The Measurement Of Capital Flight Economics Essay First of all, I would like to thank GOD for giving me the strength to complete this project. I would also like to express my sincere gratitude to my project supervisor Dr Jankee Keswar for his good guidance and help in the achievements of this project. I would like also to thanks Dr Nowbutsing Baboo for his advices. Secondly, special words of thank go to Nawshine, for her advice and support through all spheres of this study. I also wish to thank Yasseer and my friends/classmates from the BSc Economics with Management, who guided and supported me in one way or another throughout this project. Last but not least, I owe sincere and earnest thankfulness to my parents for their support. Dissertation Form Abstract List of Abbreviations ADF Augmented Dickey Fuller ARDL Autoregressive Distributed Lag BOP Balance of Payment BT Bound Testing CIF Cost, Insurance and Freight ECM Error Correction Model FDI Foreign Direct Investment FOB Freight on Board This essay is an example of a student’s work Disclaimer This essay has been submitted to us by a student in order to help you with your studies. This is not an example of the work written by our professional essay writers. Essay Writing Service Dissertation Writing Service Who wrote this essay Place an Order GDP Gross Domestic Product IMF International Monetary Fund MUR Mauritian Rupee OLS Ordinary Least Square US United States Chapter 1: Introduction 1.1 Introduction The financial crisis in the early 1980’s has raised considerable awareness on the large outflows of resident capital. This phenomenon of massive capital outflow has been termed as ‘capital flight’ and has been found to affect mostly developing countries. The size of capital flight in the developing countries in recent years has drawn the attention of many economists. There is no doubt that capital flight affects an economy negatively since it represents a loss of potential investment in a given country. The Sub Saharan African countries have not been spared from this plague. Recent studies have revealed that real capital flight from forty of these countries amounted to around $420 billion for the period 1970- 2004 (Ndikumana and Boyce 2008), which is a figure to worry about. 1.2 Statement of the problem Capital flight represents a serious threat to the sustainable growth of developing countries, especially in Africa. It is found that Africa is a net creditor to the rest of the world as the stock of capital outflows is far more than the external debt stock of Africa (Ndikumana and Boyce 2000). Therefore, it is of prime importance to know the size of capital flight in a country in order to take remedial actions. Although, various time series studies have been conducted on the determinants of capital flight for different countries, no econometric study has yet been undertaken for Mauritius. The only study of capital flight which included Mauritius was done by Ndikumana and Boyce (2002) whereby they have investigated the causes of capital flight for 30 Sub Saharan African countries for the year 1976 to 1996. The endeavor, however, was not based specifically on Mauritius. As such, to fill this shortcoming, this study is being undertaken for period 1980 to 2010. There are several definitions and thus different methods of estimating capital flight, which have been lengthily explained in the literature with their respective pros and cons. The Residual method, which is the most common one, is used in this study to compute the estimated capital flight from Mauritius. 1.3 Aims and objectives The contribution of this study is to: estimates the size of capital flight from Mauritius for the period 1980 to 2010, using the residual method. analyzes the determinants of capital flight and analyzes the impact of some of these determinants, in terms of their contribution to capital flight, using an autoregressive distributed lag (ARDL) approach based time series data for Mauritius for the period 1985-2010. 1.4 Organization of the study This paper will further down be arranged into five parts as follows:- Chapter 2- focuses on the literature reviews both on the theoretical and empirical parts. Chapter 3- proceeds with the estimation of capital flight from Mauritius. Chapter 4- explains the methodology to be used to find the impact of the determinants on capital flight. Chapter 5- addresses the discussion on the empirical results and the interpretation of the research findings. Chapter 6- concentrates on the conclusion and recommendations. Chapter 2: Literature Review 2.1 Introduction Since the debt crisis in the early 1980’s, there has been copious literature devoted on the outflows of resident’s capital in response to changes in domestic policies and political instability. This phenomenon of huge capital outflows was termed as “capital flight”. With time, capital flight became to be regarded as an indicator of a country’s economic situation. The aim of this chapter is to have an overview of relevant literature on capital flight. The chapter is divided into two main parts, as follows:- Theoretical aspect Empirical aspect 2.2 Theoretical Aspect This section looks at the various definitions of capital flight, its determinants and the possible methods for measuring its magnitude. 2.2.1 Definition of Capital Flight It is worth pointing out that there is ample controversy on the definitions of capital flight. While some analysts view capital flight as a normal capital outflow, others consider it as an abnormal capital outflow. In general, capital outflow from developed countries is regarded as foreign investment while the same activity is referred to as capital flight when undertaken by developing countries (Ajayi 1995). One reason according to Ajayi (1995) behind this dichotomy is the belief that investors from developed economies are seen as responding to opportunities abroad while investors from developing economies are said to be escaping the high risk perceived at home. It cannot be denied that usually investors, whether from developed or developing economies, base their decisions on the relative risk and return at home and abroad. Therefore, investors around the world are rational and will thus search for better risk-return trade-off and portfolio diversification. This essay is an example of a student’s work Disclaimer This essay has been submitted to us by a student in order to help you with your studies. This is not an example of the work written by our professional essay writers. Essay Writing Service Dissertation Writing Service Who wrote this essay Place an Order Capital flight is referred as “the reported and unreported acquisition of foreign assets by the non-bank private sector and elements of the public sector” by Morgan Guaranty Trust Company (1986 cited Ajayi 1995). Nowadays, capital flight can be viewed as an illegal transaction in situations where traders falsify their trade documents in order to keep their capital abroad (Schneider 2001). By doing so, they generate capital flight through export underinvoicing and import overinvoicing. For others, capital flight is the non reporting of income earned from claims on non residents, in the balance of payment system in order to escape the control of the home government (Dooley and kletzer 1994). Short term private capital outflow is also regarded as capital flight. The latter absorbs hot money that responds to political and financial crisis, heavier taxes, potential tightening of capital control, considerable devaluation of domestic currency and actual or forthcoming hyperinflation (Cuddington 1986 cited Machochekanwa 2007). This interpretation is also found in Schneider (2003) who argues that capital flight is the flow of resident’s capital from domestic country to another due to political and economic risk. It can be seen from the above that there are numerous definitions of capital flight. Nowadays, with the increased globalization, the financial markets are more conducive to the movement of capital thus causing the phenomenon of “capital flight”. Nevertheless, it is generally accepted that capital flight represents the outflow of capital from domestic financial market in order to evade losses. 