Abstract

The study investigates the effect of economic, political and social institutions on FDI in Nigeria between 1980 and 2015. The Stock-Watson Dynamic Least Squares (DOLS) is relied upon for analyzing the data extracted from the World Bank WDI. The KPSS stationarity results indicate that all the variables were mixed integrated with FDI and political integration being levels stationary while economic and social factors were first difference stationary. The Johansen cointegration test finds evidence of two cointegrating equations in the model, indicating that long run relationship exists among the variables. The estimated cointegrating regression model showed that economic and social integrations were highly significant in explaining changes in FDI. Political institution factors, on the other hand, have an insignificant positive relationship with FDI. The Wald test showed evidence to support that the coefficients of the regressors were significantly different from zero. The result of the Granger causality test reveals that political and social institution has high predictive power for FDI flows to the Nigerian economy. Based on the findings, this study recommends that Nigeria should adopt a gradual reduction in all economic, political and social constraint to FDI inflows to optimize the benefits of foreign direct investment in Nigeria.

Keywords: FDI, Institutional, factors, Cointegrating, regression,Granger, causality, WDI, and, Nigeria.

Received: 4 June 2018 / Revised: 8 August 2018 / Accepted: 12 September 2018 / Published: 22 October 2018

Contribution/ Originality

This paper is one of very few studies which have investigated the role of institutional quality in mobilizing FDI in capital deficient economies. Its measure of institutional quality from social, economic and political dimensions is noteworthy which add significantly to the existing literature.

1. INTRODUCTION

The determinants of FDI in poor countries have been characterized by controversies as suggested by empirical evidences in existing literature. The sources of these controversies could be traced to the problem associated with collecting appropriate data or using the proper proxies for the variables considered as qualitative in nature. For instance, Asiedu (2002); Anyanwu (2012) and Adeoye (2009) used telephones in measuring infrastructure development, which fails to reliably and efficiently capture the level of infrastructure in the recipient economy. Although telephone lines may be available to the majority of the population, poor transportation network, inadequate power supply and underdeveloped financial system are the major bottlenecks impeding growth and and as such contracts FDI inflow. These factors tend to limit the capacity of institutional factors ability to attract FDI inflows as outlined in Farooque and Yarram (2010).

The principles underlying the importance of institutional quality in mobilizing FDI are attributed to what make up the institution. For instance, good institutional qualities in the form of political stability, security; controlled corruption and enforcement of rule of law in the host country give good signals to foreign investors. Wei (2000) argued that absence of institutional quality due to incidence of corruption, bureaucracy and high levels of extortions tend to limit the ease of doing business both for domestic and foreign investors. Hence, cost of business operation increases due to poor institutional quality. However, non-enforcement of property rights has been identified in literature as one of the sources of risks of expropriation and hence tends to deter investors from taking part in business activities in the recipient FDI countries. It is therefore necessary to take into consideration the issues underpinning the political economy of FDI. The past studies on the impact of institutional quality on FDI flows have been relatively limited in Nigeria, regardless of the huge amount of literature that exist on the predictors FDI.

2. STATEMENT OF THE PROBLEM

High-quality institutional environment has been identified as panacea to FDI inflows. Hence, it is important for the institutional factors to be identified as they help to promote the process globalization in the recipient country. However, it is notable that institutions are borne out of human beings. Therefore, building stronger institutions rather than human beings is the prerequisite for rapid growth and development, especially in mobilizing foreign capital in Nigeria. The institutional quality suggests that Nigeria consciously or otherwise build strong personalities, thus creating a barrier to free flow of foreign investments.

Furthermore, FDI has received a wide range attention as a factor determining the pace of growth in the recipient countries, especially in the poor countries like Nigeria (Khan and Ahmad, 2008). It is theoretically established that institutional quality plays important role in driving the process of growth through inflows of FDI. The nature of these institutions in Nigeria is perceived by economists and non-economists alike as not impressive compared to what is obtainable in some emerging countries. This has continued to remain a source worry to policy makers and potential foreign investors. In view of the relevance of institutional quality in driving the investment decision of foreigners, continuous fear and uncertainty have characterized the poor dimensions of institutional quality considering the extent of FDI in Nigeria. It is based on the above that this study is designed to achieve the specific objectives such as to determine the effect of economic, political and social factors on FDI on FDI Nigeria.

