Bb

The role of financial sector development and FDI on economic growth in the BRIMCs and some selected African countries: An empirical analysis Babatunde, Bola F. E: [email protected]

Abstract

This paper will investigate the impact, cointegration and causality between financial sector development (FSD), financial sector foreign direct investment (FSFDI) and economic growth. In an attempt to determine what causes economic growth, economists have studied various channels; among them is the role of the financial sector and FSFDI in economic growth. Some argue that financial factors do not have any effect on the level of GDP growth. However, a great deal of empirical research has found a robust relationship between financial development and economic growth. Economic history confirms that less developed countries have hidden and unexploited potential for growth. In order to promote economic growth, they need to modernise their financial systems by replicating financial technology from advanced countries. The increase in financial globalisation through the liberalisation of financial sectors has brought about the integration of developing countries with the global economy. This has led to increase in capital flows (e.g. FSFDI) from developed to developing countries through mergers and acquisitions. FSFDI is assumed to benefit developing countries not only by supplementing domestic investment but also in terms of technology diffusion and increased domestic competition. Therefore, it could be concluded that the provision of incentives to attract FSFDI would lead to new technologies and in turn a more developed financial system. Using three alternative measures of FSD, this paper will examine the short run and long run relationship and causality between FSD, FSFDI and economic growth in Brazil, Russia, India, Mexico and China (BRIMCs) and some selected African countries. The analysis will be conducted using panel unit root and panel cointegration analysis techniques with the short run and long run causality tested in the panel context using a Granger causality test.

Keywords: Financial sector development, FSFDI, BRIMCs, technology diffusion, panel unit root and panel cointegration analysis.

i

Acronyms and Abbreviations:

ACs

African countries

BC

Bank credit

BRIMCs

Brazil, Russia, India, Mexico and China.

CEEC

Central eastern European countries

ECM

Error correcting model

FDI

Foreign direct investment

FMOLS

Fully modified ordinary least square

FSFDI

Financial sector foreign direct investment

FSD

Financial sector development

GDP

Gross domestic product

GMM

Generalised methods of moments

HC

Human capital

IFI

International Financial Integration

IFS

International Financial Statistics

Mcap

Stock market capitalisation

OECD

Organisation of Economic Cooperation and Development

gPcap

real GDP per capita (annual growth)

PC

Private Credit

Inf

Inflation

TV

Trade Volume

WDI

World Development Indicator

WIR

World Investment Report ii

1. Introduction: The immense difference in economic growth across countries has led to a resurgence of research interest in the determinants of economic growth, a subject which has been extensively debated. Many of the literature contain competing explanations of economic growth and the notable ones are the importance of development of financial and banking systems (King and Levine, 1993a, b; Levine and Zervos, 1996 and Levine et al, 2000), the role of foreign direct investment (hereafter, FDI) (Caves, 1982; Borensztein et al, 1995, 1998; De Mello, 1997, 1999 and Campos and Kinoshita, 2002), importance of financial sector FDI, (hereafter, FSFDI) (Herrero and Simon, 2003; BIS, 2004; Goldberg, 2004; Domanski, 2005 and Khan, 2007) the role of institutions and the quality of macroeconomic policies (Rodrik et al, 2002; Berg and Krueger, 2003 and Easterly and Levine, 2003) and international financial integration (hereafter, IFI) and globalisation (Agmon, 1972; Bordo, 2000, Kim et al, 2000 and Camilleri, 2003). A general characteristic of the different research papers that have examined this issue is that they relate to a combination of developed and developing countries. Interestingly, the results of these researches have been very convincing but not conclusive therefore leading to the need for more research in this area. One aspect of the literature that has received considerable study by researchers and academics is the importance of FSD in explaining economic growth. For example, following the seminal studies of Goldsmith (1969), McKinnon (1973) and Shaw (1973), there has been a huge amount of literature on the effects of financial sector development (FSD) on economic growth. This issue has also been studied empirically by King and Levine (1993a, b), Levine and Zervos, (1996) and Rousseau and Wachtel (2002). The general consensus is that a well functioning financial sector, by lowering transaction costs, channel capital to the investments that yield high returns and therefore promotes economic growth. Hence, the literature point out that FSD exerts a strong positive effect on economic growth.

While the majority of the research that used cross country and panel data studies found evidence that a well developed financial sector promotes economic growth, results using the time series method are contradictory. King and Levine (1993a) conclude that FSD exerts a positive impact on economic growth and Levine and Zervos, (1998) indicate that both stock market and bank development are positively related to capital accumulation, productivity growth and current and future rates of economic growth. Nourzad (2002) pointed out that financial deepening reduces productivity inefficiency in both developed and developing countries, however, this is more pronounced in the developed countries. Guillaumont, Hua and Liang (2006) found that financial development contributes to total factor productivity growth. Alternatively, Chang (2002) found no causal relationship between financial development and economic in China. Also, Ghirmay (2005) using a sample of 13 sub Saharan African countries, concluded that the causality between financial development and economic growth is bidirectional in six of the thirteen countries. The results also show that financial development granger causes economic growth in eight of the countries. The author suggests that for African countries to develop,

3

they need to improve their financial systems. Using cointegration and error-correction techniques to study the issue of causality between financial depth, savings and economic growth in Kenya, the author found that there exists a unidirectional causal flow from economic growth to financial development. In addition, it was found that FSD in Kenya is driven by savings.

