Cause of death: femicide

ABSTRACT

Labelled ‘the shadow pandemic’ by UN Women, violence against women received considerable global public attention during 2020–21. Underpinning this moment of public concern, there lies a substantial history of efforts to document the nature of, and campaign against, the extent of violence against women globally. This is also the case in relation to femicide. Whilst we recognise that this is a contested term, for the purposes of this paper we use femicide to refer to the killing of women and girls because they are female by male violence. Femicide, as a death to be specifically counted in law only exists in a small number of jurisdictions. Where it is so recognised, primarily in South American countries as feminicidio, such deaths represent only the tip of the iceberg of such killings globally. This paper, in drawing on empirical data from a range of different sources (including administrative data, media analysis, and Femicide Observatory data) gathered throughout 2020, considers: what it means to call a death femicide, what implications might follow if all the deaths of women at the hands of men were counted as femicide, and the extent to which extraordinary times have any bearing on this kind of ordinary death.

KEYWORDS:

• Femicide
• feminicidio
• counting deaths
• covid-19 pandemic
• violence against women
• media analysis

Introduction

The ongoing global coronavirus pandemic has brought to the fore the precarity of everyday life for everyone. This kind of precarious life is one in which women and children globally living with violence(s) come to accept (and resist) as normal, routine, ordinary. Sociologists such as Bauman and Beck have differently focused their theoretical attention on the liquid precariousness of life (Bauman, 2005) along with the potential consequences for everyday life of manufactured risks (Beck, 1992). Each of their respective concerns arguably had their origins in extra-ordinary times (the events of the second world war in particular) and as such provide important landscapes for understanding the present. Both precarity and risk, ordinary and routine for women living with violence, have been the visceral and ongoing focus of the human condition during the global pandemic of the second decade of the twenty-first century. Precarity and risk have been rendered ordinary for us all.

Questions remain, however. For example, what happens to understandings of the ordinary experiences of women living their lives in the constant presence of the threat of violence and death (precarity and risk) when this kind of threat is (differently) normalised for everyone? How might these women and children’s experiences be understood, measured, and responded to? This paper explores these questions by considering what lies behind understanding the cause of a death as being femicide and what might be required to recognise such deaths as the ordinary deaths of the precarity and risk of their everyday lives.

In endeavouring to answer these questions, this paper falls into four parts. The first briefly reflects upon the contested nature of femicide, the efforts underway to measure it, and overviews the associated strengths and limitations of these efforts. The second part considers these efforts in the context of what commentators perceived to be the likely consequences of the various public health policies put in place around the globe to counter the spread of coronavirus, namely, an increase in femicide. Set against the evidence to date, the third part of the paper offers a critique of the assumptions underpinning these commentaries and provides a deeper consideration of what the current approaches to counting femicide might reveal. In the fourth and concluding part of the paper the case is made that such commentaries were misplaced because they failed to appreciate the nature and extent of the marginalisation of women’s deaths at the hands of men. Such marginalisation is layered. These deaths are simultaneously marginalised from counting procedures because of varied legal and conceptual understandings of femicide and further marginalised because of the implicit conceptual framework used by many in this field to make a case for better counting practices. Once these processes of marginalisation are exposed, men’s behaviour in relation to all of them is rendered visible and arguably more accountable.

Thus, for the cause of death to be more accurately and consistently recognised as femicide, the perpetrator, committing ordinary violence in ordinary and extraordinary times in a wide range of different contexts, needs to be front and centre in our understandings of what it means to say that the cause of death is femicide. This assertion has wide-ranging implications for how women’s deaths are labelled and thereby counted as, or not as, femicide. However, it remains the case that the term ‘femicide’ is itself contested and different counting priorities emanate from this contestation. It is to a consideration of these debates that we turn first of all.