2.2.2 Measurement of capital flight Numerous definitions of capital flight as seen in the previous section, give rise to different methods to measure this phenomenon in respect of data availability and the methodology used by different countries. The following measures are found in the literature: the residual method; the Dooley method; the hot money method; the trade misinvoicing method and the asset method (Claessens and Naudé 1993). Residual Method This is the principal measure of capital flight proposed by the World Bank (1985). The residual measure, also known as the broad measure is an indirect approach to estimate capital flight. Outflow of capital is equal to the differences between sources of funds (that is, net increases in external debt and net inflow of foreign investment) and uses of these funds (that is, the current account deficit and additions to foreign reserves). Algebraically, capital flight is expressed as follows: CFr = ΔED + FDI – CAD – ΔFR Where: CFr stands for capital flight, ΔED is the change in the stock of gross external debt, FDI is the net foreign investment inflows, CAD is the current account deficit and ΔFR represents the change in the stock of official foreign reserves. The above approach is changed slightly by Morgan Guaranty Trust (1986 cited Claessens and Naudé 1993). The change in short term foreign assets of the domestic banking system (ΔB) is taken into consideration. This additional item is deducted from the residual method, showing that the banking system has nothing to do with capital flight as shown below: CFr = ΔED + FDI – CAD – ΔFR – ΔB Dooley Method The Dooley method seeks to separate the legal and illegal capital flows. Hence, under the Dooley method, capital flight is equal to that amount of income from foreign assets which are not reported to home country. Following Hermes et al (2002), according to the Dooley method capital flight is calculated as shown below: TKO = FB + FDI – CAD – ΔFR – EO – ΔWBIMF Where: TKO is the total capital outflows, FB is the foreign borrowing as reported in the BOP statistics, EO is net errors and omission and ΔWBIMF shows the difference between the change in the stock of external debt reported by the World Bank and foreign borrowing reported in the BOP statistics published by the IMF. The stock of external assets related to reported interest earnings is: ES = INTEAR/ rus Where: ES is the external assets, rus is the US deposit rate (assumed to be a representative of international market interest rate) and INTEAR shows the reported interest earnings. Thus, capital flight according to this method is measured as: CFr = TKO – ΔES Hot money method The hot money method is also referred as the narrow measure of capital flight. According to this method, capital flight represents the short term movement of capital of the non bank public sector plus the errors and omission from the BOP (Cuddington 1986 cited Makochekanwa 2007). One criticism about this method found in the literature is that the hot money method focuses only on the short term outflows of capital. Thus, capital flight is often underestimated. It is calculated as: This essay is an example of a student’s work Disclaimer This essay has been submitted to us by a student in order to help you with your studies. This is not an example of the work written by our professional essay writers. Essay Writing Service Dissertation Writing Service Who wrote this essay Place an Order CFr = SKO + EO Where: SKO is short term capital outflow of the private sector and EO is the errors and omission. Trade Misinvoicing Method Capital can move from one country to another illegally through trade. This can be calculated by taking data from both the importing and exporting countries. Therefore, capital flight arises when there are export underinvoicing and import overinvoicing. This can be illustrated as follows: Export underinvoicing = (Mw/CIFFOB) -Xc Import overinvoicing = (Mc/CIFFOB) – Xw Where, Mw: World’s import from that country Xc: Country’s export to the world Mc: Country’s import from the world Xw: World’s export to that country It is important to note that the import reported by the country and the import as reported by the world should be on a comparable basis. So, they need to be adjusted by a country specific CIF/FOB ratio. A positive sign indicates capital flight while a negative sign shows capital inflow. The net effect of misinvoicing is the capital flight. Asset method The stock of assets which are held in foreign banks by nonresident is referred to as capital flight by some authors. This is the so-called asset method (Hermes and Lensink 1992). However, given that there exist several forms in which assets can be held, this method measures only part of the capital outflows. It can be seen from the above that there are different methods to estimate capital flight. Different authors have in fact tried to measure capital flight by adding new variables to existing equation and compare the new result obtained to the previous one. 2.2.3 Determinants of capital flight Normally, individuals are regarded in most cases as being risk-averse, that is, they prefer a sure investment income. They try to avoid risks and losses as far as possible by diversifying their wealth in order to maximize their asset returns. Thus, there is a direct relationship between the behavior of a risk-averse individual and capital flight. It can be found that decision on whether to move or hold capital abroad is based on the amount of wealth, the relative risk and uncertainty and the relative rates and returns of asset (Hermes et al 2002). Capital flight is generated by either economic factors or non economic factors. Economic factors for instance consist of exchange rate, inflation, foreign borrowing, fiscal deficit, foreign direct investment (FDI) and capital outflows whereas non-economic factors can be corruption and political or social instability. These factors are discussed below:- Economic Growth An indicator of the macroeconomic environment which determines capital flight is the growth rate of the economy. Economic growth usually triggers the presence of investment opportunities in the local economy. Thus, low economic growth or recession may indicates low return on domestic assets and therefore encourage capital to flow out of the home country. Exchange Rate It is often found that one of the determinants of capital flight is exchange rate overvaluation. An overvalued exchanged rate means that economic agents would predict depreciation in the near future. Incipient depreciation would make foreign goods to appear more expensive than domestic one. Thus, in order to avoid future losses, residents will opt to hold their assets abroad which generate capital flight (Ajayi 1995). Inflation High level of inflation can also trigger capital outflows. Firstly, high inflation implies that the value of domestic assets will be eroded which provide incentive for residents to hold their assets abroad. Secondly, high inflation is closely attached to exchange rate such that it increases the expectation of future depreciation (Hermes et al 2002). External debt Several studies have shown that there is a positive relationship between external debt and capital flight. External borrowing usually reflects that an economy is not doing well or the investment climate of the country is not favorable which explain capital outflows. It is shown by Beja (2006) that there is a direct linkage between external debt and capital flight. External borrowings are often used to finance capital outflows. In some cases, the transactions are only made through the financial institutions and may not even enter the home country. However, it can also be that these capital flights stimulate more external borrowings. More often than not, government passes the burden of the external debt on the public through the imposition of high tax. In turn, the domestic residents try to avoid such taxes by placing their capital outside the country which leads to capital flight (Gulati 1988). Moreover, it is observed by Eaton (1987) that government based-guaranteed debt stimulate further capital flight. Assuming that foreign borrowings are used to finance capital outflow, nationalization of private debt implies that if one borrower fails to repay its debt, government will raise the tax burden on other borrowers. Consequently, other borrowers too will be motivated to flee the tax obligation by investing abroad. This essay is an example of a student’s work Disclaimer This essay has been submitted to us by a student in order to help you with your studies. This is not an example of the work written by our professional essay writers. Essay Writing Service Dissertation Writing Service Who wrote this essay Place an Order Fiscal deficit It can be found that in period of fiscal crisis, government usually seeks the help of foreign countries through foreign borrowing. Likewise, in this situation insolvency and default risk are likely to increase which induce capital flight (Ize and Ortiz 1987). Furthermore, if fiscal deficit is financed through the injection of more money in the economy, this can generate inflationary pressure which in turn erodes the monetary balances. As a consequence, residents will shift from domestic to foreign assets in order to avoid the inflationary tax. On the other hand, if fiscal deficit is financed through bonds sales, the effect will be the same as printing money. However, it will be in the form of higher tax liabilities (Ajayi 1995). Foreign Direct Investment (FDI) There exists an ambiguous relationship between the outflow of capital and FDI. From the investment climate perspective, capital flight is subject to the rate of return differential of