Similarly, many previous studies have recognised the importance of IFI in explaining economic growth. In the neoclassical growth model, IFI eases the flow of capital to capital scarce countries with positive output effects (Edison et al, 2002). According to theoretical literature, IFI can promote economic growth through risk sharing and it may also enhance the functioning of domestic financial systems through competition and improved financial services with a positive effect on economic growth, (Levine, 2001). However, Boyd and Smith (1992) show that IFI only promotes economic growth in countries with well developed financial institutions and sound policies.

Using FDI as a measure of capital flow which could also proxy for IFI, Blomstrom et al (1994) find that FDI is positively linked with economic growth when countries are sufficiently rich. Similarly, Borenzstein et al (1998) find that FDI is important in determining economic growth when the population is educated. However, in Africa, the story is quite different. The OECD reports that poor quality services, closed trade regimes, lack of political legitimacy and unsustainable national economic policies hinder African countries in attracting FDI which may lead to positive economic 1

growth . Furthermore, Mottaleb (2007) indicates that countries with friendly business environments receive more FDI inflow than those with less friendly business environments.

Zhang (2001) employing a cointegration and Granger causality test and using a sample of eleven countries from Latin America and Asia to study the link between FDI and economic growth, found that FDI promotes economic growth in five out of the eleven countries. Chakraborty and Nunnenkamp (2008) indicate that there is strong bidirectional causality between FDI stock and output in India in the short run, however they did not find evidence for a causal relationship in the long run.

Recent studies by Eller et al. (2005), Alfaro et al. (2006), Ljungwall and Li (2007) and Khan, (2007) add new insight to the debate on economic growth. These studies examine the view that FSD may influence FDI in promoting economic growth by taking advantage of technology and knowledge spillovers which are benefits of FDI inflow. In the literature, economic growth is positively related to FDI in the financial sector. Similarly, according to the literature on finance-growth nexus, economic growth rarely occurs without a well functioning financial sector. This is because a well developed financial sector promotes economic growth through its effect on capital accumulation (Levine, 2004). Furthermore, the literature supports the notion that a well developed financial sector positively contributes to the process of technology diffusion associated with FDI in the host country and in turn enhance economic growth, (Alfaro et al, 2006; Khan 2007 and Ljungwall and Li, 2007). Consequently,

1

OECD report 2002: 8.

4

the level of FSD is likely to influence the extent to which economic growth responds to foreign capital flow in general and FDI in particular.

According to Eller et al (2005) if high human capital is present, FDI in the financial sector can promote economic growth. The underlying principle with this line of reasoning can be traced to the belief that FDI encourages technology transfers and knowledge spillovers and in turn enhances economic growth in the host countries, (see for example, Blomstrom et al, 1996; Borensztein et al 1995; 1998; De Mello, 1997; 1999 and Campos and Kinoshita, 2002). Therefore, FDI in the financial sector may have an impact on economic growth depending on the absorptive capacity of the host country.

While the literature on FSFDI seeks to investigate how the response of economic growth to FDI varies with the level of FSD, the research has mostly concentrated on Asia, Central and Eastern European countries, (CEECs), Latin America and a few country case studies. In addition, the debate on the link between FSD, FDI and economic growth is still in its infancy. Studying the response of economic growth to FDI spillovers through the level of development of domestic financial sector in other geographic regions such as Africa and also in income groups (for example, low income, low middle income and upper middle income groups) will not only add to the body of literature but also be a valuable guide to especially policy makers.

2. Data and Sources: To investigate the relationship between FSD, FDI and economic growth, this paper uses a 2

heterogeneous and homogenous panel of 10 developing countries over the period 1980-2006. The countries are then grouped under BRIMCs and African countries (hereafter, ACs). The countries are 3

4

then divided into geographical region and income groups according to the World Bank income 5

classification .

In the literature studied, Levine (2003) explained that the degree to which financial development impacts on economic growth varies with the type of financial market indicator used and the level of financial development of a country. However, because FSD varies across countries and over time, to be able to capture the range of differences, this paper will use three different FSD indicators. The FSD indicators will be divided into two categories: the banking sector and the stock market. This paper will use two banking sector variables because FSD usually takes place in the banking sector in most ACs. The variables chosen are bank credit (BC) and private credit (PC). The bank credit refers to domestic credit provided by the banking sector as a percentage of GDP. In the literature, a higher

2

The countries include Botswana, Brazil, China, Ghana, India, Mexico, Morocco, Nigeria, Russia and Tunisia. The geographical region and sub region covered include: Africa (SSA and MENA), Asia (South and East), Europe and Latin America. 4 This is divided according to the World Bank income groups’ classification and it includes: low, low middle and upper middle income groups. 5 The World Bank only categorizes geographic regions for low and middle-income countries. Therefore, to some extent, each geographic region has homogeneity in the level of economic growth and financial development. 3

5

bank credit indicates a high level of dependence on the banking sector. The literature concludes that the development of bank credit has an important impact on economic growth (see for example, Hassan and Yu, 2007). The second variable, which is private credit (PC), equals the value of domestic credit to private sector divided by GDP. This measure of FSD was developed by Levine et al (2000) as it provides an accurate measure of overall financial development. A high ratio of PC indicates a high level of domestic investment, which in turn indicates high output (Hassan and Yu, 2007). For the second category, stock market capitalisation (mcap) will be used to proxy stock market. This variable is used to capture the overall size of the stock market and it was developed by King and Levine (1993a, b). It is measured by the average value of listed domestic shares on the domestic stock exchange in a year as a share of the size of the economy. The data on FSD indicators were collected from the World Banks‟ World development indicators (WDI, 2008) online database and World Bank Financial Structure database.

