Electoral authoritarianism and economic control

Abstract

While electoral revolutions in the Philippines and the post-Communist world have ousted dictators, autocrats from Mexico to Zimbabwe have cemented their rule through regular multi-party elections. Why do elections sometimes undermine authoritarian regimes while at other times they help sustain them? I argue that a dictator’s control over the economy conditions the effect of authoritarian elections. Where rulers command the heights of the economy, elections are more easily manipulated to sustain their rule. But where such control is lacking, elections may spur regime change. In a cross-national study of autocracies from 1970 to 2006, I find that as incumbent control over the economy increases, elections are less likely to lead to regime breakdown. Where economic control is at its lowest, elections increase the risk of regime collapse. Thus, research on authoritarianism needs to supplement the study of authoritarian regime types and institutional characteristics with a focus on the rulers’ control over the state and the economy.

Keywords

Authoritarianism elections regime breakdown economic control electoral manipulation state capacity

Introduction

Today, the label ‘electoral authoritarianism’ could be applied to the majority of the world’s autocracies, as more than half of these regimes have recently held a multi-party election. But the effect of such electoral events on regime stability is disputed. Do multi-party elections sustain authoritarian rule or promote its downfall?

Following the Serbian parliamentary and presidential elections in September 2000, the electoral commission abandoned the tallying of votes as they realized that the results were not in favor of incumbent President Slobodan Milošević, causing hundreds of thousands of protesters to storm the federal parliament building. The coal industry was paralyzed by strikes and central elites of the ruling front defected (Bunce and Wolchik, 2011: 110–111). Within weeks, Milošević gave up power in the ‘Bulldozer Revolution’. Along with a number of other cases, this has spurred a belief in elections as a force for transforming dictatorships. Elections are theorized to destabilize and sometimes even democratize autocracies (e.g. Bunce and Wolchik, 2011; Howard and Roessler, 2006; Kuntz and Thompson, 2009; Levitsky and Way, 2010; Lindberg, 2006; Schedler, 2006).

But other rulers show no signs of conceding power in spite of multi-party elections. In Zimbabwe, President Robert Mugabe cemented his rule by winning the heavily manipulated yet relatively peaceful 2013 elections. The main opposition party, the Movement for Democratic Change – Tsvangirai, lost 30 seats and split the following year. Similarly, in the authoritarian regimes of Malaysia and Singapore, Egypt under Hosni Mubarak, and Mexico in the 20th century, rulers have used regular, multi-party elections to their advantage, leading scholars to conclude that multi-party elections are just another tool adopted and adapted to sustain authoritarian rule (e.g. Blaydes, 2011; Lust, 2009; Magaloni, 2006).

But although the effects of authoritarian institutions on regime stability are widely studied, little comparative evidence exists for when, where, or why elections would sustain authoritarian rule rather than undermine it. I argue that the effect of authoritarian elections on regime stability depends on the central capacities available to the ruling front. It has recently been argued that elections are more likely to stabilize dictatorships where rulers dispose over strong states and, conversely, that elections – at least when they are somewhat competitive – may spur democratization if sufficient international pressure is applied (Donno, 2013; Levitsky and Way, 2010; Seeberg, 2014). But rulers lacking an efficient bureaucracy or facing international pressure – as is the case in Zimbabwe under the rule of the Zimbabwe African National Union-Patriotic Front (ZANU-PF) – still manage to sustain their rule through multi-party elections.

I propose that not only the administrative and repressive arms of the state and the role of the international community are important in understanding the role of elections. Rulers’ domination of the economy matters. Control over the economy, partly through means of the state such as a large public sector and regulation of private business (Greene, 2007), but also through control over natural resource rents (Levitsky and Way, 2010), may allow a dictator to manipulate elections to a degree where they sustain her rule rather than undermine it.

This article tests the claim that multi-party elections are more likely to stabilize dictatorships where rulers control the economy and more likely to break down following elections where economic control is limited or absent. First, I theorize how economic control matters regarding the effect of elections as rulers rely on economic domination to carry out electoral manipulation. Second, I test the claim on cross-national data of all authoritarian regimes from 1970 to 2006. I find that as economic control increases, the likelihood decreases that the regime breaks down following elections. Authoritarian elections, however, are not necessarily regime-stabilizing even in cases where the rulers control the economy. Rather, economic control reduces the risk of breakdown following elections. The findings imply that we need to supplement the study of authoritarian regime types and institutional characteristics with a focus on the economic capacity of the ruler.

Authoritarian elections and regime stability

Research on electoral authoritarianism has blossomed in the 21st century and a multitude of diverging arguments on the effect of elections has been presented. Elections won with supermajority victories signal invincibility (Magaloni, 2006), they are used to co-opt or split the opposition (Malesky and Schuler, 2010), and they allow the ruling party to bind voters through patron–client relationships, thus stabilizing the regime (Lust, 2009). But surprising electoral setbacks for the ruling party may also cause the elite to rebel (Magaloni, 2006: 258) and fraudulent elections can become focal points for opposition mobilization and citizen protests (Howard and Roessler, 2006; Kuntz and Thompson, 2009) making authoritarian elections inherently destabilizing. Thus, the question, do elections threaten authoritarian rule or do they help sustain it?

