Undermining Resistance

Abstract

This study examines attempts by authorities to undermine overt collective challenges, such as protests, riots, or armed attacks, by targeting activities that precede and/or support such behavior. After providing a theory of how repression and resistance develop, the study analyzes unique data drawn from the confidential records of the Guatemalan National Police to assess the use of repression during the years between 1975 and 1985. Empirical tests demonstrate that (1) government forces anticipate challenger development by identifying the mobilization activities nascent challengers rely on to initiate and sustain overt collective challenges and (2) the use of repression designed to undermine such efforts is specifically targeted against radical (i.e., highly transformative) claims making. Implications are drawn for how we understand and study political order and conflict.

Keywords

human rights repression social movements civil war Guatemala

How do governments enforce political order? Prevailing views correspond to a relatively simple sequence of events. Episodically, dissidents are able to overcome the internal barriers to collective action and engage in overt collective challenges, such as protests, strikes, riots, or armed attacks (Beyerlein and Hipp 2006; Gates 2002; McAdam 1986; McAdam et al. 2001; McCarthy and Zald 1977; Tarrow 1994). To protect political order, governments respond to overt collective challenges with repression (Carey 2010; Davenport 1995, 1996, 2007, 2010; Earl, Soule, and McCarthy 2003; Moore 2000; Pion-Berlin 1988; Poe and Tate 1994; Poe, Tate, and Camp-Keith 1999; Valentino, Huth, and Balch-Lindsay 2004).¹ Contestation then ensues as governments and challengers engage in open conflict (Cunningham, Gleditsch, and Salehyan 2009; Davenport 2014; Kalyvas 2006; Moore 1998; Pierskalla 2010; Ritter 2014).

Although useful in many ways, there are important limitations to this view. For example, while it is often suggested that a key reason challenges are not observed in a particular situation is because expectations of repression prevent dissidents from organizing (e.g., Ritter and Conrad 2012; Lacina 2014; Pierskalla 2010), existing theory provides few details on how governments might anticipate the development of overt collective challenges or how they attempt to undermine this behavior (e.g., Danneman and Ritter 2014; Herreros and Criado 2009; Nordas and Davenport 2013). Understanding how governments anticipate and/or undermine overt collective challenges is important because if authorities attempt to eliminate threats without exclusively relying upon repressive responses to overt activity, then models that fail to identify relevant state action have neglected an important channel through which governments enforce political order. Indeed, if government forces repress individuals who are hoping to challenge the regime and if they succeed at incapacitating the dissident’s organization, this form of “preemptive” repression would provide a novel mechanism for restraining threats. Under this circumstance, explanations for challenger emergence will have to address preemptive repression, while explanations for civil conflict will have to be amended to account for the selection of surviving movements into subsequent conflagrations.

This study provides insight into why repression is applied in anticipation of and in conjunction with overt collective challenges as well as when and where such repression can be expected. In addition to engaging with overt behavior, governments seek to subvert the most costly challenges by applying repression against mobilization (i.e., the formation and support of an opposition organization from which overt collective challenges can be initiated and sustained). Governments do not attempt to address all mobilization attempts, but focus their efforts on those mobilizing in support of highly transformative redistributions of political power. Upon observing “radical” mobilization, authorities repress to eliminate influential leaders, deplete organizational resources, and subvert overt challenges.

This study provides empirical evidence to support these claims by analyzing new data collected from the confidential records of the Guatemalan National Police. These records contain unprecedented details on the myriad of covert mobilization activities engaged in by challengers as well as the spectrum of repressive behavior employed by government forces. Analysis of the police data demonstrates that (1) government forces anticipate challenger development by identifying the mobilization activities nascent challengers rely on to initiate and sustain overt collective challenges and (2) the use of repression designed to undermine such efforts is specifically targeted against radical claims making.

While exploring these issues, the investigation demonstrates how misconceptions regarding the dynamics of political repression manifest themselves empirically as well as theoretically. By comparing results generated from the Guatemalan National Police records against data collected from international and Guatemalan newspapers, the analysis reveals how biases in newspaper data predispose analysts to the conclusion that governments only employ repression in response to overt collective challenges, while missing repression directed against more covert mobilization activities. The full range of repressive activity and useful insights into its application will only be revealed when new and better data are brought to bear.

The article proceeds as follows: First, I review the academic literature on political order and conflict. Second, a model of repressive action is presented that distinguishes repression targeting mobilization from repression directed against overt collective challenges. I then present the data, research design, and analysis. Finally, in the conclusion, I discuss the implications of the study for understanding political order and conflict.

Defending Political Order

Over the past few decades, research on political repression and human rights violations, social movements, and civil war has sought to explain how governments enforce political order. Each approach has had a somewhat different but generally overlapping understanding of the topic.

For instance, one body of work has emerged to focus on government coercion, such as political imprisonment, torture, disappearances, and mass killing (e.g., Carey 2006, 2010; Conrad and Moore 2010; Davenport 1995, 2007; Gartner and Regan 1996; Moore 2000; Pierskalla 2010; Poe and Tate 1994; Regan and Henderson 2002; Shellman 2006; Tilly 1978; Valentino, Huth, and Balch-Lindsay 2004). Often referred to as the “threat-response theory” (Earl, Soule, and McCarthy 2003), the dominant model expects governments to employ higher levels of repression in response to an increase in observed dissident threats.² The theory has as its underlying foundation a conceptualization of policy making in which governments respond to overt collective challenges to the regime by engaging in repressive behavior to control or eliminate challengers (e.g., Davenport 1995; Earl, Soule, and McCarthy 2003; Pion-Berlin 1989; Poe and Tate 1994; Poe, Tate, and Camp-Keith 1999). Recent studies are beginning to push in the direction of a more forward-looking model, but such work remains limited in part because the microfoundational mechanisms linking prospective decision making to the repression of specific events or individuals have yet to be well articulated (e.g., Danneman and Ritter 2014; Herreros and Criado 2009; Moore 1995; Nordas and Davenport 2013; Pierskalla 2010; Walter 2006).

At the same time, a rich literature on social movements has focused directly on the emergence of challenges to political order. Since Olson’s (1965) seminal study, it has generally been assumed that the internal barriers to collective action are so great that nonparticipation can be taken as a default position, rather than a facet of the existing interactions between governments and challengers. As a result, studies of social movements have tended to focus on factors influencing the capacity of organizations to overcome their internal collective action problems, while overlooking the competition for resources that occurs between governments and challengers during mobilization (e.g., Beyerlein and Hipp 2006; Ritter and Conrad 2012; Klandermans 1984; McCarthy and Zald 1977; Siegel 2009; Snow and Benford 1988). Where repression enters such research, causal explanations have centered on how unobservable expectations of repression hinder resolution to the collective action problem, rather than on any systematic and observable efforts authorities commit to anticipate and subvert the development of political challenges (e.g., Lacina 2014; Pierskalla 2010). And where scholars have examined the effects of repression on social movements empirically, research has not yet systematically addressed the effects of repressive behavior directed against activities other than overt collective challenges (e.g., Dugan and Chenoweth 2012; Davenport, Johnston, and Mueller 2005; Daxecker and Hess 2012; Rasler 1996).

Among scholars studying the emergence of large-scale conflicts between governments and challengers (i.e., civil wars), opportunity-based arguments stress how the ability of insurgents to challenge the government results in part from the limited coercive capacity of weak states (e.g., Fearon and Laitin 2003; Gleditsch and Ruggeri 2010). But such work sees state repression as largely structurally determined and only briefly references the strategic interplay between nascent insurgents and the government.³ By treating repression as a structure, opportunity-based theories marginalize government strategies to restrain dissent operating within the state’s structural constraints (McAdam et al. 2001, 43-45).

