Economic Policy, Political Constraints, and Foreign Direct Investment in Developing Countries

ABSTRACT

As foreign direct investment (FDI) has become increasingly important in the world economy, a large body of literature has emerged regarding the determinants of FDI flows. Some scholars argue that democracy attracts FDI through the mechanism of political constraints, which reduce the risk of negative policy changes. However, the value of policy stability should be conditional on the attractiveness of contemporary FDI-relevant policies. I therefore propose a theoretically more comprehensive argument: political constraints are attractive to investors when the host country policy environment is FDI-friendly, because these political constraints reduce the probability of negative policy changes in the future. When the policy environment is hostile to FDI, on the other hand, political constraints will have little positive effect, and, to the extent they indicate that FDI-relevant policies are unlikely to improve, may even deter FDI. This argument helps explain why the positive relationship between democracy and FDI seems to emerge after a global shift toward FDI-friendly polices. I find robust empirical support for the argument in tests covering more than 100 developing countries from 1970 to 2014, indicating significant effects using a variety of policy and political constraint measures.

KEYWORDS:

• Foreign direct investment
• multinational corporations
• political institutions
• policy stability

Acknowledgement

I wish to thank Kim Yi Dionne, Allison Woods, Barbara Geddes, Daniel Posner, Ronald Rogowski, Arthur Stein, Kathleen Bawn, Joseph Wright, Daniela Campello, Julia Gray, and anonymous reviewers for helpful comments on this article. An online appendix can be found at https://dataverse.harvard.edu/dataverse/internationalinteractions.

Supplemental infromation

Supplemental data for this article can be accessed on the publisher’s website.

Notes

¹ In manufacturing, on the other hand, multinational corporations retain their bargaining power relative to the host government if, for example, the industry is globally integrated or if intangible assets such as brand are important sources of value.

² To isolate this dynamic from the others, I exclude Soviet Bloc countries during the Cold War, conduct some of the analysis using post-Cold War data exclusively, and isolate specific policies that are independent of democratic institutions.

³ Countries such as Poland and the Czech Republic who joined the OECD and EU midway through the analysis period are included before they join the OECD.

⁴ For example, FDI per capita to Equatorial Guinea (population 665,000) swung from 1,720 to −788 to 2,360 in the years 2007-2009.

⁵ I use the IMF list of OFCs at http://www.imf.org/external/np/mae/oshore/2000/eng/back.htm. My appreciation to an anonymous referee for calling my attention to the need to exclude cases where holding companies make up the majority of FDI inflows.

⁶ For the 5-year period measure, if 1 year of data is missing in a period, I calculate a 4-year average in order to reduce missingness.

⁷ As in Busse and Hefeker (2007), FDI per capita is transformed using the following procedure: , excluding outliers, defined as observations where the dependent variable is more than three standard deviations from the mean.

⁸ All data, except for GATT/WTO membership, FDI stock, Investment Promotion Agencies, and Capital Account Openness are conveniently provided by Teorell et al. (2016). The GATT/WTO memberships were coded using the WTO website. In robustness checks in the online Appendix, FDI stock comes from the United Nations Conference on Trade and Development (UNCTAD); Capital Account Openness (KA OPEN) comes from Chinn and Ito (2008); and Investment Promotion Agency dummy comes from Harding and Javorcik (2011).

⁹ Results using the other EFW and EF components and for Capital Account Openness index (KA OPEN) from Chinn and Ito (2008) are available in the online Appendix.

¹⁰ They also test other investment-related treaties. To simplify the analysis, I only include GATT/WTO membership.

¹¹ The latter four are found in the online Appendix. Harding and Javorcik (2011) find that investment promotion agencies (DIPA) increase FDI flows to developing countries by “providing information (on business opportunities, prevailing laws and regulations as well as factor cost in a host country) and helping foreign investors to deal with bureaucratic procedures.” Moosa and Cardak (2006) and others find that telephones and exports are robust predictors of FDI.

¹² In robustness tests, I introduce country fixed effects, a quadratic time trend, and use a GLS model with Random Effects.

¹³ I use the package in Stata, with the kernel smoothing estimator. The control variables are the same as in Table 1, with standard errors clustered by country.

¹⁴ This is an example of an increase from 1-3 in the 7-point Constraints on the Executive scale.

¹⁵ One potential concern in this analysis is that past FDI may influence policy to become more FDI-friendly, and also attract future FDI inflows. The results from the analysis including FDI stock as a control help address these concerns.