3

WARTIME ORGANIZATIONAL LEGACIES

Building Proto-Party Structures

In mid-1983, the FMLN commander Nidia Díaz scribbled “¿Qué falta?”—or, “What’s missing?”—into her field diary. While we might expect rebel leaders to ponder their shortcomings during war, this question becomes puzzling when we consider the context: in 1983, when she asked what the organization was missing, the FMLN was at the top of its game. On the battlefield, it had the upper hand against the regime. Off the battlefield, it was steadily gaining both local and international support. If rebels’ concerns about their wartime organizations were limited to their capacity to wage war, Díaz’s diary would have been blank that day. Instead, it is littered with bullet points, enumerating the future political, educational, and organizational directions of the PRTC and the FMLN more broadly.

Conventional portrayals of rebel organizations do little to help us make sense of Díaz’s musing. As I demonstrated in chapter 1, many, if not most, models of rebel organizations focus on the effectiveness with which they produce violence—and the structures responsible for it.1 To be sure, a focus on the militant aspects of rebellion is crucial for getting traction on conflict strategies and variations in patterns of violence during civil war.2 However, this analytic focus on the militant subdivisions of rebel groups has inadvertently been interpreted as a holistic portrayal, rather than a scope condition. As a result, militant organizations are often treated as unitary, homogeneous actors comprising “armed fighters and the commanders who lead them.”3 This assumption pre-disposes scholars to characterize the line between war and peace as more of a discontinuity than it is in reality. Indeed, we are told rebel groups need to build party structures from scratch “because electoral politics require a different set of skills than those demanded in wartime.”4

By these accounts, rebel groups should not make particularly good political parties. They have spent years, if not decades, engaging their opposition on the battlefields. Many (though not all) have committed heinous acts of violence against not just their opponents but noncombatants as well. Thus, while we may not expect a rebel commander to pose this question at the height of their military success, it seems considerably more warranted as they grapple with party transformation. This stage is often characterized by endless and tenuous ceasefires, frustrating rounds of negotiations, and an uncertain opportunity to participate in legal politics. In light of the very different road ahead, we find again the same refrain: that rebel groups must suddenly build party structures “from scratch” and “learn the art of non-violent politics.”5 If these refrains hold, we should expect groups attempting to make the jump into the legal political sphere to have to (and want to) break from their wartime legacies as swiftly and emphatically as possible. In other words, on the cusp of party formation, the answer to “¿qué falta?” should be “everything.”

Even nominally, however, the empirical record suggests otherwise. At the most superficial level, most rebel successor parties keep their name when they decide how to present themselves on the ballot.6 Beneath the surface, many wartime organizational legacies persist as well. Yet, rather than being detractors to transformation—bygone vestiges of conflict that obstruct peacetime politics—I argue that many of the structures built during war are essential for it.

Too often, wartime organizational legacies are bracketed in favor of explanations that lie outside the rebellion or outside the conflict theater entirely. For example, consider the prewar-party hypothesis: the argument that having once operated as a political party makes rebel-to-party transformation more likely.7 In short, prewar performance is the key to postwar performance. But what do we make of everything that happened in between? Of course, from a practical vantage, some bracketing is understandable. There are important questions to be asked about the extent to which wartime structures are knowable for clandestine groups—and some structures are harder to gauge than others. However, every scholar of organizational change highlights the importance considering the organization’s recent past to explain success or failure in the organization’s new future.8

This chapter tackles Díaz’s question and the other questions that arise when we hold the empirical record up to the conventional wisdom. ¿Qué falta? is fundamentally a question about the structure, content, and evolution of the wartime organization. To fully account for the structural aspect of rebel-to-party transformation, we have to know what rebels are starting with and what they are working toward. What structures do rebels have? What structures do parties need? Which structures make transformation into successor parties easier, and which obstruct it? What traits allow some rebel groups to exploit or shed their legacies? What’s missing—and, perhaps more importantly, what isn’t—as rebel groups look ahead to a new battle on the campaign trail?

I argue that the capacity for successful rebel-to-party transformation lies in rebels’ wartime organizational legacies. Organizations need two things to insure themselves against the volatility of change: relevant skills and the ability to exploit them. Thus, one part of transformation is a function of the diversity and type of organizational subdivisions built during wartime. Moving beyond unitary, combat-centric models of rebel groups, I identify a variety of subdivisions commonly—though not universally—built during conflict, which mirror the functions and goals of structures in party organizations. Where present, these proto-party structures provide decisive advantages when rebel groups face the pressure of postconflict transformation. Specifically, wartime subdivisions dedicated to governance, political messaging, social outreach, and other noncombat tasks imbue the organization with personnel, skills, and routines that retain their value in the electoral arena. As a result, some rebellions come to the negotiating table equipped with the structural building blocks of political parties. In sum, party building often starts when the first bullet is fired, not the last.

An equally important part of transformation is rebels’ capacity to exploit these structures under pressure. Militant organizations must display a capacity for critical self-reflection and flexibility—traits upheld as paramount in explanations of organizational change, yet infrequently found in political science.9 The sanguine focus on wartime bureaucracies and institutionalization masks their pitfalls: entrenched bureaucratic procedures are the enemy of flexibility. I identify traits critical to overcoming these issues: rebels’ ability to identify shortcomings and their willingness to change paths.

This chapter develops the first piece of my three-part theory of transformation—accounting for the content of change. Using the framework outlined in chapter 1, I map the core structures and variation among both militant and party organizations. The newfound conceptual precision from taking an explicitly organizational approach allows me to specify what parties need, what rebels have, and how the gap between the two varies from one militant group to the next. More specifically, I show that the gap between rebel structures and party structures is a variable, not a constant. To test whether this conceptual break from the literature holds in practice, I trace the parallel organizational evolutions of wartime structures within the subgroups of the FMLN.

Armed with a more thorough understanding of the composition and constraints on organizations, I motivate this theory by revisiting the outcome through an organizational lens. Rebel-to-party transformation is fundamentally a puzzle about executing and surviving a massive organizational change. The difficulties of transformation occupy two sides of the same coin. Throwing out existing routines and personnel destabilizes the organization, and the complementary challenge is the difficulty of learning new skills while also adapting to a new environment.10 Accordingly, transformation is easier to the extent that groups have “preexisting competencies”—established skill sets embedded in the organization that perform similar functions to those needed in the new environment.11 Thus, it is less risky when an organization can just do an old thing in a new way. In other words, although change is inherently difficult, it is not uniformly so.12

Rebel-to-party transformation is, however, a unique type of organizational change because we have a built-in template for the outcome: political parties. This is not to say that all parties look the same. Nevertheless, we know what types of tasks parties perform and the environments in which they function, which together provide broad insight into the internal structures parties need to support their activities.

The Content of Transformation: Rebel and Party Structures

Explaining how party organizations are forged out of rebellions demands an explicitly organizational approach—one that can account for both the content of the transformation (i.e., the specific structures that form the old and new organizations) and the process of transformation (i.e., how subdivisions and decision-making structures are altered and what obstacles they encounter along the way).13 This chapter deals specifically with the former. To explain how one builds a party out of the pieces of rebels, I start by specifying the building blocks of each.

Unfortunately, the rebellion scholarship and party scholarship exhibit a parallel oversight: despite a broad agreement that it is valuable to conceive of rebels and parties in organizational terms, there is much less consensus on what that entails analytically.14 I use the framework from chapter 1 to bring to light the (often) implicit organizational insights from both literatures. To assess which wartime legacies set rebels up for success—and which do not—as they reconfigure to function in the electoral arena, I first examine party structures.

Mapping Party Organizations

To shed light on which wartime structures affect party building, let us take a step back to examine what we know about parties as organizations: how they are structured, what skills they need in order to function, and the variety of functions they perform. Political parties—like rebel groups—are multifaceted organizations. They are irreducible to a single goal (winning elections—or winning wars). To paint them as unitary actors is to overlook the variety of activities, tensions, structures, and objectives within the organization, which can simultaneously work toward and detract from these broader goals.15 Though most studies of the organizational features of political parties focus on aggregated traits (e.g., institutionalization, bureaucratization, centralization), the literature is punctuated by a few crucial works that help unpack the diversity and type of roles that make parties what they are.16

In a long, if sporadic, tradition, party scholars like V. O. Key, Samuel Huntington, Kenneth Janda, and William Schonfeld laid a critical analytic foundation for disaggregating political parties beyond their parliamentary-extraparliamentary distinction.17 A common thread runs through their work: each scholar independently notes that parties must be understood as “working coalitions” of separate organizations.18 “Variety,” Key argues, “is the reality of party organization.”19 To specify what structures parties need in order to function, I start by asking, what is the nature of this variety?

