The ORE of Politics will be available for subscription in late September. Speak to your Oxford representative or contact us to find out more.

Show Summary Details

Page of

PRINTED FROM the OXFORD RESEARCH ENCYCLOPEDIA, POLITICS ( (c) Oxford University Press USA, 2016. All Rights Reserved. Personal use only; commercial use is strictly prohibited. Please see applicable Privacy Policy and Legal Notice (for details see Privacy Policy).

date: 19 September 2017

Dynamics, Endogeneity, and Complexity in Protest Campaigns

Summary and Keywords

There is an argument that nonviolent civil resistance or protest campaigns should be studied as dynamic and complex phenomenon, rather than a single case comprised of various attributes, such as size, scale, and scope, which are then compared with other cases. As protest campaigns have increased all over the world during the last few years, international relations scholars have begun to devote more time and resources to studying them systematically with new data projects and analytical tools and methods. In light of this emerging research program, one needs to understand that protest campaigns contain large-scale processes of political contention that evolve across time and space. Such evolutionary processes are the result of the interactive relationships among multiple governmental and nongovernmental actors. These interactions reflect “highly interdependent sets of actions and reactions” that generate causal mechanisms and intersect with other large-scale processes which can produce similar and dissimilar outcomes across different political contexts. An argument will be advanced that a “relational mechanisms-process” approach articulated by McAdam, Tarrow, and Tilly provides analytical leverage over such complexity. Most, international relations scholars, unlike social movement scholars, are not familiar with this approach. So, the effort herein is to not only make the case for a relational mechanisms-process approach but also to illustrate it with a partial analysis of the Egyptian uprising on January 25, 2011, which led to Mubarak’s resignation. The end result is a call for theoretical and empirical research that bridges two communities of scholars, one that is dominated by sociologists (social movement scholars) and the other that is dominated by political science (international relations political violence scholars).

Keywords: empirical international relations theory, protest campaigns, civil resistance, relational mechanisms and processes, social appropriation of mobilizational vehicles, carrot and stick strategies, multisectoral coalition formation, diffusion, Arab Spring, Egypt January 25, 2011 uprising

A Definition of Protest Campaigns

In social movement research, a protest campaign is a “sustained, organized public effort making collective claims on targeted authorities (Tilly & Tarrow, 2015, pp. 3–25). It always links at least three parties: a group of self-designated claimants acting on behalf of a constituency, some target of their claim-making (usually governmental authorities), and a general public. A protest campaign grows out of a single protest event and evolves into a broader set of protest events (Tarrow, 2011, p. 191).

In the international relations (IR) community, Chenoweth and Stephan (2011) have produced a dataset that covers nonviolent and violent civil-resistance campaigns. Their definition of a nonviolent civil-resistance campaign is different from but overlaps considerably with the Tilly and Tarrow approach. Chenoweth and Stephan identify a campaign as a series of observable, continuous tactics in the pursuit of a political objective. A campaign can last anywhere from days to years. It has a discernible leadership and is often identified by a name with relatively clear and ending points. Their data comprise over 250 violent and nonviolent campaigns, and was originally published as the Nonviolent and Violent Campaigns and Outcomes (NAVCO) in their book, but it has since been updated in NAVCO 2.0 (Chenoweth & Lewis, 2013).

Although there is very little difference between these two approaches, social movement scholars relax the nonviolence assumption considerably since they acknowledge that protest campaigns are likely to involve both violent and nonviolent forms of protest actions that cannot and should not be disentangled. In this article, it adheres to the social movement definition since the author believes that the presence of violence and nonviolence is an important dimension of protest campaigns that needs to be studied closely.

The IR Political Violence Approach to Protest Campaigns

In the early 21st century, protest campaigns have occurred more frequently all over the world. Between 2006 and 2013, there were a total of 843 world protests, in all regions of the world and across all national income levels. There were 59 reported protest events in 2006, 160 in 2012, and 111 in the first six months of 2013 (Ortiz, Burke, Berrada, & Cortes, 2013). This upsurge in worldwide protests is an important reason why IR conflict scholars have started paying closer attention to them, especially when these events have the potential to bring down regimes and generate full-scale civil wars (Chenoweth & Stephan, 2011; Nepstad, 2011).

In particular, the Arab Spring protest campaigns in 2011, along with the timely publication of the highly visible and lauded book Why Civil Resistance Works: The Strategic Logic of Nonviolent Conflict, by Erica Chenoweth and Maria J. Stephan, generated a new effort to bring systematic, quantitative data to understanding how these campaigns emerge and contribute to domestic and international changes. In fact, special issues in two international relations journals—International Interactions in 2012—and the Journal of Peace Research in 2013—were devoted to discussing the Arab Spring campaigns and the need for new theories and datasets of nonviolent collective action. Their conclusion was that one needed to build and test better theories with more fine-tuned data in order to “help researchers and policymakers shape strategies to support the movements behind these campaigns . . . and to manage changes in the international system that result from . . . their success” (Chenoweth & Cunningham, 2013, p. 271).1 This motivation has produced two new datasets that IR scholars have used frequently in studying protest campaigns: the Nonviolent and Violent Campaigns and Outcomes, 1900–2006 (NAVCO 2.0); and the Major Episodes of Contention, 1945 to present (MEC).2

Utilizing this data, most of the recent IR research on protest campaigns have been conducted in the context of quantitative cross-sectional large n-studies. In most of these studies, IR scholars have been interested in predicting the onset or various outcomes of protest campaigns with independent variables such as: modernization, poverty, development, inflation, the diffusion of cross-border protest campaigns, repression, type of regime, dissident organizational infrastructure, protest history; armed rebellion, and oil rents (Braithwaite, Braithwaite, & Kucik, 2015; Butcher & Svensson, 2016; Chenoweth & Schock, 2015; Chenoweth & Stephan, 2011; Chenoweth & Ulfelder, 2016; Girod, Stewart, & Walters, 2016; Sutton, Butcher, & Svennson, 2014). In a couple of other studies, the type of protest campaign (violent or nonviolent) is related to greater democratization or autocratization outcomes (Bayer, Bethke, & Lambach, 2016; Celestino & Gleditsch, 2013).

Thus far, IR scholars have been overwhelmingly interested in asking basically two questions: What explains the cross-national onset of protest campaigns, and what are the potential cross-national outcomes of these campaigns? In either case, IR scholars are less interested in the internal dynamics of protest campaigns and more interested in whether campaigns and their attributes are related to cross-national variables that explain their onset and outcomes. The underlying assumption is that protest campaigns and their variables are treated as self-contained, homogeneous entities. As a consequence, protest campaigns have been treated essentially as a “black boxed” phenomenon.3

Five Theoretical Issues That IR Studies of Protest Campaigns Overlook

There are five theoretical issues that IR scholars fail to consider in their cross-national studies of protest campaigns—all of which suggest that the emergence and outcomes of protest campaigns are dependent on the interactive relationships among multiple actors that occur during the course of the protest campaigns. These dynamic relationships shape the onset, expansion, and outcomes of protest campaigns.

