Determinants of FDI in developed and developing countries: a quantitative analysis using GMM

Neha Saini (Faculty of Management Studies, University of Delhi, New Delhi, India)
Monica Singhania (Faculty of Management Studies, University of Delhi, New Delhi, India)

Journal of Economic Studies

ISSN: 0144-3585

Publication date: 14 May 2018

Abstract

Purpose

The purpose of this paper is to investigate the potential determinants of FDI, in developed and developing countries.

Design/methodology/approach

This paper investigates FDI determinants based on panel data analysis using static and dynamic modeling for 20 countries (11 developed and 9 developing), over the period 2004-2013. For static model estimations, Hausman (1978) test indicates the applicability of fixed effect/random effect, while generalized moments of methods (GMM) (dynamic model) is used to capture endogeneity and unobserved heterogeneity.

Findings

The outcome across different countries depicts diverse results. In developed countries, FDI seeks policy-related determinants (GDP growth, trade openness, and freedom index), and in developing country FDI showed positive association for economic determinants (gross fixed capital formulation (GFCF), trade openness, and efficiency variables).

Research limitations/implications

The destination of FDI is limited to 20 countries in the present paper. The indicator of the institutional environment, namely economic freedom index, used in this paper has received some criticism in calculations.

Practical implications

The paper enlists recommendations for future FDI policies and may assist government in providing a tactical framework for skill development, thereby increasing manufacturing growth rate. The paper also throws light on vertical and horizontal capital inflows considering resource, strategy, and market-seeking FDI.

Social implications

FDI may bring significant benefits by creating high-quality jobs, introducing modern production and management practices. It highlights how multinational corporations and government contribute to better working conditions in host countries.

Originality/value

The paper uncovers important features like macroeconomic variables, especially country-wise efficiency scores, policy variables, GFCF, and freedom index, for determining FDI inflows in 20 countries using panel data methods and provides a roadmap for developed and developing countries. The study highlights endogeneity and unobserved heteroscedasticity by applying GMM one- and two-step procedure.

Keywords

Citation

Saini, N. and Singhania, M. (2018), "Determinants of FDI in developed and developing countries: a quantitative analysis using GMM", Journal of Economic Studies, Vol. 45 No. 2, pp. 348-382. https://doi.org/10.1108/JES-07-2016-0138

Download as .RIS

Publisher

:

Emerald Publishing Limited

Copyright © 2018, Emerald Publishing Limited


1. Introduction

Foreign capital flows influence the growth process of both developed and developing countries. Developed countries require foreign capital inflows for sustainable development, while developing countries need it for growth and investment purpose. Implicitly, FDI is an attractive area to explore in the context of both developed and developing countries. There are several views from traditional to modern theory that explains the importance of FDI inflows. Traditional view explains FDI as a movement of capital with respect to differences in returns among countries. Neoclassical economists view foreign capital inflows as a resource for economic development and openness toward other economies. In addition, it enhances the prospects of increased standard of living (Levin, 2001). While modern theory elucidates FDI not only destined for transfer of capital, it also supplies various forms of international sponsorship to a local firm consisting transfer of proprietary and intangible assets including technology, business techniques, and skilled personal Johanson and Mattsson (2015). The upsurge of FDI leads to high (financial) growth and capital formation in host countries. FDI has witnessed dramatic growth in developed and developing countries since the 1990s (UNCTAD, 2014) and this may be credited to macro, national, and firm-level performance as displayed in Figure 1. At the global level, organizations such as WTO and IMF are working to increase investment inflow by reducing restrictions on fund transfer. Similarly at the national level, governments of different countries are designing their policies to attract more capital inflows to lift their economic growth. At the firm level, FDI is the source of higher order technology and innovation.

Dunning and Lundan (2008) developed the theory answering the question “why do companies invest abroad,” and stated “OLI” paradigm. According to this theory, FDI is observed if ownership-specific advantages (“O”), such as proprietary technology, exist with locational advantages (“L”), such as low factor costs, and potential benefit from internationalization (“I”) of production process abroad. The objective of FDI inflow in the host country may be categorized under different motives. The trade effect of FDI depends upon whether it is undertaken to gain access to natural resources, consumer markets, or aimed at exploiting locational comparative advantage and other strategic assets such as research and development capabilities. In literature, a large number of studies have appeared on determinants of FDI but no such consensus has been emerged on “true determinants of FDI” (Kok and Acikgoz Ersoy, 2009). Greene and Villanueva (1991), Singhania and Gupta (2011), Azam and Lukman (2010) and Miskinis and Juozenaite (2015) explored a diverse set of indicators but they are sensitive to specific conditions and locations, resulting in lack of robustness in results. For example, labor cost, tax regimes, R&D, GDP have been found to have both negative and positive effects of FDI depending upon the economic and political environment of the host country. The established body of literature postulates potential determinants of FDI and categorizes them as economic, political, and institutional factors. Economic factors comprising trade openness, exchange rate, infrastructure to market size have often been studied (Masuku and Dlamini, 2009; Leitão and Faustino, 2010; Alam and Zulfiqar Ali Shah, 2013). Political factors include political stability, government effectiveness, regulatory quality and level of corruption (Asiedu and Lien, 2004; Mina, 2009; Qian et al., 2010). Another important determinant may include institutional factors. Institutional factors have a rich collection of inter-reliant structures and systems within a country (Bevan et al., 2004; Asiedu, 2006; Henisz and Swaminathan, 2008). An appropriate measure of defining the institutional environment is the level of “economic freedom” enjoyed by investors in the host country.

The objective of the present study is to identify the potential determinants of FDI classified as economic, political, and institutional measures in both sets of economies. Further economic factors may be divided into three parts, namely market-seeking, efficiency-seeking/resource-seeking, and strategy-seeking FDI. Country-level efficiency is another variable that is considered as an important determinant in attracting FDI. Many study considered wage rate, output-input ratio, human capital, and technological gap as the measure of country-level efficiency and positioned efficiency either under resource-seeking (Bilgili et al., 2012; Goswami and Saikia, 2012; Alam and Zulfiqar Ali Shah, 2013) or efficiency-seeking (De Vita and Kyaw, 2008; Freckleton et al., 2012) FDI. It may also be sited as strategic consideration for lifting FDI inflow.

The paper relates to a panel of 20 countries (11 developed and 9 developing countries) and adopted rigorous econometric models to explain determinants of FDI Inflows. The study uses fixed/random effect (static models) and generalized moments of methods (GMM) one-step/two-step (dynamic models) procedure to estimate results. The rest of the paper is organized as follows: Section 2 presents a theoretical framework on determinants of FDI. Section 3 covers literature review and Section 4 describes research design. Section 5 presents the empirical analysis and findings of the study. Section 6 presents summary and concluding thoughts. Last section explains policy and managerial implications

2. Theoretical framework

The motives and drivers of FDI inflow changes over time but priority toward growth and development remains the same for developed/ developing countries. There are three theories of FDI explaining the theoretical base of this paper. It covers journey from neoclassical theory to industrial theory of FDI with spillover effects. Neoclassical theory, dependency theory, and industrialization theory are explained as follows.

2.1 Neoclassical theory

As per neoclassical theory, FDI contributes positively toward economic development of host country by increasing its well-being status. FDI leads to capital formation in host country, thereby influencing reinvestment of profits and further inflows of capital therein. Infusion of foreign capital makes lower balance of payment and provides higher order techniques of production by replacing unproductive methods. Kojima (1978, 1982), Bergten et al. (1978), Fischer (1998, 2003), Goldar (2004), Dwivedi (2012) valued FDI rewards to host countries as technology spillover, higher managerial skills, and marketing information skills.

2.2 Dependency theory

Developing countries are well endowed with natural resources and they need innovative techniques to maximize their output. Dependency theory tries to bridge this technological gap. During the 1970s, many East Asian and Latin American countries followed this principle but later these countries had to shift from dependency principle of stringent strategies to liberal policies for more capital inflows, as this theory proved unhealthy for the development of emerging countries leading to shift from closed economy to open economy (Cardoso and Faletto, 1979; Dixon and Boswell, 1996; Anoruo and Mustafa, 2007; Hein, 1992).

2.3 Industrialization theory and spillover effects

It considers FDI as transfer of “package” including capital, management, new technology, and characterized as an international extension of industrial organization theory. FDI infuses contagion effect in host country through adaption of management practices and advanced technology. It is a channel that promotes growth by technology transmission from parent firm of a multinational corporation to its subsidiary abroad (Findlay, 1978; Das, 1987; Wang and Blomstrom, 1992; Ambos et al., 2006; Behera et al., 2012; Xu et al., 2014).

3. Literature review

This section offers discussion on determinants of FDI inflow based on past evidences. FDI has fueled extensive literature because it increases the growth prospects in an economy. Alfaro et al. (2008) related economic development with foreign capital inflows and institutional quality. Many studies recognize relationship between foreign capital inflows and economic growth (Moran, 1998; Thiam, 2006; Alam et al., 2013; Majumder and Nag, 2015). Neumayer and Spess (2005) emphasized on quality of domestic institution, enforceability of law, and emphasized on bilateral investment treaties (BITs) of developing countries to have higher foreign direct inflows. BITs attract more FDI as they became the “most important international legal mechanism for encouragement and governance” of FDI (Elkins et al., 2006).

Developing countries are gaining attractiveness since the beginning of the twenty-first century due to higher institutional quality, infrastructural development, national resource availability, and presence of semi-skilled/skilled labor force. Various pull and push factors attract FDI inflows to developed and developing countries. Real GDP growth, per capita income, domestic inflation, commercial interest rates, trade openness, exchange rate, and external indebtedness play a significant role in shaping the trends of foreign capital inflows. Determinants of FDI depend on country-specific factors and hence studied by many researcher (such as Greene and Villanueva, 1991 studied seven developing countries; Singhania and Gupta, 2011 studied India; Azam and Lukman, 2010 studied India, Indonesia, and Pakistan; Miskinis and Juozenaite, 2015 studied Greece, Ireland, and the Netherlands; Rangkakulnuwat and Paweenawat, 2015 studied ASEAN countries).

The presence of FDI inflow increases income and total factor productivity (TFP) growth (Li and Liu, 2005; Woo, 2009) in a country because FDI is considered as the source of technological diffusion which in turn leads to economic growth and higher standard of living (Findlay, 1978; Borensztein et al., 1998; Chang and Luh, 2000; Xiaming et al., 2001; Zhang, 2002; Apergis et al., 2008). To get the benefit of technological diffusion, a country must be advanced enough to absorb innovative technology. Ng (2007) advocated that advance countries absorb technological up-gradation quickly as compared to less technology savvy countries. Recent research has paid increased attention to the institutional environment and the role of government in attracting FDI. Globerman and Shapiro (2003), Brewer (1993) used a number of factors that constitute the governance infrastructure of a country to explain patterns in FDI. There are number of studies available that talk about determinants of FDI but no such study has come up with the true determinants of FDI (ÇEviŞ and Camurdan, 2007; Kok and Acikgoz Ersoy, 2009). The inflow of FDI depends on so many factors which may be significant for a country but insignificant for other. While considering determinants, literature also postulates country-wise determinants including various pull factor (Onyeiwu and Shrestha, 2004; Jadhav, 2012; de Castro et al., 2013; Bokpin et al., 2015; Masron, 2017). Existing literature is enriched with macroeconomic, political, and legal factors but various determinants including productivity, economic freedom, and other sets need to be studied separately to regulate FDI inflow in developed and developing countries. The detailed chronological analysis of the literature of determinants as per country-specific studies is presented under Table I.