Drawing on this legacy, Richard S. Katz and Peter Mair synthesize organizational insights into a useful framework for disaggregating the party organization into three core structural components: the party in public office, the party on the ground, and the party in central office.20 While they refer to these branches as the three “faces” of the party, in organizational terms, these components represent broad subdivisions of the party organization. Each demands different types of personnel, with different skills, to carry out different tasks, in service of different objectives.21 Reducing the party to any one element masks the variety of tasks and dynamics that undergird party functioning. Thus, the subdivisions Katz and Mair identify shed light on what “party structures” actually refer to.

Since the objective here is to identify the types of tasks parties must be able to perform—and, by extension, the types of structures they need to support those activities—I briefly outline how each subdivision functions and fits into the organization. The party in public office refers to the governing apparatus of the party.22 It comprises the members who run in elections and eventually occupy seats in government. This wing has the most direct control over creating and enacting policy. They also interface and negotiate with other political actors both inside and outside of government.23

The party on the ground refers to the subdivision responsible for forging linkages with society. The objectives associated with this wing include managing and maintaining external relations with those needed to keep the party in office. In short, they are responsible for building a loyal constituency. The structures that compose the party on the ground are the party congresses and committees that exist at various levels of the state, as well as the ancillary membership organizations (e.g., youth wings, women’s wings, and student organizations).24

The third party structure is the party in central office. This wing comprises the national executive committee and the central party staff.25 Functionally, the party in central office coordinates campaigns, oversees media and press, and acts as a conduit between the party on the ground and the party in public office. It is largely responsible for building the party platform and deciding how the core tenets of the party will be communicated to the grassroots and represented in policy. Broadly, the formation and communication of party identity lie with the party in central office

Of course, party organizations differ in terms of the priorities, overlap, and strategies they employ across these fronts. Moreover, they differ in the extent to which they emphasize (via dedicated personnel and resources) these subdivisions at all.26 This framework is especially useful because it optimizes between nuanced and flexible insights into the types of structures and skills party organizations require. At the most general level, political party organizations will always require a governing sect, a national committee to coordinate elections and party messaging, and a division that forges linkages with society. Each subdivision allows for a variety of different structural elements depending on the party, its approach, its objectives, and its environment.27

These insights help flesh out the otherwise-vague concept of party structures. They equip scholars with the analytic and conceptual precision needed to specify what party organizations need in order to function in a competitive electoral environment. Armed with a nuanced conception of what parties do and how they are structured to do it, this framework allows me to move beyond abstract questions like “how do rebel groups transform into political parties?” and toward a set of more nuanced questions like “which wartime activities are best suited to pivoting toward these tasks?” and “which structures best prepare militants to operate in these political domains?”

Mapping Rebel Organizations

To motivate this section I revisit the hummable tune of the rebel-to-party literature: to successfully transform, rebel groups must build party structures from scratch once the war ends because “electoral politics require a different set of skills than those demanded during wartime.”28 The purpose of this chapter and the next is to treat this refrain as a set of questions rather than a set of facts. What skills—and corresponding structures—do parties need? What skills—and corresponding structures—do militants build during conflict? Armed with a more nuanced conception of what parties do and how they are structured to do it, this section asks whether the organizational legacies of war can ever work in service of the party.

It is now widely acknowledged that rebels do more than just fight battles. They govern; they provide social services; they extract, process, and sell natural resources; they provide political education, both externally and internally; and the list goes on.29 With few exceptions, however, these critical insights about what rebels do during war are seldom interpreted as insights about what rebels are during war. In other words, while these observations are treated as behavioral insights, they are also structural ones.

Any routinized (as opposed to ad hoc) task—whether it’s providing governance or selling diamonds—requires a dedicated role or subdivision to coordinate, execute, and support it. Thus, routinized behaviors reflect the structures that produce them. Take, for example, creating propaganda. To perform this task regularly, the group needs a dedicated wing of personnel who coordinate the content, writing, printing, and distribution of the group’s message. If, alternately, routinized behaviors were not direct evidence of underlying organizational structures, wartime propaganda would be generated by a different set of combatants who spontaneously perform “political-messaging duties” each day or week. It quickly becomes clear that the alternate story would be inefficient and unsustainable, if not outright ridiculous. As such, rebels’ wartime behaviors provide critical information about the skills, priorities, and personnel embedded in the organization.

Taking an organizational approach to transformation reveals that one key to succeeding lies in the diversity and type of noncombat structures rebels build during war. In a departure from most analyses of militant organizations, I look beyond combat units and identify a set of wartime subdivisions that mirror the core structures of party organizations: governance and administration, political messaging, social service provision, and citizen education and outreach. These proto-party structures imbue the organization with the relevant skills, institutions, and personnel required to operate in an electoral and governing capacity once the conflict ends. While all militant organizations have combat units, rebel groups vary considerably in the scope and development of structures dedicated to noncombat tasks.

Proto-Party Structures: The Currency of Transformation

In a departure from nebulous concepts like “political wings” and “organizational resources,” I identify specific organizational structures that map directly onto the core structures of party organizations in terms of the functions they perform and the personnel that compose them. These wartime proto-party structures can be sorted to roughly correspond to the party structures in Katz and Mair’s framework (see figure 3.1). Conceptualizing militant organizations in these terms illustrates the range of partylike activities in which rebel groups have developed competencies.

The first set of wartime structures with postconflict utility are the wings in charge of governance and administration of local populations. In terms of Katz and Mair’s framework, governance divisions broadly mirror the party in public office in that they form the public face of the nascent party and confer legitimacy on the organization. Adapting Ana Arjona, Nelson Kasfir, and Zachariah Mampilly’s definition of rebel governance to reflect the organization, governance structures refer to the subdivisions that facilitate militants’ “regulation of the social, political, and economic life of non-combatants during war.”30 Specifically, these structures carry out tasks like taxation and market regulation, land and resource allocation, building infrastructure, social service provision, and dispute resolution.31

Rebel governance tasks vary widely, but none can be accomplished without dedicated resources, established routines, and skilled personnel. Hezbollah, for example, had an extensive service sector bureaucracy, which managed everything from building and staffing schools to providing potable drinking water to running medical clinics to collecting garbage in the areas it administered.32 In El Salvador, one of the FMLN’s sects (the FPL) established local governance structures known as Poderes Populares Locales (Local Popular Power, or PPLs), in which rebels created extensive zones of democratic political control. PPLs integrated local populations into administrative structures via extralegal elections and managed dispute resolution, economic regulation, security, education, agricultural production, and a variety of other governance activities.33

FIGURE 3.1. Mapping wartime structures to party structures

To be sure, many governance tasks actually transcend the responsibilities of political parties. For example, providing security and local judiciary systems are—as Mampilly notes—more reflective of the state apparatus than a party.34 As such, every governance division may not have a clear political-party counterpart. Nevertheless, rebel organizations with governance subdivisions accrue a variety of organizational benefits relevant to postconflict transformation. First, these structures often require personnel with specialized administrative skills, whose experience in the organization with relevant political tasks provides leaders with an internal pool of viable party members. Second, a wartime legacy of governance structures imbues the organization with the personnel, skills, and experience to engage in nonviolent political coordination with local populations. Third, the challenges of coordinating among different sectors (both internally and externally) means the organization has learned at least some of the crucial skills required for political coordination in government.35 Finally and most broadly, governance structures enhance both the legitimacy and political credibility of the rebels.36

The second set of wartime structures that mirror core party structures are the subdivisions dedicated to social and political outreach (e.g., providing social services or political education) and managing ancillary organizations (e.g., youth wings, student unions, women’s wings, and trade associations). Broadly corresponding to what Katz and Mair call the party on the ground, these outreach structures forge politically salient linkages with the community on which they rely for political and logistical support. Ancillary organizations, in particular, have clear counterparts in party organizations and, as such, can form the basis of a politically mobilized and loyal constituency after the war ends without much change to the structures or relations that they comprise.

An alternative—though not entirely unwarranted—perspective views youth and student organizations (as well as social service provision) as little more than thinly veiled pools for semicoercive combatant recruitment.37 Of course, this interpretation may be true in some cases, or recruitment may be one of many functions these ancillary organizations serve. As a generalization, however, this view conflates “participation” with “fighting,” thereby reifying the assumption that combat is the only activity for which rebels recruit. After all, if all rebels do is fight, then recruitment is all about finding fighters.38 Sometimes, this characterization is accurate. As a generalization, however, it obscures the complexity and variation in recruitment tactics and rebel-citizen dynamics. Participation in rebellion takes many forms—varying both within and between conflicts.39

The final class of wartime structures with clear postconflict relevance are what I broadly refer to as political-messaging wings. Mirroring—at least in part—the party in central office, these structures manage the articulation, refinement, creation, and distribution of media conveying the group’s political message. As such, political-messaging wings are responsible for taking the group’s core ideology and translating it into a cohesive, digestible, and transmissible narrative—to members and outsiders alike.