The first issue is that IR scholars tend to observe each protest campaign as conceptually and temporally bounded with internal attributes or variables that can be compared with other campaigns. For instance, in the NAVCO 2.0 dataset, each protest campaign is described by a host of variables, such as its size, the diversity of its participants, variety of dissident tactics, level of governmental repression, presence of elite (military) defections, structure of dissident political organizations, and appearance of radical groups.4

These variables do indeed go a long way toward unpacking protest campaigns and treating them as multifaceted phenomenon. But in cross-sectional research designs, these variables are treated as static, independent entities. This perspective overlooks any deliberation on how and why the variables emerged in the first place. For instance, if some of the key factors of success for protest campaigns are associated with the participation of large numbers of diverse participants who use innovative and varied tactics, then what are the underlying causal mechanisms that produced these variables? Moreover, protest campaigns rarely start with large numbers of people demonstrating with innovative and diverse tactics. Instead, the size, scale, and scope of dissident participation changes over time and in reaction to the responses of governments, elites, political entrepreneurs, and international actors. Explaining success, therefore, depends on understanding how and why causal mechanisms work together to produce the variables that can be observed during the course of a protest campaign.

A second issue that IR scholars overlook is the theoretical interconnections of campaign episodes across time. At present, cross-sectional designs treat campaigns as independent events. Frequently, successful (and even unsuccessful) protest campaigns follow previous episodes that were smaller and less likely to reach the threshold to be included in a dataset.5 Yet these past episodes have crucial influences on subsequent campaigns, as governments, political opposition groups, and their external supporters learn, adapt, and prepare for the next confrontation. If IR scholars rely on large-scale data projects that establish high thresholds of inclusion, selection bias emerges, but equally importantly, they risk overlooking the theoretical contribution that smaller past episodes of collective action have in shaping and transforming subsequent protest campaigns.

A broader historical examination of the January 25 revolt (2011) in Egypt, for example, reveals that this campaign was preceded by other protest mobilizations: the failed prodemocracy campaign in the Kefaya Movement in 2004–2005; by the presence of and, in some cases, successful wildcat strikes waves from 2004–2008; and by the failed nonsanctioned labor strike at the Mahalla textile factory in 2008, out of which emerged the April 6 political movement, whose members were among the initiators of the revolt on January 25, 2011, that forced Mubarak’s withdrawal (Beinin, 2016).6Alexander and Bassiouny (2014) argue that the coalescence of democracy activists and striking textile workers in the 2008 “Mahalla Uprising” was a “dress rehearsal” for the uprising of January 2011 (p. 100). In short, the characteristics and even the outcomes of prior protest campaigns are linked to future episodes of contention.

A third theoretical issue is associated with the patterns of contention that vary during the lifespan of a protest campaign. Karklins and Peterson (1993) argue that the entry of protest groups, which tend to specialize in different types of contentious actions (e.g., working class groups engage in labor strikes, whereas other groups engage in demonstrations) are time dependent (early vs. late risers) and heavily affected by what other groups have done previously. Moreover, the timing of these contentious events can have differential effects on the outcomes of protest campaigns.

In the Egyptian and Tunisian Arab Spring cases, for instance, large-scale demonstrations by wide segments of the general public were important in bringing down these old autocratic regimes. Yet, in the Egyptian January 25 campaign, widespread labor and civic strikes during the last six days of the uprising were linked more closely to the actual timing of Mubarak’s resignation than to the popular street demonstrations. Within three days after nationwide strikes had shut down governmental institutions, transportation systems, and commercial businesses, Egypt’s military defected from the regime, and shortly afterward, Murbarak resigned (Alexander & Bassiouny, 2014, pp. 202–204). In Tunisia, labor representatives supported local and regional strikes on January 12, 2010, and two days later these were followed by a national strike on January 14, 2010—the day that the Ben Ali regime resigned. Both cases demonstrate that contentious events are related endogenously to other contentious events, as well as to government actions. Moreover, this expanding process of mass dissident protest actions influences not only whether regimes survive or not, but also which actors are likely to have important roles in reshaping future political institutions (Brownlee, Masoud, & Reynolds, 2015).

A fourth theoretical issue is associated with the tendency of IR scholars to treat protest campaigns as essentially nonviolent phenomenon. This assumption is reflected in Chenoweth and Stephan (2011) and in their NAVCO 2.0 dataset. Their approach is to define protest campaigns as largely nonviolent wherein unarmed civilians did not directly threaten or harm the government through the use of armed violence. Civilians rely mostly on tactics such as peaceful protests, sit-ins, boycotts, and other, noninstitutional tactics (Chenoweth & Stephan, 2011, p. 12). Although Chenoweth and Stephan (2011) and their NAVCO 2.0 data allow for the possibility that civilians may shift their nonviolent tactics to violent ones through their coding system, they still treat violent and nonviolent campaigns as conceptually distinct phenomena or as forms of behavior that exist on a continuum—each of which bears separate analysis. In fact, the NAVCO 2.0 dataset does not include violent campaigns; instead, it merely codes the presence of violence (as a binary variable) in a campaign episode. Such an approach is problematic because it obscures the underlying process that generates the violence to begin with, as well as the key role that violence plays in generating a critical mass of demonstrators.

Ketchley (2017) brings this point home through his analysis of the Egyptian uprising of 2011. He shows that street demonstrations and antiregime violence, in the form of militant attacks against police stations, were both “synergetic and complementary.” He documents that highly networked activists conducted nonviolent street demonstrations, while poor urban groups targeted their attacks on local police stations. Confronted with the challenge of dispersing demonstrations and defending their police stations simultaneously, authorities ordered their security forces to leave the streets in order to protect the police stations. This withdrawal made it possible for nonviolent demonstrations to diffuse and increase in numbers, as protesters gained occupational control over main roads and public squares. The bottom line here is that the “uncoordinated” violent attacks by the urban poor played a key role in the ability of Egyptian activists to coordinate, organize, and build on the momentum of their nonviolent tactics by reducing the level of government repression the protesters had to face (Ketchley, 2017, pp. 18–45).

Violence played a similar role in the Tunisian case. During the early days of the Tunisian campaign, footballers (the ultras), who had earlier formed links with Takriz, a “cyber think tank and street resistance network,” clashed repeatedly with security forces. Takriz and the ultras formed the protestors’ fighting core in confronting the security forces before the campaign had reached its full peak of civilian participation. Over 300 people were killed during this period. The resilience of the ultras in the face of state violence was a fundamental factor in stimulating greater dissident participation by reducing the barriers of fear (Dorsey, 2012, p. 413).7

These two cases show that violence in the early phases of a protest campaign may play a critical role by generating a bandwagon effect. If one starts with the proposition that protest campaigns are nonviolent, one is likely to neglect an important process that is critical to their emergence and evolution. Again, a dynamic analysis is crucial for understanding the evolutionary process of protest campaigns, and it is important that greater sensitivity is paid to the timing, magnitude, and intensity of the violence that occurs in these campaigns.8

This rigid distinction between violent and nonviolent protest campaigns also obscures another important theoretical relationship. More often than not, the insurgencies, rebellions, and guerrilla wars that constitute violent resistance campaigns are preceded by “nonviolent” protest campaigns, indicating that there are crucial linkages between these forms of political contention. The Arab Spring cases involving Libya and Syria are prime examples. Both situations began as protest campaigns and quickly escalated to full-scale armed conflicts—similar in nature to the violent campaigns that are described in the NAVCO dataset.9 So maintaining a distinct conceptual line between contentious events that are nonviolent relative to those that are violent draws us away from thinking about dynamic escalatory processes as well.