3.1 Research gap

Sambharya and Rasheed (2015) and Ghazal and Zulkhibri (2015) incorporated institutional factors as one of the determinant of FDI. Jadhav (2012) considered corruption and the rule of law as a proxy for institutional factors and political stability as one of the determinants of FDI. His study used the variable as “ex post facto proxy of institutional and political reform” rather than as a standalone predictor. “Economic freedom” in our study is used to capture institutional and political environment reforms. We incorporate institutional framework listed in Heritage foundations by taking the average of all individual institutional factors, namely corruption in government, barriers to international trade, income tax and corporate tax rates, government expenditures, rule of law and the ability to enforce contracts, regulatory burdens, banking restrictions, labor regulations, and black-market activities. Since all the individual institutional factors itemized in heritage database may intricately be related to each other, it is important to capture their impact collectively as one institutional variable and thereby eliminate the possibility of multi-collinearity. This, to the best of our knowledge, has not been attempted in literature of FDI determinants till date. As far as productivity factors are concerned, Piteli (2010) studied the impact of TFP and considers it as one of the important determinant for attracting FDI in selected OECD countries. He considers AMECO, 2005 database which is limited to 28 OECD countries indicating research beyond 28 countries with TFP variable is a constraint. Li and Liu (2005) enlightened absorptive capacity of foreign technology and defined technological gap as the difference of GDPUS (per capita) and GDPother (per capita), that is Technological gap = GDPUS−GDPother, considering GDPUS as a benchmark for the technically upgraded country. Extension to previous approaches, we constructed productivity variables based on output-input ratio using Fare et al. (1994) methodology, a non-parametric programming method. The three efficiency indicators, technical change (techch), TFP change (tfpch), and efficiency change (effch), are used to measure country-level efficiency. This approach may be functional for every set of the country having input and output variables. The use of three productivity-linked variables, to the best of our knowledge, is being attempted for the first time in the context of analyzing FDI. On methodological grounds, Hsiao and Hsiao (2006) suggested the superiority of panel data estimations over time series. Few studies (Aziz and Mishra, 2015; Bokpin et al., 2015) have highlighted endogeneity issue in this area of research. This paper enhances existing literature by unraveling the puzzle of endogeneity considering institutional framework, efficiency scores, and other macroeconomic variables.

4. Research design

This section provides discussion about empirical model including estimation, data sources, unit selection, and construction of variables.

4.1 Research methodology

Static and dynamic panel data modeling is used for determining inflows of foreign capital. In static panel data, estimation is undertaken using fixed and random effects. Majumder and Nag (2015) specified total capital inflows as voluminous, more volatile, and more persistent than net inflows. In such backdrop, the use of lagged independent variable adds dynamic nature to analysis. Inclusion of lagged dependent variable yields the model dynamic and leads to a problem of endogeneity, as it becomes correlated with differenced error terms, least square estimation in this case provides bias and inconsistent results (Baltagi, 2008). Arellano and Bond (1991) recommended the usage of instrumental variable including the lag of dependent and independent variables (García-Herrero et al., 2009; Athanasoglou et al., 2008), as a result differenced GMM is applied in this paper to incorporate endogeneity issue

4.1.1 Static panel data specifications

In general, the models for determinants of FDIs using static modeling may be studied using the following specifications:

(1) FDI i t = c + j = 1 J β i X i t j + k = 1 K β k Y i t k + l = 1 L β l Z i t l + e i t   e i t = υ i + u i t
where X, Y and Z are different vectors of pull and push determinants. This equation represents the static nature of model.

Four different models using static panel data approach with FDI as a dependent variable and a set of different independent variables are considered. In this section, developed and developing countries are segregated into groups. The specifications of each model are as follows:

  1. Model 1: FDI = f(GDP, gross fixed capital formulation (GFCF), Freedom index, Trade openness).

  2. Model 2: FDI = f(GDP, GFCF, Freedom index, Trade openness, Technical change, TFP change, Efficiency change).

  3. Model 3: FDI = f(GDP, GFCF, Freedom index, Trade openness, Technical change, Interest rate differential, TFP change, Efficiency change).

  4. Model 4: FDI = f(GDP, GFCF, Freedom index, Trade openness, Technical change, Interest rate differential, TFP change, Efficiency change, Crisis dummy).

The estimated coefficients and their corresponding significant results using static panel data approach are shown in Tables III and IV.

4.1.1.1 Model specification tests

Following Jensen (2003) and Ahlquist (2006), we define dependent variable as a net FDI expressed as a percentage of GDP. The process of scaling by GDP makes the series stationary. Levin et al. (2002) panel unit root test has been performed and all variables rejected the null hypothesis of the common unit root.

The paper adopts several approaches to model FDI inflows. The first method employed is pooled OLS estimation. Breusch-Pegan multiplier test rejects the null hypothesis of using pooled OLS estimation over random effects. Since LM test rejects the null hypothesis of homoscedasticity, we employ white heteroscedasticity consistent standard errors[1] tests.

Subsequently static panel models are applied. Fixed effect panel models are likely to be superior on the theoretical ground, as they control for time-invariant heterogeneity across countries and provide robust results to omit variable biasness (Hausman and Taylor, 1981). However, Baltagi (2008) referred suitability of fixed effect models where the focus is provided on a specific set of entities, while random effect model is more appropriate when inferences are based on entities randomly drawn from a large sample. Breush-Pegan Test suggests the applicability of random effect model over pooled OLS, while F-test specifies the applicability of fixed effect model over pooled OLS. The applicability of fixed effect and random effect model may be specified by Hausman Specification test[2]. Panel causality tests signify the bidirectional causality among variables especially in case of technical scores and FDI inflows. To capture endogeneity and biasness, GMM is used in this paper.

4.1.1.2 Endogeneity and biasness of static model estimations

Corporate finance literature identifies two potential sources of endogeneity arising due to unobserved heterogeneity and simultaneity. One source of endogeneity is often ignored which rises from past values of selected macroeconomic variables and foreign capital inflows. Neglecting this source of endogeneity may have serious consequences on inferences. Traditional fixed-effects estimation may potentially ameliorate the biasness arising from unobservable heterogeneity. It assumes that current observations of the explanatory variable (that is GDP, GFCF, efficiency scores, freedom index, interest rate differential, trade openness) are completely independent of past values of the dependent variable (typically foreign direct inflows), which is an unrealistic assumption. To deal with this issue, dynamic models are used.

To solve the endogeneity problem, Anderson and Hsiao (1981) suggested first differencing the model to eliminate the individual effect ηi and then using Δyit−2=yit−2yit−3 or simply yit−2 as an instrument for Δyit−1=yit−1yit−2, since the differenced variable Δyit−1=yit−1yit−2 is correlated with the differenced error term Δεit=εitεit−1. Arellano and Bond (1991) further proposed that the predetermined variables (yi1, yi2, …, yit−2) be used as the instrumental variables of Δyit−1=yit−1yit−2 and then a GMM procedure be used to estimate the coefficients. We use Arellano and Bond’s GMM estimator in the following estimation since it is more efficient due to its usage of more information (Baltagi, 2008).

The analysis incorporates four steps. First, the intuitive and theoretical arguments are built along with empirical results suggesting that foreign capital inflows are dynamically related to past performance of macroeconomic variables. The second concern is how a well-developed dynamic estimator is well suited to deal with dynamic nature of the relation between FDI and macroeconomic variables. The third step is to study the dynamic GMM estimator to estimate the relationship among various variables. And the fourth step includes discussing the implications of results with the dynamic GMM estimator to deal with endogeneity.

4.1.2 Dynamic panel data specifications

In static panel data model, estimations are undertaken using fixed or random effects model. The condition for applicability of static panel models is not often recognized. The static model prescribed under Equation (1) would be consistent only if current values of an explanatory variable are independent of past realization of a dependent variable. This means that static effect estimators are biased if lagged values of macroeconomic variables affect the current value of FDI. To obtain consistent and unbiased estimates, it is required to look at past realization of macroeconomic variables with the dependent variable. The basic estimation procedure consists of two essential steps, first, converting the model in differenced form and calculating lagged values. First differencing eliminates the potential biasness arising from time-invariant unobserved heterogeneity. After this, the lagged value of explanatory variables is considered as instruments for the model. Arellano and Bond (1991) suggested to use second lag as in instrument because it is not correlated with the current error term, but first lag may be. If the sample size is large enough, one may use all available lags (second and deeper lags) as instruments. We use FDI inflows, other macroeconomic variables, efficiency variables as instruments at lag1 and lag2. The implications drawn from the theoretical model suggest that the impact of macroeconomic variables on past FDI inflows ignores dynamic structure (as do traditional fixed-effects estimators) which yields inconsistent results. Dynamic panel GMM incorporates two aspects of foreign capital inflows: effect of macroeconomic variables on FDI inflows and determinants of FDI inflows and comparative analysis with results obtained from OLS or traditional fixed-effects estimates. Literature review postulates that most studies related to determinants of FDI consider static model only except Aziz and Mishra (2015) and Bokpin et al. (2015).

An attempt to overcome from the previous literature, the dynamic GMM panel estimator is used to account for unobservable heterogeneity, simultaneity to identify the relation between current FDI inflows and past macroeconomic variables (including efficiency variables). The focus on the relationship between efficiency scores/macroeconomic variables and FDI inflows is provided here. Host countries could increase their FDI inflows by improving macroeconomic variables. Therefore, set of macroeconomic variables should be modeled as an endogenous variable. We consider freedom index, interest rate differential, and crisis dummy as a control variable. In such backdrop, the dynamic GMM methodology is best suited, to examine the effect of different variables on FDI inflows.

Consistent with above discussion, the model may be prescribed as follows considering the lagged value of a dependent variable:

(2) FDI i t = c + δ FDI i , t 1 + j = 1 J β i X i t j + k = 1 K β k Y i t k + l = 1 L β l Z i t l + e i t e i t = υ i + u i t
where FDI denotes foreign direct investment of country i and time t with i= 1, … , N, and t=i, …, T. c is the constant term, X i t j , Y i t j , Z i t j are explanatory variables (vector of pull and push factors), eit is disturbance term with unobserved country-specific effects, and υi and uit are idiosyncratic error where υ i I I N ( 0 , σ v 2 ) and u i I I N ( 0 , σ u 2 ) . The basic estimation procedure consists of two essential steps. The model needs to be written in first difference form. First differencing eliminates one period lag of dependent variable which makes the specification dynamic and δ coefficient denotes the speed of adjustment. The following equation takes a differenced form to remove the unobserved country-specific effects:
(3) Δ FDI i t = c + δ Δ FDI i , t 1 + j = 1 J β i Δ X i t j + k = 1 K β k Δ Y i t k + l = 1 L β l Δ Z i t l + e i t

The paper uses GMM one-step and two-step procedures to get the estimates. In one-step estimators, the error terms are assumed to be independent and homoscedastic across the country and time, while two-step estimator uses the residuals of the first step to estimate consistently the variance-covariance matrix of residuals, relaxing the assumption of homoscedasticity (Favarra, 2003). If the sample size is small, the estimates using two-step GMM showed downward biasness and inferences from such estimates may lead to inaccuracy, hence one-step GMM is preferred in such cases (Arellano and Bond, 1991; Blundell and Bond, 1998; Inkmann, 2000). Two-step GMM uses consistent variance-covariance matrix from the first step GMM. Since the construction of one- and two-step GMM is different (both one and two step use different weighting matrices), they provide different estimates. However, both procedures are consistent, but latter is more asymptotically efficient.

GMM panel estimates use two basic assumptions: there is no serial correlation between error term and lagged instruments used are enough to explain the model. Therefore, AR (1) and AR (2) are tests for first-order and second-order serial correlation. In GMM, the biggest concern is related to the inclusion of lags to control the dynamics of the empirical relationship. The residuals in first differences (AR(1)) may be correlated but there should not be any serial correlation in second difference (AR(2)). Inconsistency will imply if second-order autocorrelation (AR(2)) is present (Arellano and Bond, 1991). The next specification test is Sargan test of over-identification. GMM uses multiple lags as instrument variables which make the system over-identified and provide an opportunity to carry out a test of over-identification under null that all instruments are valid.

4.1.3 Calculation of efficiency scores using Malmquist index

To calculate efficiency scores for each country, Malmquist index is adhered using the data envelopment analysis (DEA) analysis. Malmquist index is identified as distance function which considers multi-input and multi-output production technology. Following Fare et al. (1994), the Malmquist TFP change index between period t (base period) and period t+1 (next to base period) is given by:

(4) M o ( x t + 1 , y t + 1 , x t , y t ) = [ ( D o t ( x t + 1 , y t + 1 ) D o t ( x t , y t ) ) ( D o t + 1 ( x t + 1 , y t + 1 ) D o t + 1 ( x t , y t ) ) ] 1 / 2
where the notation D o t + 1 (xt,yt) represents the distance from the period t+1 observation to period t technology, and xt is the input and yt is output for period t. The value of Mo decides to increase or decrease in TFP. If the value is greater than 1, it indicates an increase in total value productivity for period t to t+1 and vice versa if the value comes out to be less than 1.

We consider productivity variables using Fare et al.’s (1994) methodology and constructed three productivity variables, namely technical change, TFP change, and efficiency change based on the input-output ratio that may be determined for every country having input and output variables. Malmquist productivity index is a product of the change in relative efficiency and change in efficiency. We obtain weights of TFP for various countries, using GDP as output and labor and capital as inputs variable. Thereafter, we regress these weights with other independent variables to study their impact on FDI.

