Linkage to the West and Electoral Manipulation

Abstract

The relationship between linkage to the West and the survival of political regimes has gained increasing attention in recent years. Despite this attention, one aspect of this linkage remains poorly understood, and that is the effect of linkage to the West on electoral manipulation. Scholars have suggested that linkage to the West raises the cost of government abuse because it increases the probability of Western governments taking action in response to reported abuse. This assumption then suggests that incumbents should choose the forms of repression more wisely. Consequently, in cases of the higher level of linkage to the West incumbents are less likely to use the more visible forms of repression and manipulation. I test the aforementioned assumptions on time series, cross-national data set with observations of 147 elections in competitive authoritarian regimes between 1990 and 2010. I find that extensive international relations to the West have only minimal and statistically non-significant effect on visible forms of repression and manipulation. These results contribute to our understanding of international linkage as they show that linkage to the West does not provide sufficient protection to opposition leaders and groups.

Keywords

international linkage competitive authoritarianism repression menu of manipulation elections

Introduction

The international Zeitgeist after the Cold War has made it much harder for explicitly authoritarian and politically closed regimes to emerge and survive. However, we cannot say that with the third wave of democratization (Huntington, 1991), all forms of authoritarianism would cease to exist around the globe. The very specific international constellation, a significant feature of which was a distinctive emphasis of the international community on free elections, led to a formation of numerous political regimes where democratic institutions exist, yet those who are at the control of power often resort to regular violations of democratic principles in order to retain the power they possess (Castañeda, 1996; Diamond, 2002; Levitsky and Way, 2010). Hence, elections in these regimes are considered competitive, but are not fair.

The common occurrence of such regimes, often referred to by the collective term of “hybrid” or “mixed,” has led to a rapid expansion of empirical research. The result of these research efforts is that today we can tell when power holders manipulate the elections (Bhasin and Gandhi, 2013; Birch, 2011; Magaloni, 2006; Przeworski and Gandhi, 2009; Simpser, 2013), strategies they employ (Schedler, 2002), and when they can be defeated (Bunce and Wolchick, 2011; McFaul, 2005). Our understanding of impacts of repression and manipulation on the party system design (Donno and Roussias, 2012) has also significantly advanced, as well as the understanding of trust in political institutions (Lehoucq 2003: 249) or forms of post-elections protests (Harvey and Mukherjee, 2018).

However, the current state of research in this field is for one reason unsatisfactory. Little is known about the effects of linkage to the West on electoral manipulation, particularly, how linkage to the West affects the selection of a concrete manipulative or repressive strategy. Earlier studies that have tried to explain on what basis the power holders choose among single strategies were mostly involved with domestic factors (Hauser, 2017; Schedler, 2013; Seeberg, 2014; Van Ham and Lindberg, 2015). The exception here are the works studying the impacts of international observers on the form of manipulation (Beaulier and Hyde, 2009; Cheeseman and Klaas, 2018; Hyde, 2007), yet these works were evaluating the effects of short-term international factors.

Based on present literature, the international constellation should play a role even in a long-term scope. Levitsky and Way assume that leaders in hybrid regimes are in their decision-making processes affected by the state’s level of linkage to the West. These linkages raise the international cost of repression because they increase the probability of Western governments taking action in response to reported abuse (Levitsky and Way, 2010: 45). Alongside with this assumption, autocratic leaders in countries with high level of linkage should, bearing in mind the possible consequences, commit the visible forms of repression and manipulation less often.

The aim of this article is to fill the gap in comprehending the relation between linkage to the West and electoral manipulation. Using empirical test and combining the original data set about linkage to the West in competitive authoritarian regimes between 1990 and 2010 from Levitsky and Way with the new data about incumbent’s abuse of power from V-dem database (Coppedge et al., 2019b), in the end I do not find support for this argument. Extensive international ties to the West reduce the level of some of the visible forms of repression and manipulation; nonetheless, this effect is small and statistically nonsignificant. These results contribute to our understanding of international linkage as they show that linkage to the West does not provide protection to opposition leaders and groups.