Finally, IR studies of protest campaigns underestimate the role of complexity and contingency. The underlying assumption in cross-sectional, campaign-centered designs is that each campaign is likely to share predictable attributes or variables (e.g., number of protesters, or broad vs. narrow based participation) that lead to likely outcomes. However, contingency plays a significant role in influencing the size, scale, and scope of these attributes and variables during the onset and expansion of protest campaigns. Unexpected pivotal or focal events, can transform bystanders into large numbers of antiregime dissenters very quickly, or demobilize them just as rapidly (Demirel-Pegg, 2017; Karklins & Peterson, 1993; Snow & Moss, 2014). Whether they do so or not depends on the presence of other mechanisms. For instance, the fall of the Ben Ali government in Tunisia became an important focal event that diffused to Egypt a few weeks later. But, the successful diffusion of protests to Egypt also depended on the presence of preexisting activist networks, social media tools, and resonating frames that were crucial in coordinating protesters (Dupont & Passy, 2011). The point here is that describing a protest campaign in terms of its participation size is a static attribute that masks crucial mechanisms that brought about this scope in the first place. The problem cross-sectional approaches face is that the intersection between focal events and these mechanisms is highly contingent and thus hard to model empirically.

Another way of looking at the issue of complexity is that protest campaigns involve strategic interactions among challengers and state and elite actors that are essentially dynamic. Not only do contingent events play a significant role in influencing the direction of contention, but “errors, false rumors, misunderstandings and inconsistent behavior” among multiple actors are also likely to impact contentious actions (Koopmans, 2004). In addition, these interactive relationships evolve over time and involve path-dependent processes. Other indeterminate factors, such as spontaneity, critical mass, and adaptive learning between and among actors, also shape and transform contentious behavior. The issue is that protest campaigns host dynamic processes that evolve over time and are highly complex affairs, and IR datasets on protest campaigns are not useful for tracing these patterns because the variables that describe these campaigns are essentially static. In sum, all these issues point to the necessity of building and testing dynamic theories of political contention that embrace both endogeneity and complexity.

Does this mean that existing IR theories, data, and quantitative tools are not suited to capturing the dynamics of political contention in protest campaigns? Not completely, but at the moment, they are more directed toward a static, cross-sectional comparison of protest campaigns than toward their internal protest dynamics that evolve over time. If there is a shift toward an evolutionary perspective, does it mean that an understanding of protest dynamics is stuck in historical description, or is there an alternative theoretical option that can complement and enrich the existing theories and quantitative evidence on protest campaigns? One promising way to bridge historical thick description that is amenable to complexity with an empirical approach that is comparative is the “relational mechanisms-process” approach advanced by the “dynamics of contention (DOC)” framework (McAdam, Tarrow, & Tilly, 2001).

“Relational Mechanisms” and “Processes” in the Dynamics of Contention (DOC) Approach

The DOC approach starts with the proposition that protest campaigns should not be considered discrete events with characteristics that can be compared with one another.10 Instead, protest campaigns are repositories of “relational mechanisms” and “processes” that are similar across various campaigns but combine together to produce different (and possibly similar) outcomes. If protest campaigns are to be compared, their processes and relational mechanisms should be the subjects of comparison. The goal is not to build general overarching theories, but to search for robust, widely applicable causal mechanisms that explain the crucial but not all the features of contention.

This emphasis on mechanisms and processes is based on the assumption that protest campaigns are the products of dynamic interactions among various actors whose “mutual orientation to each other defines a fluid, and socially constructed ‘field of contention.’” This field of contention is a “set of adversarial relationships that is embedded in a legal/institutional system that effectively constrains the strategic options available to all contenders” (McAdam & Tarrow, 2000, p. 149). Among these actors are state officials, challenging groups and non-state elites, bystander publics and representatives of the news media. In the context of protest campaigns, “relational mechanisms” alter the relationships between “people, groups, and interpersonal networks” which shape the trajectories of protest campaigns.

Although there are many types of relational mechanisms, three are important for protest campaigns. The first one is the attribution of similarity, or the successful effort of dissident entrepreneurs to frame the claims and identities of different dissident groups as sufficiently similar to mobilize against the regime. A second mechanism is the social appropriation of mobilizing vehicles, or the ability of dissidents to seize old or form new networks and institutions in order to coordinate actions and strategies with previously inert populations. These vehicles can be traditional (churches, mosques) or reflect new forms of innovation (Facebook, Twitter), but their role is to circumvent any social, political, or economic constraints that prevent mobilization. A third mechanism involves the government’s “carrot and stick” strategies of repression and accommodation, which aim to contain transgressive political contention. In this instance, governmental actors extend concessions at the same time that they repress activists with the use of force. The strategy is designed to satisfy moderate protesters with some concessions while deterring future mobilization by raising the costs of protesting. Whether this strategy works depends on whether governmental actors make adequate concessions and use proportional levels of violence and whether it is perceived as such by protesters. If not, protesters may be radicalized, generating a bandwagon effect of participation and steadfastness by activists.

These and other relational mechanisms—either by themselves or in combination with others—bring about large-scale “processes,” such as mass mobilization (size, scale, and scope of political contention), multisectoral coalition formation (broad based societal coalition), diffusion (geographic spread of contention), and scale shift (a dramatic increase in the scale and scope of contentious political actions). Together these processes shape the location, character, and consequences of contention during protest campaigns (McAdam et al., 2001, p. 26). The bottom line is that scholars can compare protest campaigns in terms of these relational mechanisms and their effects on the processes of political contention. The aim is not to approach protest campaigns solely through the lens of thick historical description but to look for recurring relational mechanisms and processes that have similar and dissimilar connections in different political or social contexts and why that might be the case.

An Application of a Mechanism-Process Approach in the Egyptian Uprising of January 25, 2011

The linkages between relational mechanisms and processes of political contention include the intersections between (a) “attributions of similarity” and multisectoral coalition formation, (b) the “social appropriation of mobilizing vehicles” and diffusion, and (c) the government’s “carrot and stick” strategies and diffusion. In the interest of brevity, the interconnections between the processes of coalition formation and diffusion are not discussed here, although this is also an important part of the evolutionary story of this protest campaign.

Attributions of Similarity: Police Brutality, Mobilization, and Multisectoral Coalition Formation

A similarity mechanism is defined by McAdam and colleagues (2001, p. 334) as the “mutual identification of actors in different sites as being sufficiently similar to justify common action.” This mechanism emerges when dissident entrepreneurs “frame the claims and identities of different actors as sufficiently similar to each other to justify a coalition formation.” The widespread dissatisfaction with police brutality was a unifying theme that brought large numbers of protesters to the streets on the first day of the Egyptian protest campaign.