The present paper uses the net inflow of FDI to proxy FDI. The International Financial Statistics (IFS) reported that FDI net inflow is the net inflow of investment required to obtain a lasting management interest (10 per cent or more of voting stock) in an enterprise operating in an economy other than that of the investor. Carkovic and Levine (2005) use FDI gross inflow, which is the sum of absolute value of inflow accounted in the balance of payments financial accounts. Because this research is focused mainly on inflows to an economy, the FDI net inflow is preferred. Also, according to et al (2006), since it is not certain that external investment can create spillovers in host economies, FDI net inflow is a better measure of FDI. The data on FDI were collected from International Monetary Funds‟ IFS, (IFS, 2008) online database and the United Nations Conference on Trade and Development‟s World Investment Report (WIR).

In this paper, real GDP per capita growth rate (annual per cent) is used to proxy for economic growth, (gpcap). The data on this variable is collected from the World Banks‟ World Development Indicator, (WDI, 2008) database online. To assess the strength of the link between FSD, FDI and economic growth variables, various conditioning information variables is used. These are in line with previous studies and are aimed to control for other factors associated with economic growth (e.g. Beck et al, 2000; Levine et al, 2000; Easterly and Rebelo, 1993; Eller et al, 2005; Mahamet, 2006; Alfaro et al, 2003 and Apergis et al, 2007). In the present paper, four variables are considered: the ratio of trade to GDP (TV) which is used to proxy openness to international trade, human capital (HC) is measured using secondary school enrolment, inflation rate (Inf) which is measured as the percentage change in the consumer price index to proxy macroeconomic stability and telephone subscribers (tel) which is used to measure the level of infrastructural development. The data on the control variables were collected from the World Banks‟ World Development Indicators (WDI, 2008) online database and Barro and Lee (2000) series. To investigate whether FSD is a factor in the link between FDI and economic growth, the present paper uses an interaction term (FSFDI) between the measures of FSD indicators and the proxy for FDI inflow.

6

3. Model specification: 3.1. Model Specification: The sub-section presents the specification of the model. The aim of this empirical analysis is to investigate whether FSD is a factor in the link between FDI and economic growth in the BRIMCs and selected ACs. In order to do this, three models will be considered. First, to consider the relationship between FSD and economic growth, the empirical analysis considers three FSD indicators which capture both the stock market and banking development. Therefore, as a starting point, this paper studies the impact of FSD on economic growth. gPCAP𝑖𝑡 = 𝛽0𝑖 + 𝛽1𝑖 𝑙𝑜𝑔 gPCAP𝑖𝑡 + 𝛽2𝑖 𝐹𝑆𝐷𝑖𝑡 + 𝛽2𝑖 𝑋𝑖𝑡 + 𝜀𝑖𝑡

(1)

where gPCAP𝑖𝑡 is GDP per capita growth (annual per cent), 𝑙𝑜𝑔(gPCAP𝑖𝑡 ) is the log of GDP per capita growth in the initial year of consideration, 𝐹𝑆𝐷𝑖𝑡 is the FSD indicator and 𝑋𝑖𝑡 is a set of control variables or other factors that affect economic growth. To address the issue of the extent to which financial development impact on economic growth, the three FSD indicators will be estimated simultaneously.

After this has been done, the paper then turns its attention to study the direct impact of FDI on economic growth by following the model specification of Mankiw et al (1992) and Alfaro et al (2006). The following model is estimated: gPCAP𝑖𝑡 = 𝛽0𝑖 + 𝛽1𝑖 𝑙𝑜𝑔(𝑖𝑛𝑖𝑡𝑖𝑎𝑙 gPCAP𝑖𝑡 ) + 𝛽2𝑖 𝐹𝐷𝐼𝑖𝑡 + 𝛽2𝑖 𝑋𝑖𝑡 + 𝜀𝑖𝑡

(2)

where gPCAP𝑖𝑡 is GDP per capita growth (annual per cent), 𝑙𝑜𝑔(gPCAP𝑖𝑡 ) is the log of GDP per capita growth in the initial year of consideration, 𝐹𝐷𝐼𝑖𝑡 is the FDI variable and 𝑋𝑖𝑡 is a set of control variables or other factors that affect economic growth.

