In 2011, Estefanía Rivera Guzmán and her colleagues went to the Mayor's office in Medellín, Colombia. They were seeking information about a number of recent murders of women in Comuna 10 – a historic zone comprising several neighborhoods in the center of the city. The office had no answers. According to Estefanía, they offered nothing and had no systematized information – not on feminicides nor on anything else. This was the moment that, as Estefanía states, "We realized that our institutions didn't have information systems."1
Estefanía was a lead organizer with a group that had been around since the late 1990s but had recently relaunched with a new name – the Red Feminista Antimilitarista (the Feminist Antimilitarist Network). Their name reflected their position against war, violence and militarism in Colombia. Composed mainly of feminists, poor women and lesbians, they were seeking "other ways of exercising and building collective power," mainly through creative political actions in public space and popular education initiatives2. Estefanía was leading a project to gather information about feminicide in Medellín, and when the team couldn't get the information from the Mayor, they decided to create their own registry of feminicides in Comuna 10.
Monitoring primarily from news media articles and supplementing with information from social media and their own feminist networks, they started logging cases in their database and quickly came to an important insight: "We began to realize that feminicides happened not only because women were targeted for being women, but because they were women from specific social classes: impoverished women, women street vendors, trans women, women who practiced prostitution."3 This insight led the Red Feminista Antimilitarista down a path of reading and discussion for the next several years. Especially influential were works by theorists such as Rita Segato, Jules Falquet and Verónica Gago, all of whom have emphasized the connections between gender violence and economic violence4. The group found that discussions centered only on intimate feminicide (where the aggressor is an intimate partner or ex-partner) were too limiting to explain violence against women in Colombia where there is a high degree of neoliberalism and militarization, even in a time of supposed "peace." 5
The group published their first report in 2014 in which they develop their concept of neoliberal feminicidal violence: "the extreme violence of capital on women who find themselves impoverished, stripped of power and significance in the modern, capitalist and patriarchal coloniality."6 Here the Red Feminista Antimilitarista is pointing to the economic roots of gender-related violence. In Colombia, they articulate how the state, paramilitaries and organized crime groups fight for control over urban spaces and terrorize poor and lower-class women and communities in the process. Likewise, the Red Feminista Antimilitarista describes the extortionary economic practices called "pagodiarios" where (disproportionately) women must borrow large sums of money from criminals at high interest rates to pay other criminals simply to avoid violence7 [cite]. For the group, even intimate feminicides are a manifestation of neoliberal feminicidal violence because the basis of intimate violence is so often connected to economic disparities, domination and control. Colombia passed a law defining femicide in 2015 but in practice the law has been applied mainly to cases of intimate partner violence, ignoring many of the cases that the group had started collecting and passing over their intersectional economic analysis of violence8 [CITE].
In 2017, the Red Feminista Antimilitarista expanded their efforts from Comuna 10 to the whole country and renamed their project to the Observatorio Feminicidios Colombia (Feminicide Observatory of Colombia)9. In this, the group is similar to the majority of our interviewees who operate counterdata projects as on-going monitoring efforts, often called observatories, that seek to document and communicate about gender-related violence. For them, the project serves two major purposes. First is to comprehend a group-based, structural phenomenon at a national scale: "Beyond the data, it is the possibility of understanding what is happening to women in our country."10 That is to say, for Estefanía, one of the goals is to examine power using concepts and theories of power appropriate to the Colombian context and introducing novel concepts where necessary. Second, "the Observatory is a tool to build protection strategies."11 While the Red Feminista Antimilitarista monitors at the national scale, they continue to run community-based programs in Comuna 10 like the "Circles of Protection" program. This works with individual women facing violence to develop a community of people in their everyday lives who are briefed on the situation and who can step in to offer support to protect them from violence12. Leveraging their knowledge from the Observatory, the group crafts these protection strategies as a way to challenge power.
This is part of the origin story of the observatory run by the Red Feminista Antimilitarista – how and why they resolved to begin counting feminicide and how their understanding of the problem shifted as they began to produce counterdata informed by a structural analysis of power. This chapter is the first of four in this book that surfaced as themes from our interviews with grassroots data activists. These themes describe workflow stages: resolving, researching, recording, and refusing and using data. Together these four stages that you can see in figure 2.4 comprise the "anatomy" of a feminicide counterdata science project – a way to study feminicide data activism practices and understand what activists' labor and agency and hardships look like across the life cycle of a project. This doesn't mean these four stages are neatly discrete and linear, like a conventional data science process diagram. The stages of a counterdata project are often a mixture of linear and circular and I will try to show their interconnections even as we travel through them one-by-one.
This chapter explores resolving – the initial stage of a feminicide counterdata science project. It surveys how and why grassroots groups begin counting feminicide – how they get started, how they conceive of the problem, and why they determine that counting and registering data will be an effective method to address feminicide. As we will see, all activists mobilize an analysis of power, though that analysis often evolves with time and learning. And all groups integrate data production and communication into their working theory of change – their idea for what types of interventions are needed to challenge feminicide.
Many monitoring efforts emerge from feminist activism, like the Red Feminista Antimilitarista, but other projects come from the nonprofit sector, from academia, or from data journalism. The latter is the case with a feminicide counterdata effort in Bolivia called Cuántas Más. The project was started by three journalists in 2014. That same year, Bolivian journalist Hanalí Huaycho was killed by her ex-partner Jorge Clavijo in front of their five-year-old son. She had reported him 14 times to the police, but he was a lieutenant in the police force and no action had been taken. Clavijo escaped, the case was mishandled in a number of ways, and a new Bolivian government reopened the investigation in 2020. Huaycho's murder was the first in Bolivia to be classified as a feminicide under Law 348, passed in 201313.
