alexa Does the Inflows of Foreign Aid Dampen or Stimulate FDI to EAC Members? Evidence From East African Community (EAC) Members | Open Access Journals
ISSN: 2162-6359
International Journal of Economics & Management Sciences
Make the best use of Scientific Research and information from our 700+ peer reviewed, Open Access Journals that operates with the help of 50,000+ Editorial Board Members and esteemed reviewers and 1000+ Scientific associations in Medical, Clinical, Pharmaceutical, Engineering, Technology and Management Fields.
Meet Inspiring Speakers and Experts at our 3000+ Global Conferenceseries Events with over 600+ Conferences, 1200+ Symposiums and 1200+ Workshops on
Medical, Pharma, Engineering, Science, Technology and Business

Does the Inflows of Foreign Aid Dampen or Stimulate FDI to EAC Members? Evidence From East African Community (EAC) Members

Masoud Mohammed Albiman*

Department Research Policy and Development, Zanzibar Revenue Board (ZRB) and Zanzibar Insitute of Business Research and Technology (ZIBRET), Tanzania, Zanzibar

*Corresponding Author:
Masoud Mohammed Albiman
Department Research Policy and Development
Zanzibar Revenue Board (ZRB) and
Zanzibar Insitute of Business Reseach
and Technology (ZIBRET), Tanzania, Zanzibar
Tel: +255242233904
E-mail: [email protected]

Received date: December 03, 2014; Accepted date: January 20, 2015; Published date: January 31, 2015

Citation: Albiman MM (2015) Does the Inflows of Foreign Aid Dampen or Stimulate FDI to EAC Members? Evidence From East African Community (EAC) Members. Int J Econ Manag Sci 4:228. doi:10.4172/2162-6359.1000228

Copyright: © 2015 Albiman MM. This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.

Visit for more related articles at International Journal of Economics & Management Sciences

Abstract

Recently, the World Bank and IMF have been more interested in understanding whether foreign aid inflows from multilateral or bilateral donors have catalyzing effect on the crowd out effect to FDI. From this fact, the paper investigates the causal relationship between foreign aid, both bilateral and multilateral aid with FDI inflows to EAC members. Using time analysis techniques of VECM and Granger causality, the results are mixed. On one hand, foreign aid (bilateral and multilateral) causes FDI in two countries. On the other hand, we found no relationship between foreign aid and FDI in other two countries. The results calls for appropriate policy implementations.

Keywords

FDI; EAC; Foreign investments

Introduction

Recently, the World Bank and IMF have been more interested in understanding whether foreign aid inflows from multilateral or bilateral donors have catalyzing effect on the crowd out effect to FDI [1]. Yet, very few studies have empirically examined the impact of foreign aid to FDI to developing countries especially the East African community members (EAC).

For long time foreign aid has become an important source of external finance to the African economics. Foreign aid as an important source of inflows especially in Africa, on average, it accounts for 12.5 percent of total GDP in Africa [2]. On the other hand, recently due to the more economic integration and promotion of open economic policies, FDI has become a second source of external finance. From this fact it is important to look for the relationship of these two variables.

According to several literatures, foreign aid has different impact of economic growth. Firstly it is argued that, foreign aid can improve the domestic infrastructures to the recipient countries [3]. Since the increase in foreign aid would necessitate marginal productivity of domestic capital thus will stimulate more FDI inflows. However, it is also argued that, when foreign aid come in the form of physical capital would increase in capital accumulation in the domestic economy as results crowd out FDI [4,5].

Furthermore, foreign aid can be used to finance budget deficits and profit outflows from FDI, as results will attract more FDI [3]. On the other hand, the increase in foreign aid can increase the supply of tradable goods and results to lower price of goods in domestic economy. From this fact, the profit of FDI would diminish due to low price of goods thus FDI will decrease or crowded out (Some time this act is tremed as “dutch diseases”).

The criteria for selecting to study the EAC members are as follows lies in the argument that, the trend of foreign aid, FDI and their ration to GDP are increasing rapidly since late 1990s (Figures 1 and 2). Furthermore, the EAC members are still in the process of monetary union and political federations by 2015. From this fact, the clear and appropriate policies are needed on the determinants of macroeconomic performance of these countries. Not only that but also according to millennium development goals (MDGs), EAC members considered as poorest countries in LDCs by the World Banks. In that sense, they receive higher aid inflows in the sense that, would help to reduce levels of poverty, budget deficits and improve the social economic well-being by 2015.

economics-management-sciences-trend-Foreign-aid-inflows

Figure 1: The trend of Foreign aid (ODA) inflows to EAC members. (Source: UNCTAD Online database and author’s calculation).

economics-management-sciences-FDI-percentage-GDP-EAC

Figure 2: FDI percentage in GDP for EAC members (Source: UNCTAD Online database, 2010 and author’s calculations).

