The Uneven Impacts of Violence against Women Reform in Guatemala: Intersecting Inequalities and the Patchwork State

In 2008 Guatemala passed legislation criminalizing various forms of violence against women and mandating the creation of courts that would specialize in such violence. This article demonstrates that these reforms’ impacts were unevenly felt, with the most marginalized benefiting the least. It explains this gap by drawing on a historical intersectional analysis, and reveals the importance of including place alongside more commonly studied categories of difference. It also illuminates the sources of the gaps between policies and their impacts. It finds that many Guatemalan women remained marginalized from security and justice, despite attempts to protect them, for two reasons. First, the reforms isolated violence against women from other structural sources of inequality, thus undercutting their impacts for multiply marginalized women. Second, the reforms’ impacts were undermined by historically constituted patterns of state-society relations and the uneven nature of the Guatemalan state.

On April 10, 2008, the Guatemalan Congress passed the Law against Femicide and Other Forms of Violence against Women (Ley contra Femicidio y Otras Formas de Violencia contra la Mujer), which established violence against women (VAW) as a crime rooted in "unequal power relations between women and men, in social, economic, judicial, political, cultural and familial realms." Aiming to overcome long-standing impunity, it criminalized various forms of VAW, established the gendered killing of women as a unique crime (femicide 1 ), and mandated the creation of specialized courts (Tribunales de Sentencia Penal de Delitos de Femicidio y Otras Formas de Violencia contra la Mujer) that would focus exclusively on VAW. A few years later, two hundred miles from the congressional floor, an indigenous teenager living in the rural municipality of Santa Nimá was assaulted by her neighbor. What were the effects of national-level reforms for this young woman and others from communities like Santa Nimá? In addressing this question, this article answers the feminist call to center policy analysis on the needs of the most vulnerable women (Smith 2006) and draws on two approaches to state reform. First, it applies a historical intersectional analysis, which acknowledges the

Historical Intersectional Analysis and VAW Reform
Much of the literature on VAW has focused on the nature and sources of gendered violence (Sanford 2008;Menjívar 2011), the variations in the timing and comprehensiveness of VAW reforms (Weldon 2002;Walsh 2008;Friedman 2009;Franceschet 2010;Htun and Weldon 2012;Weldon and Htun 2013;Htun and Weldon 2018), and to a lesser extent, the factors that affect implementation (Montoya 2013). Fewer works have moved beyond the passage and implementation of reforms to their varied effects. This is a critical gap because even well-implemented laws may not have the intended effects, as demonstrated by the negative impacts of US criminal justice responses to VAW in communities of color (INCITE! 2006;Bumiller 2008;Potter 2008;Goodmark 2017). Historical intersectional analysis focuses on how individual lived experiences and acts of violence are embedded in historically constituted structures of power organized around gender, class, ethnicity, and place (Crenshaw 1989(Crenshaw , 1991Collins 2000;Sokoloff and Dupont 2005;Deer 2015). This type of analysis is critical to understanding the reforms' impacts because the same webs of power that systematically expose some to violence also systematically limit some people's access to institutions of security and justice and put them at a higher risk of "everyday and routine abuse" in those institutions (Menjívar and Walsh 2017, 235).
While it is well documented how gender, class, race, and ethnicity shape the experiences of VAW, less recognized is the importance of place and how urban/rural inequalities shape exposure to violence or access to justice (Sandberg 2013). This is a critical deficiency as "urban and rural are not only neutral descriptions of territories but are also linked to power as well as social inclusion and exclusion" (Sandberg 2013, 251). People in rural areas are disadvantaged because they are geographically distant from critical services, and the institutions closest to them tend to be under-resourced and less specialized to meet their needs (Sandberg 2013;Sosa 2017). Small rural communities offer less privacy, increasing the social stigma women face when reporting VAW and the likelihood that they have to report abuse to someone who knows them or their abusers (Sandberg 2013).
In a rare study of the impacts of Guatemala's VAW reforms, Shannon Drysdale Walsh and Cecilia Menjívar (2016) examine two families' pursuits of justice in Guatemala City following the murder of female relatives. One family is working class, living in a semirural area outside of the capital; the other is middle class, located in the urban periphery of the capital. Walsh and Menjívar (2016, 32) explore the classed and gendered legal tolls that families face, including overcoming fear in a context without protection for victims and witnesses, and expending time and energy navigating a "legal system that is a labyrinth of bureaucracy and discrimination." They highlight similar barriers that both families face despite their differences: police who lack commitment to enforcing laws, overburdened public prosecutors who need to be pressured to pursue investigations, and delays in hearings. Yet they also implicitly demonstrate what is rendered explicit here: the need for an intersectional account of the experience of pursuing justice. For example, the authors find that the rural family confronts such under-resourced police that the victim's family must secure the police a car and food before they agree to arrest their daughter's murderer. This reflects the reality that while state institutions are under-resourced everywhere, the dearth of resources is particularly severe outside of the capital (where attention, personnel, and money are concentrated). Walsh and Menjívar also note that the middle-class family is able to make the multiple visits to state agencies to pressure for investigations in part because they have more resources than do lower-class families. They thus indicate that high legal tolls are experienced differently depending on claimants' class position and that these legal tolls may be effectively insurmountable for many poor and working-class families. Both of these findings indicate the need for an explicitly intersectional analysis that considers the ways that class, gender, and place (alongside ethnicity, which is not explored by the authors) shape access to justice.