It was in this period of intense public attention on feminicide that journalists Raisa Valda Ampuero, Ida Peñaranda and Marcelo Lazarte started Cuántas Más (How Many More [Women]). At first the team sought official data on murders and feminicide cases, however Bolivia had no open information laws which made obtaining official records nearly impossible. So, like the majority of counterdata groups in this book, they turned to the media to monitor cases of feminicide. From 2014 - 2017, the team scoured articles from Bolivian news media and extracted details about each case of feminicide for their database. This was supplemented by information and case files contributed directly by families who were seeking – and not finding – justice. Ida explains why families began to come to them: "One way to exert pressure is to make your case public. And, not everyone has been able to do it."14
While Cuántas Más aspired to collect every case they could find of feminicide during this time period, the point of the project was not only about producing numbers. As Raisa framed it, "[We wanted] not only to fill in the missing data that are urgently needed, but also to be able to give a face, an image, to those data… here we are not talking about cold data but about people, each of them is a person who has lost their life in dire circumstances."15 Cuántas Más published a map and short notes about each case and included details about the person from their life. For example, one young woman was a mariachi and they included a clip of her favorite song. Their goal was to humanize and remember victims in life but also to challenge the misogynist, victim-blaming narrative framing of feminicide coming from both police and media reports. For example, when the group went to the police for records about women murdered in domestic violence, the only data fields collected by the police were marriage status and level of intoxication, two fields that lend themselves easily to victim-blaming narratives ("she was a loose woman", "she was a public woman", "she was not in a heterosexual relationship", "she was drunk", etc). Likewise, Raisa found the media reports on feminicide from fellow journalists to be dehumanizing: "'100 deaths this year.' No, no, we're not going to do it like that. We're going to do it by telling a little more, about who are they, what situation were they in, what relationship did they have, for example." As they frame it, the goal of Cuántas Más production of counterdata was to understand and explain the relationships of power behind the phenomenon of feminicide, and to push back on the reductive and dehumanizing narratives that were widely circulating.
The stories of Cuántas Más and the Red Feminista Antimilitarista are origin stories. They describe the stage of resolving, the first stage in what I am calling the anatomy of a feminicide counterdata science project (see figure 3.3). Simply deciding to start counting can be both the hardest and the easiest part of a long and arduous process. The activists we have spoken with describe various motivations for starting to register cases of gender-related violence. In the case of Cuántas Más, coming out of journalism, these factors included recent changes in legislation, a high-profile case of the feminicide of a journalist, and increased public attention to the issue. The Red Feminista Antimilitarista, on the other hand, felt like they needed more knowledge in order to inform their community-based violence prevention programs in Medellín.
For some activists, the motivation can be deeply personal. At least five groups disclosed to us that they themselves were survivors of gender-related violence or else had close family and community members who had been killed. For example, the Néias Observatório de Feminicídios Londrina, in Brazil, is named for the sister, Néia, of one of the founders. Néia was brutally assaulted by her husband when she tried to leave him in 2019. While she survived the assault, she was left with severe cognitive and physical impairment and unable to speak again or to care for her four children. She died two years later of a heart attack at age 35. During the trial of her aggressor, Néia's sister and colleagues ran a successful campaign to attract public attention to the case. While campaigning, they were struck by the fact that they came into contact with numerous other cases of feminicide which had received no media attention and whose defendants got off with misogynist arguments about being carried away by "strong emotions." The group decided to track and systematize these cases and they founded the observatory in order to "raise a voice for so many Néias."16
Yet some data activists start more informally and come to understand the magnitude and complexity of the task at hand later in the process. For example, Annita Lucchesi was doing her masters degree in geography and needed statistics to describe cases of MMIWG2 so she decided to just start collecting them herself. As she characterizes it, "I was pretty naive in thinking that I could just kind of cross reference in a spreadsheet and come up with a number that way and that there would be a start and an end. And now, five years later, I'm still doing it and now it's my life's work."17 María Salguero had a similar start. She wanted to use her technical skills to help the Mexican investigative journalism group Periodistas de a Pie which covers deaths and disappearances. They told her there was no data about feminicides so she volunteered to check the crime tabloids and make a map18. What she thought was a one-time map has grown into a years-long mapping and data production project that has gone on to have wide impact in policy and global news media.
When feminicide counterdata projects become highly visible, like María's, they inspire and motivate others to follow their lead. Aimee Zambrano described how she followed María's work and the work of other observatories for some time before deciding to start monitoring femicide for Utopix, an activist media group based in Venezuela. Police discovered the body of a woman on a beach in Caracas that Aimee herself often frequented, but Aimee didn't see any news reports about it. This led her to ask, "how many cases are happening in the country and they are not coming out or they are not being shown through the media?"19
The tipping point for Aimee – like Cuántas Más and many other groups – was a specific case of feminicide. This is also the situation with Letra Ese, a group that monitors LGBTI+ killings in Mexico. The group had been protesting and working towards justice in a number of cases of hate crimes and were frustrated by the lack of judicial response and the impunity that surrounded the cases. They started an observatory – the Citizen Commission Against Hate Crimes – following the 1995 murder of María Elena Cruz, a trans woman. Their primary objective was to systematize cases and influence the media to report on them. According to Alejandro Brito, one of the founders, the group felt that if they were successful at influencing the media's agenda then "other agendas would be influenced: other legislative agendas, human rights, the women's movement, the LGBTI+ movement."20
In all of these cases, when the groups got started they found the existing responses – from the state, the media and the public – both inadequate and discriminatory. From the state and its institutions, they encountered narrowly framed legislation which left out many types of feminicide, families that were unable to find justice, as well as missing data – either the complete absence of official records or else data that were inaccessible, incomplete or collected with biased categories (e.g. marriage status and intoxication). From the media, Cuántas Más describes either the complete lack of reporting and/or the sensationalism and dehumanization with which cases were reported. These produced a public culture in which individual acts of violence against women were normalized and tolerated. If a woman was killed, the event was more often than not written off as an isolated "crime of passion" or else the victim was blamed for her death due to some behavior that did not conform to norms of cisgender, heterosexual, white, middle-class notions of womanness.