On the other hand, due to the small private capital inflows like FDI, especially for the LDCs, for long time foreign aid seems to play important role in supporting FDI inflows. It suggested that, foreign aid can play the important role in mitigating the supply constraints, such as poor infrastructure and communications and low capital available for the development program. For example, the key function of the Multilateral Investment Guarantees Agency (MAG) plays an important role to ensure that FDI risks are eliminated [6].

For example, MAG, provide insurance to any risk relating to expropriation, currency, profit transfer restrictions, and political risk or any government actions against private sectors. In fact, the main assumptions here is that, the foreign aid programs would help to remove obstacles relating to the inflows and performance of FDI in developing countries especially African countries. For example, Multilateral aid (MAID) helps to reduce expropriation risk of the government. On the other hand, bilateral aid (BAID) said to reduce the risk of FDI since it acts as good signal of the relationship between recipients and host country.

Literature Reviews

Very few literatures investigated the impact of foreign aid to FDI. This part presents the discussion of the previous empirical literatures. Karakaplan [7] argued that, found that foreign aid do not have any impact to FDI instead, financial development and good governance should be promoted to attract more FDI. Furthermore, Harmz and Lutz [3] supported Karakaplan [7], by suggesting that foreign aid does not have any impact to economic growth. In contrast, Blaise [8] found positive impact of aid to FDI in Japan. However, they pointed that aid has significant impact to FDI, when countries have unfavorable environments or when private investors faces heavy regulatory restrictions. All literatures of [3,7,8] used same proxies of total value of aid and FDI in their analysis.

Furthermore, Kosack and Tobin [9] supported Karakaplan [7] and Harmz and Lutz [3] by suggesting that, foreign aid does not have an impact to marginal productivity in host country since most of aid inflows are in the form of government budget deficits and human support. Simultaneous to that, Caselli and Feyrer [10] argued that foreign aid in most African countries substitute FDI rather than complement it. In supporting these results, Arellano et al. [11] pointed that, foreign aid substitute FDI in the host country.

Moreover, Kimura and To-do [12] after using a gravity model to investigate the impact of aid to FDI in Japan. He found that foreign aid does not attract FDI. This is indicated major difference with the results of Blaise [8] who did the same analysis in Japan.

In general, Kimura and Todo [12] supported the results of Karakaplan [7], Kosack and Tobin [9] and Harmz and Lutz [3]. Furthermore, Changsheng Xu et al. [4] suggested that, foreign aid in the form of human capital and infrastructure causes FDI in Bangladesh, India and Sri Lanka. On the other hand, they found that aid in the form of physical capital causes FDI only in India. On the other side, suggested that FDI causes aid of human capital and infrastructure only for Bangladesh. Generally, they found that aid is complement factor for FDI in five Asian countries.

Methodological Issues

The study uses time series data therefore we have to ensure that all time series technique are met before final conclusion. Therefore, the study used three unit roots test to test whether variables are stationary these tests including Augmented Dickey fuller [13] test (ADF) Philips and Perron. The standard test of ADF and PP test both relied on the null hypothesis that there is unit root, and thus they are not powerful for the alternative hypothesis [14]. From this fact, we use KPSS test which assumes that the null hypothesis of a series are stationary. Thus the KPSS test is suggested to eliminate near unit root process which cannot be detected in ADF and PP test. The method uses to test for Co integration is a Johansen and Juselius approach [15,16]. If variables are cointegrated we use a vector error correction term (VECM) to test for the long run and short run causality.VECM model will be specified in the following form:

equation
equation
equation
equation

For Equation 1 up to 4, INBAID Implies bilateral aid, INMAID implies multilateral aid. For other variables and symbols as previously defined. From the equations above ECMs’ is error correction terms lagged one year period. This help to differentiate between “short run” and “long run “causality. When ECM is negative and significant we conclude that there is long run causality. This is measured through the significance of the t test of ECM. The significance of the lagged changes of all independent variables (α ' s,ϕ ' s,δ ' s,λ ' s ) implies that there is short run causality. This test is performed through the significance of the F-test. However, if the variables are not Co integrated, then the Granger causality test will be conducted in first difference VAR without including the ECM In the equations.

This study used annual data with the spanning from 1970 until 2010. The data collected is as follows, Uganda (1970-2010), Burundi (1970-2006), Kenya (1970-2010)), Rwanda (1976-2010). The data collected from UNCTAD online database key indicators and World’s Bank indicators (WDI). Table 1 provides the descriptions of the variables and data source.

Variable Measurement Descriptions Sources
Economic growth GDP per capita Real GDP which includes domestic productions UNCTAD, World Bank
FDI FDI-GDP ratio Foreign direct inflows UNCTAD, World Bank
Export Export-GDP ratio   Export of goods and services UNCTAD ,World Bank
Domestic investment Capital formation-GDP ratio Outlays or additions of fixed assets in the economy plus net changes in inventories excludes all form of FDI   World Bank
Foreign aid ( multilateral nd bilateral aid) Net ODA –GDP ratio Disbursement of loans and Grants for development activities UNCTAD, World Bank
(UNCTAD refers to the United Nations Conference on trade and development

Table 1: Description of Data and Sources.

