Making Sense of a Surprise: Perspectives on the 2020 “Belarusian Revolution”

Abstract Drawing on three theoretical perspectives—“protest-democracy,” “authoritarian/patronal regime dynamics,” and “contentious politics”—developed in the study of popular protests in post-Soviet electoral autocracies, this article argues, first, that the 2020 postelection mobilization in Belarus was not to be expected for both structural and agency-related reasons. Second, by the summer of 2020, the political opportunity structure had opened up because of contingent choices by individual actors, with Alyaksandr Lukashenka committing several major mistakes, particularly on pandemic (non)control and the administration of the upcoming presidential election, and political newcomers taking on the role of challenging him. After the election, mass mobilization unfolded in two waves triggered by two additional regime mistakes: blatant electoral fraud and excessive repression. These mistakes served as focal points for spontaneous coordination, substituting for the deliberate “engineering” of protest by an organized opposition typical of the post-Soviet color revolutions. Third, Lukashenka survived in office because popular protests did not lead to elite defection. Instead, he was able to secure the loyalty of elites because he avoided gross blunders against both regime insiders and Russia. The case of Belarus indicates that hegemonic-authoritarian regimes are more vulnerable to mistakes by incumbents than to challenges from below and outside the regime.

increasing temporal distance, it becomes possible to move from mere ad hoc explanations to interpretations that use existing theoretical insights to better understand the causes, process, and outcome of this case. This article contributes to that endeavor by drawing on three theoretical perspectives that emerged in the study of the color revolutions and were further developed in light of subsequent protest events in the region. This is the first attempt to bring together the fragmented debate to explore the explanatory potential of these strands, two of which-protest-democracy and authoritarian/patronal regime dynamics-offer rival explanations, one focusing on agency the other on structure, while both get along well with the contentious politics approach, because the latter stresses micro-level explanations of protest. More specifically, the article seeks answers to three questions: Why did mass mobilization emerge in the first place? Why did the regime's early and massive repression provoke an even stronger mobilization instead of ending the protest? Why did the Lukashenka regime survive?
The article proceeds as follows: In the first section, I outline three paradigmatic tendencies that emerged from comparative studies of mass protests in nondemocratic settings in general and in post-Soviet electoral autocracies in particular. Based on congruence analysis, I then show that, in the spring of 2020, there were no conditions that indicated that the upcoming presidential election would become a major challenge for the regime. The next two sections use process tracing to show that the sudden opening up of the political opportunity structure was the result of a contingent chain of decisions by individual actors interacting with each other, whereas the subsequent mass mobilization was triggered by two specific occurrences: electoral fraud served as the focal point for the first outburst of collective action, and excessive repression by the regime for the second. While these dynamics are best explained by mechanisms associated with the toolkit of the contentious politics perspective, the survival of Alyaksandr Lukashenka's regime, which I discuss in the penultimate section, was the result of his ability to prevent elite defection and restore linkages to Russia, as emphasized by theories within the regime dynamics perspective. In the conclusion, I summarize my findings and discuss their implications for further research.

Three Perspectives in the Study of Protest in Nondemocratic Regimes
The scholarly debate on mass protests in electoral autocracies has produced a fairly extensive literature. Originating in comparative research on the "peaceful revolutions" in Eastern Europe in [1989][1990], theory building was driven by the election-triggered "color revolutions" that took place more than a decade later in several countries in southeastern Europe and post-Soviet Eurasia, where the transition to democracy stalled in the mid-1990s. The best-known examples of what have been conceptualized more generically as "electoral revolutions" (Bunce and Wolchik 2006) are the mass protests in Serbia (2000), Georgia (2003), and Ukraine (2004), but the list of cases that did not result in regime change is much longer. 1 Even after the wave of color revolutions subsided, postelection protests continued to occur in the post-Soviet space, for example in Russia in -2012and in Belarus, Georgia, and Kyrgyzstan in 2020 The background condition for an electoral revolution is that it takes place in an "electoral" or "competitive" autocracy and under a regime that features formally democratic institutions, tolerates pockets of pluralism and opposition, but falls short of democratic standards (Levitsky and Way 2010;Lührmann et al. 2018;Schedler 2006). While the chief executive and the legislature are regularly elected under universal suffrage and with a modicum of competition, electoral contests are systematically subject to manipulation so that incumbents de facto evade electoral accountability. This manipulation includes not only the falsification of results, but also medium-and long-term strategies such as designing party and electoral laws to favor regime parties and candidates, controlling the media, excluding opposition candidates from entering the electoral arena, buying votes, blackmailing voters, and the like. In most cases, these strategies are effective. In fact, empirical analysis shows that whereas elections bolster the repressive and co-optive capacities of autocrats in the long run, thereby promoting their persistence, in the short run, especially just before or after their occurrence, elections also carry a considerable risk of autocrats' being ousted from office (Knutsen, Nygård, and Wig 2017).
While there is no general theory of electoral revolutions to draw on, analysis of the Belarusian "revolution without a name" (Wilson 2021) can benefit from valuable theoretical insights and models that have grown out of the lively debate on the color revolutions and subsequent protest events in post-Soviet space. Three emerging perspectives, or "paradigmatic tendencies," can be distinguished (Hale 2019).
The first of these, the protest-democracy perspective, links the concept of large-scale mobilization to the idea of democratization, identifying any challenge to a nondemocratic regime from below or from regime outsiders as a potential "democratic breakthrough" (e.g., Wolchik 2011, 2018;McFaul 2010). Within this perspective, the mainstream considers the existence of a unified, effective opposition that implements a sophisticated "electoral model" to be the most important factor for regime change and subsequent democratization. Valerie Bunce and Sharon Wolchik (2011) derive their findings from a systematic comparison of six successful and five failed color revolutions between 1998 and 2008, discovering a range of innovative strategies that were developed with the support of transnational democracy assistance networks. During an energizing nationwide electoral campaign, the opposition forged close ties with civil society groups, including youth organizations; used unconventional means to strengthen their appeal to voters and undermine the regime's popularity; devised their own election monitoring campaigns; and prepared for nonviolent postelection mobilization. When blatantly fraudulent election results were announced on election day, the opposition called voters into the streets to demand the annulment of the poll and the holding of free and fair elections (Bunce and Wolchik 2011;cf. McFaul 2010). From this point of view, color revolutions such as the Orange Revolution in Ukraine in 2004 were "engineered by the opposition at a moment anticipated well beforehand, albeit under structurally conducive conditions" (Beissinger 2022, 132). Thanks to demonstration effects and transnational activists who passed on their revolutionary know-how to local actors, the electoral model successfully spread throughout the region until authoritarian rulers learned the critical lessons, preventing or repressing further emulation (Beissinger 2007). The debate on cross-national diffusion and its role is ongoing and has since been extended to cases of "democracy protest" around the world.