In sum, understandings of conflict could be improved with increased attention to government efforts to subvert the development of potential challenges. If governments anticipate and successfully undermine overt collective challenges without engaging this behavior directly, then existing theories will have to be revised in a number of important ways.⁴ First, prevailing models of political repression will have to be amended to recognize that government forecasts of future challenges are as important for inspiring repressive behavior as responses to ongoing threats. Second, theories of social movement mobilization will need to pay greater attention to understanding how movements struggle to outpace government attempts to undermine mobilization. Finally, rather than the dominant concern with structural variables theorized to influence challenger or state capacity, understandings of the emergence of large-scale conflict will need to pay greater attention to how the interaction of these two dynamic forces (repression and mobilization) impact prospects for conflict escalation. Should government forces successfully demobilize the opposition, an absence of overt collective challenges would occur not because movements could not organize internally, but because the government was able to outcompete challengers and reduce their capacity to engage in overt challenges. This will impact prospects for conflict escalation as well as the possibility for a change in political order.

From Mobilization to Repression

This section presents an argument for how and why governments employ repression designed to subvert the development of overt collective challenges. The argument begins with the assumption that governments wish to stay in power. They extract some rents from being in office and allocate concessions and repression in order to manage threats and maximize their net resources. In choosing to apply concessions or repression, governments consider the costs of these policy instruments and their impacts on both ongoing and future challenges.⁵

Moving beyond extant theory, this article argues that governments not only consider the presence of observable challenges (e.g., Davenport 1995; Earl, Soule, and McCarthy 2003; Valentino, Huth, and Balch-Lindsay 2004) or unobservable, latent probabilities of challenges (e.g., Danneman and Ritter 2014; Herreros and Criado 2009; Nordas and Davenport 2013) but aim to improve the specificity of their expectations by monitoring their citizenry to identify the observable indicators heralding the development of the most threatening overt collective challenges. Governments recognize that the outcome of repression targeting overt collective challenges is highly unpredictable (e.g., Davenport, Johnston, and Mueller 2005). At the same time, the repression of those not directly engaged in challenges is costly as it involves resource allocations, the possible sacrifice of human life, and potential political backlash. As a result, governments must balance the costs of repression designed to subvert challenger development against (1) the probability of a threat being realized into overt collective challenges and (2) the probable costs of dealing with those challenges, should they materialize.

To improve their forecasts of the likelihood and costs of future overt collective challenges, governments must consider two factors. First, they focus on the behaviors of potential challengers, seeking to identify the most relevant actions signaling challenger development. Second, they attend to the demands challengers are likely to express, should a challenge occur, seeking to identify the potential costs of conflict bargaining if overt collective challenges materialize. Subsequently, I present a model of challenger development in order to specify the behaviors and demands inspiring repressive efforts to subvert challengers. Based on this model, I develop hypotheses specifying where and when such repression can be expected.

Mobilization, Repression, and the Enforcement of Political Order

Political order is often conceptualized as a dyadic relationship between the powerful and the (potentially) compliant (e.g., Arendt 1970; Huntington 1968; Lukes 2005). However, it is perhaps better conceived as a triadic relationship between the powerful, the (potentially) compliant, and the (existent or nonexistent) organized opposition. This is the case because opportunities for challenging the regime vary to the extent that there is, in fact, a viable alternative to compliance. In the absence of an organized opposition, many who would otherwise defy government authority find compliance to be in their self-interest. Resistance can take the form of foot dragging or other “weapons of the weak,” but without a viable alternative to the existing power structure to align with, individuals often find open resistance personally impractical (Scott 1985, 1990).

At times, small bands of individuals are able to overcome such restraints and engage in “mobilization,” which describes the process of forming and supporting an organized alternative to the government through small, generally clandestine activities such as organizing meetings, articulating organizational objectives, distributing information, campaigning for funds, training members, and recruiting new participants (Gaventa 1982, 24-25; Tilly 1978, 54). Distinct from overt collective challenges, which are outwardly targeted at political authorities, mobilization is designed to draw individuals and resources into an organization in order to sustain a platform from which overt collective challenges can be planned, communicated, and directed (Chong 1991; McCarthy and Zald 1977).

Mobilization is difficult and requires extensive contributions from a small number of individuals (Olson 1965). These organizers must be willing to overcome incentives to comply with the regime and act as first movers in a potential challenging organization (Oliver 1984). Such actions are demanding both because they require personal contributions to a highly uncertain effort and because early action leaves organizers vulnerable to identification by the state. But the contributions of core members are critical because organizations cannot inspire others to participate in overt collective challenges without first reaching the minimal threshold of mobilized resources necessary to express collective demands and coordinate collective behavior (Marwell and Oliver 1993).

From the core group of organizers and the resources they mobilize, overt collective action can spring up, be sustained, and potentially grow. Overt collective challenges have the potential to set in motion cascades of participation (Granovetter 1978). As a result, challenges have the capacity to escalate rapidly from seemingly insignificant degrees to levels that threaten the regime (Kuran 1989). However, cascades of participation cannot proceed unless organizers are able to mobilize sufficient resources to spark and sustain such behavior.

Governments recognize how significant mobilization is for inspiring and sustaining overt collective challenges and translate their beliefs about mobilization into expectations for future threats. Prior experience dealing with challengers along with evidence from neighboring states provides the background necessary to formulate expectations for the development of challenges to political order (e.g., Moore 1995; Danneman and Ritter 2014; Shellman 2006). Sophisticated surveillance apparatuses are established to monitor social behavior and detect movements toward overt collective challenges (e.g., Lewis 2012). Governments look to identify behaviors such as meetings, seminars, fund-raising, and recruitment drives, which typically take place outside of the direct observation of the government and do not directly threaten its authority but involve the coordination of individuals organizing in support of collective challenges.

Where mobilization is identified, it presents clear incentives for government forces to intervene with repression.⁶ Allowing such behavior to progress means accepting the possibility that overt collective challenges can be coordinated and sustained, which means risking open contestation with challengers. Upon observing mobilization, governments can direct repression against organizers in an effort to drain the mobilizing organizations of critical individuals and deplete the resources available to inspire and sustain overt collective challenges.

Hypothesis 1: Mobilization is expected to be related to increased political repression.

In deciding to repress mobilization, governments must not only consider the probability that overt collective challenges develop but also weigh the costs of coercion against the predicted costs of challenges. Repression of this sort is costly, and government forces cannot reasonably be assumed to have the capacity to repress all forms of mobilization. Yet at the same time, not all forms of mobilization are equally threatening and worthy of government attention.

The anticipated demands of potential challengers are critical in this regard, as challenger demands directly impact the probability of settlement and thus the expected costs of conflict. Of greatest concern for political authorities is those mobilizing in support of “radical” transformations in political power, such as overthrowing existing elites or revolutionizing the rules governing the polity (Gamson 1975; Gartner and Regan 1996; Tilly 1978). If challenges are expected to emerge, but dissidents are expected to make demands that fall short of radical redistributions of power, governments have the opportunity to negotiate with challengers and retain power at minimal costs.⁷ Conflict may still emerge as each side faces asymmetric information on their opponent’s capacities or resolve. However, as long as there exists at least one agreement that both parties prefer to fighting, the two groups will eventually converge on a settlement (Walter 2009).

But if challengers are expected to demand radical transformations in power, conflict bargaining can be expected to center on issues considered indivisible by governments and challengers. Challengers are likely to demand that political authorities step down as a necessary condition to any settlement, which puts the government in an all-or-nothing position to defend its status.⁸ In this case, there are no agreeable settlements that are mutually preferable to conflict, and the government must fight until it is ousted or until it achieves a decisive military victory. As a consequence, conflicts over radical demands tend to be longer, more costly, and more difficult to resolve (Thomas 2012; Wood 2005; Wucherpfennig et al. 2012).

Because mobilization supporting radical ideals has the potential to spark and sustain more protracted and costly overt collective challenges, government forces weigh the utility of repressing this form of mobilization differently from repressing nonradical mobilization. In this case, governments attempt to suppress the development of the most costly challenges by targeting radical mobilization with repression.

Hypothesis 2a: Increases in political repression are expected to take place when mobilizers support radical redistributions of political power.

Hypothesis 2b: Increases in political repression are not expected to take place when mobilizers support less radical demands.