Analysis and Discussions

Unit root tests

The results of unit root tests are presented in Tables 2-5. The results are presented in two different forms, of intercept and trend and intercept. The critical value statistics are given in response of MacKinnon values. The PP statistics are obtained through Newey West adjustment of Bartlett Kernel On the other hand, the critical values for Kwiatkowski et al. [14] is given is given by Kwiatkowski et al. [13]. Special attention has been given in the process of lag length selection, so as to ensure the disturbance terms are white noise. In this consideration, we use the Schwarz Criterion method is used to select the appropriate lag length which is selected automatically using Eviews 7.

COUNTRY/VARIABLES ADF TEST PP TEST KPSS TEST
Kenya Constant Constant and trend Constant Constanta and trend Constant Constant and trend
InGDP -1.731524(1) -2.451883(1) -1.674096(1) -2.065574(1) 0.485651*(4) 0.112937(4)
InFDI -4.491753(0)* -4.531971(0)* -4.621104(4)* -4.620772(3)* 0.255018(4) 0.137697**(4)
InAID -1.906711(0) -1.793426(0) -1.994703(3) -1.888663(3) 0.175152(5) 0.141573**(5)
InEXP -2.470181(0) -2.708465 -2.556325(5) -2.819109(3) 0.274447(4) 0.104256**(4)
InLDI -1.298025(0) -1.312531(0) -1.275661(2) -1.3254981(2) 0.413098**(5) 0.168133*(5)
LPOP -1.843587(4) -2.573801(4) 0.597785(4) -2.062498(4) 0.706347*(5) 0.175043*(5)
Uganda  
InGDP -1.149078(4)* -4.551953(1)* -2.815235(9)* -3.319285(21)* 0.376751**(5) 0.200312*(5)
InFDI -8.391391(0)* -8.369213(0)* -8.747674(3)* -8.726228(3)* 0.551435*(5) 0.155731*(4)
InAID -0.727849(0) 1.019541(0) -0.727849(0) 1.336052(1) 0.173010(1) 0.170296*(2)
InEXP -6.070900(0)* -5.168852(3)* -6.322324(13)* -14.09970(38)* 0.2509285(4) 0.209309*(4)
InLDI -8.005927(0)* -8.360548(0)* -8.171282(2)* -8.360548(0)* 0.647906*(5) 0.176764(4)
LPOP -1.232883(0) -1.714150(0) -1.232883(0) -1.765305(1) 0.639866*(5) 0.126875**(5)
Notes: Asteriks* and ** implies a significant level at five percentage (5%) and ten percentage (10%). Values in ( ) implies the t statistics level. ADF implies Augmented Dickey Fuller [13] and Critical values obtained in response of McKinnon. PP implies Philips and Perron. KPSS implies Kwiatkowski et al. test. The critical values of KPSS found in Kwiakowski et al. [14].

Table 2: Unit Root Test Results in Level Form.

COUNTRY/VARIABLES ADF TEST PP TEST KPSS TEST
Rwanda Constant Constant and trend Constant Constanta and trend Constant Constant and trend
InGDPC -1.944762(0) -2.054124(0) -1.867600(1) -1.979785(1) 0.193793(4) 0.161714*(4)
InFDI -4.074611*(0) -4.037325*(0) -4.200330(3)* -4.177382*(3) 0.189170(4) 0.151584*(4)
LBAID -2.760526(0)** -2.900308(0) -2.752944(1) -2.911429(1) 0.241348(4) 0.094483(3)
LMAID -0.082679(0) -1.395741(0) -0.082679(0) -1.056673(2) 0.521105*(1) 0.113882(1)
InEXP -2.264078(0) -0.755180(2) -3.144161*(2) -4.032150*(2) 0.274447(4) 0.104256(4)
InLDI -3.142918(0)* -3.976006*(0) -3.144161(2)* -4.032150*(2) 0.166015(4) 0.167591*(4)
Burundi  
InFDIGDP -5.162049(0)* -5.162049(0)* -5.236411(3)* -6.179674(3)* 0.513015*(4) 0.132708**(3)
InGDPC -0.836574(0) -1.465642(0) -1.064946(3) -1.493653(1) 0.170201*(5) 0.273348(2)
LBAID -0.821502(7) -1.995938(7) -1.276133(3) -2.143425(3) 0.497656*(4) 0.061438(4)
LMAID -2.254999(1) -3.077111(1) -1.748873(0) -2.276275(0) 0.479080*(4) 0.059486*(4)
InEXP 4.193280(0)* -4.781145(0)* -4.282792(3)* -4.768487(1)* 0.424638**(4) 0.085762(0)
InLDI -2.354820(0) -2.310767(0) -2.154292(1) -2.100319(1) 0.146934(4) 0.149821*(4)
Notes: Asteriks* and ** implies a significant level at five percentage (5%) and ten percentage (10%).Values in ( ) implies the t statistics level. ADF implies Augmented Dickey Fuller [13] and Critical values obtained in response of McKinnon (1999). PP implies Philips and Perron (1988). KPSS implies Kwiatkowski et al. [14] test. The critical values of KPSS found in Kwiakowski et al. [14].