The second paradigmatic tendency, regime dynamics, deems the "democracy" aspects of protest in nondemocratic countries epiphenomenal and less relevant than the deeper processes of regime reproduction, putting the emphasis on structural conditions (Hale 2019(Hale : 2405. For Lucan Way (2008Way ( , 2009Way ( , 2015Levitsky and Way 2010), the color revolutions are cases of authoritarian failure rather than democratization. Way argues that the stability of an authoritarian regime depends not on electoral or diffusion dynamics but on two other factors: the strength of the regime's autocratic party or state-that is, the degree of state capacity and the presence of polarized splits within the titular nation's identity-on the one hand, and the strength of the country's ties to the West, its "linkage," on the other.
The most comprehensive explanation of the outcomes of mass protests within this perspective is offered by the theory of "patronal politics" (Hale 2006(Hale , 2015, 2 which addresses regime contexts more commonly referred to as "(neo)patrimonialism," "clientelism," or "low social capital." In patronal regimes, the coordination of individuals occurs through extended networks based on personal acquaintance, rather than impersonal formal organizations. According to Henry Hale, it is the coordination dynamics of intra-elite political-economic networks that drive regime change, collapse, and reproduction. When a country's most powerful networks are tightly integrated into a "single-pyramid system" centered on a "chief patron" who occupies unchallenged the highest position both on the informal pyramid and in the formal state (i.e., the presidency), the regime appears to be a consolidated autocracy. By contrast, when the elite splits and the power pyramid disintegrates, the regime may collapse. In such regimes, elections are important because they signal the incumbent's actual power, thus generating incentives for network coordination. They may trigger elite defection, when the president/chief patron appears to be a "lame duck" because of illness, defeat in a war, economic crisis, or the impending end of their time in office (Hale 2015, ch. 4). Therefore, popular mobilization after a rigged election will only play a critical role for regime survival if it is linked to the dynamics of intra-elite coordination, and even a successful electoral revolution will not necessarily result in a democratic breakthrough (Hale 2006(Hale , 2019. The third paradigm in the study of mass mobilization in the post-Soviet space draws on research on social movements, revolutions, and other forms of collective action in Western democracies, particularly the general research program of contentious politics (e.g., McAdam, Tarrow, and Tilly 2001;McAdam and Tarrow 2010). This broad and heterogeneous framework does not represent a theoretical alternative to the two paradigms just mentioned with respect to the causes of electoral revolutions or their outcomes. Instead, it addresses "recurring mechanisms" and their combinations in processes of mobilization and demobilization. Similar to research within the protestdemocracy framework, studies within this paradigm inquire about mobilization processes from below and usually (or implicitly) view them as drivers of democratization. However, their focus is on issues such as solutions to collective action problems, the role of micro-level mobilizational structures (e.g., families or networks of friends), the formation of collective identities, the effects of repression on protest behavior, the role of social media, and so on. In post-Soviet studies, this research program has expanded since the color revolutions (e.g., Beissinger 2007;Tucker 2007), began to flourish with the "For Free Elections" movement in Russia in -2012(e.g., Rosenberg 2017Smyth 2021) and the Euromaidan in 2013-2014 (e.g., Onuch and Sasse 2016;Pop-Eleches, Robertson, and Rosenfeld 2022), and currently predominates in research on the 2020 events in Belarus.
In what follows, I refer to all three perspectives, each of which contributes in its own way to our understanding of the 2020 protests in Belarus. Theories within the protest-democracy and regime dynamics paradigms allow us to derive a catalog of conducive or even "necessary" (McFaul 2010) conditions for postelection mass mobilization. Since these were largely absent, the focus will then switch to contingent choices of individual actors and the dynamics of large-scale mobilization. Last but not least, the regime dynamics perspective offers satisfactory explanations for Lukashenka's survival in office.
Regime and Opposition on the Eve of the 2020 Presidential Election From a structuralist perspective, the question is whether Lukashenka's regime-an electoral autocracy (Coppedge et al. 2021) that gradually evolved from a competitive into a hegemonic one (Levitsky and Way 2020)-was ready to fall as the 2020 electoral season approached. In retrospect and referring to theories of revolution, Arkady Moshes and Ryhor Nizhnikau (2021) assert that the preconditions for a revolutionary situation were already in place by the spring of 2020. According to these authors, during the previous decade or so, societal modernization had gripped large segments of the population, leading to rising demands for democracy and participation. As a consequence, the paternalist model of state-society relations lost support, especially as the regime's ability to maintain the social contract with the population deteriorated. This argument is corroborated by several studies diagnosing the increasing exhaustion of the neo-Soviet "Belarusian model," the stagnation of economic development and wages, the regime's eroding capacity to deliver social services, and shifting attitudes among the population (e.g., Grishchenko and Titarenko 2020;Krawatzek and Langbein 2022;Moshes and Nizhnikau 2019;Papko and Kozarzewski 2020).

Low Regime Vulnerability
It is striking, however, that even shortly before the 2020 election, researchers did not consider strong popular pressure from below, let alone a revolution against the incumbent, when discussing scenarios of the country's future development. On the contrary, after the protest rallies against the so-called parasite tax in the spring of 2017, Moshes (2017, 1) assumed that "the majority of the population and the regime may actually be united in their wish to resist change and preserve the status quo." A chaotic transfer of power with heightened intra-elite conflict and interference by Russia or a gradual "softening" of the regime from above, initiated by Lukashenka himself, were thought to be the more realistic options for the future (Moshes and Nizhnikau 2019, 8;Shraibman 2018, 23-28).
In fact, as of early 2020, Lukashenka's regime remained an intact police state that had marginalized any meaningful opposition . After Lukashenka won the first-and so far only-relatively free and fair election in 1994, he undertook the imposition of a strong presidency, the strengthening of the state, including its inherited repressive apparatus, and the rebuilding of a centralized economy (Way 2008;2015, ch. 5). By the late 1990s, years before the color revolutions, he had already invested in strategies to prevent the emergence of political challengers from below and outside the regime. His effective "preemptive authoritarianism" (Silitski 2009) served as a model for other autocrats in the region, when the color revolutions brought "revolutionary contagion" to neighboring countries. Lukashenka restricted the activities of independent media and civil society, attacked Western democracy assistance, hampered independent election monitoring, and intimidated potential protesters with propaganda and legal harassment. For Vitali Silitski (2009, 280), this policy was the main reason why political change in Belarus remained unlikely in the medium or even long term.
Moreover, Lukashenka's regime proved highly adaptive for over a quarter of a century (Frear 2019). It dealt particularly harshly with the opposition after each presidential election (Ash 2015). In particular, the 2010 crackdown led to fragmentation and a near-total failure of coordination within the opposition, the demobilization of socioeconomic protest activities, and the increasing politicization of the scarce protest events in subsequent years (de Vogel 2022). On the other hand, the regime had softened its initially hostile attitude toward the emerging civil society over the last couple of years, beginning to tolerate and even encourage "constructive" nonpolitical activities. In the second half of that decade, therefore, the continued suppression of opposition was accompanied by the selective co-optation of more "innocuous" segments of civil society, such as initiatives promoting the Belarusian language and heritage in the arts, education, and sports, as part of the regime's "soft Belarusianisation" (Astapova et al. 2022;Marin 2020).