Data

Identifying the repression of mobilization requires data that track the coercive actions of government forces as well as the covert and overt activities of dissidents. To acquire such information, I created an events database using records found in a unique archive of confidential police documents, the Archivo Histórico de la Policía Nacional (AHPN), which was produced by the Guatemalan National Police between 1975 and 1985 (Morales Alvarado 2009; Weld 2014). During most of the twentieth century, the Guatemalan National Police employed a central depository to store the records produced by the inner workings of their bureaucracy (e.g., memos passed up and down the chain of command, arrest records, log files summarizing daily activity, and investigative reports produced by local divisions or other specialized units). Following the signing of the peace accord and the disbanding of the national police in 1996, this warehouse and the millions of documents it contains were abandoned. After this time, the documents lingered on the outskirts of Guatemala City for approximately ten years before being discovered (Doyle 2007; Smith 2009). Following a lengthy legal struggle, the decision was made to clean and index the documents into the newly formed AHPN. The digitization and archiving process has recently been completed, and this study is one of the first to have been granted access to the archive’s full 1975–1985 collection.

Records at the AHPN were archived following the administrative structure of the national police, with each file indexed according to the office that received that particular memo or report. The national police had jurisdiction over the entire country and were divided into ten cuerpos (divisions) based on geographic location. Three additional divisions handled administrative tasks, major criminal investigations, and joint operations with the military, respectively. The director general oversaw the entire force.

A purposefully selected subset of the Archive’s collection was coded as an events database of state and dissident actions. The subset is composed of the entire selection of records from two offices, the Director General of the Police and the Commissioner for Coordinating Joint Operations with the Military.⁹ The resulting data set was constructed by reading all 300,000+ records sent to the two offices and preserved in the AHPN. All forms of political activity were coded as discrete events.¹⁰ In total, the project coded more than 7,000 acts of state and dissident behavior.¹¹

Because the AHPN records were constructed for bureaucratic performance rather than for public dissemination and were released without oversight by the agency responsible for producing the documents, the Archivo contains one of the most comprehensive and unbiased collections of information on the inner workings of the state’s repressive apparatus identified to date.¹² Still, data taken from police records are not without problems, and it is important to recognize that all sources of conflict data are likely to present incomplete and potentially biased representations of political events.¹³ With data from the police, one needs to be conscious of parochial incentives that could lead to biased reporting for professional gains. For example, with reference to the reporting of repression, given the Carter administration’s emphasis on human rights, it is clear that the regime was consciously trying to improve its international human rights reputation by concealing evidence of massacres (Doyle 1999). While the records were only accessible to those within upper levels of the police, it is still possible that they contain some bias against the reporting of atrocities. Moderating the scope of repression could bias the study against identifying a relationship between mobilization and repression.

However, with regard to the reporting of challenger behavior, the direction of bias is less clear. Organizations might develop an interest in overestimating the threat of the movement in order to increase their budget (Stanley 1996). Alternatively, they might develop an interest in downplaying movement behavior to demonstrate professional success (Reiner 2010). What is evident is that because the Guatemalan government had significantly more resources than did news organizations and because it had extensive networks designed to identify political activities, the AHPN provides better details on a far larger spectrum of political activity than other sources of information on the Guatemala conflict.

Indeed, because they contain a more extensive list of political events, government records have been used widely to help identify the form and extent of bias in newspaper coverage (e.g., Barranco and Wisler 1999; McCarthy et al. 2008; McCarthy, McPhail, and Smith 1996; Oliver and Myers 1999; Weidmann, 2015). With regard to the present topic of study, one of the factors contributing to the empirical validation of threat-response theories of political repression may be that employing traditional newspaper-based data to study the topic has biased existing understandings of when political repression is applied. While part of the difficulty in studying the repression of radical mobilization has been conceptual—scholars studying political order have focused their theories on the most overt forms of behavior—there are corresponding empirical challenges, as these overt forms of behavior are also the activities most commonly identified in events databases. If we are more aware of larger acts of challenger activity, then when it is time to articulate where and when states repress, larger and more intense incidents of overt collective challenges will be more easily conceptualized as catalysts (e.g., Danzger 1975; Drakos and Gofas 2006; Myers and Caniglia 2004). And when newspaper data are used to evaluate theoretical claims, the omission of smaller, nonviolent events from events’ databases can result in spurious correlations consistent with such theories (e.g., Barranco and Wisler 1999; Davenport and Ball 2002; Earl et al. 2004; Snyder and Kelly 1977). At the same time, description biases in newspaper coverage of contentious politics can lead to the assumption that governments respond to “public disturbances” only in self-defense (e.g., Davenport 2010; Smith et al. 2001).

To empirically evaluate how the data sources employed in the analysis of political order and conflict influence understandings of state repression, this study replicates its analysis on a separate database of political activities reported in Guatemalan and international newspapers. The expectation is that the relationship between mobilization and repression will be apparent in the analysis of the AHPN data, but we are not likely to observe this relationship using data gathered from newspapers. The newspaper data for this replication were compiled from several sources. The first is a database of contentious events reported in the Guatemalan press, which was generated by Brockett (2005) for a study on repression and dissent in Central America.¹⁴ Brockett’s data collection effort involved hand coding the largest local and national daily newspapers in Guatemala using a system for measuring government and dissident behavior that is consistent with the protocol used at the AHPN. I then supplemented Brockett’s data with my own coding of political activity in Guatemala reported in the New York Times as well as three Guatemalan newspapers, Prensa Libre, El Diario, and La Hora, using the same coding system as applied at the AHPN. When combined, the newspaper data set identifies more than 2,000 incidents of political activity.¹⁵

Repression and Dissent in Guatemala, 1975–1985

The period under investigation was an extremely turbulent time in Guatemalan history. Conflict occurred between a variety of leftist social movement organizations (including unions, peasant cooperatives, land rights groups, Catholic activists, students, urban revolutionaries, and Marxist insurgents) and a fractured military government aligned with the landowning and capitalist classes (Archdiocese of Guatemala 1999; Ball, Kobrak, and Spirer 1999; Brockett 2005; Carmack 1992; Garrard-Burnett 2011; Manz 2005; Schirmer 1998; Stoll 1993).

Guatemala’s security forces were both highly professional and relatively unconstrained in their use of coercion (Schirmer 1998). Over the course of the ten years under review, the Guatemalan government was responsible for killing more than 100,000 of its citizens (Ball, Kobrak, and Spirer 1999). Yet at the same time, there were periods of peace, both in the time leading up to and away from violence and in locations unaffected by the fighting (Gulden 2002; Stoll 1993). While any single country study faces difficulties of generalizability, looking at Guatemala provides a unique laboratory for examining an extremely broad spectrum of dissident behavior and related repression, while holding relatively constant other important factors thought to influence these behaviors. Future work will be necessary to examine the dynamics of mobilization and repression in more stable contexts.

Political conflict emerged at different points in time and with varying intensity depending on where in the country one was located, and this variation is captured by the AHPN data. In some departments, such as Guatemala City or Escuintla, the AHPN records identify conflict behavior during the first few months under review. On the other end of the spectrum lies El Petén and Zacapa, two departments that appear to have been spared from the worst forms of violence. Once initiated, mobilization, overt collective challenges, and repression each ebbed and flowed in each department as challengers attempted to press claims against authorities and the government engaged in coercion to suppress such efforts (Brockett 2005; Stoll 1993; Sullivan 2012, 2014).

Prior to 1977, a date often used to mark the onset of the civil war (e.g., Brockett 2005), mobilization and sporadic protests were organized by student organizations and labor unions operating in the capital. Acts of terror waged by urban revolutionaries and insurgent assassinations of large landowners have also been documented during this period (Brockett 2005; Garrard-Burnett 2011; Stoll 1993), while in the highlands, nascent insurgent organizations were engaged in their initial mobilization attempts (Archdiocese of Guatemala 1999). On the side of the government, surveillance of social activity, arrests of suspected dissident leaders without charge, killings, disappearances, and the application of torture all appear to have been common (Archdiocese of Guatemala 1999; Ball, Kobrak, and Spirer 1999; Carmack 1992; Commission for Historical Clarification [CEH] 1999; Falla 1994; Manz 2005; Perera 1993; Schirmer 1998).