How Linkage Affects Repression and Manipulation

The linkage to the west is supposed to affect the level and type of repression and manipulation. The origin of this explanation lies in the theory of regime changes by Levitsky and Way (2010). The authors, in their undoubtedly pathbreaking work, tried to explain different regimes trajectories of competitive authoritarianisms after the end of the Cold War. In their view, the fate of this hybrid regimes category is affected by the combination of three variables: linkage (to the West), leverage (Western), and organizational power. Among the above-mentioned triad, the linkage is clearly the most fundamental since it constitutes a fundamental condition for democratization. In their book, Levitsky and Way do not present a single case of democratization without a high level of linkage to the West. The other two variables then become a factor only in cases of low level of linkage, and they only determine how stable a form of authoritarian government establishes in a certain country (Levitsky and Way, 2010; Slater, 2011). Levitsky and Way (2010: 43)

define linkage to the West as the density of ties (economic, political, diplomatic, social, and organizational) and cross-border flows (of capital, good and services, people, and information) among particular countries and the United States, the EU (and pre-2004 EU members), and Western-dominated multilateral institutions (Levitsky and Way, 2010: 43).

Intensive ties to the West contribute to democratization in four different ways. They help to even the uneven play field between government and opposition, increase the probability of potential rupture within autocratic parties, improve the image of democratic opposition at domestic audience, and what is central for this article, they protect opposition before regime repression (Levitsky and Way, 2010: 47–48). Linkage to the West raises the international cost of repression as it increases the probability that Western governments will take action in response to reported abuse. There is a simple explanation for this. A country with international ties to the West becomes a subject of Western media more frequently. Even relatively small signs of repression and manipulation may then draw international attention. If that happens, it becomes more likely that media, international nongovernmental organizations, and civil society will push the governments of Western countries to intervene. Typical cases are the different approaches to competitive authoritarianism in Slovakia and Zimbabwe. Due to high level of linkage to the West, Vladimir Mečiar’s regime was far more often the target of criticism and kept drawing attention from the Western pro-democracy actors for even minor violations of democratic procedures such as violating informal parliamentary norms of committee assignment. By contrast, due to low level of linkages to the West, the international attention paid to Zimbabwe was sporadic and focused only on the most extreme abuses of power (Levitsky and Way, 2010).

Based on this assumption, it is possible to establish hypothesis on the relationship among the linkage to the West, and repression and manipulation. It is supposed to be risky for power holders in hybrid regimes to employ acts of fraud or repression, provided the ties of their country to the West are strong. Employment of repression and manipulation will lead to reducing legitimacy at the international landscape which may turn into imposing sanctions, limiting foreign aid or even direct international intervention (Tanneberg et al., 2013: 119).

Nevertheless, it is likely that a high level of linkage to the West will not influence all types of repressions and manipulation in the same way. That is due to the diverse visibility of individual strategies that causes the international audience being more sensitive to some expressions of electoral manipulation than others. Following Schedler’s (2013) conceptualization, this article works with a quaternion of strategies: exclusion, repression, fraud, and censorship.

The most visible form is the exclusion of majority or all of the major opposition candidates. This form of electoral manipulation is difficult to hide as it directly pertains the actors who, in the country with a high level of linkage to the West, will surely try to draw attention to their situation (Schedler, 2013: 274) as it is on the outs with the ideal of liberal democracy. Another visible form of manipulation is a direct repression of opposition representatives. Experience shows that international audience is highly sensitive to violent acts of current authoritarianists (Cheeseman and Klaas, 2018: 96–97). The electoral fraud is also a visible form of electoral manipulation because it is vulnerable to detection and high profile (Cheeseman and Klaas, 2018). On the other hand, censorship is harder to spot (Donno, 2013: 172; Schedler, 2013: 274). That is because of media censorship, if performed well, is nearly impossible to uncover, and as a bonus, it reduces the visibility of other strategies for manipulate domestic audience (Van Ham and Lindberg, 2015):

H1: Governments in high-linkage contexts are less likely to use the more visible forms of repression and manipulation (exclusion, repression and fraud).

H2: High linkage to the West does not affect the less visible form of repression and manipulation (censorship).