assets between the home and foreign countries after adjusting for exchange rate. For instance, if the investment climate is favorable, i.e. the rate of return is higher at home as opposed to the foreign country, FDI increases and capital flight is expected to reduce, thus implying a negative relationship (Pastor 1989). Alternatively, the discriminatory treatment perspective states that usually host countries provide incentives to nonresident investors such as differential taxation, investment or exchange rate guarantees which are not imparted to the domestic residents. Hence, if FDI is the outcome of those preferential treatments given to non residents, this will generate capital flight. Therefore, from the discriminatory treatment point of view, a positive relationship between outflow of capital and FDI can be established (Dooley 1988). Political instability The influence of weak political bodies on economic institutions can cause distortions and instabilities. This can be a reason for capital flight. The high perceived risks and uncertainty in the public sector causes residents to lose confidence in domestic economy as they may expect an erosion of their future assets. As a result, domestic residents may choose to hold their wealth outside the country (Hermes et al 2002, p.9). Some of the factors which can lead to instability are financial repression (artificially low interest rate), threat of expropriation, perceived policy reversal and default of government obligation. It is shown by Ndikumana and Boyce (2008) that high level of corruption reflecting a ‘sick’ economy can encourage capital flight. As stated earlier in this paper, one of the definitions of capital flight includes the view that it is an illegal transaction. Therefore, corruption facilitates the acquisition of those illegal assets. Generally, representatives of the government are implicated in such transactions thereby inducing private agents to do so. Capital outflow Capital flight can by itself be a determinant of further capital flight as pointed out in the literature. According to Ndikumana and Boyce (2002), countries with high capital outflow are expected to have a high level of capital flight in the future. In most cases, capital flight is associated with a worsening macroeconomic environment where investment is not favorable which in turn leads to further capital outflows. In addition, the presence of capital flight forces government to increase the tax liabilities on domestic residents. As a result, the low returns derived after the tax adjustment discourages private agents to invest and motivate them to look for higher return abroad (Collier at al 2001). 2.3 Empirical Aspect Over the years, there have been extensive researches which have tried to shed light on factors affecting capital flight. This section therefore points out the empirical results of some studies carried out on the determinants of capital flight in different countries. 2.3.1 Determinants of capital flight Although there had been massive determinants of capital flight, mainly because of differences in its measurement and differences in econometric techniques and specifications, some factors have become more obvious as its determinants. The key results from a selection of 17 studies on developing nations have been reviewed as shown in Table-A1. Further empirical evidences on the most common determinants of capital flight, as per the table, are highlighted. Macroeconomic Stability In general, an inverse relationship is expected between capital flight and the growth rates of an economy. This is because when an economy is doing well, the investment opportunities are expected to be positive. An endeavor was carried out by Alam and Quazi (2003), who used the ARDL procedure to investigate on the above relationship for the period 1973-99 in Bangladesh. This essay is an example of a student’s work Disclaimer This essay has been submitted to us by a student in order to help you with your studies. This is not an example of the work written by our professional essay writers. Essay Writing Service Dissertation Writing Service Who wrote this essay Place an Order The study showed that capital flight in Bangladesh was caused partly by the lower real GDP growth rate. However, in another study by Hermes et al (1996), it was found that as opposed to the above, economic growth (lagged) in Cote d’Ivoire, Uganda and Nigeria has a positive effect on capital flight. This positive relationship has been attributed to the fact that economic growth was not sufficiently significant to retain capital in those countries. Exchange Rate The degree of overvaluation of the Latin America currency was positively related to capital flight. Taking the case of Mexico, Pastor (1989) found that by decreasing the real exchange rate index by 1 % capital flight fell by $68 millions in 1985. One reason why capital fled Mexico is because the Mexican feared that their assets would be eroded in the future due to devaluation of their currency. In the same perspective, in another study, using the ordinary least squared (OLS) technique, Ayadi (2008) examined the determinants of capital flight in Nigeria. The result suggests that exchange rate devaluation significantly explains capital flight in the short run as well as in the long run. Inflation An investigation was carried out by Victor (2007) on a panel data of 77 countries from 1971 to 2000 in order to test whether post-war inflation increases annual capital flight. Using different methods of measuring capital flight and many econometric estimation techniques, evidence suggest that 1% increase in inflation is accompanied by 0.005% to 0.01 % of GDP increase in capital flight. In order to examine the impact of inflation on capital flight, Dooley (1988) used a pooled regression on five Latin American countries and Philippines for the period 1976-83. And the result is consistent with theoretical expectations, that is, a high inflation rate generates capital flight. It was found that a 1 % increase in inflation leads to 23.10 % increase in capital outflow. It is worth pointing out here that inflation is referred to as ‘inflationary tax’ as the authorities has resorted in the creation of money. External Debt A study by Ndikumana and Boyce (2008) shows that external debts encourage capital flight. Firstly, OLS estimation was used followed by adding country specific effects and lastly by taking into account change in debt as endogenous. The results show a statistically and economically positive relationship between external borrowing and capital flight. It was found that between 1970 and 2004, 62 cents out of each dollar borrowed abroad has left sub-Saharan Africa in the form of capital flight. Earlier empirical evidence by Ndikumana and Boyce (2002) approved the above statement. An econometric analysis was carried from 30 sub-Saharan Africa countries for the period 1970 to 1996. The result showed that 80 cents out of each dollar of external borrowing escape the countries in terms of capital outflow in the same year. The relationship between capital flight and external debt can be further supported by Henry (1996), who sought to find the causes of Jamaican capital flight. According to the result, Jamaican capital outflow was generated both by external debt and the economic circumstances due to these external debts. Evidences also showed that for every one dollar of external borrowing, 66 cents escape the country in term of capital flight in the same year and a further 33 cents flow out after two years. Foreign Direct Investment (FDI) A study was carried out by Kant (1996) to analyze the relationship that exists between capital flight and FDI. The latter employed the contemporaneous-correlation and principle component analysis to estimate the above relationship. Results suggest that capital flight computed under the Cline, Dooley and Hot Money methods are negatively related to FDI. It was deduced that capital flight was not caused by preferential treatment but instead by the general economic mismanagement and inefficiencies. Political and Social Instability A research was done by Hermes et al (2000) where they made use of dummy variables to evaluate problems attached to political instability such as the number of assassination and revolutions per year and indexes of political rights and civil liberties. It can be concluded from the empirical analysis- using the robustness test put forward by Sala-i-Martin (1997 cited in Hermes e al 2000) – that political instability or risk do engender capital flight. In South Korea during the mid 1980s and early 1990s, the country had undergone political turmoil and during that period capital fled the country considerably. From 1976 to 1991, the average capital flight was around 1.92 % of GDP yearly. However, in 1987 when political unrest was at its peak, it was noticed that capital flight reached 6.6% of GDP (Le and Zak 2001). This essay is an example of a student’s work