Finally, international pressure on the regime-even attention to the upcoming election-was low after the Euromaidan led to a further deterioration of relations between Russia and the West, and the European Union (EU) prioritized stability in the region over its democratization. After the 2015 presidential election, the EU lifted most sanctions against Belarus, playing an active role as a mediator in settling the 2014-2015 Russo-Ukrainian war in Donbass. Thus, Lukashenka was able to continue to refine his long-standing pattern of oscillating between East and West. On the one hand, Belarus was part of a Union State with Russia and had developed a high economic dependence on it, with Russia being the country's main trading partner and a provider of generous energy subsidies as well as preferential loans (Leukavets 2021). On the other hand, Russia's annexation of Crimea in 2014 was perceived by Lukashenka as growing security threat. He responded by prioritizing the security of the state over the security of the regime, taking specific steps to strengthen Belarus's sovereignty and territorial integrity rather than-as in the first decade after the color revolutionsinitiating counterdemocracy measures and power entrenchment (Ambrosio 2022).

A Consolidated Power Pyramid without an Imminent Succession Crisis
From the perspective of patronal politics, Lukashenka's ability to effectively prevent elite defection seems even more important than the regime's intact repressive capacity-or, rather, elite cohesion was the precondition for that capacity. As a political outsider and initially without support inside the state, it took him several years to crush rival groups and to consolidate his rule, but by the mid-2000s, all relevant elite groups had ultimately coordinated around him as president (Hale 2015, 258-266).
The core of Lukashenka's single-pyramid system consisted of the political machine of his region of origin and two groups within the state: the central bureaucracy and the siloviki, the security apparatus. Unlike most of his post-Soviet colleagues, Lukashenka did not rely on an institutionalized party but balanced the networks within the pyramid by routinely shifting preferential treatment. His staffing decisions were designed to prevent ambitious politicians from emerging and positioning themselves as feasible alternatives to his leadership. Thus, he refrained from appointing prominent figures to high political positions, pursued a policy of "cadre rotation," and initiated criminal trials against officials from time to time, mostly for corruption (Frear 2019, ch. 4;Shraibman 2018). On the eve of the 2020 election, Lukashenka appointed three representatives of the military and security services as head of the presidential administration (December 2019), chief adviser during the election campaign, and prime minister after a major cabinet reshuffle (2020), ousting so-called liberals (Shraibman 2020).
In addition to tight control over the elites, maintaining a high level of popularity among the population and avoiding the image of a lame duck were key conditions for the longevity of Lukashenka's single-pyramid system (Hale 2015, 258). In 2020, after 26 years in office, he was still, at 65, younger than most of his counterparts in the other hegemonic-authoritarian regimes in the post-Soviet sphere, and, like them, he had long since ensured that the original two-term limit on his tenure had been abolished by constitutional amendment (2004). As in 2006 and 2010, the upcoming presidential election was therefore not accompanied by expectations of succession and power transfer, and consequently not by cracks within the regime elites and heightened competition. This was a critical difference from the preelection situations in all countries where color revolutions had been victorious in the 2000s, because there, elections coincided with succession crises and upcoming changes in the presidential office.

A Marginalized and Fragmented Opposition
In contrast with structuralist approaches, the actor-focused protest-democracy mainstream of the literature on color revolutions puts the emphasis on the implementation of the electoral model as a recipe for a "democratic breakthrough" from below. Its most important prerequisite-and the "most difficult piece of the electoral model to put into place" (Bunce and Wolchik 2011, 253)-is the forging of a well-organized, united opposition collaborating with civil society organizations. Comparison of all the color revolutions shows that the capacity for cooperation among the fragmented anti-regime opposition developed everywhere over a lengthy period of time. Out of prior experiences in local and national elections, which sometimes played the role of "rehearsals," the realization gradually grew that it was necessary to agree on a single candidate, the ability to identify this "best hope" being based on data provided by pollsters. Furthermore, pressure from youth organizations and transnational democracy assistance networks also contributed to the opposition's decision to cooperate and pool resources, candidates, and campaigns (Bunce and Wolchik 2011, 252-255).
No such united opposition had emerged in Belarus in the quarter century since independence. On the one hand, unlike in Ukraine, for example, the opposition faced daunting structural conditions from the 1990s because of a weak sense of national identity and a predominantly Russophile population, which led to the marginalization of nationalist challengers (Way 2015, ch. 5). On the other hand, not only the crackdown on anti-regime candidates after the 2006 and 2010 presidential elections, but also the regime's long-term preemptive strategies, including the demobilization and depoliticization of civil society groups, were major obstacles to the organized opposition's rallying around a single candidate. When presidential elections approached, the radical wing of the opposition usually called for an election boycott, while the moderate wing made several attempts to agree on a single candidate. Yet, the only time they (almost) managed to do that was in 2001 (Ash 2015;Marples 2006). Lukashenka's competitors, who together never received more than 9 to 18 percent of the vote, included not only genuine, moderate opposition candidates but also some handpicked "fake sparring partners," and "blurred" candidates whom everyone talked about though no-one knew whether they were supported by the opposition, the regime, or Russia (Wilson 2021, chs. 10-13). By the spring of 2020, little had changed. Five opposition parties held primaries to nominate a joint candidate, while four other parties refused even to participate in the process, which was aborted in mid-May, reportedly because of the COVID-19 pandemic. A few days later, candidates from three parties registered separately with the Central Election Commission (Bykovskii 2020;Mishchenko 2020).
To summarize, none of the perspectives on previous postelection mobilization in the post-Soviet sphere provided any theoretical grounds for anticipating a highly competitive election or even large-scale protests in Belarus in 2020. While structuralist arguments do not preclude spontaneous mobilization from below against rigged elections, they implied a low likelihood that Lukashenka would face serious challengers in the contest for votes or that he would even need to engage in largescale electoral fraud. From the agency-focused perspective of the electoral model, an electoral revolution was almost completely ruled out, since there was no organized opposition capable of running a sophisticated campaign, organizing election monitoring, and mobilizing voters first at the polls and then in the streets. In no small part, this appeared to be a long-term consequence of the color revolutions in the 2000s, or rather of Lukashenka's ability to improve the effectiveness of his "preemptive authoritarianism" by drawing lessons from them and neutralizing Western democracy assistance, which diminished significantly after 2010 (Pikulik and Bedford 2019).