To examine the link between FSD, FDI and economic growth, a similar model to Alfaro et al (2006) and Aslan et al (2008) is estimated. Here, an interaction term (FSFDI) between the three FSD indicators and FDI inflow variable is included in the regression. To ensure that the interaction term does not proxy for both FSD indicators and FDI variable; both the FSD indicators and FDI variable are included in the regression independently. Therefore, the following model is estimated: gPCAP𝑖𝑡 = 𝛽0𝑖 + 𝛽1𝑖 𝐹𝑆𝐷𝑖𝑡 + 𝛽2𝑖 𝐹𝑆𝐷∗ 𝐹𝐷𝐼

𝑖𝑡

+ 𝛽3𝑖 𝐹𝐷𝐼𝑖𝑡 + 𝛽4𝑖 𝑋𝑖𝑡 + 𝜀𝑖𝑡 (3)

where gPCAP𝑖𝑡 is GDP per capita growth (annual per cent), 𝑙𝑜𝑔(gPCAP𝑖𝑡 ) is the log of GDP per capita growth in the initial year of consideration, 𝐹𝑆𝐷𝑖𝑡 is the FSD indicator, 𝐹𝐷𝐼𝑖𝑡 is the FDI variable, 𝐹𝑆𝐷 ∗ 𝐹𝐷𝐼

𝑖𝑡

is the interaction term used to measure level of FDI in the financial sector and 𝑋𝑖𝑡 is a

set of control variables or other factors that affect economic growth. Controlling for initial conditions is

7

common to cross country regressions, therefore, this study also controls for the initial income by using the log of GDP per capita growth in the initial year of consideration.

3.2.

Econometric methodology and framework:

The study follows recent development in the literature (e.g., Christopoulos and Tsionas, 2004; Alfaro et al, 2006; Hassan and Yu, 2007; Apergis et al, 2007 and Aslan, 2008) and employs the use of panel data analysis to study the link between FSD, FDI and economic growth.

3.2.1. Panel unit root test: The variables need to be stationary in order to test for causality. Recent literature suggests that a panel unit root test has higher power than unit root test on time series data. As a result, the panel unit root test will be used to make the variables stationary because the use of the Augmented Dickey Fuller (ADF) test in most cases misinterprets stationarity due to the small size of the sample. The first methods developed to tackle the issue of unit root in panel data only dealt with the use of univariate panels. However, Quah (1994) advanced this methodology by deriving standard normal asymptotic distributions for testing unit root in panels as the size grows larger. Levin and Lin (1992) and Levin, Lin and Chu (2002) also derived methods that allow for heterogeneous fixed effects and time trends. The Levin, Lin and Chu (2002) unit root test considered the basic ADF test, using the following equation; 𝑑𝐴𝑖,𝑡 = 𝛼𝐴𝑖,𝑡−1 +

𝑝𝑖 𝑗 =1 𝛽𝑖,𝑡 𝑑𝐴𝑖,𝑡−𝑗

∗ + 𝐵𝑖,𝑡 𝛾 + 𝜀𝑡

(3.2)

where 𝑑𝐴𝑖,𝑡 = differenced panel data, α = ρ – 1, 𝐴𝑖,𝑡−1 = panel data, pi = the number of lag order for the ∗ differenced terms, 𝐵𝑖,𝑡 = contains the unobserved country specific and time specific effects and 𝜀𝑡 is

the error term that contains all unexplained information in the data. Equation 3.2 can be re written as: 𝑑𝐴𝑖,𝑡 = 𝑑𝐴′𝑖,𝑡−1 +

𝑝𝑖 ∗ 𝑗 =1 𝐵𝑖,𝑡 𝑑𝐴𝑖,𝑡−𝑗

∗ + 𝐵𝑖,𝑡 𝛾 + 𝜀𝑡

(3.3)

to remove the autocorrelation and the deterministic components. where 𝐴′𝑖,𝑡−1 is defined as 𝐴′𝑖,𝑡−1 = 𝐴𝑖,𝑡−1 +

𝑝𝑖 ∗ 𝑗 =1 𝐵𝑖,𝑡 𝑑𝐴𝑖,𝑡−𝑗

∗ + 𝐵𝑖,𝑡 𝛾

(3.4)

The next step is to divide both equations 3.3 and 3.4 by the estimated standard error of the regression from the ADF equation which is represented by 𝑆𝑖 . Therefore we have; 𝑑𝐴𝑖,𝑡 = 𝑑𝐴𝑖,𝑡 /𝑆𝑖

(3.5)

𝐴′𝑖,𝑡 = 𝐴′𝑖,𝑡−1 /𝑆𝑖

(3.6)

8

Levin, Lin and Chu (2002) show that, under the null hypothesis, a modified t-statistics for the resulting 𝛼^ is asymptotical normally distributed and can be written as: 𝑡 ∗ = 𝑡 − 𝑁𝑇 𝑆𝑁𝜎 ^ − 2 𝑆𝑒 𝑎^ µ𝑚𝑇 ∗ / 𝜎𝑚𝑇 ∗ à N

(3.7)

where 𝑡 ∗ = the standard t-statistic for a^ = 0 𝜎 ^ = the estimated variance of the error term η, 𝑇 ∗ = 𝑇 −