Contesting femicide

There has been considerable debate concerning the term femicide (Dawson & Carrigan, 2020; Pierobom de Avila, 2018; Wallkate & Fitz-Gibbon, 2022) with Corradi et al. (2016) outlining the interconnections between different definitions of this term and the varied approaches taken to make sense of it. Summarising extensive work on this issue, the deaths of women at the hands of men because they are women are variously referred to as femicide (Radford, 1976, as cited in Russell, 1992), feminicide/feminicidio (terms introduced by South American feminists to capture the complexity of the state in such deaths, see as an example Lagarde y de Los Ríos, 2006), and intimate femicide (for a summary of the use and development of this term see Walklate et al., 2020). Each of these terms puts a slightly different emphasis on the different mechanisms contributing to femicide but all agree on its gendered nature. However, this definitional contest also carries consequences for how femicide is understood, responded to, and subsequently counted as the work of Corradi et al. (2016) and Walklate et al. (2020), amongst others has documented (see also the recent report produced by EIGE, 2021a). For the purposes of this paper, it will be of value to say a little more about the different counting practices generated by this debate.

There are currently two main data sources offering counts of femicide; administrative data and that data gathered by various femicide observatories around the world. Many of these observatories were formed in the aftermath of the call made by the United Nations in 2017 for the establishment of femicide watches or observatories in every country arguing that:

Having clear data helps law makers and government officials win the public’s support for tackling it through targeted prevention and investigation resources. Academic Council on the United Nations System (ACUNS, 2017: 1).

Despite this high-level recognition of femicide, divergent counting practices exist in different parts of the world and these differences make it difficult to assess the nature and extent of femicide as a global problem (EIGE, 2021b). So, for example, where counting is informed by an offence of femicide existing in law (which can vary across the Global North and the Global South, see Pierobom de Avila, 2018), this produces one form of counting.

As the work of Walby and Towers (2017) demonstrated administrative data emanating from legal frameworks does not necessarily or consistently record the sex of the victim and/or the sex of the perpetrator across different jurisdictions. In contrast data produced by femicide watches/observatories is not necessarily informed by what is legal or illegal but is more likely to be informed by what the watch or observatory judges to be an act of femicide. This can depend on their understanding of the motive of the perpetrator and the gender of the actors involved. Relying on either of these sources of data carries risks including the problems of gaps in data and gaps in understanding the meaning of the act of killing resulting in a woman’s death. Elsewhere we have argued that taking these risks together results in administrative data producing ‘thin’ counts (Wallkate & Fitz-Gibbon, 2022). Moreover, even femicide watch/observatory data fail to grasp the full picture of what Walklate et al. (2020) call ‘slow femicide’, or what has more recently been expanded as the ability to produce ‘thick’ counts (Wallkate & Fitz-Gibbon, 2022).

Thus, counting femicide is fraught with difficulties and complicated when claims about the nature and extent of the global problem of violence against women are made when based on such counts. Such counts can hide a spectrum of different counting practices, definitional issues, and data gaps. In sum, none are complete, consistent, or coherent. They each reveal varying issues associated with, and offer a different picture of, the femicide story.

None of the above means, of course, that counting is unimportant. On the contrary as, Walklate et al. (2020: 96) stated:

Data alone won’t save women’s lives but counting the killing of women by men is a way of valuing women’s lives, accounting for gendered killings and a tool for further action. Data is not sufficient to change and save women’s lives, but it is necessary.

However as Merry (2016) has argued relying on numbers is a perilous process especially if the focus of measurement is to provoke changes in policy and understanding as much work in this field is. Yet, during the global pandemic, the search for numbers (especially in relation to violence against women) as a way of influencing policy, perhaps understandably, grew apace with many regarding this moment as a policy window: an opportunity for change (see Mintrom & Jacqui True, 2022). It is to these commentaries, informed by the search for numbers on femicide during the global pandemic, that we now turn.