An Unpredictable Opening of the Political Opportunity Structure
For most of his presidency, Lukashenka had enjoyed solid popularity across the population, and scholars had long since become accustomed to the apparent paradox that he managed to combine oppressive policies toward the opposition with broad popular support (White 2011, 800). Experts held that Lukashenka could regularly count on about 50 percent "real" support and that the opposition's potential electorate was hardly larger than 30 percent (Dorokhov 2020). According to independent surveys (IISEPS 2016), between 2005 and 2016, some 44 to 60 percent of citizens trusted the president most of the time, interrupted by brief lows in the fall of 2011 (24.5 percent), late 2013, and mid-2016 (about 38 percent). By the end of 2019, however, only an estimated 26 percent of the population retained their trust (O'Loughlin and Toal 2022, 51). When the COVID-19 pandemic hit the country in the spring of 2020, Lukaschenka's approval ratings deteriorated further as he underestimated the seriousness of the disease, downplayed it, and ridiculed the victims instead of taking the political measures adopted in most European countries. In this situation, civil society stepped in and filled the void by disseminating information about the virus, mobilizing support for their fellow citizens, and providing emergency assistance to health workers (Astapova et al. 2022;Marples 2021;Moshes and Nizhnikau 2021).
From the protest-democracy perspective, the incumbent's low popularity was a necessary proximate cause of the color revolutions (McFaul 2010), while in the theory of patronal politics, it is a possible driver of elite splits threatening regime survival (Hale 2015). Yet, this does not necessarily mean that Lukashenka was on the verge of electoral defeat. First, even experts were divided over whether his ratings were actually as low as wishful thinking suggested (Dorokhov 2020;Ioffe 2020). Second, while mobilization from below prompted by COVID-19 boosted the selfconfidence and self-empowerment of "ordinary" citizens, the pattern of apolitical authoritarian co-optation remained intact. In fact, the joint effort marked the "apex of civil society's 'constructive' collective action" (Astapova et al. 2022, 20). Third, and most importantly, there was no suitable challenger to the incumbent to translate popular discontent into political action at the ballot box. Lukashenka continued to rely on his well-proven strategies for maintaining an uneven playing field, as in all previous presidential elections. However, this time he made some important mistakes, which almost proved fatal because they were matched by unpredictable decisions by other individual actors.
To understand the significance of these mistakes, it is helpful to refer to the typology developed by Daniel Treisman (2020) from a comparison of all democratizations since 1800. He found three main types of missteps by incumbents that triggered previously latent factors for authoritarian breakdown. Authoritarian rulers may take "non-optimal" actions with respect to domestic regime outsiders (i.e., the population and the opposition), regime insiders (i.e., the elites and their networks), and international actors. Lukashenka's blunders all belong in the first category and can be further differentiated.
First, his COVID-19 "nonpolicy" was an example of a "major domestic policy failure" that negatively affected the incumbent's popularity. Second, he mishandled elections: he could have used the pandemic as a pretext to postpone the vote so as to sit out his drop in popularityinstead, he moved election day forward three weeks in the hope that the vacation season would shorten the campaign and lower voter turnout. Likewise, his decisions on candidate registration turned into serious mistakes, as he did not anticipate other actors' reactions. Among the 15 candidates registered by the Central Election Commission on May 15 were one or two "fake candidates," three challengers from the "old" nationalist opposition, and two from the moderate opposition-none of them posing any threat to the regimebut also three political newcomers without any political affiliation (Wilson 2021, 280-281). They had already gained some public attention and were beyond the regime's control, so their removal from the ballot seemed entirely reasonable at first. Popular vlogger Syarhei Tsikhanouski appealed to large segments of Lukashenka's core constituency-which made him a real competitor. In early May, he was arrested on a pretext so that he could not sign the application to collect signatures for candidate registration. At the last minute, Tsikhanouski's wife Svyatlana registered in his place, a move that she intended not as a political signal to the public but as a gesture of personal support for her husband (Tsikhanouskaya 2020).
From late May to mid-June, several polls surfaced on social media that put Lukashenka's approval rating at just 1.0 percent and 6.2 percent, respectively. These figures were obviously unrealistic (Drakochrust 2020), but they created the impression among the internet-savvy, mostly urban parts of the electorate that the regime had become vulnerable. The regime responded by taking the other two challengers out of the game. On June 17, Viktar Babaryka, the former CEO of Belgazprombank, was arrested, and one month later, Valeryj Tsapkala, a former adviser to Lukashenka, ex-ambassador, and founder of a major Minsk IT hub, fled to Russia. Both advocated moderate reforms to modernize the economy and society, appealing to relatively broad sections of the population, but especially to educated and urban middle-class milieus. Moreover, the former loyalists, Babaryka in particular, were known for their close ties to the Putin regime. Consequently, the president may even have feared that Russia was looking for an opportunity to get rid of him, especially since relations between the two countries were strained (Marples 2021, 287;Wilson 2021, 282). From the perspective of patronal politics, this would indicate that the ruling elite was beginning to show signs of disintegration.
With none of the three opposition newcomers admitted to the ballot and another eight candidates failing to provide a sufficient number of signatures or dropping out of the campaign, the final list of contenders that passed the Election Commission on July 14 contained only four challengers to the incumbent. None of them was considered a strong rival. Lukashenka asserted that the voters would not support Tsikhanouskaya because the Belarusian constitution "makes it difficult even for a man to carry the burden" of the strong presidency, and scoffed, "if you put this on a woman, she will collapse, the poor thing" (Lukashenka 2020). So, his misogynous underestimation of an opponent and his decision to withdraw all three newcomers from the race, rather than exploit the fact that they appealed to different segments of the electorate, came together.
Ultimately, Lukashenka's moves turned into serious mistakes because of the decisions of his challengers. First, Tsikhanouskaya chose to accept her role as the opposition's only candidate, developing unexpected charismatic appeal. She was inexperienced but, by the same token, also untainted by the past, coming directly from the grassroots and representing a clear moral cause. Second, on July 17-just three weeks before election day-her campaign united with the electoral teams of Babaryka and Tsapkala. That, too, was anything but inevitable. It certainly helped that by chance all three electoral teams were chaired by women. In an act of female solidarity, they agreed on a pragmatic goal: Tsikhanouskaya was to become interim president in order to organize free and fair elections. It is striking that she thus apparently confirmed Lukashenka's view of the "burden of the presidency," but no doubt this also increased her acceptance. The three "girls," as Lukashenka called them, reframed the election campaign, introducing new symbols and performative elements that were clear and strong, but less confrontational than Syarhei Tsikhanouski's original campaign. With their signature gestures-fist, heart, and victory sign-and their message "Believe! Can! Win!," the three women acted less as organizers of voter mobilization than as direct representatives of the people.
All in all, this was a very peculiar path to the formation of a united opposition, remarkably different from the pattern of the color revolutions, where well-organized elite networks led by former regime insiders had been preparing for years for the electoral road to power. Nowhere before had a promising candidate emerged from "accidental and chaotic processes and the impact of a handful of people," prompted by gross blunders committed by the incumbent. Previously, this had always been the result of "planning, coordination, and hard work," as well as-in the case of the color revolutions-international support (Bunce and Wolchik 2011, 253, 277).