Much as parties do, many rebel groups produce tailored narratives for multiple audiences simultaneously. Some of the messaging is intended for local populations, some is intended for internal distribution, some is directed at the opposition (e.g., trying to encourage defection), and some is directed at foreign sponsors or the diaspora.40 The FMLN, for example, ran two radio stations and published multiple newspapers and magazines—and those cover just the domestic messaging wings. Furthermore, one of its five leaders published in Foreign Policy (in English) during the war, gracing the magazine with what might be the most compelling, if not intimidating, corresponding author footnote: “Joaquín Villalobos is a member of the directorate with the rank of commander of the Farabundo Martí National Liberation Front.” The takeaway is that political messaging is not an ad-hoc activity—particularly in the preinternet days. Even written media requires teams of (often multilingual) writers, photographers and developers, designers and illustrators, printers, and distributors.

It is worth noting that political-messaging wings as formal structures have retained their salience over time, despite the accessibility of social media and online publishing. Notwithstanding what cute epithets like “the Twitter Revolution” would have us believe, rebel groups have not interpreted the shift in the medium of publication as an easy way out. Though they may no longer need access to photographic processing or printing presses, these roles are being replaced by digital artists, graphic designers, and software engineers. Contemporary political messaging like Dabiq (an English-language publication distributed by the Islamic State of Iraq and the Levant) or Special Force 1 and 2 (Hezbollah’s video game series) require a lot more than 240 characters to get off the ground. The production quality and consistency of this propaganda speaks to the organizational diversity of the respective groups involved.

Beyond the clear applicability of skills, I expect militant organizations with well-developed messaging wings to have three advantages over their less vocal counterparts. First, these structures create and reify a core political identity, which contributes to a more cohesive organizational culture and mitigates the risk of change.41 Second, the opportunity to develop their messaging strategy over the course of the conflict gives these groups a sort of trial-and-error period. They can test what resonates and what needs refinement before theirpolitical platform has to (legally) “go live” when campaigning begins.42 Third, when campaigning starts, groups with preexisting messaging wings already have at least the skeleton of a platform on which to run. Rebel-to-party scholars often allude to the process of building a cohesive political platform as one of the primary challenges of making the jump into the political arena.43 Wartime political-messaging wings should thus provide rebels with a considerable head start on this crucial asset.

For rebel groups seeking to reinvent themselves as political parties, proto-party structures confer four advantages to the organizations that build them. First, since the organization has skilled personnel and existing routines, proto-party structures narrow the presumed gap between the skills needed during war and the skills needed to function in electoral politics.44 In organizational parlance, these “preexisting competencies” will make transformation less taxing because the group will not have to suddenly “learn the art of nonviolent politics,” or at least, the learning curve should be considerably less steep.45 The second and third advantages concern the process of transformation—how the organization undergoes and survives a massive overhaul in its structures, its objectives, and its environment. I address these advantages in greater detail in chapter 4, but briefly, proto-party structures contribute to organizational diversity (which broadly enhances groups’ resilience to shocks) and provide a less disruptive path to party building, enabling groups to repurpose existing structures rather than starting from a blank slate.

The fourth benefit is less intuitive, but equally crucial. Proto-party structures benefit the organization by forcing groups to confront relevant challenges. Returning to Katz and Mair’s framework, the authors note two challenges associated with the party in public office: (1) they must be attentive to the electorate, and (2) they face a set of constraints from needing to work collectively with “coalition partners, civil servants, [and] officials at other levels of government.”46 In the wartime context, rebels with governance structures will face similar challenges. On the one hand, of course, these challenges are organizationally taxing—which is in part why we find rebel governance a puzzle worth investigating.47 On the other hand, facing these challenges during wartime represents an early opportunity for organizational learning, thereby attenuating the gap in knowledge as they pivot to campaigning and mobilization.

Facilitating Change: Self-Reflection and Flexibility

To explain how militant organizations survive the massive change that rebel-to-party transformation entails, we need to account for more than just useful structures. As with elite athletes and musicians, talent can only take organizations so far. At a certain point, the capacity to execute a risky move is more about whether they have the traits to absorb shocks and continue functioning at a high level under pressure. Here, I identify the organizational traits most likely to facilitate rebel-to-party transformation. Two undertheorized organizational traits put some groups in a better position than others to quickly exploit their varied skill sets when the time is right. The first is the capacity for critical assessment and self-reflection; the second is what Debra Minkoff calls a “repertoire of prior flexibility.”48 Self-reflection is about consistently seeking information with an eye toward improving performance.49 From an organizational standpoint, this trait is present to the extent that leaders have institutionalized mechanisms for seeking and incorporating feedback. In other words, the group needs a culture of self-reflection.

This trait not a given. Organizations, like individuals, can tend toward openness or stubbornness. On the battlefield, stubbornness can almost be forgiven. The problem, as Vaughn Tan deftly observes, is our tendency to conflate uncertainty with risk.50 As such, the openness required for innovation can easily feel like risk acceptance, especially in unsafe environments. In turn, organizations are liable to stick with what they know, even when those strategies are no longer working. While major organizational changes will be unavoidable once the group has to pivot to electoral politics, making these changes will be easier to the extent that they know how to evaluate their shortcomings and the skills at their disposal.51

Critical self-reflection is the theoretical side of change; to actually enact it, organizations must be flexible in practice. Flexibility describes an organization’s ability to adapt under duress. While militant groups are no strangers to adversity, developing a repertoire of flexibility is less about surviving a constant threat of death (e.g., fighting consistent battles in the same way) than it is about facing new threats to their survival or operating in new conditions. For example, insurgencies often encounter boosts or tactical changes in counterinsurgency strategy to which they must respond. Or they suffer a particularly deadly incursion—perhaps one that took out a leader or a critical logistical subdivision. In these cases, flexibility means using existing organizational resources in different or innovative ways.52

The extent to which rebel groups have undergone substantial organizational changes in the past is the extent to which they are better poised to adapt and survive major changes in the future. This cycle is what Minkoff means by “developing a repertoire of prior flexibility.”53 Before party transformation, many rebellions undergo significant internal shifts (either by force or by strategic choice). These changes may arise from a variety of circumstances: shifts in sponsorship or funding, upticks in counterinsurgency efforts, the decision to build new wings of the organization to cope with new obstacles or exploit new opportunities, leadership turnover, and a variety of others. For example, following a surge in US military aide to the Salvadoran regime, the FMLN’s use of conventional warfare was no longer sufficient to keep government forces at bay. In response, the group adopted the war-of-attrition approach of one of its subdivisions and subsequently reorganized its military units across the organization to wage a guerrilla insurgency. This change was made easier because the FMLN had one subdivision already employing this tactic, and undergoing this change made the organization more open to organizational innovation in the future.

A reasonable question to ask at this point is whether organizational flexibility is realistic for a hierarchical military organization, or merely aspirational. Scholars come down quite hard on both sides. David Close and Gary Prevost explicitly argue that “political-military fronts cannot have flexible structures [because] they are command organizations.”54 They maintain that malleability is incompatible with hierarchy. Others, however, have unearthed evidence of flexibility in places we might not expect it. Sarah Parkinson’s analysis of the Palestine Liberation Organization mobilization following the 1982 Israeli invasion of Lebanon expressly contradicts Close and Prevost’s assertion. She demonstrates that clandestine and flexible structures “emerge from formal militant hierarchies.”55 Thus, rather than being antithetical to military hierarchy, organizational flexibility can be a consequence of it.

Moving off the battlefield, Anna Grzymała-Busse observes the presence and implications of flexibility in an equally surprising place: pre-1989 communist parties.56 While acknowledging that many communist parties of the era “lacked incentives” to devise flexible policy strategies and “became the stereotype of unchanging behemoths,” she observed that not all exhibited the same stubbornness. Grzymała-Busse not only demonstrates variation in flexibility, but also that the parties displaying policy responsiveness had more favorable outcomes than their rigid counterparts as they sought to reinvent themselves in the wake of 1989: “The more a party promoted policy innovation prior to 1989, the more it fostered pragmatism and flexibility in policy making. The more it had subsequently implemented these innovations, the more experience the party elites received in overcoming administrative reluctance, organizational entrenchment, and other . . . barriers to party regeneration.”57 Crucially, these traits go hand in hand: together, self-reflection and flexibility mean the organization has a culture of continuous innovation. The leaders accept the possibility of wrongdoing, search for new ways of operating, and enact changes accordingly. In short, these traits combat organizational rigidity, which is anathema to successful transformation.