Disclaimer This essay has been submitted to us by a student in order to help you with your studies. This is not an example of the work written by our professional essay writers. Essay Writing Service Dissertation Writing Service Who wrote this essay Place an Order Similarly to South Korea, factors affecting Argentina’s capital outflow was associated to political instability. During these days no particular attention was given to the economic situation of the country and the latter exacerbate such that there was massive capital outflow. Results suggest that between 1976 and 1991, Argentina’s capital flight averaged 3.4 % of GDP yearly but peaking at 10% of GDP in 1989 when severity program was put in place which caused political instability (Le and Zak 2001). Moreover, a study was undertaken in Bangladesh on the determinants of capital flight over the 1973-99 periods, employing the ARDL approach. The result suggests that political disturbances had a positive and significant impact on capital flight both in the short run and long run (Alam and Quazi 2003). Besides, social instability and capital flight are seen to be positively linked. Social unrest motivates residents to hold their capital abroad as the latter lose confidence in the domestic economy. The fact that social instability is difficult to measure; the level of unemployment is thus used as a proxy. The level of unemployment was found to be significant in explaining capital flight in Trinidad and Tobago for the period 1971-87 (Henry 1996). All the above determinants have, indeed, either a positive or negative effect on capital flight. It has been noted, however, that the impact might not be the same in all countries due to difference in political, social and macroeconomic environment. This study will therefore analyze the effect of some of the determinants stated above on capital flight in the Mauritian economy. Chapter 3: Magnitude of Capital Flight 3.1 Introduction This chapter demonstrates the magnitude, trend and burden of capital flight from year 1980 to 2010 in Mauritius employing the residual method. However, capital flight is a phenomenon which affects the whole world; the only difference lies in its magnitude. 3.2 Overview of capital flight in four regions of the world As shown in Figure 3.1, the highest magnitude of capital flight has been recorded in East Asia while for the Sub Saharan countries, it is noted that the capital flight is lower with a more stable trend. The above can be attributed to a deep financial and economic crisis in certain Asian countries including Russia during year 1997-98, which fueled capital flight. The crisis brought about devaluation in exchange rates and a decline in share prices. These situations were, at the beginning, noted in Thailand and Indonesia further to which investors, as a result of loss of confidence, have withdrawn their investments. During the same period, it can be noted that the Sub-Saharan countries experienced relatively low capital flight. It can be deduced that they were not affected by the Asian financial crisis. Given that Mauritius forms part of the Sub-Saharan African countries, therefore it is evident that capital flight is indeed present in the island’s economy. Figure 3. Magnitude of Capital flight in Four Regions of the World Source: Author’s computation (using Table A-2) Moreover, it can be drawn from figure 3.2 that the burden of capital flight in the Sub Saharan African countries was high relative to the other regions, especially during 1985-87 and 1990. Though, it was noticed that in absolute terms, the Sub Saharan African countries experienced lower capital flight. The highest burden of capital flight over the period considered occurred in 1988 in East Asia where capital flight to GDP attained 12%. It can be observed from figure 3.1 and 3.2 that capital flight in South Asia was lower when compared to the East Asia not only in absolute terms but also in relative terms (i.e. in terms of GDP). Furthermore, the burden of capital flight in Latin America was lower to East Asia from the mid 1985 to 1998. Figure 3. Capital Flight Burden in Four Regions of the World Source: Authors computation (using Table A-2) 3.3 Capital flight in Mauritius Over the years, Mauritius has shifted from a monocrop economy to a more diversified one driven by export oriented manufacturing, tourism, financial and business services sectors. According to the World Bank’s Doing 2009 Business report, Mauritius has been positioned 24th in the world and 1st in Africa in the ease of doing business. Since the late 1980’s, the Mauritian financial system has witnessed various changes as part of the economic liberalization strategy. Some of the major financial reform which took place were; the liberalization of the interest rates in the mid 1988, issue of Bank of Mauritius Bills and auction of treasury bills in 1991, abolition of ceilings on credit to priority sectors in the mid 1992, liberalization of the exchange control in the mid 1994 and the abolition of the credit-deposit ratio as well as the imposition of 15% limit on the overall foreign exchange exposure in the mid 1996 among others. This essay is an example of a student’s work Disclaimer This essay has been submitted to us by a student in order to help you with your studies. This is not an example of the work written by our professional essay writers. Essay Writing Service Dissertation Writing Service Who wrote this essay Place an Order This study employs the residual method to calculate capital flight in Mauritius for the year 1980 to 2010. The reason for adopting the residual method is that it takes into account both the reported and the unreported foreign assets for the public as well as the private sectors. Moreover, given the various drawbacks associated with the other methods used to estimate capital flight, the residual method is the most appropriate. Capital flight is obtained using the equation that follows whereby CFr stands for capital flight, ΔED is the change in the stock of gross external debt, FDI is the net foreign investment inflows, CAD is the current account deficit and ΔFR represents the change in the stock of official foreign reserves. This technique is based on the differences between sources of funds and the uses of these funds. CFr = ΔED + FDI – (CAD + ΔFR) The following analysis pursues the rule in the above literature whereby capital flight is denoted with a positive notation. In other words, positive value of CFr means capital flight while negative value shows the reverse of capital flight. 3.3.1 Data Used Table 3.1 below consists of secondary data used for this study to compute capital flight. The observed variables are assigned an acronym as shown in the second column which is used throughout the study. Moreover, the third column illustrates the term of currency in which the variables are used. Table 3.0. Description of Variables Used in Estimating Capital Flight Description Abbreviations Units Change in External Debt ΔED Current US dollars Foreign Direct Investment FDI Current US dollars Current Account Deficit CAD Current US dollars Change in Foreign Reserves ΔFR Current US dollars All the data series, except for data on changes in net reserves, have been obtained from the World Bank. The data on changes in net reserves has been obtained from International Monetary Fund’s Balance of Payments Statistics Yearbook. It is to be noted that the figure used for the change in external debt was taken from the World Bank’s Global Development Finance. The following trend has been computed using data from Table-A3. Figure .3 Magnitude of Capital Flight in Mauritius Source: Author’s computation It can be noticed that the magnitude of capital flight is surely not following a straight line trend. There are fluctuations over the years with capital flight attaining its highest level in year 2002-03 and dropping to its lowest level in year 2006-07. The events in the Mauritian economy which could explain, to some extent, the cause of the high capital flight in year 2002-03 are a low GDP rate of 2.1% accompanied with a relatively high inflation rate of 6.5% as well as a budget deficit for that year. From the figures, it can be deduced that the economy was in a bad shape thereby causing investors to lose confidence and consequently moving their money out of the country. Moreover, the decline in the capital flight noted in the mid 1980’s to 1990 can be attributed to the intervention of the Bank of Mauritius to smoothen the currency fluctuations which were being experienced by the economy in some precedent years. Figure 3.4 shows the trend of capital flight burden in Mauritius, measured as a share of GDP over the period 1980-2010 which relates outflow of capital to the size of the economy. Figure 3. Capital Flight Burden in Mauritius The capital flight burden follows a similar trend as in figure 3.3. However, comparing the highest magnitude of capital flight and the highest burden of