In response to these structural barriers, groups develop their own analysis of the problem. While data activists use diverse ideas and theories appropriate to their context, in all cases they have developed and continue to refine an analysis of structural power that undergirds their work. This power analysis is central to the resolving stage of a counterdata science project. The Red Feminista Antimilitarista did this through a deep and collective dive into feminist and decolonial theory, from which they then produced several theoretical innovations such as the concept of neoliberal feminicidal violence. Cuántas Más did this through their critique of existing media coverage and the victim-blaming culture that it produced. A group's power analysis directly informs their theory of change – the rationale and means by which they will challenge existing, biased responses to feminicide. Both groups – and all groups in this book – integrate counting into their theory of change. They resolve that counting and registering cases of gender-related killing will be a useful strategy to challenge power. Why? What purpose does counting serve across these diverse projects?
For grassroots data activists, it's not about collecting data for data's sake – in zero cases do they see data collection as a "solution" to gender-related violence. All of our interviewees were circumspect about the role of numbers and data. Estefanía asserts that the underlying analysis of inequality in Colombia is more important than counts: "But beyond the data and figures, it's the analysis of the context... [we have learned to] carry out analysis of the context on a daily basis, always critically."21 In a similar vein, Ida from Cuántas Más states, "We have to understand that numbers are important, yes, but they are not the full representation of what we are seeing, rather it's the characteristics of the violence that we were seeing."22 As I will argue in the rest of this chapter, groups don't initiate data monitoring projects to "solve" a problem, rather they produce data to remake and reframe the problem space entirely – to transmute gender-related violence from the personal realm to the political realm.
There is a famous feminist saying, used by the Comabahee River Collective and written about by activist Carole Hanisch: "the personal is political."23 This saying came out of US activism in the 1960s, where women were told not to bring "personal" problems such as sex, domestic violence, forced sterilization, childcare and abortion rights into "political" organizing work predominantly led by men. In response to the exclusion of these issues, women formed consciousness-raising groups, small-group dialogue sessions where they would gather and share personal experiences. The purpose of such sharing was not (only) for venting and support. A key insight of consciousness-raising groups was that gender oppression works by stigmatizing and privatizing a structural problem – like domestic violence – so that an individual would face her struggle alone and ashamed, blaming herself. These groups validated and named these lived experiences as oppression and worked together to build a collective understanding of them as a structural problem.
Hanisch regarded these groups as a mass organizing tool. As she writes, "One of the first things we discover in these groups is that personal problems are political problems. There are no personal solutions at this time. There is only collective action for a collective solution."24 While consciousness-raising groups for white women focused mainly on sex and gender, the Combahee River Collective ran consciousness-raising groups that also integrated discussions of race and class oppression25 [CITE]. The goal of these groups was to provide a space for processing personal shame, isolation and pain around stigmatized issues, to recognize those experiences as a structural pattern, and to channel that understanding towards collective political action and transformation. The consciousness raising group, therefore, is a space of scalar transmutation – a place where personal pain is heard and held and lifted into collective struggle26.
Counterdata groups do something similar by resolving to document each case of feminicide in painstaking detail. The Red Feminista Antimilitarista states it like this in their 2014 report, "...every violent death is a public death. Those murdered women not only deserve the grief of their relatives and loved ones, they also deserve a public grieving. Documenting femicide is a way of doing this and, at the same time, a form of social reparation."27. Groups see the production of counterdata as a way to move feminicide and gender-related violence from a personal problem to a political problem; from a private issue to a public issue; from the interpersonal domain of the matrix of domination towards the structural, disciplinary and/or hegemonic domains. It is the counterdata groups' analysis of power, combined with the case documentation and numbers that they produce, that allow them to attempt this scalar transmutation.
All counterdata projects begin by examining power, and here there is great resonance with the first principle of data feminism which states that a feminist approach to data begins by analyzing how power operates in the world. Activists examine the underlying social inequality that produces feminicide. This includes consideration of the biased and unjust responses to feminicide from the state, the media and the public. Examining power may be an explicit process that takes groups some time to develop – as did the several years of reading and discussion that the Red Feminista Antimilitarista undertook. Or a power analysis may be a shared understanding that is developed implicitly, through practice and dialogue, like the three collaborators in Cuántas Más who were determined to shift the way feminicide was narrated publicly in Bolivia as well as to use their project to show that "[feminicide] is a problem of society, it is not a personal problem."28 An examination of power may also emerge and develop more informally as activists start their work. For example, the activists from the Néias observatory had a realization about the structural nature of the problem as they campaigned for justice for a specific case and saw how many cases did not have advocates like themselves.