Empirical results are ambiguous. Cross-nationally, it has been shown that electoral autocracies are more likely to democratize when they break down (Brownlee, 2009; Miller, 2015). But electoral autocracies are neither more – nor less – likely to remain stable than are non-electoral autocracies (Brownlee, 2009). This has led some scholars to claim that elections have no effects on the stability of authoritarian regimes (Brownlee, 2007). Countless, thorough case-studies, however, suggest otherwise. They instead use case evidence to account for either the stabilizing or destabilizing effect of authoritarian elections (Bunce and Wolchik, 2011; Greene, 2007; Kuntz and Thompson, 2009; Levitsky and Way, 2010; Magaloni, 2006).

One factor whose presence (or absence) is documented in certain cases to have secured (or impeded) a stabilizing effect of authoritarian elections is the leadership’s extensive control over the economy (Greene, 2007; Levitsky and Way, 2010). The Institutional Revolutionary Party (PRI), ruling Mexico from 1929, kept its vote share above 60% until the opposition started gaining ground in the 1980s. One of the party’s main resources was its control over the economy. In the early 1980s, state-owned enterprises accounted for almost a quarter of gross domestic product (GDP) and the federal government employed more than three million people (Camp, 1990; Greene, 2007: 101–103). These resources allowed party candidates to trade infrastructure, electricity, and public sector jobs in return for votes and government resources were employed in ruling party campaigns (Fox, 1994), in turn helping to sustain electoral hegemony. In a study of eight dominant party autocracies, Greene shows that a large public sector controlled by the incumbent increases the incumbent party’s margin of victory in elections (Greene, 2010), thus supporting the notion that elections are more likely to stabilize authoritarian rule where economic control is high.

Building on the arguments introduced by Greene (2007; 2010) and Levitsky and Way (2010), I test whether economic control can account for the diverging effect of elections across cases. Rather than elections playing no role in stabilizing authoritarian regimes, it may be that authoritarian rulers’ control over the economy, whether through the state apparatus or natural resources, allows them to stabilize their regimes through multi-party elections.

Economic control and electoral manipulation

In recent years, case studies of electoral authoritarian regimes have investigated the role of incumbent economic monopoly (Greene, 2007) or citizens’ economic autonomy from rulers (McMann, 2006). It might not be the overall level of economic growth but rather the degree of incumbent control over the economy that affects the role of authoritarian elections in promoting regime stability. Whereas several studies have pointed to the importance of an incumbent resource advantage for controlling elections (Greene, 2007; Levitsky and Way, 2010), none have systematically theorized it, nor tested the claim on cross-national data.¹

I define economic control as the ruler’s domination of economic resources, including natural resource revenues, land, and employment opportunities. Thus, economic control is clearly related to administrative state capacity. When the bureaucracy is controlled by the ruler, a large public sector directly contributes to the incumbent’s degree of economic control, as it allows for some degree of control over employment opportunities, particularly when combined with a weak private sector or an economic crisis. And a politicized bureaucracy can leave the incumbent in control of import and export licenses, thus controlling international trade. But underlying economic structures independent of the state bureaucracy, such as natural resources under rulers’ control, may also contribute to an incumbent economic monopoly (Levitsky and Way, 2010: 66). Natural resources controlled by the regime have been shown to prolong authoritarian rule (Ross, 2012). Furthermore, economic resources can contribute to both administrative and repressive state capacity. Thus, economic control is in many ways related to but also distinct from state capacity in general.

I argue that control of the economy affects the authoritarian leader’s ability to remain in power not only in general but particularly in the context of authoritarian elections because it enables a number of strategies of electoral control that bind internal elites and voters and prevent mobilization of protesters and opposition members.

First, a ruler with extensive economic control has unique means of manipulating voters’ preference formation, for instance through using propaganda, abusing state spending, or preventing the opposition from campaigning. Controlling the economy provides rulers with the funds necessary to campaign. Money can be generated from state-owned enterprises, public money can be funneled to incumbents directly from the budget, and candidates may receive contributions from private businesses with links to the state (Greene, 2007: 40–41). Alternatively, natural resource wealth may be directed to party-coffers. Dictators governing resource-rich states can distribute resources in order to prevent the formation of opposition (Ross, 2012) or can use the money generated from natural resources to fund election campaigns. Thus, manipulation of voters’ preference formation may work regardless of whether the ruler controls the economy through the state apparatus or through access to natural resource wealth independent of the formal state bureaucracy.

An economic monopoly is also an effective way to put pressure on opposition-friendly media outlets, particularly when the economy is controlled through the formal state apparatus: McMann describes how private businesses in the Russian province of Ul’ianovsk, in order to stay on friendly terms with the local government, which they depend on for licenses, etc., avoid advertising in independent newspapers, thus depriving the independent media of essential sources of finance (McMann, 2006: 74). As voters affected by subtle manipulation become more likely to support the ruling front, elections are also more likely to signal invincibility and thus deter internal elites from defecting and the opposition from mobilizing. Where rulers fail to dominate the economy, voters are harder to control and elections are more likely to send signals encouraging elite defections and opposition mobilization, thus jeopardizing regime stability.