4.2 Sample countries, variables, data and period

In all, 20 countries are considered covering 80 percent of foreign capital inflows globally. UNCTAD has identified three blocks of determinants, namely policy framework, economic, and business facilitation. We tried to capture each block as the determinant of FDI in selected countries. Table II enlists the definition of variables undertaken along with their derivation (except for technical scores). Data are sourced from Bloomberg, World Development Indicators (World Bank) and Heritage Foundation from 2004 to 2013 (ten years). Efficiency change (effch), TFP change (tfpch), and technical change (techch) variables are constructed using the DEA for further analysis. Tables AV and AVI represent index scores of efficiency variables year wise.

4.3 Limitations

There are inherent limitations of data used in this research. Despite the popularity of economic freedom index and other indices published by Heritage Foundation, these have attracted considerable criticism. Scott (1997) argued that this database has missed out some important aspects of regulation. Our work is limited to ten years and 20 countries.

4.4 Scope for further research

The greater validity of results may be attained using economic freedom with the location-specific advantage that a country may possess is likely to influence FDI inflows. It is important to replicate the study for both additional time and countries to test whether the relationships uncovered by this research hold true across different economic cycles and may provide more viable results. The present paper incorporates various factors except for sociocultural factors and financial factors that may be undertaken in future as an extension of present research.

5. Analysis and findings

Figure 2 may be segregated into three main parts, namely policy, economic environment, and business facilitation, as determinants of foreign capital inflows (UNCTAD, 2006). Favorable policies, economic environment, and business facilitation attract more capital in developing as well as developed economies. Seven different factors are identified from the list, responding to the economic-policy framework and business facilitation in developed and developing countries. Trade openness is taken as a proxy of the policy framework, efficiency variables as a proxy of the economic framework, and freedom index represents both business facilitation and policy framework determinant. Country-wise efficiency scores are calculated using Fare et al.’s (1994) methodology. A non-parametric programming method is used to compute Malmquist productivity indexes that may be segregated into three parts, namely technical change, efficiency change, and TFP change. Malmquist productivity index is calculated for both developed and developing countries in relation with FDI inflows. Tables AV and AVI elaborate country-wide efficiency scores.

5.1 Descriptive statistics

Tables AI and AII show descriptive statistics of developed and developing countries, respectively, and consider fixed capital formations, trade openness, efficiency measures, reserve to import ratio, US stock market returns, home country stock market movement, and interest rate differentials. Descriptive statistics include mean, median, maximum, minimum, skewness, kurtosis, and standard deviation. The statistics of tables are self-explanatory. The mean score of FDI/GDP is 2.94 in developing countries and 3.84 in developed countries, but GDP growth rate is higher in developing countries than developed countries. During the period 2004-2013, developing countries had higher GFCF/GDP ratio and trade openness, but developed countries had higher freedom index. GDP growth rate is high in developing countries due to movement from low to medium income group countries. GFCF/GDP[3] ratio is higher in developing countries meaning that they are increasing their infrastructural base to attract foreign presence in the market. Similar results are found with efficiency variables, such as efficiency index, which are higher in developing countries than developed countries. Weak infrastructure, low investments in capital formation activities preclude developing countries in seizing business opportunities searched by foreign investors. Apparently, developing countries need to attain a certain level of developments in education, healthcare, technology, skill building, and infrastructure before being able to get benefit from foreign existence in the market.

5.2 Correlation analysis

Tables AIII and AIV show the correlation among different variables. Correlation among independent variables must be low to overcome the problem of multi-collinearity. Variance inflationary index (VIF) is calculated to support correlation test for the pooled regression model. And all VIFs of our regression fall comfortably below the threshold limit prescribed by Chatterjee and Price (1991), indicating that multi-collinearity is not present. Tables AV and AVI show a country-wide annual average of efficiency scores. During the study period, we observe that developing countries are leading developed countries in terms of efficiency scores. And this outcome of the efficiency-driven stage of development includes efficient production process and increased product quality.

5.3 Regression results of determinants of FDI in developed and developing countries

Panel OLS is applied to identify the relationship among FDI, macroeconomic, infrastructural, and efficiency variables. Analysis has been undertaken on sample data of 11 developed and 9 developing countries using fixed and random effect estimations. The choice between two estimates are based on Hausman Test. Another methodology to capture causality among the variables is Pairwise Granger causality test (Granger, 1969) which verifies the treatment of endogenous variables as exogenous one. This approach is used to find whether xt causes yt by seeing the current and past values of yt or vice versa. In econometric terms, causality simply refers to the ability of one variable to predict the other variable. However, causality among variables may be checked via Granger causality test. These tests explore the presence of endogeneity in model. To capture endogeneity in the model, GMM one- and two-step procedure is used.

While analyzing the variables in different models, efficiency scores are added to baseline model (Model 1), then interest rate differential in Model 3, and dummy variable is introduced to capture the impact of crisis period (during 2008-2010) in Model 4. The data are segregated into two parts before crisis and after crisis. Dummy variable consists of 0 before crisis and 1 after crisis to investigate the results of it on foreign capital inflows. The static results using fixed/random effect estimations are prescribed by Hausman specification test which are presented in Tables III and IV for both developed and developing countries, respectively. Models 1-3 highlight fixed effect estimates by rejecting the null hypothesis of random effect applicability through Hausman Test, while Model 4 fails to accept the fixed effect estimators.

Static results of developed countries under Table III show relationship of macroeconomic policy variables on FDI attractiveness. Foreign capital inflows are influenced by GDP growth (Models 1-3), trade openness (Model 4), and freedom index (Models 1, 3, and 4). Efficiency variable, GFCF, interest rate differential, and crisis dummy do not have a significant impact over FDI inflows. The positive association between GDP growth at level and FDI inflows supports the market-seeking hypothesis (Models 1-3). Freedom index has a positive association for FDI inflows (Models 1, 3, and 4), supporting government policy and business facilitation in attracting FDI. Trade openness shows significant relation, but Model 1 is negatively related and Model 4 is positively related to trade openness.

In developing countries (see Table IV), efficiency scores like efficiency change (Models 2 and 4), technological change (Models 1 and 2), TFP change (Models 2 and 4), trade openness (Models 1, 2 and 4), and GFCF (Models 2 and 4) specify a significant positive relationship with FDI inflow except TFP change (Models 2 and 4). The negative association result in that inputs are not used effectively in developing countries and there is a requirement of skill development measures. Positive effects of efficiency variable support resource/efficiency-seeking FDI. The negative relation of GDP growth (Model 4) with FDI inflows is very surprising, and indicates that the variable needs to be examined by taking lag of it, whether there exist any lag relationship or not. Since developing countries are equipped with the advantage of demographic dividends, when they are coupled with skill development initiatives, higher productivity index may be attained, which in return leads to higher capital inflows. Freedom index of developing countries is ineffective in attracting foreign capital because of the relatively lower index in comparison to developed countries. The results of the static model indicate efficiency variables and trade openness to be significant in developing countries and attract more of capital inflows; however, freedom index is not significant at all. It may be concluded that trends of FDI in developing countries are determined by efficiency component and trade openness policy. Developing countries need to focus on skills, efficiency, and innovative capacity to attract FDI.

Crisis during 2008-2010 did not affect FDI inflows because it is a long-term, irreversible decision. Similarly, the insignificance of interest rate differential elaborates FDI as a long-run investment decision, which is not affected by movements in short-run interest rates in developing and developed countries. The above analysis lacks significant results and indicates incomplete models; this may be because of the lagged relationship among the variables. In such kind of backdrop, there is a need to introduce lag of dependent and independent variable as instruments in the model (Arora and Sharma, 2016).

5.4 Panel Granger causality test

To check unidirectional and bidirectional causality, Granger causality test is employed. The result confirmed the role of efficiency in attracting FDI inflows and vice versa. In developed countries, FDI inflow carries growth in GDP, fixed capital, and trade as well. In addition, trade openness leads to growth in GDP. Developing countries show unidirectional causality from efficiency change to GDP growth and GFCF to trade.

From Table V, upsurge in foreign capital flows that leads to higher efficiency, GDP growth, and vice versa is recognized. In other words, efficient countries having higher GDP attract more FDI inflows (Thiam, 2006). Table V also elaborates the presence of bidirectional causality among different variables, highlighting the need for higher order regression models considering endogeneity issues.

5.5 Dynamic panel regression results

Panel causality test explains the presence of unidirectional and bidirectional[4] causality among different variables. To account endogeneity emerging from reverse causality association (Table V), dynamic panel regression (GMM) proposed by Arellano and Bond (1991) with two years lag is used.

Tables VI and VII capture the results of one-step GMM estimators for developed and developing countries, respectively. In the dynamic model of developed countries (Table VI), FDI is having persistence effect because of positive and significant past values (Models 1-4). FDIt-1 and FDIt-2 indicate pull factor for host countries and positive association with current lag specified group as an attractive destination for FDI inflows due to supportive policies and business environment in past years (ÇEviŞ and Camurdan, 2007). Efficiency scores including efficiency change and technical change (Models 1-4), the first lag of GDP growth (Models 2-4), and freedom index (Models 2-4) play a positive and significant role in attracting FDI. Tfpch is negatively related to FDI inflows stating the fact that inputs are not utilized effectively in the host country. In addition, the crisis dummy has significant negative effects in developed countries (Models 2-4), but GFCF and trade openness do not have a significant impact in attracting foreign capital. The results of Table VI elaborated higher efficiency, institutional variable (freedom index) and one period lag of GDP as a significant determinant, supporting resource/efficiency-seeking and market size-seeking FDI.

In the case of developing countries (Table VII), FDI own lag has positive significant coefficients, resulting in the presence of dynamic aspect (except Model 4). GFCF at current lag, trade openness at lag 1, GDP at lag 1 and efficiency variables are found to be positive and significant considering the market size and resource/efficiency-seeking FDI. Since the current value of GDP is negatively related to FDI inflows (Models 1, 3 and 4), but one period lag of trade openness and GDP growth affects FDI inflows positively. It may be because previous year financial openness and growth prospects significantly attract foreign capital inflow. The negative relation of tfpch with FDI highlights that inputs are not utilized effectively in the host country.

The significant coefficient of freedom index in dynamic models is consistent with the static models. The inferences from results showcase the importance of freedom index, efficiency, and past FDI performance in attracting FDI inflows in developed countries (Table VI), while GFCF, trade openness, GDP growth, and efficiency significantly attract higher FDI inflows for developing countries (Table VII).

Table VIII highlights two-step GMM estimators for both set of countries taken as a group. Lagged dependent variable (FDI) comes out to be highly significant which confirms the dynamic specifications of the model and justified the usage of two-step GMM. The coefficient of lagged dependent variable ranges from 0.253-0.502 which shows a moderate degree of persistence in FDI inflows as a pull factor. GMM two-step estimates showed GFCF, trade openness, GDP growth, efficiency change, and crisis dummy as significant determinants of FDI inflow. GFCF lag 2 and lag 1 (except Model 1) are significant for all models stating that previous year infrastructural upgradations attract more FDI inflows. Lags 1 and 2 of trade openness implicate financial openness leading to higher capital inflows and upturns the level of attractiveness of the country. Looking at growth rate, current period and lag 1 entice capital inflows; however, lag 2 does not have a significant impact on it. The current level of technical change is highly significant. FDI flows contribute to economic growth in a host country by promoting productivity growth and this productivity growth in turn leads to higher inflow of capital through technology transfer. The crisis had a significant negative impact on FDI inflows when we consider the set of 20 countries as a group, but insignificant in case of developing countries.

Post-estimation tests, AR(1) and AR(2) coefficients are found to be insignificant which implies the absence of first- and second-order autocorrelation in data. Wald test statistics gives significant values of χ2, rejecting the null hypothesis that estimated coefficients are jointly and significantly different from zero, meaning that model is having predictive power. Rejection of over identifying restrictions using Sargan test suggests that all instruments are valid to explain the model.