2.1. Theoretical and Empirical Literature

The eclectic theory also known as concept of OLI paradigm was used for this study. This theory seeks to provide better insight on the decision of a firm on locating its investment in foreign countries rather than producing locally and exporting to other countries. Past studies tried to explore relationship between institutional quality and FDI with the findings showing both direct and indirect effects. Globerman and Shapiro (1999) opined that better institutions are good for inward FDI given that they provide enabling environments for multinational companies to profitably invest outside their home countries.

In their study, Stein and Daude (2001) found that their measures of institutional variables are significant in driving inward FDI. Wei (2000) posited that poorly regulated institutions add to the constraints to bilateral FDI flows. However, Campos et al. (1999) are of the opinion that corruption stands as notable bottlenecks to FDI. This agrees with the conclusion of Brunetti et al. (1998). Additionally, Kostevc et al. (2007) contributed by arguing that FDI tends to move to different destinations when property rights are not clearly defined and enforced. Contrarily, some studies tend not to find any empirical evidence to support the claim that institutions and FDI are related positively. For instance, Wheeler and Mody (1992) did not find any strong link between institutions and FDI. Notably, human capital, healthcare, labor force and the quality of public facilities were identified by Globerman and Shapiro (2002) and Mody and Srinivasan (1998) as having indirect effects on FDI.

3. METHDOLOGY

Data required for the study was the time series data which consist of foreign direct investments as dependent variable and the independent variable which includes economic, political and social institutions. Country-specific time series data spanning from 1980-2015 sourced from World Bank World Development Indicators were utilized in this study.

3.1. Estimation Technique/ Method of Data Analysis

The Dynamic Least Squares (DOLS) methodology credited to Stock and Watson (1993) was employed in this study to estimate the cointegrating regression model. The choice of this estimation approach stems from its efficiency to produce robust estimates for both small datasets and relatively large sample. Additionally, the DOLS has the potential of producing optimal estimates of cointegration regressions and corrects for endogeneity usually associated with explanatory variables by adding lags and leads of the first differences of the explanatory variables.

3.1.1. Pre-Estimation Tests

i. Stationarity test: This study specifically employs Kwiatkowski et al. (1992) to conduct the stationarity test to determine whether or not each of the series is stationary. The KPSS has advantage over the Augmented Dickey Fuller approach as the latter tends to be associated with low power and inefficient in the evident of structural break in the series. The null hypothesis of stationarity is tested against the alternative hypothesis of no stationarity at 5 percent level. This general form of the unit root test model with a constant and trend is formalized below:

Where: Zt = Economic time series under consideration, and ci = parameter estimates, f = lag length, ∆= First difference operator and µt = Random disturbance term
ii. Cointegration test: The Johansen Maximum Likelihood (JML) test was adopted in this study to examine whether or not the variables are cointegrated. Specifically, the Trace and Max-Eigen statistics provides basis for rejecting the null hypothesis and vice versa.

3.1.2. Post- Estimation Tests

Aside the pre-estimation tests, some post estimation tests were carried out in the course of the study. Notably, Wald test and normality tests were conducted. The Wald is relied upon to examine if the estimates of the regressors are significant different from zero. The normality test on the other hand was utilized to check if the error terms are normally distributed. The probability value of the Jarque-Bera statistics forms basis for this test. Additionally, the Granger causality test is employed to determine the direction of causality among the variables. The null hypothesis of null causality is tested against the alternative hypothesis of causality at 5 percent level. Causality is established when the null hypothesis is rejected and vice versa. The rejection of the null hypothesis depends on the computed probability value (P-value) of the chi-square distributed statistic. If the p-value is less than or equal to 0.05 the null hypothesis is rejected, but if otherwise the null hypothesis cannot be rejected.