Second, rulers can use economic control to manipulate voters’ preference expression. Thus, voters may be intrinsically inclined to support the opposition, but by offering particularistic benefits, rulers can increase voters’ benefits of voting for the regime. Again, the strategy may rely on extensive state control of the formal economy or unofficial access to resource wealth. One of the most common arguments in the rentier state literature is that oil wealth allows dictators to spend more money on patronage (Ross, 2012). In relation to elections, Blaydes has demonstrated the existence of budget cycles and their connection to the centralized economy in Mubarak’s Egypt (Blaydes, 2011) and Magaloni presents convincing evidence of the Mexican PRI’s abuse of state funds through the PRONASOL, a poverty-relief program, to reward swing municipalities that did not elect an opposition candidate (Magaloni, 2006).

Furthermore, a large public sector and few options of employment outside of the reach of the ruling group may allow rulers to pressure voters into compliance. Both the United Malays National Organization in Malaysia and ZANU-PF in Zimbabwe have traditionally made sure to inform public sector workers that a vote for the opposition could entail job loss or a transfer to the outskirts of the country (Crouch, 1996: 127; Kriger, 2005: 23). Thus, rulers may use their control over the economy to pressurize voters into supporting the regime, thus reducing the probability that elites will defect and an opposition will mobilize. Where economic control is lacking, elite defections – and ultimately regime breakdown – is more likely.²

Third, the ruler may use economic harassment to prevent opposition and protesters from mobilizing. This strategy typically relies on formal state control over the economy rather than access to natural resources. McMann explains how, in Russia and Kyrgyzstan, ‘The ability to make a living independent of the state is critical to the practice of democracy; otherwise, citizens will avoid activism for fear of economic reprisals by the government’ (McMann, 2006: 4). The individual will not risk her livelihood for political activism. Thus, in order for the opposition to thrive in electoral authoritarian regimes, there must be a sufficient supply of jobs outside of the public sphere and hence outside of government control.

The dynamics are largely the same when economic harassment is used towards potential protesters. Although individual protesters may be hard to identify, leaders and organizers of protests, or entire districts in which protests emerge, may be threatened with withdrawal of public support and jobs. In Belarus, the Lukashenko government decided to limit employment at state-owned enterprises to one-year contracts that could be extended at the will of the management and declared that academic degrees could be withdrawn based on opposition activism (Silitski, 2005: 91–92). Thus, economic control can prevent elections from becoming regime-destabilizing because opposition is prevented from emerging and post-electoral protests are limited. On the other hand, where rulers’ control over the economy is limited, elections are also more likely to run out of hand, causing protests and potentially regime breakdown.

In sum, economic control, whether informal or through the state apparatus, will enable various forms of electoral manipulation that prevent voters from supporting the opposition, deter elites from defecting and the opposition from mobilizing, and deter potential protesters from taking to the street. In turn, elections become more likely to stabilize the regime and less likely to lead to regime breakdown. Thus, as economic control increases, elections should become less likely to lead to the breakdown of an authoritarian regime. It may be that we see regime-sustaining elections primarily in dictatorships where economic control, whether formal or informal, is high. And that regime breakdown will be more likely following elections where rulers’ lack economic control. Empirically, the various strategies of manipulation may be hard to separate in large-N analyses, but the theory generates three more general propositions.

H1: The effect of authoritarian multi-party elections on the likelihood of regime breakdown will decrease with higher degrees of incumbent control over the economy.

H2a: The effect of authoritarian multi-party elections on regime breakdown will be negative for high levels of economic control.

H2b: The effect of authoritarian multi-party elections on regime breakdown will be positive for low levels of economic control.

A final matter of interest is the time frame within which elections can be expected to affect regime dynamics. Elections may have a negative effect on regime stability immediately following the actual contest (for instance by provoking post-electoral protests), but they can also be expected to have stabilizing effects for years to come (for instance by generating legitimacy). I expect the effect of elections to be strongest in the first one to two years following the actual contest. But the conditioning effect of economic control may be present at least for the first five years after the elections, where rulers can refer back to a recent electoral victory as a way of deterring elite defections or generating legitimacy. After those initial years, however, unless new elections are held, I expect the effect of elections – and thus the conditioning effect of economic control – to wear off.

Method and data

To test the effect of economic control on the relationship between authoritarian multi-party elections and regime stability (defined as the absence of breakdown), I rely on data of all authoritarian regimes from 1970 to 2006. To identify authoritarian regimes, I use the data provided by Geddes, Wright, and Frantz (GWF) (Geddes et al., 2014). They code as authoritarian all regimes in which the leader did not acquire power through a reasonably competitive election; a democratically elected government that circumvented electoral competition; or where the military wielded substantial power.

The dependent variable is regime breakdown as operationalized by the GWF data. It captures transitions from authoritarian rule to democracy as well as transitions from one autocracy to the next. It is scored 1 if one or more regime breakdowns occurred in the given year.

The independent variable is multi-party elections in authoritarian regimes. I score all direct, national elections (from NELDA; Hyde and Marinov, 2012) as multi-party competitions if parties outside the ruling front existed and were represented in the legislature (data from Cheibub et al., 2010). Elections are authoritarian if they were held in an authoritarian regime (operationalized above). The detailed coding procedure is described in the online Appendix.

How recent must an election have been in order for a regime to be classified as electoral authoritarian? As the time frame during which elections are likely to provoke regime breakdown or promote regime stability is larger than the year following elections, I test for the effect of elections held within the past one, three, five, and seven years.³ Thus, depending on the model, the independent variable – authoritarian multi-party elections – is scored 1 if an election has been held within the past one, three, five, or seven years.