Counterdata groups draw from a wide array of influences to examine power. These range from social movements (#NiUnaMenos, the Latin American feminist movement, the #MMIW movement) to individual people (María Salguero, Marcela Lagarde y de los Rios and Diana Russell) to international organizations (the UN's work to define femicide/feminicide) and also to more academic bodies of knowledge like Latin American decolonial feminist thought, queer theory, Black feminism and Indigenous North American scholarship. While there is great variation in the guiding concepts groups choose to frame their analysis around, all counterdata groups seek to move from a personal to a structural explanation for gender-related violence. Rather than explaining incidents of gender-related violence as isolated events, "crimes of passion", or the work of pathological deviants, counterdata producers work to establish the deeper, more essential reality that violence is rooted in structural causes: from the feminization of poverty and deeply unequal gender structures, to white supremacy and colonialism, neoliberal economic policy, and territorial dispossession.
But for grassroots data activists this structural analysis is not enough. This is where their approach shows resonance with the second principle of data feminism: a commitment to using data to challenge unequal power structures and to work towards justice. Here is where the counting comes in. Just as counterdata groups are using conceptual and analytic tissue to link diverse incidents of violence together, logging these cases – whether in a spreadsheet or database – operates as a way to represent the power analysis of a given group empirically. Counterdata activists are not only asserting that intimate partner feminicides are related to feminicides of sex workers in public space are related to deaths from clandestine abortion, but are demonstrating that fact by placing them together and by counting them together.
Counting and quantification have the power to render diverse things alike, each item neatly placed in its own row29. Typically, in feminist theory this affordance has been viewed negatively. In the twentieth century, there was a long-running "quantitative/qualitative debate" stemming from feminist critiques of the patriarchal nature of knowledge production. Many argued that the use of quantitative methods had the effect of silencing of subjugated standpoints, though it is worth stating that feminist researchers had also always used quantitative methods30. Debates about quantification continue today. For example, Sally Merry demonstrates convincingly how global indicators about gender-based violence strip context and hide important cross-cultural variation31. Many critical data scholars have shown that quantification, especially arising from "big data", holds the potential for harm and erasure, in both novel and painfully familiar forms32. In some cases, such as facial recognition, where quantification is weaponized against minoritized groups, the path towards justice leads not through more counting but through blanket refusals, bans and moratoria33.
We can hold it as true that quantification can cause harm. Yet at the same time, what about the harm of allowing a widespread, political phenomenon like gender-related violence to be endlessly pushed into the private and domestic realm, divided into individual cases, always treated separately? There is harm in refusing to see a structural pattern right in front of your eyes. To refuse to bring related things together or to give a problem a name is a form of willful ignorance. As discussed in chapter 2, scholars have shown how ignorance is situated – how what a society doesn't know or doesn't choose to know follows the lines of gendered and racialized power34. This leads back to Shulamit Reinharz' provocation from 1992 that the social sciences should be studying the sociology of the lack of knowledge – "how and why knowledge is not produced, is obliterated, or is not incorporated into a canon."35 In other words, while there may be harm in counting and aggregating, there may also be harm in not counting and not aggregating.
This is what anti-feminicide data activists mean when they speak of trying to "visibilize" (visibilizar) feminicidal violence. As Dawn from Women Count USA states, "And it [femicide] just seemed like it was so ubiquitous that it was invisible…you get this kind of drip, drip, drip and it's so fragmented that it's hard to get a good picture of exactly what's happening – to whom and by whom and where and how."36 Counting and documenting and aggregating and naming has a way of transforming the phenomena that are being measured from discrete, individual events into stable and solid examples of a widespread pattern. Or, as Pine and Liboiron state succinctly, "measurements make things."37 Scholars have described how technical artifacts like databases have a performative power to shape the ways that we see the world38. For example, Jonathan Stray discusses how the US census brought into being the category of "Hispanic" in 198039. This is not to say people who spoke Spanish or immigrated to the US from Latin America did not exist prior to that point, but there had not previously been a state classification of that social phenomenon as an ethnic identity category40. Once social concepts are named and measured, they may scale and circulate, and, when recognized by the state, serve as evidence for policy or decision-making. For example, Merry recounts how British colonial rule played a role in solidifying and scaling the Indian idea of caste, which had previously been more local, situated and contingent: "By redefining castes in terms of categories that applied across the subcontinent, the British rendered caste a far more fixed and intractable social entity."41 This is clearly an example of a social concept that is named, measured and scaled in the service of domination, imperialism and economic extraction. Yet it is precisely this power of "thingification" that anti-feminicide data activists are seeking: to name, to measure, to bring into being and to scale the reality of systemic gender violence. As Helena Suárez Val writes, drawing from the words of Marcela Lagarde y De los Ríos, feminist activists have long understood that "Naming the violence is a way to understand the problem, to intervene in it and to contribute to its eradication."42.
