The Determinants of Foreign Direct Investment (FDI): How Important Are Institutional Factors in Developing Countries?

Since the 1980s, the attractiveness of foreign direct investment has always been one of the growth drivers for most developing countries. Indeed, the debt crises that have shaken these countries have led them to change at least partially their development strategies. Different empirical studies have tried to deal with such an issue. In this paper, we examine empirically the determinants of foreign direct investment (FDI). We employ a sample that covers a panel of 46 developing countries over the period of study from 1996 to 2018. These countries are grouped in four regions such as: Africa Sub-Saharan (12), Middle East and North Africa (6), Latin America Caribbean (20), Asia and Pacific (8). From the empirical results, we show that, in addition to some classic determinants of foreign direct investment attractiveness, institutional factors, particularly those related to economic freedoms, available to investors, are also of great importance. In terms of economic policy implications in our study, developing countries, especially Middle East and North Africa countries, must act on the level of freedoms to progress their incorporation into the global economic system but moreover to address the main contest, refining the standard of living of the population, which remains depreciated in the massive popular of countries. This obviously means creating a satisfactory business climate that can inspire together domestic and foreign investors and therefore generate a dynamics, capable of removing the challenges of development and integration.


Introduction
Despite the growing importance of international investment in economies, and despite (or because of) the impressive literature on this subject, there is no unified theoretical framework for understanding the determinants of FDI.Theoretical analyses on the multinationalization of firms are recent, their appearance dating back to the late 1950s.In fact, it is especially since the turn of the 2000s that the literature on FDI has seen a very strong acceleration, in connection with the development of the phenomenon.
In addition, the analysis of the determinants of FDI is at the intersection of the international economy and the industrial economy.The first allows us to understand these behaviours in their geographical arbitrage dimension and from the perspective of a link between trade in products and movements of capital.The second focuses more on firms' development strategies and the trade-off between various modes of organization of their activities.
FDI has been of interest to developing countries since the1980s, especially those in need of financing sources that can replace loans that have become increasingly difficult to obtain because of the conditionality, imposed by the World Bank and the International Monetary Fund.The attention, generated by FDI, is based on the theoretical arguments, put forward by liberal thought and conveyed by international institutions.
FDI is an indispensable driver of growth and development as it provides benefits for host countries in terms of job creation, human resource training, foreign exchange inflows, competition, and complementarity with domestic enterprises.There was then fierce competition between the developing countries to attract the maximum amount of investment.One of the major questions, facing researchers, but also policy makers at this level, is the determinants of FDI.This question has caused a great deal of ink both theoretically and empirically, and continues to do so.This article is an empirical contribution in this area.
FDI is seen as an indispensable driver of growth and development as it provides benefits and benefits for host countries in terms of job creation, human resource training and entry competition and complementarity with domestic companies.There was then fierce competition between the DEVELOPING countries to attract the maximum amount of investment.One of the major questions facing researchers, but also policy makers and policy makers at this level, is the determinants of FDI.This issue has generated a great deal of ink both theoretically and empirically and continues to do so.This article is an empirical contribution in this area.
In [1] the researcher found that trade GDP, reserve GDP, the exchange rate, the R and DGDP and the FDGI are the main determinants of FDI, and FDI was an important factor, influencing the level of economic growth in India.[2] has been focused on "trends in FDI inflows into India."The researcher found that the services sector is the largest sector and the banking, and insurance sector was the second fragment of what chooses growth during the second decade of reforms.Mauritius and Singapore are the top two nations with the highest FDI in India.
[3] has been analyzed "trends in FDI flow in the Indian era of later reforms and also analyzed the relationship between FDI and the country's economic growth to assess the impact of FDI on the stock market of India".The researcher found that there is a strong positive correlation between FDI and India's GDP growth and found that there is a strong positive correlation between FDI flows and SENSEX BSE movements and that GDP largely depends on FDI inflows into the country.In [4], author analyzed "trends in FDI flows in the country to discover the relationship between FDI, foreign institutional investors (FII) and the country's GDP."The researcher found that there is still a positive trend in FDI in India, a positive relationship between FDI and GDP, where is a low positive correlation between FDI and IIF.
[5] use the Vector Auto Regression (VAR) analysis and found that the combined interest rate, inflation rate and RGDP (economic growth) have a significant positive impact on FDI in Nigeria, while [6] have stated that the exchange rate is significantly correlated with the level of foreign direct investment.[7] use the OLS technique to examine the relationship between independent variables, such as trade opening, real exchange rate, financial market development and dependent variable, FDI in Nigeria.The result indicated that FDI was a positive and significant function of trade opening.The real exchange rate had a positive but not significant relationship with the FDI.
The aim of this paper is to investigate empirically the determinants of foreign direct investment (FDI).We use a sample, composed of a panel of 46 developing countries during the period of study from 1996 to 2018.These countries are grouped in four regions such as: Africa Sub-Saharan (12), Middle East and North Africa (6), Latin America Caribbean (20), Asia and Pacific (8).
From the empirical findings, we find that, in addition to some classic determinants of foreign direct investment attractiveness, institutional factors, particularly those related to economic freedoms, available to investors, are also of great importance in the determinant of FDI.
The rest of the paper is organized as follows: in a second section, and in the light of theoretical and empirical literature, but also some characteristics of the Middle East and North Africa (MENA) region, we discuss the variables able to be integrated into the model.We also show in this section the originality of our work.In a third section, we specify the model to be tested as well as the sources of the data.Finally, in a fourth section, we advance the results of the estimates.