The January 25 uprising was broader and deeper than previous recent events, not only because the scale and scope of the protests were larger, but the range of participants extended beyond prodemocracy and professional groups. According to El-Ghobashy (2012, p. 24), Egypt’s uprising occurred when three sets of societal protest sectors converged among workers, professional associations, and local neighborhood communities. In the ten years (and especially in the three years) preceding the January 25 Revolt, these groups engaged in a great deal of civil unrest. For instance, industrial workers, civil servants, and trade practitioners engaged in workplace protests. Professional groups, including lawyers’ and doctors’ syndicates and remnants of the anti-Mubarak Kefaya movement (students, Muslim Brotherhood youth, and liberal political activists) participated in antigovernment demonstrations. Lastly, neighborhood protests, whether on a single street or in an entire town, increased as Copts, Sinai Bedouins, farmers, microenterprise owners, or local citizens organized collective actions. By January 25, 2011, every protest sector had “field experience with police rule,” which was associated with excessive and arbitrary brutality (El-Ghobashy, 2012, p. 25).

During the last ten years of his tenure, Hosni Mubarak expanded the security apparatus in the Interior Ministry and the Central Security Forces.11 The police presence was ubiquitous in the everyday lives of Egyptians: the police issued passports, drivers’ licenses, and birth and death certificates; vetted political and academic appointments; monitored workplace environments; and mediated labor-management disputes. In addition, the police supplemented their wages by organizing the drug trade; collecting tributes from taxis, bus drivers, and shopkeepers; or setting up “loan shark” and protection rackets and (Amar, 2011, p. 84; El-Ghobashy, 2012, pp. 24–25).

In addition to the development of increased brutality from local police, the Interior Ministry and Central Security Services outsourced coercion to the baltagiyya (e.g., thugs or gangs) by paying and training them to use sexualized brutality to punish and deter female and male protesters alike. These “police-connected” thugs also harassed wealthy citizens for money, molested women on crowded streets, terrorized shopkeepers, and small-business owners and tortured those who refused to submit (Amar, 2011, pp. 84–85; Kandil, 2012, p. 269).

The third pillar in the security forces was the Central Security Services, which is autonomous from the Interior Ministry, and responsible for monitoring and infiltrating political-opposition and civil associations. This police force became notorious for its policies of detaining and torturing masses of dissidents. In 2010 alone, there were approximately 17,000 detention centers.12 Its presence was also everywhere including residential neighborhoods and universities alike. Over time, the Interior Ministry became one of Egypt’s most powerful branches, rivaling the army. One estimate is that the Interior Ministry had an army of approximately two million informers: one for every 40 Egyptian citizens (Shatz, 2011).

Consequently, violence permeated the relationship between the police and Egyptian citizens, who developed a profound hatred of them. By the start of the 2011 uprising, there was not a single segment of Egypt’s political and economic society that had not been touched by police brutality, and this proved to be a unifying force in encouraging large numbers people to demonstrate peacefully in opposition to it (Amar, 2011). In fact, photos and video tape of the brutal death, in 2010, of the owner of an Internet cafe, Khaled Mohammed Said, at the hands of local police in Alexandria were posted on the Internet and used to mobilize protestors in what started out to be an antipolice rally on January 25, the day on which Egypt’s annual National Police Day holiday is celebrated (Amar, 2011). The central organizing theme of the rally was “we are all Khaled.”13

Social Appropriation of Mobilizational Vehicles: The Role of Social Media in the Mobilization and Diffusion of Broad-based Dissent

In the Dynamics of Contention framework, social appropriation of mobilizational vehicles is an important mechanism driving coalition formation and diffusion. It refers to the challenger’s ability to create or utilize organizational contexts or tools that can be transformed into instruments of contention (McAdam et al., 2001, p. 47).14 In the Egyptian revolt, the effective use of the new social media was a crucial instrument in two different processes (coalition formation and diffusion).

Dissident entrepreneurs stitched together a multisectoral coalition by utilizing Facebook, Twitter, and You Tube to make broad appeals and formulate strategies of collective action.15 The use of these media platforms had three effects. First, it aided collective action by lowering information costs and enhancing coordination strategies (Howard & Hussain, 2011). Second, it brought about a “cascade of messages” that helped raise people’s expectations for success (Howard, Duffy, Freelon, Hussain, Mari, & Mazaid, 2011). Lastly, the regime’s attempt to shut down the Internet had an unintended backlash effect that brought many more participants to the streets (Hassanpour, 2011).

Regarding the first effect on political mobilization, Howard and Hussain (2011, pp. 39–40) maintain that Facebook provided a logistical infrastructure for the initial stages of the Egyptian protest. For instance, Google executive and Internet activist, Wael Ghonim, set up the “We are all Khaled Said” Facebook page, which documented the Said’s death. The webpage would become a virtual forum for protest against police brutality and human rights violations. Facebook members eventually totaled over 500,000, as users posted photographs, videos, and lists of the names of corrupt police officers. After the Tunisian uprising, youth activists used the Facebook page to announce a massive demonstration protesting police brutality in Cairo on January 25, 2011, National Police Day. More than 50,000 members confirmed that they would participate (Lim, 2012, p. 242). The “We are all Khaled Said” Facebook page allowed activists to build solidarity around shared grievances and to identify collective political goals (Howard & Hussain, 2011, p. 42).

Meanwhile, coordination during the January 25 street protests was facilitated by activists and their followers twittering and sending cell-phone text messages to amass large numbers of protestors in Tahrir Square in Cairo. The same strategies were played out in other Egyptian cities and towns. The youth activists were also careful to direct large crowds to attend protest sites in working-class neighborhoods, not only to mobilize wider numbers of participants, but also to highlight the connection between their living conditions and the political environment. As one young activist, Amr Salah, said, “We needed to make the connection between liberty and bread to attract broader support” (Jensen, 2011).

Social media networks also allowed for instant, on-the-ground reports that activists could use to adapt their mobilization strategies to both overwhelm and avoid security forces (Lim, 2012, p. 243). The new social media helped to “turn individualized, localized and community-specific dissent into a structured movement with a collective consciousness about shared plights and opportunities for action” (Howard & Hussain, 2011, p. 41). The remarkable aspect of the political mobilization during the 18 days of revolt was the lack of leadership associated with the demonstrations.

The second important impact that the new social media had was its role in influencing people’s expectations of success (Howard et al., 2011). Empirical evidence of the messages sent in Egypt via Twitter and in blog posts show that a spike in online revolutionary conversations often preceded major protest events on the ground. In-country Twitter traffic peaked on the day the street protests reached into the thousands and then peaked again during the last days of Mubarak’s reign. Much of the Twitter traffic and blog posting dealt with political change. Meanwhile, Twitter traffic inside and outside the North African region increased substantially, especially after Tunisia’s president, Ben Ali, resigned. Much of the Twitter traffic produced a cascade of messages across the region that centered on themes of “freedom, liberty, and democracy” and raised public expectations that political uprisings could be successful (Howard et al., 2011).

