Exports as a Mediator Variable Between Foreign Direct Investment Inflows and GDP in Low and Low-Middle Income African Countries

Table of Contents

Abstract

1. Introductıon

2. Literature Review
2.1. Definition and Previous Findings
2.2. Previous Methodology

3. Methodology

4. Research Results
4.1. Descriptive statistics
4.2. Correlation Analysis
4.3. Mediation Analysis Result

5. Result Discussion and Conclusion
5.1. Results Interpretation

6. Conclusion

References

Abstract

The relationship between foreign direct investment (FDI) inflows, exports and economic growth as measured by gross domestic product (GDP) has been a global interest of academics and policy-makers, but research methods did not allow the characterization of the indirect mediating effects that exports have on that relationship. Therefore, this study aims identifying the mediation effect of export in the relationship between FDI and GDP in low and middle low-income African countries. The study uses correlation analysis, Baron and Kenny method, Bootstrap procedure and Sobel test to investigate the significance of the indirect effect. The result of the analysis shows a partial mediation of exports in the relationship between FDI and GDP. The study demonstrates the indirect effect caused by FDI through export. It is therefore recommended that low and middle low-income African countries should stimulate foreign direct investment to boost their exports, and gross domestic product. Additionally, these countries should find new ways of financing exports as FDI are predicted to fall due to the Covid-19 pandemic during 2020.

Key Words: Foreign Direct Investment Inflows, Exports, Gross Domestic Product, Mediation, low and middle low-income African countries.

1. Introductıon

The analysis of the relationship between FDI exports and economic growth as measured by GDP has attracted academic world and policy-makers. Nevertheless, little is known about the way FDI transfers effects on GDP through EXP. Some researchers found a positive impact of FDI on GDP, while others found a negative impact. For instance, Rubens (1999) indicated that FDI in Africa can contribute significantly to the economic development of the continent. George, James and Peter (2013) recognized that FDI is an essential in developing countries, as FDI increases resources available for investment and capital formation, facilitates transfers of technology, increases skills and innovative capacity, and enhances organizational and managerial practices. Joseph (2015) used correlation and regression analysis to find relationship between GDP per capita and FDI in Rwanda, and found a strong positive relationship between the FDI inflows, and the GDP per capita. Christie (2018) analyzed the long-run positive relationship between foreign agricultural investment and economic growth in Sub-Saharan Africa, and found a positive link between foreign agricultural investment and economic growth in the long-run. Arafatur and Summit (2015) analyzed the causal relationship between FDI and GDP in Bangladesh. The cointegration test confirmed the existence of a long-run equilibrium relationship between the two variables. The use of a Granger causality test showed unidirectional causality which runs from FDI to GDP. Nadeem, Naveed, Zeeshan and Sonia (2013) analyzed the relationship between FDI and GDP in Pakistan. Using correlation analysis and a regression model, they found a positive relationship between FDI and GDP. Qaiser, Salman, Ali, Hafiz, and Muhammad (2011) analyzed the impact of FDI on GDP of SAARC countries from 2001 to 2010. The result of multiple regressions showed a positive and significant relationship between GDP and FDI while an insignificant relationship was found between GDP and inflation.

However, other studies showed the negative side of FDI in the host country. Suzana, Zoran, and Sinisa (2015) found very few positive effects of FDI in Croatia, where, economic performance did not improve, in the period from 1999 to 2014. Anne and Luis (2008) concluded that the neo-liberal practices such as those enforced in developing countries like Colombia, while seeking to attract foreign investment to push their economies, tend to produce a false aggregated demand growth that is not sustainable in the long-term, increases global unemployment, unleashes destructive competitive processes and weaken government’s ability to regulate business in the citizens` best interests. Obi, Anthony and Felicia (2020) using data from 1981 to 2018 concluded that FDI had an insignificant positive impact on industrial growth, and an insignificant negative impact on domestic investment in Nigeria. David (2013) recognized that FDI indirectly does not always have a good effect on the host-country. However, he joined other researchers and affirmed that any international investment should bring development benefits to the receiving country regarding technology transfer, employment creation, and provides upstream and downstream linkages that these investments are to be “win-win” rather than “neo-colonialism”. Moreover, beneficial flows are not automatic, and care must be taken in the formulation of investment contracts and selection of business model (David, 2013, p.xiii). Nunnenkamp (2001) indicated that fiscal and financial incentives offered to foreign investors may do more harm than good by giving rise to costly bidding wars. In addition, policymakers should not expect too much from FDI inflows. Rui and Rosa (2009, p.7-15) showed the risks associated with the potential advantages that can be brought by foreign direct investment. It has been said that foreign direct investment can make transfer of technology. Rui and Rosa (2009) showed that the risk may result in the fact that host country can become dependent on technologies introduced by multinationals and other developed countries. They added there is a decline in local firms’ interest in the production of new technologies. Multinationals may have an adverse reaction to host country research to continue to hold a technological advantage compared to local firms. They argued that multinationals only transfer inappropriate technologies, and the host country dependence from multinationals technology will be perpetuated.