capital flight, it can be noted that they did not occur in the same year. The highest magnitude of capital flight was recorded in 2002 while the highest burden of capital flight in Mauritius was felt in 1987 amounting to 23.95%. One possible reason for the highest peak in the burden was that capital flight was quite high and GDP was relatively low in that year. 3.4 Conclusion It is clear that, indeed, capital flight is present in the Mauritian economy. From the trend displayed above, it can be concluded that economic indicators are the drivers of capital flight, each having its own effect. The next two chapters shall shed more light on the variables which are affecting capital flight in Mauritius. This essay is an example of a student’s work Disclaimer This essay has been submitted to us by a student in order to help you with your studies. This is not an example of the work written by our professional essay writers. Essay Writing Service Dissertation Writing Service Who wrote this essay Place an Order Chapter 4: Methodology 4.1 Introduction Notwithstanding differences in the methods used to calculate capital flight, empirically, there is a common agreement on the determinants of capital flight. Therefore, this chapter describes the econometric technique that will be used to analyze the impact of some of the determinants on capital flight. 4.2 Data Sources Typically, all the data used are of secondary nature. All the data, except for unemployment rate and capital flight, were obtained from the World Bank. The unemployment rate was extracted from the IMF’s World Economic Outlook while capital flight is computed. Table 4.1 below summarizes the expected signs of the regression coefficients based on the model specification provided below. 4.3 Model Specification Along the lines of the above discussion regarding the numerous causes of capital flight, this study seeks to find the impact of the following determinants on capital flight burden which could be modeled as follows: = (1) Where, CF stands for capital flight as a percentage of GDP; GW is the annual percentage growth rate ;ED is external debt as a percentage of GDP; FDI represents foreign direct investment as a percentage of GDP;UN is the unemployment rate which is used as a proxy of social instability; ER is the exchange rate of local currency per US dollar; D is the dummy variable for the liberalization of the exchange control; is the constant term, (i=1,…6) are the regression coefficient, t denoted the time period and ε is the error term. 4.3.1 Definition of variables Capital Flight as percentage of GDP To account for the dependent variable that is capital flight, it has instead been used as a percentage of GDP which is referred to as the burden of capital flight. The figures for capital flight were computed from chapter 3 using the residual method. Annual Growth rate The annual percentage growth rate of GDP at market prices is used to show the health of the economy. Aggregate are based on constant 2000 U.S. dollars. Mauritius has experienced a relatively stable growth rate, at an average of 5% from 1983 to 2010. The relationship between capital flight and the GDP growth rate is expected to be negative. Conversely, study carried out by Hermes et al (1996) suggests a positive relationship between outflow of capital and the growth rate. External debt as a percentage of GDP External debt is the amount of money borrowed from nonresidents and which is repayable in foreign currency, goods or services. It comprises of IMF credit, publicly guaranteed and private nonguaranteed long term debt as well as the short term debt. The data are in current US dollar. External debts can channel easy access to foreign exchange rate which in turn facilitate capital flight. Evidence showed that, indeed, external debt fuelled capital flight in Jamaica for the period 1971-87 (Henry 1996). Foreign Direct Investment as a percentage of GDP Foreign Direct Investment (FDI) represents the net inflows of investment. It is calculated by summing up equity capital, reinvestment of earnings and long term as well as short term capital in the balance of payments. Mauritius is regarded as an island having a favorable investment climate, sound political and social position. Recently, it has been noted that the Mauritian financial sector is used as a platform and the country is seen as a gateway to invest in Africa. The relationship between capital flight and FDI is quite complex as seen in chapter 2. Therefore, this study shall shed more light on the effect of FDI on the outflow of capital. Exchange rate The exchange rate used is the annual official exchange rate that is determined by the national authorities or in the legal sanctioned exchange market. It is based on monthly averages of the Mauritian rupees relative to the US dollar. Expected devaluation trigger capital flight and this has been encountered in countries like Jamaica, Trinidad and Tobago (Henry 1996). Unemployment rate The unemployment rate as a percentage of the labour force is used as a proxy for social instability. In Mauritius, the unemployment rate consists of those people who are willing to work but are unable to find one. It does not include children, retired and people who are not searching for a job. The relationship between the unemployment rate and capital flight is expected to be positive since social instability can motivate residents to take out their money out of the economy, which was used by Henry (1996) Dummy for liberalization of exchange control A dummy for liberalization of exchange control has been used. Mauritius has liberalized its exchange control in the mid 1994’s and has put in place an interbank foreign exchange market. It can be noted that Mauritius has maintained a managed floating exchange rate system. It is expected that the liberalization of the exchange control shall lead to more capital flight. Note that the Dummy variable, D, is used for the liberalization of exchange control which assumes values as follows: This essay is an example of a student’s work Disclaimer This essay has been submitted to us by a student in order to help you with your studies. This is not an example of the work written by our professional essay writers. Essay Writing Service Dissertation Writing Service Who wrote this essay Place an Order D=0 (Pre-reform period 1980-1994) D=1 (Post-reform period 1995-2010) Table 4.0. Description of the Independent Variables Used in the Regression Description Abbreviations Units Expected Signs Growth rate GW % per annum Positive/Negative External debt ED % of GDP Positive Foreign Direct Investment FDI % of GDP Positive/Negative Exchange rate ER Mauritian currency per US dollar Positive Unemployment rate UN % per annum Positive 4.4 Tests for Stationarity It is likely that macroeconomics time series are either non stationary or have a unit roots. Hence, there is a need to check the order of integration of the above variables. A given series is said to be integrated of order d, denoted as I (d), if it must be in differenced d times for it to become stationary. So, if a variable is found to be integrated of order zero, that is, the variable is I (0), it is said to be stationary in its level forms. On the other hand, if it is integrated of order one, that is, I (1), it is assumed to be stationary after its first difference. By definition a series is stationary if it has a constant mean and a constant finite variance. On the contrary, a non-stationary series contains a clear time trend and has a variance that is not constant overtime. Although, there are many methods that can be used to check for stationarity and the presence of unit roots, the Augmented Dickey Fuller (ADF) test is adopted to check for stationarity of the variables. 4.5 Exploring the ARDL Model The Bounds testing (BT) approach was recently developed by Pesaran et al (2001) which is based on the ARDL model. According to Pesaran et al (2001), the ARDL technique has many advantages. Firstly, the cointegration relationship can be estimated using the OLS, once the lag order of the model is known. Secondly, it is valid irrespective of the order of integration of variables considered. Lastly, it is appropriate for small sample size. Since this study is analyzing a small sample size of 26 observations of annual capital flight burden from Mauritius, thus the bound test is the most relevant testing procedure. 