Reassessing the relationship between elections and democratization

AuthorNam Kyu Kim
Published date01 June 2020
Date01 June 2020
DOI10.1177/0192512119846144
Subject MatterArticles
https://doi.org/10.1177/0192512119846144
International Political Science Review
2020, Vol. 41(3) 305 –320
© The Author(s) 2019
Article reuse guidelines:
sagepub.com/journals-permissions
DOI: 10.1177/0192512119846144
journals.sagepub.com/home/ips
Reassessing the relationship
between elections and
democratization
Nam Kyu Kim
Sungkyunkwan University, South Korea
Abstract
Do elections improve a country’s prospects for democratization? The theory of democratization by elections
proposes that reiterated multiparty elections foster gradual institutional changes leading to progress toward
democracy. Edgell etal. provide the most systematic evidence for this theory, but substantial disagreements
remain among scholars studying the effects of authoritarian elections on democratization. In this study, I
reassess the relationship between elections and democratization. In doing so, I account for the temporal
structure of democratic qualities that are highly persistent. When I control for the dynamics of democratic
qualities by adding a lag (or lags) of the dependent variable, the democratizing effects of repeated elections
disappear; indeed, most evidence indicates a negative relationship between elections and democratization.
This finding calls into question the causal role of elections in the process of democratization and demonstrates
the importance of modeling the persistent nature of democracy in exploring the effect of a variable on
democratization.
Keywords
Elections, democracy, authoritarianism, democratization, regime change
Introduction
Multiparty elections have become commonplace, even in autocracies. The number of regimes
with multiparty elections increased sharply after the end of the Cold War, and electoral authori-
tarianism has become the modal form of autocracy. As of 2015, about nine out of ten countries
held multiparty legislative elections. Only countries with either single party dictatorships, such
as China, Cuba, and North Korea, or monarchies, such as Qatar and Saudi Arabia, refused to
hold multiparty elections.
Corresponding author:
Nam Kyu Kim, Department of Political Science and Diplomacy, Sungkyunkwan University, 25-2 Sungkyunkwan-Ro,
Jongno-Gu, 03063 Seoul, Korea.
Email: namkkim1@gmail.com
846144IPS0010.1177/0192512119846144International Political Science ReviewKim
research-article2019
Article
306 International Political Science Review 41(3)
Given this trend, comparative politics scholars have studied the consequences of authoritarian
elections. In particular, the question of how authoritarian elections, often flawed and manipulated,
influence the prospects for democratization has drawn a great deal of scholarly attention. One
group of scholars emphasizes the possibility of democratization by elections (Bratton and Van de
Walle, 1997; Donno, 2013; Howard and Roessler, 2006; Lindberg, 2006, 2009). They contend that
elections, even when held in authoritarian regimes, promote democracy. Among others, the democ-
ratization-by-elections theory, developed by Lindberg (2006, 2009), proposes that ‘the longer an
uninterrupted series of elections a country has, the more its society will become imbued with
democratic qualities’ (Lindberg, 2006: 99). This theory has played a seminal role in ongoing
debates about the impact of elections on democracy. However, substantial disagreements remain
concerning the effects of authoritarian elections on democratization. Several scholars argue that
authoritarian elections serve to stabilize and strengthen incumbent regimes rather than promote
democratization (Blaydes, 2010; Gandhi and Lust-Okar, 2009; Gandhi and Przeworski, 2007;
Geddes, 2006; Magaloni, 2006, 2008). Moreover, some scholars are skeptical that multiparty elec-
tions play any causal role in democratization (Brownlee, 2007; Carothers, 2002; Levitsky and Way,
2010). Studies have documented the conditional role of authoritarian elections in democratization
(Howard and Roessler, 2006; Morgenbesser and Pepinsky, 2019; van Ham and Seim, 2018).
Against this background, Edgell et al. (2018) provide the most systematic evidence for the
theory. Using a global sample covering the period 1900–2010, they show that reiterated sequences
of multiparty elections are associated with gradual improvements in non-electoral components of
democracy, as measured in the Varieties of Democracy (V-Dem) dataset (Coppedge et al., 2017).
They also find that the relationship between repeated elections and democracy is only pronounced
in the period since the third wave of democracy.
However, they do not account for the temporal dynamics in democratic qualities though there
are important theoretical and econometric reasons for doing so. This omission is problematic
since past democracy levels significantly influence future democracy levels. For example, as
Gerring et al. (2012: 2) state, ‘regimes do not begin again, de novo, with each calendar year.
Where one is today depends critically upon where one has been.’ Additionally, as Wilkins (2018)
shows, ignoring dynamics in highly persistent series, such as democratic qualities, can lead to an
overestimation of the effect of a variable on the outcome. This suggests the possibility that Edgell
et al. (2018) may overestimate the effect of elections and mistakenly attribute democratizing
effects to repeated elections. While the recent seminal studies on democratization emphasize the
importance of modeling the dynamics of democracy (Acemoglu et al., 2008; Boix, 2011; Gerring
et al., 2012; Treisman, 2015), the dynamics of democracy are often overlooked in the scholarship
on democratization by elections.
Using the data provided by Edgell et al.’s (2018) replication files, I revisit their analyses. My
empirical analysis shows that when I account for the dynamics of democratic qualities by adding a
lagged dependent variable (LDV) or more lags of the dependent variable, the democratizing effects
of repeated elections disappear. In contrast to Edgell et al. (2018), I find a negative relationship
between repeated elections and various democratic qualities. This suggests that their finding is mainly
driven by a failure to control for the persistent nature of democratic qualities. Consistent with previ-
ous studies, I find that the level of democracy is highly persistent and tends to revert to the mean.
This finding stands, even when I limit the analysis to the period since the third wave of democ-
racy as well as when I focus on specific regions, including sub-Saharan Africa, which was the main
empirical basis for the theory. The main result is highly robust to the use of alternative measures of
democracy or democratization, temporal aggregation strategies, and modeling strategies. Last, I
examine whether democratic context, state capacity, or economic development serves as a precon-
dition for the democratizing power of elections in authoritarian regimes. I find that repeated

To continue reading

Request your trial

VLEX uses login cookies to provide you with a better browsing experience. If you click on 'Accept' or continue browsing this site we consider that you accept our cookie policy. ACCEPT