Understanding femicide during the Covid-19 pandemic

Femicide was already a feature of the global agenda, as articulated by the UN Sustainable Development Goals (SDGs), by the time the pandemic made its presence felt in early 2020. For example, in 2018 the UN Office on Drugs and Crime (UNODC) found that 87,000 women were victims of femicide in 2017 alone of which 30,000 women were killed by a current or former intimate partner (United Nations Office on Drugs and Crime (UNODC), 2018). This UNODC (2018) report also provided country-level data for Europe highlighting a higher-than-average risk of intimate femicide for women living in Albania, Croatia, Hungary, and Lithuania when compared with other European countries. According to this, same UNODC data, women in Asia were the least likely globally to be killed by their male intimate partner. Statistics such as these, and other country level data, had already started to fuel the view of femicide and violence against women (and children) more generally, as a global crisis. Thus, when the global pandemic hit, it did so in a context in which the private violence(s) by men against women were already widely recognised. Arguably, this recognition, in part, was what led UN Women (2020) to name the possible consequences of the stay-at-home and similar restrictions introduced in many jurisdictions as likely to result in a ‘shadow pandemic’ of violence against women including femicide. This term gathered significant momentum throughout the pandemic and became the go-to term used by the media, policymakers, academics and advocates worldwide to capture concerns surrounding the impact of the global health crisis on the relevance, and nature of men’s violence against women (in Australia, see for example, Pfitzner et al., 2020b, 2022 in United Kingdom, see for example, Navarre, 2021; and in United States, see for example, Gabriel, 2020).

These concerns build on an already existing body of the literature pointing to the relationship between different kinds of crises and increases in gender-based violence(s). Research following the 2008 Global Financial Crisis (Bhattacarya, 2013–2014), the Australian bushfires (Parkinson & Zara, 2013), natural disasters in several countries (Harville et al., 2011), work in India (Rao, 2020), the Philippines and Vietnam (Nguyen & Rydstrom, 2018), Iran, (Sohrabizadeh, 2016) and Japan (Yoshihama et al., 2019), all consistently demonstrated the heightened risk of intimate partner violence and increased stresses in family life during and post-crisis. However, these studies are all largely focused, almost exclusively, upon the increase of non-fatal violence against women. It remains unclear to what extent these heightened risks have extended to an increase in femicide during the global pandemic though the media in several countries reported such a link. It is to this question and coverage that we now turn.

Throughout 2020, media coverage collected by the authors between January and September 2020 of the coronavirus crisis suggested a link between the pandemic restrictions and the prevalence of femicide. Specifically, using a range of open access search engines (i.e. Google, Bing) and databases (i.e. OpenGrey, NGO search, IGO search), a systematic search as undertaken using the following combination of search terms to identify relevant media articles: ‘femicide covid’, ‘femicide coronavirus’, ‘murder women covid’ and ‘murder women coronavirus’. The search was then confined to English language articles.

This search strategy resulted in the collection of 97 media articles and public reports relevant to femicide and the pandemic, of which 37 articles specifically explored the relationship between the pandemic and the killing of women. For example, Grierson (2020) reported in April of that year, the findings of ‘Counting Dead Women’, a UK based campaign group, who had identified 16 alleged domestic abuse killings over a 10-day period. This report suggested that this rate was three times higher than the average for the same time period in other years. Media headlines covering these deaths included, ‘Domestic abuse killings “more than double” amid Covid-19 lockdown’ (Grierson, 2020) and ‘3 times more women killed by men than average during U.K. coronavirus lockdown, data show’ (Ott, 2020). However, as Karen Inglis Smith (one of this campaign group’s leaders) explained:

We have to be cautious about how we talk about increases in men killing women … But we can say that the number of women killed by men over the first three weeks since lockdown is the highest it’s been for at least 11 years and is double that of an average 21 days over the last 10 years. (cited in Grierson, 2020)

Similar media coverage emerged in South Africa, also during 2020, but interestingly rather than being attributed to a period of COVID-19 restrictions, the coverage coincided with the easing of restrictions. For example, one journalist reported:

A wave of killing of women and children has horrified the country in recent weeks since the gradual easing of coronavirus lockdown restrictions on June 1. The police say the end of the nine-week ban on alcohol sales contributed to a spike in crime and gender-related violence directed at women and children. (Egwu, 2020)

Reflecting on such concerns more broadly, in June 2020 media reporting following the publication of a report by Action Aid warned that the world was ‘sleep walking into the shadow pandemic of global femicide’ (Action Aid, 2020).