𝑖 𝑝𝑖

𝑁

− 1

Se (𝛼^) = the standard error of a^. The null hypothesis for this test is 𝐻0 : 𝑝1 = 0 panel data has unit root and the alternative 𝐻𝑎 : 𝑝1 < 0 panel data has no unit root. This test assumes that 𝑝1 is the same for all members in the panel as the test is done on a pooled data. If 𝑡 ∗ is significant, then we reject the conclusion that the panel data has no unit root. Otherwise, if 𝑡 ∗ is not significant, then we accept the null hypothesis that the panel has a unit root. Im, Pesaran and Shin (1996) focused on small sample properties of unit root tests in panels with heterogeneous dynamics and developed an alternative test based on the group mean statistics. In 2003, Im, Pesaran and Shin (2003, Ipshin hereafter), proposed a t-bar unit root test that allows for complete heterogeneity units in the dynamic panel framework. The test is based on individual ADF regressions: 𝑦𝑖,𝑡 = 𝜌𝑖 𝛾𝑖,𝑡−1 +

𝑝 𝑗 =1 ∅𝑖𝑗

𝛾𝑖,𝑡−𝑗 + 𝑧𝑖,𝑡 𝛾 + 𝜀𝑖,𝑡

(3.8)

where i= 1,.....N, t = 1,.....T. The null hypothesis in the Ipshin test is 𝐻0 : 𝛽𝑖 = 0 for all 𝑖 which is tested against the alternative 𝐻𝑎 : 𝛽𝑖 < 0, 𝑖 = 1,2, … . . 𝑁, 𝛽𝑖 = 0, 𝑖 = 𝑁𝑖 + 1, 𝑁𝑖 + 𝑁. In a traditional ADF, the t values are compared to a critical value while in the Ipshin unit root test the sample mean (t-bar) is estimated as 𝑡=

1 𝑁

𝑁 𝑖=1

𝑡𝑝1

(3.9)

𝑡𝑝1 is the individual t- statistics for testing the null hypothesis. Whilst this study will use the Levin, Lin and Chu (2002) and the Ipshin test for unit root, it will also employ the use of another unit root test, the Maddala and Wu (1999) method to confirm the results of the previous tests.

9

3.2.2. Panel cointegration test: To study the long run relationship between the variables, this study will use panel cointegration analysis. There is a need to estimate the order of integration of the variables by first conducting a panel unit root analysis, after which, the panel cointegration technique can be used. To determine if a long run relationship exists between variables, Pedroni (1999) developed a panel cointegration technique; this extends the two step residual based strategy of Engle and Granger (1987). The first step is to test for the hypothesis of no cointegration by regressing residuals from the hypothesised cointegration regression. Using a panel regression: 𝑦𝑖𝑡 = 𝛼𝑖 + 𝛽1 𝑋1,𝑖,𝑡 + 𝛽2 𝑋2,𝑖,𝑡 … … … . + 𝛽𝑛 𝑋𝑛,𝑖,𝑡 + 𝑢𝑖𝑡

(4)

where 𝑋𝑖,𝑡 are the independent variables for N cross sections. We then perform another regression on the residuals from the previous regression: 𝑢𝑖,𝑡 = 𝜌𝑖 𝑢𝑖,𝑡−1 + 𝑣𝑖,𝑡

(4.1)

A fixed or random effect can be used when estimating the regression. Based on the assumption that cointegration between variables is heterogeneous, Pedroni (1999) performed seven different tests. Four of these tested for cointegration “within” dimensions (panel cointegration statistics) with the null hypothesis as 𝐻0 : 𝜌𝑖 = 0 for all 𝑖 with the alternative as 𝐻𝑎 : 𝜌𝑖 = 𝜌𝑖 < 1 for all 𝑖. The last three tested for cointegration “between” dimensions (group mean cointegration statistics). The null hypothesis is given as 𝐻0 : 𝜌𝑖 = 0 for all 𝑖 and the alternative is 𝐻𝑎 : 𝜌𝑖 < 1 for all 𝑖. After being able to establish cointegration between (among) variables, we can then apply the fully modified ordinary least square (FMOLS) technique proposed by Pedroni, (1996; 2000). The FMOLS is used in order to control for endogeneity of explanatory variables. However, the generalised methods of moments GMM error correction model (ECM) can also be applied to control endogeneity of the regressors, (Eller et al, 2005). Westerlund (2007) developed a more general test to capture the issue of endogeneity; this is referred to as the four panel cointegration test. This method tests for the absence of cointegration by determining whether individual panel members are error correcting. In other words, this is an error correcting panel based cointegration test. He considered an error correcting model

`

𝑦𝑖𝑡 = 𝜙1𝑖 + 𝜙2𝑖 𝑡 + 𝑧𝑖𝑡

(4.2)

𝑥𝑖𝑡 = 𝑥𝑖𝑡 −1 + 𝑣𝑖𝑡

(4.3)

Where t = 1,......., T for the time series dimension and N = 1,......, N for the cross section dimension. 𝑥𝑖𝑡 is the K dimensional vector which is modelled as a random walk, while 𝑦𝑖𝑡 consists of both deterministic trend 𝜙1𝑖 + 𝜙2𝑖 𝑡 and a stochastic term, 𝑧𝑖𝑡 . This can be re written as: ′ ′ 𝛼𝑖 𝐿 ∆𝑍𝑖𝑡 = 𝛼1 𝑍𝑖𝑡 + 𝛽𝑖,𝑥𝑖𝑡 −1 + 𝛾1 𝐿 + 𝑣𝑖𝑡 + 𝑒𝑖𝑡