6. Summary and concluding thoughts

The present study attempts to enrich existing literature by using efficiency scores, economic freedom and identifying the potential issues related to endogeneity among different variables. So far, several economic, political, and institutional determinants have been analyzed where the data sets are limited to OECD and developed countries, especially in the case of efficiency and governance indicators. The present study offered a novel line of research by constructing efficiency scores outlined in Fare et al.’s (1994) methodology and using a standalone predictor of economic freedom in the form of freedom index. The panel of 20 developed and developing countries is analyzed, and we found that efficiency scores play a significant role in explaining the behavior of FDI inflows in developed and developing countries. The significance of economic freedom is observed in developed countries only. The positive coefficients of efficiency change in static and dynamic model (in case of developing countries) signify the importance of efficiency in attracting FDI (Dunning, 1998; Kinoshita and Campos, 2003; Gourinchas and Jeanne, 2006). Nevertheless, the coefficient of TFP change is negatively related to FDI indicating that inputs are not used efficiently and intensively in production function (Table VII). TFP is a Solow residual but technically it is a portion of output that is not explained by inputs. It seems that developed and developing countries are not currently meeting the level of expertise required for output orientation; hence, more proficiency is required to solve this agenda. Negative TFP encourages more training and development activities to personnel and innovative research activities to minimize capital inputs. Negative TFP coefficient should not be taken as a destructive measure, rather it should be used as one of the implications to provide skill-based training to demographics of a country.

FDI inflows have been categorized under two heads: horizontal (market seeking) and vertical (resource seeking/efficiency seeking) investment. As discussed earlier, FDI is attracted by the characteristics of the host country, and the flows of FDI to developed (developing) countries are usually horizontal (vertical) in nature. The static models of developed countries (Table III) support market-seeking hypothesis where GDP and freedom index are positively related to FDI, and the results of dynamic model (Table VI) support market-seeking as well as resource/efficiency-seeking hypothesis where GDP growth, freedom index, and efficiency scores are among the major determinants of FDI. The coefficients of efficiency, capital formulation, and trade openness are found to contribute significantly in attracting FDI for developing countries in resource-seeking/efficiency-seeking FDI (Tables III and VI). FDI is a long-term irreversible decision and involves a huge investment in the host country. The results implicate ineffectiveness of interest rate differentials in both sets of countries. The crisis dummy is negatively related to a developed country (Table VI) and insignificant in the case of developing countries (Table VII), inferring that developing countries are preferred choice for FDI investors even in the crisis period. The overall impact of crisis dummy was seen negative for both set of firms taken together (Table VIII), stating that the negative effect of the global financial crisis in developed countries appears to have a progressively negative impact on the whole group.

7. Policy and managerial implications

The results confirmed the presence of regional and income group heterogeneity in FDI inflows. In the context of policy formulation developing countries need progress in different areas such as improvement in freedom index, openness in economy, skill-based training to workers to increase their technical efficiency and stability in growth rate. These benefits should be clubbed with the liberal economic framework. Political, government, and legal factors need to be addressed to upsurge the level of foreign capital inflow in developing countries. Higher infrastructural development leads to higher capital inflows, and developing countries should upgrade an infrastructural framework for ease in doing business. Policy measures such as promoting domestic manufacturing, financial assistance, and skill-based trainings and educating workforce about new avenues, technology transfer requirements are strategic tools for attracting FDI. Another aspect that needs to be addressed in both set of countries is related to improvement in production network and supply chain through economic agglomeration to increase efficiency and factor productivity. Most developing countries are well equipped with the dividend of demographics and have potential to become “human resource capital” of the world. Efficient demography has the power to upgrade developing countries in higher income grade. Hence, productivity improvement and skill development program need to be strengthened in the near future.

Figures

Attributions to macro, national and firm level performance

Figure 1

Attributions to macro, national and firm level performance

Determinants of foreign capital inflows

Figure 2

Determinants of foreign capital inflows

Chronological analysis of literature review

Author(s) of study Title of study Sample data No. of sample country Methodology/Tools adopted for data analysis Determinants Findings and conclusion
 1. Tsai (1994) Determinants of FDI and its impact on economic growth 1975-1978/1983-1986 62/51 Simultaneous equation model using 2SLS using SAS Economic: trade openness, export growth As determinant of FDI, market size hypothesis is having favorable evidences than growth size hypothesis. Trade balance, cheap labor and rate of export growth are key factors for attracting FDI in a country
 2. Shamsuddin (1994) Economic determinants of foreign direct investment in less developed countries 1 36 Cross-sectional regression Economic: wage cost, per capita GDP
Policy/Political: economic instability, per capita public aid
Per capita GDP, wage cost, per capita debt, per capita public aid and volatility of prices played a significant role in attracting FDI
 3. Aristotelous and Fountas (1996) An empirical analysis of inward foreign investment flows in the EU with emphasis on Market enlargement hypothesis 1980-1990 9 Cross-sectional regression Economic: GDP
Policy/Political: exchange rate
Institutional: European Act
Locational determinants of FDI inflows are considered in EU. Market size, real exchange rates and single European Act influences FDI inflows
 4. Dees (1998) Foreign direct investment in China: determinants and effects 1983-1995 11 Panel data regression Economic: patents, labor, relative wages, GDP, production
Policy/Political: exchange rate, imports/GDP
Traditional determinants of FDI seem to be relevant for China considering domestic market size, cost advantages and openness to rest of world
 5. Yang et al. (2000) The determinants of FDI in Australia 1985-1994 1 Multiple regression analysis Economic: wage rate, GDP
Policy/Political: trade openness, exchange rate
Interest rate, wage changes, openness in economy and industrial dispute settlement strategies are major determinants of FDI in Australia
 6. Sun et al. (2002) Determinants of foreign direct investment across China 1986-1998 30 Pooled regression model Economic: retail sales, railway, highway, no. of engineers, no. of patents, GDP, GDP per capita
Policy/Political: trade openness, country risk
Wage rates have positive relationship with FDI inflow before 1991 but thereafter showed a negative association. Labor quality and infrastructural facilities are important determinants throughout period of study
 7. Yuqing (2006) Why is China so attractive for FDI? The role of exchange rates 1981-2002 1 Panel data and pooled regression are used Economic: GDP
Policy/Political: Exchange rate
Real exchange rate between Yuan and Yen is one of the significant variable determining Japanese direct investment in China. Devaluation of Yuan substantially enhanced inflows of direct investment from Japan
 8. Hsiao and Hsiao (2006) FDI, exports, and GDP in East and Southeast Asia – panel data vs time-series causality analyses 1986-2004 8 Panel data, time series, Granger causality and VAR model Economic: real domestic investment, real FDI inflows, real exports, real imports, GDP, real consumption
Policy/Political: interest rates, exchange rates
Time series analysis has been observed individually using granger causality and VAR, followed by panel data analysis for set of 8 countries. Findings include analysis through panel data causality to be superior over time-series causality analysis
 9. Herzer and Klasen (2008) In search of FDI-led growth in developing countries: the way forward 1970-2003 28 Panel co-integration Economic: GDP, FDI Moving from FDI to GDP, no single country is having positive unidirectional long-term effects. No clear association between growth impact of FDI and level of per capita income, level of education, degree of openness and level of financial market development in developing countries is found
10. Kok and Acikgoz Ersoy (2009) Analyses of FDI determinants in developing countries 1983-2005 24 Panel data and cross-sectional data using FMOLS and cross-sectional SUR Economic: technological gap, GDP, GFCF
Policy/Political: inflation
In developing countries, interaction of FDI with select determinants has a significant positive relation with economic progress while the interaction of FDI with Total debt service/GDP and inflation has a negative impact. Communication variable is major determinant of FDI
11. Hanousek et al. (2011) Direct and indirect effects of FDI in emerging European markets: a survey and meta-analysis 2000-2007 11 Primary survey Primary survey FDI inflow leads to direct and indirect spillover effects. Forward spillover are found to be negative and significant while backward spillover are positively related to FDI inflows
12. Nyen Wong and Cheong Tang (2011) Foreign direct investment and employment in manufacturing and services sectors 1997-2005 1 ARDL, causality and co-integration Economic: employment in services, employment in manufacturing, FDI Interaction between FDI inflows and employment services in Singapore provides useful insights toward promoting foreign investment in emerging areas and work force development
13. Bilgili et al. (2012) The determinants of FDI in Turkey: a Markov regime-switching approach 1988-2010 1 Markov regime-switching models Economic: GDP growth, labor cost, the electricity price growth, the growth
Policy/Political: export growth, import growth
Institutional: discount rate and country risk indexes
Turkish FDI growth equation has significant structural changes at level or/and trend. It has significant coefficient shifts in explanatory variables. Explanatory variables include GDP growth, labor cost, electricity, sulphur, oil prices, natural gas, export growth and import growth
14. Goswami and Saikia (2012) FDI and its relation with exports in India, status and prospect in north east region 1991-2011 1 Granger causality Economic: GDP, in average prices of high sulphur fuel oil, cooking coal, steam coal and natural gas, FDI
Policy/Political: export
Long-run and short-run causality among the FDI, GDP and exports on Indian data is reported
15. Temiz and Gökmen (2014) FDI inflow as an international business operation by MNCs and economic growth: an empirical study on Turkey 1992-2007 1 Co-integration analysis and Granger causality Economic: GDP, FDI As per econometric analysis, the country does not observe significant correlation between FDI entry and GDP growth (economic growth) in short and long run
16. Lansbury et al. (1996) Foreign direct investment in Central Europe since 1990: an econometric study 1991-1993 4 Multiple regression analysis Economic: labor-related factors Capabilities of labor force attract FDI. The study focused mostly on efficiency and labor specifications, namely labor cost, research intensity and skilled workforce attracts FDI
17. De Vita and Kyaw (2008) Determinants of capital flows to developing countries: a structural VAR analysis 1976-2001 5 VAR model Economic: GDP, productivity, FDI Foreign output shock have a negative effect on capital flows while a foreign interest rate shock is positively related. Domestic productivity shock seems to cause an increase in FDI
18. Fetscherin (2010) The determinants and measurement of a country brand: the country brand strength index” 1 year 31 Standardized CBSI technique is used Branding index is prepared This study highlights tourism, export, FDI, government policies for developing the brand confidence index and strong country brand help in increasing exports, invite tourism, investment and immigration
19. Freckleton et al. (2012) Economic growth, foreign direct investment and corruption in developed and developing countries 1998-2008 70 PDOLS Economic: GDP, human capital, labor Capital flows, labor and human capital are found to be significant in long run for both developed and developing countries. Results indicate importance of profitability, government directed incentives, local institutional and human capital effectiveness rather than corruption in country
20. Alam and Zulfiqar Ali Shah (2013) Determinants of foreign direct investment in OECD member countries 1985-2009 10 Panel data, Granger causality for long term and for short term, co-integration and VECM Economic: quality of infrastructure, labor cost, labor productivity, GDP
Policy/Political: trade openness, political stability, exchange rate, inflation
Fixed effect estimations specify market size, labor cost and quality of infrastructure as significant coefficients in relation to FDI. A bidirectional short-run relationship between market size and labor cost is found in short run whereas short term unidirectional causalities are found amongst quality of infrastructure, market size and labor costs
21. Belloumi (2014) The relationship between trade, FDI and economic growth in Tunisia: An application of the autoregressive distributed lag model 1970-2008 1 ARDL, Co-integration and VECM Economic: GDP, cost, infrastructure Results indicate that there is no significant Granger causality running from FDI to economic growth, from economic growth to FDI, from trade to economic growth and from economic growth to trade. Even though there is a widespread belief, that FDI can generate positive spillover externalities for host country
22. Chan et al. (2014) Foreign direct investment and its determinants: a regional panel causality analysis 1995-2010 1 Granger causality (linked to 4 ways) Economic: GDP, capital, road, wage, education, telephone This study showcases short and long-run flows of causality involving FDI and a comprehensive set of possible determinants like FDI, GDP, domestic investment, infrastructure and quality of labor force. In short and long run, growth in GDP directly influences FDI while growth in local infrastructure and local investment provide indirect influence
23. Wisniewski and Pathan (2014) Political environment and foreign direct investment: evidence from OECD countries 1975-2009 33 Pooled regression analysis, fixed effect model and dual fixed effect models Policy/Political: party in power, party age, government spending, military expenditure Political factors play a significant role in FDI determinants as they strongly affects investment location decisions of MNEs
24. Sambharya and Rasheed (2015) Does economic freedom in host countries lead to increased foreign direct investment? 1995-2000 95 GLS Policy/Political: political freedom, corruption
Institutional: economic freedom, economic management, state interference
Better economic management, less government participation in economy, less state intervention, absence of wage and price controls and higher levels of political freedom lead to higher FDI inflows
25. Aziz and Mishra (2015) Determinants of FDI inflows to Arab economies 1984-2012 16 Economic: financial development
Policy/Political: trade openness
Institutional: trade agreement
Better financial institutions and educated labor force may play a key role in attracting FDI. It is suggested that Arab economies should sequence their economic policy measures with institutional policies, starting with privatization, trade liberalization and improvement in economic growth
26. Ghazal and Zulkhibri (2015) Determinants of innovation outputs in developing countries: Evidence from panel data negative binomial approach 1996-2010 18 Panel data, negative binomial approach Economic: economic freedom, Government spending
Policy/Political: political stability, government effectiveness, rule of law, corruption
Institutional: property rights, Business freedom, trade freedom, governance indicators
The study supports the role of institutions and governance indicators for increasing innovating activities. But impact of freedom indicators on improving innovation output is mixed
27. Onyeiwu and Shrestha (2004) Determinants of foreign direct investment in Africa 1975-1999 29 Panel regression analysis using fixed and random effects Economic: GDP Growth, natural resource
Policy/Political: inflation rate, real interest rate, openness, international reserves, external debt
Economic growth, inflation, openness of economy, international reserves and natural resource availability were found to be significant while political rights and infrastructure were found unimportant for FDI flows in Africa
28. Jadhav (2012) Determinants of foreign direct investment in BRICS economies: analysis of economic, institutional and political factor 2000-2009 5 Panel data multiple regression Economic: GDP
Policy/Political: inflation rate, FDI policy as policy variable, political stability and govt. effectiveness as political factors
Institutional: corruption, rule of law as institutional variable
Economic factors are more significant than institutional and political factors in BRICS economies. Market size measured by real GDP is a significant determinant of FDI
29. de Castro et al. (2013) The determinants of foreign direct investment in Brazil and Mexico: an empirical analysis 1990-2010 2 VECM and VAR Economic: GDP, exchange rate, FDI
Institutional: trade openness
In Brazil, main multinationals’ strategy is market seeking, linked to size of domestic market. In Mexico, dominant strategy seems to be efficiency seeking, related to importance of trade liberalization. Brazilian FDI is affected by more of GDP while in Mexico it is based on liberalized policies
30. Bokpin, Mensah and Asamoah (2015) Foreign direct investment and natural resources in Africa 1980-2011 49 System GMM Economic: natural resource, GDP This study is based on association between natural resources and FDI inflows. Results indicated regional economic cooperation for facilitating FDI to small size African countries
31. Masron (2017) Relative institutional quality and FDI inflows in ASEAN 1996-2003 8 Dynamic OLS, FMOLS Institutional: relative institutional variable Relative IQ score in comparison to China is analyzed to explore the level of competition among ASEAN countries. And the study suggest that apart from improvement in IQ per se, ASEAN countries need to insure significant improvement in relative IQ scores by considering IQ of competing countries to increase the level of FDI inflows