Finally, a study by Hassanpour (2011) shows that the Egyptian government’s decision to shut down the Internet, which had its full impact on January 28, increased the dispersion of the protesters across Cairo and overwhelmed the numbers of the police forces. In short, “there were too many protests in too many places” (local blogger, quoted in Hassanpour, 2011, p. 36). Dispersion of the protest action occurred because the lack of cell-phone coverage and an Internet connection forced Egyptians who were worried about their friends and family members to join the crowds in the streets to find out about them. As more people entered the streets for face-to-face communications, the uprising swelled and became more decentralized, which made it harder for the Egyptian government to control and repress it (Hassanpour, 2011, p. 28).

These new technologies played an important role in creating networks across families, neighborhoods, mosques, and other local civic associations instead of the typical political associations or parties. Hence, they brought together otherwise remote and disparate groups to form a broad-based coalition. It was especially important in sustaining wildcat strikes as rank-and-file workers coordinated their own lines of formal leadership and union support independent of their traditional union leaders, who had been coopted by the regime. Lastly, the digital technologies made it possible for workers, who had been traditionally insulated from other sectors of the economy, to coordinate their wildcat strikes with other workers from a variety of different settings, as well as with protesters. The result was a broad-based coalition that transcended sectoral and class divisions (Johnston, 2011, p. 16).

Government’s Carrot and Stick Strategies and the Diffusion of Protests

Carrot and stick” strategies refer to a government’s attempt to demobilize protesters by offering limited concessions that are then coupled with repression. The government’s strategy is to appeal to moderates to go home with the future expectation that their demands will continue to be met while leaving the radicals in the streets. This divide-and-rule strategy is designed to break down the multisectoral coalition with the intention of separating the protesters and easing the dispersal of the remaining protesters.

However, this is a strategy that could misfire because the combination of concessions and a government crackdown can have a backfire effect. Concessions (especially superficial ones) signal to protesters that the government is weakening and capable of being pressured to provide deeper reforms. This often leads challengers to expand their demands while sustaining and increasing their opposition. If the concessions are combined with a government crackdown, the situation becomes highly combustible. More precisely, a government crackdown in the midst of protester activity can incite the participation of bystanders—which in turn can escalate the scale and scope of the opposition. This backfire effect occurs when governments miscalculate the level and range of sympathy the protesters have garnered among the rest of the population.16 Therefore, diffusion may occur. In this instance, diffusion occurs when collective actions are emulated by other groups across different geospatial locations.

These series of concatenating mechanisms occurred in the Egyptian case. Two prearranged events scheduled on January 28 and February 1, 2011, demonstrate these sequences. On January 25, the first day of the campaign, a rally was organized by April 6 Youth Movement, prodemocracy activists and other youth groups linked to opposition political parties, groups advocating labor rights, and the Muslim Brotherhood. They also announced the sites and time of the demonstrations on the Internet and Facebook.17

By the evening, tens of thousands of protesters had made it to Tahrir Square. Meanwhile, thousands of demonstrations also occurred in Alexandria and other major Egyptian cities. By the second day, however, the government began to crackdown on these demonstrations. Riot police used tear gas and water cannons along with thousands of thugs, who wielded knives and sticks against the protesters. By the third day, January 27, the government had shut down the Internet and text-messaging services that connected Egyptian businesses, banks, Internet cafes, schools, embassies, and government offices. In response, the opposition groups called for a “Day of Anger” demonstration after Friday prayers on January 28, largely in reaction to civilian deaths caused by police brutality.

By the end of the day on Friday, January 28, after bloody battles that had been over several hours long and left hundreds of civilians dead, the outnumbered police retreated, and army tanks rolled into the streets of Cairo. Similar actions occurred in Suez, Alexandria, and the towns and small cities of the Nile Delta. By the end of the evening, 365 citizens had died and another 5,000 had been injured. The police lost 32 personnel and reported 1,079 injuries. Meanwhile, 99 police stations and 3,000 police trucks were burned throughout Egypt, as was the National Democratic Party headquarters in Cairo (El-Ghobashy, 2012, p. 38). During this period, neighborhood groups that had defeated the police established citizen governance committees and citizen patrols to protect their residential areas from looters.

The diffusion of collective action between January 25 and January 27 and across different locations in Cairo and elsewhere (and especially in neighborhood areas that were unused to such displays) “emboldened Egyptians who revised their calculations of what was possible and reduced their uncertainty about the consequences of action.” It broke down the divisions between different groups, forcing the regime to confront them simultaneously when “for thirty years, it had done so serially” (El-Ghobashy, 2012, p. 33).

On January 29, Mubarak appeared on state television and announced political concessions: (a) the dismissal of the cabinet, (b) the appointment of a new vice president; and (c) a change in the prime minister. According to El-Ghobashy (2012), if Mubarak had extended the offer four days earlier, it would have been a shrewd move; by the time he did extend it, however, the proposed cabinet reshuffle only “sharpened the population’s apprehension of imminent victory, spurring them to stay out doors and demand nothing less than the ouster of the president” (pp. 38–39).

By January 30, the coalition of protesters had expanded to include the Muslim Brotherhood and secular opposition groups (including the individual participation of Mohammed ElBaradei) who demanded Mubarak’s resignation. Subsequently, protest organizers called for a “million-man march” in Cairo on February 1, and the Army publicly announced that it would not use force against the demonstrators.

As expected, on February 1, between 250,000 and a million protesters filled the streets of Cairo as Mubarak announced, in another television address, that he would not run for re-election in 2011, and that he would reform two articles of the constitution to ensure this.18 On February 2, the morning after the million-man march, the government organized over 3,000 attackers (baltagiyya) armed with guns, knives, machetes, pipes, brass knuckles, and tear-gas canisters, who entered Tahrir Square and attacked protestors accelerating a new round of demonstrations in the days to follow. After the initial round of government attacks, dozens of men riding horses and camels charged at the demonstrators, who fought back with their hands and bodies. The third government attack began during the evening when thousands of baltagiyya, plainclothes policemen, and snipers confronted the demonstrators, who then moved behind the barricades they had built around Tahrir Square, throwing stones and Molotov cocktails. The battle lasted through the night, and by the morning the police and baltagiyya had fled in the aftermath of what would later be called the Battle of the Camels. Six hundred people were wounded and three killed in the day’s battle (International Crisis Group Report, 2011a, pp. 10–13).

From February 3 to February 6, an “uneasy and unstable détente” emerged between the protestors and the regime as both sides struggled to find a strategy to break the impasse. As the regime initiated discussions with political groups on February 6, Tahrir Square protest leaders worried that they would lose their momentum and that splits within their ranks over the political concessions would encourage people to leave the streets.19 They were aware that protestors were getting weary and, after weeks of not working, wanted to get back to work. In an effort to regain lost momentum, they successfully escalated the protests in the form of “strikes and joint actions with other political and labor groups” (International Crisis Group Report, 2011a, pp. 10–13).

