The effects of economic freedom on FDI inflows: an empirical analysis

ABSTRACT

There is a regular emphasis on the significant role of inward Foreign Direct Investment (FDI) in promoting economic growth. This favourable relationship has induced many governments to adopt policies intended to increase FDI inflows and, thereby, to create conducive business and economic conditions for Multinational Enterprises (MNEs). This paper examines the effects of Economic Freedom (EF) and its sub-components reflecting the Quality of Institutions (QIs) on FDI inflows, using indices derived from the Fraser Institute and from the Heritage Foundation. The empirical analysis is carried out for a panel dataset using different econometric methodologies and empirical specifications. The results underline positive effects of EF on FDI inflows. They reveal that EF sub-components have varying impacts on FDI inflows, where rule of law, market openness, and less-restrictive regulatory environment stand out as the major FDI-promoting institutional factors. Also, there is an empirical evidence that the effects of EF sub-components on FDI inflows exhibit variations through the economic characteristics of the host countries and across geo-economic regions. The results suggest that governments should pursue EF-improving policies, which should be tailored according to the economic and geo-economic characteristics of the host countries, to increase FDI inflows.

KEYWORDS:

• Economic freedom
• foreign direct investment
• institutions
• multinational enterprises

JEL CLASSIFICATION:

• F21
• F23
• O43

Disclosure statement

No potential conflict of interest was reported by the authors.

Notes

¹ Inward FDI could also promote economic growth by generating backward and forward linkages through the development of business networks between domestic firms and foreign affiliates of MNEs (Ezemenari, Tiruneh, and Wamboye 2016).

² Saltz (1992) finds that FDI could exercise a negative impact on economic growth in developing countries, including those that lessened the constraints on the repatriation of profits. She displays empirical evidence that FDI increases the income of the richest developed countries, but tends to decrease the income of developing countries, particularly those with the lowest income. She notes that FDI raises the price of capital, and tends to crowd-out domestic investment.

³ In this context, Gwartney (2009) points out that private investment, including FDI, is relatively higher in countries with higher levels of EF, whereas governmental investment is relatively higher in countries with lower levels of EF. These observations emphasize the entrepreneurial choices in selecting the locations of productive activities. Thus, countries with better institutions tend to attract more private investment, including FDI. Also, Gwartney, Holcombe, and Lawson (2006) find that higher levels of EF are associated with higher levels of investment productivity, as expressed through stronger impacts of private investment on economic growth. Gwartney (2009) notes that better QIs would eventually lead to faster economic growth and higher income per capita levels. Also, Azman-Saini, Baharumshah, and Law (2010) report that the impact of inward FDI on economic growth is accelerated with higher levels of EF in the host country.

⁴ Other studies investigate the role of EF/QIs in catalysing knowledge spillovers from inward FDI and, hence, in promoting economic growth (e.g. Fosfuri, Motta, and Rønde 2001; Javorcik 2004; Azman-Saini, Baharumshah, and Law 2010).

⁵ Some studies (e.g. Cebula and Clark 2011; Cebula 2014) indicate that migration is significantly inclined toward economic locations (countries, states) that enjoy higher levels of economic and personal freedom.

⁶ Mathur and Singh (2013) show that foreign investors and MNEs are more concerned about economic freedom than political freedom in formulating and undertaking their international investment decisions. These outcomes are associated with the difficulties encountered through the democratization process in developing countries, particularly when implementing FDI-promoting economic reforms due to competing political interests.

⁷ The Fraser Institute’s EF and EF sub-component indices were first published in the year 1970, and they have been reported over a five-year interval up to the year 2000. Afterwards, these indices have been reported annually. The Heritage Foundation’s EF and EF sub-component indices have been reported annually starting in the year 1995.

⁸ Heritage Foundation’s Rule of Law, Market Openness, Regulator Efficiency, and Government Size roughly correspond to Fraser Institute’s Legal System, Freedom to Trade, Regulation, and Size of Government, respectively.

⁹ Justesen (2008) indicates that weak property rights’ systems would lead to distortions in the allocation of resources, and would encourage immediate consumption at the expense of long-term investment.

¹⁰ In this context, Gwartney (2009) finds that more open economies tend to have higher rates of private investment. This is because lower transportation and communication costs have led to more selective locations of production facilities. As a result, more open economies would naturally become more attractive for private investment, including FDI.

¹¹ Gwartney (2009) notes that heavy regulations would eventually lead to higher corruption levels, since they place officials in a position to grant favours or impose costs on businesses.