Empirical Analysis: Tracing Wartime Legacies

The central argument of the book is that the organizational legacies of war shape rebels’ prospects for and path to party formation at the war’s end. This chapter develops the first part of the theory: explicating the organizational content that facilitates change. I argue that proto-party structures and institutionalized flexibility together facilitate postwar transformation by imbuing rebel groups with relevant skill sets and the capacity to leverage them into the electoral arena.

The next logical question is whether and to what extent rebels actually build these proto-party structures and adopt a culture of flexibility in practice. Crucially, it is worth noting that phantom structures do exist. As William Reno documents in the National Patriotic Front of Liberia (NPFL), leaders constructed “the veneer of a government administration, legislature, and courts . . . primarily . . . to gain international recognition and for additional opportunities to collect bribes.”58 Since many proto-party structures concern the nature of interactions with local populations, it would be reasonable to expect leaders to exaggerate the extent to which those interactions are characterized by service and downplay the extent to which those interactions are characterized by violence.

Thus, before I test their relevance to party transformation, I first ask, what do rebels’ wartime organizations look like on the ground? In this section, I conduct process tracing using a wide variety of primary source material from across the FMLN’s subgroups. To make the argument that wartime organizational legacies are central to postwar party formation, I must be able to demonstrate the following: First, rebel groups must build functioning proto-party structures that operate according to the theory. In short, at least some militant organizations must build structures that actually function—in contrast to the “phantom structures” Reno documents in the NPFL. Moreover, these structures must vary from one group to the next. If everyone builds them, then they will not account for variation in rebel-to-party outcomes. Second, these structures must persist until the end of war. If instead proto-party structures exist for a short while, but are dismantled long before the end of the conflict, this discontinuity would call into question the mechanism of my argument.

Beyond the structural considerations, substantiating this theory requires evidence that groups undergo organizational changes in the way that I expect. Are leaders thinking about organizational dynamics and flexibility? Or are their decisions ad hoc or based on some other calculus entirely? If we consider the business world, we know that C-level executives think deeply about their organizations in the face of change: they are considering growth dynamics; they are bringing in organizational consultants; they are talking about resilience to economic downturns. Do these conversations happen as much on the battlefield as they do in the boardroom? To assess the FMLN’s capacity for flexibility, I look for evidence of how leaders respond to failure: Do they admit it? Do they attribute failures to organizational shortcomings or external shocks? How do they respond?

Recall that at this point in the war, the strategic directive from the FMLN General Command still de-emphasized local political engagement in favor of military recruitment. Obviously, this directive has clear organizational implications accompanying the behavioral ones: if the emphasis is exclusively on fighting, organizational resources should be dedicated to combat divisions and logistical support. Yet, I expect that the extent to which this decree played out in practice will vary considerably as a function of the groups’ respective organizational ideologies of war. In light of their founding principles and the unique geographical dispersion that occurred in the wake of the failed insurrection, I expect that the groups’ initial ideological and structural disparities will be further entrenched in this period.

The FMLN on the Heels of Failure

By all scholarly, military, and internal accounts, the Final Offensive was an abject failure. Some of the problems were attributable to logistical blunders (e.g., the FMLN broadcasted information about the insurrection but the state cut them off before they could give a date or time).59 Some were due to naive miscalculation—the ERP vastly overestimated the number of people who would defect from the army; others overestimated how many would turn out for the general strike.60 Yet, the lion’s share of the failure was organizational. Despite agreement among some FMLN commanders and their patrons about the plan, top leaders remained divided over the appropriate military strategy. Neither the FPL nor the RN favored a violent insurrection, and many cadres from these OP-Ms neglected to participate in what was supposed to be coordinated military action.61 Indeed, scholars have argued explicitly that were it not for the vitriolic relations between leaders, the FMLN would likely have achieved a swift victory during that period.62

Infighting, however, was not the group’s only organizational gaffe. After nearly a decade spent prioritizing mass mobilization and political cadres (with the exception of the ERP)—the FMLN had only a few months to build up and train a standing army.63 Even the ERP, which favored violent insurrection from the start, lacked the tactical and strategic training needed to mount an insurrection at the scale they had planned. At this point, the FMLN faced a major organizational dilemma. The most readily available source of military recruits was its affiliated mass organizations, yet, as Bracamonte and Spencer note, “stripping the political wing of its key cadres meant that there was no guarantee that the organizations would come out into the streets for the insurrection.”64 Nevertheless, they took the risk. On the advice of the Sandinistas, FMLN leaders opted to repurpose key cadres from its mass organizations into combat units ahead of the insurrection.65

Seventy-two hours after the Final Offensive, FMLN combatants withdrew to their respective bases to regroup, rebuild, and try again. While the General Command held firm to its military-first strategy, the prolongation of the war (beyond a single insurrection) meant that each group was to focus on building up a strategic rearguard. This directive had important implications both for the organization and the analytic purchase of studying it. Recall that, from the beginning, each OP-M retained its own organizational structure. Thus, the diversity and nature of subdivisions, activities, institutions, and external relations were still under the jurisdiction of the OP-Ms’ respective leaders. This organizational decision in conjunction with their postfailure retreat created a unique geographical phenomenon that persisted throughout the war: five groups, with five different structures, and five different interpretations of revolution settled in distinct regions of the country. As a result, the war is characterized by geopolitical conditions in which each OP-M receives the same directives, but functions (largely) autonomously.

Although the January offensive was a failure any way you turn it, the manner in which the FMLN leadership reacted to the blow had beneficial organizational repercussions that shaped its evolution. In the aftermath of the insurrection, the General Command held numerous meetings to debrief, reflect, and identify tactical errors in its approach.66 By and large, the leaders converged on what went wrong:67 the strategic approach was a good one, but the organization lacked the training, cohesion, and communication to execute it. Eager to take advantage of a good revolutionary moment, they sacrificed preparedness for enthusiasm. The PCS leader Schafik Hándal reflected on this failure in an interview: “The delay in unifying the revolutionary organizations did not allow us to take advantage of the . . . situation of 1980. Unity should have been consummated a year earlier, but analysis is slower than objective reality.”68 Joaquín Villalobos (ERP) corroborated this point, attributing the failure to take power at that time to “the lack of unification upon a strategic line within the revolutionary movement.”69

Though the FMLN was still growing, the failed offensive paved the way for the organization to consolidate around a core principle that shaped its trajectory: espíritu autocrítico—literally, the spirit of self-critique. In essence, autocrítico is the organizational equivalent of the “growth mindset”: an approach to problem solving in which individuals view failure as a learning opportunity rather than evidence of deficiency.70 To be sure, this concept was not new. Each of the OP-Ms boasted an explicit commitment to autocrítico in their formative years.71 And they embraced it both privately—as in the case of Nidia Díaz scribbling “¿que falta?” in her field diary—and publicly—as in the many cases of leaders writing openly about this approach in popular publications during the war.72

The pernicious effects of rivalry dealt a sobering blow to the leadership. But even this awareness was no guarantee of future harmony. In the wake of such a profound failure to execute a plan only one of the five groups truly endorsed, the General Command could have gone one of two ways. One option was that all of the members bring to the table their commitment to autocrítico, and ask, in good faith, what went wrong? In an equally plausible world, however, egos could easily trump principles. Even one or two people doubling down on avoiding blame could prevent a culture of self-reflection from sticking. From an organizational standpoint, having a consensus-based leadership structure was likely quite helpful in preventing the latter.73 Requiring consensus meant that accountability for the failure was dispersed across many people rather than concentrated in a single individual. Psychologically, this dispersal of responsibility made it easier for leaders to ask what went wrong without needing to point (only) to themselves.

While consensus is not a necessary condition for critical self-evaluation, it was helpful for institutionalizing the practice during the organization’s rocky adolescence. Moving forward, the General Command more explicitly incorporated the OP-Ms’ varied strategic approaches into a holistic arsenal. As Bracamonte and Spencer argue, this intellectual shift allowed the FMLN to “change the priority depending on the current strategic situation.”74 The question, of course, is whether this commitment bore fruit in practice.

Organizational Evolution, 1981–1984

In the run-up to the first incursion, the OP-Ms had neither the time nor the resources to see their organizational plans to fruition. For all except the ERP, the combat-centric directive pushed them even farther afield as they were forced to temporarily repurpose political cadres into military roles.75 In the wake of the retreat, however, the OP-Ms had the time to regroup and rebuild their organizations in accordance with their founding principles. As the FMLN entered the new strategic phase in preparation for the “Second Offensive,” their geopolitical dispersion affords a unique opportunity to trace how organizational structures evolve during war. The ERP leader Villalobos recounts the circumstances leading up to the rematch:

This offensive put the army on the defensive by concentrating it in its own strategic areas. That gave us a few months of peace of mind and allowed us to create the seven strategic fronts, the seven concentrations of forces and the existence of that rearguard that gave us the possibility of preparing people. . . . Even the enemy’s own offensives became a preparation school combative. All those months, the months in which we resisted in those positions forced us to learn. Not only did we have the terrain to prepare the men, but we were also forced to solve the problem of military tactical learning in the concrete, facing the enemy.