Thus, the counting of feminicide, the assembling of spreadsheets and databases under that name and frame, does more than simply "count the dead." Rather, counterdata actions also play a narrative, affective and political role. This is to say that they constitute a form of necropolitical struggle – a discursive struggle over the widespread deaths of women and what they mean and who is responsible. Counting inserts into the public realm the groups' analysis of power and their uniting of disparate events under the concept of feminicide – or the concept of femicide, or MMIWG2, or Black women killed in police violence, as the case may be. As Estefanía stated when she was explaining how the Red Feminista Antimilitarista defines and counts feminicide, "It is like a political and structural and public health issue that should be included in structural criminal policies, such as extortion and robbery. It should also be in state policies. That is why we expand it so much, and that is how we classify whether or not it is feminicide."43 One of the ideas that activists were most adamant about was how missing data about gender-related violence were related to the lack of a name for the phenomenon and the lack of a power analysis that would link disparate events together. For example, Fabiola from the Grupo Guatemalteco de Mujeres (Guatemalan Group of Women) states, "Giovanna was mentioning that moment where there was no official data. There was no acceptance that there was this problem of violence."44 This naming and framing, backed by counting and systematizing, is a central part of the necropolitics of anti-feminicide data activism.
Naming structural violence and assembling counts and registries does not immediately move the actors that counterdata groups care about influencing the most: the state, the media, the families of victims and the people in their everyday lives. Groups find that they must use creative and intentional strategies to circulate data to influence these specific actors. These methods – along with their necropolitical effects – are addressed in more detail in Chapter 6 about Refusing and Using Data. As Mariana from the Manuela Ramos Foundation in Peru described, "...there are some media outlets that still don't consider feminicides as such. They don't write about them, they don't name them. They just put 'murdered woman'."45 Audrey from Counting Women Kenya recounted how femicide is not a known word in her country and how she is constantly teaching people about its meaning46. And yet, Ida from Cuántas Más also sees social transformation happening in regards to solidifying feminicide as a concept: "Before 2013, in Bolivia, there was talk of 'crimes of passion'... It's a fairly short history of how the fact of using a word changes a country."47
The grassroots groups we interviewed often alluded to the fact that death is the most visible tip of an iceberg (figure 3.4); or, another common metaphor, it was an extreme of a continuum of violence. Violence, despite its prevalence, is a notoriously difficult concept to define and measure because it is "multifaceted, socially constructed, and highly ambivalent"48. The highly influential concept of a "continuum of violence", proposed in 1988 by Liz Kelly following her survivor-centered research, describes the idea that acts of sexual violence are not necessarily discrete events, but shade and blend into each other, and almost inevitably reinforce each other. The iceberg metaphor was created by feminist activists to represent this idea and proposes that some parts of the continuum are more explicit and visible in public consciousness (murder, rape, insults, threats) while others remain below the waterline and are rarely seen as violence (sexist jokes, ads, microaggressions) either in law or in culture.
Myrna Dawson, founder of the Canadian Femicide Observatory for Justice and Accountability puts it this way, "we all know that many of these cases start with violence, previous violence. They don't always become fatal right off the bat, whether it's intimate partner or non-intimate partner."49 This awareness – of the connectedness of fatalities to prior, less sensational forms of violence – was an explicit part of the analysis offered by more than half of the groups we interviewed. While they focus their attention and counting efforts on feminicide, they also try to use public attention on that concept to talk about less visible, less spectacular, and harder-to-count forms of gender violence: misogyny, sexist language, street harassment, harmful and degrading cultural stereotypes, economic violence, disrespect, dehumanization, racialization, erasure of trans and LGBI+ people, failure to hire and elect women leaders, microaggressions, exclusion of women and gender-non-conforming people from masculinized professions, and so on.
This is where counterdata about feminicide and gender-related killing becomes strategic and constitutes a form of charismatic data. This is a concept introduced by Max Liboiron and developed by Pine and Liboiron to denote measurements by activists to make a previously invisible form of harm visible, as well as to focus on metrics that promote dramatic or emotionally-moving visions of that issue. Charismatic data functions simultaneously as "scientific evidence and proof of a moral imperative, and thus has the potential to launch action the way other data may not"50. As the most extreme form of gender-related violence, feminicide is a (sort of) tractable measurement problem, but it is also a public-attention-gathering gateway to discussing the forms of violence that precede it and their structural roots. In this way activists can use the frame of feminicide to expand public dialogue to other forms of gender violence. This rationale is embodied in the way feminist journalist María Florencia Alcaraz explains the hashtag #NiUnaMenos: "It means 'no more femicides' but it is also a demand to stop all forms of oppression, from the most extreme violence to the invisible unpaid work of women; from unsafe abortions to daily harassment on the streets. 'Not One Woman Less' amalgamates each and every one of the forms of gender violence suffered by women, lesbians, travestis and trans people."51 Each issue that Alcaraz mentions here is complex and multifaceted. Many of these issues – e.g. invisible unpaid work of women – are less charismatic and harder to measure than feminicide. Yet connecting the charismatic issue to the less charismatic, but related, issues opens a gateway to a more comprehensive, structural public understanding of both the depth and the breadth of gender-related violence.
Some counterdata groups do in fact monitor other forms of violence. For example, Mumalá collects data about hate crimes against LGBTQ+ people and has run surveys about street harassment in Argentina. The Observatory for Gender Equity in Puerto Rico has conducted studies about domestic violence during COVID. Counterdata groups expressed interest in collecting other types of data in order to reinforce their structural analysis of gender violence, but are often constrained by time and resources. For example, the Argentinean observatory Ahora que sí nos ven (Now that they see us) stated that although they work hard to build data on feminicide, "...we would like to have the time, tools and resources necessary to be able to work on other types of violence. About economic violence, about sexual violence, to be able to work on the issue of gender violence in a more structural way, and not just the tip of the iceberg."52 And the Red Feminista Antimilitarista aspires to expand their feminicide observatory to track cases as they proceed through the justice system so that the database doesn't only represent a registry of murders but a registry of justice.