In Guatemala, inequalities based in gender, ethnicity, class, and place overlap to limit the rights and well-being of poor indigenous women living in rural areas. Roughly half of Guatemalans live in poverty, but economic vulnerability disproportionately affects the country's indigenous populations, which are overrepresented in rural areas where state services are missing, under-resourced, and/or corrupted. The intersection of indigeneity, rurality, and poverty can be seen by comparing rates of extreme poverty (Figure 1).
These inequalities intertwine with gender to further disadvantage women. Guatemala trails all Latin American countries on the gender inequality index, which considers gaps in employment, education, representation, and reproductive rights (UNDP, n.d.). Guatemalan women have the lowest rates of labor market participation regionally (37.4 percent); in rural areas these rates drop to 28.1 percent. More than half of women working in agriculture do not earn their own income because their labor is seen as part of that of their male kin (USAID 2018). Ethnic discrimination also hurts women disproportionately, as can be seen by comparing literacy rates: 88.9 percent of nonindigenous men and 81.1 percent of nonindigenous women are literate, compared to 74.6 percent of indigenous men and 51.9 percent of indigenous women (Orozco and Valdivia 2017). Everywhere, women experiencing abuse struggle to exit relationships on which they are economically dependent. Yet dependence is heightened in rural areas, where there are larger gender gaps in income and education, and in indigenous areas, where women face additional linguistic and cultural barriers to leaving their communities.
Reforms to combat VAW and increase access to justice in Guatemala were embedded in this broader context, which robbed them of their full impact. The government pursued criminal justice reform related to VAW but failed to tackle widespread discrimination or establish a functioning welfare state. Many VAW reforms focused on protecting women as a category, overlooking the ways that aspects of marginalization  (UNDP 2009, 9). based on gender, class, ethnicity, and place might intertwine to limit access to state institutions. 3 In contexts where systems of domination converge, such interventions to address VAW that artificially separate genderbased marginalization from that rooted in other categories of difference ineffectively address the obstacles to security and justice faced by multiply marginalized women (Crenshaw 1991(Crenshaw , 1246.
By highlighting states' uneven protection of rights across groups, historical intersectional analysis challenges unitary views of the state. It reveals the Guatemalan state to be uneven, rather than "strong" or "weak," in its capacity and will to perform its functions across groups, space, and institutions. The Guatemalan state has been capable of eliminating over six hundred villages, monitoring even remote communities, and achieving goals consistent with neoliberal democracy (Burrell 2013, 118). Yet it also underfunded and underinstitutionalized social services in rural and indigenous areas and failed to uphold the rights of its female, poor, and indigenous subjects. Instead, communities like Santa Nimá and people like Virginia fell into "brown areas"-the "frontiers of various regions and class, gender and ethnic relations" in which "components of democratic legality and, hence, of publicness and citizenship, fade away" (O'Donnell 1993(O'Donnell , 1360(O'Donnell -1361. Brown areas are difficult to eliminate because they are rooted in historical processes that have shaped the frameworks and expectations of state and societal actors alike. Menjívar and Walsh (2016, 10) argue that state institutions reproduce inequalities even in the face of equalizing reforms because state officials have internalized societal frameworks and "have decades of experiences applying discriminatory [laws]." This institutional "stickiness" is all the more likely when reforms touch on the most intimate aspects of people's lives-their relations with their partners and kin. Disrupting perverse state-society relations is also difficult because, just as many officials have been socialized in discriminatory institutional and societal cultures, many citizens-most notably the poor, indigenous people, women, and those living in rural areas-have learned to see state institutions as irrelevant or threatening to their needs.
The government tried to circumnavigate the distrust of existing institutions, and problematic cultures within them, by creating new institutions with targeted mandates and specially trained employees. Many officials in new institutions challenge broader societal frameworks as well as those of under-reformed state agencies. Yet because of the lack of more widespread reforms, women experience a "patchwork" state in which institutions that have internalized VAW reforms exist alongside institutions that have not. Women's experience of this patchwork depends on their geographic and sociocultural location. Notably, women living in rural indigenous communities have the most contact with state institutions that are inadequately resourced and reformed to meet their particular needs, and they are therefore unlikely to change their view of the state as ineffective or untrustworthy.

The Roots of Marginalization and Violence
This account moves past essentialist depictions of "Guatemalan culture" (Narayan 1998) to trace the historical roots of inequalities based in gender, ethnicity, class, and place that have intertwined into a complex knot that is difficult, if not impossible, to untangle. It demonstrates that these inequalities were imbricated in the lives of women like those of Santa Nimá in ways that disproportionately exposed them to violence in public and private spheres, excluded them from state and nonstate systems of justice, and established patterns of state-society relations characterized by discrimination and distrust.
In colonial and postcolonial Guatemala, gender, class, and ethnic oppressions did not exist in parallel. Instead, each was "intimately bound up with the others, resting on the same scaffolding of structural inferiority and manifested in daily violence that enforced domination and submission" (Forster 1999, 58). In indigenous communities like Santa Nimá, minority ladino (nonindigenous) elites dominated local politics and economics, leveraging gendered, racialized, and classed violence. They forced indigenous labor on public works and plantations, where overseers used sexual violence against women as a tool of ethnic and class domination (Sieder 2013). Despite their prevalence, cases of sexual violence were rarely heard in official courts. When they were, prosecutions were limited to cases in which victims were young virgins and focused on the loss of virginity rather the lack of consent (Forster 1999;Komisaruk 2008). The criminal code reflected this history, labeling rape as a crime against women's honor (pudor), leaving unprotected those who were seen as lacking honor-indigenous women and lower-class ladinas-who courts allowed to be abused with impunity (England 2014). During proceedings, even monolingual Mayan victims were forced to undergo court-ordered gynecological exams performed by Spanish-speaking male doctors (Carey 2013), based on the assumption that "girls' bodies-the appearance of their hymens … had greater evidentiary and legal status than did [their] narratives" (Komisaruk 2008, 280). Such practices, which alienated indigenous victims and prioritized bodily evidence over testimonies, are mirrored today.