It was not a school in which we graduated people and then took them to a theater of operations, but rather we were with the rearguard and the theater of operations intermingled there, because sometimes the enemy took us out of certain areas and then we returned to retake ground control. This also implies that people learned engineering work to protect themselves from artillery bombardments, from aviation, on the basis that this was the order of the day and happened on a daily basis. In other words, what made good military contingents form in Morazán, in Chalatenango, in Guazapa was the fact that for many months in those areas they had to fight almost daily against the enemy’s effort to annihilate us.76

The groups’ physical and ideological isolation from one another were compounding forces, allowing five different groups to evolve according to their own visions, without interference from the others. I trace whether and how those principles translated into measurable organizational outcomes.

The ERP: Militancy in Morazán

After the January offensive failed to overthrow the government, the ERP retreated to its base in Morazán department in eastern El Salvador. To the ERP, retreat meant an opportunity to build and train its armed forces and prepare for another shot at overthrowing the regime through violent insurrection. Thus, their goals upon returning to Morazán were to liberate the northern half of the department by wiping out any government and paramilitary presence, and to establish a rearguard dedicated to military training.77 To be sure, the ERP did not take a completely maladaptive approach. In responding to a question about how the ERP pivoted to build a popular army in the wake of the January failure, the leader Villalobos explicitly identified where adjustment was needed: “Our current conditions [in the wake of the retreat] mean having to build an army. Once the door is closed on the possibility of an insurrection, we face the need to achieve greater attrition of the army, which in turn forces us to fine-tune our military structures.”78

In line with my expectations, the ERP’s organization evolved in lockstep with its organizational ideology of war. It focused almost exclusively on building a standing army, which it rapidly achieved.79 Despite near-constant incursions from the government and paramilitary forces, the rebel army managed to gain the upper hand. By 1983, northern Morazán was decisively under the control of the ERP.80 While Villalobos was hesitant to denote Morazán as a fully liberated zone, this region was by far the most autonomous rearguard the FMLN had in the early stages of the war.81 To the extent that the ERP diversified, its focus was on building logistical subdivisions in support of the war effort: “establishing hospitals, production brigades, and training camps.”82

What were the organizational implications beyond the battlefield? The level of autonomy the ERP exercised in the region would lead some to expect governance structures and mass political mobilization to follow in turn.83 As Reyko Huang argues, “Where rebels tap into civilians as a significant war-making resource,”—which the ERP did—“the latter become politically mobilized”—which locals did not.84 In reality, the militarization of Morazán was as evident in the ERP’s relationship with the masses as it was in its internal dealings.

It is here where opening the black box of rebel organizational structures really matters. To be fair, assuming that political mobilization follows necessarily from territorial control is easy because the organizational structures rebels (often) build are so robust that mobilization appears seamless. In reality, political engagement is effortful and the variation across cases is salient and systematic. Consider the account of Miguel Ventura, a local priest who worked closely with the ERP in Morazán throughout the war. He writes the following about the population living in ERP-controlled territory in the first phase of the war: “Some of the people [in northern Morazán] had been involved in the liberation process during the 1970s, while others had lived under the control of the military and did not have a clear vision of the revolutionary process. . . . There was little sense of community” among civilians living in the area.85

While the ERP was not isolated from the masses, its engagement was uni-dimensional. The people existed to support the war effort, not the other way around.86 For example, when asked why the army must move with the masses, Villalobos responds predictably: “[We move with the masses] because we depended on them. In other words, we in that rearguard, without the masses, would not have had the best chance, neither of having human reserves nor of having supplies. This forces us to depend on the masses in order to survive. We had to protect the masses and take it into consideration for any plans for military maneuvers.”87

Crucially, relegating the masses to a support role was a choice rather than an inevitability. Recall that, like the FPL, the ERP was built in the context of the Christian base communities (CEBs) of the 1960s and ’70s. As such, it drew its support from a well-politicized social infrastructure—it just chose not to use it as such. Binford’s research further corroborates this argument, noting that “between 1981 and 1983 even the church was subordinated to the revolutionary process and assigned the task of preparing the population for the insurrection that the ERP believed was imminent.”88

Diversification, like most things, is a dilemma rather than an unencumbered virtue. Rebels face a trade-off between putting all their eggs in one basket and being jacks of all trades and masters of none. In the first phase of the war, the ERP came down hard on the former. The decision stemmed from two beliefs: First, that a second offensive executed with a better army and more extensive communication system would level the regime. Second, that political mobilization would follow necessarily from a successful insurrection.89 The decision to continually shirk the political side of the fight had problematic consequences, however, which the ERP was starting to feel. As Binford notes, the ERP’s military emphasis turned it into “the best-organized and most formidable military force of the guerrilla groups,” but it “lacked a mass political base.”90 Moreover, as 1983 progressed, the ERP faced an intensifying counterinsurgency strategy from the government. It was losing personnel faster than it could recruit and train replacements, at which point the group briefly turned to forced conscription.91 While it abandoned the practice after only a few months in response to criticism, ideological support for the ERP dropped precipitously in this period.92

The FPL: Diversification in Chalatenango

The progress of the war and development of the ERP in Morazán were largely representative of the FMLN more broadly during this phase; yet if we turn to Chalatenango, a very different story emerges. In the aftermath of the failed insurrection, the FPL retreated to Chalatenango where—much like the ERP in Morazán—it had originated out of the CEBs of the 1970s. Yet, the similarities between OP-Ms stop here. In contrast with many operations elsewhere, FPL insurgents in Chalatenango “placed great emphasis” on developing grassroots governance structures in cooperation with locals.93 Rather than simply controlling territory, the FPL’s priority was to “establish the infrastructure of resistance” among the peasantry.94 This approach to territorial administration followed from the FPL’s ideological orientation emphasizing “social and community organizations” in service of a prolonged popular war.95 Here, I detail the social and political structures that the FPL built during this time and their implications for both the FPL and the future of the broader movement.

Before I begin, however, it is worth addressing a potential misinterpretation of the FPL’s activities during this period. Keeping in mind that the General Command’s directive was to build up large battalions for engaging the armed forces in conventional warfare, one might interpret the FPL’s political activities as insubordination, and thus, not worthy of comparison. I argue, however, that the FPL’s approach represents not insubordination, but a logical interpretation of orders filtered through the FPL’s organizational ideology. The FPL emphasized a revolution through a popular war achieved by politicizing and mobilizing peasants in the countryside.96 In essence, FPL ideology does not understand an insurrection as something that can be divorced from political mobilization. According to Jenny Pearce, the FPL interpreted the concept of “rearguards” (which OP-Ms were instructed to establish in the countryside) as something more than a military designation. She argues that FPL leaders viewed rearguards and liberated territory “within a broader framework of political mobilization.”97

Moreover, while the FPL dedicated significant organizational resources to bolstering its administrative, social service, and press divisions, it did not neglect its army. Recovered documents from the FPL’s armed forces reveal concrete plans to augment its battalions. The plans lay out broad military objectives (e.g., sabotaging strategic highways and communication lines, setting fire to oligarchs’ properties), detail various workshops on how to make weapons, and even assign teachers to the different workshops.98 Yet, even in its military plans, the FPL’s political through line is clearly embedded. Figure 3.2 depicts a section of a document in preparation for what the authors called the Second Offensive.99 While the majority of the content concerns armed operations, the two lines with arrows indicate plans to “politically consolidate our forces,” and to “create local popular power structures (to provide services to the people).”

FIGURE 3.2. Political priorities in military plans (FPL 1981). Orden de Preparar Segunda Ofensiva, David Spencer Collection, box 5, folder 4.5, Hoover Institution Library & Archives.

Even in the early stages of the war, FPL diversified into numerous proto-party domains. The scope and content of the FPL’s structural diversification adheres closely to the organizational ideology the group articulated—but could not fully realize—before the war. Indeed, the breadth and depth of organizational development it achieved was no small feat given both its dearth of resources and the fact that it never managed to secure a fully autonomous rearguard in Chalatenango comparable to the ERP’s territory in Morazán. Here, I focus on three initiatives and corresponding subdivisions that together illustrate the FPL’s organizational development in this period: the creation of local governance structures (PPLs), its education and mass literacy programs, and its press and propaganda wing.