I proxy the conditioning variable – the incumbent’s economic control – through three measurable dimensions. First, the size of the public sector gauges citizen’s dependence on the rulers both for jobs and benefits. If the public sector is larger, citizens will be more likely to earn their living in public jobs. Furthermore, money can be directed from the state budget to the rulers and be spent for partisan purposes. I use the log of government spending as a share of GDP (Penn World Table, 2013).

Second, where private business is heavily controlled, citizens are more dependent on the government. To measure this, I use an index of regulation of credit markets, labor markets, and business, including information on the costs of starting a business, licensing restrictions, and ownership of banks (Gwartney et al., 1996: 243–248).⁴

Third, income from natural resources can be extracted with little reliance on labor and provide rent windfalls to the dictator that can be distributed to loyal followers (Ross, 2012). To measure this, I use the log of total income from natural resources per capita (Haber and Menaldo, 2011).

To test the hypotheses, I examine the effect of the separate measures of economic control as well as an index comprising all three. As an autocrat may control one dimension without the regime scoring highly on the others (Pearson’s correlations vary between −0.19 and 0.21, see online Appendix), the index is additive: it is constructed by summing all three variables and dividing by three. Medium degrees of economic control may be attained by scoring high on one or two dimensions or by having mean values on all three.

To address concerns of endogeneity, Figure B (online Appendix) shows that elections are neither more nor less common in dictatorships with high or low degrees of economic control. It is thus unlikely that elections are merely introduced by dictators who control the economy – or forced upon those who do not. Further, I control for the first multi-party election held by every authoritarian ruling group (data on ruling groups from GWF). The logic is that whereas elections are hard to abandon once institutionalized, the choice to introduce elections may be endogenous. As dictators may be forced to introduce elections once they are already weak – and thus more likely to experience regime breakdown – including first elections may contribute to overestimating the relationship between elections and breakdown.

Finally, I control for factors that may affect dictators’ decisions to hold elections, regime stability, and economic control, namely wealth and economic growth (Haber and Menaldo, 2011). As I do not expect economic control to be the only factor that affects the role of elections, I include controls for the tax-to-GDP ratio (to proxy administrative capacity) (Kugler and Tammen, 2012, available in Andersen et al., 2014) and the logarithm of military spending/capita (to proxy coercive capacity) (Singer et al., 1972). I also include a cubed version of time and age of the regime. In alternative specifications, I control for authoritarian regime type (Geddes et al., 2014), protests (Banks and Wilson, 2012), media freedom (Whitten-Woodring and Belle, 2014), executive constraints (Marshall et al., 2011), and the Cold War period.

Does economic control condition the effect of authoritarian elections?

I test the argument that economic control conditions the effect of authoritarian multi-party elections on regime stability relying on random- and fixed-effect logistic regression models. I first focus on the index of economic control before turning to the separate effects of the subcomponents. Model 1 in Table 1 investigates the effect of having held at least one multi-party election within the past year on the likelihood of regime breakdown. Focusing on changes within countries (fixed effects) rather than comparing across countries, the model shows that having held a very recent multi-party election significantly increases the risk of breakdown for an authoritarian regime while increasing growth reduces the risk of regime collapse. When proxies for administrative capacity, coercive capacity, and economic control are introduced in Model 2, the effect of elections is unchanged. Surprisingly, the proxies for authoritarian capacities do not reduce the risk of regime breakdown.

Table 1.

Elections, regime breakdown, and economic control (fixed effects).

	(1)	(2)	(3)
Multi-party elections (within the past year)	1.354^*** (0.354)	1.471^*** (0.382)	1.387^*** (0.385)
Economic control index		0.973(0.785)	1.128(0.796)
Multi-party elections* Economic control index			−1.048^* (0.609)
Military expenditure/capita		0.321(0.461)	0.321(0.464)
Tax-to-gross domestic product (GDP) ratio		0.424(3.542)	0.464(3.533)
GDP/capita	0.556(1.486)	0.304(1.574)	0.359(1.583)
Growth	−0.056^** (0.023)	−0.054^** (0.023)	−0.059^** (0.024)
Observations	975	975	975
Breakdowns	66	66	66

Notes: Dependent variable is regime breakdown. Explanatory variable is coded 1 if one or more multi-party elections were held in the past year. All covariates except time trends are lagged one year. Controls for time and regime age (cubed) are included but not shown. Logit model with country fixed effects and robust standard errors clustered on country.

p < 0.10; ^**p < 0.05; ^***p < 0.01.

Model 3 provides the first test of H1 as it explores whether the effect of a very recent election changes as economic control increases. The coefficient on elections represents the effect of a recent multi-party election in cases where economic control is 0 (roughly its average), revealing that at average levels of economic control, introducing multi-party elections significantly increases the risk of regime breakdown in the following year. The direct effects of coercive and administrative capacity remain insignificant, but the interaction effect of economic control is negative and significant (p < 0.1). Thus, the positive effect of elections on the likelihood of regime breakdown is reduced as economic control increases. The finding supports H1. Controlling for administrative and coercive capacity, a dictatorship which has held an election within the past year is less likely to collapse when control over the economy increases.⁵

Table C (online Appendix) shows that the finding is robust to control for regime type, protests, media freedom, executive constraints, and the Cold War period. The finding also holds when controlling for first elections indicating that endogeneity of the election variable is not driving results.⁶

Does economic control also condition the effect of more distant elections? As the time frame during which elections are likely to provoke regime breakdown or promote regime stability is larger than the year following elections, I test the relationship across a variety of time horizons. The interaction effect of economic control is negative throughout and significant when comparing effects both within and across countries for elections held in the past five and seven years (p < 0.1) (see Table D, online Appendix). But the results are not robust to the three-year interval or relying on fixed effects – potentially because the combination of fixed effects and three-, five-, and seven-year running averages of economic control reduces the sample size markedly.⁷ There is thus some evidence that economic control also conditions the effect of authoritarian elections in the longer term.