Lukashenka committed the next grave miscalculation when, on August 9, he had the Election Commission announce results that were blatantly fraudulent. While Tsikhanouskaya's election staff claimed about 80 percent of the vote for her, conservative estimates suggest that a runoff would have been necessary because neither candidate received more than 50 percent of the vote (Golos 2020). Only at this point could it have been anticipated that an "electoral revolution" was imminent.

The Micro-Level Logic of Protest Mobilization
The "Belarusian revolution" unfolded in five phases. The first, from early May to July 17, 2020, brought about a unified opposition, while the second involved voter mobilization throughout the election campaign and continuing until election day. The third phase lasted from August 9 to 11 and was characterized by the outbreak of spontaneous mass protest that was violently suppressed by the regime. After that, a new and even stronger mobilization ensued, to which the regime responded with mass arrests but less brutality. This fourth phase continued until about late October. After that, demonstrations petered out. While everyday life returned to the provinces, decentralized actions continued in some Minsk neighborhoods for months, organized with "partisan tactics" so that they were difficult to control by the security forces. I argue that the two waves of mobilization in the third and fourth phases were triggered by different mechanisms.

Electoral Fraud as a Focal Point for Solving the Problem of Collective Action
Any postelection mass protest in electoral autocracies is in need of explanation because of the well-known problem of collective action, a central topic within social movement research and the study of contentious politics. What has been called the "paradox of revolution" (Tullock [1974] 2005) or the "rebel's dilemma" (Lichbach 1995) is the most critical obstacle for large-scale mobilization in rational-choice-informed research on collective action. Since social movements -or postelectoral anti-regime protests, for that matter-deliver collective goods when they are successful, rational individuals will typically prefer not to bear the costs of collective action due to the lack of individual incentives. If they believe that the costs of participation exceed the benefits they will receive regardless of their individual participation, that the likelihood of achieving the collective goal is low, or that their individual behavior will not be pivotal, they will prefer to "shirk." Obviously, this kind of reasoning is likely to discourage collective action even more strongly in repressive contexts, where the expected individual costs are significantly higher than in democracies.
However, protest events such as the color revolutions prove that collective action problems in electoral autocracies can be mitigated by major electoral fraud, because this-fairly typicalmistake on the part of authoritarian rulers may provide a "focal point" for coordination. Widely shared information about large-scale falsification of election results helps individuals decide to voice discontent because they can assume that other people will do the same. Therefore, the perceived likelihood of individual punishment decreases, while that of success increases. The efficacy of this mechanism is enhanced by the limited time frame (action is required before the self-proclaimed winner takes office), the clear message that the fraud conveys to citizens, and the international attention the incumbent's deception attracts (Tucker 2007).
Against the background of this theoretical insight, the protest mobilization of the first days does not seem particularly surprising because the official results were grotesquely distorted, and this was evident to everyone. But a glance at the color revolutions shows that the situation in Belarus was different. In the former cases, the uncovering of electoral fraud was considered "an occasion rather than a driver of mobilization" (Beissinger 2011, 36), and the detection of fraud was a component of the electoral model, deployed by an opposition that had thoroughly prepared for protest actions in the event that the incumbent did not concede defeat on election day. The strategy included the use of exit polls, parallel vote tabulation, and international monitoring reports, as well as establishing links with independent media and-as in the Orange Revolution in Ukraine-arrangements to provide food and shelter during protests (e.g., Beissinger 2011; Bunce and Wolchik 2011;McFaul 2010;Tucker 2007). In Belarus, by comparison, electoral fraud emerged much more unequivocally as a focal point for collective action in the strict sense-that is, as a driver of spontaneous, unorganized, large-scale mobilization as a result of a multitude of individual decisions.
The Belarusian opposition-unlike the regime challengers in the successful color revolutionshad made little use of independent election observers or exit polls in 2006 and 2010 (Bunce and Wolchik 2011, ch. 7), and even less in 2020. But now citizens took election monitoring into their own hands. Voting itself, normally an individual activity, turned into a collective action aimed at preventing or detecting the falsification of election results. To outsmart the authorities, who severely restricted access to polling stations by volunteer observers, many voters wore white wristbands or folded their ballots on the spot to publicly signal their support. More than half a million of the 5.8 million voters sent photos of their ballots to several internet platforms, which had only recently been launched by volunteers. They analyzed the official protocols of almost a quarter of the polling stations, matched them with the photos sent in, and published the results on social media channels (Golos 2020). In the Belarusian case, therefore, the actions of citizens who had politicized themselves during the election campaign substituted for the organized opposition considered crucial to the color revolutions in the respective literature.
The apparent overreach of the officially announced outcome of the ballot called tens of thousands onto the streets, although Tsikhanouskaya had not encouraged a protest; on August 11, she even released a video, apparently recorded under duress, in which she urged her sympathizers to stop contesting the results and not to resist the police. In terms of scale, this mobilization was comparable to those that took place after the 2006 and 2010 presidential elections. Lukashenka responded just as he had done then (Marples 2006, 363), overreacting with brutal police violence. During the first three days after the election, security forces arrested approximately 6,700 people, who were detained for days and often severely mistreated. Tsikhanouskaya and her team were taken into police custody or forced into exile. Lukashenka thus demonstrated that if the elections were a public signal of the incumbent's popularity, he was not inclined to draw any conclusions from it.

Repression and Mass Mobilization: The Role of Social-Identity Dynamics
Under these circumstances, and given the experience of the color revolutions, it was to be expected that the use of massive police force would put a swift end to the protests. In Serbia (2000), Georgia (2003), Ukraine (2004), andKyrgyzstan (2005), the opposition had been successful largely because the military and security services quickly disintegrated in the face of mass demonstrations (Bunce and Wolchik 2011). By contrast, postelection protests stopped after clashes with the police in cases such as Azerbaijan (2003,2005), Armenia (2008( ), and Belarus in 2006 This time, however, violence failed to achieve its goal. The crackdown on protesters turned out to be another Lukashenka mistake, this time of the "counterproductive violence" variety (Treisman 2020). Instead of abating, popular mobilization gained even more momentum. By the end of October, more than 100,000 people participated in demonstrations in Minsk on every Sunday, reaching as many as an estimated 250,000 to 300,000 at the peak. Meanwhile smaller protests were taking place throughout the country and on other days of the week. The demonstrators formed a temporary cross-class and cross-cleavage coalition, with women, often with their children, forming human chains in all major cities, and doctors, factory workers, pensioners, and athletes staging walkouts and strikes (Minakov 2021;Onuch and Sasse 2022).
This dynamic cannot be explained by the "focal point" model of electoral fraud. Rather, the mechanism that set the second protest wave in motion was "backlash mobilization," following an act of government repression (Sullivan and Davenport 2017). In the post-Soviet region, the most prominent example of such a dynamic was the Euromaidan in Ukraine in 2013-2014, where protest also unfolded in two waves. The first was triggered by President Viktor Yanukovych's announcement that he would not sign the Association Agreement with the European Union, but it was only the brutal repression of the Maidan activists ten days later that led to a large-scale explosion of protest (Beissinger 2022, 133;Onuch and Sasse 2016).
