 Trade Openness (TRDOP) ....................................................................................... 25 CHAPTER FOUR ..................................................................................................................... 27 RESULTS PRESENTATION AND DISSCUSSION ............................................................. 27 4.1 Introduction ..................................................................................................................... 27 4.2 Descriptive Statistics ........................................................................................................ 27 vi University of Ghana http://ugspace.ug.edu.gh 4.3 Ghana External Debt Trend Analysis ............................................................................... 29 4.4 Stationarity Test ............................................................................................................... 30 4.5 ARDL Bounds Tests ........................................................................................................ 32 4.6 Long Run Estimation ....................................................................................................... 33 4.7 Short Run Estimation ....................................................................................................... 35 4.8 Model Diagnostic and Stability Test ................................................................................ 37 CHAPTER FIVE ...................................................................................................................... 39 SUMMARY, CONCLUSION AND RECOMMENDATION ................................................ 39 5.1 Introduction ..................................................................................................................... 39 5.2 Summary ......................................................................................................................... 39 5.3 Conclusion ....................................................................................................................... 40 5.4 Recommendation ............................................................................................................. 41 REFERENCES ......................................................................................................................... 42 vii University of Ghana http://ugspace.ug.edu.gh LIST OF FIGURES Figure 4.1: A Graph of Ghana’s External Debt……………………………………………… 29 Figure 4.2: A Graph of Cumulative Sum of Squares of Recursive Residuals……………… 38 viii University of Ghana http://ugspace.ug.edu.gh LIST OF TABLES Table 4.1: Descriptive Statistics ................................................................................................ 27 Table 4.2: ADF Unit Root Test Results ..................................................................................... 31 Table 4.3: ARDL Bounds Test Results ...................................................................................... 32 Table 4.4: Long Run Estimation Results.................................................................................... 33 Table 4.5: Short Run Estimation Results ................................................................................... 36 Table 4.6: Breusch-Godfrey Serial Correlation LM Test........................................................... 37 ix University of Ghana http://ugspace.ug.edu.gh ABSTRACT This study investigated the long and short-run impact of external debt on economic growth in the context of Ghana using the sample period spanning from 1980 to 2016. The study also examined the Ghana’s external debt trend analysis over the sample period considered for this study. Theoretical and empirical evidences indicates that external debt can either exert negative or positive impacts on economic growth depending on how the external funds acquired is injected into the economy. The variables employed for this were found to be I(0) and I(1) and as result Autoregressive Distributed Lag bounds test to co-integration econometric technique was adopted. The results of this study revealed that the external debt of Ghana has increased by 1,430% from 1980 to 2016. Also, external debt of Ghana was found to have positive significant impacts on economic growth of Ghana in the long run. The external debt servicing which is the cost of external debt was found to have negative insignificant impact on growth of Ghana. The study therefore recommend to policymakers to ensure that all external funding are injected into productive sectors of the economy in order to stimulate economic growth. x University of Ghana http://ugspace.ug.edu.gh CHAPTER ONE INTRODUCTION 1.1 Background of the Study Governments across the globe have a responsibility of financing their expenditures. They mostly depend on tax from citizens to finance expenditures, but revenue from tax alone seems to be insufficient to finance such expenditures. Therefore, governments mostly rely on public borrowing (Public debt) to bridge the gap between what they receive and their expenditures (Safdari, Keramati, & Mahmoodi, 2011). Public debt is the total of a country’s debts which includes domestic debts and external debts (Maku, 2009). Domestic debt is a stock of fund which is borrowed by the government inside the country while an external debt is a stock of fund lent by foreign countries or international financial institutions to another country per a period of time. The 2006/2007 global credit crunch put many economies into latent danger. Private debts which significantly lubricate the wheels of economies have fallen drastically and left several economies near collapse (Reinhart & Regoff, 2010). To resuscitate their economies and rescue the financial sector from collapse, many governments increased their debt. Consequently, the financial crisis resulted in more or less a government debt crunch. The theory on macroeconomics indicated that greater degree of public debt results in stagnation of economic growth through crowding out private investment, greater inflation, greater future nuisance taxes, higher volatility and general vulnerability of the economy to little shocks such as changes in commodity prices world-wide (Cecchetti et al., 2011). The empirical results lend credence to this. According to Reinhart and Regoff (2010), moderate levels of debts improve economic growth; nevertheless, too much debt 1 University of Ghana http://ugspace.ug.edu.gh could be detrimental to welfare. In line with this, the Eastern and Central European economies signed the Maastricht agreement to restrict member countries from exceeding certain levels of debts. The Maastricht agreement requires member countries not to accumulate debts exceeding 60 percent (%) of GDP. Although the economy of Ghana differs in terms of size, legal and institutional arrangement compared to the European economies, its debt level in recent years is very alarming. According to Bank of Ghana, Ghana’s public debts reached a record high of 111.9 percent in 2000. The country by then was in a debt crisis. Consequently, in 2006 Ghana asked for a relief of its debt under the Heavily-Indebted Poor Countries (HIPC) project brought by the International Monetary Fund (IMF) and World Bank. This made the Ghana’s external debts decline sharply to 26.20 percent (%) at the end of 2006. Unfortunately, few years after Ghana benefited from the debt relief programme, the country has returned to its tragic condition of high public debt. Between 2006 and 2016, Ghana’s debt to gross domestic product (GDP) rose from 26.2 percent (%) to 73.1 percent (%). Even though, it dropped to 69.2% in 2017, it increased to 70.7% in 2018. In 2018 alone, the additional debt accumulated was 19.8 percent (%). According to the IMF (2018) report on the extended credit programme, Ghana has continuously breached the public debt yardstick of 56% of GDP and the apparent absence of fiscal adjustment might escalate the public debt profile. The report based on the standard stress test shows Ghana is susceptible to depreciation of its currency (IMF, 2015). This couple with the fact that its debt sustainability is largely susceptible to exchange rate tremors as well as its unsustainable debt trajectories makes Ghana to be heading towards debt 2 University of Ghana http://ugspace.ug.edu.gh distress. In view of the fact that the country still relies on support from other countries and international bodies to fund most of its programmes and projects and has failed consistently to maintain fiscal discipline (IMF, 2015), there is high tendency to raise the country’s indebtedness. Consequently, it is vital to check the impacts of Ghana’s debt on Ghana’s economic growth in both short-run and long-run terms. 1.2 Statement of the Problem Every country aims at attaining sustainable economic growth. This however makes countries responsible for boosting their economic growth to lower their debt burden. Countries according to Baum et al. (2013) are faced with different challenges in their efforts to boost economic growth. Most countries especially developing ones such as Ghana are characterised by high degree of indebtedness, high inflation figures, high unemployment rate, high government spending and absolute poverty coupled with poor economic performance. Moki (2012) highlighted that almost all governments experience budget deficit as a result of high expenditure and lower revenues. Developing economies such as Ghana are faced with low tax collection leading to low revenues, and so opt for debt as a best alternative for financing government budget (Rosen & Gayer, 2008). Considering this, it is evident that public debt plays a significant role in developing countries. According to the Caribbean Development Bank (2013) public debt enables individuals or fiscal authorities to play their role in stabilizing their economies and also stimulate aggregate growth. 