Table 3: Unit Root Test Results in Level Form.

COUNTRY/VARIABLES ADF TEST PP TEST KPSS TEST
Kenya         Constant Constant and trend
InFDI -7.640860(0* -7.539890(1)* -11.48543(4)* -12.00144(5)* 0.068201(4) 0.044010(2)
InGDP -4.116716(0)* -4.046209(0)* -3.985447(4)* -3.901202(4)* 0.148200(0) 0.148200(0)
LBAID -3.688012*(1) -3.649542*(1) -7.456828*(4) -7.391203*(4)    
LMAID -8.684503*(0) -8.623060*(0) -8.684503*(0) -8.680327*(1) 0.085435(3) 0.060293(3)
InLDI -6.543565(0)* -5.032387(1)* -6.550380(1)* -6.793171(4)* 0.163232(0) 0.129231(0)
LEXP            
Uganda  
InFDI -8.391391(0)* -8.369213(0)* -8.747674(3)* -8.726228(3)* 0.177549(5) 0.088330(6)
InGDPC -1.149078(4)* -4.551953(1)* -2.815235(9)* -3.319285(21)* 0.654680(4) 0.114284(4)
LBAID -3.985545*(3) -3.953470*(3) -5.540887*(3) -5.463748*(3) 0.082933(1) 0.084072(1)
LMAID -6.956248*(0) -6.978830*(0) -6.959008*(2) -6.982372*(1) 0.168527(2) 0.047265(0)
InEXP -6.070900(0)* -5.168852(3)* -6.322324(13)* -14.09970(38)* 0.426540(11) 0.500000(39)
InLDI -8.005927(0)* -8.360548(0)* -8.171282(2)* -8.360548(0)* 0.279799*(1) 0.138764(2)
Notes: Asteriks* and ** implies significant level at five percentage (5%) and ten percentage (10%).Values in ( ) implies the t statistics level. ADF implies Augmented Dickey Fueller [13] and Critical values obtained in response of McKinnon. PP implies Philips and Perron. KPSS implies Kwiatkowski et al. [14] test. The critical values of KPSS found in Kwiakowski et al.[14].

Table 4: Unit Root Test Results after First Difference.

COUNTRY/VARIABLES ADF TEST PP TEST KPSS TEST
Rwanda Constant Constant and trend Constant Constant and trend Constant Constant and trend
InFDI -10.24709(0)* -10.12138(0)* -12.50651(5)* -14.38733(7)* 0.142941(7) 0.124823(8)
InGDPC -7.009409(0)* -5.668109(1)* -7.381535(7)* -7.90344*(12) 0.188740(10) 0.160761(14)
LBAID -6.826276*(0) -6.718878*(0) -7.257424*(5) -7.129368*(5) 0.241348(4) 0.094483(3)
LMAID -1.542577(0) -1.486560(0) -1.469814(1) -1.357543(1)    
InEXP -5.702638(1)* -6.525035(1)* -7.712599*(0) -18.8233*(21) 0.054959(2) 0.048817(2)
InLDI -9.826658(0)* -9.796309(0)* -11.72239(5)* -14.94311(7)* 0.165855(4) 0.130902(5)
Burundi  
InGDPC -5.294493(0)* -5.433210(0)* -5.318789(2)* -5.433210(0)* 0.273348(2) 0.083974(0)
InFDI -7.340173(1)* -7.280233(1)* -18.59025(8)* -20.29324(7)* 0.053057(3) 0.053057(3)
LBAID -6.083343*(6) -5.937659*(6) -5.541868*(3) -5.515516*(2) 0.091274(4) 0.068571(4)
LMAID -4.745933*(1) -4.680378*(0) -4.645795*(0) -4.568240*(0) 0.054166(3) 0.053749(3)
InEXP -9.612456*(0) -9.459687*(0) -12.14575*(6) -11.90680*(6) 0.167416(8) 0.147186(9)
InLDI -9.543565(0)* -9.3211662(0)* -0.895686(3)* -8.754315(3)* 0.164485(3) 0.145432(3)
Notes: Asteriks* and ** implies significant level at five percentage (5%) and ten percentage (10%).Values in ( ) implies the t statistics level. ADF implies Augmented Dickey
Fueller and Critical values obtained in response of McKinnon. PP implies Philips and Perron. KPSS implies Kwiatkowski et al. test. The critical values of KPSS found in
Kwiakowski et al.

Table 5: Unit Root Test Results after First Differences.

Each test among these three tests has advantages and disadvantages at different circumstance. For example the standard test of ADF and PP test fails to detect the structural break that is common in time series moving average. From this fact, we assume that variables are not stationary at level or my (0), once it is accepted for all three tests in both with intercept and intercept with the trend. Otherwise, for any conflicting results we assume that variables is not stationary at levels I (0), hence we continue with first difference “The same reasons has been given to justify that variables are stationary after first difference by several studies and used Johansen and Juselius co integration method (See, Das and Choudhary, Malik)”.