The purpose of the study is to identify the mediation effect of export in the relationship between FDI and GDP in 28 low and low middle African countries on one hand. In the other hand, the study investigates whether exports (EXP) mediate the relationship between FDI and GDP, and tests the significant indirect effect. The main question to be answered is: what are the direct and indirect effects produced by FDI on GDP in low and middle low-income African countries? This question was not addressed by researchers. They focused on direct effect produced by FDI on GDP. The following literature review attempts to present studies that showed impact on FDI and exports on economic growth.

2. Literature Review

2.1. Definition and Previous Findings

The World Bank (2020) provides indicators and defines each indicator. The following variables are defined according to World Bank. Gross domestic product (GDP) (constant 2010 US$): GDP at purchaser's prices is the sum of gross value added by all resident producers in the economy plus any product taxes and reducing possible subsidies not included in the value of the products. It is calculated without reducing depreciation of produced assets or for depletion and degradation of natural resources. Foreign direct investment net inflows (current US$) (FDI): is the sum of equity capital, reinvestment of earnings, and other capital. Direct investment refers to a category of cross-border investment linked with a resident in one economy, having control or a significant degree of influence on the management of an enterprise that is resident in another economy. Having at least 10 percent of the ordinary shares of voting stock is the criterion for determining the existence of a direct investment relationship. Exports of goods and services (current US$) (EXP): is the value of all goods and other market services provided to the rest of the world. They do not consider the value of merchandise, freight, insurance, transport, travel, royalties, license fees, and other services, such as communication, construction, financial, information, business, personal, and the government services. Compensation of employees and investment income and transfer payments are excluded.

Many studies were conducted to analyze the impact of FDI on GDP. Many methods were used to identify the relationship between both variables. The findings showed a positive impact of FDI on GDP.

The Organization for Economic Co-operation and Development “OECD” (2002) indicated that FDI triggers technology spillovers, assists human capital formation, contributes to international trade integration, helps improve environmental and social conditions, helps create a more competitive business environment, leading to more socially responsible corporate policies and enhances enterprise development. In addition, all of these contribute to higher economic growth, which is the most potent tool for alleviating poverty in developing countries. An economy is an interdependent system where one action on one variable may have spillovers and cyclic impacts on other variables. For instance, an increase in the unemployment rate may create a vicious circle by increasing the dependency ratio and reducing household income, thus reducing both demand and government tax income, as well as reducing corporate investment and job creation and finally increasing the unemployment rate again.

Assiobo and Fang (n.d.) made analysis in Togo using Granger-causality to test and determine the causal relationship between FDI and Economic Growth. The result of his analysis concluded on the existence of a unidirectional relationship between FDI and GDP. Obi, Agu and Ezebest (2020) used Augmented Dickey-Fuller (ADF) and Philips-Perron unit root tests, Engle-Granger two-step approach and error correction mechanism to investigate the impact of FDI on domestic investment in Nigeria. Data collected cover the period time from 1981 to 2018. They found that FDI had an insignificant positive impact on industrial growth in Nigeria. However, research conducted by Muhia (2019) in Kenya found that FDI oriented in infrastructure sector impact positively, and significantly economic growth even though the impact is insignificant on manufacturing, and agriculture sectors. Olawumi and Olufemi (2016) analyzed the impact of FDI in randomly selected African countries from the period 1980 to 2013. Data were analyzed using ordinary least squares and generalized methods of moments. They noticed a limited or negligible impact of FDI on African countries economic growth. The impact is different from a country to another. Their results show that an increase at one percent in FDI result in .12% increase in GDP in South Africa, .05% increase in Egypt, .03% in Nigeria, and .02% in Kenya. Patrick, Emmanuel and Edmond (2013) used Johansen cointegration analysis to investigate the impact of foreign trade in Ghana. They found a long run and short run relationship among real GDP, FDI, exports, imports and FDI in Ghana. Additionally, they concluded that in the overall, export had a positive effect on real GDP, and an increase in exports leads to an improvement in real GDP. Ogbokor and Meyer (2017) investigated the impact of foreign trade on economic performance using the economy of South Africa. They used vector autoregression (VAR) as method. They found that export contributes more towards economic performance compared to openness of the economy and exchange rates. Mukhtarov, Alalawneh, Ibadov and Huseynli (2019) on contrary of the little impact found a positive, and statistically significant impact of FDI on export in the long-run in Jordan. They show that 1% increase in FDI increases exports by 0.13%. Same conclusion has been reached by Nasir, Kushtrim and Luljeta (2016). These authors investigated the impact of FDI on western Balkan Countries. Export and FDI data were collected for the period from 1996 to 2013. Using panel regression techniques and least square dummy variable (LSDV) regression method, they concluded that FDI have positive effect on export performance in the sample countries in various model specifications.