On February 7, 2011, thousands of workers joined wildcat strikes in a part of the country that had been untouched by the protest movement. Wildcat strikes occurred across wide sectors of the Egyptian economy: public electricity, communications, sanitation, railways, buses, and oil. Doctors, lawyers, and pharmacists joined in as well, along with workers in state-owned companies. Workers pressed for higher wages, fairer working conditions, a minimum-wage law, and free trade unions. Strikes by industrial labor and by civil servants diffused to include education, energy, agriculture, transportation, the ports, and the state-owned press (El-Ghobashy, 2012, pp. 4–5).20 By the end of the week, more than 300,000 workers were on strike (Alexander & Bassiouny, 2014, pp. 200–202), and millions of people all over Egypt were involved in demonstrations. Meanwhile, the strikes spread across main industrial cities and zones from Alexandria in the north to Aswan in the south. They also spread into military-owned factories that were run by Egypt’s generals (Naguib, 2011).

On February 10, Mubarak made another televised speech and announced additional concessions.21 However, he refused to step down formally. Strikers and protesters alike stayed in the streets, and on February 11, more than 15 million people were estimated to have been involved in organized demonstrations all over the country. Workers continued to organize demonstrations from their workplaces, signaling that they would paralyze the country of Mubarak refused to back down. On the evening of February 11, the new vice president Omar Suleiman announced that Mubarak had resigned and that his powers had been transferred to the military.

To put it briefly, the regime’s strategy over the course of the protest wave involved a “carrot and stick” approach that combined violent repression, both official and unofficial, with limited concessions in an effort to “peel off” parts of the opposition. The slow, frequently piecemeal forms of concessions that Mubarak offered over the course of those 18 days undermined the government’s survival. Protest leaders viewed the concessions as encouraging signs that they were close to toppling Mubarak and pressed harder for their central demand. A broader segment of the protesters may have been swayed to go home if a “contrite” Mubarak had offered substantive concessions early in the protest wave. Meanwhile, Mubarak’s use of indiscriminate violence also backfired as outraged demonstrators increased in numbers in its aftermath (International Crisis Group, 2011a, pp. 6–7)

In summary, this application of a “mechanism-process” approach to the Egyptian 2011 case has demonstrated how three relational mechanisms (attributions of similarity, the social appropriation of mobilizing vehicles, and the government’s repression and accommodation strategies) were each responsible for the broader processes of mobilization, multisectoral coalition formation and diffusion. A cross-sectional design would certainly have neglected the dynamics of the Egyptian uprising in 2011, and even a longitudinal analysis of the associations between data on strikes, demonstrations, and protests would have overlooked the underlying relational mechanisms that would have accounted for the empirical evidence.

Do these three mechanisms need to be present in every protest campaign? Not necessarily. Many small, episodic protest campaigns that are short-lived with unsuccessful outcomes are unlikely to manifest all of these mechanisms. That is not to say that protest campaigns cannot occur without them, but in cases of large-scale protest campaigns with successful outcomes, such as regime change, the probability that these relational mechanisms will be present is high. In the Tunisian uprising, in 2011, for instance, all three of these mechanisms were present, although the way they played out was not identical to how they played out in Egypt (International Crisis Group, 2011b; Lynch, 2012). In fact, these three mechanisms are not the only ones advanced by the dynamics of contention research program. In fact, McAdam et al. (2001) documented more than two dozen relational mechanisms—all or some of which have the potential to be important in understanding the dynamics of protest campaigns. The effort should be to systematically identify when, where, and how these mechanisms unfold in the context of many protest campaigns with the aim of understanding how they operate in similar or dissimilar political contexts. The aim is not to build an overarching theory about how protest campaigns evolve dynamically, but to support an agenda for middle-range theories that emphasize these relational mechanisms.


It is important to reorient the theoretical thinking about protest campaigns from a static, cross-national orientation to one that is rooted in understanding the dynamics of political contention that occur in the context of these campaigns. The argument is that protest campaigns are endogenous and complex episodes of political contention, and thus theories need to incorporate the causal mechanisms that can be found in a mechanisms-process approach advanced by the dynamics of contention framework. The brief case study on the Egyptian revolt demonstrates the utility of understanding a protest campaign from this vantage point. This effort is a call for bridging two communities of research that seem isolated from each other. At this point, IR scholars tend to be interested in protest campaigns for two theoretical reasons. First, they are interested in whether “nonviolent” protest campaigns are successful, a perspective that reflects the influence of Gene Sharp’s (1973) works on strategies of nonviolence. Second, IR scholars are interested in the regional and international effects of protest campaigns, which reflects traditional theoretical concerns about neighborhood effects, stability, democratic transitions, and rivalries between states. Yet IR scholars risk neglecting another serious and rich vein of research that is represented by social movement scholars (Schock, 2013) whose work will have important links to their two theoretical interests above.


Alexander, A., & Bassiouny, M. (2014). Bread, freedom, social justice: Workers and the Egyptian revolution. London: Zed.Find this resource:

Alimi, E. Y., Bosi, L., & Demetriou, C. (2012). Relational dynamics and processes of radicalization: A comparative framework. Mobilization, 17, 7–26.Find this resource:

Almeida, P. (2005). Multisectoral coalitions and popular movement participation. Research in Social Movements, Conflicts and Change, 26, 65–99.Find this resource:

Almeida, P. (2008). Waves of protest: Popular struggle in El Salvador, 1925–2005. Minneapolis: University of Minnesota Press.Find this resource:

Amar, P. (2011, February 1). Why Mubarak is out. Jadaliyya. Online publication. Retrieved from this resource:

Bayer, M., Bethke, F. S., & Lambach, D. (2016). The democratic dividend of nonviolent resistance. Journal of Peace Research, 27, 5–40.Find this resource:

Beinin, J. (2016). Workers and thieves: Labor movements and popular uprisings in Tunisia and Egypt. Stanford, CA: Stanford University Press.Find this resource:

Beissinger, M. R. (2002). Nationalist mobilization and the collapse of the Soviet State. Cambridge, U.K.: Cambridge University Press.Find this resource:

Braithwaite, A., Braithwaite, J. M., & Kucik, J. (2015). The conditioning effect of protest history on the emulation of nonviolent conflict. Journal of Peace Research, 52, 697–711.Find this resource:

Brownlee, J., Masoud, T., & Reynolds, A. (2015). The Arab Spring: Pathways of repression and reform. Oxford: Oxford University Press.Find this resource:

Butcher, C., & Svensson, I. (2016). Manufacturing dissent. Journal of Conflict Resolution, 60, 311–339.Find this resource:

Celestino, M. R., & Gleditsch, K. S. (2013). Fresh carnations or all thorn, no rose? Nonviolent campaigns and transitions in autocracies. Journal of Peace Research, 50, 385–400.Find this resource:

Chenoweth, E. (2011). Nonviolent and violent campaigns and outcomes dataset (v. 1.1). University of Denver. Retrieved from

Chenoweth, E., & Cunningham, K. (Eds.). (2013). Understanding nonviolent resistance [Special Issue]. Journal of Peace Research, 50, 271–276.Find this resource:

Chenoweth, E., & Lewis, O. A. (2013). Unpacking nonviolent campaigns introducing the NAVCO 2.0 Dataset. Journal of Peace Research, 50, 415–423.Find this resource:

Chenoweth, E., & Schock, K. (2015). Do contemporaneous armed challenges affect the outcomes of mass nonviolent campaigns? Mobilization: An International Quarterly, 20, 427–451.Find this resource:

Chenoweth, E., & Stephan, M. J. (2011). Why civil resistance works: The strategic logic of non-violent conflict. New York: Columbia University Press.Find this resource:

Chenoweth, E., & Ulfelder, J. (2016). Can structural conditions explain the onset of nonviolent uprisings. Journal of Conflict Resolution, 61, 298–324.Find this resource:

Demirel-Pegg, T. (2014). From the streets to the mountains: The dynamics of transition from a protest wave to an insurgency in Kashmir. Mobilization: An International Quarterly, 19, 309–327.Find this resource:

Demirel-Pegg, T. (2017). The dynamics of the demobilization of the protest campaign in Assam. International Interactions, 23(2), 175–216.Find this resource:

Davenport, C., & Moore, W. H. (2012). The Arab Spring, Winter, and Back Again? (Re)Introducing the Dissent-Repression Nexus with a Twist. International Interactions, 38(5), 703–733.Find this resource:

Dorsey, J. M. (2012). Pitched battles: The role of ultra soccer fans in the Arab Spring. Mobilization: An International Journal, 17, 411–418.Find this resource:

Dupont, C., & Passy, F. (2011). The Arab Spring, or how to explain those revolutionary episodes? Swiss Political Science Review, 17, 447–451.Find this resource:

El-Ghobashy, M. L. (2012). The praxis of the Egyptian revolution. In J. Sowers & C. Toensing (Eds.), The journey to Tahrir: Revolution, protest, and social change in Egypt (pp. 21–40). New York: Verso.Find this resource:

Dugan, L., & Chenoweth, E. Government Actions in a Terror Environment – Israel (GATE-Israel). hdl:1902.1/18903, Harvard Dataverse, V2, UNF:5:jPRWAw/AaA2IITkaFQJfPg==.Find this resource:

Girod, D. M., Stewart, M. A., & Walters, M. R. (2016). Mass protests and the resource curse: The politics of demobilization in rentier autocracies. Conflict Management and Peace Science. Advance online publication.Find this resource:

Ghonim, W. (2012). Revolution 2.0: A memoir. New York: Houghton Mifflin Harcourt.Find this resource:

Hassanpour, N. (2011, August). Media disruption exacerbates revolutionary unrest: Evidence from Mubarak’s natural experiment. Paper presented at the Annual Meeting of the American Political Science Association, Seattle, WA. Retrieved from this resource:

Howard, P. N., Duffy, A., Freelon, D., Hussain, M., Mari, W., & Mazaid, M. (2011). Opening closed regimes: What was the role of social media during the Arab Spring? (Working Paper 2011.1). Retrieved from the Project on Information Technology and Political Islam website. Available at

Howard, P. N., & Hussain, M. M. (2011). The role of digital media. Journal of Democracy, 22(3), 35–48.Find this resource:

Ill-Treatment and Torture (ITT) Dataset (2012). Principal Investigators: Courtenay R. Conrad and Will H. Moore. Retrieved from

International Crisis Group (2011a). Popular protest in North Africa and the Middle East (I): Egypt victorious? (Report No.101). Retrieved from the International Crisis Group website.

International Crisis Group. (2011b). Popular protest in North Africa and the Middle East (IV): Tunisia’s way (Report No. 206). Retrieved from the International Crisis Group website.

Jensen, J. (2011, February 13). Behind Egypt’s revolution: Youth and the internet. Global Post Special Report on the Arab Awakening. Retrieved from this resource:

Johnston, T. (2011). What’s old is new again: Social media, protests and the enduring role of civil society. Unpublished paper. Retrieved from

Kandil, H. (2012). Soldiers, spies and statesmen: Egypt’s road to revolt. New York: Verso.Find this resource:

Karklins, R., & Peterson, R. (1993). Decision calculus of protesters and regimes: Eastern Europe 1989. Journal of Politics, 55, 588–614.Find this resource:

Ketchley, N. (2017). Egypt in a time of revolution: Contentious politics and the Arab Spring (Cambridge Studies in Contentious Politics). Cambridge, U.K.: Cambridge University Press.Find this resource:

Koopmans, R. (2004). Protest in time and space: The evolution of waves of contention. In D. A. Snow, S. A. Soule, & H. Kriesi (Eds.), The Blackwell companion to social movements (pp. 19–46). Oxford: Blackwell.Find this resource:

Lim, M. (2012). Clicks, cabs and coffee houses: Social media and oppositional movements in Egypt, 2004–2011. Journal of Communication, 62, 231–248.Find this resource:

Lynch, M. (2012). The Arab uprising: The unfinished revolutions of the new Middle East. New York: Public Affairs.Find this resource:

Major Episodes of Contention (MEC) data. Principal investigator: Erica Chenoweth, University of Denver. Retrieved from

McAdam, D., & Tarrow, S. (2000). Nonviolence as contentious interactions. PS: Political Science and Politics, 33(2), 149–154.Find this resource:

McAdam, D., Tarrow, S., & Tilly, C. (2001). Dynamics of contention. Cambridge, U.K.: Cambridge University Press.Find this resource:

McLaughlin, T., & Pearlman, W. (2012). Out-group conflict, in-group unity? Exploring the effect of repression on movement fragmentation. Journal of Conflict Resolution, 56, 41–66.Find this resource:

Naguib, S. (2011, September). Egypt’s unfinished revolution: Egypt since the fall of Mubarak. International Socialist Review, 79. Retrieved from this resource:

Nepstad, S. E. (2011). Nonviolent revolutions: Civil resistance in the late 20th century. Oxford: Oxford University Press.Find this resource:

Ortiz, I., Burke, S., Berrada, M., & Cortes, H. (2013). World Protests, 2006–2013 (Working Paper 2013). Initiative for Policy Dialogue and Friedrich-Ebert-Stiftung New York. Retrieved from this resource:

Raleigh, C., Linke, A., Hegre, H., & Karlsen, J. (2010). Introducing ACLED: Armed conflict location and event data. Journal of Peace Research, 47, 651–660.Find this resource:

Rod, E. G., & Weidmann, N. B. (2013). Protesting Dictatorship: The Mass Mobilization in Autocracies Database. Working Paper, Department of Konstanz. Cited in Weidmann, Nils B. and Rod, Espen Geelmuyden (2015) Making Uncertainty Explicit: Separating Reports and Events in the Coding of Violence and Contention. Journal of Peace Research, 52, 125–128.Find this resource:

Rasler, K. (1996). Concessions, repression, and political protest in the Iranian revolution. American Sociological Review, 61, 132–152.Find this resource:

Schock, K. (2013). The practice and study of civil resistance. Journal of Peace Research, 50, 277–290.Find this resource:

Sharp, G. (1973). The politics of nonviolent action. Boston: Porter Sargent.Find this resource:

Shatz, A. (2011). After Mubarak. London Review of Books, 33(4), 7–8.Find this resource:

Social Conflict in Africa Database. Principal Investigators: C. Hendrix & I. Salehyan. University of Denver. Retrieved from

Social, Political and Economic Event Database (SPEED) Project. Cline Center for Democracy, University of Illinois at Urbana-Champaign. Retrieved from