3 University of Ghana http://ugspace.ug.edu.gh Again, the public debt (whether internal or external debt) impact on growth of economies in both the short-run and long-run is not clear. For instance, a study by Chudik et al. (2013) show that public debt enhances growth in the short-run and debts exerts harmful effects on growth in the long-run. However, Pegkas (2018) found a long-run positive impact of debt on growth. Also, Ajayi and Oke (2012) indicated that debt has destructive influence on per capita income. According to the Ministry of Finance, as at the close of 2014, Ghana’s debt totaled $24,787.4 million comprising $10,915.6 million domestic debt and $13,871.8 million external debt. The level of public debt rose to $26,403.3 million at the end of 2015 representing 6.51% increase. The rise in public debt was mainly due to an increase in external debt because domestic debt witnessed a decline of 2.7%. The upsurge in Ghana’s debt continued from 2014 through to 2018. However, it is external debt that has troubled policymakers because Ghana’s unstable exchange rate depreciation raises the volume of external debt (Ghana Cedi equivalent). That is, with the cost of borrowing externally at hand, is it worth borrowing externally or does the cost of these loans exceed the benefits to economic growth. Again, several studies conducted focused on cross-country investigations which do not offer explanations that reflect the actual situation in individual countries. It is against all these background that this study investigates the impacts of debts on growth in the context of Ghana. However, the unavailability of data on internal debt meant that the study focuses on the effects of external debt on Ghana’s economic growth. 4 University of Ghana http://ugspace.ug.edu.gh 1.3 Objectives of the Study Overall, the study aims to ascertain the influence debt has on growth in the Ghanaian economy with a specific focus on the external debt component. In order for this broad objective to be realised, the ensuing specific objectives were developed. 1. To examine trends in Ghana’s external debt from 1980 to 2016. 2. To examine the effects (both short-run and long-run) of Ghana’s external debt on Ghana’s economic growth. 1.4 Research Questions 1. What are the trends in Ghana’s external debt from 1980 to 2016? 2. What is the effect of Ghana’s external debt on the growth of the Ghanaian economy in both the short-run and long-run? 1.5 Significance of the Study One of the main issues that negatively impact Low and Middle Income Countries (LMIC) such as Ghana is the unsustainable level of public debts. Therefore, carrying a study of this nature would assist the nation to know whether external debts are good for growth of the economy. The findings from the study in no small way would be relevant to the general public, Policy makers and development institutions. The finding would inform these individuals and institutions in their crusade for or against public debt based on empirical data from Ghana’s history. It will help policymakers to ascertain the volume of public debt that would result in stimulating growth as this would serve as a guide to authorities in charge of the country’s public debt management. It 5 University of Ghana http://ugspace.ug.edu.gh would also be beneficial in terms of shaping, framing, implementation, monitoring and evaluation polices on debt management. The study would also serve as an academic reference material for students, researchers, and other scholars interested in studying similar subject area. 1.6 Research Scope and Limitations This study is restricted to assessing the influence of Ghana’s external loans on the growth of the Ghanaian economy. The data for the analysis spans the period 1980 to 2016. The concentration on only external debts and selection period of the study and variables is greatly influenced by the availability of data. The study adopts an ADRL to remedy the short time period for a time series study. 1.7 Organization of the Study This current study is structured into five chapters. The first chapter provides an introduction to this research. It presents the study background, its purpose and objectives and the questions that the study seeks to answer. It further highlights the importance of the study as well as its scope and limitation. Chapter two looks at the reviews of the comprehensive literature relevant to this study. Chapter three deals with data, variables and methods of data analysis. It also provides justifications for the selected methods and the analysis. In the fourth chapter, the results of data analysis and a discussion of implications are presented. Finally, chapter five highlights the key issues in the study, findings, conclusions and recommendations based on the findings of the study. 6 University of Ghana http://ugspace.ug.edu.gh CHAPTER TWO LITERATURE REVIEW 2.1 Introduction The present chapter reviews the theories relevant for this study as well as the empirical literature on the nexus between public debts on the growth of economies. The summary of gaps in the literature and the main points are presented at the end. 2.2 Theoretical Review Economic growth is viewed as a rise in the level of economic activities (measured by the GDP) over a time period. In other words, economic growth is a quantitative change in national output. It can also be described as a process of “steady increase in the productive capacity of the economy over a given time period which leads to higher national production and income” (Todaro & Smith, 2006). Growth is characterized by increased productivity, structural transformation, increases in per capita income and international flow of factors of production (Ajibike, 2016). In the theoretical literature on economic growth, one of the main dialogues is about the sources of growth. Theorists have sought to find answers to the question: where does growth come from? Two dominant viewpoints give answers as “outside” and “inside”. The earliest theorists who attempted to answer the question on determinants of economic growth were of the view that the factors that cause economies to grow are not factors that are found in the economic system; that is they are outside the system. This viewpoint is collectively called the neoclassical theories. These theories effectively argue that growth is exogenous to the economy. The implication of 7 University of Ghana http://ugspace.ug.edu.gh this theoretical perspective is that policy measures which target economic variables are not potent in bringing about growth (Ngepah, 2017; Laeven, Levine & Michalopoulos, 2015; Bayraktar-Saǧlam & Yetkiner, 2014; Ayadi, Arbak, Ben-naceur, & Groen, 2013). They implied that for an economy to grow, we just have to sit down and wait for it to happen. It is easy to see that scientists and economists would want to deliberately cause growth to happen rather than wait for the “gods” to give us luck. Thus, a new school of thought evolved, they developed theories of economic growth that made it possible for intervention to be possible. This school, known as the Endogenous growth theories, argues that growth can be induced by implementing measures that affect one or more economic variables (Azam & Ahmed, 2015; Bayraktar-Saǧlam & Yetkiner, 2014; Ganegodage & Rambaldi, 2014; Laeven, Levine & Michalopoulos, 2015). Their models allow for economic growth to be influenced by policies aimed at changing economic variable hence, growth can come from “inside” factors. One such variable is financial development (such as banking sector development). An important implication of the neoclassical perspective was that the growth rate of economies would eventually converge. New data however showed that countries which had better technological innovations were growing relatively faster. 2.2.1 Neoclassical Growth Theories -Solow Growth Model Neoclassical theories try to elucidate economic growth and income distribution symmetrically given the relative scarcities of land, labour and capital. A major shortfall of these theories is that they fail to explain the sources of productivity. Also, they assume that growth is influenced by 8 University of Ghana http://ugspace.ug.edu.gh exogenous factors. The Solow model is the workhorse of neoclassical growth models and all other neoclassical models represent variations of the Solow model. Solow (1956) developed what is now known as the Solow growth model to describe how higher levels of capital stock can yield higher per capita output. The starting point of his proposition is the assumption that national savings is a constant proportion of income and that supply of labour (population growth) is at a constant rate and the capital per worker (capital intensity) can be regulated. He argued that the capital per worker depended on the factors of production relative prices. Also, the theory states that increasing capital per worker would lead to ever decreasing rate of increase to the output level (diminishing returns). This implied that the growth rate of labour, capital and total output will become identical in the long term given no technical progress. Thus, per capita output and real wage will not increase so growth will not be continued in the long term by increasing the savings rate. The model implies that countries that have similar technologies and preferences will converge to the same constant level of GDP. However, wealthy economies that have greater savings rate will experience greater productivity and greater real income and a falling marginal product of capital. By implication, a great impact will be observed in developing economies when capital moves from wealthy nations to less endowed nations. Thus, this speeds growth of economies in developing countries. Furthermore, the model argues that without technological progress, the rate of growth will be constant regardless of the rate of savings, hence, growth will be dependent exclusively on increased labour supply, which is a factor oriented in technological progress. Good health and education are the major variables that can help a nation to harness the gains in human capital 9 University of Ghana http://ugspace.ug.edu.gh innovation (Serieux & Samy, 2001). Nevertheless, the high burden of public debt and public debt servicing on governments in less developed economies makes it difficult for these governments to allocate the rightful resources to improve human capital thereby creating a negative connection amid public debt servicing and human capital development (Serieux & Samy, 2001). 