Theoretical Perspectives,Empirical Findings,and Policy Implications of Research on Femicide Before,During,and After the COVID-19 Pandemic: Beyond a Special Issue Introduction

Nullius in Verba

In May 2020, one of the coeditors of this special issue and cowriter of this introduction was teaching an online class to students of a master’s program organized by several Catalan universities. We were at the height of the pandemic, and he shared his worries about the potential harmful effects of the lockdowns on violence against women. According to opportunity and strain theories, which seemed the most relevant in that context in terms of their potential for explanation and prediction, he mentioned that an increase in this kind of violence was foreseeable. He stated the limits of official crime indicators, such as police statistics, in measuring violent behaviors and concluded that the only offense that could be reliably measured was homicide—because the records of people killed are relatively reliable—provided the country made a distinction according to the gender of the victim and, ideally, the characteristics of the offender. In that way, one could try to identify the cases where the victims were women, and particularly those that could be considered as femicides, even if the definition of the latter varies from country to country. Then, loyal to the motto of the Royal Society (Royal Society, 2024) that gives title to this section and that he made his own years ago, he shared his screen and began searching for data on “femicidios” in Spanish-speaking countries, given that the class was taught in Spanish. To his surprise, the available data for March and April 2020 from several countries with femicide observatories showed no increase in femicide. This unexpected finding gave him one of the most memorable moments in his research career. Not only were human lives being spared, but for once, the outcome contradicted well-established criminological theories. Predictably, he proposed studying these trends further.

The result is an article titled “Against All Odds, Femicide Did Not Increase During the First Year of the COVID-19 Pandemic: Evidence from Six Spanish-Speaking Countries,” published in issue 37/4 of this same Journal of Contemporary Criminal Justice (Aebi et al., 2021). That paper, which tries to contribute to the “southernization” of criminology by focusing on Latin American countries that are often neglected in international criminological literature, is the cornerstone of this special issue. It tests the situational hypothesis, which postulates that the number of femicides should increase as an unintended consequence of the COVID-19-related lockdowns. Using monthly data on femicides from 2017 to 2020 collected in six Spanish-speaking countries (Argentina, Chile, Paraguay, Panama, Mexico, and Spain) and analyzed using threshold models, the authors reject the situational hypothesis. In fact, the total number of femicides in 2020 was similar to that recorded during each of the three previous years, and femicides did not peak during the months of the strictest lockdowns. Furthermore, the monthly distribution of femicides in 2020 did not differ from the seasonal distribution of femicides in any former year. That finding challenges explanations of femicide based on routine activities theory (Cohen & Felson, 1979), suggesting that a concrete event, such as the perpetrator’s awareness of the end of the relationship, may be a missing element in them. The article criticizes the current state of research on femicide and its inability to inspire effective criminal policies, highlighting the need for a more holistic approach that incorporates biology, neurosciences, and alternative sociological and cultural explanations. We propose three lines of intervention: placing the study of femicide in the context of crimes against persons, incorporating biology and neuroscience approaches, and expanding current cultural explanations of femicide. It is precisely to provide that more holistic approach that we proposed to Chris Eskridge, editor-in-chief of this Journal of Contemporary Criminal Justice, to edit this special issue on “Trends in Intimate Partner Homicide during the COVID-19 pandemic: Theoretical explanations and criminal policy implications.”

Toward a Unified Criminology

This special issue is inspired by Robert Agnew’s (2011) concept of Unified Criminology:

Most theories appear to have some merit, explaining a portion of the variation in some crime. Criminologists also attempt to integrate certain of these theories. But none of the integrations has attracted wide support; partly because they reflect the divided nature of the field, combining a small number of related theories and ignoring others. And, perhaps most commonly, criminologists set up shop in their own corner of the discipline; mostly ignoring the work of criminologists in other areas, but occasionally drawing on or attacking it. [. . .] [T]here is reason to believe that each of the theories and perspectives in criminology has some relevance. Therefore, the best explanation of crime will be one that incorporates all major theories and perspectives. (Agnew, 2011, pp. 191, 201)

These words of Robert Agnew were our guide while inviting, selecting, and reviewing papers for this special issue. We were lucky enough to receive almost all positive answers to our invitations.

