Institutional quality and inward FDI: empirical evidence from GCC economies

Abstract

ABSTRACTThis paper studies the determinants of inward foreign direct investment (FDI) across the Gulf Corporation Council (GCC) countries between 2009 and 2017 using a modified knowledge-capital model of the multinational firm. In particular, we investigate the importance of institutional quality factors. We document the significant effects of institutional characteristics such as government effectiveness, control of corruption, political stability and rule of law, while regulatory quality was found to be less important. Moreover, our study finds that the GCC countries’ FDI can be explained by horizontal market seeking rather than efficiency-seeking vertical motives. Finally, the extended specification results highlight the significant effects of colonial relationships, common language and contiguity.KEYWORDS: Foreign direct investmentGCC regioninstitutional qualitypseudo-poisson maximum likelihoodknowledge capital modelJEL CLASSIFICATION: F23 AcknowledgmentThe authors would like to thank two anonymous reviewers for their constructive comments and suggestions. We are also gratefully acknowledging Oleg Gurshev in the Department of Economic Sciences at the University of Warsaw for reading and commenting on the previous drafts of the paper.Disclosure statementNo potential conflict of interest was reported by the author(s).Notes1 They argue that the GCC region features a lower institutional quality relative to the other regions around the world and suffers from flawed institutions at numerous levels, including the malfunctioning of public administration, shortages in political rights, and ineffective laws and regulations.2 These reforms incorporated incentives such as tax waivers, relaxed restrictions on foreign ownership, and removal of trade barriers.3 The GCC countries have signed over 200 BITs, where Kuwait is the leader followed by the UAE and Qatar.4 The examples include Méon and Sekkat (Citation2004), Mina (Citation2007), Aziz and Mishra (Citation2016), and Aziz (Citation2018).5 Their seminal models have been extended by, inter alia, Horstmann and Markusen (Citation1987), Markusen & Venables (Citation1998; Citation2000), Helpman et al. (Citation2004), Cieślik and Ryan (Citation2012).6 In the horizontal type of FDI, firms face the trade-off between maximizing proximity to households and concentrating production to achieve economies of scale. In contrast, the vertical type of FDI is related to countries’ differences in relative factor endowments. Thus, the decision of firms to engage in horizontal FDI would be driven by the size and growth of the host country, whereas vertical FDI seeks cost competitiveness and other factors such as quality of domestic institutions, political risk, and physical infrastructure.7 The GCC countries include Bahrain, Kuwait, Oman, Qatar, Saudi Arabia, and the United Arab Emirates (UAE).8 Our bilateral data covers Saudi Arabia and the United Arab Emirates due to nearly all of the GCC countries do not report outward FDI statistics (e.g. Bahrain, Oman, Qatar and Kuwait). Our final bilateral dataset has 5037 observations, 3031 (65%) of which constitute ‘zero’ FDI.9 The complete list of investment partners is shown in Table A2 in the Appendix.10 The following facts dictate the use of FDI stock data. First, FDI stock is significantly more across-the-board and readily available for a large economy group than affiliate sales. Second, unlike flow data (net FDI), which is essentially a sum of equity and debt instruments, stock data holds only the reported accurate capital equity owned by foreign affiliates in a given host economy. Besides, FDI stock data represent ‘the long-run factors that explain the distribution of FDI’ (Blonigen & Piger, Citation2014, p. 782). Finally, the extant empirical literature has taken advantage of the use of stock data, (e.g. Awokuse et al., Citation2012; Blonigen & Piger, Citation2014; Camarero et al., Citation2019; Cieślik et al., Citation2021).11 These observations are omitted from the estimation as we rely on the logarithmic transformation.12 The detailed definitions of our explanatory variables are provided in Table A3 while pairwise correlations are reported in Table A4 reports in the Appendix.13 In Table A5 in the Appendix, we use the average years of schooling as an alternative measure of human capital. The average number of years spent in school is a commonly used measure of a population’s education level. It allows aggregation of attainment across education levels and allows an analysis of the ‘stock of human capital’ that a population has at any given point in time. The data refer to Our World in Data based on Lee and Lee (Citation2016).14 The six governance indicators are based on 31 underlying data sources reporting the governance perceptions of a large number of survey respondents worldwide. Details on the underlying data sources, the aggregation method, and the interpretation of the indicators can be found in Kaufmann et al. (Citation2009; Citation2010).15 These include bilateral investment treaties (BIT), contiguity (Contig), common spoken language (ComLang), and common colonizer post-1945 (ComCol). These variables are widely used in the FDI literature (e.g. Bergstrand & Egger, Citation2013; Blonigen & Piger, Citation2014; Nguyen et al., Citation2020).16 The PPML has several desirable properties. First, it assists in mitigating Jensen’s inequality, which is the following: E[g(X)]≠ln[E(y)]. One of the implications of Jensen’s inequality is that when we take the logarithm of the explanatory variables, lower values and disinvestment values in the data values automatically become zero. In the empirical analysis, the OLS estimators cannot read zero values that may generate inconsistent and biased estimation results. It also increases the heteroscedasticity in the empirical analysis that E[yi|x]=exp⁡(xiβ)∝V[yi|x], where βs can be estimated by solving the following set of first-order conditions: ∑i=1n⁡[yi−exp⁡(xiβ~)]xi=0.17 This result differs from existing studies, which typically find a negative relationship between FDI and the interaction between BIT and colonial past such as Gurshev and Hamza (Citation2021).18 The RESET test in our estimations is higher than in previous studies such as Nguyen et al. (Citation2020).