2.2.2 The Dual Gap Theory The economic growth of every country is supported and maintained by increased investment in both public and the private sector. As investment is spurred by increase in savings, it implies that nations with low levels of savings, a common feature of developing countries, will have a very feeble economic growth (Hunt, 2007). There is a certain level of capital threshold that will guarantee sustained economic growth. This means that if domestic savings fall short of this threshold, the dawn of foreign inflows (external public debt) should result in automatic rise in growth due to an increase in savings during the period. This conduit is what is known as the dual gap theory. This gap is dire in developing countries hence Mckinnon (1964) suggest developing countries should embrace external debt as an urgent funding source to fill this gap. However, how this support is put to use determines the outcome of results. For borrowed funds to provide maximum benefits, the funds need to be invested in areas that will generate sufficient returns to be able to satisfy the debt servicing obligation at maturity. To throw more light on this theory, Laurenceson (2002) espoused a two-gap model. The model means that capital flowing from foreign economies result in economic growth in developing economies. This capital inflows aid developing countries to experience higher investment 10 University of Ghana http://ugspace.ug.edu.gh compared to the local savings. By supposing that foreign exchange gap is binding, the rise in import is brought by a foreign capital inflow which suggests growth. 2.3 Empirical Literature Review The dawn of the 2007/8 financial crunch reignited the need to revisit debt and economic growth connection and identify the conduit via which debt impact economies. Nevertheless, the empirical evidences on how economic growth is influenced by public debt predated the financial crunch. To start with, to ascertain how public debt impact growth via total factor productivity and factor accumulation, Patillo et al. (2004) employed a sample of 61 developing economies. Their study reports that minimum levels of debt increases economic growth, however, debt beyond a given level does not promote growth. They further found that if the debt level is doubled for countries that have high debt to GDP, such countries’ GDP would fall by 1 percent. But productivity of the factors and physical capital would even fall more than 1 percent. To extend the work of Patillo et al. (2004), Schclarek (2004) included another dimension of debt, that is private debt into the analysis. The sample for the study consists of 24 developed countries and 59 least developed economies. Using the general method of moment (GMM), the study showed that small level of external debt improves GDP growth but no effect on factor productivity growth was observed. In order to establish whether threshold levels exist for the efficacy of debt accumulation, Reinhart and Rogoff (2010) group debt as a ratio of GDP of the nations into 4 categories for both developing and advanced countries. The initial category is the period at which public debt to 11 University of Ghana http://ugspace.ug.edu.gh GDP is less than 30% (low debt); next category is the period where debt to GDP ranges between 30% and 60% called “medium debt”; the third category is the period at which debt to GDP ratio ranges between 60% and 90% known as “high debt”; the final range is where debt to GDP ratio is beyond 90%, it was found out that for both advanced and emerging countries whose ratio of debt-GDP is above 90 percent threshold they experience lower GDP growth compared to the other categories. On the contrary, Kumar and Woo (2010) using diverse assessment methods like fixed effects, system GMM and pooled OLS revealed that if the debt to GDP ratio level is beyond 90%, the decrease in growth is high for emerging economies than advanced economies. Similarly, Pattillo et al. (2011) conducted on 93 developing economies provided evidence that debt to GDP level as low as 35 percent to 40 percent exerts negative impacts on growth. This is lower as compared to the conclusion from past studies, showing there is no clear empirical evidence on what debt-GDP ratio is detrimental to growth. Fiscal policy is somewhat linked to public debt because future tax obligations are required to offset public debt. As a result Aghion, Hemous and Kharroubi (2009) assessed the extent to which the cyclical fiscal policy could influence growth. In a study to ascertain the association between the burden of debt accumulation and growth of 16 economies over the period 1971- 2007, Cunningham (1993) provided evidence which suggests debt accumulation has a detrimental consequence on growth. On the contrary, the Chowdhury (1994) rejected this finding. The author argues that the decrease in growth in emerging economies cannot be linked 12 University of Ghana http://ugspace.ug.edu.gh to debt accumulation and that it is a mere exaggeration therefore, there is no need to establish debt relief institutions. Using instrumental variables regression approach, Panizza and Presbitero (2012) analysed the consequence of debt upon growth on OECD economies. The instrumental variable method was selected because it is perceived that when foreign debt is denoted in foreign currency, variation in the rate of exchange the country has an undeviating as well as mechanical impact on debt- GDP ratio. Their data was separated into high and low debt economies. Their study shows the harmful influence of debt on growth vanishes as debt is instrumented with variables that capture valuation effects caused via interacting foreign currency as well as exchange rate fluctuation in economies that have low debt. This finding is inconsistent with previous studies and therefore much need to be done about this subject matter. In a cross-country study for ten Central and Eastern European nations during the period 2001 to 2012, Vosyliūtė (2014) examined how debt influence growth. The study used several methods including system GMM, dynamic OLS and fixed effect model. The results indicate that debt significantly affect real capital growth as well as total factor productivity. The results were explained by the neoclassical Solow growth model. In a related study in Sub-Saharan Africa, Iyoha (1999) used a simulation method from the time 1970-1994 to investigate external debt-GDP association. The study findings buttress the fact that external debt accumulation is deleterious to investment and that a decline in debt accumulation improves economic growth as well as investment. 13 University of Ghana http://ugspace.ug.edu.gh Similarly, Fosu (1999) also studied the impact of debt obtained externally on growth of Sub- Saharan Africa economies. The author used data spanning from 1970 to 1996 and sourced from World Bank to examine this nexus. Employing “augmented production function”, the findings of the paper revealed that debt using both debt service and outstanding debt as proxy has deleterious effects on growth of Sub-Saharan economies. They found weak evidence of debt on investment. Other studies have also investigated the debt long period impact on growth. For instance, to ascertain the debt long period effect on Greece’s growth after the financial crisis, Pegkas (2018) employed data on trade openness, government consumption, debt to GDP ratio, investment, and population. The data was sourced from AMECO database. The study found among other things, the evidence of long-run nexus among the employed variables. Specifically, trade openness, government consumption and investment have a favourable link to growth. However, population and debt exert an unfavourable impact on economic. The author further established that the connection amid debt to GDP growth relies on debt thresholds. For instance, when the proportion of debt to GDP rose before the year 2000, the debt increment did not have significant deleterious impact on growth. Nevertheless, increment in debt level significantly decreased the growth when public debt increased after the year 2000. Until now the study has reviewed previous studies that have conducted cross –country analysis. The study will now review recent empirical studies on single countries. Saifuddin (2016) 14 University of Ghana http://ugspace.ug.edu.gh investigated the impacts of public debt on Bangladesh’s economic growth. The study used growth and investment model. The data employed for the study spans from the period 1974 to 2014. The TSLS regression was methodology employed. Results point out that debt has positive significant effect on