Research on contentious politics is well aware of the complex relationship between repression and mobilization protest (Beissinger 2022, 219-228;Davenport, Johnston, and Mueller 2005;Opp and Roehl 1990). Inquiries into the mechanisms of backlash mobilization, usually conducted in single-case studies, have produced various explanations for the reverse effect of repression. Thus, it has been theorized that the deterrent effect as a cost can be overcome: by social incentives, such as expectations that important others will not abstain, because repression is regarded as illegitimate; by moral incentives, such as protest norms and norms of violence; and by public goods incentives, that is, discontent with political institutions (Opp and Roehl 1990). A related explanation draws on the social-identity model in social movement research. This approach keeps to the premise that individual identity can be transformed by participation, stressing the socio-psychological dynamics of mass mobilization rather than rational-instrumental calculations (e.g., Drury and Reicher 1999;Klandermans 2014;Klandermans and Stekelenburg 2020). According to this view, changes in the external environment or some kind of triggering event may strengthen group identification, which enhances the likelihood of people participating in collective action on behalf of that group, which, in turn, further strengthens their identification with this group.
In Belarus, processes of collective identity formation had already been underway during the election campaign, when previously apolitical civil society experienced its "political reincarnation," reallocating its resources from the co-optation period for political purposes and quickly forming support groups for Tsikhanouskya (Astapova et al. 2022, 22-24). On the one hand, her team won sympathizers who identified with the candidate precisely because she was an unlikely one-a housewife instead of an established politician. On the other, the participants in the signaturecollection gatherings and preelection campaign rallies recognized themselves as citizens with a common goal. Among other things, this was expressed through the use of the white-red-white flag of the Republic of Belarus of 1918Belarus of and 1991Belarus of -1995, which soon became a symbol of the struggle for "justice," "truth," and a "country that is worth living in." During the fall of 2020, this politicization of a new, pluralist, and inclusive civic identity of a Belarusian "demos" that was different from the cultural and political patterns of the ethnocultural national movement of the 1990s became even more explicit and dynamic (Astapova et al. 2022;Gerasimov 2020).
It is likely that the harsh repression during the first days of protest drove many new citizens onto the streets because they felt part of a broader collectivity. Given the less adversarial policies toward civil society during the previous couple of years, hardly anyone had expected the level of violent resistance from the authorities during the first three days, which provoked strong feelings of anger and solidarity (Nikolayenko 2022). This is consistent with the findings of social movement research in democratic contexts, where police action has created a strong unified crowd out of a fragmented collectivity (Drury and Reicher 1999;Drury, Reicher, and Stott 2003). Empirically, the effect of violence on a decision to protest is corroborated by an online survey among urban Belarusian adults, according to which probably more than half of the protesters took to the streets only after August 11, meaning that they did not react immediately to electoral fraud. Moreover, as many as 98 percent of the respondents said their most important motive for participation was opposition to the unjustified use of force in the first three days after the election (Rudnik 2020).
As, for example, during the Euromaidan in Ukraine (MacDuffee, Metzger and Tucker 2017;Onuch 2015), social media platforms played a special role in protest mobilization. They facilitated the exchange of information about protest activities, but also the exchange of emotional and motivational content, and they had an important effect on the organization of protest as a result of their structural characteristics as well as their political opinions (Greene 2022;Wijermars and Lokot 2022). Compensating for temporary internet shutdowns in the first days, hundreds of thousands of people switched to VPN services and Telegram channels. The Warsaw-based channel nexta_life, whose number of subscribers grew from 300,000 to about two million within days, received about 200 messages per minute from all parts of the country in the first weeks of protest, updating its account of the course of events every five minutes. Based on this information, which was transmitted "horizontally" in the absence of protest leaders and organization, countless people not only learned what actions were taking place that they could engage with and where, but also developed similar perceptions of the situation and a belief that success was within reach, besides being presented with enhanced opportunities for identification with other participants (Gabowitsch 2021).
All in all, the protest dynamic that unfolded in fall 2020 can be adequately captured by drawing on the distinction between "top-down" and "bottom-up" identity formation. In the former, group identity is implied by the organizers encouraging participants to identify with the leadership ("dynamics of mobilization"), while in the latter participants identify with each other, and their group identity evolves inductively by their very interaction ("dynamics of participation"). Usually, both types of identity formation emerge at the same time (Klandermans and Stekelenburg 2020). In the Belarusian case, however, the second type clearly prevailed. In addition, the protesters were driven by expressive rather than instrumental motives. These lead people to see participation less as a means of achieving some goal than as a goal in itself, that of making known that they are upset about the current state of affairs (Klandermans and Stekelenburg 2020, 5-6). Evidence to support this interpretation is provided, for example, by the findings of Emma Mateo (2022) about the importance of preexisting local networks for mass mobilization and the observations of Mischa Gabowitsch (2021) on the style of the Belarusian protest and its specific "regimes of engagement and coordination." Marches as the paradigmatic protest form created a horizontal spatial structure in which all participants appeared as equals; they did not culminate in meetings with prominent speakers; the content and symbolic meaning as well as the materiality and presentation of slogans and protest symbols, the white-red-and-white flag among them, represented a civic community taking part in joint collective action rather than individuals striving for self-expression and a creative personalization of their protest participation.

Why Did Lukashenka's Regime Not Collapse?
For a few weeks in the early fall of 2020, it seemed not unlikely that Lukashenka's regime might be doomed. His mistaken decisions with regard to candidate registration, committing massive fraud, and unleashing excessive repression during the first three days after the election led to the emergence of a united opposition, followed by an unprecedented politicization of citizens and a large-scale, sustained mass mobilization. No less than 15 to 20 percent of the country's urban adults took part in at least one demonstration, putting this episode on a par with, for example, Ukraine's Orange Revolution in 2004 and the Euromaidan in 2013-2014 (Beissinger 2013;Gerasimov 2020;Minakov 2021). Yet Lukashenka did not resign, nor were new elections announced. The only promise he made, vaguely and under pressure from Russia, was that of constitutional reform. On February 27, 2022, in the shadow of the Russian invasion of Ukraine and under repressive conditions, constitutional amendments were finally approved in a referendum in which, according to official figures, 78.6 percent of the country's eligible voters participated, 82.9 percent of whom supported the reform. The most important changes involved the abolition of Belarus's non-nuclear and neutral status, paving the way for an even closer alliance with Russia, and the reintroduction of the constitutional limit on presidential terms to a maximum of two five-year terms. Since this will not take effect until after the next presidential election, Lukashenka's tenure could hypothetically extend to 2035. No protests were reported.