Data and Methods for the Explanatory Analysis¹

In this text, hybrid regimes are called all the regimes that Levitsky and Way (2010) regard as competitive authoritarianisms in their book. Hence, the time range is limited to the years 1990–2010. There are two reasons for this limitation. The first one is the fact that the book is trying to verify Levitsky and Way’s theoretical assumptions on cases that the authors call competitive authoritarianisms. That is mainly given by Levitsky and Way (2010: 34) saying that the validity of their theoretical assumptions is restrained by the time they are focused on in their book. In my view, expanding the time range would not then be quite fair toward the authors. Considering the tested hypotheses, it would be undeniably worse if they do not stand against the cases that should be valid by all means, in other words the competitive authoritarianisms right after the end of Cold War. Second reason is that for these cases, the same authors offer data on the linkage to the West. The aim of this work is then as much as possible to replicate Levitsky and Way’s original data. Nevertheless, competitive authoritarianisms are obviously not the only type of hybrid regime. It is still the regime category that is the most common in the current world and for which the aforementioned theoretical hypothesis should be most applicable (Levitsky and Way, 2010: 13–16).

In this work, it is the single repressive and manipulative strategies that are the dependent variables. The formulations of the data come from the V-Dem database. This is a new comprehensive data set on democracy that collects data on more than 450 indicators of democracy in 173 countries around the world from 1789 until 2018, engaging over 3000 country experts worldwide to collect the data (Coppedge et al., 2019b). As was already noted in the preceding part, this article works with a quaternion of strategies: exclusion, repression, fraud, and censorship. The advantage of the V-Dem strategy is that the value of all variables is given on the scale from 0 to 4 including decimals.² That is a major improvement in the data quality as earlier researches often must have worked only with dichotomous variables.³ The overview of the used variables may be found in Table 1.

Table 1.

Overview of the Depend Variables.

Dependent variable	Code	V-Dem source
Exclusion	Ban	Party ban
Exclusion	Barriers	Barriers to parties
Repression	Violence	Physical violence index
Repression	Intimidation	Election government intimidation
Fraud	Fraud	Election Management Body autonomy
Censorship	Censor	Government censorship effort — Media
Censorship	Bias	Media bias

To formulate the level of exclusion, two variables from the V-Dem database are suitable to use. The first is Party Ban that gives the number of banned parties. Barriers to parties then illustrate how high the obstacles for starting a political party are (Coppedge et al., 2019a: 88–89). In order to capture the level of repression, two variables are used again—Physical violence index⁴ and Election government intimidation. The first one involves political murders and tortures, in other words the violent forms of repression, while intimidation looks at whether members of political opposition are intimidated or harassed by the ruling party (Coppedge et al., 2019a: 58).

To identify the level of electoral fraud,⁵ EBM autonomy variable—indicator of the independence of electoral committees—is selected (Coppedge et al., 2019a: 53). The V-Dem database includes also other variables that might be put in use here; however, this is the closest to delineating the electoral fraud in Schedler’s sense. In his view, electoral fraud is a situation when electoral bodies convey unusual attraction toward power holders (Schedler, 2013: 99). Finally, two variables are selected in order to identify censorship level. The paper then, similarly like Levitsky and Way (2010: 11), distinguishes between the state when media liberty is restrained (censorship) and situation when media are biased toward power holders.

The explanatory variable is the level of linkage to the West. Source of the data is the previously mentioned Levitsky and Way’s (2010) work. Apart from the linkage to the West, some types of electoral manipulation are likely to be influenced by other contextual factors, as well. To address this, I include several control variables. The first one is a dichotomous variable determining whether a government holds a majority in the parliament.⁶ Power holders with a legislative majority should then favor exclusion and censorship over violence and fraud (Schedler, 2013: 198). The second control variable gives the level of state capacity.⁷ Earlier works indicate that in non-democratic regimes with a lower level of state capacity, power holders still favor repression as they are not able to enjoy support of more “sophisticated” forms of manipulation (Seeberg, 2014: 1269–1271). Another control variable is the election type (legislative, executive, and concurrent). At executive and concurrent elections, media manipulation often takes place (Birch, 2011). The overall level of repression should be affected by contextual factors, anyway. First, one may assume that electoral manipulation is supposed to be more frequent at executive and concurrent elections since such elections are more important for rulers in hybrid regimes (Schedler, 2013: 287). Next to that, electoral manipulation is expected to be more frequent in countries affluent with mineral resources. A dictator in a country affluent with mineral resources⁸ tries to retain his power as long as possible (Michalik, 2015: 53–85). For similar reasons, electoral manipulation is supposed to be more frequent also in countries with a high level of social inequalities (Henderson, 1991: 125). It is because both situations raise the stakes of authoritarian politics seeing that in the case of electoral defeat, a dictator has more to lose (Schedler, 2013: 284). Finally, I use data on gross domestic product (GDP) growth in order to control for economic situation. The electoral manipulation surge may be a reaction to worsening of economic situation (Dresden and Howard, 2015).