4.5.1 The Long Run Relationship The bound test approach is divided into two stages. The first stage is about testing the existence of a long run relationship in the levels among the variables. In this stage the ARDL-ECM given in equation (1) can be estimated using the OLS in order to test for the existence of long run relationship among variables. = + + + + + + ++ + + + + (2) Where, is a drift component and is the white noise error. It is to be noted that are the long run multipliers while,,,,, and represent the short run impacts on capital flight. There is a need to test the presence of any level relationship through the exclusion of all lagged level variables equal to zero. Thus, the null and the alternative hypotheses are as follows: H0: = 0 H1: The F-test with an asymptotic non standard distribution can be applied to test the above hypotheses (Pesaran et al 2001). There are two asymptotic critical value bounds- the lower bound when all regressors are I(0) and the upper bound when all regressor are I(1)- which specify the existence of a long run relationship among variables. If the F-test statistic lies above the upper critical values, the null hypothesis is rejected, indicating there is a long run relationship among the variables irrespective of the order of integration. On the other hand, the null hypothesis is not rejected if the F-test statistic lies below the lower critical values, implying that there is no evidence of long run relationship. If the statistics lies within the lower and upper critical values, inference would be inconclusive. 4.5.2 The Short Run Dynamic Once the existence of a long run relationship of the variables is known, the second stage is performed whereby the parameters of the long run relationship are estimated including the short run ECM. This method has the advantage of generating consistent estimates of the long run coefficients that are asymptotically normal, irrespective of whether the underlying variables are I(0) and I(1) (Pesaran 1997). The orders of the lags in the model are selected by the Schwarz Bayesian Criterion (SBC). A lagged error correction term is constructed, denoted by ECMt-1, out of the long run coefficient to substitute all lagged level variables. Hence, it is possible to obtain the short run coefficients by estimating an error correction model which is associated with the long run estimates as shown below: This essay is an example of a student’s work Disclaimer This essay has been submitted to us by a student in order to help you with your studies. This is not an example of the work written by our professional essay writers. Essay Writing Service Dissertation Writing Service Who wrote this essay Place an Order = + (3) Where, represent the error correction term and its coefficient,, shows the speed of adjustment while the coefficient and are the short run dynamics that cause the model to converge to equilibrium. 4.6 Stability of the Relationship Besides, this study also conducts the Cumulative sum (of square( CUSUMSQ) test proposed by Brown, Durbin and Evans (1975 cited Deng and Perron 2005) which can be used to check the goodness of fit of the ARDL model. In other words, to infer whether the long run and short run relationships are stable or not. The CUSUMSQ compares the cumulative sum of squares of the residual over time within a set of confidence intervals. If the sum of residuals at a given point lies outside the given intervals, then it can be deduced that the model is unstable. The next chapter focuses on the results based on the different stages of the bound testing procedure. The results will serve to analyze the impact of the considered determinants on capital flight burden in Mauritius. Chapter 5: Findings and Analysis 5.1 Introduction The previous chapter explored the ARDL model and explained its different steps. This chapter shall adopt these steps to analyze the short run and long run relationship that exists between the variables considered in chapter 4 in the Mauritian context during the period 1983 to 2010. The layout for this chapter shall be as follows:- Section 5.2 focuses on the test for stationarity; section 5.3 depicts the long run relationship; section 5.4 concentrates on the short run relationship; section 5.5 examines the stability of the relationship while the last section shall conclude. 5.2 Tests for Stationarity The Augmented Dickey Fuller (ADF) test is used to know the order of integration of the variables. The results obtained are presented in Table 5.1 with their respective order of integration. Table 5.0. Unit Root Test for Variables Variables Level form ADF Test Variables 1st Difference ADF Test Decision CF -3.3975 … … I(0) GW -3.1288 … … I(0) FDI -2.7748 ΔFDI -6.2272 I(1) ED -1.1692 ΔED -4.5471 I(1) UN -4.1074 … … I(0) ER -0.50609 ΔER -4.7596 I(1) D -4.1074 … … I(0) The ADF results indicate that variables CF, GW, UN and DUM are I(0) while FDI, ED and ER are integrated of order I(1). Since the ARDL bounds tests procedure is valid only if all variables in the model are integrated of order I(d), where 0 ≤ d ≥ 1, this method is sufficiently justified. Table 5.2 displays an ARDL process based on the regression equation (2) in chapter 4. The lag estimates in the table reveals an ARDL (1,2,1,2,2,2,0) process. The diagnostic tests obtained usually show how appropriate and effective the regression is. It can be found that the diagnostic test statistics are satisfactory and the R-Squared is high at nearly 95%, which demonstrates the overall significance of the regression. The R-Squared implies that the explanatory variables together account for 95% of the variation in the capital flight burden. Table 5.0. Lag Estimates for Dependent Variable CF ARDL Estimates(1,2,1,2,2,2,0) Dependent variable is CF 26 observations used for estimation from 1985 to 2010 Regressor Coefficient T-Ratio Prob CF(-1) -0.48339 -2.7346 0.023 GW -0.85969 -1.0527 0.320 GW(-1) -1.8254 -2.5388 0.032 GW(-2) -2.0113 -2.9676 0.016 ED 1.0845 2.9976 0.015 ED(-1) 0.68589 2.2034 0.055 FDI -0.68160 -1.0091 0.339 FDI(-1) -1.1817 -1.6501 0.133 FDI(-2) 2.1977 3.3498 0.009 ER -0.28233 -0.45614 0.659 ER(-1) 1.5476 2.4474 0.037 ER(-2) 1.6259 2.0718 0.068 UN -7.7565 -4.4061 0.002 UN(-1) -0.95829 -0.42166 0.683 UN(-2) 3.8228 2.8794 0.018 D 4.0096 1.1192 0.292 Diagnostic Tests Serial Correlation(1) = 3.48681 [0.062] Normality-n.a Functional Form(1) = 7.4991 [0.006] Heteroscedasticity(1) = 0.0052586 [0.942] After estimating the lag length, the F-test is undertaken to test for the joint significance of the lagged level variables’ coefficients and the results are shown in Table 5.3. The null of no cointegration is rejected as the test statistic value is above the upper bound critical value, indicating the presence of a long run relationship among the variables. The fact that there exists a long run relationship, the second stage can now be carried out to examine the long run and short run impacts of the determinants on the capital flight burden. Table 5.0. F-test Result Model Lower bound Upper bound Computed F-statistic Result Equation(2) I(0) I(1) 8.4963(0.003) Pass 2.86 4.35 Note: lower and upper bound was obtained from the table CI (iii) in pesaran et al(2001) 5.3 Long run relationship Using equation (2) in the previous chapter, the long run estimates of the model is shown as per table 5.4 below. This essay is an example of a student’s work Disclaimer This essay has been submitted to us by a student in order to help you with your studies. This is not an example of the work written by our professional essay writers. Essay Writing Service Dissertation Writing Service Who wrote this essay Place an Order Table 5.0. The Long Run Coefficients Estimate Estimated Long Run Coefficients using ARDL (1,2,1,2,2,2,0) Dependent Variable is CF 26 Observations used for estimation from 1985 to 2010 Regressor Coefficient T-Ratio Prob GW -3.1660 -2.6763 0.025 ED 1.1934 5.8269 0.000 FDI 0.22544 0.32262 0.754 ER 1.9490 3.8539 0.004 UN -3.2978 -6.4073 0.000 D 2.7030 1.1131 0.295 From the results obtained, it is found that all variables except growth rate and unemployment have a negative impact on capital flight burden. Analyzing the variables individually, a p-value less than 0.05 should indicate which variables are significant to the model at 5% significance level. Therefore, it can be concluded that external debt and unemployment are highly significant. In the Mauritian context, economic growth turns out to be a significant determinant of capital flight burden. It can be noted that a 1% increase in the growth rate result in approximately 3.2% fall in the capital flight burden. This suggests that when the economy is in good shape, Mauritian residents are less willing to move their money out of the economy. Therefore, economic growth curtails capital flight in the sense that it boosts the confidence of the residents in the local economy. This result supports the findings of Alam and Quazi (2003) which suggest that capital flight in Bangladesh was caused partly by the lower real GDP growth rate as opposed to Hermes et al (1996) findings. External debt as a share of GDP has a positive and highly significant impact on capital flight burden in Mauritius. A 1% increase in external debt is accompanied by an approximately 1.2% increase in capital flight burden in the long run. This can be explained through the investment climate perspective which states that high external debt indicates that an economy is not doing well which in turn triggers capital flight. In addition, it could be that external debts