Nonstate institutions also affected indigenous women's lives. Especially in rural areas, indigenous leaders up until the mid-twentieth century could wield spiritual and communal authority and resolve many intracommunity conflicts if they avoided challenging ladino elites' power (Hessbruegge and Ochoa García 2011). Yet this alternative sphere of justice, while less affected by ethnic and class discrimination, failed to adequately address VAW. Instead, customary law often recognized a man's right to abuse his wife and implicitly taught women "that they must always obey their husbands and endure occasional beatings" (Carey and Torres 2010, 146). Thus both official and communal institutions encouraged indigenous women to endure VAW as a fact of life.
The armed conflict between military-led governments and leftist guerrillas  transformed gendered, classed, racialized, and place-based violence; weakened communal authorities; and taught women, the poor, and indigenous rural communities to distrust the state. At the height of violence (late 1970s/early 1980s) the state deployed scorched-earth strategies in indigenous, rural areas to exterminate presumed guerilla sympathizers. These genocidal strategies "deliberately destroyed the social fabric" of communities, "supplanting locally legitimate institutions with militarized alternatives" (Godoy 2006, 34). During this time, the military targeted Santa Nimá, burning houses, raping women, torturing and/ or disappearing those suspected of subversion, and forcibly enlisting men into paramilitary organizations, known as Civil Self-Defense Patrols (Patrullas de Autodefensa Civil, PACs). Some residents, as part of family or land disputes, informed on their neighbors to members of the PACs, military, or guerrillas, decreasing levels of intracommunity trust for years to come. Militarization of authority in this municipality, alongside the suppression of communal systems, left residents here and elsewhere "without peaceful and culturally appropriate means of regulating their coexistence" following the conflict (Sieder 2011, 50).
Violence rooted in ethnicity, class, and place intertwined with gendered violence. In Santa Nimá, police, military, and paramilitaries used VAW to punish women's male kin for suspected acts of subversion, contributing to a distrust of state forces and further normalization of VAW. The report of the United Nations Commission for Historical Clarification (Comisión para el Esclarecimiento Histórico, CEH), Guatemala: Memory of Silence, found that military and paramilitary forces across Guatemala used sexual violence as a weapon of war, often during community sweeps and massacres in rural areas like the one in Santa Nimá. Over one hundred thousand women and girls were raped in Guatemala, over 99 percent of them at the hands of state or PAC forces. The sexual violence deployed during this time was racialized; 90 percent of victims were Mayan (CEH 1999).

Postconflict Reforms' Uneven Effects
The democratic opening (1985) and the 1996 Peace Accords generated pressures to enhance the rights of those most marginalized and affected by violence. Yet subsequent reforms created institutions that inadequately met indigenous, rural communities' needs; concentrated the most effective institutions in urban centers; and were pursued without addressing structural inequalities. This reality undercut the impacts of subsequent VAW reforms for poor, indigenous women in rural communities like Santa Nimá, which were traumatized by war, robbed of effective communal authority structures, and distrustful of state institutions.

Postconflict state-society relations in rural municipalities
Postconflict governments established new institutions of security and justice in rural communities, with mixed results. In the area of security, the government dissolved military commissions and PACs and dispatched officers of a reorganized National Civilian Police force (Policía Nacional Civil, PNC) to even remote municipalities. Despite Article 2 of the National Civilian Police Law, which called for police to reflect "the multiethnic and pluricultural character of Guatemala," officers were overwhelming ladino, male, and Spanish-speaking. Indigenous, rural communities were unenthusiastic about police given the force's makeup and their wartime experiences. Santa Nimá residents found the police to be understaffed, undertrained, corrupt, and lacking the linguistic and cultural capacity to effectively provide security. The linguistic and cultural barriers disproportionately affected women, who were less likely to speak Spanish or feel comfortable seeking help from ladino men. Police also abused their authority. As elsewhere, they demanded and accepted bribes (CNSAFJ 2007, 26). They were also accused of assaulting women. According to asylum declarations from former Santa Nimá residents, in a ten-year period there were over twenty accusations of police raping or threatening to rape women in Santa Nimá, mostly women who were reporting domestic violence or whose kin were perceived to be former guerrilla sympathizers.