However, when looking beyond media reporting specifically, other data sources are less clear on the link between fatal violence and the impact of confinement or stay-at-home directives. A UNODC Research Brief published in November 2020 reported a drop in alleged femicides in Mexico in May 2020 (when stay at home restrictions were in place) with an increase in June 2020 (when restrictions were relaxed). Comparing data for 2019 with 2020 in Brazil, that same report shows a 2.2% increase in femicide for Brazil. Data for Peru and Ecuador were more mixed (a decrease in Ecuador in March 2020 during confinement measures followed by an increase in May, whereas in Peru the decrease in these cases seems to have been longer lasting through to August 2020). In Paraguay, no changes were recorded. In European countries, the data are also very mixed. For example, a report from the Observatory of Murdered Women in Portugal suggested the statistics during 2020 were comparable with earlier years at the time of reporting, being August 2020 (OMA UMAR, 2020) while a study conducted in England and Wales examining lockdown restrictions on domestic homicide during 2020 reported that:

Despite the unprecedented circumstances of the last year, domestic homicides do not appear to have increased substantially in the twelve months since the start of the Covid-19 pandemic and related restrictions in the UK. (Bates et al., 2021: 69)

So, as this brief overview suggests, and in line with the indications of the UNODC Research Brief, whilst there is plenty of evidence that women reached out for support and sought help in greater numbers during 2020 (see inter alia Peterman & Zara, 2013, Peterman, et al., 2020a; Pfitzner et al., 2020b, 2022) the jury remains out on the impact the global pandemic has had to date on the rate of fatal violence perpetrated by men on women. This observation is not intended, however, to imply that femicide has receded as an issue of concern. As Bates et al. (2021: 69) go on to state:

these numbers [in their study] do confirm that domestic homicides and suspected victim suicides with a history of domestic abuse in England and Wales are an entrenched and enduring problem.

Indeed, a key word Google search of 192 countries conducted for the authors in November 2020 revealed that 32 countries had femicide counts publicly available. Most of these counts came from Femicide Observatory data though in some cases additional data were publicly available via criminal justice statistics. These counts did not include violent deaths which did not comply with the local (legal) definition of femicide so excluded in some instances intimate partner homicide, and other deaths which may have been classified as misadventure, accidents, or were the result of conflicts with, and/or, at the hands of men by other violent means. These 32 counts accounted for 4604 deaths of women from 1 January, 2020, to 30 November, 2020. Deaths which happened simply because the victim was a woman. Moreover, a UNODC Research Brief published in November 2021 states that:

Some 47,000 women and girls worldwide were killed by their intimate partners or other family members in 2020. This means that, on average, a woman or girl is killed by someone in her own family every 11 minutes. (UNODC, 2021)

This report goes on to suggest however, that:

At the national level, monthly data from 14 countries in various regions show high variability in trends across countries but suggest that, overall, female intimate partner/family-related homicides remained relatively unaffected by the lockdowns in those countries. (UNODC, 2021, p.20).

Thus, the question remains, given the widely expressed expectations that the data would tell a different story, how might it be possible to make sense of the disparity between these expectations and the prevalence data on femicide so far available? Is it simply a matter of it being too soon to tell whether the Covid-19 pandemic will shift the dial on the global prevalence of femicide, or is there a deeper and more complex narrative to unravel underpinning the debate surrounding counting femicide at play here? This is the focus of the next part of this paper.

Examining the conflation between process and escalation

To make some sense of the disparities examined thus far, it is necessary, in part, to unpack some of the assumptions underpinning the connections between the nature and extent of violence against women and the lethal act of femicide. Two of these, which are inter-connected, are of note at this juncture. The first assumption to make explicit is that violent relationships, like all other relationships, are processual. In other words, the incidents that occur between people need to be understood in the context in which they occur, not as isolated individual incidents of abuse. The second assumption, which is frequently made in relation to violent relationships, is that of escalation: that is, that violence gets worse over time. Interestingly, the assumption of escalation has rarely, if ever, been subjected to conceptual or empirical interrogation (Boxall & Lawler, 2021) and its conflation with understanding relationships as processual is resultantly problematic. Put simply, not all relationships in which violence has occurred, or does occur, escalates into more violence or lethal violence and not all deaths that do occur because of such escalation are counted as acts of femicide. This observation is not only important in policy and practice terms in relation to knowing how, where, and when intervention might be appropriate or not, as well as for reporting and recording purposes, it is also potentially important for understanding the disjunction between the expectations of an increase in femicide during 2020 and the failure for this to have been realised in the data available to date.