Variables and notations

Variable Type of variable Definition/Explanation Formula Supported studies
 1. FDI/GDP Nominal FDI inflows as percentage of GDP (FDI inflow/GDP) × 100 All studies take FDI as dependent variable
 2. GDP growth Real Percentage growth in GDP ((GDPt/GDPt-1)−1) × 100 Bahmani-Oskooee and Niroomand (1999), De Vita and Kyaw (2008), Aw and Tang (2010), Bilgili et al. (2012), Belloumi (2014), Chan et al. (2014)
 3. GFCF/GDP Nominal Gross fixed capital formation as percentage of GDP (GFCF/GDP) ×100 Blomstrom et al. (1989), Kok and Acikgoz Ersoy (2009), Pradhan and Kelkar (2014).
 4. Interest rate differential Real Difference of Interest rates in home country and LIBOR US Home country interest rate-LIBOR Reinhart (2005), Dua and Garg (2015)
 5. Technical change (techch) Nominal Measure of efficiency describing change in output with same input units Calculated from DEA programming using Malmquist index Simar and Wilson (2007), Chortareas et al. (2013), Dell’Atti et al. (2015)
 6. Total factor productivity change (tfpch) Nominal Measure of economy’s long-term technological change. Also, called multifactor productivity
 7. Efficiency change (effch) Nominal Measures growth in productivity due to changes in input-output ratio
 8. Trade openness Nominal Total Trade as percentage of GDP (Trade/GDP) ×100 Banga (2006), Leitão (2010), Aw and Tang (2010), Singhania and Gupta (2011), Alam et al. (2013)
 9. Freedom index Nominal Weighted average of all freedom indexes including fiscal freedom, government spending, business freedom, labor freedom, trade freedom, investment freedom and financial freedom Extracted from Heritage Foundation Sambharya and Rasheed (2015)
10. Crisis dummy Real Measure of crisis effect on FDI inflows Data have been segregated by introducing crisis dummy Ersoy and Erol (2016), Desbordes and Wei (2017)

Static panel results of developed countries

Model 1 (baseline model) Model 2 (with efficiency Scores) Model 3 (with interest rate differential) Model 4 (with crisis)
Dependent variable as FDI GDP
Independent variables Coefficient t-statistic Coefficient t-statistic Coefficient t-statistic Coefficient t-statistic
GDP growtht 0.590047*** 3.973813 0.555768*** 3.508063 0.554941*** 3.311964 0.439176 0.163934
GFCF/GCPt −0.10339 (−1.540958) −0.064256 −0.920793 −0.05706 −0.758053 −0.07129 (−1.021012)
Efficiency changet −140.7316 −0.98159 −59.3977 −0.381337 −64.6244 (−0.447508)
Technological changet −141.2999 −0.989671 −60.6584 −0.390454 −66.6406 (−0.462314)
TFP changet 148.5746 1.037435 −60.6584 −0.390454 78.90849 0.547802
Trade opennesst −0.04749* (−1.888049) −0.032723 (−1.23322) −0.04109 (−1.49103) 0.052793*** 12.98031
Freedom Indext 0.292927** 2.094342 0.228138 0.0791 0.282287** 1.997612 0.060602** 2.100572
Interest rate differentialt −0.23232 −0.941881
Crisis Dummyt −0.73668 (−1.330349)
Constant −15.46193 (−1.496962) 121.7379 0.847237 38.21138 0.243515 47.69695 0.329989
R2 0.817347 0.811463 0.72305 0.764759
F-statistic 33.88114 27.01954 26.35799 45.51341
Prob (F-statistic) 0.0000 0.0000 0.0000 0.0000
Hausman specification test (χ2 statistics) 17.212706 19.545398 34.732885 3.000000
Prob 0.0018 0.0066 0.0000 0.9670
Applicability of model Fixed effects Fixed effects Fixed effects Random effects
No. of observations 110 110 110 110

Note: *,**,***Significant at 10, 5 and 1 percent levels, respectively

Static panel results of developing countries

Model 1 (baseline model) Model 2 (with efficiency Scores Model 3 (with interest rate differential) Model 4 (with crisis)
Dependent Variable as FDI GDP
Independent variables Coefficient t-statistic Coefficient t-statistic Coefficient t-statistic Coefficient t-statistic
GDP growtht 0.046139 1.098692 0.017463 0.355303 0.042901 0.972749 −0.11162*** (−3.242757)
GFCF/GCPt 0.008595 0.493515 0.033163** 1.667754 0.018844 1.036955 0.058984*** 3.582392
Efficiency changet 44.61656** 2.717905 17.5728 0.092 32.49702** 1.988638
Technological changet 42.20586** 2.615478 17.0174 0.0952 0.03277 (0.401711)
TFP changet −40.74532** (−2.545555) −26.9035 (−1.58365) −27.2614* (−1.737642)
Trade opennesst 0.02287* 1.845888 0.019098** 1.993498 0.018298 1.41629 30.12489* 1.899645
Freedom Indext −0.02075 (−0.240079) −0.088964 −0.947689 −0.02405 −0.275975 −0.00199 (−0.787848)
Interest rate differentialt 0.011301 0.7492
Crisis dummyt −0.28954 (−1.197971)
Constant 0.862934 1.059894 −44.93296*** −2.691981 −31.07671* −1.757111 −32.331* −1.94
R2 0.561363 0.703172 0.594612 0.207156
F-statistic 7.252137 10.80836 5.867094 2.351535
Prob (F-statistic) 0.0000 0.0000 0.0000 0.006309
Hausman specification test (χ2 statistics) 10.914886 78.451472 60.036917 2.685765
Prob 0.0275 0.0000 0.0000 0.9125
Applicability of model Fixed effects Fixed effects Fixed effects Random effects
No. of observations 90 90 90 90

Note: *,**,***Significant at 10, 5 and 1 percent levels, respectively

Granger causality results among the variables of developed and developing countries

Developed countries Developing countries
Null hypothesis F-statistic Prob. F-statistic Prob.
Efficiency change does not Granger cause FDI/GDP 3.79692** 0.0256 2.98797* 0.0564
FDI GDP does not Granger cause efficiency change 3.40621** 0.0369 2.97987* 0.0568
GDP growth does not Granger cause FDI/GDP 1.49634 0.2288 0.32547 0.7232
FDI/GDP does not Granger cause GDP growth 8.78808*** 0.0003 1.68830 0.1918
TFP change does not Granger cause FDI/GDP 3.12730** 0.048 2.04955 0.1359
FDI/GDP does not Granger cause TFP change 0.56938 0.5676 1.47121 0.2362
TRADE does not Granger cause FDI/GDP 22.8442*** 0.000 0.39783 0.6732
FDI/GDP does not Granger cause trade openness 8.19597*** 0.0005 0.17523 0.8396
GDP growth does not Granger cause efficiency change 3.48444*** 0.0343 1.43253 0.2452
Efficiency change does not Granger cause GDP growth 5.40674*** 0.0058 3.58204** 0.0327
GFCF/GDP does not Granger cause efficiency change 5.48569* 0.0054 0.26211 0.7701
Efficiency change does not Granger cause GFCF/GDP 0.24274 0.7849 0.47249 0.6253
Technical change does not Granger cause efficiency change 2.10619 0.1269 0.15360 0.8579
Efficiency change does not Granger cause technical change 0.25527 0.7752 0.41548 0.6615
TFP change does not Granger cause efficiency change 1.94950 0.1476 0.01753 0.9826
Efficiency change does not Granger cause TFP change 0.09110 0.913 0.09737 0.9073
Trade openness does not Granger cause efficiency change 1.83252 0.1652 0.35544 0.702
Efficiency change does not Granger cause trade openness 9.26271*** 0.0002 0.86049 0.4271
trade openness does not Granger cause GFCF GDP 4.45579** 0.0139 1.09111 0.3411
GFCF/GDP does not Granger cause trade openness 2.16907 0.1195 3.94735** 0.0234

Note: *,**,***Significant at 10, 5 and 1 percent levels, respectively

Dynamic panel results of developed countriesa (GMM one step)

Model 1(baseline model) Model 2(with crisis) Model 3(with trade openness) Model 4(with all variables)
Independent variables Coefficient t-statistics Coefficient t-statistics Coefficient t-statistics Coefficient t-statistics
FDIt−1 0.243* 1.69 0.143* 1.68 0.186** 1.98 0.2394856** 2.25
FDIt−2 0.557* 2.04 −0.16 (−0.59) −0.00851 (−0.03) 0.0964524* 1.90
GFCF/GCPt −0.068 (−0.32) 0.0687 −0.66 −0.107 −0.98 0.1644982 −1.08
GFCF/GCPt−1 −0.539 (−0.36) −0.267 (−0.37) −0.643201 (−0.68)
GFCF/GCPt−2 −0.152 (−0.16) −0.025 (−0.07) −0.3955497 (−0.77)
Trade opennesst 2.214 0.21 −2.4 (−0.46) −1.129 (−0.59) −2.228172 (−0.38)
Trade opennesst−1 1.225 (−0.14) −2.272 (−0.74) −5.849824 (−1.04)
Trade opennesst−2 −3.136 (−0.56) −0.54 (−0.23) 0.487694 −0.15
GDP growtht 0.491 (−0.68) −0.363 (−0.77) −0.295 (−1.05) −0.5635959 (−0.94)
GDP growtht−1 −0.131 (−0.06) 0.767* 1.83 0.506* 1.86 1.627642* 1.71
GDP growtht−2 0.471 0.26 −0.0556 (−0.08) −0.0108 −0.04 −0.5596971 −0.59
Efficiency changet 7.828* (−2.05) 17.2* 2.08 16.1* 2.36 26.9934** 2.03
TFP changet −5.863* (−2.00) −16.1* (−2.06) −14.0* (−2.35) −13.0679** 2.01
Technological changet 10.47* 2.02 16.3* 2.1 15.1* 2.38 25.4758** 2.04
Interest rate differentialt 15.26 −0.55 9.831 0.77 −14.53 −1.6 12.08442 0.84
Crisis dummyt −2.064* (−2.52) −2.216** (−2.82) −12.08442* (−2.25)
Freedom Indext 2.064** 2.52 2.216** 2.82 0.4599103** 2.34
Constant −65.99 (−0.46) −30.03 (−0.43) −44.63 (−1.07) −19.22357 (−0.24)
Observations 18 18 18 18
Wald test χ2=16.92 (0.3238), df=15 χ2=80.53 (0.00), df=14 χ2=105.5 (0.00), df= 14 χ2=67.05 (0.00), df=16
Sargan test χ2=1.678784 (0.6417), df= 16 χ2=2.654197 (0.6173), df= 4 χ2= 2.990 (0.5594), df= 3 χ2=1.075857 (0.5840), df= 2