10

(4.4)

where 𝛼𝑖 𝐿 = 1 −

𝑝𝑖 𝑗 𝑗 =1 𝛼𝑖𝑗 𝐿

and 𝛾1 𝐿 = 1 −

𝑝𝑖 𝑗 𝑗 =0 𝛾𝑖𝑗 𝐿

are scalar and K dimensional polynomials in

the lag operator, L. By substituting equation (4.2) into equation (4.4), we have: ′ 𝛼𝑖 𝐿 ∆𝑦𝑖𝑡 = 𝛿1𝑖 + 𝛿2𝑖 + 𝛼1 𝑦𝑖𝑡 + 𝛽𝑖,𝑥𝑖𝑡 −1 + 𝛾1 𝐿 ′ + 𝑒𝑖𝑡

(4.5)

The error correcting model is re written as: ∆𝑦𝑖𝑡 = 𝛼𝑦 − 𝛾𝛼𝑥 𝑦𝑖𝑡 −1 − 𝛽1𝑥𝑖𝑡 −1 + 𝛾∆𝑥𝑖𝑡 + 𝑒𝑖𝑡

(4.6)

∆𝑥𝑖𝑡 = 𝛼𝑥 𝑦𝑖𝑡 −1 − 𝛽1𝑥𝑖𝑡 −1 + 𝑣𝑖𝑡

(4.7)

The null hypothesis here is 𝐻0 : 𝛼𝑖 = 0 for all 𝑖

which is tested against the alternative 𝐻𝑎 : 𝛼𝑖 <

0 for all 𝑖. In the event that 𝐻0 of no cointegration is rejected, this also affects the whole dimension. The method provides flexibility as it allows for heterogeneous specification of both the long and short run parts of the error-correction model, where the latter can be determined from the data. The series are allowed to be of unequal length. If the cross-sectional units are suspected of being correlated, robust critical values can be obtained through bootstrapping. In addition, a GMM error correction model will be used in order to deal with causality issue. In summary, the methodology which will be employed to examine the link between FSD, FDI and economic growth will follow these steps: 1. Test the properties of the data; specifically to check whether or not the series is stationary. 2. To check through the Levin, Lin and Chu (2002), Ipshin test (2003) and Maddala and Wu (1999) the order of integration of the variables. 3. To check through Pedroni (1999) and Westerlund (2007) test whether or not the series are cointegrated. If the series are cointegrated and a long run relationship is established, then a causality test can be carried out using the FMOLS estimator by Pedroni (2000) and the GMM error correction model of Arellano and Bond, (1991) and Arellano and Bover (1995). 4. To determine whether the causality occurs in the short run or long run. 5. To investigate the direction of causality to check if it suggests a long run causality.

4. Conclusion: There are no initial findings for the study. All the required data has been gathered and treated and is now ready for empirical analysis. It is hoped that the findings that are obtained in this study are relevant to policymakers in the sample in general and the selected African countries in particular.

11

References and Bibliographies: Agmon T. (1972), “The relations among equity markets: a study of share price co-movements in the United States, United Kingdom, Germany and Japan”, Journal of Finance, Vol 27 (4), p839-855. Alfaro, L., A. Chanda, S. Kalemli-Ozcan and S. Sayek, (2003), “FDI and economic growth: The role of the local financial markets”, Journal of International Economics, Vol 64(1), p 89-112. Alfaro, L. A. Chanda, S. Kalemli-Ozcan and S. Sayek, (2006), “How does Foreign Direct Investment promote economic growth? Exploring the effects of financial markets on the linkages” [Online], available on accessed on 20/01/09. Arellano M. and S. Bond, (1991), “Some tests of specification for panel data: Monte Carlo evidence and an application to employment”, Review of Economic Studies, Vol 58, p 277-297. Arellano M. and O. Bover, (1995), “Another look at the instrumental-variable estimation of errorcomponents models”, Journal of Econometrics, Vol 68, p 29-52. Aslan, A. (2008), “Financial development and economic growth: A panel cointegration analysis of Middle East countries”, [Online], available at accessed 01/09/08. Apergis, N, I. Filippidis and C. Economidou, (2007), “Financial deepening and economic growth linkages: A panel data analysis”, Review of World Economics, Vol 143(1). Bank for International Settlement, BIS, (2004), “Foreign direct investment in the financial sector of emerging market economies”, [Online], available on report submitted by a working group established by the committee on the Global Financial System accessed 20/01/09. Barro, R. J. and Jong-Wha Lee, (2000), “International Data on Educational Attainment: Updates and Implications”, Centre for international development, Working Paper No. 42, Harvard University. Beck, T., R. Levine and N. Loayza, (2000), “Finance and the Sources of Growth”, Journal of Financial Economics, Vol 58, p 261-300. Berg, A., and A. Krueger, (2003), “Trade, Growth, and Poverty - A Selective Survey”, IMF Working Paper 03/30, Washington: International Monetary Fund. Blomstrom, M and E.N Wolff (1994), “Multinational Corporations and Productivity Convergence in Mexico,” National Bureau of Economic Research working paper no 3141. Blomstrom, M and A. Kokko (1996), “How foreign investments affect host countries”, The World Bank International Economics Department, International Trade Division, policy research working paper no. 1745, [Online], available on accessed 16/02/09. Bordo M.D, (2000), “The Globalization of International Financial Markets: What Can History Teach Us?” [Online] available from < http://econweb.rutgers.edu/bordo/global.pdf> accessed 25/11/08.