This conflation, in directing the counting gaze to criminal justice data, frames and limits femicide counts to those incidents of lethal violence visible in such data. Interestingly in so doing, this perpetuates an incident focused approach to violent relationships at the expense of a process focused one. Why does this matter?

Arguably this matters not only because a denial of process results in a partial vision of the social reality of women’s deaths but also because it fails to make women’s lives count. Some time ago Weber and Pickering (2011: 8) observed that, ‘The body cannot be understood without recognising the life that was lost’. Reflecting the meaning intended by this observation in the context of femicide counts, [dead] bodies cannot be understood without recognising the lives that were lived in leading to that death. In other words, the processes that resulted in this death. For women living with violence, the consequences of these lives so lived can be found in deaths counted in other ways and by different means. For example, the study conducted by Bates et al. (2021) paid attention not only to domestic homicides during 2020 but also to other differently categorised deaths where intimate partner violence had been evidenced prior to the death recorded. They comment:

Most police domestic abuse leads we surveyed or interviewed said that they were relatively confident in identifying unexplained deaths and suspected victim suicides where there was a history of domestic abuse in order to submit them to this project. However, logically, unexplained deaths or suspected victim suicides where a history of domestic abuse was not known to police would not be flagged to this project. So, it is likely that each year there are more suspected victim suicides with a history of domestic abuse than identified by this project and analysis alone. (Bates et al., 2021: 70, our emphasis)

This observation provides one clue in the femicide counting puzzle and points to the need to look beyond the pure counting of homicides involving a female victim as equating to the total prevalence of femicide. Interestingly, several femicide observatories gather data on the suicide of the perpetrator (often termed femicide-suicide) as do some criminal justice statistics. See for example, the Canadian Femicide Observatory for Justice and Accountability and the Argentina Observatory on Femicide, with the suicide of the perpetrator being recorded by eight of the European member states, EIGE (2021a). Yet there is little systematic collation of data where the suicide of the woman can be attributed to domestic abuse. From what research there is a pattern connecting a woman’s suicide with her experiences of domestic abuse is observed (see inter alia, Aitken & Munro, 2018; MacIsaac, Bugeja, Weiland, 2017a, MacIsaac, Bugeja, Jelinek, 2017b). When considered within the confines of ‘traditional’ counting practices the cause of death for these women would be suicide, not femicide. We argue the latter may more appropriately reflect the factors contributing to the death than a classification of death which invisibilises the impact of male violence in the woman’s life. Similar observations can be made about deaths occurring as a result of dependent drug and alcohol use, which has occurred on the backdrop of intimate partner violence victimisation.

When considered within the context of understanding rates of femicide during the COVID-19 pandemic, this research on femicide-suicide and suicide in and of itself – as excluded from the counting of femicide – has potentially interesting implications. While not evident globally, some countries have observed an increase in suicides rates since the outset of the pandemic (Santoni et al., 2021). Concerningly, 2020 data from several countries including the United States and South Korea shows an escalation in the suicide rates for women and adolescent girls (Santoni et al., 2021). Given what is known about the high rates of violence for women and girls, and the expected escalation of violence during the pandemic period, an exploration of the degree to which some of these suicides could be more accurately recorded as femicide is an important one and one which would ensure the cause of death can be more appropriately reflected.

The limits of criminal justice administrative data in understanding femicide

Importantly, and moving beyond the invisibility of some suicides in the counting of femicide, the conflation between process and escalation carries consequences for another reason. It directs the counting gaze to look for statistical data on femicide in particular ways, namely as the lethal outcome of violent relationships per se. This gaze implicitly affords a pre-eminence to criminal justice administrative data, such as homicide incidents as recorded by police and homicide convictions as constituted by the courts (much of which either implicitly or explicitly also informs femicide watch/observatory data) as the key mechanism for counting an act of femicide – and by extension who counts as a victim and who is counted as a perpetrator of such violence. This leaves less visible a wide range of other administrative data sources in which such fatal violence might also be recorded but remain excluded from official counts of femicide. This assertion requires further explication.