PODERES POPULARES LOCALES

Recall that the FPL organized around the principle that mass participation in a prolonged popular war was the only tenable path to revolution. In stark contrast to the ERP, “the FPL never saw the zones of control as mere military rearguards[;] . . . rather, they saw them within a broader framework of political mobilization.”100 Thus, immediately following their retreat to Chalatenango, the FPL conducted an extensive analysis of the masses vis-à-vis the war.101 The authors of the report concluded that “any revolutionary movement seeking the definitive liberation of the people must develop an adequate structure allowing for the broadest participation of the masses in the war.”102 To achieve this goal, the masses needed to be organized, ideologically cohesive, and self-governing.

In the rural areas the FPL controlled, this directive took the form of what the group called Poderes Populares Locales. PPLs were grassroots structures through which locals could organize, determine the needs of the community, and devise plans for attending to those needs.103 While PPLs are “democratically elected . . . by the [local] population from their own ranks, the FPL had a heavy hand in coordinating and managing the structures.”104 Leaders across various PPLs formed the Boards of Sub-Regional Popular Power and Boards of Local Popular Power, both of which also included members of the FPL’s mass organization and the chief of the local popular militias, and they had regular meetings with subdivisions of the FPL coordinating local administration.105 The organizational complexity of the PPLs can hardly be understated, as illustrated in figure 3.3.106

The PPLs were central to the FPL’s vision of how a prolonged popular war would integrate the masses to slowly remake the country in its vision. The PPLs were explicitly forged, as one former FPL member put it, “to create real links between our zones and the rest of society.”107 As such, these local governance structures differed markedly from what minimal local organizing arose in Morazán under the ERP.108 The central role of the PPLs was not lost on the people participating, the rebels coordinating the them, or even the other OP-Ms. In what would prove to be a prescient analysis, Pearce notes that “many peasants, like the revolutionary organizations, see the PPLs as the embryo of future forms of popular local government in a liberated El Salvador, not simply as a means of solving the material problems that arise from war.”109 Fermán Cienfuegos—one of the leaders of the RN—later reflected on the importance of the PPLs and other comparable structures built during the conflict. At the midpoint of the war he referred to the FMLN as having “dispersed territorial power.” The problem, he noted, was that the breadth and quality of power they had was inconsistent, even across territories they technically controlled: “To structure this power it is necessary to create new forms of organization that reflect our true capacity, because I can have 100 people organized in a hamlet but if there is no organizational form that reflects that power objectively, then it is not real, they are just scattered masses without links or permanent contacts.”110

FIGURE 3.3. Organizational structure of PPLs illustrated in Pearce (1986). Copyright © Latin America Bureau.

In other words, organized masses were of little use unless the FMLN built dedicated structures to coordinate and integrate them into the movement. Thus, the PPLs not only occupied the critical nexus between the FPL’s organizational development and the relations they forged with the masses, but the leaders in charge of them understood their organizational and mobilizational value.

It is crucial, however, not to underestimate the challenge of creating participatory democratic institutions in a place with no legacy of political organization and a mostly illiterate population. To deal with the fist obstacle, the FPL leveraged its prewar social networks in CEBs. By fostering community, leadership, and political consciousness among campesinos, the CEBs provided an ideal medium for transferring the roles and skills from religious communities into an explicitly political sphere. In the search of PPL leaders, the rebels turned to catechists—the unordained “lay preachers,” who had received political training in the previous decade. Rather than merely recruiting from among a known network (as the ERP did in Morazán), PPLs were constructed by taking collections of individuals with relevant skill sets and transferring them into new but related roles.

You would think it was difficult to understand the tasks of administration of a place, especially for the peasants, many of them illiterate. But it really wasn’t because the people knew what the needs were. We really know the function of each secretary. For the secretary of production and popular economy, we look for and elect a compañero who knows about farming, who himself uses a machete, a plough, who understands commerce. The compañeros involved in health need preparatory courses; the auxiliary health workers know how to manage a pharmacy and give talks on health. In education, there’s usually a compa who can read, who has been to school until fifth or sixth grade and has shown willingness to teach those who don’t know.111

Indeed, the way in which the FPL leveraged extant roles and relevant skill sets from church networks to construct local political organizations unfolded in lockstep with my theoretical expectations about how proto-party structures facilitate party transformation after war.

M IS FOR MANZANA MATANZA: FPL LITERACY PROGRAMS

Existing church networks helped the FPL overcome obstacles to mobilizing an otherwise politically inexperienced community. Yet, it still had to contend with the fact that more than half of rural Salvadorans were fully illiterate.112 Given the FPL’s objective of building a highly diversified organization capable of administering territory, mobilizing the masses, and, ultimately, taking over in a governing capacity, literacy was crucial.113 In response to the inundating demand for literacy (from the organization and the people alike), the FPL forged one if its most robust social service divisions to provide literacy classes to local populations.114

Prior to the insurrection (and, thus, prior to the existence of PPLs), the FPL attempted to set up literacy programs in Chalatenango.115 These early programs were ad hoc, lacking consistency in both scheduling and curriculum. By late 1981, however, the FPL was able to leverage three organizational resources in service of creating more widespread literacy programs in Chalatenango: (1) its affiliation with ANDES (a progressive teachers association), which provided access to experienced educators, (2) the PPLs, which provided an ideal organizational medium for administering literacy programs, and (3) its international networks, which provided educational resources.116

Once literacy programs were organized and routinized, they gave rise to a positive-feedback loop between education and political mobilization in the region. A community organizer in Chalatenango recounts this process:

At first, there was an emphasis on the children . . . but many others said they wanted to learn too and the PPLs put literacy into their programmes as a fundamental element. We saw that literacy had a great deal to do with the organization of the people. . . . For the first time, there was popular government with a minimum programme which gained legitimacy as schools began to function. . . . The programme contributed to the development of the PPLs because, through literacy, people could be brought into more tasks; it mobilizes the population.117

Thus, the PPL structures not only acted as an organizational medium through which the FPL could provide education, but the education they provided strengthened the administrative wings by bringing more advanced and diverse skills. To be clear, the education-mobilization feedback loop was by design. The literacy curriculum drew on Paolo Freire’s approach, using economically and politically salient concepts to first teach the basic tenets of literacy (writing, syllables, conjugation) and then to foster conversation about the concept itself.118

PRESS AND PROPAGANDA

Beyond its administrative and social service divisions, the FPL dedicated substantial organizational resources to building a robust political-messaging wing. Even in the early years of the war, the FPL’s political messaging was not only extensive but also highly specialized. Both internal directives and the messaging that ultimately went to press suggest that the organization customized its messages and delivery to attend to the various groups of people it sought to influence. the following propaganda directive from an internal memorandum is one example:

“It is necessary to increase our propaganda, but it must be propaganda that really touches the feelings of the people, their needs and interests. . . . [We must] direct propaganda to each level: to the popular masses (the different sectors in general) and to the puppets of the regime (the leaders and authorities). . . . We must focus and concretize the content of propaganda: differentiate the treatment of each sector.”119

Crucially, this directive was not just wishful thinking. The FPL’s press division was multimedia and multifaceted, comprising a radio station (Radio Farabundo Mart´ı), a domestic press wing (The Salvadoran Press Agency, or SalPress), the Revolutionary Cinematographic Institute, and numerous divisions dedicated to internal messaging (e.g., the Farabundo Martí Documentation Center).120

Moreover, in the spirit of autocrítico, the FPL’s internal memoranda consistently pushed to evaluate and improve its political-messaging operations. In a document titled “Assessment of the Politico-Military Work in the Disputed Zones,” two pages out of six are dedicated expressly to plans for improving propaganda, political education, and political work among the masses.121 Plans for improvement range from the organizational (“we’ve had improvement in the propaganda infrastructure . . . but we still have organizational difficulties . . . in the propaganda teams that prevent us from maximizing exploitation of [the infrastructure]”) to the logistical (“we are lacking in typewriters . . . [and] we need to expedite the plans to obtain mimeographs”) to the content itself (“there’s been an improvement in the presentation and content, but we need more concrete and updated content . . . as well as clear follow-up”).122

During the first few years of its life, the FMLN was a chimera—some parts of it bore so little resemblance to other parts that one would hardly recognize them as coming from the same animal. The early organizational trajectory of different OP-Ms sheds new light on the scope and implications of suborganizational variation. This analysis focused on the FPL and ERP, but the structural divergence was by no means unique to these two groups. Rather, the FPL and ERP are analytically instructive because they exhibited negligible differences in their origins: both groups emerged in rural areas among poor campesinos, both had strong ties to CEBs, and both had well-educated leadership. It is precisely the extent of their similarities that throws their differences into even sharper relief.