Notes: aModel 1-4 consider GDP growth rate, trade openness (except Model 3) as endogenous variable and efficiency variables, interest rate differential, freedom index as control variable. *,**,***Significant at 10, 5 and 1 percent levels, respectively

Dynamic panel results of developing countriesa (GMM one step)

Model 1(baseline model) Model 2(with crisis) Model 3 (with efficiency change) Model 4 (with technological change)
Independent variables Coefficient t-statistics Coefficient t-statistics Coefficient t-statistics Coefficient t-statistics
FDIt−1 0.255* 1.69 0.28* 1.79 0.114* 1.71 −0.502 (−0.53)
FDIt−2 0.0746 0.34 0.0485 0.19 0.602* 1.68 0.134 0.18
GFCF/GCPt 0.352** 2.09 0.362** 2.01 0.148 0.54 0.203 0.64
GFCF/GCPt−1 0.441* 1.78 0.432* 1.66 0.693** 1.99 0.890** 2.12
GFCF/GCPt−2 −0.0549 (−0.23) −0.0401 (−0.16) −0.623 (−1.63) −0.295 (−0.58)
Trade opennesst −4.097** (−2.40) −3.974** (−2.14) 8.682** 3.28 −8.925** (−3.05)
Trade opennesst−1 2.95* 1.88 2.809 1.61 7.089** 3.1 6.381** 2.5
Trade opennesst−2 0.42 −0.36 0.401 0.34 −0.706 (−0.50) −0.333 (−0.21)
GDP growtht −0.971** (−1.96) −0.91 (−1.55) −1.801** (−2.31) −1.693** (−1.98)
GDP growtht−1 1.502** 2.7 1.420** 2.06 2.905** 3.03 3.218** 2.9
GDP growtht−2 −0.947 (−1.51) −0.883 (−1.24) −1.015 (−1.00) −0.344 (−0.28)
Efficiency changet −31.04 (−1.30) −29.66 (−1.16) 62.71** 2.05 86.78** 2.24
Efficiency changet−1 −1.512 (−0.36) 60.68* 1.74
Efficiency changet−2 −6.097 (−1.54) −30.9 (−0.81)
Technological changet 32.74 1.37 31.61 1.25 59.45** 2 85.67** 2.2
Technological changet−1 62.13 1.21
Technological changet−2 28.35 0.73
TFP changet −31.99 (−1.28) −30.64 (−1.16) −64.39** (−2.02) −88.39** (−2.21)
TFP changet−1 4.46 0.79 −54.76 (−1.11)
TFP changet−2 −3.579 (−0.71) −28.11 (−0.73)
Crisis dummyt 0.078 0.21 −0.406 (−0.71) −0.0377 (−0.05)
Interest rate differentialt −0.101 (−0.32) −0.129 (−0.37) 0.179 0.45 −0.0469 (−0.10)
Freedom indext 3.535 0.76 3.306 0.68 1.453 0.26 5.326 0.79
Constant −10.71 (−0.60) −9.706 (−0.51) 3.277 0.14 −12.14 (−0.43)
Observations 29 29 29 29
Wald test χ2=20.77 (0.0875), df=16 χ2=19.55 (0.2980), df=17 χ2=29.41 (0.1044), df=21 χ2=26.53 (0.2766), df=23
Sargan test χ2=15.59579 (0.2716), df=13 χ2=14.59798 (0.2642), df=12 χ2=6.704249 ( 0.5689), df=8 χ2=4.236281 (0.6447), df=6,

Notes: aModel 1-4 consider GDP growth rate, trade openness and efficiency as endogenous variable and interest rate differential, freedom index as control variable. *,**,***Significant at 10, 5 and 1 percent levels, respectively

Dynamic panel results of developed and developing countriesa (GMM two step)

Model 1 (baseline model) Model 2(with GFCF) Model 3 (with crisis dummy) Model 4 (with all variables)
Independent variables Coefficient t-statistics Coefficient t-statistics Coefficient t-statistics Coefficient t-statistics
FDIt−1 0.683** 2.22 0.553** 2.98 1.155*** 3.30 0.597** 2.80
FDIt−2 0.714** 2.05 0.653** 2.74 0.962** 2.92 0.202** 2.82
GFCF/GCPt 0.0760 (−1.29) −0.135 (−1.54) −0.240* (−1.77)
GFCF/GCPt−1 −0.251* (−1.92) 0.965** 2.25 1.427** 2.38
GFCF/GCPt−2 0.285*** 3.34 0.292** 3.10 0.348** 1.99
Trade opennesst 1.100 1.07 0.961 0.98 2.917** 2.14 4.008* 1.91
Trade opennesst−1 −1.200 (−0.53) −0.637 (−0.30) −1.956 (−0.90) −1.424 (−0.68)
Trade opennesst−2 4.014** 1.97 3.227** 2.58 6.698** 2.64 10.10** 2.31
GDP growtht 0.322** 2.35 0.214** 3.16 0.396*** 3.52 0.735* 1.72
GDP growtht−-1 1.691** 2.42 1.320** 3.25 3.187** 2.79 5.089** 2.29
GDP growtht−2 −0.116 (−0.21) −0.157 (−0.57) −0.502 (−1.15) −0.660 (−1.36)
Efficiency changet 0.302 0.18 0.951 0.95 1.580 0.69 −0.584 (−0.17)
Technological changet 6.020** 2.27 5.410** 2.76 14.52** 3.19 18.13** 3.13
Crisis dummyt −0.590* (−1.81) 0.757 1.60
Interest rate differentialt 0.426 0.88
Freedom indext 7.152 0.43
Constant −14.39 (−0.88) −12.58 (−1.00) −29.63 (−1.48) −82.15 (−0.96)
Observations 47 47 47 47
AR(1) p-value 0.5097 0.3430 0.3220 0.2369
AR(2) p-value 0.8656 0.8028 0.4613 0.7238
Wald test χ2=20140.1 (0.000), df=13 χ2=1354.6 (0.000), df=13 χ2=144.4 (0.000), df=13 χ2=2275.9 (0.000), df=13
Sargan test χ2=4.909039 (0.997), df= 21 χ2=3.98919 (0.999), df= 12 χ2=3.176632 (0.999), df= 22 χ2=2.6138794 (0.2716), df=22

Notes: aModel 1-4 consider GDP growth rate, trade openness as endogenous variable and efficiency variables, interest rate differential, freedom index as control variable. *,**,***Significant at 10, 5 and 1 percent levels, respectively

Descriptive statistics of FDI in developed countries

FDI/GDP GDP growth GFCF/GDP Trade openness Freedom index techch tfpch effch Interest rates differential
Mean 3.841813 2.145486 2.432899 66.80251 72.97833 1.016117 1.017100 1.001433 −0.02546
Median 2.183080 2.194351 2.906593 45.25903 73.70000 1.014000 1.017850 1.000000 −0.242
Maximum 26.52124 15.24038 15.97063 345.4244 88.90000 1.150000 1.143000 1.151000 4.823
Minimum −3.619176 −5.637953 −16.96551 18.45755 57.20000 0.919000 0.735000 0.743000 −3.895
SD 5.377358 2.716732 5.536143 73.33063 8.895858 0.040220 0.055936 0.050844 1.70761
Skewness 2.465560 0.531777 −0.793037 2.905547 −0.137968 0.693285 −1.22073 −0.96141 0.66076
Kurtosis 8.971056 7.758552 4.388379 10.04379 1.825983 4.251566 8.407378 9.447069 3.31835
Observations 110 110 110 110 110 110 110 110 110

Descriptive statistics of FDI in developing countries

FDI/GDP GDP Growth GFCF/GDP Trade openness Freedom Index techch tfpch effch Interest rates differential
Mean 2.947787 5.546972 7.806251 73.74505 56.72400 1.017790 1.024770 1.009680 4.30611
Median 2.842785 5.624162 9.129935 51.02315 56.00000 1.011000 1.024500 1.000000 4.205
Maximum 9.663039 14.16240 24.15719 185.8072 66.40000 1.274000 1.349000 1.237000 16.41
Minimum 0.056694 −7.820885 −14.4 17.69373 46.10000 0.877000 0.805000 0.788000 −3.62
SD 1.704843 3.141020 7.051113 46.05723 5.290035 0.079306 0.091212 0.087712 4.17053
Skewness 1.092375 −0.799765 −0.281137 0.808156 0.302561 0.361501 0.285490 0.051214 0.73687
Kurtosis 5.267706 6.009180 3.574991 2.276471 1.789773 2.753801 3.791492 3.443038 3.73782
Observations 90 90 90 90 90 90 90 90 90

Correlation matrix of listed variables of FDI in developed countries

FDI/GDP GDP Growth GFCF/GDP Trade openness Freedom index techch tfpch effch Interest rate differential
FDI/GDP 1
GDP growth 0.59 1
GFCF/GDP 0.32 0.75 1
Trade openness 0.84 0.53 0.31 1
Freedom index 0.45 0.37 0.28 0.42 1
techch 0.03 −0.19 −0.29 −0.02 −0.07 1
tfpch 0.02 −0.06 −0.29 −0.16 −0.02 0.48 1
effch 0.01 0.08 −0.09 −0.17 0.03 −0.24 0.53 1
Interest rate difference 0.03 0.15 0.2 −0.08 0.26 −0.02 0.14 0.17 1

Correlation matrix of listed variables of FDI in developing countries

FDI/GDP GDP growth GFCF/GDP Trade openness Freedom index techch tfpch effch Interest rate differential
FDI/GDP 1
GDP growth −0.19 1
GFCF/GDP 0.14 0.3 1
Trade openness −0.05 −0.07 0.03 1
Freedom Index −0.02 −0.2 −0.19 0.18 1
techch 0.16 −0.33 −0.11 0.16 0.13 1
tfpch 0.18 −0.21 −0.31 0.16 0.03 0.52 1
effch 0.06 0.09 −0.23 0.01 −0.09 −0.36 0.61 1
Interest rate difference −0.07 0 −0.1 0.11 −0.17 0.02 0.02 0.01 1

Average annual efficiency scores of developed countries

Country effch techch pech sech tfpch
USA 1.002 1.016 1 1.002 1.018
UK 1 1.014 1 1 1.014
Australia 1.028 1.02 1.032 0.996 1.049
Singapore 0.973 1.016 0.983 0.989 0.989
Japan 0.994 1.013 0.995 0.999 1.007
Israel 0.992 1.005 1.004 0.989 0.997
Canada 0.996 1.014 0.995 1.001 1.009
Switzerland 1 1.033 1 1 1.033
New Zealand 1.014 1.012 1.01 1.004 1.026
Germany 1.001 1.004 0.998 1.004 1.005
France 0.995 1.02 0.995 1 1.015
Italy 1.006 1.016 1.006 1 1.022

Notes: Effch, efficiency change; Techch, technical change; Tfpch, total factor productivity change

Source: Author’s own computation following Fare et al. (1994) methodology

Average annual efficiency scores of developing countries

Country effch techch pech sech tfpch
Brazil 1 1.018 1 1 1.018
Russian Federation 1.026 1.052 1.009 1.016 1.079
China 1.045 1.061 1 1.045 1.076
India 0.994 0.99 0.98 1.014 0.984
South Africa 0.992 1.042 1 0.992 1.034
Indonesia 0.985 0.991 0.985 1.001 0.976
Thailand 0.997 0.99 0.997 1 0.987
Malaysia 0.988 1.092 1 0.988 1.079
Philippines 1.029 0.989 1.007 1.022 1.017
Vietnam 1.034 0.988 1 1.034 1.022

Notes: Effch, efficiency change; Techch, technical change; Tfpch, total factor productivity change

Source: Author’s own computation following Fare et al. (1994) methodology

List of developed and developing countries undertaken

Developed countries Code Developing countries Code
USA USA Brazil BRA
UK UK Russian Federation RUS
Australia AUS China CHN
Singapore SGP India IND
Japan JPN South Africa SAF
Canada CAN Indonesia IDN
Switzerland SWL Thailand THA
New Zealand NZL Malaysia MYS
Germany GER Philippines PHL
France FRA
Italy ITA

Notes

1.