in Mauritius directly finance capital flight. From Table 5.4, it is apparent that 1% increase in FDI as a share of GDP is associated with an approximately 0.2 % increase in capital flight burden. This positive relationship can be supported from the discriminatory treatment perspective. In Mauritius, there are a series of FDI incentives provided by the government to investors such as the corporate and income tax of 15%, tax free dividends, no capital gains tax, up to 100 percent foreign ownership and a wide tax treaty network with various countries amongst others. Moreover, foreign investors in Mauritius do not need any approval for the repatriation of their profits, dividends, and capital gains since the liberalization of the exchange control. Therefore, from the discriminatory treatment perspective, the fact that Mauritius has many policies in place to promote investment in the island, capital flight is generated in the island. However, it is to be noted that FDI as a percentage of GDP is statistically significant at 10% level. Furthermore, exchange rate is found to have a positive impact on capital flight and is highly significant in the model. Therefore, it implies that in Mauritius, when the exchange rate of the Mauritian rupees adopts a depreciating trend relative to the US dollar, economic agents tend to hold more US dollar or simply send money out of Mauritius which causes an increase in the capital flight. Usually, a positive relationship is expected between unemployment and capital flight since unemployment generally occurs during unfavorable investment climates and periods of low economic growth. However, the above results indicate that as unemployment increases by 1%, capital flight burden is reduced by approximately 3.2%. Though not in line with the expected result, the above findings can be justified to some extent. Faced with high unemployment rate, the government is likely to adopt measures and thus provide incentives to employers with the motive to create jobs. In so doing, the government is in turn helping to retain capital in the country and thus foster growth. Attract FDI to reduce unemployment…having created a favorable clime for investment…capital flight burden is likely to be low. In addition, it has been found that liberalization of exchange rate has led to a higher level of capital flight. With the removal of barriers to trade and exchange controls, capital is able to move more freely thus justifying the positive relationship between the liberalization of exchange rate and capital flight. This essay is an example of a student’s work Disclaimer This essay has been submitted to us by a student in order to help you with your studies. This is not an example of the work written by our professional essay writers. Essay Writing Service Dissertation Writing Service Who wrote this essay Place an Order 5.4 The Short Run Dynamics On the other hand, the short run dynamics are estimated using equation (3) in the previous chapter. These are reported in table 5.5 as the error correction representation. Table 5.0. Error Correction Representation Error Correction Representation for ARDL (1,2,1,2,2,2,0) Dependent variable is CF 26 observations used for estimation from 1985 to 2010 Regressor Coefficient T-Ratio Prob ΔGW -0.85969 -1.0527 0.310 ΔGW1 2.0113 2.9676 0.010 ΔED 1.0845 2.9976 0.010 ΔFDI -0.68160 -1.0091 0.330 ΔFDI1 -2.1977 -3.3498 0.005 ΔER -0.28233 -0.45614 0.655 ΔER1 -1.6259 -2.0718 0.057 ΔUN -7.7565 -4.4061 0.001 ΔUN1 -3.8228 -2.8794 0.012 ΔD 4.0096 1.1192 0.282 ecm(-1) -1.4834 -8.3918 0.000 ecm = CF + 3.1660GW – 1.1934ED – 0.22544FDI -1.9490ER + 3.2978UN – 2.7030D Interestingly, it can be noticed that the impact of some of the variables on capital flight burden is different in the short run as compared to the long run. From the table above, immediate change in the growth rate has a negative effect on capital flight burden but is insignificant at 5 per cent level in the short run. However, the change in the lag of the growth rate (ΔGW1) indicates a positive effect on capital flight burden and is found to be statistically significant. Therefore, in can be deduced that in the short run, change in the lag of growth rate has a positive impact on capital flight burden whereas the growth rate has a negative impact in the long run. Results also suggest that the immediate change in external debt is positive and significant as in the long run. Note that the coefficient of external debt to GDP is approximately the same both in long run and short run. On the other hand, both the change and the change in the lag FDI as a share of GDP have a negative impact on capital flight burden. This relationship can be explained to some extent through the investment climate perspective which states that high FDI in an economy indicates that the economy is favorable for investment and therefore discourages residents to hold assets abroad which in turn curtail capital flight. However, the change in FDI as a percentage of GDP is insignificant in the short run. Wrapping up the results above, it can be concluded that FDI as a share of GDP in Mauritius is found to have a negative impact in the short run whereas a positive effect in the long run on capital flight burden. Moreover, the figures for change and change in the lag of exchange rate report that exchange rate affect capital flight burden negatively. However, it can be drawn from the table that exchange rate appeared to be insignificant in the short run at 5 per cent level. In the short run unemployment and liberalization of exchange control are significant and insignificant, respectively. The immediate effect of both unemployment and liberalization of the exchange control on capital flight burden are the same as in the long run. Besides, the equilibrium correction coefficient (ecm) is estimated at -1.4834 and is highly significant. The ecm coefficient has the correct sign (negative) which implies the speed of adjustment to equilibrium. In other words, the ecm term helps to strengthen the long run equilibrium relationship between the variables in the model. The fact that the value of the ecm coefficient lies between -1 and -2, the error correction process will create dampened oscillations around the long run value before converging to the equilibrium path (Alam and Quazi 2003). 5.5 Stability of the relationship Finally, the Cumulative sum (of square (CUSUMSQ) result is represented in Figure5.1. The figure indicates that the CUSUMSQ plot lies within critical bound of five percent level of significance, which point out that the coefficients in the given regression are stable, that is they are not affected by any significant parameter drift or structural instability over the period being considered. Figure .1 Plot of CUSUMSQ of recursive Residuals 5.6 Conclusion These findings reveal that among the possible determinants considered in the model, external debt as a share of GDP, unemployment rate and the growth rate have the greatest impact on capital flight burden in Mauritius in the long run. It is noteworthy that the change in the lag of FDI as a percentage of GDP has significant impact on capital flight burden in the short run. The effects of some of the variables considered can both be explained from the investment climate and discriminatory point of view. This essay is an example of a student’s work Disclaimer This essay has been submitted to us by a student in order to help you with your studies. This is not an example of the work written by our professional essay writers. Essay Writing Service Dissertation Writing Service Who wrote this essay Place an Order Chapter 6: Conclusion This study attempted to estimate the magnitude of capital flight and highlight possible determinants of capital flight burden in Mauritius, by adopting method used in the literature and empirical studies. Capital flight was thus calculated using the Residual method while capital flight burden was measured as a percentage of GDP. This phenomenon of capital flight is found to be well present in Mauritius. A bound testing (ARDL) approach was used to examine the long run and short run relationship between capital flight burden and various possible determinants of outflow of capital over the period 1985 -2010. The associated error correction procedure is negative and significant confirming relatively high speed of adjustment to equilibrium in long run relationship. The analysis showed that, in the long run, capital flight burden in Mauritius was strongly caused by external debt and unemployment, followed by growth rate and exchange rate. However, it is worth noting that the negative relationship between unemployment and capital flight was not consistent with empirical studies, such as Henry (1996). This could be explained to some extent… Recommendation and policy implication Reduce unemployment and external debt Limitation of Research One constraint in using the residual method when examining the effect of fdi on capital