1. Econometric Methodology
The economic literature is marked by an abundance of analyses, aimed at explaining the operations of international business.Foreign direct investment as one of the forms of foreign direct investment and as a much-needed method of implementation in recent decades is no exception.So, theories to explain FDI can be categorized on the basis of several criteria: pro-FDI theories/theories, completely hostile to the operations of multinational firms (this is the reference to liberal analysis/Marxist analysis), theories, advancing macroeconomic factors/theories, suggesting micro-economic variables, theories, favouring localization factors/those, focusing on strategies.
These different theories have highlighted a multitude of factors that can explain FDI flows.Thus, on the one hand, the theories of perfect markets have advanced the rate of return, portfolio diversification, market size.On the other hand, imperfect market theories have advocated other types of factors in this case the specific advantage, the internalization advantage and the localization factors.Dunning (1981) considered through a synthetic theory from the early 80s that one can speak of FDI, only if the last three factors are combined, namely, the specific benefits, the benefits of internalization and finally the attractive territory characteristics.
Other theories have focused on the notion of risk (political, financial and economic) as the basis for any explanation of the distribution of FDI around the world.
However, what we can see is that, on the one hand, there is not a single theoretical basis, capable of explaining all FDI flows.On the other hand, we have no idea about the relative importance of the various factors advanced.In the following, we will discuss, the variables, selected as well as the specificity of our model in relation to the existing work in this area.At the end of the theoretical analyses on FDI, and the various studies, carried out in the MENA region, we can consider the following factors that can be grouped together in this way: 1.The control variables selected

Growth prospects
The large influx of FDI to dynamic economies within the global system, in this case some countries in Asia, Latin America and more recently China, supports the idea of a very likely link between internal dynamism or growth prospects and investment attraction.Lim (1983) considered that the strong economic growth in a country offers better opportunities for investors than a low-growth or non-growth country.Growth prospects also give a company a signal about the potential demand for its products and/or services.
The growth rate of real gross domestic product per capita (GROWTH it ) can be used as indicators in our specification.We retain this variable, following [8][9][10][11].Indeed, the first study of a panel benchmarking of 36 PVDs over the period 1975-2006 found a positive and significant correlation in the growth rate.Similarly, the second study reached the same conclusion, based on a cross-sectional analysis of 38 DEVELOPING countries over the 2000-2004 period.Finally, the third, based on a cross-sectional analysis of 18 MENA countries and 35 regressions, revealed that the growth rate is one of the most significant variables (3/4 of the 35 regressions).

Domestic investment
We retain this variable, following [10].Domestic investment gives an idea of the growth of the capital stock in a host country.We use as a proxy the gross formation of fixed capital (FBCF).

Macroeconomic stability
It is a defining one of the pillars of the competitiveness of nations.KEYNES considers that inflation confiscates the wealth of economic agents.To take into account the macro-economic situation, which has a considerable impact on investment operations, we take into account the level of inflation, estimated by the change in the consumer price index and the country's external solvency [total external debt relative to GDP (% EDETT)].
In addition to these control variables that have proven significant in most empirical FDI studies, the ultimate goal of our work is to test the significance of two categories of indicators, referred to as variables of interest, in 46 PVDs, including six MENA countries.