Results of all specification tests can be provided on request.

2.

The results of Hausman specification test has been shown in Tables III and IV.

3.

GFCF/GDP is a ratio for developing countries that is higher because of low GDP (constant in US dollar) as compared to developed countries.

4.

Reverse causality is the another name of bidirectional or two-way causality which causes endogeneity.

Appendix 1. Descriptive statistics of FDI in developed and developing countries

Table AI

Table AII

Appendix 2. Correlation matrix of listed variables of FDI in developed and developing countries

Table AIII

Table AIV

Appendix 3. Average annual efficiency scores of developed and developing countries

Table AV

Table AVI

Appendix 4

Table AVII

References

Ahlquist, J.S. (2006), “Economic policy, institutions, and capital flows: portfolio and direct investment flows in developing countries”, International Studies Quarterly, Vol. 50 No. 3, pp. 681-704.

Alam, A. and Zulfiqar Ali Shah, S. (2013), “Determinants of foreign direct investment in OECD member countries”, Journal of Economic Studies, Vol. 40 No. 4, pp. 515-527.

Alam, A., Arshad, M.U. and Rajput, W. (2013), “Relationship of labor productivity, foreign direct investment and economic growth: evidence from OECD countries”, Journal of Business and Management Sciences, Vol. 1 No. 6, pp. 133-138.

Alfaro, L., Kalemli-Ozcan, S. and Volosovych, V. (2008), “Why doesn’t capital flow from rich to poor countries? An empirical investigation”, The Review of Economics and Statistics, Vol. 90 No. 2, pp. 347-368.

Ambos, T., Ambos, B. and Schlegelmilch, B.B. (2006), “Learning from foreign subsidiaries: an empirical investigation of headquarters benefits from reverse knowledge transfers”, International Business Review, Vol. 15 No. 3, pp. 294-312.

Anderson, T.W. and Hsiao, C. (1981), “Estimation of dynamic models with error components”, Journal of the American Statistical Association, Vol. 76 No. 375, pp. 598-606.

Anoruo, E. and Mustafa, M. (2007), “An empirical investigation into the relation of oil to stock market prices”, North American Journal of Finance and Banking Research, Vol. 1 No. 1, pp. 22-36.

Apergis, N., Economidou, C. and Filippidis, I. (2008), “Innovation technology transfer and labor productivity linkages: evidence from a panel of manufacturing industries”, Review of World Economics, Vol. 144 No. 3, pp. 491-508.

Arellano, M. and Bond, S. (1991), “Some tests of specification for panel data: Monte Carlo evidence and an application to employment equations”, The Review of Economic Studies, Vol. 58 No. 2, pp. 277-297.

Aristotelous, K. and Fountas, S. (1996), “An empirical analysis of inward foreign direct investment flows in the EU with emphasis on the market enlargement hypothesis*”, JCMS: Journal of Common Market Studies, Vol. 34 No. 4, pp. 571-583.

Arora, A. and Sharma, C. (2016), “Corporate governance and firm performance in developing countries: evidence from India”, Corporate Governance, Vol. 16 No. 2, pp. 420-436.

Asiedu, E. (2006), “Foreign direct investment in Africa: the role of natural resources, market size, government policy, institutions and political instability”, The World Economy, Vol. 29 No. 1, pp. 63-77.

Asiedu, E. and Lien, D. (2004), “Capital controls and foreign direct investment”, World Development, Vol. 32 No. 3, pp. 479-490.

Athanasoglou, P.P., Brissimis, S.N. and Delis, M.D. (2008), “Bank-specific, industry-specific and macroeconomic determinants of bank profitability”, Journal of International Financial Markets, Institutions and Money, Vol. 18 No. 2, pp. 121-136.

Aw, T.Y. and Tang, T.C. (2010), “The determinants of inward foreign direct investment: the case of Malaysia”, International Journal of Business and Society, Vol. 11 No. 1, pp. 59-76.

Azam, M. and Lukman, L. (2010), “Determinants of foreign direct investment in India, Indonesia and Pakistan: a quantitative approach”, Journal of Managerial Sciences, Vol. 4 No. 1, pp. 31-43.

Aziz, O.G. and Mishra, A.V. (2015), “Determinants of FDI inflows to Arab economies”, The Journal of International Trade and Economic Development, Vol. 25 No. 3, pp. 1-32.

Bahmani-Oskooee, M. and Niroomand, F. (1999), “Openness and economic growth: an empirical investigation”, Applied Economics Letters, Vol. 6 No. 9, pp. 557-561.

Baltagi, B. (2008), Econometric Analysis of Panel Data, Vol. 1, Wiley, New York, NY.

Banga, R. (2006), “The export-diversifying impact of Japanese and US foreign direct investments in the Indian manufacturing sector”, Journal of International Business Studies, Vol. 37 No. 4, pp. 558-568.

Behera, S.R., Dua, P. and Goldar, B. (2012), “Foreign direct investment and technology spillover: evidence across Indian manufacturing industries”, The Singapore Economic Review, Vol. 57 No. 2, pp. 1250011.

Belloumi, M. (2014), “The relationship between trade, FDI and economic growth in Tunisia: an application of the autoregressive distributed lag model”, Economic Systems, Vol. 38 No. 2, pp. 269-287.

Bergten, C.F., Horst, T. and Moran, H. (1978), American Multinationals and American Interests, Brookings Institute, Washington, DC.

Bevan, A., Estrin, S. and Meyer, K. (2004), “Foreign investment location and institutional development in transition economies”, International Business Review, Vol. 13 No. 1, pp. 43-64.

Bilgili, F., Tülüce, N.S.H. and Doğan, İ. (2012), “The determinants of FDI in Turkey: a Markov regime-switching approach”, Economic Modelling, Vol. 29 No. 4, pp. 1161-1169.

Blomstrom, M., Lipsey, R.E. and Ohlsson, L. (1989), “What do rich countries trade with each other? R&D and the composition of US and Swedish trade (No. w3140)”, National Bureau of Economic Research.

Blundell, R. and Bond, S. (1998), “Initial conditions and moment restrictions in dynamic panel data models”, Journal of Econometrics, Vol. 87 No. 1, pp. 115-143.

Bokpin, G.A., Mensah, L. and Asamoah, M.E. (2015), “Foreign direct investment and natural resources in Africa”, Journal of Economic Studies, Vol. 42 No. 4, pp. 608-621.

Borensztein, E., De Gregorio, J. and Lee, J.-W. (1998), “How does foreign direct investment affect economic growth?”, Journal of International Economics, Vol. 45 No. 1, pp. 115-135.

Brewer, T.L. (1993), “Government policies, market imperfections, and foreign direct investment”, Journal of International Business Studies, Vol. 24 No. 1, pp. 101-120.

Cardoso, F.H. and Faletto, E. (1979), Dependency and Development in Latin America, University of California Press, Berkeley, CA.

ÇEviŞ, İ. and Camurdan, B. (2007), “The economic determinants of foreign direct investment in developing countries and transition economies”, The Pakistan Development Review, Vol. 46 No. 3, pp. 285-299.

Chan, M.L., Hou, K., Li, X. and Mountain, D.C. (2014), “Foreign direct investment and its determinants: a regional panel causality analysis”, The Quarterly Review of Economics and Finance, Vol. 54 No. 4, pp. 579-589.

Chang, C.C. and Luh, Y.-H. (2000), “Efficiency change and the growth in productivity: the Asian growth experience”, Journal of Asian Economics, Vol. 10 No. 4, pp. 551-570.

Chatterjee, S. and Price, B. (1991), Regression Diagnostics, Wiley, New York, NY.

Chortareas, G.E., Girardone, C. and Ventouri, A. (2013), “Financial freedom and bank efficiency: evidence from the European Union”, Journal of Banking and Finance, Vol. 37 No. 4, pp. 1223-1231.

Das, S. (1987), “Externalities, and technology transfer through multinational corporations a theoretical analysis”, Journal of International Economics, Vol. 22 Nos 1-2, pp. 171-182.

de Castro, P.G., Fernandes, E.A. and Campos, A.C. (2013), “The determinants of foreign direct investment in Brazil and Mexico: an empirical analysis”, Procedia Economics and Finance, Vol. 5 No. 1, pp. 231-240.

De Vita, G. and Kyaw, K.S. (2008), “Determinants of capital flows to developing countries: a structural VAR analysis”, Journal of Economic Studies, Vol. 35 No. 4, pp. 304-322.

Dees, S. (1998), “Foreign direct investment in China: determinants and effects”, Economics of Planning, Vol. 31 Nos 2/3, pp. 175-194.

Dell’Atti, S., Pacelli, V. and Mazzarelli, G. (2015), “The efficiency of the European banking groups and its determinants”, Managerial Finance, Vol. 41 No. 7, pp. 734-751.

Desbordes, R. and Wei, S.J. (2017), “Foreign direct investment and external financing conditions: evidence from normal and crisis times”, The Scandinavian Journal of Economics, Vol. 119 No. 4, pp. 1129-1166.

Dixon, W.J. and Boswell, T. (1996), “Dependency, disarticulation, and denominator effects: another look at foreign capital penetration”, American Journal of Sociology, Vol. 102 No. 2, pp. 543-562.

Dua, P. and Garg, R. (2015), “Macroeconomic determinants of foreign direct investment: evidence from India”, The Journal of Developing Areas, Vol. 49 No. 1, pp. 133-155.

Dunning, J.H. (1998), “Location and the multinational enterprise: a neglected factor?”, Journal of International Business Studies, Vol. 29 No. 1, pp. 45-66.

Dunning, J.H. and Lundan, S.M. (2008), Multinational Enterprises and the Global Economy, Edward Elgar Publishing.

Dwivedi, A. (2012), “Effect of FDI and trade on productivity in Indian electronics firms”, The Economic Journal, Vol. 60 No. 3, pp. 76-90.

Elkins, Z., Guzman, A.T. and Simmons, B.A. (2006), “Competing for capital: the diffusion of bilateral investment treaties, 1960-2000”, International Organization, Vol. 60 No. 4, pp. 811-846.

Ersoy, İ. and Erol, K.O. (2016), “The impact of Euro-zone crisis on foreign direct investment inflows in the EU-15”, Global Business and Economics Review, Vol. 18 No. 2, pp. 216-226.

Fare, R., Grosskopf, S., Norris, M. and Zhang, Z. (1994), “Productivity growth, technical progress and efficiency change in industrialized countries”, The American Economic Review, Vol. 84 No. 1, pp. 66-83.

Favarra, G. (2003), “An empirical reassessment of the relationship between finance and growth”, No. 3-123, International Monetary Fund, pp. 14-16.

Fetscherin, M. (2010), “The determinants and measurement of a country brand: the country brand strength index”, International Marketing Review, Vol. 27 No. 4, pp. 466-479.

Findlay, R. (1978), “Relative backwardness, direct foreign investment, and the transfer of technology: a simple dynamic model”, Quarterly Journal of Economics, Vol. 92 No. 1, pp. 1-16.

Fischer, S. (2003), “Globalization and its challenges”, American Economic Review, Vol. 93 No. 2, pp. 1-30.

Fischer, S. (1998), “The Asian crisis: a view from the IMF”, Journal of International Financial Management & Accounting, Vol. 9 No. 2, pp. 167-176.

Freckleton, M., Wright, A. and Craigwell, R. (2012), “Economic growth, foreign direct investment and corruption in developed and developing countries”, Journal of Economic Studies, Vol. 39 No. 6, pp. 639-652.

García-Herrero, A., Gavilá, S. and Santabárbara, D. (2009), “What explains the low profitability of Chinese banks?”, Journal of Banking & Finance, Vol. 33 No. 11, pp. 2080-2092.

Ghazal, R. and Zulkhibri, M. (2015), “Determinants of innovation outputs in developing countries: evidence from panel data negative binomial approach”, Journal of Economic Studies, Vol. 42 No. 2, pp. 237-260.

Globerman, S. and Shapiro, D. (2003), “Governance infrastructure and US foreign direct investment”, Journal of International Business Studies, Vol. 34 No. 1, pp. 19-39.

Goldar, B. (2004), “Indian manufacturing: productivity trends in pre-and post-reform periods”, Economic and Political Weekly, Vol. 39 Nos 46/47, pp. 5033-5043.

Goswami, C. and Saikia, K.K. (2012), “FDI and its relation with exports in India, status and prospect in north east region”, Procedia-Social and Behavioral Sciences, Vol. 37, pp. 123-132.