In any conflicting results, we also assume that, variables are not stationary at levels, because most of the macroeconomic variables were found to have a unit root problem at their level form. The results in Tables 2 and 3 present the results of unit root test in level form. After considering all the three tests, in both conditions of intercept with and without trend, it is concluded that that, all variables are not stationary in level form. On the other hand, Tables 4 and 5 presents results of unit root after first difference.

Multivariate cointegration analysis

Before, estimating the long run relationship between Foreign aid and FDI we have to ensure that all variables in the system have Co movement towards long run equilibrium. The optimal lags selected we rely on two information criteria Akaikes and Schwarz criteria. The lag selected for each country are in brackets, Rwanda (2), Uganda(1), Burundi(2) and Kenya (1). The Co integration test was estimated in the system of six variables, namely, FDI (LFDI), economic growth (LGDP), bilateral aid (LBAID), Multilateral aid (LMAID). For six variables included, the maximum Co integrating relationship is utmost five.

From Table 6 (Rwanda), the results show that there are, one Co integrates vectors in case of maximum Eigen test (n-r) =1, whilst Trace test indicates two Co integrating vectors (n-r) =2. On the other hand, from Tables 7 (Uganda) both tests show four Co integrating vectors (n-r) =4). Furthermore, from Tables 8 (Burundi) and 9 (Kenya) both tests of max Eigenvalue and Trace test suggest one Co integrating vectors (n-r) =1. In general both tests suggest that there is a long run relationship for all variables in FDI model. This shows that, all variables included in the model are important determinants of FDI inflows in the long run. On the other hand, data indicate that, the models are far from spurious regression.

Rwanda (2)
Variables: LFDI   LGDP   LBAID  LMAID  LEXP   LDI
Ho: Maximum Eigen value 95% critical value Trace test 95%critical value
r=0 57.91033* 40.07757 132.5956* 95.75366
r ≤1 30.86456 33.87687 74.68530* 69.81889
r ≤2 23.95607 27.58434 43.82074 47.85613
r ≤3 10.88805 21.13162 19.86467 29.79707
r ≤4 6.846837 14.26460 8.976625 15.49471
r ≤5 2.129788 3.841466 2.129788 3.841466

Table 6: Johansen Multivariate Cointegration Test Results.

Uganda (3)
Variables:           LFDI   LGDP   LBAID  LMAID  LEXP   LDI
Ho: Maximum Eigen value 95% critical value Trace test 95%critical value
r=0 114.6813* 40.07757 256.1311* 95.75366
r ≤1 57.87435* 33.87687 141.4498* 69.81889
r ≤2 36.28126* 27.58434 83.57546* 47.85613
r ≤3 33.77517* 21.13162 47.29420* 29.79707
r ≤4 13.48425 14.26460 13.51903 15.49471
r ≤5 0.034784 3.841466 0.034784 3.8414661)
Notes: LGDP (economic growth), LFDI (FDI), LBAID (bilateral aid), LMAID (multilateral aid), LEXP (export), LDI (Domestic investment), LPOP (labor force). * and ** refers to significant at 5 and 10 percent level respectively. Number in ( ) implies the optimum lag selected by both AIC and Schwarz Criterion

Table 7: Johansen Multivariate Cointegration Test Results.

Burundi (2)
Variables:           LFDI   LGDP   LBAID  LMAID  LEXP   LDI
Ho: Maximum Eigen value 95% critical value Trace test 95%critical value
r=0 113.6289* 40.07757 182.1067* 95.75366
r ≤1 25.62332 33.87687 68.47783 69.81889
r ≤2 19.34886 27.58434 42.85451 47.85613
r ≤3 14.65725 21.13162 23.50564 29.79707
r ≤4 8.835404 14.26460 8.848396 15.49471
r ≤5 0.012992 3.841466 0.012992 3.841466
Notes: LGDP (economic growth), LFDI (FDI), LBAID (bilateral aid), LMAID (multilateral aid), LEXP (export), LDI (Domestic investment), LPOP (labor force). * and ** refers to significant at 5 and 10 percent level respectively. Number in ( ) implies the optimum lag selected by both AIC and Schwarz Criterion

Table 8: Johansen Multivariate Cointegration Test Results.

KENYA (1)
Variables:           LFDI   LGDP   LBAID  LMAID  LEXP   LDI
Ho: Maximum Eigen value 95% critical value Trace test 95%critical value
r=0 30.40181 40.07757 93.82097** 95.75366
r ≤1 25.34509 33.87687 63.41916 69.81889
r ≤2 20.11653 27.58434 38.07407 47.85613
r ≤3 8.738821 21.13162 17.95754 29.79707
r ≤4 7.248945 14.26460 9.218719 15.49471
r ≤5 1.969774 3.841466 1.969774 3.841466
Notes: LGDP (economic growth), LFDI (FDI), LBAID (bilateral aid), LMAID (multilateral aid), LEXP (export), LDI (Domestic investment), LPOP (labor force). * and ** refers to significant at 5 and 10 percent level respectively. Number in ( ) implies the optimum lag selected by both AIC and Schwarz Criterion.