3.2. Model Specification

The functional relationships between the dependent variable and regressors are formalized as:

Where: FDI = Foreign direct investment, ECGO = Economic factor, PLGO = political factor, SCGO = social factor.

Equation (3.2) is expressed as a cointegrating model below:

Where: FDI, ECGO, PLGO and SCGO are as earlier defined in equation (3.2), is the constant term, are long run multipliers which captures long run effects of variations in the regresors on the dependent variable, b and y are the maximum lag and lead lengths respectively, and µt is random error term.
The a priori expectations of the require that: >0, >0.

4. TREND ANALYSIS

The trends of foreign direct investment, economic, political and social factors in Nigeria based on data adapted from notable sources, especially World Bank world development indicators and KOF institutional index are depicted in figures 1 to 4

Figure 1: Trends of FDI in Nigeria, 1980-2015

Source: Author’s estimation based on underlying data adapted from World Bank (2016) World Development Indicators (WDI)

Figure 2: Trends of FDI in Nigeria, 1980-2015

Source: Author’s estimation based on underlying data adapted from KOF (2013, 2014 and 2015) Index of Globalization

Figure-3. Trends of FDI in Nigeria, 1980-2015

Source:Author’s estimation based on underlying data adapted from KOF (2013, 2014 and2015) Index of Globalization*.

Figure-4. Trends of FDI in Nigeria, 1980-2015

Source: Author’s estimation based on underlying data adapted from KOF (2013, 2014 and 2015) Index of Globalization*.

4.1. Unit Root Test

Before estimating the model, the time series characteristics of each of the underlying series is examined via unit root test. Specifically, the approach to unit root test proposed by Kwiatkowski et al. (1992) was adopted to examine the time series properties of the underlying series. The result is report in Table 1 hereunder.

Table-1. Results of the KPSS unit root test on the underlying series

KPSS UNIT ROOT TEST
Variables	Levels			First Difference
	KPSS test stat	Test critical value (5%)	Inference	KPSS test stat	Test critical value (5%)	Inference	Order of integration
Log (FDI)	0.0649	0.146	S	0.1872	0.146	NS	I(0)
Log (ECG0)	0.1817	0.146	NS	0.1023	0.146	S	I(1)
Log (PLG0)	0.1429	0.146	S	0.1397	0.146	S	I(0)
Log (SCG0)	0.2913	0.146	NS	0.1187	0.146	S	I(1)

Source: Authors computation using E-views 9

NB: NS and S respectively imply non-stationary and stationary at levels. I(0) and I(1) denote integrated of order zero and one respectively.

4.2. Johansen Conitegration Test

The Johansen coitegration approach was relied upon for examining if the series are cointegrated or not. The null hypothesis of no cointegration was tested against the alternative hypothesis of cointegration at 5 percent level. Evidence of at least one cointegrating equation provides basis for rejecting the null hypothesis. The result of the cointegration test is presented in Table 2.

Table-2. Johansen Cointegration Test Result

Series: LOG(FDI) LOG(ECGO) LOG(PLGO) LOG(SCGO)
Trace test
Hypothesized	Eigenvalue		Trace		0.05	Prob.**
No. of CE(s)			Statistic		Critical Value
None *	0.615431		70.42499		47.85613	0.0001
At most 1 *	0.552041		39.84477		29.79707	0.0025
At most 2	0.325981		14.14704		15.49471	0.0790
At most 3	0.046483		1.523132		3.841466	0.2171
Max-Eigen test
Hypothesized	Eigenvalue	Max-Eigen		0.05		Prob.**
No. of CE(s)		Statistic		Critical Value
None *	0.615431	30.58022		27.58434		0.0200
At most 1 *	0.552041	25.69773		21.13162		0.0106
At most 2	0.325981	12.62391		14.26460		0.0894
At most 3	0.046483	1.523132		3.841466		0.2171