Grassroots activists produce data about feminicide in order to solidify and stabilize the concept of feminicide, and to bring a structural analysis of gender violence into law, policy, media and culture. But which feminicide? Who gets to define, describe and typify the concept? Groups often mobilize their power analysis and their data to push back against narrow or exclusionary definitions of the concept coming from the state and the media, but also from other activists and advocates. For example, while the Red Feminista Antimilitarista wanted to connect their feminicide observatory to legal and international conceptions of feminicide, Estefanía describes how they found the emphasis on intimate partner feminicide – prevalent in Latin American legislation and activism – to be limiting in for women in Colombia: "[We do] a lot of context analysis. It is what we are passionate about. And it is a lot of analysis of our context in which we women live in Colombia, in a country that emerged from more than 30 years of armed conflict, the redistribution of men in arms, and so for us it was very limiting just to talk about feminicide in the framework of erotic-affective relationships, with the level of militarization."53 Given this imperative to understand feminicide in their region, the group created the aforementioned concept of neoliberal feminicidal violence (see figure 3.1) as well as created categories of feminicide in their database related to organized crime and to the Venezuelan refugee crisis. Estefania sees these theoretical innovations as a way to localize to the particulars of Colombia as well as an offering back to the broader dialogue on the definition, categories, and causes of feminicide: "For us, in structural terms, we have to expand the analysis of feminicide, but it is not about expanding it for the sake of expanding it. It is that it is necessary to do so."54
The desire to use counterdata to deepen and contextualize a geographic and regional analysis of violence is echoed by other groups. Sociologist Mariana Mora runs the project Cartografía de Femicidios en Costa Rica (Cartography of Femicides in Costa Rica) and described how feminist scholars in the country such as Montserrat Sagot had been instrumental in formulating early concepts of femicide drawing from the Central American experiences of gender violence. However, in Mariana's estimation there had been less scholarship since the passage of Costa Rica's legislation in 2007, and the circumstances of gender inequality had changed. This motivated her to initiate her mapping project: "[there are] new contexts of violence and inequality in which feminicides are now happening in the country."55 This includes rising violence against women from organized crime and drug trafficking. Likewise, in Mexico, María Salguero explained how she uses the UN's Latin American Model Protocol – the international standard described in Chapter 2 – as a foundation for defining and categorizing feminicide in her database and yet "the Mexican reality is different, it overflows any protocol."56 This is specifically due to the prevalence of feminicide due to organized crime and narcotrafficking in Mexico, so in recent years María has focused her attention on analyzing how women's bodies are used as weapons of war, profit, and territorial domination57.
Activists use their counterdata projects not only to contest, expand and adapt definitions of feminicide to their unique regions and geographies, but also to center the importance of intersectional differences along axes of race, class, sexual orientation, gender identity and/or indigeneity. For example, Rosalind Page is a Black nurse based in Arkansas. Over her decades of professional experience doing intake with patients in hospitals and clinics, she routinely asked questions about sexual abuse and domestic violence and was struck by how many of her Black women patients had experienced violence. And yet the issue was not discussed publicly: "I noticed that within the Black community, the focus is not on women and girls. The focus is on Black men and boys who are victims of police brutality, you know, any kind of law enforcement-involved violence. And, not to take away from that issue, but I feel that within the Black community, there is a deeper conversation that we need to have."58 Rosalind started Black Femicide US – a database and social media presence – to try to instigate such a public conversation about gender violence specifically within the Black community.
While Rosalind uses the broad frame of femicide, and sees her work in dialogue with the work of activists in Latin America, other groups choose to be more specific in how they name the violence. For example, as part of the #SayHerName campaign, the African American Policy Forum compiles a database of Black women killed in police violence, and is primarily focused on highlighting the ways in which Black women are left out of the national US narrative around police brutality. The Sovereign Bodies Institute (introduced more in-depth in the next chapter), centers their data production efforts on Missing and Murdered Indigenous Women, Girls and Two Spirit people (MMIWG2) across the Americas. Annita Lucchesi describes how they have, in fact, had challenges collecting data in Latin America where femicide is so central to the discussion of gender violence because "indigeneity or even race in general is just kind of left out of that narrative. There's lots of national and international discussions of femicide, but the racially specific portion of it is missing."59 While data activists seek to stabilize and scale the concept of feminicide, groups also use their counterdata to contest the concept and draw attention to who, what or where is disregarded and marginalized within that frame.
Thus, as part of their examination of power, counterdata groups are enacting what Patricia Hill Collins calls the "heuristic" use of intersectionality: asking who and what and where is centered in activist knowledge production about feminicide. For Collins, this kind of critical analysis is generative because it can help identify areas of overemphasis and underemphasis – to reveal, for example, that the specific violence that Colombian women face is not captured in dominant framings of feminicide or that Indigenous women are marginalized within activist discourses around feminicide.60 Counterdata groups' analysis of power – in their context, for their communities – serves as both a reinforcement of the importance of analyzing structural violence but also, and crucially, a critique of which nations, communities, races, and genders are centered by reigning concepts such as "feminicide" or "police violence". They are simultaneously critiquing the framing of these concepts (drawing the focus towards an intersectional analysis that includes gender and nation, gender and racial domination, gender and settler colonialism) and expanding or contracting the frame to include people who have been excluded. What this action-tension enables is, potentially, what Collins calls "new forms of transversal politics" to confront violence61. Transversal politics include retaining groups' rootedness in their own contexts and insistence on serving their communities while simultaneously building coalitions across differences. As we will see in Chapter 5 on Recording, these tensions play out around categorizing and classifying, aggregating and standardizing data about feminicide and fatal gender-related violence. Fundamentally, they are questions of what – and who – counts as feminicide.