¹² The overall EF indices of the Fraser Institute and the Heritage Foundation exhibit a high level of correlation that reaches 0.82. Also, the correlation coefficients among the corresponding sub-components are generally high. For example, the correlation coefficients between Legal System and Rule of Law, and between Freedom to Trade and Market Openness are 0.84 and 0.72, respectively. It is worth noting that the correlation coefficient between Regulation and Regulatory Efficiency is comparatively lower, standing at 0.55.

¹³ EF indices for Cuba and North Korea are missing from the Fraser Institute’s dataset.

¹⁴ Alternatively, some studies examine the determinants of sales of foreign affiliates of MNEs (e.g. Brainard 1997; Ghazalian and Furtan 2008, 2009).

¹⁵ For instance, Tobin and Rose-Ackerman (2011) indicate that Bilateral Investment Treaties (BITs) are more effective in attracting inward FDI when the quality of domestic institutions is adequate.

¹⁶ Population size is often used as a proxy for domestic market size to dampen endogeneity concerns associated with the use of RGDP measures (Habib and Zurawicki 2002). Also, these market size variables are complemented with the inclusion of the RGDP per Capita (RGDPC) to account for consumption potentials.

¹⁷ Some empirical studies found that the importance of domestic market size in determining inward FDI has decreased over time, particularly in the case of developed host countries (see, for example, Loree and Guisinger 1995). Blomström and Kokko (2003) note that globalized trade have allowed MNEs to set international production networks that serve foreign markets or foreign affiliates located outside the host country, resulting in lower significance of domestic market size in the decision-making of MNEs about the location of their foreign investments.

¹⁸ Higher economic growth rates could also signal steady improvements in the quality of infrastructure.

¹⁹ It is worth noting that inflation is used in the construction of the Fraser Institute’s Sound Money. The inflation rate is included in the benchmark empirical equations, and the results are subsequently checked when excluding this variable from the empirical equations.

²⁰ Following Bénassy‐Quéré, Coupet, and Mayer (2007), five-year lag is selected through the benchmark empirical analysis. Also, this selection is in accordance with the fact that Fraser Institute’s EF indices were recorded in a five-year frequency before 2000. Alternative lags are also examined through the empirical analysis.

²¹ It is worth noting that Bénassy‐Quéré, Coupet, and Mayer (2007) carried out the empirical analysis using bilateral observations. They constructed the predicted residual ${\hat{v}}_{ij}=\left|{\hat{v}}_i-{\hat{v}}_j\right|$, and defined the 3 × 1 vector $V_{ij}=\left({\hat{v}}_i,{\hat{v}}_j,{\hat{v}}_{ij}\right)$, which is interpreted as the institutional ‘qualities’ and ‘distance’ not related to RGDPC.

²² Clustered standard errors are associated with consistent estimates across a broad range of possible forms of correlation, but they may not be efficient when the exact form of correlation is known. Meanwhile, random effects model yields more efficient estimates for a correct form of correlation, but it produces biased estimates and standard errors when the form of correlation is incorrectly specified.

²³ It is worth noting that the size of the panel dataset is based on the availability of observations for the key variables under consideration. The countries covered in our dataset are listed in Table A1 of the Appendix.

²⁴ This approach is also used to attenuate potential endogeneity concerns that could prevail between FDI inflows and some explanatory variables (Egger and Winner 2005).

²⁵ This approach is consistent with the recommendations of Imbens and Wooldridge (2007).

²⁶ It is worth noting that the moderate differences between the estimates on the Fraser Institute’s index and those on the Heritage Foundation’s index could be partly attributed to the differences in the time periods that are covered in the corresponding regressions (1970–2015 and 1995–2016, respectively). Subsequent regressions will shed light on this point.

²⁷ For instance, in the case of the Fraser Institute’s EF sub-components, the correlation coefficients of Freedom to Trade with Regulation and with Legal System are 0.64 and 0.62, respectively. In the case of the Heritage Foundation’s EF sub-component indices, the corresponding correlation coefficients of Market Openness with Regulatory Efficiency and Rule of Law are 0.71 and 0.73, respectively. The correlation coefficients of Size of Government/Government Size with the other sub-component indices are found to be relatively smaller, falling below 0.35 (in absolute terms).

²⁸ These results complement Gwartney’s (2009) observations, which indicate that poor legal structure, restrictive regulations, and trade barriers significantly contribute to explaining lower income per capita levels in SSA countries.