The survival of an authoritarian regime despite mass protests is not surprising in itself. They are generally not an effective means of regime change, as shown by a statistical analysis of 310 electionrelated mobilization events in 92 nondemocratic countries worldwide between 1989 and 2011, which found that autocratic elites made political concessions in only about a quarter of all cases (Brancati 2016, 179). In this respect, the 2020 mobilization in Belarus is a "typical case" (Onuch and Sasse 2022), even if-according to the same study-concessions are much more likely when protests are large-in other words, when they involve more than 100,000 participants (Brancati 2016, 179), as was the case in Belarus. In fact, none of the theoretical perspectives I have referred to allowed a clear prediction of whether Lukashenka's regime would survive or collapse. Nonetheless, they contribute valuable arguments to the discussion about why Lukashenka has not only remained in office, but also seems to have reconsolidated his rule, something few anticipated.
According to the protest-democracy mainstream, "extreme" degrees of political repression preclude opposition victory, and "insufficient" protest mobilization contributed to the failure of color revolutions in Armenia, Azerbaijan, and Belarus in the 2000s (Bunce and Wolchik 2011, 332;2018, 169). This notwithstanding, from this perspective, it is neither the strength of the regime prior to elections nor the scale of mobilization, but the opposition's implementation of "all aspects of the electoral model" that is held to be crucial to success in repressive electoral-authoritarian systems such as Lukashenka's. International factors, particularly Western support, are considered important secondary factors as they facilitate the diffusion of the electoral model across countries (Bunce and Wolchik 2011, ch. 12;2018).
From this perspective again, the function of the electoral model is to highlight to the incumbent the costs of a continued exercise of power and the credibility of the opposition. This strengthens citizens' optimism that change is possible, which mobilizes them to vote and later to participate in street demonstrations (Bunce and Wolchik 2011, 333). However, the case of the 2020 "Belarusian revolution" shows that many elements of the electoral model do not necessarily require long-term preparation by a well-organized opposition. They can also develop spontaneously or through preexisting social networks and the activities of social media actors, which also boosts optimism about the possibility for change. Yet all of this was not enough to convince Lukashenka that the opposition was credible and he should step down because of the unbearably high costs of continuing in power. Ultimately, it was not harsh repression-as in the first three days-that put an end to the street protests, but their slow waning in the face of large-scale but targeted arrests, Lukashenka's stubbornness in sitting it out, the dwindling prospects for success, and the fading of the mobilizing appeal of emotions.
The main problem with the electoral model theory is the underlying assumption that authoritarian regimes collapse when they are exposed to strong pressure from a united front of "the people" and "the opposition" calling for democratization. For theories of the regime dynamics paradigm, on the contrary, it is the condition of the authoritarian regime itself that is critical to whether it crumbles under pressure from below and from outside, while aspects of democracy are secondary. In this regard, Hale's (2019) argument that mass protests are only effective in patronal regimes when they are associated with elite splits is instructive. First, protest may be an intended product of such splits, in other words, warring political-economic networks may foment mass mobilization to bolster their legitimacy claims and marshal additional resources, which increases the perceived costs of staying in power for the ruling group. This is the typical pattern seen in the post-Soviet color revolutions, which took place under "competing-pyramid" conditions. According to Hale (2006;, 2408-2409, they were, in essence, succession struggles taking place on the occasion of elections won by challengers who mobilized popular support rather than "democratic breakthroughs." By contrast, Lukashenka's regime did not encounter an imminent succession crisis in 2020 (and had never done so before). He stood firmly at the top of an integrated "single-pyramid" power structure both formally (as the president) and informally (as the chief patron). Unlike the "lame ducks" in countries with successful color revolutions, he did not face competing power claims from political heavyweights of a younger generation, such as former prime ministers or holders of other key posts in government.
Second, mass protest can precede and thereby induce divisions among previously loyal networks, as evidenced by the Euromaidan in Ukraine in 2013-2014, where excessive, counterproductive violence ultimately spurred the defection of key elites from the regime (Hale 2019(Hale , 2407(Hale -2408. Although Lukashenka initially made the same mistake as Yanukovych and also triggered a second wave of mobilization, his single-pyramid system did not collapse because he quickly managed to create the credible perception that he would not concede. Unlike his Ukrainian colleague, who fled to Russia in the face of the Euromaidan uprising, Lukashenka projected vitality, ruthlessness, and a resolve to keep the situation under control at all costs. Even seemingly bizarre actions, such as appearing before a squad of riot police officers with an automatic weapon, were part of an "inventive political spectacle" that conveyed his determination to remain in power and to evoke the perception of the absence of alternatives (Morozov 2021). Even more than the public, such performances were staged for the benefit of the regime elites and international actors, especially President Vladimir Putin of Russia, whose continuing support was the decisive factor in his survival in office. By September 2020, defections from the regime had already largely come to a halt (Way 2020, 26), having been only sporadic even before then.
From this perspective, the greatest weakness of the regime challengers around Tsikhanouskaya was not so much their strategic mistakes or programmatic vagueness, as some authors contend (e.g., Moshes and Nizhnikau 2021;Mudrov 2021). Rather, their tragedy was that they did not belong to the elites. It is hard to overestimate how much this fact may have contributed to Lukashenka's stubborn refusal to enter into negotiations with the opposition's Coordination Council. Likewise, potential defectors from within Lukashenka's power pyramid had no incentive to align with the newly emerged opposition, which consisted of regime outsiders who did not appear to be credible contenders for, or even aspirants to, political power.
The same logic applies to Russia's calculations. For Putin's own patronal regime, it is important that the person who replaces Lukashenka be not only loyal to the regional hegemon, but also a regime insider, since any possible transfer of power would need to happen swiftly and smoothly with a minimum of network reshuffling and the related political instability. When Lukashenka shifted toward Putin for support and financial aid in mid-August 2020 (Leukavets 2021), he not only contrived his own political survival. He also tilted the balance between state security-which he had prioritized after 2014-and the security of his regime clearly in favor of the latter (Ambrosio 2022). With the far-reaching loss of Belarus's sovereignty that this move entailed, he paid a high price for remaining in power, the consequences of which are still unpredictable at present. For Belarusian elites, this geopolitical rapprochement, or rather, Lukashenka's unequivocal commitment to submit to Russia in exchange for Putin's backing, was another powerful signal that defection would not pay off. In turn, Putin's unequivocal support for and adherence to his inconvenient colleague was an important precondition for Russia's invasion of Ukraine in 2022.

Conclusion
In empirical terms, interpreting the 2020 events in Belarus through the lens of three paradigms that have emerged in research on mass protest in the post-Soviet space, makes it possible to highlight important (dis)similarities with other protest events in the region. Like the color revolutions in the 2000s, the "Belarusian revolution without a name" erupted on the occasion of a major electoral fraud, but, unlike them, the opposition this time consisted of political newcomers who had neither split from the regime elites nor prepared over the long term to seize power through elections or, if necessary, postelection protests. Unlike the color revolutions but like Russia's "For Free Elections" movement in 2011-2012, the protesters faced a consolidated hegemonic electoral-authoritarian regime in which all major political-economic networks were integrated into a coherent singlepower pyramid whose repressive capacities were intact and did not dissolve when confronted with mass protest. The protesters themselves, on the other hand, were acting in a geopolitical situation in which the West was much less supportive of democratization abroad than it had been in the first decade of the twenty-first century and were careful not to geopoliticize their protest. Unlike his colleagues who were toppled by color revolutions, Lukashenka managed to stabilize his endangered position as president/chief patron because, first, no significant part of the ruling coalition opted to split off and instrumentalize the protest to further its own power ambitions, and, second, he was able to secure Russia's backing. Consequently, the regime was neither ripe for collapse from within, nor was pushed to do so from outside.