Disappointed with police, residents of Santa Nimá created community watch groups (comités de seguridad) with mandatory rotating participation for men. While some residents appreciated the groups' linguistic, cultural, and local knowledge, others complained of their use of corporal punishments, illegal detention of suspects, and harassment of those who refused to participate. Asylum cases from the area indicate that among watch leaders were former PAC members and that like police, watch members sexually harassed or assaulted women, at times to punish their male kin who criticized them. 4 Postconflict attempts to enhance access to justice also failed to fully address the needs of poor indigenous women in rural communities. Justices of the peace were created in every municipality to adjudicate minor crimes and disagreements. These authorities-mostly ladino and Spanish-speaking-usurped the roles of indigenous authorities, sometimes causing local conflicts (CNSAFJ 2007, 16-17;Hessbruegge and Ochoa García 2011). In the early 2000s, only nine of the twenty-one municipalities in Quiché and only two of the thirty-two municipalities in Huehuetenango had bilingual judges despite the fact that 89 percent of Quiché residents and 65 percent of Huehuetenango residents were indigenous (DEMI 2007, 30). Some Santa Nimá community leaders distrusted the newly established justice of the peace because judges were ignorant of their local customs and rarely employed people who spoke their language. This linguistic gap disproportionately affected women who were less likely to speak Spanish. While some indigenous women appreciated the new avenue for child support and alimony claims, they found that judges offered inconsistent support in the face of VAW. Some judges encouraged reconciliation with abusive partners, even after the 2008 law barred such practices. The head of a state oversight agency admitted in 2017 that municipal-level judges sometimes acted in contravention of the law because they were "structurally … products of this society, of machismo, of a patriarchal society and a patriarchal system"demonstrating how uneven internalization of reforms across state institutions can actively undermine them.
Some in Santa Nimá looked to communal authorities such as auxiliary mayors to resolve conflicts, but the armed conflict had dramatically weakened these structures, thus robbing Santa Nimá of meaningful alternatives. Even municipalities with more robust indigenous systems of justice, like Joyabaj and Santa Cruz del Quiché, also inadequately addressed VAW. There, residents were reluctant to report, and indigenous authorities were reluctant to hear, cases of rape or domestic abuse. Those who reported these offenses to state or indigenous authorities experienced criminalization and secondary victimization (ASIES 2008, 53-57). Some indigenous leaders across the country have since begun to creatively combine women's rights with reconstituted indigenous justice systems (Sieder 2013), though these innovations have not impacted Santa Nimá.
Postconflict governments also established municipal women's offices (Oficina Municipal de la Mujer, OMM) 5 to increase women's political participation. In Santa Nimá, women experiencing abuse looked to the OMM because they trusted employees who were community members and spoke the local language. Employees provided informal support, often challenging other local institutions in doing so. They often directed women to institutions in the departmental capital rather than the local justice of the peace, who might encourage reconciliation. Employees occasionally accompanied women on the two-hour trip until the mayor banned the practice to avoid getting involved in family conflicts. Santa Nimá OMM staff thereafter occasionally and secretly donated their personal money to cover women's transportation costs. Still, the OMM's effectiveness was undercut by frequent turnover and inadequate funding (MESECVI 2017, 12). Conflicts between the OMM, mayor, and justices of the peace demonstrate how state institutions often work at cross-purposes, giving women an uneven experience of the state.
While the postconflict period saw some institutional creation and reform, attempts to challenge structural inequalities affecting communities like Santa Nimá were lacking. Recognizing the roots of the conflict in inequality, the Peace Accords mandated the government to increase tax collection to spend more on health, education, and other equalizing social services. Yet tax reform and social spending stalled even as neoliberal reforms focused on privatization and trade liberalization succeeded. Guatemala continued to collect one of the lowest tax revenues worldwide (13 percent of GDP) and to have the lowest levels of social spending in the region (7 percent of GDP). Fiscal policy and social spending had essentially no effect on overall or ethnic inequality (Cabrera, Lustig, and Morán 2015) and contributed to the systematic under-resourcing of the institutions of security and justice that were critical to addressing VAW. In Santa Nimá, economic assistance came only from international nonprofits that promoted the cultivation of export crops, which served to increase the community's vulnerability to international price fluctuations and failed to make a dent in rates of unemployment (66 percent), child malnutrition (67 percent), or poverty (89 percent) (2009 figures).
Women in this context remained economically dependent and struggled to escape abusive situations or pay the costs associated with pursuing justice.

Out-of-reach specialized institutions to address VAW
Despite changing rhetoric around women's rights, many postconflict state institutions remained stubbornly complicit in VAW and impunity. The criminal code reflected early twentieth-century frameworks, defining sexual assault as a crime against women's honor until 2009 and allowing rapists to escape charges if they married their victims until 2005 (Godoy-Paiz 2008). One of the first significant pieces of VAW legislation, the 1996 Law to Prevent, Punish, and Eradicate Intrafamilial Violence, established victims' rights to restraining orders and mandated police interference in abuse. But it failed to criminalize VAW and was only weakly enforced by police and judges who admitted reluctance to interfering in "private matters" (Walsh 2008;Trujillo 2010). Police and investigators often dismissed women's murders as unimportant and departed from investigatory protocols in femicide cases (Sanford 2008). In 2007, over six hundred women were killed; 70 percent of their murders went uninvestigated, and no arrests were made in 97 percent of cases (Prieto-Carrón, Thomson, and Macdonald 2007, 31). In the face of such apathy, violent men learned that they could act with impunity, and women learned that it was not worth reporting abuse. Reluctance to apply the 1996 law and investigatory and prosecutorial protocols demonstrates how state officials, conforming to prevailing social frameworks, can undermine reforms and teach societal actors damaging lessons about the state.
Finally, in response to grassroots and international pressure, Congress passed the 2008 law that criminalized femicide and physical, sexual, psychological, and economic violence. The law mandated the strengthening of existing institutions and the creation of new institutions that would specialize exclusively in VAW, including specialized courts. These and other postconflict reforms to enhance women's legal access generated institutions that challenged VAW's normalization and impunity. Yet new institutions were systematically under-resourced, concentrated in urban centers, and connected to insufficiently reformed institutions that undercut their effectiveness, limiting their impacts for poor, indigenous women in rural communities.