This special issue brings together a diverse set of articles that explore the trends, theoretical explanations, and policy implications of intimate partner violence (IPV) and femicide during the unprecedented global crisis provoked by the COVID-19 pandemic. These articles apply various methodological approaches, such as time series analysis, comparative analysis of official data, and examination of media reports and hotline data to shed light on the complex dynamics of femicide and IPV during the pandemic. Some articles delve into the specific characteristics of femicide cases, such as victim–perpetrator relationships and weapons used, whereas others focus on broader trends and policy implications. The articles draw on insights from criminology, sociology, psychology, and public health, demonstrating the value of a multifaceted approach to understanding and addressing the phenomena under study.

The theoretical contributions of this special issue are noteworthy. The articles make use of a diverse range of perspectives, including modern control theory (MCT; Gottfredson & Nielsen, 2024), developmental and life-course (DLC) criminology (Narvey et al., 2024), situational action theory (SAT; Treiber, 2024), and general strain theory (GST; Cunha et al., 2024). By applying these frameworks to the analysis of femicide and IPV trends during the pandemic, the authors provide valuable insights into the underlying mechanisms and risk factors associated with these forms of violence. Similarly, the empirical contributions are outstanding. From the global estimates provided by Dunaiski et al. (2024) to the country-specific analyses of Hacin and Meško (2024), Yehuda and Ein-Tal (2024), and Cocco et al. (2024), the studies highlight the challenges and opportunities for enhancing our understanding of femicide through comprehensive, reliable data. In addition, Strang (2024) not only proposes an analysis of trends in the United Kingdom but also presents research that has concrete policy implications in terms of femicide prevention.

Contents of the Special Issue

The papers included in this special issue are presented in an order that allows for a logical progression from theoretical foundations to empirical analyses, from global estimates to detailed regional investigations, and from broad comparative studies to focused case studies. The issue begins with an article by Michael R. Gottfredson and Mikaela S. Nielsen that provides a strong theoretical foundation and raises relevant issues for research and policy on the topic under study. The authors make a case for the relevance of Gottfredson and Hirschi’s MCT (Gottfredson & Hirschi, 2019) in understanding and preventing IPV and femicide through and analysis of empirical patterns and theoretical assumptions. They argue that general theories like MCT provide valuable explanations that are more comprehensive than those proposed by crime-specific theories focused solely on these offenses. From the perspective of MCT, IPV is predicted to be more likely among those with low self-control and weak social bonds to others. In that perspective, Gottfredson and Nielsen highlight empirical evidence showing correlations between IPV and factors like prior violence, alcohol and substance abuse, and weapon access, which align with MCT’s predictions. Furthermore, they argue that long-term trends in IPV parallel trends in other crimes and risky behaviors, suggesting common underlying causes as predicted by a general theory. In terms of prevention, they suggest focusing resources on early childhood intervention programs to build self-control, as well as restricting alcohol and firearm access for high-risk individuals.