economic growth via positive impacts on investment. In Bangladesh, another investigation by Hassan and Akhter (2012) examined the influence of debt on growth. The period of sample was 1980 to 2011. The study’s outcome revealed there is no significant adverse connection between loans contracted outside the country and GDP growth. It was further revealed that loans contracted within the country have little statistical significant destructive influence on growth. To assess the influence of loans contracted outside a country on Nigeria’s economic growth, Ajayi and Oke (2012) showed that loans contracted outside a country had a destructive influence on the country’s per capital income. It was further added that high levels of loan contracted outside the country resulted in a fall in value of the Nigeria’s currency, increased employees cost-cutting, and result in incessant work strikes. Weak educational system was also found to be the product of servicing high level external debt. Studies in the context of Ghana on the effects of debt on economic growth were found to be few and in the studies, the researcher found used outdated and a short time period for the analysis. For instance, Frimpong and Oteng–Abayie (2006) investigated the influence of loans contracted outside Ghana on Ghana’s growth. The authors used Johansen cointegration method and Vector Error Correction Model (VECM) approaches over the period 1970-1990. The study 15 University of Ghana http://ugspace.ug.edu.gh found indication of long run connection, foreign debt and growth consistent with what was found in Nigeria by Egbetude (2012). The study further reveals that the inflow of external loans is favourable to growth in Ghana, however, debt servicing has a harmful consequence on growth. Also, the study shows the evidence of crowding out consequence. Most of these studies have been concentrated on cross-studies and this gave just average impact on the countries in which the studies have been conducted and this do not allow better analysis of the impact of debt on each country. Again, single studies conducted on Ghana’s economy used data that is quite outdated as well as relatively short times and this may not reflect recent happenings in debt profile of Ghana’s economy thereby creating a gap for this current study to use updated data, to assess not only long period and short period effects of loans contracted outside a country on economic growth but public debt in general. 16 University of Ghana http://ugspace.ug.edu.gh CHAPTER THREE RESEARCH METHODOLOGY 3.1Introduction This chapter addresses the data sources, data definition, research design and the empirical model employed for this study to avoid invalid and inconsistent conclusions. 3.2 Research Design The study adopts a quantitative research design. This is most appropriate in order to fulfil the purpose of the research and to help achieve the research objective. The researcher upon considering the objective of the study adopted the causal research design. Causal research, also called explanatory research is the investigation of cause-and–effect relationship. In order to determine causality, it is important to observe variation in the variables that is assumed to cause the change in the other variables(s), and then measure the change in the other variables (s). Other confounding influences need to be controlled for so they don’t distort the results, by holding them constant in the experimental creation of data. The design is appropriate for the study since it shows the effect of external debt on economic growth in Ghana. 3.3 Data Sources In this study, a secondary data of annual external debt, debt serving, real GDP, FDI, exchange rate, financial development, government spending, gross domestic investment and trade openness were sourced from the World Development Indicator (WDI) 2017. The WDI database which is hosted by World Bank is one of the most reliable and credible data sources and is used by several researchers and scholars. This secondary data span from 1980 to 2017. The data used 17 University of Ghana http://ugspace.ug.edu.gh were in annual frequency and as a result approximately 37 observations of all the variables were used. 3.4 Empirical Methodology 3.4.1 Model Specification The basic objective of this current study is to examine the effect of Ghana’s external debts on Ghana’s GDP growth in both the short-run and the long-run. For the stated objective to be addressed, the general time series regression model employed is given in equation3.1. 𝐆𝐃𝐏𝐭𝒇 = (𝐄𝐃𝐄𝐁𝐓𝐭, 𝐄𝐃𝐄𝐁𝐓𝐒𝐑𝐭, 𝐄𝐗𝐂𝐭, 𝐅𝐈𝐍𝐃𝐕𝐓𝐭, 𝐆𝐎𝐕𝐒𝐏𝐃𝐭, 𝐆𝐑𝐃𝐈𝐍𝐕𝐓𝐭, 𝐓𝐑𝐃𝐎𝐏𝐄𝐭) − 𝟑. 𝟏 where GDP = Gross Domestic Product; EDEBT = External Debt; EDEBTSR = External Debt Servicing; EXC = Exchange rate; FINDVT = Financial Development; GOVSPD = Government Spending; GRDINVT = Gross domestic investment; TRDOPE = Trade Openness; 𝑡(𝑡 = 1,2,3,4, … ) represents the variables time variant nature. The variables introduced into this model were based on theoretical and empirical review. From equation 3.1 above, it is observed that economic growth (GDP as proxy) is the endogenous variable whiles the exogenous variables are made up of exchange rate, trade openness, external debt, debt servicing, financial development, government spending and gross domestic investment. 18 University of Ghana http://ugspace.ug.edu.gh Based on the Equation3.1, an estimable version which contains one dependent variable and seven independent variables in the logarithm form are specified as follows: 𝐈𝐧 𝑮𝑫𝑷𝒕 = 𝜶 + 𝜷𝟏𝐈𝐧 𝐄𝐃𝐄𝐁𝐓𝐭 + 𝜷𝟐𝐈𝐧 𝐄𝐃𝐄𝐁𝐓𝐒𝐑𝒕 + 𝜷𝟑𝐈𝐧 𝐄𝐗𝐂 + 𝜷𝟒𝐅𝐈𝐍𝐃𝐕𝐓𝒕 + 𝜷𝟓𝐈𝐧 𝐆𝐎𝐕𝐓𝐒𝐏𝐃𝒕 + 𝜷𝟔𝐈𝐧 𝐆𝐑𝐃𝐈𝐍𝐕𝐓 + 𝜷𝟕𝐓𝐑𝐃𝐎𝐏𝐄𝒕 + 𝜺𝒕 (𝟑. 𝟐) where In = natural logarithm, 𝛼 = constant, 𝜀𝑡 = error term, 𝛽1, 𝛽2, …, 𝛽6 are the coefficients of the exogenous variables and all other variables follows the definition in equation 3.1. 3.4.2 Estimation Methodology 3.4.2.1 Unit Root Test One of the major characteristics of financial data is the non-stationarity of the data. Brooks (2008) indicated that using non-stationary data for regression will yield a spurious regression result. That is, to avoid spurious regression, the data must be stationery and this can be achieved by performing the unit root test to ensure stationarity of the data. However, there are several statistical tools for testing the unit root test of a dataset. These include “Augmented Dicker- Fuller tests”, “Phillips and Perron test”, “Kwiatkowski-Phillips-Schmidt-Shin test”, and “Ng and Perron test”. However, the Augmented Dicker-Fuller test is adopted to test the stationarity of the datasets employed due to the fact that it is the most widely used in the literature. 3.4.2.2 The Autoregressive Distributed Lagged (ARDL) approach Several quantitative methods have been adopted to explore the long-run and short-run nexus among variables in the literature. These methods include “Engle-Granger”, “Johansen cointegration” and “Autoregressive Distributed Lag (ARDL)” econometric techniques. The 19 University of Ghana http://ugspace.ug.edu.gh ARDL econometric approach is employed in this study to explore the Ghana’s external debt effects on Ghana’s growth in both the short-run and long-run. The ARDL model is an econometric model that contains the dependent variable’s lagged values and the regressors’ current and lagged values. The ARDL was adopted due to some advantages reported in the literature over the other models. For instance, the ARDL bounds test do not demand pre-testing of the variables to establish their integration order since the test is valid for series with purely I (0), purely (1) or mix of both but not I (2) or more. Also, the endogeneity challenge of not being able to test hypotheses on the long run estimated coefficients is solve by ARDL model. According to Pesaran, Shin and Smith (2001), modelling using ARDL with the correct lags has the benefits of correcting endogeneity and serial correlation shortfalls. Finally, the ARDL is credited with superior small sample size results characteristics compared to the other models (Pesaran et al., 2001). Following Pesaran et al. (2001), the generalised ARDL (p, q) model is given as: 𝒀 𝒍𝒕 = 𝜷𝟎 + ∑𝒊=𝟏 𝜷𝒊 𝒚𝒕−𝒊 + ∑ 𝒌 𝒊=𝟎 𝝈𝒊 𝒙𝒕−𝒊 + 𝜺𝒕 (𝟑. 𝟑) where Y is the endogenous variable, 𝑥 are the exogenous variables, 𝛽𝑖 and 𝜎𝑖are the short-run coefficients, 𝛽0 is the intercept and 𝜀𝑡 is the disturbance term, and 𝑙 and 𝑘 is the optimal lag lengths. The application of the ARDL technique involves two levels. Firstly, the examination of the long- run nexus existence by estimating “ARDL bounds test for cointegration”. The bounds test report the F-statistics. To perform the bounds test for cointegration, the ARDL model in equation (3.3) is specified as: 20 University of Ghana http://ugspace.ug.edu.gh 𝒍 𝒌 ∆𝒀𝒕 = 𝜷𝟎 + ∑ 𝜷𝒊 ∆𝒚𝒕−𝒊 + ∑ 𝝈𝒊 ∆𝒙𝒕−𝒊 + 𝜽𝟏𝒚𝒕−𝟏 + 𝜽𝟐𝒙𝒕−𝟏 + 𝜺𝒕 (𝟑. 𝟒) 𝒊=𝟏 𝒊=𝟎 where 𝑌 = target variable; 𝑥 = exogenous variables; 𝛽𝑖 and 𝜎𝑖 = short-run coefficients; 𝜃1 and 𝜃2 = long-run coefficients; 𝜀𝑡 = disturbance (white noise) term; 𝑛, 𝑘 = lag length; 𝛽0 = constant; and ∆ = differences operator. Pesaran et al. (2001) offered two types of appropriate critical values with the assumption that the models have an intercept or trend or both taking into consideration the number of regressors. One section of critical values postulates I (0) for the regressors whiles the other one assumes I (1) for the regressors. The decision criterion is that, for a given level of significance, there exists a significant statistical evidence of cointegration when the F-statistic value obtained from the bounds test lies “above the critical value of the upper-bound”. On the other hand, given the significance level, there is enough statistical evidence to show the absence of long-run nexus among the variables whenever the F-statistics value recorded from the bounds test “lies below the critical value of the lower bound”. However, there will be inconclusive result when the F- statistics value obtained from the bounds test “lies within the upper- and lower-bound”. The null and alternative hypothesis for the F-test is given as: 𝑯𝟎: 𝝈𝟏 = 𝝈𝟐 = 𝝈𝟑 = 𝝈𝟒 = 𝝈𝟓 = 𝟎 against 𝑯𝟏: 𝝈𝟏 ≠ 𝝈𝟐 ≠ 𝝈𝟑 ≠ 𝝈𝟒 ≠ 𝝈𝟓 = 𝟎 (𝟑. 