The support for H1 is somewhat less consistent across time horizons than when looking at the year immediately following elections. But this picture changes when focusing on the subcomponents of economic control: Regulation (of business, labor and credit markets) and natural resource wealth have no conditioning effect on the relationship between elections and regime stability (see Tables E and F, online Appendix). Including these two parameters in the index of economic control has thus reduced the effects displayed in Table 1 and Table D.

Table 2 shows the effect of the subcomponent of government spending on the relationship between elections and regime stability across various time horizons. Increasing government spending consistently reduces the destabilizing effect of authoritarian elections regardless of whether they were held one, three, five, or seven years prior and this conditioning effect is statistically significant with both random and fixed effects across nearly all time intervals. The results are substantially unchanged when controlling for a regime’s first elections (not reported). Thus, the effect of the incumbent’s control over the economy is driven by government spending rather than resource wealth and focusing solely on this parameter lends more consistent support to H1.

Table 2.

Government spending, elections, and regime stability – varying time horizons.

Elections in the past	(1)	(2)	(3)	(4)	(5)	(6)	(7)	(8)
	1 year		3 years		5 years		7 years
	Re	Fe	Re	Fe	Re	Fe	Re	Fe
Multi-party elections	3.293^*** (1.190)	4.439^*** (1.553)	1.419(1.281)	2.956^* (1.566)	1.668(1.287)	4.448^** (1.869)	2.147(1.315)	5.133^** (2.034)
Government spending	0.121(0.177)	0.643(0.560)	0.040(0.201)	0.541(0.639)	0.174(0.232)	0.638(0.842)	0.205(0.221)	0.370(1.005)
Multi-party elections* Government spending	−0.795^* (0.424)	−1.298^** (0.570)	−0.328(0.434)	−0.933^* (0.530)	−0.423(0.447)	−1.509^** (0.645)	−0.558(0.441)	−1.638^** (0.698)
Military expenditure	0.012(0.053)	0.297^* (0.154)	−0.028(0.066)	0.308(0.229)	−0.025(0.069)	0.586^** (0.290)	−0.043(0.070)	0.275(0.240)
Tax-to-gross domestic product (GDP) ratio	−3.440^*** (1.208)	0.029(2.444)	−4.431^*** (1.418)	−2.802(3.266)	−4.312^*** (1.425)	−2.598(3.955)	−4.331^*** (1.475)	−1.167(4.032)
GDP	−0.521^*** (0.174)	−0.620(0.733)	−0.445^** (0.175)	−0.198(0.906)	−0.357^** (0.167)	−0.179(1.037)	−0.360^** (0.182)	0.740(1.031)
Growth	−0.013^* (0.007)	−0.013(0.008)	−0.039^*** (0.015)	−0.034^** (0.015)	−0.038^** (0.016)	−0.035^** (0.016)	−0.028(0.018)	−0.036(0.023)
Constant	4.567^** (2.230)		4.798^** (2.292)		3.903^* (2.338)		2.760(2.982)
Observations	2550	1787	2411	1617	2238	1414	2056	1291
Breakdowns	109	109	104	104	94	94	82	82

Notes: Dependent variable is regime breakdown. Explanatory variable is coded 1 if one or more multi-party elections were held in the past one, three, five, and seven years respectively. Government spending, military expenditure, and tax-to-GDP ratio are running averages for the past one, three, five, and seven years respectively. Other covariates are lagged one year. Controls for time and regime age (cubed) are included but not shown. Logit model with random (re) and fixed effects (fe) and robust standard errors clustered on country.

p < 0.10; ^**p < 0.05; ^***p < 0.01.

This picture fits cases such as PRI-Mexico and Mubarak’s Egypt where budget cycles around elections indicate that government spending is used to manipulate elections. But there is no general evidence for dynamics such as those in Zimbabwe where official state coffers are often virtually empty but resources from diamonds are allegedly appropriated by the ruling party to abuse elections (International Crisis Group, 2013: 24–25). Economic control seems to work through the state apparatus rather than outside of it.

The initial results support H1: As authoritarian incumbents’ tighten their grasp of the economy, the risk of regime breakdown following multi-party elections drops. But the regression coefficients themselves do not offer a test of H2a–b, namely whether elections are indeed stabilizing at high levels of economic control and destabilizing at low levels. I therefore proceed to graphically test H1 and H2a–b.

Figure 1 visualizes the conditional relationship.⁸ Panel A shows the marginal effect of having held an election within the past year (column 1) and the past five years (column 2) as compared to having no multi-party elections as economic control increases. In both cases, the risk of experiencing regime breakdown following multi-party elections clearly drops as economic control increases. Panel B, which plots the same effect on the logit scale to allow for correct estimation of confidence intervals, shows that even the one-year random effect which was statistically insignificant (Table D, online Appendix) is in fact significant for a large part of the spectrum. This finding lends additional support to H1. Even for the models where the conditional effect of economic control was initially insignificant (see Table D, online Appendix), increasing economic control – at least for low and medium levels – significantly reduces the risk of breakdown following elections:⁹ For 85% of the included country-years, increasing control over the economy makes regime breakdown significantly less likely following multi-party elections (p < 0.1) (column 1).