Table 9: Johansen Multivariate Cointegration Test Results.

Vector error correction model (vecm) and granger causality

As we have said earlier within VECM model we can estimate the long run and short run relationship among the variables. After having found the long run relationship between FDI and Foreign aid now, here again we consider the short run interaction among the bilateral aid, multilateral and FDI inflows.

However, we include other important variables like economic growth, domestic investment and human capital. The lag was selected using AIC and Schwarz Criterion. The results are presented in Tables 10-13. The results of error correction terms (ECTs) shown by the t statistics value whilst Granger causality is shown by Chi square statistics. For the diagnostic test most of the results show that our models are free from serial correlation and normally distributed.

In the case of Kenya, in Table 10 results show that, error correction term of bilateral aid (LBAID) and economic growth (LGDP) are significant at the five percent level of significance. This implies that, they make the adjustment towards long run equilibrium, after any short run deviations. For remaining variables do not make adjustment for any shock in the model. The results show that, at the five percent level of significance, bilateral aid (LBAID) Granger cause exports (LEXP). On the other hand, bilateral aid (LBAID) causes multilateral aid (LMAID) at ten percent level of significance. We also found that, economic growth granger causes exports.

Dependent variables Independent variables
X2 statistics of lagged first differenced term (p values)
ECTt-1
[ t-ratio]
  ΔLFDI ΔLGDP ΔLBAID ΔLMAID ΔLEXP ΔLDI  
ΔLFDI - 0.37(0.53) 0.22( 0.63) 0.60(0.43) 0.86(0.35) 0.68(0.40) 0.05
(0.36)
ΔLGDP 0.28(0.59) - 0.90(0.34) 0.11(0.73) 6.25*(0.01) 0.40(0.52) 0.00*
(3.6)
ΔLBAID 0.21(0.64) 1.95(0.16) - 3.39**(0.06) 1.38(0.23) 0.07(0.78) -0.05*
(-2.33)
ΔLMAID 2.55(0.10) 6.64*(0.00) 0.34(0.55) - 0.54(0.45) 0.30(0.58) 0.01
(0.16)
ΔLEXP 0.45(0.49) 6.36*(0.01) 0.06* (0.80) 0.02(0.86) - 1.19(0.27) -0.02
(-1.59)
ΔLDI 1.94(0.16) 1.81(0.17) 0.00*(0.98) 0.02(0.86) 0.17(0.67) - -0.05
(-1.80)
GDP MODEL,  AR(1)=86.84, JB=13, FDI MODEL,  AR(1)= 86.84, JB=14, EXP MODEL, AR(1)=86.84, JB=19, AIDMODEL, AR(1)=86.84, JB =12.44, LPOP MODEL, AR(1)=86.84, JB =12.44, LDI MODEL, AR(1)=86.84, JB=83.30
Notes: Asteriks * and ** refer to significant levels at five (5) and ten (10) significant level. Numbers in ( ) and [ ] refers to the chi square value and t statistics respectively. AR(1) refers to autocorrelation test at order one.JB refers to Jaque Berra normality test.

Table 10: Results of Vector Error Correction Model (VECM) (Kenya).

In the case of Rwanda (Table 11) the error correction term of multilateral aid (LMAID) and domestic investment (LDI) and export are significant at the five percent level of significance. Any short run deviations from the long run, those variables which are significant will make a correction in the next period and the model will return to the equilibrium. In this model, we found that, all variables including bilateral and multilateral aid granger cause FDI. Multilateral aid (LMAID) Granger causes export and economic growth. We also found that, domestic investment (LDI) granger cause bilateral id (LBAID).

Dependent variables Independent variables X2
statistics of lagged first differenced term (p values)
ECTt-1
[ t-ratio]
  ΔLFDI ΔLGDP ΔLBAID ΔLMAID ΔLEXP ΔLDI  
ΔLFDI - 0.80
(0.66)
2.44
(0.29)
1.79
(0.40)
0.10
(0.40)
3.85
(0.14)
0.88
(1.05)
ΔLGDP 11.27
(0.00)*
- 3.11
(0.21)
5.2
(0.07)
0.01
(0.99)
5.88
(0.05)**
0.09
(1.61)
ΔLBAID 6.93
(0.03)*
0.96
(0.61)
- 2.89
(0.23)
1.14
(0.56)
5.07
(0.07)**
  -0.36
(-1.68)
ΔLMAID 7.96
(0.01)*
4.92
(0.08)**
1.49
(0.47)
- 5.51
(0.00)*
1.86
(0.39)
4.97*
(3.11)
ΔLEXP 4.57
(0.10)*
0.17
(0.91)
0.66
(0.71)
3.38
(0.18)
- 1.86
(0.39)
-0.23
(-1.79)
ΔLDI 7.87
(0.01)*
6.09
(0.04)
9.69
(0.00)*
3.04
(0.21)
1.43
(0.4)
-   0.33*
(2.47)
GDP MODEL,  AR(1)=75.84, JB=13, FDI MODEL,  AR(1)= 75.84, JB=19, EXP MODEL, AR(1)=75.84, JB=29, AIDMODEL, AR(1)=75.84, JB =22.44, LPOP MODEL, AR(1)=75.84, JB =22.44, LDI MODEL, AR(1)=75.84, JB=43.30
Notes: Asteriks * and ** refers to significant level at five (5) and ten (10) significant level. Numbers in ( ) and [ ] refers to chi square value and t statistics respectively. AR (1) refers to autocorrelation test at order one. JB refers to Jaque bera normality test.