𝟓) When cointegration is established then the second level of the analysis is performed. The second level involves the long-run and short-run nexus between the variables under study estimation. However, it is important stating that the error correction model (ECM) assists the researcher to identify how disequilibrium in the past is corrected in current periods. 21 University of Ghana http://ugspace.ug.edu.gh 3.4.2.3 Diagnostic and Stability test In order to ensure reliable and robust estimation results, reliability and diagnostic tests were conducted. To ensure functional correctness form, the Ramsey’s RESST test was adopted. Similarly, autocorrelation and heteroscadasticity of the residuals were tested using Lagrange multiplier test and the regression of squared residuals respectively. The CUSUM of Squares test was employed to exam the model’s stability. Variables 3.4.3 Economic Growth (GDP) Economic growth refers to “market value of all goods and services produced by the factors of production within an economy annually” (WDI, 2017). In mathematical terms, it is the summation of the aggregate consumption, investment, expenditure of government and net exports. Economic growth is employed by researchers and policy makers to measure the performance of economies around the globe. The economic growth variable is the endogenous variable employed in this study. Following Muhanji and Ojah (2011) and Egert (2012), the real GDP per capita sourced from the WDI is used as the measure for economic growth. 3.4.4 External Debt (EDEBT) Total external debt refers to “the debt an economy owed to non-residents which are usually repayable in good or services or in monetary terms. That is, external debt of an economy is the aggregate of public, publicly guaranteed, and private nonguaranteed long term debt, IMF credit and short term debt” (WDI, 2017). External debt impact on growth of economies reported in the 22 University of Ghana http://ugspace.ug.edu.gh literature is ambiguous. That is, some empirical studies reported negative impacts whiles other studies (such as Abass & Christensen, 2007) reported positive external debt influence on growth of economies. Therefore, the impact of debt on growth in this study is expected to be positive or negative. The external debt stock, total variable sourced from WDI is used as proxy for debt in this study by following Hassan and Akhter (2012) and Quresi and Ali (2010). 3.4.5 External Debt Servicing (EDEBTSR) Debt service is the amount of money required to make payment of the interest and the principal within a period of time. Debt servicing is the proportions of the revenues employed to settle the debt. It is explained in the literature that debt servicing can exert positive and negative influence on the economic growth. The economy would experience negative impacts of debt servicing when there is debt overhang. That is, debt overhang is observed when the benefit accruing from debt is lower than the service of the debt. On the contrary, debt servicing has positive impacts on growth if it facilitates a profit atmosphere between creditors and other investors in the economy. Debt service on external debt, as indicated in WDI is employed as proxy for debt servicing in this study. Therefore, in this study, debt servicing estimate is expected to be either positive or negative. 3.4.6 Exchange Rate (EXC) Exchange rate “is the price of one country’s currency in terms of another country’s currency” (WDI, 2017). According to Korsi (2015), when there is fall in value of domestic currency, the burden or the cost of the external debt increases. That is, domestic currency depreciation makes 23 University of Ghana http://ugspace.ug.edu.gh the debt expensive to meet. The exchange rate estimate is expected to be either positively or negatively significant in this study. 3.4.7 Financial Development (FINDVT) The financial development of an economy as reported in the literature exerts significant impacts on growth of economies (Beck & Levine, 2004). The impacts of financial development in an economy mostly rest on the macroeconomic policies and the quality level of the financial institutions. Following Beck and Levine (2004), the proxy for financial development is domestic credit to private sector by banks (% of GDP). Financial development would trigger positive growth when financial institutions perform their duties of directing investment to the productive sectors of the economy. In the contrary, financial development would trigger negative economic growth when the financial intermediaries fail to channel investment to the appropriate industries of the economy. Hence, the estimates of the financial development variable are anticipated to positively or negatively significant. 3.4.8 Government Spending (GOVSPD) There are two schools of thought with regard to government expenditure and economic growth nexus in the literature. Firstly, the Keynesian school of thought argues that a rise in the rate of government expenditure implies the expansion in growth (Salih, 2012). On the contrary, the Wagner theory argues that the growth of the economy triggers similar growth in public expenditure. General government final consumption spending is employed as the measure of 24 University of Ghana http://ugspace.ug.edu.gh government spending. Therefore, the coefficient of the government expenditure is expected to be significant and can be either positive or negative 3.4.9 Gross Domestic Investment (GRDINVT) Investment is documented to have both positive and negative impacts on growth. The positive impacts on growth are experienced in the form of competitive experience, technology advancement and knowledge which enhances production and as result trigger economic growth. Gross domestic investment “consists of outlays on additions to the fixed assets of the economy plus net changes in the level of inventories. Fixed assets include land improvements (fences, ditches, drains, and so on); plant, machinery, and equipment purchases; and the construction of roads, railways, and the like, including schools, offices, hospitals, private residential dwellings, and commercial and industrial buildings. Inventories are stocks of goods held by firms to meet temporary or unexpected fluctuations in production or sales, and work in progress” (WDI, 2017). The GDI coefficient in the study is also expected to be significant and can be less than or greater than zero. GDI is introduced into the model because empirical evidences (such as Misztal, 2011) have documented its impact on growth. 3.4.10 Trade Openness (TRDOP) The nexus between trade openness and growth has been established in the literature. Some empirical evidences (such as Chang & Mendy, 2012; Rao & Rao, 2009) have reported positive effects of trade openness on growth of economies. Contrary, some researchers (such as Zanohogo, 2017) indicated that trade openness exerts significant negative impacts on growth. 25 University of Ghana http://ugspace.ug.edu.gh However, some studies (such as Eris & Ulasan, 2013; Babatunde, 2011) also reported no link and little influence of trade openness on growth. Based on this background, trade openness is introduced into this model as a control variable. Following Chang and Mendy (2012) and Zanohogo (2017), the proxy for trade openness is the ratio of exports plus imports to GDP. 26 University of Ghana http://ugspace.ug.edu.gh CHAPTER FOUR RESULTS PRESENTATION AND DISSCUSSION 4.1 Introduction The previous chapter discussed the research methodology used for this study. This current chapter covers the discussion on the results analysis, presentation and interpretation. 4.2 Descriptive Statistics The descriptive statistics of the variables employed are depicted in Table 4.1. Specifically, the mean, maximum and minimum values, standard deviation, skewness, Jarque-Bera normality test and number of observations on the GDP, external debt, external debt servicing, exchange rate, financial development, government spending, gross investment and trade openness are depicted in Table 4.1. Table 4.1: Descriptive Statistics LGDP LEDEBT LEDEBTSR LEXC LFINDVT LGOVSPD LGRDINVT LTRDOPE Mean 6.916013 22.40788 19.68580 −2.09858 1.980354 2.444762 2.801924 3.986766 Maximum 7.442880 23.78647 21.34281 1.363486 2.963600 3.039173 3.458987 4.754008 Minimum 6.553259 21.05857 18.52552 −8.19928 0.433254 1.768370 1.217176 1.843774 Std. Dev. 0.258756 0.720486 0.613211 2.794504 0.782116 0.286963 0.602020 0.713082 Skewness 0.760171 −0.06564 0.278974 −0.84262 −0.47479 0.110099 −1.30092 −1.44101 Jarque-Bera 3.990748 0.314351 0.678164 4.497611 3.350249 0.137924 11.09304 15.68131 Probability 0.135963 0.854554 0.712424 0.105525 0.187285 0.933362 0.003901 0.000393 Observation 37 37 37 37 37 37 37 37 27 University of Ghana http://ugspace.ug.edu.gh Source: Researcher’s own calculations (2019) using Eviews 9.0 (Data from WDI); Note: L= natural log of the series; GDP = Gross Domestic Product; EDEBT = External Debt; EDEBTSR = External Debt Servicing; EXC = Exchange rate; FINDVT = Financial Development; GOVSPD = Government Spending; GRINVT = Gross domestic investment; TRDOPE = Trade Openness; Std. Dev. = Standard Deviation From Table 4.1, the mean of the natural logarithm of the Gross Domestic Product (LGDP) of Ghana used as a measure for Ghana’s economic growth