Political Investment Cycles in Democracies and Autocracies

ABSTRACT

Extant research has shown considerable interest in whether host countries’ political uncertainty impedes foreign direct investment (FDI). Building upon the scholarly consensus on the adverse impact of political uncertainty on FDI, this article demonstrates that the extent to which investment climates are unpredictable varies cyclically, on the basis of election timing in democracies and leadership turnover in autocracies. The empirical results show that in presidential democracies, FDI tends to slowly increase after an executive election and then decline as the next executive election nears. However, I find that an electoral investment cycle is not found in parliamentary democracies where election timing is irregular, less predictable, and endogenous to domestic economic conditions. I also find that a similar political investment cycle exists in autocracies not through electoral cycle but through leadership tenure cycle. The level of FDI inflows tends to be relatively low early in autocrats’ tenure when political uncertainty is high and rise as autocratic leadership tenure increases over time but eventually wane again as autocratic leadership is destabilized in the late period of power transition. The findings indicate the existence of heterogeneous political investment cycles, depending on regime type.

KEYWORDS:

• Autocracy
• cycle
• democracy
• FDI
• uncertainty

Acknowledgments

An earlier draft of this article was presented at the International Studies Association meetings, Atlanta GA, March 2016. I thank James Hollyer, Hoon Lee, Chungshik Moon, Stephen Nelson, and the anonymous reviewers for helpful comments.

Supplementary data

Supplemental data for this article can be accessed on the publisher’s website www.tandfonline.com/gini.

Notes

¹ Some firms might adopt more aggressive and risk-taking investment strategies to preoccupy the markets of host countries.

² The theoretical framework of irreversible investment under uncertainty was introduced in detail in Dixit (1992), Dixit and Pindyck (1994), and Pindyck (1991). Refer also to Pindyck and Solimano (1993) for the early literature on the theory of irreversible investment.

³ This delaying effect of policy uncertainty may well be mitigated when firms observe the sufficient benefits of early entry in times of severe competition (Rivoli and Salorio 1996; Trigeorgis 1996).

⁴ Relying on a similar mechanism, Canes-Wrone and Park (2014) found, in their housing market analysis, that domestic economic activities decline before elections.

⁵ Smith (2004) finds empirical evidence that improving economic performance does induce electoral benefits in early elections.

⁶ Alexander (2002) argues that democratic uncertainty should be interpreted in a technical sense rather than in a substantive sense because uncertainty over election outcome does not necessarily undermine the preexisting rules and norms of democratic governance.

⁷ Furthermore, they argue that “Dictators interested in gaining access to international funds possess a strong interest in adopting multiparty elections because donors generously reward dictators who hold elections” (Magaloni and Kricheli 2010:135).

⁸ The different temporal dimensions result from the data availability of leadership tenure from the Archigos database (~ 2004) for the autocracy sample (Goemans, Gleditsch, and Chiozza 2009).

⁹ The formula used for the logarithmic transformation is ln(FDI + 1) if FDI ≥ 0, and −ln(1 + |FDI|) if FDI < 0.

¹⁰ This measure is a composite index that records the number of different types of mass and elite unrest such as purges, riots, government crises, assassinations, strikes, guerrilla wars, revolts, and anti-government demonstrations.

¹¹ Although the literature reveals the positive impact of BITs on FDI in general (Buthe and Milner 2008; Egger and Pfaffermayr 2004; Milner 2014; Neumayer and Spess 2005; Salacuse and Sullivan 2005), the empirical evidence is far from definite, as Newcombe and Paradell (2009) noted. Thus focusing on BITs with OECD countries is useful because OECD countries are largely responsible for global FDI outflows, and past studies found that BITs with OECD countries do have positive and significant effect on FDI (Neumayer and Spess 2005; Tobin and Rose-Ackerman 2006).

¹² I examined whether the time series is stationary using augmented Dickey-Fuller (DF) tests (Dickey and Fuller 1979, 1981) and nonparametric Phillips and Perron (PP) tests (Phillips and Perron 1988) and found that the FDI time series in the sample is stationary.

¹³ I also examine whether the effect of electoral cycle is conditioned by a democratic country’s economic conditions in the online appendix: For example, the effect of electoral cycle may be smaller when a country is experiencing significant economic growth because a leader may have less incentives to reverse existing economic policies that have been proven to be successful. However, I find minimal empirical evidence for this conjecture.

¹⁴ In addition, I test if the empirical results for the autocracy sample are driven by major Asian economies under autocracy that have been popular destinations of FDI. The results in the online appendix show that my findings hold, excluding potential outliers such as Taiwan (~1995), China, South Korea (~1987), and Singapore.

¹⁵ For example, the reflection point appears around 3.2 years after an executive election. The average length of an inter-executive election period is about 7.2 years in the autocratic sample.

¹⁶ I further discuss an alternative measure, autocratic time horizon, which captures autocratic vulnerability in greater detail in the online appendix.