As the violence moved into the rural highlands (beginning in the late 1970s), the available indicators document a sharp increase in repression (Ball, Kobrak, and Spirer 1999; CEH 1999). Human rights organizations identify how rampant abuse occurred during this period; tens of thousands were killed and many more displaced from their homes (Archdiocese of Guatemala 1999; Ball, Kobrak, and Spirer 1999; CEH 1999; Sanford 2004). The strong overlap between identified political conflict and racial segregation led to charges of acts of genocide levied against the government (CEH 1999).

As government forces reestablished control throughout the country, estimates of insurgent mobilization subsequently dropped rapidly, from a high of more than 3,000 troops in 1982 to fewer than 300 in 1984 (Central Intelligence Agency estimates from Doyle 1999). Measures of repression similarly decreased steadily in 1983 and 1984 (Ball, Kobrak, and Spirer 1999). With the imposition of constitutional rule in 1985, it appears that most of the violence had ended (Schirmer 1998).

Operationalization

For the purposes of this study, mobilization is operationalized as small, generally clandestine efforts to influence individuals affiliated or unaffiliated with a social movement to participate in collective action. Examples include mobilization by radical student and labor organizations, separatist groups, and Marxist insurgent organizations. Mobilization is distinct from Overt Collective Challenges, which are operationalized as public, outwardly focused demonstrations of organizational strength targeted at political authorities. Examples include strikes, demonstrations, marches, roadblocks, targeted killings, arson, kidnapping, and the taking of hostages.¹⁶

Mobilization was further divided into two subtypes based on the claims expressed by the organizations responsible for coordinating these activities. Radical mobilization is operationalized as those activities carried out in support of expressed demands for overthrowing or displacing the political system.¹⁷ Practically, this list includes all mobilization activities engaged in to support the overthrow of the government and/or division of the state’s territorial integrity. Examples include radical student and labor organizations, separatist groups, and Marxist insurgent organizations.¹⁸ Nonradical mobilization is operationalized as those events carried out in support of demands that do not directly call for overthrowing the political system and/or a division of the state’s territorial integrity. Examples of nonradical mobilization include mobilization to support ideals and demands such as community development, land rights, and freedom from persecution.

Empirically, there are nearly ten times as many acts of mobilization captured in the AHPN data (751 events) compared to the data collected from the news media (88 events). Of these, roughly 80 percent of the mobilization events recorded in the AHPN data are categorized as pursuant of radical goals, which is slightly higher than the comparable figure for the newspaper data. This descriptive evidence is consistent with arguments that (1) government strategies are far more concerned with mobilization attempts than is often acknowledged and (2) governments are more likely to devote attention to mobilization when such activity pursues radical demands.

Finally, political repression is operationalized as the politically motivated use of force committed by representatives of the state against individuals under their political jurisdiction. Consistent with the articulation expressed in the theoretical arguments, this operationalization encapsulates both overt forms of repression (e.g., death threats, torture, disappearances, shootings, raids, protest policing, and politically motivated arrests) and more covert activities (such as surveillance, wiretapping, informing other officers, initiating an investigation, and drafting security plans). The AHPN data document more than 2,500 repressive actions committed during this time period, while the news material includes reports on less than a quarter of that figure.

For the analysis, both the AHPN-data and the newspaper-data event catalogs were transformed into a cross-sectional time series of department months. For each of Guatemala’s twenty-two departments, the time series identifies monthly event counts of each specific form of activity (as reported in the two data sets) between January 1975 and December 1985.¹⁹

Analysis

In line with the above-mentioned arguments, it is expected that repression is related to challenger mobilization and that increased repression can be expected to follow mobilization in support of radical ideals. To evaluate these contentions, the analysis identifies factors related to the use of political repression using a series of econometric analyses.

The methods most commonly employed to identify the causes of political repression are cross-sectional time-series models examining correlations between lagged measures of dissent and subsequent levels of repressive activity (e.g., Davenport 1995, 2007; Poe and Tate 1994). A similar method is employed here. The functional form is specified as negative binomial due to the nature of the dependent variables, which are count data that display over dispersion (Winkleman 2008).

As designed, the empirical analysis first investigates the relationship between all forms of mobilization and political repression before examining how the repression of mobilization might be conditional on the radical or nonradical ideals around which groups are mobilizing. The aim here is to identify whether mobilization is having an independent effect on decisions to engage in political repression, while controlling for the ebb and flow of overt collective challenges. The models also include a number of additional controls out of concern that they might influence both mobilization and ensuing repression. These controls can be divided roughly into two categories. The first includes structural characteristics that are conceptualized as relatively static or slow-moving conditions that influence the underlying expectations for mobilization and repression. Included in this category are a department’s literacy rate, indigenous population rate, population size, and population density, all calculated using census measures from 1973. Also included in this category are measures of democracy, measured annually using the Coppedge, Alvarez, and Maldonado (2008) standardized inclusion metric.²⁰ Literacy rates are taken as an indicator of the capacity of the state in a given territory, which is likely to influence both decisions to mobilize against the government and the government’s use of repression (e.g., Fearon and Laitin 2003). Indigenous population rates are included to control for the strong ethnic divisions that categorized the Guatemalan conflict (e.g., Perera 1993). Population size is included as this variable can be expected to influence the scale of dissident and government activities (e.g., Raleigh and Hegre 2009).²¹ Population density is provided as an indicator of the degree of urbanization in a territory and thus proxies the department’s relative level of economic development (e.g., Collier and Hoeffler 2004).²² Finally, democracy provides an indicator of the underlying regime structure, which can provide an indicator for national-level political opportunities as well as constraints on repressive behavior (e.g., Davenport 2007). This variable is measured annually and does not vary across departments.

A second category of controls is included to capture important behavioral factors likely to impact decisions to mobilize against the government or commit repression. The first is a measure of prior repression, which is recorded as the number of repressive actions occurring in a department during the previous month. The second behavioral control is a spatial lag of repression, which is measured as the number of repressive actions occurring in adjacent departments (weighted by the number of adjacent departments). This measure is also lagged one period and is included to control for the spatial clustering of repressive behavior (e.g., Danneman and Ritter 2014). An additional measure is included to capture variation in the overall intensity of violence in the department. The variable, department violence, is a dummy variable measuring whether ten or more individuals were killed by any force (or a combination of forces) in a department during the previous month.²³ Finally, the models include a control for campaign activity, which provides a more disaggregated measure of democratic participation. Campaign activity is measured as an event count of public actions taken by representatives of political parties. It is lagged one month for each department.

In selecting the control variables, effort was made to identify observable confounding factors that could theoretically bias the analysis. It is also important to control for those confounding factors that are potentially unobservable. The approach adopted to account for such factors is to replicate the analyses using a series of unconditional fixed effects (FEs) models (Allison and Waterman 2002; Angrist and Pischke 2009). One set of these models includes departmental FE dummy variables to account for any unobserved department-specific factors that could influence observations of both mobilization and repression. The second set of models includes monthly FE dummies to account for any temporal trends common both to repression and to dissent.

The results of the first set of analyses are displayed in Table 1. Models 1 through 3 present the results of analyses estimating the effects of all forms of mobilization on repression. Models 4 through 6 replicate this analysis, except that they divide mobilization activities into radical and nonradical subtypes as specified earlier. To aid in interpretation, the table presents the incident rate ratio (IRR) for each coefficient along with the standard error and level of statistical significance. An IRR above one represents a percentage increase in predicted counts of repression, while an IRR below one indicates a percentage decline in repression.

Table 1.

Evidence of the Causes of Repressive Action: Negative Binomial Regression Models using the AHPN Data.