between 1980 and 2016 is approximately 6.9 with 7.4 and 6.6 as the maximum and minimum values respectively. The GDP recorded a standard deviation value of 0.26 which implies the volatility of the Ghana’s GDP from 1980 to 2016 was not strong compared to its mean value. The GDP series is positively skewed which implies that the LGDP is skewed to right and leptokurtic. The Jarque-Bera normality test result shows that the LGDP is normally distributed at 1% significant level. Also, the log of Ghana’s external debt (LEDEBT) recorded an average value of 22.4 with 23.8 and 21.1 as the maximum value and minimum values respectively. The LEDEBT recorded a standard deviation value of 0.72 which implies the volatility level of Ghana’s external debt was low compared to the mean. The LEDEBT series is negatively skewed (-0.066) which suggest that the LEDEBT variable is skewed to the left. However, LEDEBT is normally distributed based on the Jarque-Bera test result. The summary statistics on the control variables, namely, natural log of Ghana’s external debt servicing (LEDEBTSR), natural logarithm of exchange rate (LEXC), natural logarithm financial development variable (LFINDVT), natural logarithm of government spending (LGOVSPD), natural logarithm of gross domestic investment (LGRDINVT) and natural logarithm of trade openness (LTRDOPE) are also depicted in Table 4.1. From Table 4.1 shows LEDEBTSR, LEXC, LFINDVT, LOVSPD, LGRINVT and LTRDOPE recorded mean values of 19.7, -2.1, 2.0, 2.4, 2.8 and 4.0 respectively. Similarly, LEDEBTSR, LEXC, LFINDVT, LOVSPD, 28 University of Ghana http://ugspace.ug.edu.gh LGRDINVT and LTRDOPE recorded standard deviation values of 0.6, 2.8, 0.8, 0.3, 0.6 and 0.7 respectively. The standard deviation of the LEXC (2.8) is high compared to its mean (-2.1) which implies the Ghana’s exchange rate is more volatile. The Jarque-Bera test results shows that all the control variables employed are normally distributed with the exception of LGRINVT and LTRDOPE. 4.3 Ghana External Debt Trend Analysis To address the research objective two, a trend analysis of Ghana’s external public debt from 1980 to 2016 was conducted and is depicted in Figure 4.2. It is observed from the graph below that the external debt of Ghana increased from 1980 to 2016. It increased from 1980 value of $1,398.362 to a value of $21,396.063 million by the end of 2016. In percentage form, the external debt of Ghana increased by approximately 1,430% from 1980 to 2016. This large percentage may be due to Ghana’s unstable exchange rate depreciation which increases the burden of external debt because the cedi equivalent amount increases for same amount of a foreign currency. Figure 4.1: A Graph of Ghana's External Debt 24.0 23.5 23.0 22.5 22.0 21.5 21.0 80 82 84 86 88 90 92 94 96 98 00 02 04 06 08 10 12 14 16 29 University of Ghana http://ugspace.ug.edu.gh Source: Researcher’s Own Construct (2019) using Eviews 9.0 (Data from WDI) A close observation of the graph shows a steady increase in the external debt of Ghana from 1980 to 2003. The external debt increased from 1980 value of $1,398.362 million to a value of $8,301.007 million as at 2003 representing approximately 493.6% increase in Ghana’s external debt from 1980 to 2003. Beginning from 2003 to 2006, Ghana witnessed a decrease in its external debt from $8,301.007 million to $3,697.633 respectively. This represent approximately 55.5% fall in the external debt of Ghana from 2003 to 2006. From the year 2005 to 2006 for instance, Ghana’s external debt decline from $7334.163 million to $3697.633 million representing approximately 50% change which represent the highest change in Ghana’s external debt from 1980 to 2016. This fall in Ghana’s debt could be as a result of the declaration of Ghana as Highly Indebted Poor Country (HIPC) that reduced the external debt burden on Ghana drastically. However, within a ten year span, from 2006 to 2016, the external debt of Ghana rose sharply from the 2006 value of $3,697.633 million to $21,396.063 million as at 2016. For instance, Ghana’s external debt witnessed 36% change from 2006 to 2007 from 2006 external debt of $3,697.633 million to 2007 value of $5041.282 million. This implies that, from 2006 to 2016, Ghana documented approximately 478.6% increase in its external debt. 4.4 Stationarity Test For the ARDL cointegration test to be performed, the variables must be integrated at an order of zero or one or mixture but must not be higher than one. Also, Brooks (2008) indicated that series 30 University of Ghana http://ugspace.ug.edu.gh that are not stationary leads to spurious regression results. As a result, the integration features of the variables are tested using ADF test. The ADF test’s outcomes are depicted in Table 4.2. Table 4.2: ADF Unit Root Test Results Variables With constant and trend Stationarity Status t-statistics P-value LGDP −1.707061 0.7317 ∆LGDP −5.703208∗∗∗ 0.0001 LEDEBT −1.828551 0.6745 ∆ LEDEBT −6.705205∗∗∗ 0.0000 LEDEBTSR −1.887930 0.6446 ∆ LEDEBTSR −6.845515∗∗∗ 0.0000 LEXC −1.491166 0.8180 ∆ LEXC −4.048665∗∗∗ 0.0140 LFINDVT −2.190223 0.4835 ∆LFINDVT −6.588488∗∗∗ 0.0000 LGOVSPD −2.973772 0.1507 ∆LGOVSPD −5.705072∗∗∗ 0.0001 LGRDINVT −2.705997 0.2391 ∆ LGRDINVT −5.829397∗∗∗ 0.0001 LTRDOPE −5.020306∗∗∗ 0.0010 I(0) Source: Author’s own calculations (2019) by using E-views 9.0; Note B: L= natural log of the series; ∆ = first difference of the series; GDP = Gross Domestic Product; EDEBT = External Debt; EDEBTSR = External Debt Servicing; EXC = Exchange rate; FINDVT = Financial Development; GOVSPD = Government Spending; GRINVT = Gross domestic investment; TRDOPE = Trade Openness; *** indicates significant at 1%. From the Table 4.2, all the variables employed for this study (LGDP, LEDEBT, LEDEBTSR, LEXC, LFINDVT, LGOVSPD, LGRINVT, LTRDOPE) are not stationary at level except trade openness (LTRDOPE). That is, at level, LGDP, LEDEBT, LEDEBTSR, LEXC, LFINDVT, LGOVSPD and LGRINVT were non-stationary. However, the first difference of these variables 31 University of Ghana http://ugspace.ug.edu.gh (∆LGDP, ∆LEDEBT, ∆LEDEBTSR, ∆LEXC, ∆LFINDVT, ∆LGOVSPD and ∆LGRDINVT) were documented to be stationary at 1% significance level. Thus, LTRDOPE was I(0) whiles LGDP, LEDEBT, LEDEBTSR, LEXC, LFINDVT, LGOVSPD and LGRDINVT were I(1). Therefore, the mix of I(0) and I(1) are the integration features of the variables and as a result the study adopts ARDL econometric model to co-integration since none of the variable is integrated above I(1). 4.5 ARDL Bounds Tests Co-integration using ARDL approach was employed to examine the existence of long-run nexus among the variables under consideration. ARDL (1, 0, 0, 0, 1, 0, 0) model was the best model according AIC selection. The results of the ARDL bounds test estimation is depicted in Table 4.3. The results show that the F-statistic value (4.67) lies above the upper bound critical values at 10% (3.3), 5% (3.5) and 1% (4.26) significance level so therefore the null of no existence of cointegration among the variables is rejected at 1% significance level. This implies that there is statistical evidence of co-integration relationship. That is, there exist a long-run nexus among LGDP, LEDEBT, LEDEBTSR, LEXC, LFINDVT, LGOVSPD, LGRDINVT and LTRDOPE. This also implies both short-run and long-run nexus can be estimated. Table 4.3: ARDL Bounds Test Results Test Statistic Value K F-statistic 4.669939 7 Significance I0 Bound I1 Bound 10% 2.03 3.13 5% 2.32 3.5 2.5% 2.6 3.84 1% 2.96 4.26 Source: Researcher’s own calculations (2019) by using Eviews 9.0. 32 University of Ghana http://ugspace.ug.edu.gh 4.6 Long Run Estimation To address the objective two of this study, the long-run estimation among the variables was performed. The result of the long-run estimation is presented in Table 4.4 below. The variables are in their lag one form since one lag is the optimal lag length structure for this study. Table 4.4: Long Run Estimation Results Variable Coefficient Std. Error t-Statistic Prob. C 3.886698∗∗∗ 0.830280 4.681190 0.0002 LDEBT(-1) 0.210250∗∗∗ 0.054029 3.891398 0.0010 LEDEBTSR(-1) −0.018210 0.036592 -0.497642 0.6244 LEXC(-1) 0.105652∗∗∗ 0.026610 3.970428 0.0008 LFINDVT(-1) 0.137026∗∗∗ 0.063718 2.150504 0.0446 LGOVSPED(-1) 0.098455∗∗ 0.067300 1.462929 0.0698 LGRDINVST(-1) 0.144461 0.087000 1.645592 0.1163 LTRDOPE(-1) −0.510812∗∗∗ 0.070664 -7.228727 0.0000 Source: Researcher’s own calculations (2019) using Eviews 9.0; Note B: L= natural log of the series; ∆ = first difference of the series; GDP = Gross Domestic Product; EDEBT = External Debt; EDEBTSR = External Debt Servicing; EXC = Exchange rate; FINDVT = Financial Development; GOVSPD = Government Spending; GRINVT = Gross domestic investment; TRDOPE = Trade Openness; *** and * indicates significant at 1% and 10% respectively From Table 4.4, it is observed that in the long-run, Ghana’s external debt exerts a significant positive impacts on Ghana’s growth. Specifically, at 1% significant level, a 1% rise in Ghana’s external debt triggers 0.210250% rise in the Ghana’s growth. The result implies that more external debt will contribute to significant growth in the context of Ghana. The positive coefficient implies that the external debts are injected in productive economic activities and also there is effective debt management team. The findings follow the findings of empirical studies of 33 University of Ghana http://ugspace.ug.edu.gh Nwannebuike et al. (2016) and Abass and Christensen (2007). On the contrary, this findings diverge from the findings of the empirical studies of Hassan and Akhter (2012) and Quresi and Ali (2010). The results also reveal that Ghana’s external debt servicing exerts negative insignificant effect on Ghana’s economic growth. Also, the results show a significant impact of Ghana’s exchange rate on Ghana’s growth in the long run 1% significance level. This result follows the findings of Nwannebuike et al. (2016). That is, 1% rise in the exchange rate triggers 0.105652% rise in economic growth. The long-run results indicate that financial development exerts positive significant impact on Ghana’s economic growth at 5% level. Therefore, 1% increase in the financial development will trigger 0.137026% increase in the Ghana’s economic growth. Similarly, government spending was documented to exert positive insignificant impacts on Ghana’s growth. At 10% significant level, the results shows that by increasing government expenditure by 1% result in 0.098455% rise in the Ghana’s economic growth. This findings is consistent with the empirical findings of Amin and Audu (2006). However, a positive insignificant long-run nexus exists between gross domestic investment and growth. Long-run nexus between trade openness and growth was found to be negatively significant at 1%. That is, 1% rise in the trade openness of Ghana triggers 0.510812% decline in growth. The result follows the results of Eris and Ulasan (2013) and Babatunde (2011). 