As has already been asserted, relying on numbers to inform any form of counting is fraught with difficulties which require an untangling and attention to be given to a wider range of deaths than has traditionally been in view when we count femicide. Such difficulties are well known amongst those voices committed to counting as one way of raising the profile of femicide as a global problem (see, among others, Walklate et al., 2020). Moreover, as has also already been stated, different criminal justice jurisdictions may count women’s deaths differently depending on the legal frameworks in operation. Indeed, these same voices are also aware of the presences and absences that criminal justice data presents, driven by, for example, the processes associated with how a death might be defined as an honour crime (Gill & Aujla, 2014; Shalhoub-Kervorkian & Daher-Nashif, 2013); whether a death has been concealed to avoid prosecution (Bitton & Dayan, 2019) or how a death might be staged as something other than murder (Ferguson, 2015). In these varying ways, the deaths of individual women at the hands of men become categorised differently, not counting as femicide. Hiding women’s deaths as a result of male violence from view in this way can also happen to particular groups of women, as for example in Cuidad Juarez (Leal, 2008; Livingston, 2004), or as in the case of the missing and murdered Indigenous women and girls in Canada (see Corradi et al., 2016; Dawson, 2018; The National Inquiry into Missing and Murdered Indigenous Women and Girls, 2019). The killing of older women by male violence has also often failed to count as femicide in criminal justice administrative data (Roberts, 2021). The ways in which these data sources are blind to intersectionality have been discussed by Cullen et al., (2021). However, while these deaths are absent in official counts of femicide, they may well appear in other data sources, including accidental death because of a fall, for example.

Connecting back to our examination of the importance of viewing relationships as processional, it is possible to push the search for more accurate and inclusive data based on an understanding of the processual consequences of violence for women further. In a comprehensive and comparative analysis of hospital data comparing those who had hospital stays because of domestic and family violence (DFV) with those who had hospital stays for other reasons, the Australian Institute of Health and Welfare (2021: 56) found that:

From 2010 to 2019, 5.7% of the FDV hospital cohort died, compared with 4.4% of the comparison group.

The FDV group were 10 times as likely to die due to assault, 3 times as likely to die due to accidental poisoning or liver disease, and 2 times as likely to die due to suicide, as the comparison group.

The comparison group were more likely than the FDV group to die due to cancers: breast cancer (2.8 times), colorectal cancer (2.5) and lung cancer (2.0).

This data suggests that not only were the causes of death different for these groups but it also, by implication, points to two other factors; the wider health consequences and deaths resulting from domestic abuse (see also Abrahams et al., 2009 and Ayre et al., 2016 on the health consequences of living with violence), and that many of these deaths would not have been recorded as connected with such abuse if one is to look solely to criminal justice administrative data. Thus, the cause of death would not appear as femicide. Moreover, if it were possible to add to these figures the early deaths endured by women occurring as a result of living with violence but likely to be categorised as stress related (for example, heart attacks, stroke, and other stress-related diseases), the femicide picture is enhanced and a more accurate prevalence of femicide is achieved. Counting these deaths as femicide might also improve the ability to better capture the processual consequences of the ordinary lives lived under the conditions seen as ‘just part of life’ (Genn, 1988).

Taken together, these absences reveal a picture of the marginalisation of women deaths and as a result their lives. Marginalised by the way in which the counting gaze gives prominence to criminal justice data (and the associated legal frameworks which may or may not include the category of femicide to facilitate counting) and marginalised because this gaze renders less visible other data sources around which the counting voices are also less audible. Importantly, this gaze, directed in this way, marginalises process and the processual consequences of living with violence which may manifest itself (or not) in a range of ways other than specific acts of lethal violence. Ultimately of course, what this gaze does, and the counts which emanate from it contribute to, is to minimise our understandings of the prevalence of men’s violence(s) and the wide range of impacts this violence has on women’s lives, including in causing their deaths from femicide. For Shalhoub-Kevorkian (2003) these form part of the constituent elements in the wider social processes which exclude gendered analysis, contribute to its invisibility, and facilitate ongoing social abuse.

Concluding thoughts. The global pandemic and femicide: too soon to count?

The purpose of this paper has been to excavate the tensions between advocacy and media expectations that femicides would increase during the global pandemic (consequent to public health stay at home directives) and what the data available to date has so far revealed in relation to these expectations. Such expectations in many ways implicitly accepted the view that the extra-ordinary times of the global pandemic would result in extra-ordinary consequences. This view was adopted with good reason and are reflective of prior research which has shown an escalation of violence against women, both in terms of frequency and severity, during times of crisis.

