The Two-sided Effect of Elections on Coup Attempts

• Published date: 01 August 2020
• Author: Lasse Lykke Rørbæk,Suthan Krishnarajan
• Date: 01 August 2020
• DOI: 10.1177/0022002719900001
• Subject Matter: Articles

Article

The Two-sided Effect

of Elections on

Coup Attempts

Suthan Krishnarajan

1

and Lasse Lykke Rørbæk

1

Abstract

In this article, we investigate the relationship between elections and coup attempts.

We argue that elections have opposing effects on the risk of coup attempts,

depending on the state of the economy in which they are held. Elections occurring in

conditions of economic crisis spur anti-government mobilization and high levels of

state repression. This increases the subsequent risk of coup attempts. Conversely,

elections held during economic expansion induce pro-government mobilization and

waning repression, which reduces the subsequent risk of coups. We find strong

support for these propositions in a statistical analysis of 130 countries that con-

ducted contested elections in the period 1952 to 2013. The results are robust to an

array of model specifications, including when we account for election outcome,

postelection economic performance, and the possibility that both elections and

economic performance are endogenous to coup attempts.

Keywords

coup attempts, elections, economic crisis, protest, state repression

Introduction

Recent coups in Thailand (2014), Burkina Faso (2015), Turkey (2016), and

Zimbabwe (2017) illustrate that military interventions in politics remain key for

understanding regime developments around the world—and this both in democratic

1

Department of Political Science, Aarhus University, Aarhus, Denmark

Corresponding Author:

Suthan Krishnarajan, Aarhus University, Bartholins All´

e 7, Aarhus 8000, Denmark.

Email: suthan@ps.au.dk

Journal of Conflict Resolution

2020, Vol. 64(7-8) 1279-1306

ªThe Author(s) 2020

Article reuse guidelines:

sagepub.com/journals-permissions

DOI: 10.1177/0022002719900001

journals.sagepub.com/home/jcr

and autocratic regimes.

1

An extensive literature has examined how the risk of coups

is influenced by regime characteri stics such as institutional coup-pro ofing (e.g.,

Belkin and Schofer 2003; Bo¨hmelt and Pilster 2015; Powell 2012) and institutiona-

lized succession rules (Frantz and Stein 2017). However, little is known about the

extent to which one of the key institutional features of modern regimes—national

elections—affects the risk of coups. In fact, although the association between elec-

tions and other types of regime instability has received much attention (e.g., Ceder-

man, Gleditsch, and Hug 2013; Fjelde and Ho¨glund 2016; Gandhi and Przeworski

2007; Magaloni 2008; Schedler 2002; Tucker 2007), only one study has system-

atically assessed the effect of elections on coups, and this only in autocratic regimes

(Wig and Rød 2016).

One explanation for this shortcoming is that scholars have failed to identify the

relevant conditions under which elections trigger coups. In this study, we argue that

electoral contestation is an important determinant of coup attempts but that elections

have opposite effects depending on the economic situation in which they are held.

More specifically, we argue that elections taking place during economic decline—

what we term “crisis elections”—increase the subsequent likelihood of coup

attempts compared to crisis periods without contested elections. Conversely, elec-

tions occurring during economic expansion—what we term “growth elections”—

should not only mitigate the potentially destabilizing effects of elections but actually

reduce the risk of coup attempts relative to a period with the same economic per-

formance but with no contested election.

Theoretically, we propose that this ambiguous effect of elections on coup

attempts can be explained by variation in mass mobilization and state repression.

First, since economic crisis induces popular discontent and electoral contestation at

the same time helps solve people’s collective action problems, we expect cri sis

elections to increase the likelihood of mass-based political protest (see Haggard and

Kaufman 1995; Brancati 2016, chap. 3; Tucker 2007). Coup attempts, in turn,

become more likely because the armed forces are motivated to intervene in order

to restore law and order and because they consider it more likely that the public will

accept an extra-constitutional change of government (Galetovic and Sanhueza 2000;

Geddes 2006; Huntington 1968, chap. 4). On the other hand, electoral contestation

that occurs during a period of economic expansion will likely drum up support for

the incumbent government, thereby lowering the military’s incentive to intervene.

Second, during economic turmoil, fear of unfavorable election results may con-

vince incumbent leaders to increase the level of repression as a means to reduce

electoral competition. However, incumbents who increasingly rely on the military to

maintain political control become more vulnerable to coup attempts, either because

they fail to reach a compromise on the military’s deg ree of policy influence or

because the military chooses to intervene instead of jeopardizing its institutional

interests by using violence against unarmed protesters (Bellin 2012; Magaloni 2010;

Svolik 2013; Wig and Rød 2016). By contrast, when the economy is performing well

and an electoral campaign is approaching, incumbents have incentives to lessen

1280 Journal of Conflict Resolution 64(7-8)