The unit of analysis is the individual elections. There are 147 cases of elections⁹ in total between 1990 and 2010 in 34 countries. Panels are unbalanced with uneven gaps in the time-series as observations per group range from 1 to 8 with an average of 4.31. Given that the observations are nested within specific countries,¹⁰ the most suitable method to take such features into account is multilevel analysis. I analyze the data using multilevel regression model (random intercept). I opted for the random-effects models because of the value of the linkage to the West being constant on country level, and random effect models are able to account for this situation (Clark and Linzer, 2015).

Results¹¹

The analysis results are presented in Table 2. Three different models were created for each of the dependent variables. Basic models show the result without the control variables, while full models include all the control variables. Beside these two, I also included the null models. According to the logic of the first hypothesis, governments in hybrid regimes with a high level of linkage to the West are supposed to employ exclusion, repression, and fraud on a lesser scale. Howbeit, the results of the performed analysis show a different picture. Linkage to the West lowers the level of repression, nonetheless the effect is very small and with one exception statistically nonsignificant. The effect of linkage to the West is statistically significant but only in the basic model of repression. In case of exclusion, the detected effect is again negligible. What is even more interesting is that the level of one its components—party ban—rises. As in the case of repression, none of these results reach conventional levels of statistical significance.¹² In countries with a high level of linkage to the West the level of electoral frauds decreases, but this effect is again small and statistically nonsignificant. The Akaike information criterion (AIC) and Bayes information criterion (BIC) scores show that the null models outperform the other two models for each dependent variable, which also confirms that there is no significant effect of linkage. Linkage to the West, without doubt, does not lower the level of repression and manipulation. The presented hypothesis then must be rejected. Two explanations may be offered here. The first one is that international reputation is not as important for power holders as the recent literature assumes. In numerous cases, that may be caused by specific domestic political factors being more important than linkage to the West. An example of such factor may be the level of economic freedom. In one of his studies, Weyland (2013) shows that hybrid regimes with high level of economic nationalism are more resistant toward international pressures for democracy. In other cases, international reputation does not have to be so significant such as in a situation when a government in hybrid regime may lean on autocratic sponsors like Russia or Venezuela (Tansey, 2016). The second explanation points out that Western countries do not always care about spreading democracy. Levitsky and Way (2010) only mention this option marginally in case of France’s actions in some of African countries. However, in recent years, other scholars also mentioned this possibility in cases of the United States’ actions in Egypt (Brownlee, 2012) and Nicaragua (Thaler, 2016).

Table 2.

Analysis explaining the effect of linkage to the West on electoral manipulation, 1990-2010.