Even prior to the 2008 law, the Peace Accords mandated the creation of legal aid offices for indigenous women, recognizing their vulnerability "in the face of gender and ethnic discrimination, and poverty and exploitation" (Sieder and Sierra 2010, 33). These Offices for the Defense of Indigenous Women (Defensoría de la Mujer Indígena, DEMIs) were established in the early 2000s in departmental capitals, housing social workers, psychologists, and lawyers who provided holistic support to indigenous women experiencing family, legal, or mental health challenges. Subsequently, Article 17 of the 2008 law mandated the further strengthening of DEMIs to support indigenous women victims of VAW. Yet, in 2017 the country's fourteen DEMI offices were allocated less than half of the funding required to carry out their mandates (CIDH 2017, 57). International bodies expressed concern over the "limited resources, authority and capacity" (CEDAW 2017, 4;MESECVI 2017) of DEMIs and other women's institutions, attributing the systemic under-resourcing of institutions in part to anemic tax revenue, thus demonstrating how the combination of successful neoliberal reform and failed tax reform undercut the VAW-related reforms' impacts. DEMIs' location in departmental capitals and lack of resources undercut their impacts. Women from Santa Nimá, for example, would have to travel two or more hours to reach the nearest DEMI. There, they would find four dedicated but overloaded employees. In 2016 alone, the sole lawyer single-handedly managed over two hundred legal cases. 6 In order to expand their impact, employees used their own personal money and time to support women, created informal relationships with justices of the peace and OMMs, and lodged complaints against other state institutions (police and forensic experts) that were undermining their efforts.
The 2008 law mandated the creation of specialized VAW courts. They housed psychologists and daycare facilities; employed judges trained in gender analysis; incorporated techniques to reduce revictimization, and were paralleled by specialized public prosecutors. By 2017 there were thirteen specialized courts nationwide-all located in departmental capitals, including one a two-hour drive from Santa Nimá's center, further from its remote villages. In practice, accessing specialized courts required indigenous, poor women from rural areas to interface with many state institutions. Because they were often uninformed about the 2008 law and specialized courts, these women were only likely to access specialized courts if directed by other institutions such as police, justices of the peace, the OMM, or DEMI. Subsequently, they would interact with the public prosecutor's office and forensics experts before entering courts. Some of the institutions that women experienced had internalized VAW reforms, but others had not. Few had adequate funding, linguistic, or cultural competency. Additionally, specialized courts and related institutions were in departmental capitals, introducing additional barriers for poor women from rural areas.
In the departmental capital, women from Santa Nimá encountered a women's public prosecutor's office that did not have interpreters for all of the department's languages. This office had only eight employees and two vehicles to investigate crimes across a geographically dispersed department and to prosecute all cases, leading to endless backlogs. One employee reported that in 2016 they directly received over 3,656 criminal reports, with almost twice that number from other institutions across the department (hospitals, police, etc.). From there, victims were directed to the regional National Forensics Institute (Instituto Nacional de Ciencias Forenses de Guatemala, INACIF) office for medical or psychological exams. While those who work in specialized courts underwent training in avoiding revictimization, the same could not be reliably said of members of such critical support institutions outside of the capital city (CEDAW 2017). The closest INACIF office for women of Santa Nimá employed only Spanish-speaking doctors and barred interpreters or relatives from medical exams. DEMI employees complained about this office's treatment of indigenous women, explaining that "if a young girl or woman goes in there … and they cannot even talk, then they can get revictimized. The doctors are all-powerful. They will not let anyone go with that young girl. That just increases the damage for the girl. Those doctors do not have any cultural competency." INACIF, like all supporting institutions, was overloaded with cases. Nationally, over half of the medical evaluations INACIF performed at the request of public prosecutors in 2016 were related to VAW. Of these, 47.3 percent were conducted in Guatemala City even though only 22 percent of VAW complaints were filed there, showing uneven reach of state institutions critical to specialized justice (INE 2017).
The government failed to outline investigation protocols related to VAW with a gender focus (MESECVI 2017), and the institutions responsible for investigation (INACIF and the specialized public prosecutors) were under-resourced and concentrated in urban centers. This produced significant delays and weaknesses in VAW investigations, especially in rural municipalities. For example, a judge recounted an instance in which investigators merely photographed the outside of a house in which a femicide occurred to document the crime scene, even though protocols mandate the documentation of the specific location of the crime (in this case, the bedroom). Another judge said he was forced to render a not guilty verdict because investigators had failed to secure the crime scene until a week after a femicide, allowing the perpetrator to destroy evidence. In the face of such ineptitude, the tasks of contacting witnesses and gathering evidence all too often fell to victims and their families. This trend led the DEMI to conclude that in cases of sexual violence, for example, victims learned to "appeal to justice if and only if you can prove you were raped" (DEMI 2007, 79).