12

Borensztein, E, J. De Gregorio and J. W. Lee, (1995), “How does FDI affect growth?” [Online], available on National Bureau of Economic Research working paper no W5057 accessed 20/01/09. Borensztein, E, J. De Gregorio and J. W. Lee, (1998), “How Does Foreign Direct Investment Affect Economic Growth?” Journal of International Economics, Vol 45, p115-135. Boyd, J. H., and B. D. Smith, (1992), “Intermediation and the equilibrium allocation of investment capital: implications for economic development”, Journal of Monetary Economics, Vol 30, p 409-432. Brada, J.C, A.M. Kutan and T.M, Yigit, (2006), “The effects of transition and political instability on foreign direct investment inflows.” The Economics of Transition Vol 14(4), p649-680. Camilleri S.J (2003), “Globalisation of international financial markets”, [online] available from accessed on 26/11/08. Campos N. F and Y. Kinoshita (2002), “Foreign Direct Investment as Technology Transferred: Some panel evidence from the transition economies” William Davidson working paper no 438. [Online] available at

accessed 02/01/09.

Carkovic, M. and R. Levine, (2005) “Does Foreign Direct Investment Accelerate Economic Growth?” University of Minnesota, Working Paper. Caves, R. E. (1982), Multinational Enterprise and Economic Analysis, Cambridge, U.K.: Cambridge University Press. Caves, R.E., (1996), Multinational Enterprise and Economic Analysis, 2 Press, Cambridge.

nd

ed, Cambridge University

Chang T, (2002), “Financial development and economic growth in mainland china: a note on testing the demand-following or supply-leading hypothesis”, Applied Economic Letters, Vol 9, p 869-873. Chakraborty C. and P. Nunnenkamp (2008) “Economic Reforms, FDI, and Economic Growth in India: A Sector Level Analysis”, World Development, Vol 36 (7), p 1192-1212. Christopoulos, D. and E. Tsionas (2004), “Financial development and economic growth: Evidence from panel unit root and cointegration tests”, Journal of Development economics Vol 51(2), p 387411. De Mello, L.R. (1997). “Foreign direct investment in developing countries and growth: A selective survey”, Journal of Development Studies, Vol 34, p 1-34. De Mello L.R (1999), “Foreign direct investment-led growth: evidence from time series and panel data,” Oxford Economic Papers 51, p 133-151. Domanski, D. (2005), “Foreign banks in emerging market economies: changing players, changing issues” [Online], available on BIS quarterly review accessed on 25/01/09. Easterly, W and R. Levine, (2003), “Tropics, Germs and Crops: How Influence Economic Development”, Journal of Monetary Economics Vol 50, p 3-39.

13

Endowments

Edison H, M. Klein, L. Ricci and T. Slok, (2002), “Capital account liberalization and economic performance: Survey and Synthesis”. IMF working papers 02/120. Eller M, P. R Haiss and K. Steiner, (2005), “Foreign direct investment in the financial sector: the engine of growth for Central and Eastern Europe” [Online], available on accessed on 20/01/09. Engel R.F and C.W.J Granger (1987), “Cointegration and error correction: representation, estimation and testing”, Econometrica, Vol 55, p 251-276. Easterly, W, and S. Rebelo, (1993), “Fiscal policy and economic growth. an empirical investigation”, Journal of Monetary Economics Vol 32 (3): p 417-458. Ghirmay, T, (2005), “Financial development and economic growth in sub-Saharan African countries: evidence from time series analysis”, African Development Review, Vol 16(3), p 415-432. Goldberg, L.S, (2004), “Financial sector FDI and host countries: New and Old lessons”, [Online], available on accessed 20/01/09. Goldsmith R.W, (1969). Financial structure and development, New Haven: Yale University Press. Guillaumont, S, P. Hua and Z. Liang, (2006), “Financial development, economic efficiency, and productivity growth: Evidence from China”, Developing Economies, Vol 44(1), p 27-52. Hassan, K.M and Jung-Suk Yu, (2007), “Financial sector reforms and economic growth in Morocco: An empirical analysis”, Networks Financial Institute Working Paper No. 2007-WP-28. Herrero, A.G and D.N Simon, (2003), “Determinants and impact of financial sector FDI to emerging economies: A home country‟s perspectives”, [Online], available on accessed 09/02/09. Im, K.S, M. H Pesaran and Y.Shin, (1996), “Testing for unit roots in heterogeneous panels”. Journal of Econometrics. Vol 115(1), p 53-74. Im, K.S, M. H Pesaran and Y. Shin, (2003), “Testing for unit roots in heterogeneous panels”, Journal of Econometrics Vol 115, pgs 53-74. Khan, M.A, (2007), “Foreign direct investment and economic growth: The role of domestic financial sector”, [Online], available on Pakistan Institute of Development Economics PIDE Working Paper no 2007:18 accessed 20/01/09. King, R.G and R. Levine, (1993a) “Finance and growth: Schumpeter must be right”, Quarterly Journal of Economics, Vol 108(3), p 717-737. King, R.G and R. Levine, (1993b) “Finance, Entrepreneurship and Growth: Theory and Evidence”, Journal of Monetary Economics, Vol 32(3), p 513-542. Kim E.H and S. Singal (2000), “Stock Market Openings: Experience of Emerging Economies”, Journal of Business 73(1) p 25-66. Levine A and C.F Lin, (1992), Unit root tests in panel data: Asymptotic and finite sample properties, Mimeo, University of California San Diego.