	All Mobilization			Radical and Nonradical Mobilization
	(1)	(2)	(3)	(4)	(5)	(6)
Mobilization _t ₋₁	1.150	1.088	1.080
	[.041]	[.026]	[.035]
	(.001)	(.001)	(.016)
Radical mobilization _t ₋₁				1.156	1.099	1.064
				[.054]	[.034]	[.030]
				(.002)	(.002)	(.030)
Nonradical mobilization_t−1				1.115	1.009	1.134
				[.200]	[.106]	[.134]
				(.542)	(.927)	(.284)
Overt collective challenges_t−1	1.272	1.147	1.118	1.272	1.146	1.083
	[.055]	[.031]	[.034]	[.055]	[.031]	[.021]
	(.001)	(.001)	(.001)	(.001)	(.001)	(.002)
Prior repression _t ₋₁	1.069	1.030	1.069	1.070	1.030	1.069
	[.029]	[.017]	[.024]	[.029]	[.018]	[.021]
	(.015)	(.077)	(.004)	(.015)	(.083)	(.001)
Spatial lag of repression _t ₋₁	1.126	1.186	1.019	1.126	1.185	1.053
	[.022]	[.027]	[.020]	[.023]	[.027]	[.020]
	(.001)	(.001)	(.331)	(.001)	(.001)	(.007)
Campaign activity_t−1	1.380	1.323	1.039	1.384	1.331	1.040
	[.183]	[.165]	[.118]	[.188]	[.168]	[.112]
	(.015)	(.025)	(.735)	(.017)	(.023)	(.713)
Population	765.361		6,531.28	775.919		929.458
	[801.221]		[6371.94]	[818.151]		[876.158]
	(.001)		(.001)	(.001)		(.001)
Population density	1.009		1.041	1.010		1.992
	[.061]		[.061]	[.062]		[.185]
	(.871)		(.489)	(.866)		(.001)
Literacy rate	76.739		317.033	76.991		3470.26
	[61.375]		[222.690]	[61.677]		[2658.04]
	(.001)		(.001)	(.001)		(.001)
Indigenous population rate	7.159		8.474	7.153		18.896
	[2.229]		[2.134]	[2.228]		[4.909]
	(.001)		(.001)	(.001)		(.001)
Department violence_t−1	2.998	2.611	2.887	2.999	2.605	3.808
	[1.306]	[.846]	[.509]	[1.302]	[.845]	[.726]
	(.011)	(.025)	(.001)	(.011)	(.003)	(.001)
Democracy	.799	.792		.799	.791
	[.035]	[.034]		[.035]	[.034]
	(.001)	(.001)		(.001)	(.001)
Constant	.000			.000
	[.000]			[.000]
	(.001)			(.001)
Department FE		Y			Y
Month FE			Y			Y
N	2,420	2,420	2,420	2,420	2,420	2,420

Note: IRR, [Huber–White standard error], and (two-tailed p value) reported. The IRR is calculated as the rate of predicted counts of repressive actions when a variable is increased one unit over the rate of predicted counts when that variable is held at zero. All other variables are held at their means. FE omitted from presentation. AHPN = Archivo Histórico de la Policía Nacional; IRR = incident rate ratio; FE = fixed effects.

Looking at the control variables’ IRRs and levels of statistical significance, the results are largely consistent with the existing work. Most notably, overt collective challenges are strongly and positively associated with increased repression. The analysis thus supports the conventional argument that governments respond to ongoing challenges by increasing repression. Also consistent with existing work, repression is strongly and positively associated with prior repressive behavior and negatively and significantly related to the existence of democratic institutions.

What is uniquely revealed through the analysis is the relationship between mobilization and repression. Results disclose that government forces systematically increase repression as a function of the covert mobilization activities that inspire and sustain overt behavior. This evidence does not contradict the conventional wisdom that overt behavioral threats prompt repression, but it reveals that the conventional understanding of the topic only captures part of the story, missing the less-overt aspects of contention. A single mobilizing activity is predicted in model 1 to increase the amount of repression in the department during the next month by around 15 percent. And a standard deviation increase in mobilization is predicted to result in an average of approximately 1.5 additional acts of repression. This effect is comparable to the substantive impact of overt collective challenges, in which a standard deviation increase in challenges increases the expected number of repressive actions by just over 1.5 events. This suggests that governments do not simply respond to overt collective challenges, but that they also attempt to anticipate and undermine challengers by identifying and repressing the mobilization activities that inspire and sustain overt behavior.

But not all forms of mobilization are robustly related to repression. As expected, mobilization supporting radical redistributions of political power is related to positive and statistically significant increases in repressive activity. A single radical mobilization event is predicted to increase repression by 16 percent in model 4. The estimated effect is substantial when one considers that, in sites radical mobilization occurred, an average of three radical mobilization events were observed. And where a department experienced three instances of radical mobilization in a month, it is predicted to be associated with, on average, one additional act of repression.²⁴

Nonradical mobilization, by contrast, is not significantly related to repression in any of the specifications. In other words, government forces do not always respond to mobilization with repression. Instead, increases in repression are specifically related to the ideals around which organizers are mobilizing. This suggests that government forces aim to anticipate and/or suppress the most costly challenges by increasing repression where mobilization supports of radical transformations in political power.

Robustness Checks

An extensive set of robustness checks can be found in the Supplementary Appendix.

A set of robustness checks analyzes the relationship between mobilization and subcategories of overt and covert forms of political repression. Additional tests also probe the sensitivity of the analysis to a division of overt collective challenges into radical and nonradical subcategories.

Another set of checks delves into the temporal dynamics at operating within the analysis. First, contemporaneous overt collective challenges are included as a control along with lagged overt collective challenges. Second, the main results are replicated while excluding all department months jointly experiencing mobilization and overt collective challenges. Finally, the analysis is limited to “preconflict” periods preceding the first instance of overt collective challenges observed in each department.

Another set of tests investigates alternative strategies for estimating conditional and hybrid FE specifications within a negative binomial framework.

A fourth set of robustness checks utilizes a series of instrumental variables models to examine the effect of exogenous variation in mobilization on repression. These analyses exploit the time series dynamics operating within the data to instrument seemingly exogenous variation in different forms of mobilization.

Finally, the analysis checks the robustness of the findings to a model in which governments are forward-looking in their application of repression. This replication employs a set of dynamic time series vector-error-corrected models.

Across all of the specifications, radical mobilization proves significantly and positively related to increased repression. Combined, the results of these tests provide strong evidence affirming the relationship between increased repression and mobilization supporting radical ideals.

Replication with Newspaper Data

In addition to the theoretical limits recognized in the literature, there are issues related to the data employed in the empirical analysis of the topic that may have contributed to the dominance of threat-response theories of political repression within previous work. To explore this idea, Table 2 replicates the analysis from Table 1 using data taken from local and international newspapers.

Table 2.

Replication: Negative Binomial Regression Models using Newspaper Data.

	All Mobilization			Radical and Nonradical Mobilization
	(1)	(2)	(3)	(4)	(5)	(6)
Mobilization _t ₋₁	1.281	.926	1.339
	[.467]	[.319]	[.544]
	(.498)	(.824)	(.472)
Radical mobilization _t ₋₁				1.416	1.017	1.541
				[.519]	[.351]	[.653]
				(.343)	(.959)	(.307)
Nonradical mobilization_t−1				.000	.000	.000
				[.000]	[.000]	[.000]
				(.001)	(.001)	(.001)
Overt collective challenges_t−1	1.171	1.055	1.247	1.169	1.052	1.242
	[.063]	[.036]	[.123]	[.063]	[.037]	[.124]
	(.003)	(.114)	(.025)	(.004)	(.145)	(.031)
Prior repression _t ₋₁	1.570	1.120	1.277	1.559	1.107	1.277
	[.356]	[.151]	[.291]	[.360]	[.148]	[.293]
	(.047)	(.400)	(.284)	(.055)	(.450)	(.289)
Spatial lag of repression _t ₋₁	1.425	1.771	1.331	1.420	1.765	1.319
	[.712]	[.889]	[.710]	[.708]	[.884]	[.708]
	(.478)	(.255)	(.593)	(.482)	(.257)	(.606)
Campaign activity_t−1	1.207	.908	.854	1.186	.884	.828
	[.386]	[.325]	[.344]	[.383]	[.313]	[.338]
	(.557)	(.790)	(.696)	(.596)	(.729)	(.645)
Population	699,305		2,179,671	730,862		2,605,597
	[1,871,161]		[5,795,930]	[1,958,924]		[6,965,757]
	(.001)		(.001)	(.001)		(.001)
Population density	.765		.760	.765		.760
	[.109]		[.106]	[.109]		[.106]
	(.060)		(.050)	(.060)		(.049)
Literacy rate	605.105		768.053	625.355		865.463
	[1,307.46]		[1,551.782]	[1,353.67]		[1,757.78]
	(.003)		(.001)	(.003)		(.001)
Indigenous population rate	12.442		13.832	12.561		14.220
	[11.829]		[11.781]	[11.923]		[12.111]
	(.008)		(.002)	(.003)		(.002)
Department violence_t−1	5.687	2.439	5.454	5.670	2.413	5.582
	[2.768]	[1.144]	[2.227]	[2.774]	[1.130]	[2.302]
	(.001)	(.057)	(.001)	(.003)	(.060)	(.001)
Democracy	.712	.740		.712	.739
	[.070]	[.074]		[.070]	[.074]
	(.001)	(.003)		(.001)	(.003)
Constant	.000			.000
	[.000]			[.000]
	(.001)			(.001)
Department FE		Y			Y
Month FE			Y			Y
N	2,420	2,420	2,420	2,420	2,420	2,420