This analysis of the FMLN’s postinsurrection trajectory has a few important takeaways. First, this section provides concrete empirical evidence that rebel groups build functioning (rather than nominal) proto-party structures and that their leaders approach organizational development with clear intentions. Second, the divergent trajectories of the ERP and FPL provide strong evidence that the process of organizational development proceeds in a manner consistent with my theoretical framework. Even in similar contexts, different leaders will exploit the same types of networks to very different organizational ends, which we observed in the differential use of CEBs in Morazán and Chalatenango. The third and broadest implication of this analysis is that it highlights the importance of expanding our analytic vocabulary to get traction on consequential differences both within and between organizations. Limiting ourselves to conventional organizational descriptors like command-and-control relations and cohesion, would erroneously lead us to conclude that the ERP and FPL are considerably more similar than they are. In reality, the two have wildly different organizational resources, advantages, and weaknesses as they head into the second phase of the war.

Organizational Adaptation (1984–1989)

In the first phase of the war, the FMLN made considerable military gains against the Salvadoran armed forces. Initially, its steadfast military focus paid off, and by 1983, the army was on the defensive. As a result, the insurgents had vastly expanded the territory under their control.123 Heading into 1984, however, two related shocks to the wartime environment posed severe threats to the rebellion. Beginning in the mid-1980s, the United States underwrote a massive counterinsurgency campaign on behalf of the Salvadoran government. The influx of material support and training resulted in the government gaining the upper hand on the battlefield. To boot, US support came at a cost to the Salvadoran regime: Reagan put pressure on the Salvadorans to open the political sphere and curtail (or at least hide) human rights abuses.124 The government complied, albeit begrudgingly. Forced disappearances declined tenfold and the 1984 elections were the most competitive in over fifty years.125 In their wake, José Napoleón Duarte—the new president—reinstated the right to organize opposition groups, which had been illegal since 1931. Significant military, political, and social shocks followed, and, as we’ll see, the FMLN was forced to adapt in the face of adversity.

The augmented counterinsurgency strategies and Duarte’s political reforms posed three challenges to different dimensions of the FMLN’s power: one military, one political, one social. From a combat standpoint, the FMLN found itself on the defensive for the first time in two years. Given the military’s sudden capacity to wage a high-technology war, the FMLN’s large, stationary battalions left them vulnerable to massive losses from a single strike. Even OP-Ms with popular support could not recruit fast enough to replace the fallen, and recruitment was even harder in Morazán where the ERP’s ties to local populations had atrophied substantially.

On the political side, the most consequential shock was the newfound right to organize political opposition groups.126 Of course, the FMLN’s expansive political roots may cast the political opening in an auspicious light. However, it would be a mistake to assume that legal protections for political organizing were a foregone benefit to the rebels. In reality, thawing political rights could easily go one of two ways. On the one hand, of course, the FMLN could try to adopt a new strategy emphasizing political engagement and taking advantage of a newfound ability to mobilize supporters. On the other hand, the opportunity to organize legally meant that leftists could view the political opening as a reason to shy away from supporting a violent revolution, and opt instead to work toward gaining representation through legal channels.

Finally, on the social front, subsiding government violence resulted in an influx of former refugees returning to their homes. Individuals in the refugee camps had experienced their own organizational renaissance—spearheading literacy programs, popular education, self-governance (autogestacíon)—and they maintained close ties to the FMLN.127 Moreover, they harbored fairly strong antigovernment sentiments. Still, their return was not a boon by necessity. In many cases, former refugees were returning to territory now under FMLN control, which meant that the social and administrative tasks of repatriation fell to the OP-M in charge. If managed well, returnees brought with them a host of “advanced technical skills” that the FMLN desperately needed as it sought new recruits.128 Yet, the complications were not merely logistical. Research at the fore-front of refugee studies suggests that repatriation often gives rise to local conflict between those who left and those who stayed, which places additional demand on the repatriating communities to mitigate potential conflict.129

This period was a multidimensional stress test for an organization that had only recently found its footing. While the FMLN successfully reinvented itself after the failed insurrection in 1981, this situation was different. Rather than facing one immediate and profound failure head on, it now confronted a series of disparate challenges (and opportunities), any one of which could easily fly under the radar. Without radical change, the FMLN was, as Bracamonte and Spencer described it, on “a collision course with disaster.”130

THE SHIFT TO “A PEOPLE’S WAR”

Depending on their capacity to assess and adapt to this new environment, the sociopolitical changes the rebels faced could either be an asset to or the Achilles heel of the revolution. Whether the FMLN could exploit the opportunities was a matter of their ability to see what they needed and their ability to see what they had. The question is, how fine-tuned was their practice of self-reflection? I argue that two organizational features paved the way for the FMLN’s successful navigation of Duarte’s rise to power (and everything that came with it). First, their long-standing commitment to autocrítico created an institutionalized practice of searching for ways to perform better. Second, notwithstanding the official orders to focus on conventional military engagement, the FMLN comprised groups with diverse tactical, political, and social approaches to conflict. As a result, quickly changing its strategy meant it had the organizational resources to pivot to a different yet existing approach, rather than designing a new one from scratch.131

In keeping with the culture of autocrítico, the General Command quickly acknowledged the need to change strategies. It concluded that in spite of the rebels’ territorial gains, the directive guiding the first phase came at substantial political costs. Realizing that they had focused too heavily on combat to the exclusion of mass mobilization, the new strategic directive took the FPL’s approach as a blueprint for the second phase of war.132 In an internal memorandum describing their new approach, FMLN leaders describe their shortcomings with astonishing clarity:

Accepting the illegality of the masses was always a wrong criterion [on which to base our fight]. At the time, our approach was successful . . . but thousands of people were left without real perspective of the fight, many of whom remained with the FMLN only out of fear of repression or because they were forced by their circumstances. . . . Fundamentally, they have not had a concrete political practice to raise their levels of consciousness—they have only known a life of hiding and flight.133

Of course, recognizing their weaknesses was only half the battle. What remained was the Herculean task of overhauling the organization in the midst of war. The directive was twofold: restructure the military to fight a war of attrition and bolster political subdivisions to foster more informed political engagement from both members and the masses.

On the battlefield, zonal leaders were ordered to break up large battalions into small cells in service of implementing the FPL’s prolonged popular war strategy.134 The changes that the FMLN made to its military apparatus were substantial and the results were even more so. Not only were large battalions broken into smaller units, but the shifting emphasis from the battlefield to the political sphere meant that the FMLN drastically reduced its armed wing from around twelve thousand members at its peak in 1983 to only six thousand by 1987.135 Despite halving their forces, the FMLN’s reach was far greater in this period. With smaller, more mobile units, the group was able to infiltrate (to some degree) every department of El Salvador including the capital and other major cities.136

The most profound shift during this period was the FMLN’s pivot toward local political engagement, which itself entailed multiple organizational changes.137 The new phase of the war was characterized by a massive uptick in the FMLN’s investment in its political subdivisions. Their goals were twofold: bolster political education within their own ranks, and use these new structures to forge links with and mobilize civilian populations. Implementing this plan entailed building structures dedicated to expanded political messaging, political education, and administration and governance. Yet, these structures were not born overnight. To varying degrees, the FPL, the PRTC, the RN, and PCS had spent the last decade building up politically relevant subdivisions, thereby laying a crucial foundation for the new directive.138 Confirming the mechanism of organizational change, Wood explicitly argues that the FMLN’s capacity to pivot toward political engagement was “the diversity of relations between the FMLN and various organizations.”139

The FMLN’s approach to mobilizing the Salvadoran population in this period was expertly crafted to exploit the new political freedoms accompanying Duarte’s inauguration. The General Command displayed a keen awareness of the potential ramifications that followed from legalizing opposition groups. As they saw it, the only way out was through: “Legality then becomes the fundamental weapon of the masses.”140 Rather than passively allow competing opposition groups to spring up and distract from the revolution, the plan was to create those groups themselves. Across the country, the FMLN enacted a strategy called poder de doble cara (literally, “two-faced power”). It encouraged the masses to organize around locally salient issues—unions and workplace associations in the cities and agricultural cooperatives in the countryside. Then, the organizations would present the neutral, legal face of the organization to the state while maintaining clandestine ties to the FMLN.141

The organizational demands of this new approach cannot be understated. Across the FMLN, the OP-Ms had to build structures dedicated to providing internal political education, mass political education, and enhanced propaganda, as well as administration and coordination of the new mass organizations. Evidence of this structural augmentation abounds.142 Figure 3.4 illustrates expansion plans for the PRTC. They address education, propaganda, further integration with the University of El Salvador, mass organization, and reintegration plans for returning refugees. Moreover, despite building myriad structures with fairly specialized tasks, we observe explicit care to avoid organizational rigidity—a known impediment to future changes. In her diaries, Díaz writes that the goal is to build structures that are flexible pero solidas—flexible, but solid.143 A later, more formal memorandum from the PRTC echoes this sentiment:

FIGURE 3.4. PRTC organizational evaluation and plans (Díaz, 1982–1984). Wartime Notebook, Nidia Díaz Papers, box 1, folder 8, Hoover Institution Library & Archives.