	Repression						Exclusion						Fraud			Censorship
	Intimidation			Violence			Party ban			Barriers to parties			Election Management Body autonomy			Censorship			Media bias
	Null	Basic	Full	Null	Basic	Full	Null	Basic	Full	Null	Basic	Full	Null	Basic	Full	Null	Basic	Full	Null	Basic	Full
Explanatory variables
Linkage		−0.104	−0.087		−0.255^†	−0.158		0.060	0.130		−0.120	−0.070		−0.171	−0.161		−0.001	0.054		−0.041	0.059
		(0.136)	(0.147)		(0.139)	(0.143)		(0.103)	(0.120)		(0.121)	(0.131)		(0.135)	(0.147)		(0.137)	(0.145)		(0.132)	(0.142)
Control variables
Legislative majority			−0.074			−0.008			0.129			0.009			−0.020			−0.039			0.021
			(0.066)			(0.058)			(0.094)			(0.083)			(0.070)			(0.076)			(0.070)
State capacity			−0.063			−0.141^†			−0.051			0.253**			0.087			0.154			0.076
			(0.086)			(0.077)			(0.100)			(0.097)			(0.090)			(0.095)			(0.089)
Main elections			−0.034			−0.064			−0.122			−0.063			−0.059			−0.103			−0.087
			(0.056)			(0.049)			(0.083)			(0.071)			(0.059)			(0.064)			(0.059)
GINI			−0.051			0.259**			0.066			0.072			0.042			0.101			0.173
			(0.103)			(0.093)			(0.113)			(0.113)			(0.107)			(0.113)			(0.106)
Natural rent			0.178			0.044			0.152			0.107			0.064			0.166			0.098
			(0.111)			(0.101)			(0.116)			(0.119)			(0.115)			(0.121)			(0.114)
GDP growth			−0.006			−0.028			0.024			0.097			−0.018			0.016			0.073
			(0.081)			(0.076)			(0.087)			(0.083)			(0.083)			(0.085)			(0.081)
Constant	−0.098	−0.080	−0.066	−0.053	−0.006	−0.031	−0.035	−0.039	−0.045	−0.082	−0.067	−0.050	−0.174	−0.145	−0.139	−0.107	−0.107	−0.093	−0.087	−0.081	−0.078
	(–0.146)	(0.149)	(0.149)	(0.156)	(0.153)	(0.147)	(0.106)	(0.106)	(0.109)	(0.128)	(0.130)	(0.125)	(0.147)	(0.147)	(0.147)	(0.145)	(0.148)	(0.144)	(0.140)	(0.144)	(0.142)
AIC	335.265	338.840	365.429	313.848	314.686	334.695	423.151	427.511	448.864	393.529	396.924	415.579	347.628	350.197	377.904	372.741	376.900	397.417	350.033	354.163	377.543
BIC	344.236	350.382	395.333	322.820	326.647	364.599	432.122	439.472	478.768	402.500	408.886	445.483	356.599	362.159	407.808	381.712	388.862	427.321	359.005	366.125	407.447
Log likelihood	−164.632	−165.420	−172.714	−153.924	−153.343	−157.348	−208.575	−209.755	−214.432	−193.765	−194.462	−197.790	−170.814	−171.099	−178.952	−183.370	−184.450	−188.708	−172.017	−173.081	−178.771
Number of observations.	147	147	147	147	147	147	147	147	147	147	147	147	147	147	147	147	147	147	147	147	147
Number of groups: country	34	34	34	34	34	34	34	34	34	34	34	34	34	34	34	34	34	34	34	34	34
Var: country (intercept)	0.633	0.641	0.634	0.755	0.699	0.630	0.160	0.162	0.186	0.397	0.399	0.363	0.625	0.612	0.607	0.579	0.601	0.553	0.558	0.575	0.551
Var: residual	0.333	0.334	0.339	0.266	0.266	0.258	0.867	0.870	0.846	0.601	0.600	0.586	0.372	0.372	0.382	0.468	0.468	0.463	0.390	0.390	0.389

GDP: gross domestic product; AIC: Akaike information criterion; BIC: Bayesian information criterion.

Multilevel regression analysis, clustered by country with random effects. Robust standard errors in parentheses. P-values: †p < 0.1; *p < 0.05. **p < 0.01. ***p < 0.001.

When it comes to the second hypothesis, the results confirm more the expectations concerning the effect of linkage on the less visible form of repression and manipulation. High linkage to the West certainly does not affect the level of censorship. The controls suggest that several factors are not related to these outcomes in the ways the literature suggests. The most surprising result appears to be that a legislative majority in government hands does not affect the form of manipulative strategy. This relationship may be valid for several specific cases like Nicaragua after Daniel Ortega’s return to power (Martí i Puig, 2016), but it cannot be generalized to a broader field of hybrid regimes. Another surprising finding is that executive and concurrent elections do not increase the level of any tested form of manipulation. Conversely, the results confirm some assumptions concerning the effect of the level of state capacity.

Conclusion

Does the level of linkage to the West affect the form of electoral manipulation? Theoretical literature suggests so. Governments in hybrid regimes with a high level of linkage to the West, in lines with this, commit the visible forms of repression and manipulation less often.

