The Politics of Foreign Direct Investment in Authoritarian Regimes

ABSTRACT

While political scientists find that democracy reduces political risk, little scholarship analyzes how authoritarian regimes attract foreign direct investment (FDI). This article argues that while authoritarian countries are generally risky, this risk can be minimized when authoritarian regimes are constrained from both “above” and “below.” Signing international investment treaties are critical for authoritarian countries to signal a commitment to FDI-friendly policies. However, only authoritarian signatories that allow some degree of public deliberation in their policymaking are then constrained from deviating from the policies of the international investment treaties. Panel statistical regressions and a case analysis support the hypothesis.

KEYWORDS:

• Foreign direct investment
• international political economy

Acknowledgments

I would like to sincerely thank Nita Rudra, Luke Condra, Jude Hays, and Kevin Morrison for their thoughtful, insightful, and comprehensive comments. I would also like to thank Scott Cook, Stephen Chaudoin, Matthew Fails, Steve Finkel, Ben Graham, Ryan Gruaer, Pierre Landry, Anibal Pérez-Liñán, Shawna Metzger, Sarah Patton, Evgeny Postnikov, Peter Rosendorff, Dan Tirone, Carey Treado, the Department of Political Science at Fordham University, the participants of the Doctoral Research Symposium at the University of Pittsburgh’s Graduate School for Public and International Affairs, and four anonymous reviewers for their feedback.

Funding

The University of Pittsburgh’s International Business Center and Fordham University’s Faculty Research Grant provided funding for the survey research.

Notes

¹ Please note that oil-rich states make up less than 20% of the authoritarian countries receiving FDI inflows.

² A debate exists on the effectiveness of BITs. For example, Yackee (2008) argues that BITs do not promote greater investment inflows. To my knowledge, Rosendorff and Shin (2011) are the only scholars who assess the effectiveness of BITs in authoritarian countries, and they do find a positive relationship with FDI inflows.

³ The strength of this signal increases as countries enter into more treaties because of an increase in the visibility of (and information flows on) a dedication to FDI friendly policies (Allee and Peinhardt 2011).

⁴ Yackee (2010) provides evidence that BITs are not associated with reductions in political risk.

⁵ Jensen et al. (2014) do find a statistically significant correlation between authoritarian legislatures and corporate government regulations or investor protections.

⁶ Linz (2000) describes authoritarian and totalitarian regimes: The former are characterized by limited pluralism and the latter repression. Leaders choose these different survival methods based on their revenue sources and availability, the strength of the opposition, and the size of their support base (Bueno de Mesquita and Smith 2010; Conrad 2011; Desai, Olofsgard, and Yousef 2009).

⁷ Participatory budgeting, the phenomenon of citizens and elected officials discussing resource constraints and allocations (including investment decisions), is an example of economic policymaking involving extensive public deliberation that is practiced across the authoritarian world, such as Zimbabwe, Egypt, Tanzania, Cameroon, China, and Russia (Pateman 2012; Sintomer, Herberg, and Allegretti 2010; Shah 2007).

⁸ As a metric for comparison, support for FDI by citizens in democratic countries is 65% and 59% in the PEW and ASES respectively. My original survey results corroborate these findings: Close to 60% of respondents indicated moderate to very strong citizen support in the authoritarian country in which their company invests (see question 18 of the survey questionnaire in Appendix C2). Additionally, a multilevel order probit estimation of FDI inflows on confidence in government (World Values Survey 2000) reveals that greater inflows of FDI are associated with greater confidence in government, thereby lending support to the microfoundation of my argument that citizens find FDI inflows to be desirable.

⁹ Simmons (2014) discusses public backlash against bilateral investment treaties.

¹⁰ While authoritarian regimes with public deliberation forums do incorporate the wider citizenry in policymaking, not all voices and partisans are necessarily included or weighted equally in the negotiating and deliberating (Chandra and Rudra 2015).

¹¹ Workers’ organizational capacities in authoritarian countries may be limited to prevent strikes that disrupt production and macroeconomic stability (Krueger and Mas 2004).

¹² Between 1970 and 2011, the average outward FDI from Western developed economies reached 90% of total worldwide FDI outflows (UNCTAD 2012), thus multinational concern about human rights and publicity is plausible for the majority of FDI flows.

¹³ Following Haftel (2010) and Tobin and Rose Ackerman (2011), I employ total BITs in force as the primary operationalization: The credible signal of the BIT occurs when the treaty is legally bound and implemented or in force. Results are robust using total BITs signed. See Appendix A3.

¹⁴ In the developing world, the level of public deliberation is not time invariant. Average difference in PARCOMP scores in a country over 10 years is close to half a point with a standard deviation of one point.

¹⁵ Other measures of formal institutions present similar trends. Since 1990, 59% of authoritarian country-years have multiple parties in the legislature, while only 15% and 26% have no legislature or the regime party only in the legislature respectively (Cheibub et al. 2009). Similarly, 75% of authoritarian governments have no veto players (Henisz 2000).