Women who successfully navigated these interactions and whose cases met the required burden of proof often faced long delays in their trials. Crimes that fell under the 2008 law became the most reported crimes in the country, reflecting an increased willingness to report VAW. This led to backlogs in specialized institutions. For example, in 2018, there were 34,422 reports of VAW in the thirteen departments with specialized courts (INE 2018, 131). That same year just 9,851 cases entered pretrial hearings and just 3,045 advanced to trial (UJEFEM 2019, 88) in specialized courts. 7 Delays disproportionately affected women in rural areas, where limited police presence made restraining orders more difficult to enforce and where the lack of privacy heightened social stigma. Once they started, trials were divided into multiple hearings that unfolded over the course of weeks or longer. Women from Santa Nimá traveled over four hours round trip to attend hearings, which were often cancelled because the public prosecutor or defender had not arrived or an earlier hearing went long (in 2016, 25 percent of hearings in the closest specialized court to Santa Nimá were suspended, according to the judiciary). Relatives often accompanied indigenous women to offer emotional and linguistic support, increasing costs.
Article 15 of the National Language Law mandated that the population "be informed and attended to in their own language" (Decree 19-2003). Specialized courts provided interpreters for indigenous languagespeaking defendants, victims, and witnesses. Yet in practice, they, like other judicial institutions, had insufficient interpreters. For a population of 16.5 million people speaking twenty-three languages, in 2016 the Attorney's General Office employed 35 interpreters, the Indigenous Affairs Unit had 101 interpreters for the entire judicial system, and the Public Defender's Office employed nine lawyers who spoke an indigenous language (López 2017). Interpreters varied in their effectiveness. For example, I observed one witness unable to communicate effectively with his assigned interpreter because of differences in regional dialects, resulting in a rescheduled hearing. In another case a defendant's interpreter only translated statements directed at the defendant, rather than the whole hearing. The Academy of Mayan Languages found that some of interpreters employed by the judiciary did not display adequate verbal comprehension, and the majority did not display adequate reading/writing literacy in their assigned languages (CNSAFJ 2007, 30). In a 2017 interview, an indigenous judge highlighted additional miscommunications. He explained that indigenous women from places like Santa Nimá "are not going to say, he raped me. Rather, they say, he touched me; he hurt me…. One cannot do a literal interpretation; one has to know the context." Thus, even with an interpreter, indigenous women's experiences were less likely to be adequately conveyed in courtrooms.
In sum, attempts to render the state more accessible to rural communities and better able to address VAW produced mixed results. Reforms were unevenly internalized across institutions, and the institutions closest to indigenous women in rural communities tended to be those that were inadequately reformed to meet their needs, as can be seen in Table 1. Instead, officials often reproduced discriminatory practices rooted in classism, racism, and misogyny, rarely spoke the local language, and often reinforced women's distrust of state institutions. As one indigenous woman explained, "because we are women [state officials] do not listen to us, and if they do it is to chastise us and tell us to come back another time or to tell us to go back to the man. Because we are indigenous and poor it is even more difficult to obtain justice. Sometimes we turn to one place and from there they send us to another, and so on. So, what use is it? It is just a waste of our time" (quoted in DEMI 2007, 94). Officials in other institutions (OMM, DEMI, specialized courts) challenged prevailing frameworks that normalized gendered, classed and racialized violence and discrimination. They addressed women's needs drawing on their formal mandates and informal practices and, at times, directly challenging other underperforming institutions. Yet their coexistence with insufficiently reformed institutions, along with the lack of structural reforms to adequately address inequality or fund key institutions, meant that the poorest and most marginalized residents were forced to pay the highest "legal tolls" to access support .

Experience Navigating the Guatemalan State
Virginia's story reveals how gender, class, ethnic, and place-based marginalization intersects with the uneven patchwork of reformed and under-reformed institutions to profoundly shape access to, and experiences of, security and justice. Virginia was an indigenous teenager living in a village an hour outside Santa Nimá's center, three hours from the departmental capital. Most of the people there, like Virginia's family, subsisted on agriculture and primarily spoke the local indigenous language. When she was twelve years old, a masked man grabbed her while she was walking home. In the struggle, his mask was removed and Virginia realized that her attacker was her neighbor. He demanded she have sex with him and punched her multiple times in the face when she resisted. Virginia lost consciousness. When she awoke, her clothing was disheveled, and her attacker had fled. Virginia ran home and told her parents about the assault. Her father accompanied her to report the crime, bypassing the local community watch, which was seen as ineffective in addressing sexual violence, and going directly to the police substation. There, the male, ladino officer took Virginia's report in Spanish, writing that Virginia's neighbor attempted to rape her but did not succeed, contradicting Virginia's later statements that indicated she had been raped. Virginia returned home and, not realizing she was destroying evidence, burned her clothing.
Four days later, Virginia and her family traveled three hours to the departmental capital so that Virginia could undergo a physical exam with INACIF. The exam was conducted in Spanish and focused on her "extragenital, para-genital, genital" body parts, according to the report. This exam found severe bruising on Virginia's face and evidence of a "deflowering" in the previous ten days, echoing the long-standing focus on bodily evidence and virginity in cases of sexual assault. INACIF was unable to collect any bodily fluids, likely because of the exam's delay. The following week Virginia and her family traveled again to the departmental capital where Virginia underwent a psychological exam in Spanish. The psychologist reported Virginia's symptoms, including intense fear, depression, loss of appetite, and insomnia. The psychologist wrote "the patient has a humble appearance," presenting "adequate" clothing and personal hygiene, and that the family had "limited economic resources [as] the father works in agriculture and the mother does housework." The examiner additionally noted that Virginia was suffering from "cultural patterns of shame for having lost her virginity in a way that is seen as dishonorable and shocking" and that affected "her future marital unhappiness." While the report portrayed Virginia as a "worthy" victim, it relied on condescending, racialized, and classed language to do so, demonstrating the way that state actors viewed indigenous, poor, and rural bodies.