Figure 1.

Marginal effect of elections for increasing economic control.

This finding also supports H2b: for low levels of economic control, holding an election increases the risk of regime breakdown. Electoral autocracies that exhibit low levels of economic control, such as Kenya in the 1990s, will be more prone to breakdown than will their non-electoral counterparts. Furthermore, Panel A shows that whereas having held a multi-party election within the past year at extremely low levels of economic control increases the probability of regime breakdown by approximately 30%, having held one within the past five years raises the risk by 20%. Thus, elections pose a higher risk for autocrats in their immediate aftermath.

The finding that elections are more likely to lead to regime breakdown when the incumbent loses control of the economy is in line with what was witnessed in Mexico from the 1980s onwards. When the debt crisis sparked a turn towards neo-liberal policies, causing state-owned enterprises to be privatized and public sector employment to be reduced, the regime lost control over elites and voters and eventually lost power (Greene, 2007; Magaloni, 2006).

The expectation that elections should reduce the risk of breakdown at high levels of economic control (H2a), however, does not find support. Although the marginal effect is negative for high levels of economic control in both the one- and five-year models, it is statistically insignificant. There is no immediate evidence that dictators – even with high levels of economic control – can use elections to stabilize their rule: Their risk of breakdown is neither substantially nor significantly different from high-control regimes without elections.

To illustrate the results, the predicted probability of breakdown for an electoral and a non-electoral regime at high and low levels of economic control respectively is pictured in Figure 2.

Figure 2.

Predicted probability of regime breakdown.

Like Figure 1, it shows that at low levels of economic control, electoral regimes are indeed more likely to break down than are non-electoral regimes (Figure 2, Panel A). At levels of economic control corresponding to those in Kenya in the early 1990s (bottom fifth percentile of sample), the risk of breakdown when a multi-party election has been held in the past five years is – all else equal – around 10%. If no elections were held, it is only 2.5%. When incumbents increase their economic domination, the risk of breakdown following elections decreases dramatically, as was also documented in Figure 1. For levels of economic control corresponding to China in the 1980s (the 95th percentile), the risk of breakdown is reduced to 2.3%. For these regimes, however, having held a recent election does not significantly reduce the risk of breakdown (Figure 2, Panel B).

Whether focusing on an index of economic control or its individual components, increased incumbent control over the economy does indeed make breakdown following elections less likely (H1). At low levels of economic control, holding elections increases the risk of regime breakdown in the years that come (H2b). But there is no general evidence that elections become stabilizing at high levels of economic control (H2a).

The rejection of H2a, however, is reliant on the definition of authoritarian elections which in this article includes elections in which the incumbent lost and peacefully handed over power. But since the incumbent peacefully gave up power, this decision may already have been made prior to the election and the capacities of the ruler may not have affected electoral dynamics. If this is the case, the inclusion of such elections in the sample will boost the relationship between elections and authoritarian regime breakdown while the breakdown was in fact underway before the actual elections. I therefore run an additional analysis where elections in which the incumbent ran, lost, and peacefully gave up power (such as in Senegal in 2000) are not included but the breakdown is instead recorded in the year prior to elections.

In correspondence with the previous results, Table G (online Appendix) reveals that the interaction effect of the index of economic control is negative and significant (apart from the three-year interval) when excluding elections where the ruler voluntarily handed over power. These results are robust to a control for first elections, indicating that the results are not driven by weak dictators being forced to introduce elections and breaking down shortly thereafter (results not reported).

The effects are visualized in Figure 3. For this sub-sample of elections, H2a finds support: For high levels of economic control, electoral regimes are significantly less likely to experience breakdown (2.1%) than non-electoral regimes (6.1%) (Panel B).¹⁰ The results are intuitive: When more democratic elections are included, elections are generally more likely to cause regime breakdown. Relying on a stricter definition of authoritarian elections, such elections are likely to stabilize dictatorships. Whereas Tables 1 and D reveal that when including all elections in the sample, elections are – for average levels of economic control – significantly associated with regime breakdown (the positive coefficient on elections), Table G shows no such effects for the sub-sample of strictly authoritarian elections. In fact, when elections have been held within the past three years, they significantly reduce the risk of breakdown (Table G, Models 3–4). Thus, H2a finds support for a certain type of authoritarian elections. Excluding cases such as the elections in 2000 in Ghana, Senegal, and Mexico where the authoritarian incumbent peacefully handed over power following electoral defeat, elections make autocracies more stable if rulers dominate the economy.

Figure 3.

Predicted probability of regime breakdown (peaceful incumbent exit excluded).

Conclusion

Researchers have shown diverging effects of multi-party elections on authoritarian regime stability: such elections seem to both sustain and undermine authoritarian rule. The divergent empirical findings have led some scholars to suggest that authoritarian elections have no substantial effects on regime stability. But analyzing the capacities available to autocrats may allow us to better understand the potentially stabilizing effect of elections under authoritarian rule. I have argued that extensive economic control through natural resources, business regulation, or a large public sector allows an autocrat to subtly manipulate elections. Voter manipulation and economic harassment prevent authoritarian regimes from breaking down following elections. Thus, the regime-toppling elections we have witnessed across the globe may have occurred under dictators endowed with limited economic capacities.