Yet, the analysis results do not encourage such deductions. Extensive international relations to the West have only minimal and statistically nonsignificant effect on the visible forms of repression and manipulation. Acquired findings clearly illustrate that even in the first two decades after the end of Cold War, an intense level of linkage to the West did not increase the international cost of repression, as many recent influential authors led Levitsky and Way (2010) hypothesized. It was this very hypothesis that in their view presented one of the ways of how the key variable of their theory—linkage to the West—contributed to the democratization of competitive authoritarianisms after the end of Cold War. Obviously though, disproving such hypothesis does not automatically mean that the whole theory of regime changes by Levitsky and Way is invalid. It only shows that the linkage to the West works differently than its authors intended because it does not provide protection to opposition leaders and groups.

Footnotes

Appendix A

Universe of Cases.

Case	Elections	Case	Elections	Case	Elections
Guyana	1992	Gabon	1996	Georgia	1999
Croatia	1992	Gabon	1998	Georgia	2000
Croatia	1995	Gabon	2001	Georgia	2003
Croatia	1997	Gabon	2005	Georgia	2004
Croatia	2000	Gabon	2006	Georgia	2008
Slovakia	1994	Moldova	1994	Zimbabwe	1990
Slovakia	1998	Moldova	1996	Zimbabwe	1995
Mexico	1991	Moldova	1998	Zimbabwe	1996
Mexico	1994	Moldova	2001	Zimbabwe	2000
Mexico	1997	Moldova	2005	Zimbabwe	2002
Mexico	2000	Russia	1995	Zimbabwe	2005
Macedonia	1994	Russia	1996	Zimbabwe	2008
Macedonia	1998	Russia	1999	Kenya	1992
Serbia	1992	Russia	2000	Kenya	1997
Serbia	1997	Russia	2003	Kenya	2002
Serbia	2000	Russia	2004	Kenya	2007
Nicaragua	1990	Russia	2007	Malawi	1992
Albania	1991	Russia	2008	Malawi	1994
Albania	1992	Senegal	1993	Malawi	1999
Albania	1996	Senegal	1998	Malawi	2004
Albania	1997	Senegal	2001	Malawi	2009
Albania	2001	Senegal	2007	Mozambique	1994
Albania	2005	Ukraine	1991	Mozambique	1999
Dominican Republic	1994	Ukraine	1994	Mozambique	2004
Dominican Republic	1996	Ukraine	1998	Benin	1991
Romania	1990	Ukraine	1999	Benin	1995
Romania	1992	Ukraine	2002	Benin	1996
Romania	1996	Ukraine	2004	Benin	1999
Romania	2004	Botswana	1994	Benin	2003
Haiti	1990	Botswana	1999	Madagascar	1992
Haiti	1995	Botswana	2004	Madagascar	1998
Haiti	2000	Belarus	1994	Madagascar	2001
Haiti	2006	Belarus	1995	Madagascar	2002
Malaysia	1990	Belarus	2000	Madagascar	2006
Malaysia	1995	Belarus	2001	Madagascar	2007
Malaysia	1999	Belarus	2004	Mali	1992
Malaysia	2004	Belarus	2006	Mali	1997
Malaysia	2008	Belarus	2008	Mali	2002
Peru	1995	Ghana	1992	Zambia	1991
Peru	2000	Ghana	1996	Zambia	1996
Armenia	1995	Ghana	2000	Zambia	2001
Armenia	1996	Cameroon	1992	Zambia	2006
Armenia	1998	Cameroon	1997	Cambodia	1993
Armenia	1999	Cameroon	2002	Cambodia	1998
Armenia	2003	Cameroon	2004	Cambodia	2003
Armenia	2007	Cameroon	2007	Cambodia	2008
Armenia	2008	Georgia	1991	Tanzania	1995
Gabon	1990	Georgia	1992	Tanzania	2000
Gabon	1993	Georgia	1995	Tanzania	2005

Appendix B

Summary Statistics.

Variable	Observations	Mean	Standard deviation	Minimum	Maximum
Linkage	147	0.432	0.295	0	0.97
Ban	147	0.396	0.497	0.07	3.86
Bias	147	1.5	0.83	0.17	3.92
Violence	147	0.385	0.218	0.05	0.96
Intimidation	147	1.47	0.73	0.08	2.91
Fraud	147	2.34	0.917	0.5	3.96
Censor	147	2.01	0.87	0.33	3.93
Legislative majority	147	0.417	0.493	0	1
State capacity	147	91.1	8.11	70	100
Main elections	147	0.74	0.433	0	1
GINI	147	42.35	11.39	20.4	73.3
Natural rent	147	6.5	8.33	0	40.1
GDP grow	147	2.78	6.33	−21.1	15.37

GDP: gross domestic product.