In theoretical terms, I have shown, first, that by the spring of 2020, none of the familiar perspectives on mass mobilization in the post-Soviet space suggested that Belarus would soon see street protests with several hundred thousand participants over several months and throughout the country. The traditional opposition parties had been marginalized due to Lukashenka's highly adaptive and effective "preemptive authoritarianism," the regime's repressive capacities were intact, there was little pressure from the West to hold elections in accordance with democratic standards, and no apparent regime or succession crisis that could have fueled intra-elite competition for power.
Second, the sudden opening of the political opportunity structure in the early summer of 2020 as well as the outbreak of mass protests in August 2020 were the result of interacting contingent choices by individual actors and, therefore, truly unpredictable. Lukashenka's miscalculations about his competitors in the upcoming election (which Tsikhanouskaya countered by accepting the challenge, and the representatives of two stymied candidates by joining her), the egregious falsification of election results, and the excessive violence during the first three days of protest set in motion a process that brought the regime to the brink of collapse. The dynamics of collective action were triggered by two mechanisms familiar from social movement studies. The first was electoral fraud, a mechanism typical of electoral autocracies; the second, excessive repression, is found in democracies and nondemocracies alike. Both provided focal points for spontaneous coordination, which substituted for the deliberate "engineering" of protest by an organized opposition.
Third, Lukashenka's ultimate survival is best explained not so much by any failure to implement the electoral model or the weakness of the opposition, but by structural factors, as set out in Way's concept of regime vulnerability-including the strength of the regime and its geopolitical situation -and even more compellingly in Hale's theory of patronal politics, which links regime strength to the structure of its key political-economic networks. In this view, Lukashenka's regime is not just a repressive electoral autocracy but, more precisely, an intact single-pyramid system in which all major networks are coordinated around an informal chief patron who is also the formal head of state. This regime survived because mass protests did not lead to the defection of larger segments of the elite. In contrast, all successful color revolutions took place in electoral autocracies with competing-pyramid systems, in which relevant elite groups split from the regime, allied with ordinary citizens, and mobilized them on the occasion of an election, thereby tilting the balance against the incumbent.
Fourth, looking at the case of Belarus 2020 from three different perspectives does not resolve the long-standing controversy between the agency-focused protest-democracy approach and the more structuralist "regime dynamics" theories concerning the causes of the successful color revolutions in the early twenty-first century. Bunce and Wolchik (2006, 2018 apply their theory of the electoral model exclusively to the postelection protests between 1998 and 2008 in the postcommunist region, thus implicitly acknowledging its narrow geographical and temporal scope. The simultaneous absence of regime defeat and a well-organized opposition equipped with a sophisticated electoral model in Belarus in 2020 does not confirm the theory, as the case took place under different regime conditions. In order to specify these, Hale's (2015) conceptualization of the two patterns of patronal regimes offers a useful starting point for further research. This distinction is consonant with the differentiation between competitive and hegemonic electoral authoritarianism sometimes drawn in the literature (Brownlee 2009;Donno 2013). However, it is not based on different degrees of electoral uncertainty, but on different structural patterns of elite coordination, which may also represent different stages within the life cycle of a patronal regime. Most mobilization events, as well as all regime breakdowns in post-Soviet Eurasia, have so far occurred in the context of competingpyramid structures. According to Hale (2019), this is also the condition under which mass protests can become effective drivers of regime change, either because they are fueled by warring politicaleconomic networks (as in the successful color revolutions) or because they cause latent elite divisions to erupt openly (as in the Euromaidan). Since elites are not divided in single-pyramid systems like that of Lukashenka's Belarus in 2020 or Putin's Russia in 2011-2012, the only chance for mass protests to become effective is to cause substantial parts of the ruling coalition to defect, in other words, to push the regime into the stage of intensively competing networks.
For further research, this means that a comparison of protests within and between the two patterns of patronal regimes is needed, as the mechanisms of regime collapse may differ systematically. This should include more recent cases, which are currently mostly examined in single-case studies. The Belarusian case, for example, suggests that the (un)popularity of the national leader and the role of elections as triggers for heightened elite competition are far less relevant under singlepyramid conditions than under those of competing pyramids. The importance of strong popular support or the credibility of an electoral victory was neutralized because neither factor seriously altered elite assumptions of who the real chief patron of the polity was. So, if an unpopular president proves capable of decoupling elite loyalty from popular support, this might be sufficient for his survival.
In this context, Treisman's argument about the mistakes of authoritarian incumbents being proximate causes of regime breakdown deserves further attention. He stresses the interaction between idiosyncratic decisions made by political actors and structural factors conjecturing that these errors may serve as triggers or "common cause" for other factors in regime breakdown, such as economic crises, violent coups, and leader turnovers. The Belarusian case shows that Lukashenka exposed his regime to the risk of defeat by committing a series of serious blunders against regime outsiders in the spring of 2020. However, this stopped after mid-August 2020. He continued to crack down on challengers from below and outside the regime, but he did so in a more targeted and less excessive manner. Most importantly, at no point did he make gross mistakes affecting regime insiders and Russia, the only international actor he could rely on. More precisely, he credibly conveyed the message that he would not leave his office and that there was no alternative to him.
While it remains to be seen whether developing a theory of when incumbents commit mistakes and when they matter (Treisman 2020, 807) is a realistic goal, one can plausibly speculate that in coherent single-pyramid systems equipped with effective tools to prevent the formation of a meaningful opposition, as in today's Belarus, Russia, and probably most of post-Soviet Central Asia, the mistakes incumbents can commit are the most important-if not the only relevanttriggers for regime collapse. Since such mistakes are as possible as they are unpredictable, challenges to single-pyramid, hegemonic-authoritarian regimes or even their breakdown will continue to come as a surprise.

Notes
1 Valerie Bunce and Sharon Wolchik (2006, 284) were the first to propose the concept of "electoral revolutions," that is, "attempts by opposition leaders and citizens to use elections, sometimes in combination with political protests, to defeat illiberal incumbents or their anointed successors; to bring liberal oppositions to power; and to shift their regimes in a decidedly more democratic direction." More recently, they have refined the concept by distinguishing between "elections," for example, in Slovakia (1998), and "electoral challenges combined with large-scale mobilization," such as the color revolutions, these representing two of four contemporary "modes of popular mobilizations against authoritarian rulers" (Bunce and Wolchik 2018). 2 For a discussion of the differences between the two models see Hale and Way (2017).