Note: IRR, [Huber–White standard error], and (two-tailed p value) reported. The IRR is calculated as the rate of predicted counts of repressive actions when a variable is increased one unit over the rate of predicted counts when that variable is held at zero. All other variables are held at their means. FEs omitted from presentation. IRR = incident rate ratio; FE = fixed effects.

In the analyses, overt collective challenges remain related to increases in the subsequent application of repression, while dissident mobilizing activities either are not significant or predict decreased repression. A single additional act of overt collective challenges is associated with around a 17 percent increase in the predicted rate of subsequent repression, a slightly lower figure than the estimate generated using the AHPN data. p values remain at or near conventional standards of statistical significance. At the same time, the newspaper data do not show any significant relationships between repression and aggregate measures of mobilization or the subset of radical mobilization activities. Nonradical mobilization is negatively related to subsequent levels of repression. This effect is highly significant and is expected to reduce repression substantially. However, care needs to be taken when interpreting the impact of this coefficient as the newspapers combined record only thirteen acts of nonradical mobilization.

When combined with the earlier results, this replication analysis shines light on significant limitations in analyses of repression and dissent conducted using newspaper data. When news sources are used in the analysis, the robust correlations between radical mobilization and repression disappear. Analyzed in this fashion, the results would thus fail to identify efforts by government forces to repress dissidents before they can engage in overt collective challenges.

Conclusion

This study has argued that the range of repressive activity employed to protect political order is significantly broader than the conceptualization found in much of the literature on political conflict. While it is often suggested that dissidents’ fear of repression deters them from challenging the state, existing research commonly portrays the government as committing repression only in response to some ongoing challenge. Instead, the study argues that in addition to responding to ongoing challenges, governments aim to anticipate challenger development and direct repression against those mobilizing in support of radical redistributions of political power. Analysis of unique data on the behavior of the Guatemalan security apparatus confirms these expectations and reveals how government forces anticipate and attempt to undermine overt collective challenges by repressing radical mobilization.

Future research will need to examine these relationships in more democratic or less violent contexts than those examined in this study. The key to doing this successfully will be to collect new and better data on the behavior of government forces and on those who would hope to challenge them. The relationship between mobilization and political repression was obscured when the analysis was replicated using newspaper data, which suggest that the news media incorrectly portray governments as disinclined to commit “preemptive” acts of repression. Though it is rare to find data that are comparable in scope and quality to the documents employed in this study, recent efforts to collect conflict data from unconventional sources have the potential to open up new avenues for studying repression and mobilization (e.g., Davenport, 2014; King, Pan, and Roberts 2013).

If the relationship between mobilization and repression does hold, research will need to more thoroughly examine the empirical effects of this type of coercion. Combined with the results of this study, evidence that repression targeting mobilization effectively limits subsequent dissent should inspire major implications for theories of political order and conflict. Dominant models of repression will need to incorporate signals about future contestation into government decision calculi, while social movement theories will have to be amended to consider how organizers attempt to outpace government efforts to suppress mobilization. Finally, by advancing understandings of early contestation between governments and nascent challengers, it could be possible to develop more dynamic models of repression and mobilization that can push beyond the existing structural understandings of civil conflict to improve forecasts for when such encounters are likely to escalate to widespread violence.

Footnotes

Acknowledgments

I would like to thank the following people for comments and suggestions: Mark Beissinger, Charles Brockett, Courtenay Conrad, Christian Davenport, Javier Osorio, Rory McVeigh, Emily Ritter, Naunihal Sigh, Allan Stam, Kiyoteru Tsusui, Elizabeth Wood, two anonymous reviewers, and the JCR editor. I also thank Alfredo Garcia and Mario Muralles for their tireless research assistance.

Declaration of Conflicting Interests

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

Funding

The authors disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: Yale University Order, Conflict, and Violence Fellowship; National Science Foundation Doctoral Dissertation Research Improvement Grant; Social Science Research Council International Dissertation Research Fellowship; Kellogg Institute Dissertation Research Grant.

Supplemental Material

The online [appendices/data supplements/etc.] are available at .

Notes

References

Allison

Paul D.

Waterman

Richard

. 2002. “Fixed effects negative binomial regression models.” Sociological Methodology 32 (1): 247–65.

Angrist

Josh

Pischke

Jorn-Steffen

. 2009. Mostly Harmless Econometrics. Princeton, NJ: Princeton University Press.

Archdiocese of Guatemala. 1999. Guatemala: Never Again. New York: Orbis Books.

Arendt

Hannah

. 1970. On Violence. Orlando, FL: Harcart Brace and Company.

Ball

Patrick

Kobrak

Paul

Spirer

Herbert

. 1999. State Violence in Guatemala, 1960–1996: A Quantitative Reflection. Washington, DC: AAAS.

Barranco

José

Wisler

Dominique

. 1999. “Validity and Systematicity of Newspaper Data in Event Analysis.” European Sociological Review 15 (3): 301–22.

Beyerlein

Kraig

Hipp

John

. 2006. “A Two-stage Model for a Two-stage Process: How Biographical Availability Matters for Social Movement Participation.” Mobilization 11 (3): 299–320.

Brockett

Charles D

. 2005. Political Movements and Violence in Central America. Cambridge: Cambridge University Press.

Bueno de Mesquita

Bruce

Smith

Alistair

Silverson

Randolph M.

Morrow

James D.

. 2003. The Logic of Political Survival. Cambridge, MA: MIT Press.

10.

De Mesquita

Bruce Bueno

Smith

Alastair

. 2010. “Leader Survival, Revolutions, and the Nature of Government Finance.” American Journal of Political Science 54 (4): 936–50.

11.

Carey

Sabine

. 2006. “The Dynamic Relationship between Protest and Repression.” Political Research Quarterly 59 (1): 1–11

12.

Carey

Sabine

. 2010. “The Use of Political Repression in Response to Domestic Dissent.” Political Studies 58 (1): 167–78.

13.

Carmack

Robert

, ed. 1992. Harvest of Violence: The Mayan Indians and the Guatemalan Crisis. Norman: University of Oklahoma Press.

14.

Chong

Dennis

. 1991. Collective Action and the Civil Rights Movement. Chicago: University of Chicago Press.

15.

Collier

Paul

Hoeffler

Anke

. 2004. “Greed and Grievance in Civil War.” Oxford Economic Papers 56 (4): 563–95.

16.

CEH (Commission for Historical Clarification). 1999. Memory of Silence. Washington, DC: American Association for the Advancement of Science.

17.

Conrad

Courtenay

Moore

Will H.

. 2010. “What Stops Torture.” The American Journal of Political Science 54 (2): 459–76.

18.

Coppedge

Michael

Alvarez

Angel

Maldonado

Claudia

. 2008. “Two Persistent Dimensions of Democracy: Contestation and Inclusiveness.” The Journal of Politics 70 (3): 632–47.