Roughly a week later, Virginia and her father again made the six-hour round-trip journey to give a statement at the women's public prosecutor's office. Virginia stated she wanted justice for herself and her family, in part based on the danger they felt living so close to her attacker: "I want him to be held responsible for what could happen to us, to me and my family, because we are neighbors." They connected with the • None • Patrol community and detain suspected criminals • Accused of abuse of power and acts of gendered violence.

Justice of the Peace
Santa Nimá (mid-1990s) • Expand access to justice.
• Address minor crimes • Take complaints of VAW (barred from promoting reconciliation) • Lacks linguistically and culturally competent staff • Offers women support in some areas (alimony, child support) • Sometimes encourages reconciliation, in violation of law where Virginia received additional psychological support from an indigenous woman-the first she had encountered thus far. The DEMI lawyer eventually joined the case representing Virginia's father as a joint plaintiff, suing for financial reparations for the lost wages and expenses that accompanying his daughter entailed. Virginia did not feel comfortable going to members of the OMM for support as others might have done because one of the staff members was her attacker's cousin, demonstrating a potential barrier to accessing support that is exacerbated in small, rural communities. Twenty days after the attack, the overloaded public prosecutor's office sent an investigator to interview potential witnesses. Four months after the attack a forensic photographer finally arrived to photograph the crime scene, too long after the attack to document any evidence. The public prosecutor's and INACIF offices' lack of adequate personnel and resources (including vehicles) and the distance that investigators had to travel exacerbated the delay in the investigation and documentation. This delay meant that when a formal arrest warrant was finally issued roughly eight months after the attack, the accused had had time to flee and could not be located. Throughout the entire process, the attackers' family harassed Virginia and her family, shaming Virginia for her own assault. Although social stigma around sexual violence was common throughout the country, it was likely heightened in Virginia's small community, which offered little privacy. Virginia stopped leaving her house and attending school.
Four years after the assault, the police finally arrested Virginia's attacker, launching another period in which Virginia and her family traveled back and forth between their village and the departmental capital to attend court hearings, despite having scarce resources to do so. The women's public prosecutor petitioned the specialized courts to hear the case, accusing Virginia's attacker of rape and psychological violence, oddly failing to include physical violence (for which there was the most evidence) in their claim. The case was admitted, with the women's public prosecutor representing Virginia and the state, and the DEMI lawyer representing Virginia's father. Despite this support, the continued harassment, constant travel, and slow progress of her case eventually became too much for Virginia to handle, and she fled to the United States, where she got lost in the desert before being detained. Back in Guatemala, lawyers from the DEMI and the office of the women's public prosecutor moved forward with the case despite Virginia's absence. Remarkably, they located Virginia in detention and arranged for her video testimony so the trial could continue, demonstrating these officials' dedication. In the United States, an immigration judge found Virginia did not have a credible fear of return and ordered Virginia deported before her video testimony could occur. Returned to Guatemala, Virginia testified in the specialized court with the assistance of court interpreter, as did her parents, a neighbor who had heard Virginia's screams, and expert witnesses from INACIF and DEMI. Among the defendant's supporting witnesses was a member of the village's communal authority structure (the auxiliary mayor's office) who had also been the head of the village's community watch at the time of the attack-demonstrating how communal authorities may act as either resources or obstacles to security and justice. Based on these testimonies and the evidence introduced, a verdict was finally rendered-five years after the assault. 8 The judge explained that there was a lack of forensic evidence to prove rape and that he had to throw out parts of Virginia's testimony because of inconsistencies between the police report, her declaration, and her court testimony regarding the sexual assault-possibly because of the lack of linguistic and cultural competency of the police taking Virginia's statement, the trauma of the attack, and the time lapse before her testimony. The judge-one of the few indigenous judges in the country-also did not accord the testimony of Virginia's parents any probative value because of the technicality that they both claimed to have finished their statements at the exact same time. The judge explained in his sentence, "logic and experience indicate that the witnesses were instructed about the way in which they had to testify and they took it literally" or, given they were illiterate, it was also possible that people without legal expertise could have given them inaccurate instructions. The judge stated that the prosecutors had failed to prove that the accused's actions were aimed at intimidating the victim, diminishing her self-esteem, or controlling her. Thus, he could not convict the defendant for sexual or psychological violence.
Yet the male judge showed the benefit of his specialized training and commitment, referencing "the methodology of legal gender analysis, whose end from the gender perspective … is to work to democratize the law." He stated that there was sufficient evidence to convict the defendant for physical violence, even though the prosecutor had not argued for a conviction on these grounds. Finding Virginia's attacker guilty of physical violence in the public sphere, the judge sentenced him to six years in prison and mandated that he pay $1,350 to cover Virginia's legal costs. Although Virginia's father challenged societal norms in his open support of his daughter and incurred significant financial and emotional costs in assisting her, the judge denied his requested reparations, arguing that there was inadequate proof of costs. Given the informal nature of agricultural work and the lack of receipts for transportation and other services in rural areas, it would have been difficult to prove financial costs. The decision to deny Virginia's father reparations overlooked this reality and also failed to acknowledge that the trauma was felt at a group level and that Virginia's navigation of the state required extensive familial support.