34 University of Ghana http://ugspace.ug.edu.gh 4.7 Short Run Estimation The short-run nexus among the variables employed for this study estimation results is reported in Table 4.5. From Table 4.5, the error correction term (ECM(-1)) which is employed to examine the speed of adjustment to long-run equilibrium and also how a shock in the exogenous variables is responded by the endogenous variable is also reported. The ECM(-1) coefficient is negative which is a confirmation to the co-integration found among the variables. Also, the ECM(-1) negative significant value implies that the model is stable at 1% significant level. Thus, the significant coefficient of the ECM(-1) implies that the external debt, the exchange rate, the financial development, government spending, the external debt servicing, gross investment and trade openness converge to the long-run equilibrium after the shock has occurred in the short- run. More significantly, the long-run equilibrium’s speed of adjustment is 77%. The short-run estimation results reveal that the lag of the growth has positive significant influence on the current growth at 1% significant level. However, the external debt have insignificant positive influence on growth. The finding is consistent with the long-run results of external debt. Also, the external debt servicing has negative insignificant influence on the growth. 35 University of Ghana http://ugspace.ug.edu.gh Table 4.5: Short Run Estimation Results Variable Coefficient Std. Error t-Statistic Prob. C 0.006010 0.007697 0.780785 0.4423 ∆ LGDP(-1) 0.885296∗∗∗ 0.153949 5.750576 0.0000 ∆ LEDEBT(-1) 0.022403 0.022414 0.999527 0.1498 ∆ LEDEBTSR(-1) −0.000889 0.011415 −0.077899 0.9385 ∆ LEXC(-1) 0.014723 0.022825 0.645047 0.5248 ∆ LFINDVT(-1) 0.047465∗∗ 0.019140 2.479877 0.0202 ∆ LGOVSPD(-1) 0.039836∗∗ 0.029980 1.328719 0.0459 ∆ LGRDINVT(-1) 0.011529 0.021402 0.538689 0.5949 ∆ LTRDOPE(-1) −0.103187∗∗∗ 0.031286 −3.298173 0.0029 ECM(-1) −0.776990∗∗∗ 0.215536 −3.604927 0.0014 Source: Researcher’s own calculations (2019) using Eviews 9.0; Note B: L= natural log of the series; ∆ = first difference of the series; GDP = Gross Domestic Product; EDEBT = External Debt; EDEBTSR = External Debt Servicing; EXC = Exchange rate; FINDVT = Financial Development; GOVSPD = Government Spending; GRINVT = Gross domestic investment; TRDOPE = Trade Openness; ***, ** and * indicates significant at 1%, 5% and 10% respectively. The results show that financial development exerts significant positive impacts on growth at 5% level. Thus, in the short-run, 1% rise in the financial development triggers approximately 0.047465% rise in growth. However, it is observed that the long-run result has larger impact than the short-run. Similarly, the government spending was documented to exert positive significant impact on growth. Thus, 1% rise in government spending triggers 0.039836% rise in the economic growth. The result is consistent with the long-run analysis results but the long run results shows larger impact. The short-run trade openness and economic growth nexus was documented to be negatively significant at 1% level. Specifically, a rise of 1% in the trade openness triggers 0.103187% decline in growth. The result also confirms the long-run results obtained. 36 University of Ghana http://ugspace.ug.edu.gh 4.8 Model Diagnostic and Stability Test From this model, a serial correlation diagnostic test and CUSUM Sum of squares model stability test of the residuals were conducted to ensure the robustness. Using Breusch-Godfrey serial correlation test, it was found that the residuals from the analysis are not serial correlated due to the fact that the p-value of the F-statistics and the Chi-square is above 0.05 alpha level (see Table 4.6). Table 4.6: Breusch-Godfrey Serial Correlation LM Test F-statistic 0.126584 Prob. F(1,24) 0.7251 Obs*R-squared 0.183633 Prob. Chi-Square(1) 0.6683 Source: Researcher’s own calculations (2019) by using Eviews 9.0. Also, Cumulative Sum of Squares of Recursive Residuals (CUSUM Squares) stability was also performed to examine model stability and the outcome is presented in Figure 4.2. it is observed from Figure 4.2 that the residual lies within the 5% critical value bounds which imply that the model employed for this study is stable. 37 University of Ghana http://ugspace.ug.edu.gh Figure 4.2: A Graph of Cumulative Sum of Squares of Recursive Residuals 1.4 1.2 1.0 0.8 0.6 0.4 0.2 0.0 -0.2 -0.4 92 94 96 98 00 02 04 06 08 10 12 14 16 CUSUM of Squares 5% Significance Source: Researcher’s own construct (2019) using Eviews 9.0. 38 University of Ghana http://ugspace.ug.edu.gh CHAPTER FIVE SUMMARY, CONCLUSION AND RECOMMENDATION 5.1 Introduction The last chapter of this study focuses on the summary of this study, conclusion and recommendations. 5.2 Summary The current study investigated the impacts of external debt on economic growth in the context of Ghana from 1980 to 2016. To achieve the purpose, two main objectives were formulated. The first objective focused on the trend analysis of Ghana’s external debt from 1980 to 2016. The second objective looked at examining the impacts of external debt on growth in both the short- run and long-run. To address the first objective, graphical analysis was employed. Similarly, the second objective was addressed by employing ARDL econometric technique due to the I(1) and I(0) nature of the variables used and also as it is reported in the literature to produce more robust results. The graphical trend analysis revealed that the external debt of Ghana increased steadily from 1980 to 2003. From 2003 to 2006, the external debt of Ghana decreased due to cancelation of Ghana’s debt as a result of HIPC. Also, from 2006 to 2016, the external debt of Ghana increased continuously. Specifically, from 1980 to 2016, the external debt of Ghana increased by approximately 1,430% change. 39 University of Ghana http://ugspace.ug.edu.gh The result from the ARDL estimations revealed significant external debt positive impacts on Ghana’s economic growth in the long-run. Specifically, in the long-run, a rise in the debt contracted by Ghana externally triggers a positive growth in the context of Ghana. That is, when external debts are obtained in the context of Ghana, growth increases due to injections of these funds into the economy’s productive sectors. However, in both the short-run and long-run, external debt servicing was documented to exert insignificant negative impacts on growth. Also, in the long-run, exchange rate was documented to exert significant positive influence on growth but insignificant positive effects in the short-run. In both the short-run and long-run, financial development and government spending were also observed to have positive impacts on growth of Ghana. In both the short-run and long-run, the gross domestic investment was documented to exert no significant impacts on growth of Ghana’s economy. In both the short-run and long-run, the trade openness was observed to exert negative significant impacts on growth of Ghana’s economy. 5.3 Conclusion The findings of the study indicate that Ghana’s external debt has a significant long-run impact on growth of Ghana’s economy. This implies that external debts contracted are injected into productive sectors of the Ghanaian economy. This implies, more external debt in the case of Ghana will trigger positive economic growth all other things being constant. However, the costs of these debts were documented to exert insignificant negative effects on the growth of Ghana’s economy. Also, trade openness, financial development, governments spending and exchange rate were observed to exert significant impact on growth. 40 University of Ghana http://ugspace.ug.edu.gh 5.4 Recommendation The findings of the study revealed a positive long-run nexus between Ghana’s debt obtained externally and Ghana’s growth. As a result, external debt can be considered as a catalyst in the economic growth of Ghana. Therefore, it is recommended to policy makers that the funds sourced externally should be invested in productive sectors of the Ghanaian economy in order to ensure these funds trigger economic growth. It is also recommended to policy makers to ensure effective and efficient prudent macroeconomic policies to control the macroeconomic variables including exchange rate, financial development, government spending and trade policies since they exert significant impact on the growth of the Ghanaian economy. It is also recommended to future researchers to study the nexus between external debt and growth in other LMICs to determine if other LMIC economies behave as this research has found for Ghana. Future research using different sample periods and different econometric tool is also recommended to determine if different econometric variables would lead to a different conclusion. 