The claim finds support in cross-national analyses. Economic control significantly conditions the effect of elections. At low levels of economic control, holding elections increases the risk of regime breakdown in the following years (H2b) but as control over the economy increases, elections become less likely to topple a dictatorship. This may be part of the explanation for why dictators lost their grip on elections – and eventually power – in electoral revolutions in the Philippines, Serbia, and Georgia, while other autocrats, such as Zimbabwe’s Robert Mugabe, have managed to cling on. Thus, Greene’s argument (2007) that the loss of monopolistic control over the economy in 20th century Mexico explains the PRI-regime’s downfall finds cross-national support. The results are robust to control for state capacity, both administrative and coercive, which is also likely to condition the effect of elections.

The conditioning effect of economic control is largely driven by government spending rather than by income from natural resources. This finding resonates with the studies analyzing the tactics of the PRI in Mexico that delivered patronage through the formal state apparatus (i.e. Greene, 2007; Magaloni, 2006). But the effect of government spending on electoral authoritarian stability does not capture the tactics of incumbents who rely on natural resources that are appropriated outside of the official state apparatus and deliver and withdraw goods from supporters and opponents outside of the formal system. Such dynamics are extremely hard to gauge, particularly in cross-national studies. The Mugabe regime in Zimbabwe may prevail over a collapsing formal economy but the access to diamond wealth (the size of which is hard to document through official statistics), as well as the arbitrary control over land – a valuable resource in rural Zimbabwe – has proven to be an important patronage resource during election campaigns. A blossoming literature on the importance of land rights as a source of political patronage promises to advance this field further and can contribute to our knowledge of the various tools autocrats use to hold on to power via elections (Albertus and Menaldo, 2012a; Boone, 2014). This study underscores the importance of finding ways to document such informal tools of authoritarian control in cross-national analyses.

H2a, stating that for high levels of economic control, elections may in fact work to stabilize elections, found support for a particular type of elections: when restricting the sample to those multiparty elections in which the rulers did not peacefully give up power, electoral autocracies with high levels of economic control are indeed more stable than those without elections.

The findings have implications for the literature on electoral authoritarianism and authoritarian institutions. This literature has greatly advanced our understanding of the effects of authoritarian regime types (i.e. electoral versus non-electoral, party regimes versus military regimes and monarchies, etc.) (e.g. Brownlee, 2009; Miller, 2015; Geddes, 2003). Here, however, I find that the degree of rulers’ control of the economy matters to electoral dynamics. Thus, it is not elections per se that determine authoritarian regime dynamics – the effects of elections are shaped by the circumstances under which they are held. Furthermore, the conditioning effect of economic control is robust to control for authoritarian regime type. Thus, the type of ruling group (party, monarchy, military, personalist) is not the only – and perhaps not even the most important – difference between autocracies.

Building on existing knowledge about authoritarian regime types and institutions, it is time to look into factors such as economic and state capacities. Such capacities are commonly accepted as drivers of regime development in both democratizing regimes and established democracies (Andersen et al., 2014; Mansfield and Snyder, 2007). But although repressive capacity and resource income are generally known to inhibit democratization (Albertus and Menaldo, 2012b; Ross, 2012), the formal and informal ways in which autocrats control the state and the economic sphere are difficult to get at (see, however, Blaydes, 2011; Boone, 2014; Greene, 2007; Magaloni, 2006; McMann, 2006; Slater, 2010), particularly in cross-national studies. Unpacking such dimensions of authoritarian control, however, is crucial to understanding the dynamics of authoritarian rule in both electoral and non-electoral regimes.

Footnotes

Funding

The author received financial support for this research from Innovationsfonden (19789).

Notes

Author biography

Merete Bech Seeberg is assistant professor of political science at Aarhus University. Her work spans across the fields of democratization, dictatorship and political violence and she is currently employed with the CODE project ().

References

Albertus

Michael

Menaldo

Victor

(2012a) If You’re Against Them You’re With Us. Comparative Political Studies 45(8): 973–1003.

Albertus

Michael

Menaldo

Victor

(2012b) Coercive Capacity and the Prospects for Democratization. Comparative Politics 44(2): 151–169.

Andersen

David

Møller

Jørgen

Rørbæk

Lasse

Skaaning

Svend-Erik

(2014) State Capacity and Political Regime Stability. Democratization 21(7): 1305–1325.

Banks

Arthur

Wilson

Kenneth

(2012) Cross-National Time-Series Data Archive. Jerusalem: Databanks International. Available at: http://www.databanksinternational.com

Blaydes

Lisa

(2011) Elections and Distributive Politics in Mubarak’s Egypt. New York: Cambridge University Press.

Boone

Catherine

(2014) Property and Political Order in Africa. New York: Cambridge University Press.

Brownlee

Jason

(2007) Authoritarianism in an Age of Democratization. New York: Cambridge University Press.

Brownlee

Jason

(2009) Portents of Pluralism: How hybrid regimes affect democratic transitions. American Journal of Political Science 55(3): 515–532.

Bunce

Valerie

Wolchik

Sharon

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

10.

Camp

Roderic

(1990) Camarillas in Mexican Politics. Mexican Studies/Estudios Mexicanos 6(1): 85–107.