Appendix C

Correlation Between Independent Variables (Pearson’s Correlation Coefficient).

	Main elections	Legislative majority	GINI	Capacity	Growth	Linkage	Rent
Main		0.158	0.02	0.013	−0.02	0.156	−0.122
Legislative majority	0.158		0.238**	0.082	−0.060	0.058	0.356**
GINI	0.02	0.238**		0.196*	−0.071	−0.353**	0.207*
Capacity	0.013	0.082	0.196*		−0.080	0.127	0.082
Growth	−0.02	−0.060	−0.071	−0.080		−0.176*	−0.083
Linkage	0.156	0.058	−0.353**	0.127	−0.176*		−0.137
Rent	−0.122	0.356**	0.207*	0.082	−0.083	−0.137

P-values: *p < 0.05. **p < 0.01. (2-tailed)

Appendix D

Models With Alternative Measures of Electoral Fraud.

Explanatory variables	Vote registry		Irregularities		Fraud^a
Linkage	−0.085	−0.091	−0.164	−0.120	−0.014	−0.025
	(0.089)	(0.100)	(0.110)	(0.121)	(0.047)	(0.045)
Legislative majority		−0.028		0.133*		−0.071^†
		(0.035)		(0.062)		(0.042)
State capacity		−0.054		0.023		0.099*
		(0.048)		(0.079)		(0.040)
Main elections		0.039		0.044		−0.089*
		(0.029)		(0.053)		(0.040)
GINI		−0.017		0.028		0.050
		(0.058)		(0.093)		(0.045)
Natural rent		0.007		−0.068		−0.079^†
		(0.063)		(0.100)		(0.045)
GDP growth		0.017		−0.076		0.025
		(0.051)		(0.070)		(0.036)
Constant	2.75***	2.75***	1.59***	1.57***	0.262***	0.281***
	(0.098)	(0.104)	(0.119)	(0.121)	(0.048)	(0.043)
Number of observations	147	147	147	147	122b	122
Number of groups:country	34	34	34	34	33	33
R² level 1 (within)	0.000	0.000	0.000	0.0481	0.000	0.047
R² level 2 (between)	0.0225	0.0253	0.0834	0.140	0.0017	0.177

GDP: gross domestic product.

Fraud allegations by Western monitors.

The number of observations in the last two models is lower because V-dem (Coppedge et al., 2019b) does not provide data for all cases.

Multilevel regression analysis, clustered by country with random effects. Robust standard errors in parentheses. P-values: †p < 0.1; *p < 0.05. **p < 0.01. ***p < 0.001.

Acknowledgements

The author thanks Jakub Lysek, Editor Justin Fisher and an anonymous reviewer for their insightful criticisms and constructive suggestions.

Declaration of conflicting interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) received no financial support for the research, authorship and/or publication of this article.

ORCID iD

Jaroslav Bílek

Notes

Author Biography

Jaroslav Bílek is a postdoctoral research fellow at the Department of Politics, University of Hradec Králové. His research interests cover hybrid regimes, electoral manipulation, and civil-military relations.

References

Beaulieu

Hyde

(2009) In the Shadow of Democracy Promotion: Strategic Manipulation, International Observers, and Election Boycotts. Comparative Political Studies 42 (3): 392–415.

Bhasin

Gandhi

(2013) Timing and Targeting of State Repression in Authoritarian Elections. Electoral Studies 32 (4): 620–631.

Birch

(2011) Electoral Malpractice. Oxford: Oxford University Press.

Brownlee

(2012) Democracy Prevention: The Politics of the U.S.-Egyptian Alliance. New York: Cambridge University Press.

Bunce

Wolchick

(2011) Defeating Authoritarian Leaders in Postcommunist Countries. New York: Cambridge University Press.

Castañeda

(1996) The Mexican Shock: Its Meaning for the United States. New York: The New Press.

Cheeseman

Klaas

(2018) How to Rig an Election. New Haven, CT; London: Yale University Press.

Clark

Linzer

(2015) Should I Use Fixed or Random Effects? Political Science Research and Methods 3 (2): 399–408.