Table 11: Results of Vector Error Correction Model (VECM) (Rwanda).

On top of that, Uganda (Table 12) economic model shows that, error correction terms of the normalized variables (LFDI) are significant implies the presence of long run causality from all variables in the system including bilateral (LBAID) and multilateral aid (LMAID). Furthermore, the error correction term of economic growth (LGDP) and bilateral aid (LBAID) are significant at five and ten percent level of significance level. This implies that, all variables make adjustment to the long run, for any shocks in the short run. The results show that, only bilateral aid (LBAID) causes economic growth (LGDP).

Dependent variables Independent variables
X2statistics of lagged first differenced term (p values)
ECTt-1
[ t-ratio]
  ΔLFDI ΔLGDP ΔLBAID ΔLMAID ΔLEXP ΔLDI  
ΔLFDI - 1.10
(0.77)
7.66*
(0.05)
5.56
(0.13)
6.59**
(0.08)
16.5*
(0.0)
0.09
(0.69)
ΔLGDP 0.15
(0.98)
- 5.20
(0.15)
3.65
(0.30)
0.71
(0.86)
19.2
(0.00)
0.01*
(1.29)
ΔLBAID 2.28
(0.51)
6.34**
(0.09)
- 14.7*
(0.00)
0.11
(0.98)
7.13*
(0.06)
0.37*
(3.37)
ΔLMAID 2.73
(0.43)
4.68
(0.19)
8.15*
(0.04)
- 2.48
(0.47)
10.1*
(0.01)
-0.10
(-0.91)
ΔLEXP 1.47
(0.68)
1.79
(0.61)
6.19
(0.10)
2.29
(0.51)
- 1.79
(0.61)
-0.07
(-0.07)
ΔLDI 1.02
(0.79)
1.32
(0.73)
3.20
(0.36)
5.79
(0.12)
2.57
(0.46)
- -0.00
(-0.19)
GDP MODEL,  AR(1)=65.94, JB=14, FDI MODEL,  AR(1)= 65.94, JB=39, EXP MODEL, AR(1)=65.84, JB=39, AIDMODEL, AR(1)=65.94, JB =22.44, LPOP MODEL, AR(1)=65.94, JB =22.22, LDI MODEL, AR(1)=65.84, JB=53.30
Asteriks.* and ** implies five (5) and ten(10) significant level. Numbers in ( ) and [ ] refers to chi square value and t statistics respectively. AR (1) refers to autocorrelation test at order one.JB refers to Jaque bera normality test.

Table 12: Results of Vector Error Correction Model (VECM) (Uganda).

On the other hand, we found that, multilateral aid (LMAID), foreign direct investment (LFDI) both Granger cause bilateral aid (LBAID). In turn, multilateral aid (LMAID) Granger causes bilateral aid (LBAID). This indicates that, both bilateral and multilateral aid, in the short run; they are dependent on their performance (Table 13). On the other hand, the results show that FDI (LFDI) causes export growth in the short run.

Dependent variables Independent variables
X2 statistics of lagged first differenced term (p values)
ECTt-1
[ t-ratio]
  ΔLFDI ΔLGDP ΔLBAID ΔLMAID ΔLEXP ΔLDI  
ΔLFDI - 0.30(0.85) 3.17(0.20) 0.01(0.99) 0.77(0.67) 14.0*(0.0) -0.48*
[-2.52]
ΔLGDP 3.49(0.17) - 7.68*(0.02) 1.38(0.49) 6.06*(0.04) 32.4*(0.0) 0.00
[0.35]
ΔLBAID 0.80(0.66) 0.19(0.90) - 3.95(0.13) 8.65*(0.01) 6.0*(0.04) 0.15*
[2.26]
ΔLMAID 3.28(0.19) 1.36(0.50) 1.63(0.44) - 3.80(0.14) 14.8*(0.0) -0.12
[-1.06]
ΔLEXP 0.33(0.84) 0.10(0.94) 3.03(0.21) 0.77(0.67) - 4.49(0.10) -0.03
[-0.55]
ΔLDI 0.33(0.84) 1.69(0.42) 3.86(0.14) 0.39(0.82) 6.01*(0.04) - 0.29*
[5.57]
GDP MODEL,  AR(1)=75.84, JB=33, FDI MODEL,  AR(1)= 75.84, JB=79, EXP MODEL, AR(1)=75.84, JB=29, AIDMODEL, AR(1)=75.84, JB =42.55, LPOP MODEL, AR(1)=75.84, JB =72.44, LDI MODEL, AR(1)=75.84, JB=7.30
Notes: Asteriks * and ** refer to significant levels at five (5) and ten (10) significant level. Numbers in ( ) and [ ] refers to the chi square value and t statistics respectively. AR(1) refers to autocorrelation test at order one. JB refers to Jaque bera normality test.