Topi (2011) showed how GDP can be calculated using three different ways. It can be obtained by value added (or production) approach, difference between gross output of different industries and intermediate inputs, to avoid double counting. Using income (by type) approach, it is the sum of all income earned by different factors of production. Finally the GDP can be determined by final demand (or expenditures) approach, which measures the activities, such as investment and consumption across different industries and imports deducted from exports. Using the final demand approach, the equation become: GDP = C + I + G + (X – M) where C is consumption of final goods and services by the households, I is investment in things such as plants, equipment and software, G is government expenditures on goods and services, X is exports and M is imports.

The choice of Exports as mediator variable is because in low and lower-middle income countries, where domestic consumption, government spending and investment are still weak, exports play a crucial role. Exports expand the market for domestic goods financed by FDI, and help a country to get international currencies. In addition, earning generated by exports increase capital accumulation, similar to the FDI.

Mamingi and Martin (2018) affirmed that even though FDI positively affects growth, its impact is minimal when considered in isolation. They confirmed that the significant effect of FDI is rather indirect than direct. The report of the United Nations (2001) enumerated supporting arguments that exports may promote economic growth by generating a greater capacity utilization; taking advantage of economies of scale; bringing about technological progress; creating employment and increasing labour productivity; improving allocation of scarce resources throughout the economy; relaxing the current account pressures for foreign capital goods by increasing the country’s external earnings and attracting foreign investment; and by increasing the well-being of the country.

The reason the reason of omitting other variables like imports (IMP), gross capital formation (GCF), house hold consumption (HHC), and governance expenses (GE) as used in GDP equation, is to avoid the problem of multicollinearity in the regression of the step 3 explained before (GDP = b03 + M + 'FDI + e3), according to Baron and Kenny procedure. Those variables are highly correlated to FDI. The correlation between FDI and IMP is 0.7, FDI and GCF is 0.8, FDI and HHC is 0.6, FDI and GE is 0.6.

2.2. Previous Methodology

There is variety of methods used by previous researchers according to the objective of the study. Some methods used are multiple regression models (Abbas, Akbar, Nasir, Aman & Naseem, 2011), correlation and regression model (Nadeem, Naveed, Zeeshan & Sonia, 2013), cointegration analysis (Nosheen, 2013), simple regression (Tamilselvan, & Manikandan, 2015), generalized least squares estimator (Khan &Mehboob 2014), dynamic panel VECM technique (Christie,2018), system generalized method of moment (SYS-GMM) estimators (Steve,2014), ordinary least squares regressions and cointegration tests (Tsatsaridis, 2017). All these methods used allow analyzing direct effect of FDI on GDP only. However, little is known about the indirect effects of FDI on GDP. Not analyzing the existence of indirect effect that FDI may have on GDP through a third variable may lead to a biased conclusion. One variable may have effect on the other variable trough a third variable, without this there may be no direct effect. In the present study, the way FDI influences GDP through Exports is investigated.

Moderation analysis allows analyzing the extent to which an independent variable can impact a dependent variable through a third variable. Borau, Akremi, Gambier, Kidar and Ranchoux (2018) explained that mediation effect relates to the mechanism through which an independent variable X has an impact on a dependent variable Y via an intermediary variable M located between X and Y. In other word mediation indicates that the effect of an independent variable (X) is transmitted to a dependent variable (Y) through a third variable considered a mediator (M) (Pardo & Román, 2013). As result mediation analysis provides a better understanding of the relationship between variables.

3. Methodology

Secondary data related to 28 low and low-middle income African countries were collected directly from World Bank web site for the period 2017. Those countries have gross national income under 3 975 $. Countries included are Burundi, Central African Republic, Congo Democratic Republic, Niger, Liberia, Malawi, Mozambique, Sierra Leone, Togo, Guinea-Bissau, Madagascar, Burkina Faso, Ethiopia, Uganda, Chad, Rwanda, Mali, Benin, Guinea, Gambia, Comoros, Tanzania, Kenya, Senegal, Lesotho, Cameroon, Cote d'Ivoire, Mauritania. Low income countries have gross national income per capita equals to 1005 $ or less. Lower-middle-income countries have gross national income per capita between $1,006 and $3,975.