As grassroots activists resolve to monitor violence, develop their analysis of power and begin researching cases, they do not always foresee the challenges involved in doing this work. Here the story of Cuántas Más is emblematic of these struggles. First and foremost, Raisa and Ida described how their work required tremendous emotional labor, which is a theme that we will return to often in this book. "It is not easy to read ten cases of femicide and put them on a table, disaggregate them, to have to put a name, age, circumstances and all that detail, without it affecting you emotionally," explained Raisa62. Caring for one's own mental health and for one's team as they encounter and document stories of brutal violence becomes paramount. Ida described that if they were to do it again they would allocate resources differently to prioritize team care. What might this mean?
Helena Suárez Val has written about data production as care work – where this may mean caring for oneself in the process of numbers production, caring for one's team, or caring for the data themselves63 . Almost every group we interviewed raised the issue of emotional labor and self- and team-care. Most anti-feminicide data activists have developed strategies for sustaining the work amidst the significant exposure to violence that it entails. In the essay, "Feminicide data, emotional labor and self-care," Helena and members of our team synthesized these activist strategies into themes such as daily maintenance, managing the affective relationship and feminist practices of care [CITE]. But sometimes the realization of the need for such strategies can come too late. In retrospect, Ida says, while the project was allocating the bulk of their funds towards trainings for journalists and public policy advocacy, they should have been putting at least 20% of their available resources towards taking care of their team64.
While the project was active, Cuántas Más gained the attention of the popular press, who often cited their figures; municipal governments, who sought them out for collaborations; and feminist journalists, who came to them seeking trainings and workshops. But this attention also proved to be challenging for a small team. Families reached out to the team with case files and requests and the team got involved with policy advocacy. "We wanted to do everything," recounts Ida, "We wanted to respond to the dizzying array of questions and proposals that came our way."65 They ended up spending a lot of time on capacity-building for individual journalists, teaching them how to report on feminicide using a gender lens, but then those journalists would leave the news outlet and "your work has gone down the drain because you have to start over."66 They realized that the people that really call the shots over cultural narratives are the owners of the media companies. The increased visibility of the project also led to media outlets using their data without crediting their work, which was extremely frustrating.
A final and significant challenge was economic. Raisa asked, "How do we keep ourselves afloat – not only us, but the team of people who supported us at that time? And how do we support them financially so that the project doesn't collapse, right?"67 Cuántas Más was sustained by small grants from international development organizations, but these came with onerous reporting requirements. Estefanía affirms that economic security has also been one of the biggest challenges for the Red Feminista Antimilitarista. The Observatory has been the least funded of their programs: "We don't know what is going to happen next year, if we are going to have resources or not, if we are going to continue or not. What is going to happen with the Observatory? All of that is very emotionally distressing, beyond the emotional weight of seeing death all the time, murder all the time."68
For these reasons and more, the labor of counterdata production is significant and difficult to sustain. This is true even (and especially) once projects acquire legitimacy and visibility in the eyes of media, policymakers, and families. Cuántas Más paused their feminicide monitoring effort in 2017 for all of these reasons. Nevertheless, some groups have found a way to persist across years and even decades. Some of the longest-running efforts include the Casa del Encuentro in Argentina, which has been counting femicide since 2008 and has had the benefit of a secure funding stream from a partnership with a national foundation. In Mexico, Letra Ese has been monitoring LGBTI+ fatal violence since 1996. Similarly, the Grupo Guatemalteco de Mujeres (Guatemalan Group of Women) has been monitoring femicide for more than 25 years through a combination of small grants and tenacious volunteer labor.
As scholars like Paola Ricaurte have argued, feminicide data activism represents one approach to developing alternative data epistemologies "that are respectful of populations, cultural diversity, and environments.”69 This is in contrast to what I have been calling hegemonic data science – mainstream data science which works to concentrate wealth and power; to accelerate racial capitalism, perpetuate patriarchy, sustain settler colonialism; and to exacerbate environmental excesses and social inequality. What exactly is it about these activist data practices that diverge in such a fundamental way from hegemonic data science? What might mainstream data scientists stand to learn from grassroots feminist activists, at each stage of their process? In each of the next four chapters, I will surface some of the key ways that – from my perspective – feminicide counterdata epistemologies differ from those used in hegemonic data science.
The motivations described in the resolving stage surface some key differences between the counterdata science projects discussed here versus hegemonic data science projects. First, feminicide counterdata science projects begin with a power analysis and the motivation to move a problem from the personal to the political – or perhaps more precisely, to move a problem from the interpersonal domain of the matrix of domination to the hegemonic, disciplinary and structural domains. Counting and aggregating the phenomenon in spreadsheets and databases becomes a way to render diverse events alike, to give them a name (e.g. "feminicide" or "femicide" or "MMIWG2"), to place them under the same power analysis, and to assert them into the public realm as empirical evidence of a structural problem. In contrast to such considered, intentional and political production of numbers, hegemonic data science has virtually nothing to say about the production of data. Data are assumed to exist as raw, apolitical observations and are either harvested through distant processes of extraction (e.g. scraping) or else produced by other people and institutions (i.e. not the data scientists themselves).