Table 13: Results of Vector Error Correction Model (VECM) (Burundi).

On the other hand, in Burundi, the results of error correction term show that, FDI (LFDI), bilateral aid (LBAID) and domestic investment (LDI) significantly make adjustments in equilibrium for any short run deviations. On the other hand, in the short run we found that, economic growth (LGDP), Granger cause bilateral aid (LBAID). Moreover, economic growth (LGDP), bilateral aid (LBAID), domestic investment (LDI) Granger causes exports. Furthermore, in the short run FDI, export, bilateral and multilateral aid Granger causes domestic investment (LDI).

Generally, for all countries we found that bilateral aid and multilateral aid cause FDI in two countries out of four namely Burundi and Rwanda. These results are supported by Changesheng Xu et al. [4] and Blaise [8]. This implies that the increase in foreign aid would stimulate more FDI through alleviating the constraints of FDI such as poor infrastructures and increasing more access of finance. On the other hand we found that, FDI granger cause bilateral aid in Rwanda which implies that the increase in FDI would stimulate more familiarity between host countries and donors which results to more FDI inflows [17].

On the other hand we found that foreign aid both bilateral and multilateral aid do not support FDI in Kenya and Uganda. These results conform to the results of Karakaplan [7], Harmz and Lutz [3]; Kimura and Todo [12]. This implies that, the increase in foreign aid in these countries would not support FDI inflows; instead the government has to rely on alternative factors to increase FDI. In general the results are summarized in Table 14.

Country FDI  Bilateral Aid Multilateral Aid Control Varaibles
                           FOREIGN DIRECT INVESTMENT MODEL ( FDI )
Kenya No causality Cause multilateral aid and export No causality GDP causes export
Burundi Cause LDI Cause LFDI, LDI Cause LFDI, LDI Export, LDI  cause FDI
Rwanda Cause bilateral aid Cause FDI  Cause FDI, Export and economic growth( LGDP)  LDI cause LBAID
Uganda Cause export Cause multilateral aid Cause bilateral aid LGDP cause LBAID

Table 14: Summary of Causalitry Results.

Conclusion

Recently, the World Bank and IMF have been more interested in understanding whether foreign aid inflows from multilateral or bilateral donors have catalyzing effect on the crowd out effect to FDI [1]. Yet, very few studies have empirically examined the impact of foreign aid to FDI to developing countries especially the East African community members (EAC). From this fact, the main intention of this research is to analyze the role of foreign aid to FDI inflows in EAC members.

Several literatures found mixed results concerning the impact of foreign aid to FDI. However at large extent, they found that, foreign aid does not have any impact to economic growth. Given that, FDI and foreign aid inflows are the large external source of EAC members it is worthwhile to determine their relationship especially in terms of causality. Using the time series techniques of vector error correction term (ECM) and Granger causality test we found mixed results for all EAC members.

Generally, for all countries we found that bilateral aid and multilateral aid cause FDI in two countries out of four namely Burundi and Rwanda. This implies that the future FDI inflows in Burundi and Rwanda largely depend on foreign aid inflows. On the other hand we found that, FDI granger cause bilateral aid in Rwanda which implies that the increase in FDI would stimulate more familiarity between host countries and donors which results to more FDI inflows. On the other hand we found that foreign aid both bilateral and multilateral aid do not support FDI in Kenya and Uganda. This implies that, the increase in foreign aid in Kenya and Uganda would not support FDI inflows; instead the government has to rely on alternative factors to increase FDI.

In the future studies, it will be important for the researchers to analyze the relationship between foreign aid and FDI in different branches such as looking the impact of human support aid and physical capital aid to FDI.

References

Select your language of interest to view the total content in your interested language
Post your comment

Share This Article

Relevant Topics

Article Usage

  • Total views: 11695
  • [From(publication date):
    February-2015 - Aug 22, 2017]
  • Breakdown by view type
  • HTML page views : 7880
  • PDF downloads :3815
 

Post your comment

captcha   Reload  Can't read the image? click here to refresh

Peer Reviewed Journals
 
Make the best use of Scientific Research and information from our 700 + peer reviewed, Open Access Journals
International Conferences 2017-18
 
Meet Inspiring Speakers and Experts at our 3000+ Global Annual Meetings

Contact Us

 
© 2008-2017 OMICS International - Open Access Publisher. Best viewed in Mozilla Firefox | Google Chrome | Above IE 7.0 version
adwords