¹⁶ Fixed effects estimations indicate that PARCOMP is positive and statistically significant in predicting favorable FDI policies (Fraser Institute’s [2012] economic freedom and Allee and Peinhardt’s [2010] strength of enforcement in the BIT) and economic stability (Chandra and Rudra’s [2015] and Quinn and Woolley’s [2001] measure of the standard deviation of the growth rate). See Appendix A1. Further, Li’s (2009a) data provide preliminary evidence that the average number of expropriations is lower in authoritarian regimes with higher levels of public deliberation. Since 1990, 5% of BITs signed by authoritarian regimes with suppressive levels of public deliberation have been taken to ICSID, while only 3% have been taken to ICSID among autocracies with transitional levels of public deliberation (no cases exist for authoritarian regimes with competitive public deliberation levels). These preliminary data suggest higher levels of BIT compliance in authoritarian countries with public deliberation forums.

¹⁷ Other measures of democracy correlate very strongly with this classification of authoritarianism (0.82 with Marshall and Gurr [2008]’s polity).

¹⁸ Preliminary regression analysis indicates that bilateral investment treaties are statistically insignificant in predicting PARCOMP levels in authoritarian countries. Results available upon request.

¹⁹ China is an outlier, as it has a repressive level of public deliberation, above-average BITs in force (90 in 2008), and high levels of foreign direct investment inflows (in current US dollars). Results are robust upon excluding China from the sample. Results available upon request.

²⁰ King Hussein chartered a path championing world trade, finance, and investment in 1989, and King Abdullah II further adopted liberalization reforms throughout the early 2000s (Milton-Edwards and Hinchcliffe 2009:78, 89; Stevenson 2010).

²¹ Pro-investment policies include privatization, liberalization, deregulation, and market-enhancing initiatives as well as contract enforcement and legal infrastructure measures. Jordan’s National Agendas routinely emphasize these policies. For example, the Investment Development Theme in the 2006–2014 National Agenda calls for facilitating access to financial and capital markets, deregulating the communications sector, creating commercial law courts and arbitration tribunals specializing in investment disputes, negotiating international economic agreements, and developing and publishing competitiveness reports (National Agenda Steering Committee 2005).

²² Jordan has five concluded cases at the ICSID between 1972 and 2014 (out of 305 of the concluded cases registered). This number can be compared to a country, such as the Democratic Republic of Congo (with far fewer BITs and low levels of public deliberation), which has been taken to the ICSID eight times during this same time frame. Jordan’s caseload is higher than the average of one concluded case among all authoritarian countries. Yet, of Jordan’s five cases, one was dismissed in favor of the Jordanian government, three were settled, and one was rendered an award in favor of the MNC by the tribunal.

²³ Despite the relatively high levels of citizen participation and freedom of association in Jordan compared to other authoritarian countries, limits on participation and associational activity do exist (Milton-Edwards and Hinchcliffe 2009; Ryan 2010; Wiktorowicz 2000).

²⁴ Furthermore, professional and employer organizations and nongovernmental organizations are active throughout Jordan (for example, see http://www.civilsociety-jo.net/en/ and Ryan 2010).

²⁵ Moore (2000) discusses unproductive activities of the Jordanian business sector in the parliament (188). In contrast, Carroll (2003) finds that business associations and other economic organizations have been productive in their negotiations occurring “largely outside of Jordan’s newly-active democratic institutions” (4). Doner and Schneider (2000) provide further evidence on market-supporting and -complementing activities of business associations in the developing world.

²⁶ Fixed effects panel estimations are robust. See Appendix A6.

²⁷ I include country fixed effects to account for possible omitted variable bias. Country-specific factors such as geography and business culture influence investment flows (Jensen 2011; Vanhonacker and Pan 1997). According to Nickell (1981), bias is a concern because of the inclusion of a lagged dependent variable and fixed effects. Alternative model specifications reveal robust results; see Appendix A6 and A8.

²⁸ Models are consistently run with and without the rule of law to ensure that the estimations either control for country property right protections or maximize observations.

²⁹ Results are also consistent in a sample of developing-country democracies. Chandra and Rudra’s (2015) theory of policy stability associated with public deliberation applies to both democracies and nondemocracies; thus this statistically significant result provides further support to this theory. Results available upon request.

³⁰ Simmons (2010) addresses endogeneity concerns associated with international treaty research.

³¹ Following Roodman (2008), results are robust upon limiting the number of instruments by setting the maximum number of lags equal to two. Results available upon request.

³² Models predicting selection into the signing of BITs and models predicting selection into high levels of public deliberation are also robust. Results available upon request.

³³ Models with the mode of legislature selection present similar results—legislature*BITs is not statistically significant, while PARCOMP*BITs is positive and statistically significant. Results available upon request.

³⁴ Haggard (1990), Li and Resnick (2003), and O’Donnell (1978) posit theories for why foreign investors prefer autocratic institutions and regimes, focusing on financial incentives, strong autocrat-multinational alliances, and limited political coalitions.

Additional information

Funding

The University of Pittsburgh’s International Business Center and Fordham University’s Faculty Research Grant provided funding for the survey research.