Virginia's attacker served three years in an overcrowded and poorly regulated prison. The verdict mandated him to receive professional assistance so that upon the completion of his sentence he could be resocialized "for his personal benefit and that of society." But the penitentiary system, which is at 300 percent capacity and in practice governed by organized crime, has effectively "abandoned social rehabilitation" (CIEN 2013, 24). Virginia's attacker was paroled early and returned to Virginia's village, where he and his family continued to threaten Virginia and her family, even spreading rumors that he would rape Virginia's sisters. Virginia's father estimated it would take the police at least an hour to arrive if he called them. He added that they were reluctant to do so because the police could easily be bribed, they were afraid of angering the man, and their previous legal experience had been so tiresome and expensive.
Virginia's experience demonstrates how abstractions-in scholarly accounts and laws-that separate marginalization rooted in gender from those rooted in ethnicity, class, or place "damage the real-life permeability of these various oppressions" (Forster 1999, 58). Virginia did not face one form of marginalization at a time but rather experienced them multiply and simultaneously (Sosa 2017). For example, the distances she had to travel were rooted in place-based marginalization that concentrated more effective institutions and resources in urban centers. Her class made the costs associated with this travel all the more difficult to bear. And Virginia's experience of traveling three hours outside of her community was additionally colored by her indigeneity: linguistic and discrimination-based barriers in state institutions meant that she could not take the trip alone but rather had to be accompanied by her father. This increased the impact of travel on Virginia's family in ways that were illegible to courts, which saw VAW as a crime between individuals, rooted solely in gender.
Virginia's experience was relatively successful. Although she confronted numerous linguistic, cultural, and geographic barriers to accessing justice and experienced her interactions with state institutions as uncomfortable and tiresome, she received unconditional support from her father, she and her family trusted the state enough to report the attack, a number of institutions went above and beyond to support her, and a guilty verdict was eventually reached by a male judge trained in gender analysis in a court specifically designed to reduce revictimization-all of which would have been unlikely a decade earlier. These advances should not be understated. Yet Virginia's experience also demonstrates that in the context of intersecting sources of marginalization and an uneven Guatemalan state, reforms are likely to differently affect individuals who are differently placed in gender, class, ethnic, and place-based webs of power. Indeed, Virginia's experiences demonstrate that even cases that yield a conviction can be experienced as a failure for victims and their families, who expend considerable energy and resources and must overcome obstacles in their search for justice, only to face continued threats to their security even after a verdict is rendered.

Implications
Prioritizing the experiences of marginalized communities like Santa Nimá and individuals like Virginia reveals the nature and sources of gaps between policies and their impacts. In Guatemala, the impacts of VAW reforms were unevenly felt, with those who are already the most marginalized benefiting the least. This uneven impact can be explained by the failure to address broader structures of power rooted not just in gender, but also class, ethnicity, and place, both within state institutions and in the broader society. These historically constituted power structures have inculcated discriminatory institutional cultures, fomented distrust of the state, and introduced barriers to accessing state services. The creation of new institutions that challenged normalization of VAW and impunity, such as DEMIs and specialized courts, was a significant step forward. But their impacts were undercut by their systematic underfunding, concentration in urban centers, and connections with insufficiently reformed state institutions working at cross-purposes.
Although Guatemala can be seen as an "extreme case" of gendered violence, structural inequalities, and uneven state-society relations, many of the obstacles encountered by Guatemalan victims are paralleled in other countries, even those with more resources. What lessons, then, can we draw from Guatemala for addressing VAW and enhancing access to justice generally? First, the analysis demonstrates that historical patterns of state-society relations matter because they affect contemporary state institutional cultures and societal expectations. In Guatemala, historical and ongoing experiences with state institutions help to explain why women in rural, indigenous communities often see police (the state's closest "face") as a threat rather than a resource, in ways that parallel black women's mistrust of the police in the United States (Potter 2008). This implies that VAW reforms should pay special attention to transforming localized interactions at points of first contact in order to overcome this history and instill trust. Second, paralleling long-standing claims of feminist women of color, policies that attempt to artificially separate axes of marginalization will be inadequate for meeting the needs of the multiply marginalized. Whereas feminists have highlighted the importance of jointly considering gender, racial, and class-based discrimination, this article highlights the importance of additionally considering place-based marginalization, which likely requires specifically targeted resources and infrastructure to be incorporated in VAW reforms. Finally, the analysis presented here demonstrates that in contexts of severe economic, racial, gender, and/or place-based inequality, criminal justice reform without social justice reform will inevitably fall short in addressing VAW (Godoy-Paiz 2008;Godoy 2006, 22). In Guatemala, fiscal and social policies failed to challenge devastating levels of poverty. This, in turn, reduced the likelihood that many women could leave abusive situations or afford the high costs of pursuing justice, and ensured that the institutions to which they could turn would be under-resourced. In this view, reforms that are not focused on VAW, such as tax reform, land redistribution, and increases in social spending, are necessary to reduce women's exposure to violence and to increase their effective access to justice.

Author Information
Erin Beck is associate professor of political science at the University of Oregon, focusing on international development, NGOs, violence against women, and access to justice in Guatemala. She is interested in how governmental and nongovernmental actors design institutions to enhance the well-being of marginalized populations and how these institutions actually affect people's lives. She is author of the award-winning book How Development Projects Persist: Everyday Negotiations with Guatemalan NGOs (2017) and her work has appeared in journals such as World Development, Latin American Politics and Society, and Studies in Comparative International Development.