19.

Cunningham

David

Gleditsch

Kristian

Salehyan

Idean

. 2009. “It Takes Two: A Dyadic Analysis of Civil War Duration and Outcome.” Journal of Conflict Resolution 53 (4): 570–97.

20.

Danneman

Nathan

Ritter

Emily

. 2014. “Contagious Repression and Preemptive Repression.” Journal of Conflict Resolution 58 (2): 254–79.

21.

Danzger

Herbert

. 1975. “Validating Conflict Data.” American Sociological Review 40:70–84.

22.

Davenport

Christian

. 1995. “Multi-dimensional Threat Perception and State Repression: An Inquiry into Why States Apply Negative Sanctions.” American Journal of Political Science 39 (3): 683–713.

23.

Davenport

Christian

. 2007. The Domestic Democratic Peace. Cambridge: Cambridge University Press.

24.

Davenport

Christian

. 2010. Media Bias, Perspective and State Repression. Cambridge: Cambridge University Press.

25.

Davenport

Christian

. 2014. To Kill a Movement: Mobilization, Repression and Demobilization of the Republic of New Africa. Cambridge: Cambridge University Press.

26.

Davenport

Christian

Ball

Patrick

. 2002. “Views to a Kill: Exploring the Implications of Source Selection in the Case of Guatemalan State Terror, 1977–1995.” Journal of Conflict Resolution 46 (3): 427–50.

27.

Davenport

Christian

Johnston

Hank

Mueller

Carol

. 2005. Repression and Mobilization. Minneapolis: University of Minnesota Press.

28.

Daxecker

Ursula E.

Hess

Michael L.

. 2012. “Repression Hurts: Coercive Government Responses and the Demise of Terrorist Campaigns.” British Journal of Political Science 43 (3): 559–77.

29.

DeNardo

James

. 1985. Power in Numbers: The Political Strategy of Protest and Rebellion. Princeton, NJ: Princeton University Press.

30.

Doyle

Kate

. 1999. Guatemalan Documentation Archive. Washington, DC: The National Security Archive.

31.

Doyle

Kate

. 2007. “THE ATROCITY FILES: Deciphering the Archives of Guatemala’s Dirty War.” Harpers 1891:52–64.

32.

Drakos

Konstantinos

Gofas

Andreas

. 2006. “Distorting Effects on the Study of Terrorism the Devil You Know but Are Afraid to Face: Underreporting Bias and its Distorting Effects on the Study of Terrorism.” Journal of Conflict Resolution 50 (5): 714–35.

33.

Dugan

Laura

Chenoweth

Erica

. 2012. “Moving Beyond Deterrence the Effectiveness of Raising the Expected Utility of Abstaining from Terrorism in Israel.” American Sociological Review 77 (4): 597–624.

34.

Earl

Jennifer

Soule

Sarah

McCarthy

John

. 2003. “Policing Under Fire.” American Sociological Review 68:581–606.

35.

Earl

Jennifer

Martin

Andrew

McCarthy

John

Soule

Sara

. 2004. “The Use of Newspaper Data in the Study of Collective Action.” Annual Review of Sociology 30:65–80.

36.

Falla

Ricardo

. 1994. Massacres in the Jungle. New York: Westview Press.

37.

Fearon

James

Laitin

David

. 2003. “Ethnicity, Insurgency and Civil War.” American Political Science Review 97 (1): 75–90.

38.

Gamson

William

. 1975. The Strategy of Social Protest. Homewood, IL: Dorsey.

39.

Garrard-Burnett

Virgina

. 2011. Terror in the Land of the Holy Spirit: Guatemala under General Efrain Rios Montt, 1982–1983. Oxford, UK: Oxford University Press.

40.

Gartner

Scott

Regan

Patrick

. 1996. “Threat and Repression: The Non-linear Relationship Between Government and Opposition Violence.” Journal of Peace Research 33 (3): 273–88.

41.

Gates

Scott

. 2002. “Recruitment and Allegiance: The Microfoundaitons of Rebellion.” Journal of Conflict Resolution 46 (1): 111–30.

42.

Gaventa

John

. 1982. Power and Powerlessness: Quiescence and Rebellion in an Appalachian Valley. Champaign: University of Illinois Press.

43.

Gleditsch

Kristian Skrede

Ruggeri

Andrea

. 2010. “Political Opportunity Structure, Democracy, and Civil War.” Journal of Peace Research 47 (3): 299–310.

44.

Granovetter

Mark

. 1978. “Threshold Models of Collective Behavior.” American Journal of Sociology 83 (6): 1420–43.

45.

Gulden

Tim

. 2002. “Spatial and Temporal Patterns in Civil Violence.” Politics and the Life Sciences 21 (1): 26–36.

46.

Herreros

Francisco

Criado

Henar

. 2009. “Preemptive of Arbitrary: Two Forms of Violence in Civil War.” Journal of Conflict Resolution 53 (3): 419–45.

47.

Huntington

Samuel

. 1968. Political Order in Changing Societies. New Haven, CT: Yale University Press.

48.

Kalyvas

Stathis

. 2006. The Logic of Violence in Civil War. Cambridge: Cambridge University Press.

49.

Klandermans

Bert

. 1984. “Mobilization and Participation: Social-Psychological Expansions of Resource Mobilization Theory.” American Sociological Review. 49:583–600

50.

King

Gary

Pan

Jennifer

Roberts

Margaret

. 2013. “How Censorship in China Allows Government Criticism but Silences Collective Expression.” American Political Science Review 107 (2): 1–18.

51.

Kuran

Timur

. 1989. “Sparks and Prairie Fires: A Theory of Unanticipated Political Revolution.” Public Choice 61 (1): 41–74.

52.

Lacina

Bethany

. 2014. “How Governments Shape the Risk of Civil Violence: India’s Federal Reorganization, 1950–56.” American Journal of Political Science 58 (3): 720–38.

53.

Lewis

Janet

. 2012. “How Rebellion Begins: Insurgent Group Formation and Viability in Uganda.” PhD diss., Harvard University, Cambridge, MA.

54.

Lukes

Steven

. 2005. Power: A Radical View. New York: Palgrave Macmillan.

55.

Manz

Beatriz

. 2005. Paradise in Ashes: A Guatemalan Journey of Courage, Terror and Hope. Berkeley: University of California Press.

56.

Marwell

Gary

Oliver

Pamela

. 1993. The Critical Mass in Collective Action: A Micro-social Theory. Cambridge: Cambridge University Press.

57.

Marx

Gary

. 1988. Undercover: Police Surveillance in America. Berkeley: University of California Press.

58.

McAdam

Doug

. 1986. Political Process and the Development of Black Insurgency: 1930–1970. Chicago: University of Chicago Press.

59.

McAdam

Doug

Tarrow

Sidney

Tilly

Charles

. 2001. Dynamics of Contention. Cambridge: Cambridge University Press.

60.

McCarthy

John

Zald

Mayer

. 1977. “Resource Mobilization and Social Movements: A Partial Theory.” American Journal of Sociology 82:1212–93.

61.

McCarthy

John

McPhail

Clark

Smith

Jackie

. 1996. “Images of Protest: Dimensions of Selection Bias in Media Coverage of Washington Demonstrations, 1982 and 1991.” American Sociological Review 61:478–499.

62.

McCarthy

John

Titarenko

Larissa

McPhail

Clark

Rafail

Patrick

Augustyn

Boguslaw

. 2008. “Assessing Stability in the Patterns of Selection Bias in Newspaper Coverage of Protest During the Transition from Communism in Belarus.” Mobilization 13 (2): 127–46.

63.

Moore

Will H.

1995. “Action-Reaction or Rational Expectations? Reciprocity and the Domestic-International Conflict Nexus during the ‘Rhodesia Problem.’” Journal of Conflict Resolution 39 (1): 129–67.

64.

Moore

Will H.

1998. “Repression and Dissent: Substitution, Context and Timing.” American Journal of Political Science 42 (3): 851–73.