Gourinchas, P.O. and Jeanne, O. (2006), “The elusive gains from international financial integration”, The Review of Economic Studies, Vol. 73 No. 3, pp. 715-741.

Granger, C.W.J. (1969), “Investing causal relations by econometric and cross-spectral methods”, Econometrica, Vol. 37 No. 1, pp. 424-438.

Greene, J. and Villanueva, D. (1991), “Private investment in developing countries: an empirical analysis”, IMF Staff Papers, Vol. 38 No. 1, pp. 33-58.

Hanousek, J., Kočenda, E. and Maurel, M. (2011), “Direct and indirect effects of FDI in emerging European markets: a survey and meta-analysis”, Economic Systems, Vol. 35 No. 3, pp. 301-322.

Hausman, J.A. and Taylor, W.E. (1981), “Panel data and unobservable individual effects”, Econometrica: Journal of the Econometric Society, Vol. 49 No. 6, pp. 1377-1398.

Hein, S. (1992), “Trade strategy and the dependency hypothesis: comparison of policy, foreign investment, and economic growth in latin America and East Asia”, Economic Development and Cultural Change, Vol. 40 No. 3, pp. 495-521.

Henisz, W. and Swaminathan, A. (2008), “Introduction: institutions and international business”, Journal of International Business Studies, Vol. 39 No. 4, pp. 537-539.

Herzer, D. and Klasen, S. (2008), “In search of FDI-led growth in developing countries: the way forward”, Economic Modelling, Vol. 25 No. 5, pp. 793-810.

Hsiao, F.S. and Hsiao, M.C.W. (2006), “FDI, exports, and GDP in East and Southeast Asia – panel data versus time-series causality analyses”, Journal of Asian Economics, Vol. 17 No. 6, pp. 1082-1106.

Inkmann, J. (2000), “Finite sample properties of one-step, two-step and bootstrap empirical likelihood approaches to efficient GMM estimation”, CoFE Discussion Paper No. 00/03, Universität Konstanz, pp. 15-23.

Jadhav, P. (2012), “Determinants of foreign direct investment in BRICS economies: analysis of economic, institutional and political factor”, Procedia-Social and Behavioral Sciences, Vol. 37 No. 2012, pp. 5-14.

Jensen, N.M. (2003), “Democratic governance and multinational corporations: political regimes and inflows of foreign direct investment”, International Organization, Vol. 57 No. 3, pp. 587-616.

Johanson, J. and Mattsson, L.G. (2015), Internationalisation in Industrial Systems – A Network Approach, Palgrave Macmillan, London.

Kinoshita, Y. and Campos, N.F. (2003), “Why does FDI go where it goes? New evidence from the transition economies (No. 3-228)”, International Monetary Fund.

Kojima, K. (1978), Direct Foreign Investment: A Japanese Model of Multinational Business Operations, Vol. 116, Croom Helm, London, pp. 388-390.

Kojima, K. (1982), “Macroeconomic versus international business approach to foreign direct investment”, Hitotsubashi Journal of Economics, Vol. 23 No. 1, pp. 1-19.

Kok, R. and Acikgoz Ersoy, B. (2009), “Analyses of FDI determinants in developing countries”, International Journal of Social Economics, Vol. 36 Nos 1/2, pp. 105-123.

Lansbury, M., Pain, N. and Smidkova, K. (1996), “Foreign direct investment in central Europe since 1990: an econometric study”, National Institute Economic Review, Vol. 156 No. 1, pp. 104-114.

Leitão, N.C. (2010), “Foreign direct investment: the Canadian experience”, International Journal of Economics and Finance, Vol. 2 No. 4, pp. 82-88.

Leitão, N.C. and Faustino, H.C. (2010), “Determinants of foreign direct investment in Portugal”, The Journal of Applied Business and Economics, Vol. 11 No. 3, pp. 19-26.

Levin, A., Lin, C.F. and Chu, C.S.J. (2002), “Unit root tests in panel data: asymptotic and finite-sample properties”, Journal of Econometrics, Vol. 108 No. 1, pp. 1-24.

Levin, R. (2001), “International financial liberalization and economic growth”, Review of International Economics, Vol. 9 No. 4, pp. 688-702.

Li, X. and Liu, X. (2005), “Foreign direct investment and economic growth: an increasingly endogenous relationship”, World Development, Vol. 33 No. 3, pp. 393-407.

Majumder, S.B. and Nag, R.N. (2015), “Capital flow in India: facts, features, and analysis”, Decision, Vol. 42 No. 1, pp. 19-31.

Masron, T.A. (2017), “Relative institutional quality and FDI inflows in ASEAN”, Journal of Economic Studies, Vol. 44 No. 1, pp. 115-137.

Masuku, M.B. and Dlamini, T.S. (2009), “Determinants of foreign direct investment inflows in Swaziland”, Journal of Development and Agricultural Economics, Vol. 1 No. 5, pp. 177-184.

Mina, W. (2009), “External commitment mechanisms, institutions, and FDI in GCC countries”, Journal of International Financial Markets, Institutions and Money, Vol. 19 No. 2, pp. 371-386.

Miskinis, A. and Juozenaite, I. (2015), “A comparative analysis of foreign direct investment factors”, Ekonomika, Vol. 94 No. 2, p. 7.

Moran, T.H. (1998), Foreign Direct Investment and Development, Institute for International Economics, Washington, DC, pp. 3-55.

Neumayer, E. and Spess, L. (2005), “Do bilateral investment treaties increase foreign direct investment to developing countries?”, World Development, Vol. 33 No. 10, pp. 1567-1585.

Ng, T.H. (2007), “Foreign direct investment and productivity: evidence from Sub-Saharan Africa”, unpublished paper, UNIDO, Vienna.

Nyen Wong, K. and Cheong Tang, T. (2011), “Foreign direct investment and employment in manufacturing and services sectors: fresh empirical evidence from Singapore”, Journal of Economic Studies, Vol. 38 No. 3, pp. 313-330.

Onyeiwu, S. and Shrestha, H. (2004), “Determinants of foreign direct investment in Africa”, Journal of Developing Societies, Vol. 20 Nos 1/2, pp. 89-106.

Piteli, E.E. (2010), “Determinants of foreign direct investment in developed economies: a comparison between European and non-European countries”, Contributions to Political Economy, Vol. 29 No. 1, pp. 111-128.

Pradhan, A.K. and Kelkar, S. (2014), “Macroeconomic determinants of foreign direct investment in India: an empirical investigation (1991-2012)”, Journal of Commerce and Management Thought, Vol. 5 No. 4, pp. 530-544.

Qian, G., Khoury, T.A., Peng, M.W. and Qian, Z. (2010), “The performance implications of intra‐and inter‐regional geographic diversification”, Strategic Management Journal, Vol. 31 No. 9, pp. 1018-1030.

Rangkakulnuwat, P. and Paweenawat, S.W. (2015), “The determinants of foreign direct investment in ASEAN”, International Journal of Social Economics, Vol. 42 No. 3, pp. 239-250.

Reinhart, C. (2005), “Some perspective on capital flows to emerging market economies”, NBER Reporter Online, Summer, pp. 8-11.

Sambharya, R.B. and Rasheed, A.A. (2015), “Does economic freedom in host countries lead to increased foreign direct investment?”, Competitiveness Review, Vol. 25 No. 1, pp. 2-24.

Scott, B.R. (1997), “How do economies grow?”, Harvard Business Review, May, pp. 156-164.

Shamsuddin, A.F. (1994), “Economic determinants of foreign direct investment in less developed countries”, The Pakistan Development Review, pp. 41-51.

Simar, L. and Wilson, P.W. (2007), “Estimation and inference in two-stage, semi-parametric models of production processes”, Journal of Econometrics, Vol. 136 No. 1, pp. 31-64.

Singhania, M. and Gupta, A. (2011), “Determinants of foreign direct investment in India”, Journal of International Trade Law and Policy, Vol. 10 No. 1, pp. 64-82.

Sun, Q., Tong, W. and Yu, Q. (2002), “Determinants of foreign direct investment across China”, Journal of International Money and Finance, Vol. 21 No. 1, pp. 79-113.

Temiz, D. and Gökmen, A. (2014), “FDI inflow as an international business operation by MNCs and economic growth: An empirical study on Turkey”, International Business Review, Vol. 23 No. 1, pp. 145-154.

Thiam, N. (2006), “Foreign direct investment and productivity: evidence from east Asian economies”, Staff Working Papers No. 3, United Nations Industrial Development Organisation, Research and Statistic Branch, Vienna, pp. 1-22.

Tsai, P.L. (1994), “Determinants of foreign direct investment and its impact on economic growth”, Journal of Economic Development, Vol. 19 No. 1, pp. 137-163.

UNCTAD (2006), “World investment report 2006, FDI from developing and transition economies: implications for development”, United Nations, New York, NY and Geneva.

UNCTAD (2014), “World investment report 2014”, New York, NY.

Wang, J.Y. and Blomstrom, M. (1992), “Foreign investment and technology transfer: a simple model”, European Economic Review, Vol. 36 No. 1, pp. 137-155.

Wisniewski, T.P. and Pathan, S.K. (2014), “Political environment and foreign direct investment: evidence from OECD countries”, European Journal of Political Economy, Vol. 36 No. 1, pp. 13-23.

Woo, J. (2009), “Productivity growth and technological diffusion through foreign direct investment”, Economic Inquiry, Vol. 47 No. 2, pp. 226-248.

Xiaming, L., Parker, D., Kirit, V. and Yingqi, W. (2001), “The impact of foreign direct investment on labour productivity in the Chinese electronics industry”, International Business Review, Vol. 10 No. 4, pp. 421-439.

Xu, H., Wan, D. and Sun, Y. (2014), “Technology spillovers of foreign direct investment in coastal regions of east China: a perspective on technology absorptive capacity”, Emerging Markets Finance and Trade, Vol. 50 No. S1, pp. 96-106.

Yang, J.Y.Y., Groenewold, N. and Tcha, M. (2000), “The determinants of foreign direct investment in Australia”, Economic Record, Vol. 76 No. 232, pp. 45-54.

Yuqing, X. (2006), “Why is China so attractive for FDI? The role of exchange rates”, China Economic Review, Vol. 17 No. 2, pp. 198-209.

Zhang, Z. (2002), “Productivity and economic growth: an empirical assessment of the contribution of FDI to the Chinese economy”, Journal of Economic Development, Vol. 27 No. 2, pp. 81-94.

Further reading

Chakrabarti, A. (2001), “The determinants of foreign direct investments: sensitivity analyses of cross‐country regressions”, Kyklos, Vol. 54 No. 1, pp. 89-114.

Dell’Atti, L., Fabiani, A., Marconi, A., Mantovani, P. and Muzzonigro, G. (2005), “Reliability of echo-color-Doppler in the differential diagnosis of the ‘acute scrotum’ Our experience”, Archivio Italiano di Urologia, Andrologia: Organo Ufficiale [di] Societa Italiana di Ecografia Urologica E Nefrologica/Associazione Ricerche in Urologia, Vol. 77 No. 1, pp. 66-68.

Ho, O.C. (2004), Determinants of Foreign Direct Investment in China: A Sectoral Analysis, Department of Economics, University of Western Australia, Perth.

Huyen, H.L.B. (2015), “Determinant of the factors affecting foreign direct investment (FDI) flow to Thanh Hoa province in Vietnam”, Procedia – Social and Behavioral Sciences, Vol. 172 No. 1, pp. 26-33.

Khachoo, A.Q. and Khan, M.I. (2012), “Determinants of FDI inflows to developing countries: a panel data analysis”, No. 37278, Pondicherry University, Puducherry.

Masron, T.A. and Masron, T.A. (2017), “Relative institutional quality and FDI inflows in ASEAN countries”, Journal of Economic Studies, Vol. 44 No. 1, pp. 115-137.

Trevino, L.J., Daniels, J.D. and Arbelaez, H. (2002), “Market reform and FDI in Latin America: an empirical investigation”, Transnational Corporations, Vol. 11 No. 1, pp. 29-48.

UNCTAD (1998), “World investment report: trends and determinants”, Table IV.1, New York, NY and Geneva, p. 91.

Xaypanya, P., Rangkakulnuwat, P. and Paweenawat, S.W. (2015), “The determinants of foreign direct investment in ASEAN: the first differencing panel data analysis”, International Journal of Social Economics, Vol. 42 No. 3, pp. 239-250.

Corresponding author

Neha Saini can be contacted at: nehasaini.phd@fms.edu