In hegemonic data science, the "science" part supposedly comes via the sophisticated methods for combining and analyzing and operating on data, not with a rigorous, scientific analysis of the conditions of its production nor of what and who is left out, erased, and misrepresented in available data, nor of how data and AI may produce harm. This is changing with work in which investigates sources of harm in machine learning and datasets, as well as work that proposes more transparency about their conditions of production and their limitations70[CITE]. But there is still little understanding in computer science that to produce a dataset is to bring a world into being – again, following Pine & Liboiron's statement that "measurements make things"71. This to say that data do not represent reality in some kind of 1:1 way. Rather, they play a strong role in producing, naming, and framing reality (and, likewise, in casting aside other possible realities) so this makes them political no matter what the circumstances. Datasets, then, are more similar to other forms of representation and communication - such as photographs and maps – than technical disciplines would like to believe. "Maps make worlds," asserted geographer John Pickles in a 2004 book on cartographic representation72. Counterdata science groups leverage this world-making affordance of data production strategically, in our case to insert a feminist frame on structural violence.
A second key difference in counterdata science about feminicide has to do with the relationship of the numbers-producers to the people they are counting. In his 1995 now-classic book "Trust in Numbers", Theodore Porter asserted that "quantification is a technology of distance," by which he meant that using numbers was a standardization technique, meant to overcome both geographic distance as well as distrust of doing business with strangers73. This holds true in many cases, particularly for knowledge produced by the state or corporations, where quantification is a management and disciplinary strategy to tame vast territories, resources and populations74. It often holds true in hegemonic data science which values distance in knowledge production and prioritizes disinterested inquiry and sees subjectivity and proximity to the object of study as a hazard. This stance is what Ruha Benjamin frames as "imagined objectivity" and is one of the many layers of "the ideology of white male supremacy" according to Adrienne Rich75.
However quantification is not always a technology of distance. For feminicide counterdata science groups, project leaders and organizers are most often from the communities that they are quantifying. Some are themselves survivors of gender-related violence, and many, like the Red Feminista Antimilitarista, work directly in their communities on a variety of violence prevention, justice-seeking and power-building projects. In these efforts, the positionality of the knower – as a person in the community and of the community – is crucial. The knower is proximate to what is being investigated. They are not disinterested, they are deeply and intimately interested. Project leaders' lived experience is valued, and they draw from this to build trust as well as to deepen their power analysis. Instead of quantification functioning as a technology of distance, then, it becomes a technology of proximation and pedagogy: a method for undertaking consciousness-raising in their community and working towards a collective understanding of structural injustice.
One final point on this matter, emphasized by both Cuántas Más and the Red Feminista Antimilitarista, is that the most technically challenging part of the work is not managing or analyzing their data, nor producing data visualizations. Both groups felt that the hardest part of the work to orient newcomers to is in the structural and feminist examination of power at the heart of the work. "Sure, you can automate certain things," reflects Ida, but there are other things that you can't, and you have to invest in your team's understanding of this feminist discussion which is so important."76 For Estefanía, "the Excel sheet is very easy. You go along putting the person who killed her, it's very easy. But to understand which case is or isn't feminicide requires this conversation, this analysis of context, this theoretical comprehension."77 Here both groups articulate the importance of developing their own and their teams' collective understanding and analysis of power. This is a theoretical, interpretive and pedagogical challenge – a consciousness-raising challenge – not a "can-you-use-this-api" or "can-you-make-a-predictive-model" challenge. This is a third and final difference with hegemonic data science, which tends to focus downstream in the data science process, developing innovative methods for analyzing data, but not for analyzing power, framing problems, or producing data. Drawing out a lesson from data activists, then, the resolving stage would ask mainstream data scientists to back up from a narrow technical focus on innovation in data operations and invest time in examining power upstream and building capacity on the technical team to navigate informational inequalities downstream.
In his 2022 book, Design as Democratic Inquiry, Carl DiSalvo makes an eloquent case for exactly such problem-making (rather than problem-solving) as part of an experimental and democratic design practice: "The outcome of inventive problem-making often directs us someplace other than design to address problems." This is to open up the possibility that specific methods – such as design, computing, data – may be used to reframe problems rather than to accept dominant problem-framings. Feminicide counterdata activists are using data precisely in this way to refuse and reframe gender-related violence, remaking the problem itself rather than "solving" it with data78.
Resolving is the first stage of a feminicide counterdata science project in which activists decide to address structural violence and determine how and why counting and registering data will be one effective method to do so. As they get started, all activist projects examine power: they develop a structural analysis of inequality. They also begin to challenge power with data: they develop a theory of change that involves systematically documenting cases. While activists have a variety of motivations for initiating their projects, I have made the case that they use counts and registers of violence not to "solve" the problem of feminicide but to remake the problem, to give it a name, and to transmute it from the personal realm to the political realm. That said, feminicide itself remains a contested concept and counterdata projects do not only use their analysis of power to challenge narrow state and media conceptions of the issue, but also to contest, expand, or narrow dominant activist framings of the term.
While I have characterized resolving as a "first" stage, it is important to note that groups' analysis of power is continually evolving as they research and record cases, and as they circulate their data to various audiences. Each of these stages feeds back into the resolving stage, leading groups to solidify or modify their analysis of feminicide, or to allocate resources differently, or to form new alliances. Resolving, then, describes a dynamic and on-going process of problem-making, where activists are continually refining their analysis, seeking new strategies to acquire and record information, and to influence the state, the media, and the broader public with counterdata.