Do the sector-orientation and the origin of foreign direct investment explain types of corruption in Sub-Saharan Africa?

Abstract

This paper focuses on analysing the effects of sector orientation and the origin of FDI on the types of corruption. The GMM technique and the IV-2SLS on a panel of Sub-Saharan African countries were used. Globally, we find that FDI increases political corruption in SSA. Moreover, the study shows that sector orientation and the origin of FDI have significant effects on each type of corruption. Primary sector FDI enhances public sector corruption and reduces judicial corruption. Secondary sector FDI increases executive, public, and judicial corruption. In the tertiary sector, executive and public corruption are enhanced while judiciary corruption is reduced. Regarding the origin, FDI from France increases public sector corruption whereas FDI from China and the USA reduces it. Executive corruption is reduced by FDI, regardless of its origin. FDI from France and the USA discourages judiciary corruption while FDI from China increases it. FDI from China reduces legislative corruption, while FDI from the USA tends to increase it. Furthermore, the analysis of the transmission channels shows education and development levels as important channels through which FDI could reduce corruption in SSA. Relevant policy implications derived from this study include the necessity for policy-makers to combat all types of corruption and mostly public corruption.

KEYWORDS:

• Origin of FDI
• sector of FDI
• types of corruption
• GMM
• IV-2SLS

JEL Classifications:

• F21
• D73
• C36

Notes

1 The CPI is an index provided by the Transparency international on the different rate of perceptions of corruption that exist in a wide range of countries both developed and developing. The index ranges from 0 (high corruption) to 100 (low corruption).

2 Recall that the CPI ranges from 0 to 100. Lowest scores are related to very high corruption perceptions.

3 We use the log values of GDP per capita. Population is measured by the population growth rate as defined in Appendix.

4 Recall that there are at least five sources of endogeneity: (a) the simultaneity that appears when a dependent variable (Y) is explained by an independent variable (X) and vice versa i.e. when X causes Y and Y causes X; (b) measurement errors on the independent variables and/or the dependent variable; (c) the omission of relevant explanatory variables in the model specification; (d) presence of auto correlation in time series; and (e) Sampling bias.

5 We use the 2SLS estimation when estimating the effect of sectors orientation FDI on types of corruption. The main justification is due to the sample size dimension and data availability. One condition to apply the GMM is that the number of individuals (countries) N should be greater than the time period (T). In our case, sector-orientation FDI data in SSA are available for 12 countries which is less than the temporal dimension (18 years). This condition, however, does not apply for the 2SLS reason why this method was found appropriate for these regressions.

6 The instrumental variable is a variable that is correlated with the endogeneity source variable, but not correlated with the error term and the dependent variable.

7 The effect of corruption on population size is indirect through GDP growth.

8 Too many instruments can severely weaken and bias the Hansen over-identifying restrictions test and therefore, the rule of thumb is that the number of instruments should be less than the number of countries (Roodman 2009).

9 The threshold in this case is obtained by dividing ${\alpha }_2/{\alpha }_4$ (see Asongu, Le Roux, and Biekpe (2017) for a more description on this methodology).