The interest variables selected
The business climate or economic freedom It is often said, that a country's economic performance is intimately linked to the business environment of economic agents.Indeed, the costs, incurred by starting business operations, differ from one country to another.They depend to a large extent on the quality of governance (institutions, policies).
The World Bank's studies, since 2004 in the "Business Practices" project clearly show that poor countries with huge development needs are also the ones that pose more barriers to business practice.Thus, according to the 2005 report, these countries are characterized by administrative costs three times higher than rich countries, administrative delays twice as high and half as many guarantees for the protection of property rights.To measure this business climate in our study, we use the Economic Freedom Index: 'Index of Economic Freedom' (IEF).The latter, as defined by its designers, allows an assessment of the obstacles, faced by economic agents, by continuing their actions within an economy.
It is, in fact, a synthetic index, based on the analysis of fifty independent economic variables and grouped into ten categories (trade policy, government tax burden, regulation, the informal sector, prices and wages, finance and banking, capital flows, property rights, monetary policy and state intervention in the economy).The index aims to describe the realities and business climate in 161 countries of the world as they are and does not purport to determine optimal policies.
The index also reflects to some extent the level of economic and political risk in each country, as risks are generally high in countries, where institutions are unable to act favourably on the prevailing climate in the economy.
To complement the various aspects of governance, mentioned above, we can introduce two other indices into the model.This is the index of political freedoms and the civil liberties index that will allow us to test the impact of individual and collective freedoms on IDE integration.We consider that the weaker the latter (characteristics of non-democratic regimes), the less creativity there is and the less incentive for agents to invest, given the transaction costs that arise.

Political and civil liberties
We use FREEDOM HOUSE (FH) indicators to make a judgment on democracy in our sample.These are two indices 'Civil Liberties' (CL it ) and 'Political rights' (PR it ) that reflect political rights and civil liberties.They range from 1 to 7. Index values are negatively correlated with the level of freedoms and rights.The FREEDOM HOUSE indicators will then allow us to complete the economic freedom index and thus test the contribution of freedom in the broadest sense in terms of investment in our sample of countries.This notion of freedom in the broadest sense allows a large measure of the business climate, governance, the place of the state, its actions, its attitudes, its effectiveness, and the place of risk in the economies in question.

The World Bank's governance indicators
To test the robustness of freedom indicators, we introduce Kaufmann's governance indicators into our modelling.We believe that both reflect the same reality.Such indicators address institutional aspects that may constitute benefits or barriers to investment.These include the effectiveness of government, the rule of law, regulation, corruption, political stability, and political and civil freedoms.Indicators vary between -2.5 and 2.5 with high values, synonymous with a better quality of the aspect in question.
In conclusion, our specification differs from the studies, carried out on the region, by considering the issue of economic freedom, but also political and civil freedom.We consider freedom in the broadest sense to be a necessary condition for business practice and its level in each country largely reflects the quality of governance and the business climate.Logically, the freer the country is, the more active economic agents are, the more incentives to invest as they face lower transaction costs, compared to a constrained non-free environment.We then test the contribution of this hypothesis in the determination of FDI.We then try to test the hypothesis that the attractiveness of countries differs according to several location characteristics.We consider that among the most important is the business climate or the level of freedoms that prevails in a specific territory.

The specification of the model and data sources
In our case, it is a question of applying Panel econometrics to compare the significance of our interest variables in our group of countries.In total, the analysis covers 46 Developing countries, including 6 MENA countries (Table 1).The selected period is 1996-2018.Table 2, for example, defines the different variables as well as the sources of the underlying data: where C is the constant, a 1 ,..., a 8 present the coefficients of explanatory variables, i designs the individual dimension or countries with i=1,..., 46, t measures the annual temporal dimension with t=1, ..., 23, and α it is the term error.

Results
Modeling, based on panel data, allows for at least two main advantages.On the one hand, a greater number of observations, allowed by the two individual and temporal dimensions of the analysis.This allows, among other things, better statistical inference.On the other hand, the Panel model specification allows consideration of the possible heterogeneity in the individual and/or temporal dimension.This ultimately avoids or at least minimizes the omission bias of certain variables.Before going any further in our econometric approach, we represent the correlation matrix of our variables before starting the modelling itself.