When the presumption of violence escalation is taken as a given, the early reports from China in 2020 pointing to an increase in domestic abuse incidents there, raised sensitivities to the potential consequences of such increases. However, evidence in relation to such increases reported during 2020 and 2021 from other jurisdictions has been rather more mixed, as explored in the first half of this paper. In some jurisdictions, the stay-at-home directive resulted in more calls for advice/support from support services (which did not necessarily translate into increased reports to the police) with other jurisdictions reporting a different pattern (see inter alia Peterman & Zara, 2013, Peterman, et al., 2020a; Pfitzner et al., 2020b). Against this backcloth, and the obvious concern that the ordinary escape mechanisms and safety planning available to those living with violence were curtailed, added to raising the awareness of the possibilities for an increase in fatal violence(s). At the same time, of course, it is equally possible that in relationships in which the controlling presence of different kinds of violence(s) are a constant feature (digital surveillance, stalking and so on) the presence of these behaviours might have receded since such tactics were no longer necessary during a period where many victim-survivors were isolated at home, within the everyday controls, of their abusers. The jury is still out on these issues. It is indeed likely that the public health policy consequences of the global pandemic have yet to be played out in full and as the pandemic continues there will likely be additional women’s safety policy and practice considerations.

It remains the case that since 2020 the global pandemic has created a testing time for many people and for their relationships: the precarity and risk routinely experienced by women living with violences has become mundane and ordinary for everyone. The possibility that this time has increased the risk and use of violence for some (predominantly women) in their relationships may yet reveal itself in femicide data. In other words, it could be too soon to count. However, in unpacking the tensions between expectations and data, as this paper has endeavoured to do, some deeper issues have risen to the surface as we have attempted to show more significant limits in the ways in which traditional approaches to counting have failed to account for all deaths of women from men’s violence.

To re-iterate, violence for many women (and children) across the globe is mundane, every day, and ordinary (Genn, 1988). It is not borne out of exceptional events or circumstances. As Das (2007: 14) would say, it is ‘folded into everyday life’. If such violence and the fatalities which follow from it are everyday and ordinary, as Walklate et al. (2020: 97) have asked, what are the dangers in only capturing those deaths counted at the extremes, that is, those counted in criminal justice data? What happens to the ‘texture and complexity’ (ibid) of everyday life as a result? Or as Merry (2016: 219) might say, does this kind of measurement make some things visible ‘while the measured disappear’ and in disappearing the measured, does the nature and extent of the violence(s) of the perpetrators also disappear? For the cause of death to count as femicide then more work needs to be done on ensuring that all mortality data reflect the processes underpinning many women’s lives and deaths, so that such counts ensure that all women’s lives count. As Fragoso (2018: 926) in talking of feminicide states:

Feminicide has been the fate of countless girls and women in our communities over a long period of time … This unpunished violence has and will have serious implications for the kind of society that we continue to create. It is time to call it for it is – feminicide.

Causes of death that render male violence towards women invisible carry similar implications. It is time to be explicit. To call cause of death femicide in all those contexts in which the ordinary risks and precarities of the everyday lives of women become normalised when in the counting of their death the cause is rendered absent.

Disclosure statement

No potential conflict of interest was reported by the authors.

Additional information

Funding

The work was supported by the Australian Research Council [DP170100706].

Notes on contributors

Kate Fitz-Gibbon

Kate Fitz-Gibbon is Director of the Monash Gender and Family Violence Prevention Centre and a Professor of Social Sciences in the Faculty of Arts at Monash University (Victoria, Australia). Kate conducts research in the field of domestic and family violence, femicide, responses to violence against women, and the impact of criminal law reform in Australia and internationally.

Sandra Walklate

Sandra Walklate is Eleanor Rathbone Chair of Sociology at Liverpool conjoint Chair of Criminology at Monash University in Victoria, Australia. She is a lead researcher within the Monash Gender and Family Violence Prevention Centre. Sandra has worked within the field of victimology since the 1980s and much of her current work in this area is focused on gender-based violence(s) and criminal justice policy responses to this.