“These organizational forms will have to be flexible. We must seek to insert them into the state structures of the regime and take advantage of that infrastructure to bolster the security of our cadres and the masses. From there, we can exercise judicial and political control, not as an end, but as a means to facilitate and open more space in service of the organizational political work of the masses.”144

In addition to catalyzing and incorporating new forms of mass organization, the General Command and the individual OP-Ms were quick to address the organizational challenge of repatriation.145 As early as mid-1983, the FPL had already developed an elaborate plan addressing the relocation and reintegration of refugees.146 Figure 3.5 suggests that the PRTC had as well. These plans had social, political, military, and economic components—ensuring that the repatriated families had appropriate provisions (e.g., clothing and shoes), were integrated into the PPL structures, and were integrated into and under the protection of the popular militias.

Poder de doble cara had two important implications for the organization’s trajectory. First, the directive meant that building proto-party structures was now a matter of compliance—rather than defiance. To implement doble cara properly, the OP-Ms had to “create wings” to “integrate and mobilize the masses” and to build “schools for political instruction.”147 Implementing these structures in practice meant that the organizations also had to focus recruitment efforts on targeting high-skilled individuals, which the PRTC documents directly.148 Thus, by expressly seeking to staff technical subdivisions of the organization, the FMLN’s recruitment profile in the second phase of the war departs sharply from our conventional understanding of who insurgents target.

FIGURE 3.5. PRTC organizational evaluation and plans (Díaz, 1982–1984). Wartime Notebook, Nidia Díaz Papers, box 1, folder 8, Hoover Institution Library & Archives.

A Glimpse Back at Morazán

The political focus of the FMLN’s new strategy (to say nothing of its explicit pivot toward the FPL’s approach) raises an important question about the extent to which it was evenly implemented across OP-Ms. While cohesion among the groups increased in this period, differences on the ground persisted. Given the ERP’s continued reluctance to engage in substantial political work on top of its atrophied ties with local populations in its rearguard, I anticipate that its organizational development will not follow the same trajectory as the others. Evidence suggests that while the ERP ramped up political activity, it exhibits far less evidence of substantial organizational changes than the other OP-Ms.

Keeping in line with the doble cara directive, the ERP strategically exploited land reform measures and encouraged campesinos to form agricultural cooperatives, which ultimately resulted in some of the most influential peasant cooperative organizations in the country.149 However, as Binford and others suggest, the ERP’s relationship with these organizations was more hegemonic than in other regions.150 For example, Wood observes that many peasant cooperatives were run by “strategically placed ERP political officers” and, at least initially, were fronts to “expand ERP influence.”151 Binford even goes as far as to argue that the “authoritarianism practiced by the ERP . . . restricted the political space within which relatively autonomous organizations might have proliferated,” thereby suggesting that Morazán’s newfound level of community organization occurred despite, not because of, the ERP’s involvement.152 Moreover, the ERP was more reactionary and opportunistic than other OP-Ms: community programs would emerge independently of the ERP (e.g., literacy programs or the Pan y Leche program aimed at feeding schoolchildren), and only “once programs demonstrated tangible benefits” would the ERP “defend them from assaults by the state.”153 Thus, while the ERP’s implementation of doble cara and its broader move to embrace political action succeeded in forging ties to political organizations, the nature of those ties is better characterized in terms of exerting local control than integrated participation.

Tracing the FMLN’s wartime evolution reveals a host of unforeseen nuance in its organizational composition. The divergent trajectories of the OP-Ms in the first phase of the war illustrates how and where proto-party structures took root. In this era, the insurgency’s diversification into political domains was present, but uneven. Every OP-M under the FMLN umbrella built robust press and propaganda wings.154 The FPL as well as the PRTC built subdivisions dedicated to mass mobilization, governance, and social service provision (especially education and health care).155 The ERP, however, focused mainly on building out combat units and logistical subdivisions to support the fight. As such, this analysis shows that rebel organizations invest heavily in building structures that mirror the content of party organizations, but not uniformly so.

The FMLN’s pivot in the second phase demonstrates the continuity and expansion of proto-party structures across the organization, as well as its capacity for flexibility and adaptation under duress. Many groups expanded their political influence in this era, laying the groundwork for future party capacity and voter mobilization. Additionally, the high command’s institutionalized willingness to critique its performance and seek alternative paths to success not only served the organization well during the war, but I argue it also sets it up for success beyond the battlefield. Reflecting on the FMLN’s performance at the five-year mark, the ERP leader Villalobos remarked, “We have been able to learn from every error . . . based on a serious self-critical spirit.”156 In the face of three major shocks to their environment, a set of disparate organizations—purportedly pegada con saliva—were nonetheless able to identify their shortcomings and devise a series of plans to combat them. The FMLN’s deft reaction to facing a radically altered environment demonstrates not only its flexibility, but also what we might call its organizational proprioception: Leaders consistently demonstrate a keen awareness of what their organizations have, what they lack, and where their different “limbs” exist in the spaces they occupy.

Finally, this analysis of the FMLN’s parallel organizational trajectories yields a set of clear predictions about which groups will be most equipped to pivot from the battlefield to the ballot box, which I explore in the following chapter. Given that organizational change is easier to the extent that rebel groups have preexisting competencies, OP-Ms with robust proto-party structures should be the ones spearheading the transformation. Specifically, I expect to see the FPL, PRTC, and PCS come out on top, while the ERP—despite its auspicious beginnings—is more likely to be marginalized in the party-formation process.

Tracing the Development of Proto-Party Structures

Understanding rebel-to-party transformation first requires moving away from the tendency to bracket the organizational legacies of war. Features of rebels’ wartime organizations not only facilitate party building, but also may well endure beyond the conflict theater. Taking a more comprehensive approach to modeling the wartime evolution of rebel organizations reveals unforeseen variation on two important dimensions, which, together, will explain the rebel-to-party process in the following chapter. First, many rebel groups build structures that mirror the form and function of those in political parties. Second, some organizations exhibit and institutionalize flexibility, which makes adaptation to shocks more efficient and less risky.

Beyond elaborating the explanatory variables of the theory, the framework and analysis in this chapter highlight the importance of expanding our organizational vocabulary. Critically examining the content of rebels’ noncombat structures reveals stark and systematic variation in how rebels organize, operate, and evolve. These insights put flesh on the bones of existing explanations of militant group structure and civilian mobilization. Absent the nuance the new organizational approach affords, even scholars with extensive case knowledge are liable to draw incomplete or misleading conclusions. For example, in describing the FMLN’s organizational structure, Jocelyn Viterna reports the following: “The FMLN guerrilla army was exceedingly hierarchical. At its apex was the ‘General Command,’ comprising the top commander from each of the five component groups. . . . Although each branch raised its own money, secured its own weaponry, and structured its own forces, they appear to have developed very similar organizational structures.”157

Again, Viterna is not wrong, per se. The FMLN’s subdivisions do exhibit structural similarities on a number of core dimensions: their hierarchical arrangement; the division of OP-Ms into political, military, and mass organizations; their disciplinary institutions. What this chapter reveals, however, is that on many other dimensions, those similarities are nowhere to be found.

Similarly, tracing the development of proto-party structures in Chalatenango, for example, elucidates one path by which the mechanism of Huang’s civilian mobilization argument may play out on the ground. Yet the parallel trajectories of the FPL and ERP reveal a source of systematic variation in where that mechanism plays out—as local mobilization was considerably more robust under FPL rule. She argues, “Where rebels tap into civilians as a significant war-making resource, the latter become politically mobilized.”158 The framework developed here demonstrates that civilian mobilization is likely achieved through the structures rebels build to interface with local populations. Where those structures are less developed (let alone exploitative), civilian mobilization likely wanes as well.

The framework developed here sheds light on why Díaz scribbled “¿Que falta?” at the peak of the FMLN’s military success. In contrast, existing conceptions of rebel organizations are not equipped to answer “What’s missing?” and even more importantly, they are not equipped to shed light on what’s not as militant groups stand on the line between military campaigns and political ones. The central takeaway from this chapter is not that not all rebels suddenly need to “learn the art of non-violent politics,” as Close and Prevost suggest. Rather, it is that they often learn the art of nonviolent politics while simultaneously practicing the art of war.