Building on that theoretical groundwork, the special issue continues with a paper by Chelsey S. Narvey, Katie Kaukinen, Nicole Leeper Piquero, and Alex R. Piquero that introduces a complementary perspective, the DLC approach. The authors provide a comprehensive analysis of how DLC theories, specifically Sampson and Laub’s age-graded theory (Laub & Sampson, 1993) and Moffitt’s dual taxonomy (Moffitt, 1993, 2018), can contribute to understanding IPV and femicide, while acknowledging the challenges in fully addressing the gendered nature of these issues. First, the authors examine the overlap between general offending and IPV, and how DLC theories can provide insights into nonlethal IPV and femicide. In that context, Laub and Sampson’s (1993) age-graded theory suggests that strong social bonds, especially those formed during adulthood, can influence the likelihood of engaging in or refraining from IPV perpetration. This aligns with findings from feminist IPV scholarship on the importance of employment and income in reducing IPV risk. In parallel, Moffitt’s dual taxonomy distinguishes between adolescence-limited and life-course-persistent offenders. Life-course-persistent offenders, who exhibit early onset antisocial behavior and have neuropsychological deficits, are more likely to engage in chronic and severe IPV. Empirical research supports the connection between general offending and IPV, with chronic offenders being more likely to perpetrate violence both inside and outside of intimate relationships. Narvey and colleagues highlight the need for further theoretical development surrounding femicide, integrating DLC perspectives with the IPV literature on escalation, situational context, and psychopathology. They point out that during the COVID-19 pandemic, nonlethal IPV increased due to factors like financial stress, social isolation, and reduced access to interventions. However, evidence on changes in femicide rates is mixed, possibly because the pandemic did not significantly affect the mental health and neurological conditions associated with lethal IPV. Finally, the authors emphasize the importance of interdisciplinary collaboration, refining theoretical frameworks, and implementing evidence-based policies to address IPV and femicide globally.

A third theoretical perspective is then introduced by Kyle Treiber, who applies SAT (Wikstrom et al., 2024) to analyze the mixed findings regarding changes in IPV and femicide during the COVID-19 pandemic, hence starting to build a bridge between the theoretical and empirical articles in this special issue. Treiber addresses the issues of intimate partner abuse (IPA) and intimate partner homicide (IPH) during the pandemic and argues that, compared with opportunity and strain theories, SAT provides a more nuanced explanation for the mixed findings regarding changes in IPA and the lack of expected changes in IPH during COVID-related stay-at-home restrictions. SAT emphasizes the importance of understanding how motivation translates into action, why some people perceive IPA and IPH as acceptable alternatives, and how exposure to different settings shapes perceptions and action choices. The author suggests that people’s proclivities (desires and sensitivities) likely did not change dramatically during stay-at-home restrictions, but people may have been exposed more frequently to relevant motivators, activating existing proclivities. Similarly, personal morality and moral contexts likely did not change significantly, but those prone to IPA and living in environments conducive to IPA may have experienced heightened exposure, leading them to perceive IPA as an acceptable response. Treiber argues that IPH would be less likely to increase than IPA, as people rarely see IPH as acceptable, and stay-at-home restrictions were unlikely to create circumstances leading to the perception and choice of IPH as an acceptable and preferred alternative. Furthermore, people’s experiences and responses during the pandemic differed based on their circumstances and personal characteristics, and SAT offers a nuanced explanation of the conditions under which changes in IPA might be expected. Treiber concludes that only people susceptible to IPA inducements were prone to IPA under stay-at-home restrictions, and changes in their IPA depended on changes in their exposure to these inducements. The same applies to IPH, but with fewer people being prone to IPH and fewer changes in IPH inducements, SAT would not predict the same increase in IPH as IPA.