41 University of Ghana http://ugspace.ug.edu.gh REFERENCES Abbass, S., & Christensen, J. (2007). The role of domestic debt market in economic growth: An Empirical Investigation for low Income Countries and Emerging Markets. IMF, working Paper No. 07/127, Wishington, D.C. Adhikary, B. K. (2011). FDI, trade openness, capital formation and economic growth in Bangladesh: a linkage analysis. International Journal of Business and Management, 6(1), 16-28. Aghion, P., Boustan, L., Hoxby, C., & Vandenbussche, J. (2009). The Causal Impact of Education on Economic Growth: Evidence from U.S. Ajayi, L. B., & Oke, M. O. (2012). The effect of external debt on economic growth and development of Nigeria. International Journal of Business and Social Science, 3(12), 297-304. Ajibike, A. O. (2016). The effects of banks profitability on economic growth in Nigeria. Journal of Business and Management, 18(3), 1-9. Amin, A. A., & Awudu, I. (2006). External debt, Investment and Economic growth: Evidence from Nigeria. Economic and Financial Review, 44(1), 81-113. Ayadi, R., Arbak, E., Ben-naceur, S., & Groen, W. P. De. (2013). Financial development , bank efficiency and economic growth across the Mediterranean. MEDPRO Policy Paper, 7(30), 1–15. Azam, M., & Ahmed, A. M. (2015). Role of human capital and foreign direct investment in promoting economic growth. International Journal of Social Economics, 42(2), 98–111. Babatunde, A. (2011). Trade openess, infrastructure, FDI and economic growth in Sub-Saharan African countries. Journal of Management Policy and Practice, 12(7), 27-36. Bayraktar-Saǧlam, B., & Yetkiner, H. (2014). A Romerian contribution to the empirics of economic growth. Journal of Policy Modeling, 36(2), 257–272. Beck, T., & Levine, R. (2004). Stock markets, banks, and growth: Panel evidence. Journal of Banking & Finance, 28(3), 423-442. Brooks, C. (2008). Introductory econometrics for finance. Cambridge University Press. Caner, M., Grennes, T., & Koehler-Geib, F. (2010). Finding the Tipping Point when Sovereign Debt Turns Bad. World Bank Policy Research Working Paper Series, 5391. Chang, C., & Mendy, M. (2012). Economic growth and openness in Africa: what is the empirical relationship? Applied Economics Letters, 19(18), 1903-1907. 42 University of Ghana http://ugspace.ug.edu.gh Chowdhury, K. (1994). A structural analysis of external debt and economic growth: some evidence from selected countries in Asia and the Pacific. Applied Economics, 26(12), 1121-1131. Chudik, A., Mohaddes, K., Pesaran, M. H., & Raissi, M. (2013). Debt, Inflation and Growth Robust Estimation of Long-Run Effects in Dynamic Panel Data Models. CESifo Working Paper, 4508. Cunningham, R. T. (1993). The effects of debt burden on economic growth in heavily indebted developing nations. Journal of economic development, 18(1), 115-126. Egert, B. (2012). Public debt, economic growth and nonlinear effects: myth or reality? OECD Economic Department Working Paper, No. 993. Ejigayehu, D. A. (2013). The effect of external debt on Economic Growth: A panel data analysis on the relationship between external debt and economic growth. Unpublished work, p59. Eris, M. N., & Ulasan, B. (2013). Trade openness and economic growth: Bayesian model averaging estimate of cross country growth regressions. Economic Modelling, 33, 867- 883. Fosu, A. (1999). The impact of external debt on economic growth in sub-Saharan Africa, Journal of development Economics, 29(3), 253-268. Frimpong, J. M., & Oteng-Abayie, E. F. (2006). The impact of external debt on economic growth in Ghana: A Cointegration Analysis. Journal of Science and Technology (Ghana), 26(3), 122-131. Ganegodage, K. R., & Rambaldi, A. N. (2014). Economic consequences of war: Evidence from Sri Lanka. Journal of Asian Economics, 30, 42–53. Gomez-Puig, M. & Sosvilla-Rivero, S. (2015). Short-run and long-run effects of public debt on economic performance: Evidence from EMU countries. Working paper, 2015 1(37). Hassan, M. H. & Akhter, T. (2012). Impact of Public Debt Burden on Economic Growth: Evidence from Bangladesh. Journal of Finance and Banking, 10(1), 57-69 Hunt, S. D. (2007). Economic growth: should policy focus on investment or dynamic competition? European Business Review, 19(4), 274-291. Iyoha, M. A. (1999). External debt and economic growth in sub-Saharan African countries: An econometric study (No. RP_090). Jhingan, M. L. (2006). The economincs of development and planning, 38th Edition. Virnda Publication (P) Ltd, Delhi. 43 University of Ghana http://ugspace.ug.edu.gh Jones, E., & Manuelli, E. (1997). Endogenous growth theory : An introduction. Journal of Economic Dynamics and Control, 21(7), 1–22. Kopf, D. A. (2007). Endogenous growth theory applied : Strategies for university R & D. of Journal Business Research, 60(December 2006), 975–978. Korsi, F. (2015). The effect of external debt on economic growth in sub-Saharan Africa, University of Ghana, unpublished work, http://ugspace.ug.edu.gh Kumar, M., & Woo, J. (2010). Public debt and growth. IMF working papers, 1-47. Laurenceson, J. (2002). External Financial Liberalization and Foreign Debt in China (No. Discussion Paper No 304), p.53. Laeven, L., Levine, R., & Michalopoulos, S. (2015). Financial innovation and endogenous growth. Journal of Financial Intermediation, 24(1), 1–24. Legesse, T. (2012). Establishing financial markets in Ethiopia: the environmental foundation, challenges and opportunities. Journal of Business and Administrative Studies, 4(1), 1-44. Pattillo, C., Poirson, H. and Antonio Ricci, L. (2011). External Debt and Growth. Review of Economics and Institutions, 2(3), 55-67. Pattilo, C., Poirson, H., & Ricci, L. (2002). What are the Channels through which External Debt Affects Growth? IMF Working Paper, 04/15 Panizza, U., & Presbitero, A. F. (2013). Public Debt and Economic Growth in Advanced Economies: A survey. Pesaran, M. H., Shin, Y., & Smith, R. J. (2001), “Bounds testing approaches to the analysis of level relationships”, Journal of applied econometrics, 16(3), 289-326. Pegkas, P. (2018). The effect of government debt and other determinants on economic growth: the greek experience. Department of computer science, school of science, University of Thessaly, 35100 Lamia, Greece. Maku, O. E. (2009). Government spending and economic growth. Applied economics 26, 84-94. Masika, K. E. (2016). Empirical investigation of the impact of external debt on economic growth in the democratic republic of the Congo (DRC). Diss. 2016. McKinnon, R. I. (1964). Foreign exchange constraints in economic development and efficient aid allocation. The Economic Journal, 74(294), 388-409. 44 University of Ghana http://ugspace.ug.edu.gh Misztal, P. (2011). The relatioship between savings and economic growth in countries with different level of economic development. E-Finanse: Financial Internet Quartely,1, 17- 29. Mohd Dauda, S. N., Ahmad, A. H., & Azman-Saini, W. N. W. (2013). Does External Debt Contribute to Malaysia Economic Growth?. Economic Research-Ekonomska Istraživanja, 26(2), 51-68. Muhanji, S., & Ojah, K., (2011), Management and sustainability of external debt: A focus on the emerging economies of Africa. Review of Development Finance, 1(3), 184-206. Murdoch, J. C., & Sandler, T. (2002). Economic growth, civil wars, and spatial spillovers. The Journal of Conflict Resolution, 46(1), 91–110. Ngepah, N. (2017). A review of theories and evidence of inclusive growth : an economic perspective for Africa. Current Opinion in Environmental Sustainability, 24(3), 52–57. Nwannebuike, U.S., Ike, U. J., & Onuka, O. I. (2016). External debt and economic growth: the Nigeria experience. European Journal of Accounting Auditing and Finance Research, 4(2), 33-48. Parker, J. (2012). Theories of endogenous growth. Economic 314 Coursebook, 25. Qureshi, M. N., & Ali, K. (2010). Pubic Debt Burden and Economic Growth: Evidence from Pakistan. International Research Journal of Finance and Economics, 53(1), 100-108. Rao, B. B., & Rao, M. (2009). Openness and growth in Fiji: some time series evidence. Applied Economics, 41, 1653-1662. Reinhart, C. M., & Rogoff, K. S. (2010). Growth in a Time of Debt. National Bureau of Economic Research, Cambridge. Safdari, M.., Keramati, J. & Mahmoodi, M. (2011). The relationship between military expenditure and economic growth in four Asia countries. Chinese Business Review, 10(2). Saifuddin, M. (2016). Public Debt and Economic Growth: Evidence from Bangladesh. Global Journal of Management and Business Research: Economics and Commerce, 16 (5), 57- 69. Salih, M. A. R. (2012). The relationship between economic growth and Government expenditure: Evidence from Sudan. International Business Research, 5(8), 40. Schclarek, A. (2004). Debt and Economic Growth in Developing and Industrial Countries. Lund University Department of Economics Working Paper, 34. 45 University of Ghana http://ugspace.ug.edu.gh Solow, R.M. (1956). A Contribution to the Theory of Economic Growth. The Quarterly Journal of Economics, 70(1), 65-94. Todaro, M. P., & Smith, S. C. (2006). Economic development (9th ed.). Pearson Education Ltd., England. World Development Indicator (WDI), (2007). Retrieved from www.wdi.worldbank .org Zahonogo, P. (2017). Trade and economic growth in developing countries: evidence from Sub- Saharan Africa. Journal of African Trade, 3(2), 41-56. 46