65.

Moore

Will H.

2000. “The Repression of Dissent: A Substitution Model of Government Coercion.” Journal of Conflict Resolution 44 (1): 107–27.

66.

Morales Alvarado

Sergio F

. 2009. El Derecho a Saber: Informe Especial del Archivo Histórico de la Policía Nacional. Guatemala City: Prosecurado de los Derechos Humanos de Guatemala.

67.

Myers

Daniel J.

Caniglia

Beth S.

. 2004. “All the Rioting That’s Fit to Print: Selection Effects in National Newspaper Coverage of Civil Disorders, 1968–1969.” American Sociological Review 69 (4): 519–43.

68.

Nordas

Ragnhild

Davenport

Christian

. 2013. “Fighting the Youth: Youth Bulges and State Repression.” American Journal of Political Science 57 (4): 926–40.

69.

Oliver

Pamela

. 1984. “If You Don’t Do It, Nobody Else Will: Active and Token Contributors to Local Collective Action.” American Sociological Review 49 (5): 601–10.

70.

Oliver

Pamela E.

Myers

Daniel J.

. 1999. “How Events Enter the Public Sphere: Conflict, Location, and Sponsorship in Local Newspaper Coverage of Public Events.” American Journal of Sociology 105:38–87.

71.

Oliver

Pamela E.

Maney

Gregory

. 2000. “Political Processes and Local Newspaper Coverage of Protest Events: From Selection Bias to Triadic Interactions.” American Journal of Sociology 106 (2): 463–505.

72.

Olson

Mancur

. 1965. The Logic of Collective Action Public Goods and the Theory of Groups. Cambridge, MA: Harvard University Press.

73.

Perera

Victor

. 1993. Unfinished Conquest: The Guatemalan Tragedy. Berkeley: University of California Press.

74.

Pierskalla

Jan

. 2010. “Protest, Deterrence and Escalation: The Strategic Calculus of Government Repression.” Journal of Conflict Resolution 54 (1): 117–43.

75.

Pion-Berlin

David

. 1989. The Ideology of State Terror: Economic Doctrine and Political Repression in Argentina and Peru. Boulder, CO: Lynn Reinner.

76.

Poe

Steven.

Neil Tate

. 1994. “Repression of Human Rights to Personal Integrity in the 1980s.” American Political Science Review 88 (4): 853–72.

77.

Poe

Steven

Neil Tate

Camp-Keith

Linda

. 1999. “Repression of the Human Right to Personal Integrity Revisited: A Global Cross-National Study Covering the Years 1976-1993.” International Studies Quarterly 43 (2): 291–313.

78.

Raleigh

Clionadh

Hegre

Håvard

. 2009. “Population Size, Concentration, and Civil War. A Geographically Disaggregated Analysis.” Political Geography 28 (4): 224–38.

79.

Rasler

Karen

. 1996. “Concessions, Repression and Political Protest in the Iranian Revolution.” American Sociological Review 61 (1): 132–52.

80.

Regan

Patrick

Henderson

Errol

. 2002. “Democracy, Threats and Political Repression in Developing Countries: Are Democracies Internally Less Violent?” Third World Quarterly 23 (1): 119–36.

81.

Reiner

Robert

. 2010. The Politics of the Police. Oxford, UK: Oxford University Press.

82.

Ritter

Emily

Conrad

Courtenay

. 2012. “International Law and Mobilized Challenges against the State.” Unpublished Manuscript. University of California-Merced.

83.

Ritter

Emily

. 2014. “Policy Disputes, Political Survival and the Onset and Severity of Repression.” The Journal of Conflict Resolution 58 (1): 143–68.

84.

Ross

Michael

. 2004. “What Do We Know about Natural Resources and Civil War?” Journal of Peace Research 41 (3): 337–56.

85.

Sanford

Victoria

. 2004. Buried Secrets: Truth and Human Rights in Guatemala. New York: Palgrave Macmillan.

86.

Schirmer

Jennifer

. 1998. The Guatemalan Military Project. Philadelphia: University of Pennsylvania Press.

87.

Scott

James

. 1985. Weapons of the Weak: Everyday forms of Peasant Resistance. New Haven, CT: Yale University Press.

88.

Scott

James

. 1990. Domination and the Arts of Resistance. New Haven, CT: Yale University Press.

89.

Siegel

David

. 2009. “Social Networks and Collective Action.” American Journal of Political Science 53 (1): 122–38.

90.

Shellman

Steven

. 2006. “Process Matters: Conflict and Cooperation in Sequential Government-Dissident Interactions.” Security Studies 15 (4): 563–99.

91.

Smith

Jackie

McCarthy

John

McPhail

Clark

Augustyn

Boguslav

. 2001. “From Protest to Agenda Building: Description Bias in Media Coverage of Protest Events in Washington, DC.” Social Forces 79 (4): 1397–423.

92.

Smith

Julian

. 2009. “A Human Rights Breakthrough in Guatemala.” Smithsonian Magazine. October.

93.

Snow

David

Benford

Robert

. 1988. “Ideology, Frame Resonance and Participant Mobilization.” International Social Movement Research 1 (1):1977–217.

94.

Snyder

David

Kelly

William

. 1977. “Conflict Intensity, Media Sensitivity and the Validity of Newspaper Data.” American Sociological Review 42:105–23.

95.

Stanley

William

. 1996. The Protection Racket State. Philadelphia, PA: Temple University Press.

96.

Stoll

David

. 1993. Between Two Armies in the Ixil Towns of Guatemala. New York: Columbia University Press.

97.

Sullivan

Christopher Michael

. 2014. “The (In)Effectiveness of Torture for Combating Insurgency.” The Journal of Peace Research 51 (3): 388–404.

98.

Sullivan

Christopher Michael

. 2012. “Blood in the Village: Local-Level Determinants of State Massacres.” Conflict Management and Peace Science 29 (4): 373–96.

99.

Tarrow

Sidney

. 1994. Power in Movement. Cambridge: Cambridge University Press.

100.

Thomas

Jakana

. 2012. “Don’t Stop Till You Get Enough: Rebel Group War Aims and Conflict Resolution in African Civil Wars.” PhD. diss., Penn State University, State College, PA.

101.

Tilly

Charles

. 1978. From Mobilization to Revolution. New York: Paradigm Press.

102.

Tilly

Charles

. 2004. Social Movements. Boulder, CO: Paradigm Publishers

103.

Toft

Monica

. 2003. The Geography of Ethnic Violence. Princeton, NJ: Princeton University Press.

104.

Valentino

Benjamin

Huth

Paul

Balch-Lindsay

Dylan

. 2004. “‘Draining the Sea’: Mass Killing and Guerrilla Warfare.” International Organization 58 (2): 375–407.

105.

Vreeland

James Raymond

. 2008. “The Effect of Political Regime on Civil War: Unpacking Anocracy.” Journal of Conflict Resolution 52 (3):401–25.

106.

Walter

Barbara

. 2006. “Building Reputation: Why Governments Fight Some Separatists But Not Others.” American Journal of Political Science 50 (2): 313–30.

107.

Walter

Barbara

. 2009. “Bargaining Failures and Civil War.” Annual Review of Political Science 12:243–61.

108.

Weidmann

Nils

. 2015. “On the Accuracy of Media-based Conflict Event Data.” Journal of Conflict Resolution 59 (6): 1129–1149.

109.

Weld

Kirsten

. 2014. Paper Cadavers: The Archives of Dictatorship in Guatemala. Durham, NC: Duke University Press,

110.

Winkleman

Rainer

. 2008. Econometric Analysis of Count Data. Berlin, Germany: Springer.

111.

Wood

Elisabeth Jean

. 2005. “Modeling Robust Settlements to Civil War: Indivisible Stakes and Distributional Compromises.” Unpublished Manuscript. New Haven: Yale University.

112.

Wucherpfennig

Julian

Metternich

Nils W.

Cederman

Lars-Erik

Gleditsch

Kristian Skrede

. 2012. “Ethnicity, the State, and the Duration of Civil War.” World Politics 64 (1): 79–115.