This paper adopted mediation approach to analyze the relationship between FDI and GDP. Borau, Akremi, Gambier, Kidar and Ranchoux (2018) explained that mediation effect relates to the mechanism through which an independent variable X has an impact on a dependent variable Y via an intermediary variable M located between X and Y. In other word mediation indicates that the effect of an independent variable (X) is transmitted to a dependent variable (Y) through a third variable considered a mediator (M) (Pardo & Román, 2013). Baron and Kenny’s method has not only been the most widely used method in the last years to demonstrate mediation in social and health sciences, but it is also very possibly the most used method to test mediation (Pardo & Román, 2013). There are four steps in Baron and Kenny’s method as explained by (Pardo & Román, 2013): variables X and Y must be related, as consequence, coefficient c in Figure 1 must be different to zero in the expected direction. Variables FDI and M must be related, this implies, coefficient a from figure 1 must be different to zero. Variables M and Y must be related once the effect of X is controlled, it means, coefficient b from figure 1 must be different to zero. The relationship between X and Y must be significantly reduced when controlling the effect of M. A result, coefficient c’ (direct effect in figure 1) must be smaller than coefficient c (total effect in Figure 1). Tingley, Yamamoto, Hirose, Keele and Imai (2014) indicated that the standard procedure for analyzing causal mechanisms in applied research is called mediation analysis, where a set of linear regression models are fitted and then the estimates of “mediation effects” are computed from the fitted models. This implies that this study will use linear regression in the process of identifying relationship between FDI and GDP. To analyze relationship between FDI and GDP, causal steps procedures recommended by Baron and Kenny (1986) will be used as presented in the previous paragraph.

Two main methods are generally used to test the mediation effects. The first most influential and widely used is the approach presented by Baron and Kenny (1986). The second is a resampling method which is based on the bootstrap resampling procedure, and performs better than the Baron and Kenny method in small sample size studies (Omokri, Agbedey and Nwajei, 2018). The following figure shows the research conceptual analysis.

Abbildung in dieser Leseprobe nicht enthalten

Figure 1: Mediation path analysis (source: Author’s computation)

To decide on the type of mediation, first the relationship between FDI and GDP must be significantly reduced when controlling the effect of M. This is, coefficient c’ (direct effect in Figure 1) must be smaller than coefficient c. In case of c coefficient is non-significantly different from zero, results indicate a full mediation. If coefficient in path b is significant after controlling for the direct effect of X (path c), but path c is still significant, the model indicates a partial mediation (Omokri, Agbedey and Nwajei, 2018).

4. Research Results

This chapter collects data to analyze the relationship between FDI and GDP in 28 low and low middle African countries on one hand. Data collected allow investigating whether exports (EXP) mediate the relationship between FDI and GDP, and test the significant indirect effect. The chapter aims to answer the following: what are the direct and indirect effects produced by FDI on GDP in low and middle low-income African countries? The chapter presents a descriptive analysis of the collected data, then analysis correlation between variables, and ends with the result of the regression analysis.

4.1. Descriptive statistics

Before analyzing mediation, the paper begins with descriptive analysis to investigate the characteristics of the collected data. It gives an idea of the data distribution. The minimum value, the mean value, the maximum value, and the standard deviation are given in the table 1.

Table 1: Descriptive analysis

Abbildung in dieser Leseprobe nicht enthalten

Source: Author’s computation

4.2. Correlation Analysis

Correlation analysis is essential tool to identify relationship between two variables. Correlation helps to identify the strength of the association, and the direction of the relationship between two variables (Zaid, 2015). The following table shows the degree and the direction of the correlation between the variables under study.

Table 2: Correlation matrix

Abbildung in dieser Leseprobe nicht enthalten

Source: Author’s computation

The correlation coefficients, and correlation test between GDP and EXP, are significant. As result, there is sufficient evidence to conclude that there is a significant linear relationship between GDP and EXP because the correlation coefficient is significantly different from zero. The Pearson's product-moment correlation test show a p-value < 5% (0.00000005343). In other words there is a significant linear relationship between EXP and GDP. The analysis of the correlation between GDP and FDI shows a high correlation between the variables. The Pearson's product-moment correlation test rejects the null hypothesis of zero correlation with a p-value under 5%, (0.0001137). The correlation between FDI and EXP is significantly different from zero. The Pearson's product-moment correlation test rejects the null hypothesis of no correlation between FDI and EXP with p-value under 5% (0.01364). There is a linear relation between the variables GDP, FDI, and EXP.

[...]