14

Levine, R. and S. Zervos (1996), “Stock Market Development and Long-Run Growth”, The World Bank Economic Review, Vol 10(2), p 323-339. Levine, R. and S. Zervos, (1998), “Stock Markets, Banks and Economic Growth”, American Economic Review, Vol 88(3), p 537-558. Levine, R, N. Loayza and T. Beck, (2000), “Financial intermediation and growth: Causality and causes”, Journal of Monetary Economics, Vol 46 (1), p 31-77. Levine, R., (2001), “International Financial Liberalization and Economic Growth”, Review of International Economic, Vol 9, p 688-702. Levine, A., C.F. Lin and C. Chu, (2002), “Unit root tests in panel data: Asymptotic and finite sample properties”. Journal of Econometrics Vol 108, p 1-24. Levine, R., (2003), “More on Finance and Growth: More Finance, More Growth?” „[Online], available on , The Federal Reserve Bank of St. Louis accessed 02/12/08. Levine, R, (2004), “Finance and growth: theory, evidence and mechanisms”, [Online], available on accessed on 20/01/09. Ljungwall, C and J. Li, (2007), “Financial sector development, FDI and Economic growth in China”, [Online], available on China centre for Economic research, Working paper no E2007005 accessed 04/12/08. Maddala, G.S. and S.Wu, (1999). Comparative study of unit root tests with panel data and a new simple test. Oxford Bulletin of Economics and Statistics Vol 61, p 631-652. Mahamet, D, (2006), “Foreign direct investment in Africa: Mali attracts foreign investment”, Journal of Applied Sciences, Vol 6(4), p 972-978. Mankiw, N.G, D. Romer and D.N Weil, (1992), “A contribution to the empirics of economic growth”, The Quarterly Journal of Economics, Vol 107(2), p 407-437. McKinnon, R. I. (1973), Money and Capital in Economic Development. Washington D.C.: Brookings Institution. Mottaleb, K.A, (2007), “Determinants of Foreign Direct Investment and Its Impact on Economic Growth in Developing Countries”, [Online], available on accessed on 16/02/09. Nourzad, F, (2002) “Financial development and productive efficiency: A panel study of developed and developing countries”. Journal of Economics and Finance Vol 26(2), p 138-148. Organisation for economic co-operation and development, OECD, (2002), “Foreign direct investment for development: Maximising benefits, minimising costs”, [Online], available on accessed 09/02/09. Pedroni, P., (1996), “Fully modified OLS for heterogeneous cointegrated panels and the case of purchasing power parity”, Indiana University Working Papers in Economics no 96-020. Pedroni, P., (1999), “Critical values for cointegration tests in heterogeneous panels with multiple regressors”, Oxford Bulletin of Economics and Statistics, Vol 61 (4), p 653-670. Pedroni, P. (2000), “Fully modified OLS for heterogeneous cointegrated panels”. Advances in Econometrics, Vol 15 p 93-130.

15

Quah, D., (1994) Exploiting cross-section variations for unit root inference in dynamic data. Economic letters, Vol 44 p 9-19. Rodrik, D., A. Subramanian and F. Trebbi, (2002), “Institutions Rule: The Primacy of Institutions over Geography and Integration in Economic Development”, NBER Working Paper 9305. Rousseau P.L and P. Wachtel, (2000), “Equity Markets and Growth: Cross country Evidence on Timing and Outcomes, 1980-1995”, Journal of Banking and Finance, Vol 24(12) p 1933-1957. Shaw E.S, (1973), Financial deepening in economic development, New York: Oxford University press. Westerlund, J., (2007), “Testing for error correction in panel data”, Oxford Bulletin of Economics and Statistics, Vol 69, p 709-748. Thorsten, B, A. Demirguc-Kunt and R. Levine, (2000), “A New Database on Financial Development and Structure”, World Bank Economic Review 14, p 597-605. UNCTAD, (2008), World Investment Report 2008, [Online], available on <www.unctad.org/wir> accessed 12/02/09. World Bank, (2008), World Development Indicator (WDI), [Online], available on accessed on 03/03/09. Zhang, K.H., (2001), “Does Foreign Direct Investment Promote Economic Growth? Evidence from East Asia and Latin America”, Contemporary Economic Policy, Vol 19 (2), p 175–85.

16