11.

Cheibub

Jose

Gandhi

Jennifer

Vreeland

James

(2010) Democracy and Dictatorship Revisited. Public Choice 143(1): 67–101.

12.

Crouch

Harold

(1996) Government and Society in Malaysia. Ithaca, NY: Cornell University Press.

13.

Donno

Daniela

(2013) Elections and Democratization in Authoritarian Regimes. American Journal of Political Science 57(3): 703–716.

14.

Fox

Jonathan

(1994) The Difficult Transition from Clientelism to Citizenship. World Politics 46(2): 151–184.

15.

Geddes

Barbara

(2003) Paradigms and Sand Castles. Ann Arbor, MI: University of Michigan Press.

16.

Geddes

Barbara

Wright

Joseph

Frantz

Erica

(2014) Autocratic Breakdown and Regime Transitions: A new data set. Perspectives on Politics 12(2): 313–331.

17.

Greene

Kenneth

(2007) Why Dominant Parties Lose. Cambridge: Cambridge University Press.

18.

Greene

Kenneth

(2010) The Political Economy of Authoritarian Single-Party Dominance. Comparative Political Studies 43(7): 807–834.

19.

Gwartney

James

Lawson

Robert

Block

Walter

(1996) Economic Freedom of the World, 1975–1995. Vancouver: Fraser Institute.

20.

Haber

Stephen

Menaldo

Victor

(2011) Do Natural Resources Fuel Authoritarianism? American Political Science Review 105(1): 1–26.

21.

Hanmer

Michael

Kalkan

Kerem

(2013) Behind the Curve: Clarifying the Best Approach to Calculating Predicted Probabilities and Marginal Effects from Limited Dependent Variable Models. American Journal of Political Science 57(1): 263–277.

22.

Howard

Marc

Roessler

Philip

(2006) Liberalizing Electoral Outcomes in Competitive Authoritarian Regimes. American Journal of Political Science 50(2): 365–381.

23.

Hyde

Susan

Marinov

Nikolay

(2012) National Elections Across Democracy and Autocracy (NELDA). Available at: http://www.nelda.co/

24.

International Crisis Group (2013) Zimbabwe’s Elections: Mugabe’s last stand. 95. Africa Briefing. Johannesburg/Brussels. Available at: https://www.crisisgroup.org/africa/southern-africa/zimbabwe

25.

Kriger

Norma

(2005) ZANU(PF) Strategies in General Elections, 1980–2000: Discourse and coercion. African Affairs 104(414): 1–34.

26.

Kugler

Jacek

Tammen

Ronald

(2012) The Performance of Nations. Lanham: Rowman & Littlefield Publishers.

27.

Kuntz

Philipp

Thompson

Mark

(2009) More than Just the Final Straw. Comparative Politics 41(3): 253–272.

28.

Levitsky

Steven

Way

Lucan

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

29.

Lindberg

Staffan

(2006) Democracy and Elections in Africa. Baltimore, MD: The John Hopkins University Press.

30.

Lust

Ellen

(2009) Competitive Clientelism in the Middle East. Journal of Democracy 20(3): 122–135.

31.

Magaloni

Beatriz

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

32.

Malesky

Edmund

Schuler

Paul

(2010) Nodding or Needling: Analyzing Delegate Responsiveness in an Authoritarian Parliament. American Political Science Review 104(3): 482–502.

33.

Mansfield

Edward

Snyder

Jack

(2007) The Sequencing Fallacy. Journal of Democracy 18(3): 5–10.

34.

Marshall

Monty

Jaggers

Keith

Gurr

Ted

(2011) Polity IV Project. Vienna: Center for Systemic Peace. Available at: http://www.systemicpeace.org/polity/polity4.htm

35.

McMann

Kelly

(2006) Economic Autonomy and Democracy. Cambridge and New York: Cambridge University Press.

36.

Miller

Michael

(2015) Democratic Pieces: Autocratic elections and democratic development since 1815. British Journal of Political Science 45(3): 501–530.

37.

Penn World Table (2013) Penn World Table. Available at: http://cid.econ.ucdavis.edu/pwt.html

38.

Reuter

Ora John

Gandhi

Jennifer

(2011) Economic Performance and Elite Defection from Hegemonic Parties. British Journal of Political Science 41(1): 83–110.

39.

Ross

Michael

(2012) The Oil Curse: How Petroleum Wealth Shapes the Development of Nations. Princeton: Princeton University Press.

40.

Schedler

Andreas

(2006) The Logic of Electoral Authoritarianism. In Schedler

Andreas

(ed.) Electoral Authoritarianism. Boulder, CO: Lynne Rienner Publishers, 1–23.

41.

Seeberg

Merete Bech

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

42.

Silitski

Vitali

(2005) Preempting Democracy: The case of Belarus. Journal of Democracy 16(4): 83–97.

43.

Singer

David

Bremer

Stuart

Stuckey

John

Russett

Bruce

(1972) Capability Distribution, Uncertainty, and Major Power War, 1820–1965. Peace, War, and Numbers. Beverly Hills, CA: SAGE Publications.

44.

Slater

Dan

(2010) Ordering Power. New York: Cambridge University Press.

45.

Whitten-Woodring

Jenifer

Van Belle

Douglas

(2014) Historical Guide to Media Freedom. Los Angeles, CA: CQ Press.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

0.74 MB