Coppedge

Gerring

Knutsen

, et al. (2019a) V-Dem Codebook V9. Varieties of Democracy (V-Dem) Project.

10.

Coppedge

Gerring

Knutsen

, et al. (2019b) V-Dem [Country-Year/Country-Date] Dataset v9. Varieties of Democracy (V-Dem) Project.

11.

Cruz

Keefer

Scartascini

(2018) Database of Political Institutions 2017 (DPI2017). Numbers for Development, Inter-American Development Bank.

12.

Diamond

(2002) Thinking About Hybrid Regimes. Journal of Democracy 13 (2): 21–35.

13.

Donno

(2013) Defending Democratic Norms: International Actors and the Politics of Electoral Misconduct. Oxford: Oxford University Press.

14.

Donno

Roussias

(2012) Does Cheating Pay? The Effect of Electoral Misconduct on Party Systems. Comparative Political Studies 45 (5): 575–605.

15.

Dresden

Howard

(2015) Authoritarian Backsliding and the Concentration of Political Power. Democratization 23: 1122–1143.

16.

Harvey

Mukherjee

(2018) Methods of Election Manipulation and the Likelihood of Post-Election Protest. Government and Opposition. Epub ahead of print 12 November. DOI: 10.1017/gov.2018.38.

17.

Hauser

(2017) Does Electoral Manipulation Vary? Examining the Conditions for Instrumental and Informational Manipulation in Post-Soviet Elections. The Soviet and Post-Soviet Review 45 (2): 5–50.

18.

Henderson

(1991) Conditions Affecting the Use of Political Repression. Journal of Conflict Resolution 35 (1): 120–142.

19.

Huntington

(1991) The Third Wave: Democratization in the Late Twentieth Century. Norman, OK: University of Oklahoma Press.

20.

Hyde

(2007) The Observer Effect in International Politics: Evidence from a Natural Experiment. World Politics 60: 37–63.

21.

Lehoucq

(2003) ELECTORAL FRAUD: Causes, Types, and Consequences. Annual Review of Political Science 6: 233–256.

22.

Levitsky

Way

(2010) Competitive Authoritarianism: Hybrid Regimes after the Cold War. Cambridge: Cambridge University Press.

23.

McFaul

(2005) Transitions from Postcommunism. Journal of Democracy 16 (3): 5–19.

24.

Magaloni

(2006) Voting for Autocracy. New York: Cambridge University Press.

25.

Martí i Puig

(2016) Nicaragua: Desdemocratización Y Caudillismo. Revista de Ciencia Política 36 (1): 239–258.

26.

Michalik

(2015) Multiparty Elections in Authoritarian Regimes: Explaining Their Introduction and Effects. Mannheim: Springer.

27.

Przeworski

Gandhi

(2009) Holding onto Power by Any Means? The Origins of Competitive Elections. Working Paper, Department of Politics, New York University, New York.

28.

Schedler

(2002) The Menu of Manipulation. Journal of Democracy 13 (2): 37–50.

29.

Schedler

(2013) The Politics of Uncertainty: Substaining and Subverting Electoral Authoritarianism. Oxford: Oxford University Press.

30.

Seeberg

(2014) State Capacity and the Paradox of Authoritarian Elections. Democratization 21 (7): 1265–1285.

31.

Simpser

(2013) Why Governments and Parties Manipulate Elections: Theory, Practice, and Implications. New York: Cambridge University Press.

32.

Slater

(2011) Book Review of Levitsky and Way (2010). Perspectives on Politics 9 (2): 385–388.

33.

Tanneberg

Stefes

Merkel

(2013) Hard Times and Regime Failure: Autocratic Responses to Economic Downturns. Contemporary Politics 19 (1): 115–129.

34.

Tansey

(2016) The International Politics of Authoritarian Rule. New York: Oxford University Press.

35.

Thaler

(2016) Nicaragua: A Return to Caudillismo. Journal of Democracy 28 (2): 157–169.

36.

Van Ham

Lindberg

(2015) From Sticks to Carrots: Electoral Manipulation in Africa, 1986-2012. Government and Opposition 50 (3): 521–548.

37.

Weyland

(2013) The Threat from the Populist Left. Journal of Democracy 24 (3): 18–32.