flight burden is that fdi is also a component of the residual measure ( result can be bias). Also, results depend mostly on the data used to compute capital flight and its burden on the economy. Appendix A: Selected Empirical Studies on Determinants of Capital Flight TableA- Selected Empirical Studies on Determinants of Capital Flight Authors Sample & Capital flows Macroeconomic Fiscal policy Risk and Financial depth Political and method environment returns to governance investments factors A. Studies on Sub-Saharan Africa 1. Hermes 6 SSA Debt flows (+) Growth (0); Budget surplus Interest rate and Lensink countries, inflation (0) (0); tax/GDP (0) differential (0); (1992) 1976-1987: exchange rate pooled data overvaluation analysis (+) 2. Murinde, 6 SSA Debt flows Growth (+/0/-); Interest rate Hermes, and countries, 1976- (+/0); grants inflation (+/0) differential (0); Lensink 1991: time- (+/-/0) exchange rate (1996) series analyses overvaluation (+/0) 3. Lensink, 9 SSA Debt flows (+) Inflation (+); Deposit rate (-); Lagged demand Hermes, and countries, lagged capital expected change deposits (-) Murinde 1970-1991: stock (-) in exchange rate (1998) pooled data (+) 4. Olopoenia Uganda, 1971- Growth (0); Parallel market (2000) 1994 inflation (+) premium (0) 5. Nyoni Tanzania, 1973- Debt flows (0); Growth Parallel market Political shock (2000) 1992: past capital differential (+); premium (0); dummy (0) regressions in flight (-) inflation (0) interest rate first differences differential (0) Table A1 (continued) Selected Empirical Studies on Determinants of Capital Flight Authors Sample Capital flows Macroeconomic Fiscal policy Risk and Financial depth Political and environment returns to assets governance factors 6. Ng’eno Kenya, quarterly Real GDP (+) Interest rate (2000) data 1981-1995 differential (-); exchange rate (+) B. Studies on other countries (some samples including SSA countries) 7. 7 Latin Debt flows Inflation (+/0) Real exchange Cuddington American (+/0); past rate (+); US (1987) countries, 1974- capital flight interest rate 1984: Time- (+/0) (+/0) series analyses 8. Dooley 5 Latin Inflation (+) Financial (1988) American repression (+); countries + risk premium on Philippines, external debt (-) 1976-1983: pooled data 9. Pastor 8 Latin Debt flows (+) Growth Change in Interest rate (1990) American differential (-); tax/GDP (0) differential (+); countries, 1973- inflation (+/0) exchange rate 1986: pooled overvaluation data (+) Table A1 (continued) Selected Empirical Studies on Determinants of Capital Flight Authors Sample Capital flows Macroeconomic Fiscal policy Risk and Financial depth Political and environment returns to assets governance factors 10. 22 developing Debt flows (+); Growth (-) Expected Mikkelsen countries, 1978- past capital relative returns (1991) 1985: pooled flight (+) on foreign vs. data + time- domestic assets series analysis (+) for Mexico 11. 4 Latin Inflation (+/0) Budget surplus Interest rate (- Anthony American (-/0) /0); exchange and Hollett countries + rate (+/0); (1992) Philippines, returns on 1976-1988: foreign assets time-series (+/0) analysis 12. Boyce Philippines, Debt flows (+); Growth (0) Budget surplus Interest rate (1992; 1962-1986 past capital (-) differential (+) 1993) flight (0) 13. Vos Philippines, Debt flows (+); Inflation (0) Tax/GDP (0) Interest rate (1992) 1972-1988 debt stock (0); differential (+); past capital exchange rate flight (+) undervaluation (-) Table A1 (end) Selected Empirical Studies on Determinants of Capital Flight Authors Sample Capital Macroeconomic Fiscal policy Risk and Financial depth Political and flows environment returns to assets governance factors 14. Henry Barbados, Jamaica, Debt flows Growth (-/0); Budget surplus Interest rate (1996) and Trinidad, 1971- (+) inflation (-/0) (-/0) differential (+); 1987: time-series exchange rate (- analyses /0) 15. Hermes 84 developing Bank Policy Political and countries, 1971- lending uncertainty: instability (+) Lensink 1991: cross-section (+/0); government (2000) analysis foreign aid consumption (+) (+); tax (+); deficit (+); interest rate (+); inflation (0) 16. 84 developing Bank and Political Lensink, countries, 1971- trade-related instability (+); Hermes and 1991: cross-section lending (+); democracy and Murinde analysis aid (+); FDI political (2000) (0); freedom (-); war (+) 17. Collier, 50 countries Debt stock Capital stock Dollar distortion M2/GDP (0) Governance Hoeffler, (including sub-set of (squared) (+/0) index (squared) indicators (0) and Pattillo 22 SSA countries); (+) (+); investor risk (2001) 1980-1990; cross- (residuals) (0) section analysis Sources: Ndikumana, L. and Boyce, J. K. (2002) Notes: Symbols in parentheses denote a statistically significant positive effect (+); no statistically significant effect (0); or a statistically significant negative effect (-). Where more than one symbol appears in parentheses, this indicates that different specifications yielded different results or that the results vary by country. Appendix B: Magnitude and Burden of Capital Flight :Flows for 1983-98 TableA- Magnitude and burden of capital flight: flows for 1983-98 East Asia South Asia Sub-Saharan Africa Latin America CF CF to CF CF to CF CF to CF CF to (US$ bn) GDP (%) (US$ bn) GDP (%) (US$ bn) GDP (%) (US$ bn) GDP (%) 1983 5 1.1 5 1.7 -0.9 -2.4 4 0.9 1984 7 1.3 0.2 0.1 0.5 1.3 17 3.9 1985 3 0.5 3 0.8 5 11.4 9 2.0 1986 10 1.9 5 1.6 5 10.8 -8 -1.5 1987 37 6.4 6 1.7 6 12.1 15 2.8 1988 33 4.8 -4 -1.2 -2 -3.1 -3 -0.5 1989 17 2.2 8 2.1 2 4.4 5 0.8 1990 50 5.8 -0.1 -0.03 4 8.1 26 3.5 1991 51 5.4 7 2.0 1 2.3 19 2.5 1992 42 4.1 5 1.3 -0.4 -0.8 40 4.9 1993 46 4.1 10 2.7 -0.1 -0.3 16 1.8 1994 140 10.5 17 4.0 2 5.2 30 2.9 1995 113 6.9 -16 -3.3 4 6.9 28 2.3 1996 102 5.6 -6 -1.2 0.7 1.3 52 3.9 1997 103 5.6 -4 -0.7 -3 -5.6 33 2.3 1998 189 12.2 8 1.5 0.4 0.6 24 1.7 Source: Hermes et al (2002) Appendix C: Capital Flight from Mauritius (US$ and MUR): 1980-2010 Year Change in External Debt (ΔED) Net FDI (FDI) Current Account Balance(CAD) Changes in Net Reserves (ΔFR) Capital Flight (US $) Capital Flight(MUR) 1980 114,037,000 1,171,221 (117,122,000) (18,184,390) 16,270,611 124,958,292.48 1981 84,527,000 671,401 (146,812,900) 111,484,400 (173,098,899) (1,547,504,157.06) 1982 45,279,000 1,747,520 (41,112,710) 20,296,330 (14,382,520) (156,337,992.40) 1983 (38,003,000) 1,623,072 (19,135,160) 30,358,930 (85,874,018) (1,005,584,750.78) 1984 (1,916,000) 4,927,417 (51,085,720) (18,608,780) (29,465,523) (406,624,217.40) 1985 67,640,000 7,965,040 (29,528,930) (18,553,910) 64,630,020 997,887,508.80 1986 56,464,000 7,425,913 94,086,310 (122,005,600) 279,981,823 3,771,355,155.81 1987 151,010,000 17,160,760 65,071,120 (219,014,800) 452,256,680 5,825,066,038.40 1988 38,795,000 23,590,300 (56,185,100) (185,068,000) 191,268,200 2,570,644,608.00 1989 (11,112,000) 35,213,650 (103,542,600) (145,588,700) 66,147,750 1,008,753,187.50 1990 119,379,000 40,434,710 (119,285,800) (231,945,200) 272,473,110 4,048,950,414.60 1991 59,480,000 6,516,611 (16,610,970) (190,820,500) 240,206,141 3,759,226,106.65 1992 (48,058,000) (28,593,080) (128,508) (43,329,360) (33,450,228) (520,485,547.68) 1993 (42,611,000) (18,472,320) (92,021,630) (7,004,618) (146,100,332) (2,578,670,859.80) 1994 197,435,000 18,930,570 (232,066,500) 43,505,400 (59,206,330) (1,063,345,686.80) 1995 316,860,000 15,069,320 (21,856,270) (108,824,600) 418,897,650 7,284,630,133.50 1996 (16,413,000) 33,986,950 33,986,950 (48,315,110) 99,876,010 1,792,774,379.50 1997 (59,310,000) 52,096,050 (88,947,950) 34,595,160 (130,757,060) (2,753,743,683.60) 1998 (58,517,000) (1,542,139) 3,292,675 65,383,940 (122,150,404) (2,930,388,191.96) 1999 (9,897,000) 42,921,000 (124,196,900) (189,746,900) 98,574,000 2,483,079,060.00 2000 (325,011,000) 252,689,300 (36,940,500) (230,591,700) 121,329,500 3,184,899,375.00 2001 (98,729,000) (30,543,570) 276,080,900 51,776,330 95,032,000 2,768,282,160.00 2002 62,189,000 23,375,340 249,387,200 (341,082,600) 676,034,140 20,253,982,834.40 2003 70,220,000 68,648,320 93,176,700 (222,386,500) 454,431,520 12,678,639,408.00 2004 (55,794,000) (17,891,790) (111,788,500) 27,464,230 (212,938,520) (5,855,809,300.00) 2005 (141,639,000) (5,187,357) (323,951,500) 164,953,500 (635,731,357) (18,754,075,031.50) 2006 (153,486,000) 97,167,380 (604,410,300) 140,118,800 (800,847,720) (25,394,881,201.20) 2007 38,653,000 281,189,300 (433,934,900) (436,018,800) 321,926,200 10,079,509,322.00 2008 (35,623,000) 325,298,200 (975,764,100) (177,881,600) (508,207,300) (14,458,497,685.00) 2009 193,789,000 218,842,100 (654,971,600) (384,653,000) 142,312,500 4,548,307,500.00 2010 249,863,000 301,650,700 (799,573,100) (209,045,900) (39,013,500) (1,200,835,530.00) Table A- Capital Flight from Mauritius (US$ and MUR): 1980-2010 Source: Authors Computation Total capital flight: 1980-1990: US $1,040,207,234 or MUR Rs 15,231,564,088 1990-2000: US $487,218,947 or MUR Rs 8,657,975,085 2000-2010: US $ (507,002,037) or MUR Rs (15,335,377,523)