1. The correlation matrix
As can be seen, the Economic Freedom Index, which reflects the investment climate or the obstacles, faced by economic agents in a country that is well-defined in their actions, is negatively linked to the percentage of net investment flows received (Table 3).The intuition behind this correlation is that the more economically free the country is (the index is small), the higher the percentage of net investment flows received.It should be noted, however, that some countries, despite their non-free nature, receive significant FDI flows, particularly in the hydrocarbon and mining sectors.As a result, the share, received by these economies, is not entirely proportional to the level of the business climate they display.
The second observation, displayed by the correlation matrix above, is that the more integrated the country is from an international trade point of view, the more integrated from the less foreign direct investment view.In other words, trade and FDI go hand in hand.This correlation is consistent with several empirical analyses of the determinants of FDI, which conclude that the most open countries, particularly from a trade point of view, are more attractive than the least open countries.The third observation noted shows that the freer the country, the higher its level of GDP/h.Unlike previous relationships, the latter is much clearer.This is probably since the impact of the level of economic freedoms in wealth creation is not only through FDI but also through domestic investment.
The latest observation shows that the level of civil and political freedoms is negatively correlated with the level of wealth (GDP/h), but the correlation seems less important in comparison with that of economic freedoms.In summary, and despite the intuitions provided, the correlation matrix unfortunately does not give enough information on the meaning of established correlations.Therefore, econometric analysis is essential.

2. 1. Estimate of the fixed-effect and random-effect model
There is clearly a large difference between the parameters estimated, based on different assumptions about the nature of individual heterogeneity (random or fixed).It is also found, that the parameters are more significant in the random effect model.To choose, we perform the HAUSMAN test.

2. 2. Test d'HAUSMAN
The HAUSMAN test shows that the differences in parameters between the two models are real.The most appropriate model is the random effect model.We correct heteroscedasticity and self-correction of errors that impair parameter efficiency.Before interpreting the different results, we begin several specifications and see the significance of our freedom variables with and without a certain number of control variables, advanced in the empirical literature as determinants of FDI. 4, as well, summarizes the results obtained.The following conclusions can be drawn: suivantes: the first specification shows that economic freedoms have a considerable effect on net flows of foreign direct investment.Indeed, any improvement in the economic freedoms index of 1 point allows an improvement in the investment ratio of 1.94 points.In the second and third specifications, we have successively added an indicator of civil liberties and political freedoms.The results show very significant parameters (-0.54 and -0.34 respectively).However, it should be noted in comparison with economic freedoms, that these are more decisive in terms of the impact on net FDI flows.In the fourth and fifth specifications, respectively, we introduced several variables into a model of freedom levels (advanced indicators very frequently in the empirical literature as determinants of FDI).These include yield, economic openness, macroeconomic stability (inflation, external debt), physical infrastructure and the level of financial development.The aim is to see to what extent our freedom variables remain significant as determinants of FDI flows after the introduction of control variables.The results show a very high significancee of our indicators as well as the control variables except those, relating to macroeconomic stability and financial development.In the sixth and seventh specifications, we removed non-significant variables.

2. The results of the analysis Table
To test the robustness of our results, we will study the correlations between interest indicators and the World Bank's governance indices in 2000, 2010 and 2015.

2. 4. The robustness of the results
According to Tables 5-7, it appears from the matrix of correlations that the index of economic freedoms in 2000, 2010 and 2018, ishighly correlated with the quality of regulation, the effectiveness of government and the rule of law while indicators of civil liberties and policies are highly correlated with the indicator 'voice and accountability'.

Discussion
In conclusion, the analysis of FDI integration in 46 developing countries over the period 1996-2018, six of which belong to the MENA region, leads to the fact that institutional factors, particularly those that directly determine levels of economic, but also political and civil freedoms, major components of the business climate, prevailing over a well-defined territory, have a considerable impact on net FDI flows.They are fundamental territory characteristics in the same way as other characteristics, related to the fundamentals of a country, and which have been emphasized in the framework of the theoretical and empirical literature of FDI.These include physical infrastructure, human infrastructure, macroeconomic stability, trade ratio, yield.
In terms of economic policy implications, developing countries, especially MENA countries, must act on the level of freedoms to improve their integration into the global economic system but also to address the major challenge, improving the standard of living of the population, which continues to deteriorate in the vast majority of countries.This obviously means creating a favourable business climate that can encourage both domestic and foreign investors and therefore create a dynamics, capable of removing the challenges of development and integration.

Table 2
Selected Variables and Data Sources

Table 3
The Correlation Matrix

Table 4
The Dependent Variable: Net FDI-to-GDP Flows Note: The sign means (***) a 99 % confidence level.Source: Elaborated by authors

Table 5
The Correlation Matrix for 2000

Table 7
The Correlation Matrix for 2018