Snow, D. A., & Moss, D. M. (2014). Protest on the fly: Toward a theory of spontaneity in the dynamics of protest and social movements. American Sociological Review, 79, 1122–1143.Find this resource:

Sutton, J., Butcher, C. R., & Svennson, I. (2014). Explaining political jiu-jitsu: Institution-building and the outcomes of regime violence against unarmed protests. Journal of Peace Research, 51, 559–573.Find this resource:

Tarrow, S. (2011). Power in movement: Social movements, collective action and politics (3d ed.). Cambridge, U.K.: Cambridge University Press.Find this resource:

Terrorism Network Project (TNP). Principal Investigators: E. Chenoweth, M. Horowitz, & P. Potter. University of Denver. Retrieved from

Tilly, C., & Tarrow, S. (2015). Contentious politics. Oxford: Oxford University Press.Find this resource:


(1.) Data projects involving contention have also been expanded to include terrorist group behavior. The Terrorism Network Project (TNP) will collect data on alliance networks among terrorist groups in order to understand the diffusion of their tactics, the type of tactics that they use, and the lethality of their attacks (led by Erica Chenoweth at University Denver, Michael Horowitz at University of Pennsylvania and Philip Potter at the University of Michigan). Besides the data projects on political contention and terrorism, there is also the effort to collect data on government repression. For instance, the Government Actions in Terror Environments (GATE) data has information on state actions toward terrorist organizations or their constituencies for a small number of countries in the Middle East and South Asia from 1987–2004 (led by Erica Chenoweth at University of Denver and Laura Dugan at the University of Maryland. In addition, the Ill-Treatment & Torture (ITT) Dataset has information on the incidence of torture by country from 1995–2005 (led by Courtney Conrad at the University of California, Merced and Will Moore at Arizona State University).

(2.) Other new datasets are the Raleigh et al. (2010) Armed Conflict Location and Event Data Project on political violence and protests in Africa and Asia (ACLED), and Social Conflict in Africa Database (SCAD); Social, Political and Economic Event Database (SPEED); Government Actions in Terror Environments (GATE); and the forthcoming Mass Mobilizations in Autocracies Database(MMAD) by Weidmann and Rod (2013). The NAVCO 2.0 and MMAD datasets treat protest campaigns as the unit of analysis; in these other datasets, protests, demonstrations, riots, strikes and other forms of internal governmental and nongovernmental actions are treated as the units of analysis.

(3.) The exception to this is Chenoweth and Stephan (2011), who correlate various characteristics of protest campaigns with regime overthrow. Some of these correlates include the size, scale, and scope of dissident protests, as well as the variety of protest tactics, the severity of government repression, and the presence of elite defections.

(4.) These variables for protest campaigns between 1946 and 2006 can be found in the NAVCO 2.0 dataset (Chenoweth & Lewis, 2013).

(5.) For instance, the NAVCO 2.0 dataset establishes the threshold of a protest campaign at 1,000 observed participants and it ends when the number of participants decreases below that figure.

(6.) The Kefaya Movement is included in the NAVCO 2.0 dataset, but the labor strikes, especially the 2008 Mahalla labor strikes, were not included.

(7.) A similar dynamic occurred in the Egyptian case as footballers repeatedly clashed with security forces and patrolled perimeter areas of Tahrir Square in the early phase of the campaign.

(8.) A similar case can be made for the emergence of violence that follows on the heels of movement fragmentation (McLaughlin & Pearlman, 2012). Nonviolent protest campaigns are nonviolent largely because the social movement actors are cohesive with a singular aim. Yet the nonviolence may be short-lived if over the course of the campaign movement splintering or factionalism occurs, encouraging some groups to use extremist tactics. What might account for this dynamic? Governments may accommodate some groups with the goal of isolating and eliminating extremist groups. Or, concessions may by themselves splinter movements because some groups are willing to entertain political reforms, while others demand deeper, more radical reforms. Political violence is likely to emerge later over the course of the protest wave as governments try to adapt to the protesters’ demands. And, by its very nature, extreme political violence may turn off moderately oriented protesters who return home, thereby de-escalating the intensity and magnitude of the campaign. Again, the timing of violence, who perpetrates it, and its consequences require a more nuanced, dynamic analysis.

(9.) In yet still another example, Demirel-Pegg (2014) provides quantitative and qualitative evidence that the onset of the Kashmir civil war that started in 1989 was preceded by an escalation of conflict between governmental and nongovernmental actors that emerged out of a nonviolent protest wave of civil resistance. She demonstrates that the intersections of governmental repression, nongovernmental radicalization, and militarization contributed to the escalation.

(10.) The DOC approach along with key concepts and their definitions are obtained from two sources: McAdam, Tarrow, and Tilly (2001); and Tilly and Tarrow (2015).

(11.) Over these ten years, Mubarak increased the number of security departments and swelled the police force from 150,000 personnel in 1974 to more than a million in 2002, a 21 percent increase. By 2011, the number of police personnel was approximately two million (Kandil, 2012, p. 265).

(12.) Kandil (2012, p. 270) reports this figure and cites human rights organizations that document torture rituals among male and female detainees: they were blindfolded, handcuffed, stripped naked to their underwear, beaten with sticks or batons, and subjected to electric shocks and stress positions.

(13.) Ghonim (2012) posted the photos and video tape on a Facebook page which quickly amassed hundreds of thousands of readers and sympathizers. The January 25 day of anti-police protest was organized via this Facebook page.

(14.) For instance, during the US civil rights movement, challengers relied on black churches to aid mobilization and collective action. Similarly, in the Iranian Revolution, radical clerics used the mosques and their local networks to build antigovernment dissent.

(15.) See Lim (2012) for a fuller discussion of the role of the April 6th Youth Movement’s introduction of Facebook and Twitter in its efforts to mobilize street demonstrations in the period leading up to the January 25 Egyptian uprising. Also, see Ghonim (2012) for a personal view of the use of social media and the Egyptian revolt.

(16.) Rasler’s (1996) empirical analysis of the dynamics of the Iranian Revolution in 1979 demonstrates this pattern.

(17.) In early January, the core leadership of these groups chose 21 protest sites, usually connected to mosques and located in densely populated working-class neighborhoods around Cairo. Youth leaders scattered the protest sites in order to draw larger numbers to the rallies, overburden the security forces, and increase the likelihood that some of the protesters would be able to reach Tahrir Square.

(18.) Observers put the number of people demonstrating in Tahrir Square closer to two million. In addition, one million people were estimated to be demonstrating in Alexandria, 750,000 in Mansoura, and 250,000 in Suez (Naguib, 2011).

(19.) At the regime’s initiative, discussions began between the government and hand-picked groups, including the National Democratic Party, the official opposition political parties, and some representatives of the protesters and the Muslim Brotherhood. These initial discussions failed.

(20.) In Port Said, 5,000 shantytown dwellers camped in front of and then later stormed the governor’s headquarters in an effort to get information about their applications for public housing. In Suez, 50,000 people demonstrated along with local factory workers (El-Ghobashy, 2012, pp. 4–5).

(21.) Mubarak offered to transfer some of his presidential powers to his vice president and said that he would reform six and possibly more articles of the constitution.