Strengthening the bridge between theory and practice, the fourth paper of this special issue is written by Olga Cunha, Sónia Caridade, and Rui Abrunhosa Gonçalves, and focuses on Robert Agnew’s GST (Agnew, 2012) as a theoretical framework for an in-depth analysis of femicide trends in Portugal during and after the pandemic. The authors explore intimate partner femicide (IPF) and attempted intimate partner femicide (AIPF) in Portugal from 2018 to 2022 using a comprehensive strain-based approach. IPF rates in Portugal during the COVID-19 pandemic remained relatively stable, following a pattern similar to other countries, whereas AIPF rates increased, with a slight uptick in IPF toward the end of the pandemic period in 2022. The prevalence of prior violence in IPFs and AIPFs was high, and a substantial number of IPFs ended in suicide or attempted suicide by the perpetrator. The authors propose that the GST offers a valuable perspective for explaining IPF and AIPF, emphasizing the role of emotions and mental health in influencing these dynamics. Various strains related to the COVID-19 pandemic, such as social isolation, economic losses, anxiety, and fear, can generate negative emotions that may lead individuals to adopt destructive behaviors, including IPF and AIPF, as coping mechanisms. The pandemic restrictions also limited victims’ access to support and protection services, exacerbating the challenges faced in seeking help. Cunha and colleagues emphasize the need for a comprehensive approach involving social, medical, and legal responses, as well as community-based initiatives and public media campaigns, to effectively prevent AIPF and IPF. Practical implications for shaping public policy and preventive programs include implementing educational and awareness initiatives, investing in victim support services tailored to the specific needs of vulnerable women during crises, and expanding awareness campaigns to eradicate attitudes and behaviors that perpetuate violence against women.

Shifting focus to a specific risk factor that was highlighted also in Cuhna et al. (2024) findings—suicidal behavior—Heather Strang offers in the fifth article of this special issue insights into the prediction and prevention of domestic homicide, while also discussing broader trends in femicide in the United Kingdom. Strang focuses on suicidal behavior as a risk factor for domestic homicide and the need for collaborative efforts between various agencies to identify and intervene in high-risk cases. First, she challenges the assumptions that domestic homicides are always foreseeable due to escalating violence and that police or other agencies should have predicted that risk. In fact, research shows that a significant proportion of high-harm cases have no prior police contact. Against that background, Strang highlights the difficulty in predicting serious harm from the large number of domestic abuse calls received by the police and the unreliability of risk assessment tools like DASH. Conversely, several studies suggest that suicidal ideation and self-harm by perpetrators are important markers for domestic homicide risk, with perpetrators showing higher prevalence rates compared to the general population. In that perspective, Strang emphasizes the need for greater engagement and cooperation between police and non-police agencies (e.g., health and social services) in identifying high-risk individuals and families, as these agencies may be more likely to become aware of precursors to domestic homicide. She also highlights the role of general practitioners in identifying and intervening in cases of emotional distress associated with domestic violence. In a coda on trends in femicide before, during, and after the COVID-19 pandemic in the United Kingdom, despite concerns about increased risk of harm to domestic abuse victims during the lockdowns, the number of homicides in the United Kingdom did not significantly increase or decrease in 2019 to 2022. The lockdowns appeared to change the type of reporting and the kinds of relationships where abuse occurred, with an increase in reports of abuse by current partners and a decrease in reports of abuse by ex-partners. Consequently, Strang points out that the impact of social upheaval, like the COVID-19 lockdowns, on domestic abuse turns out to be far more complex than usually thought.

The two parts of Strang’s article complete the transition from the theoretical papers to the empirical articles of this special issue. Hence, the following article by Maurice Dunaiski, Camelia Abdelgelil, Andrada-Maria Filip, Natalia Ivanova, David Rausis, Sonya Yee, and Angela Me uses data collected by the United Nations Office on Drugs and Crime (UNODC) to present global and regional estimates of female intimate partner and family-related homicide, as well as the impact of the COVID-19 pandemic on trends in these killings. This overview of the scope and scale of the problem shows that nearly 49,000 women and girls worldwide were killed by their intimate partners or other family members in 2022, amounting to a rate of 1.2 victims per 100,000 women. Africa recorded the largest number and rate of female intimate partner/family-related killings in 2022, surpassing Asia for the first time since 2013. Women and girls are disproportionately affected by domestic homicidal violence, representing an estimated 53% of all victims of killings in the home and, on average, 66% of all victims of intimate partner killings. Monthly data from 14 countries show high variability in trends across countries but suggest that female intimate partner/family-related homicides remained relatively unaffected by lockdowns in the short term. Annual data from Europe and the Americas indicate that trends in female intimate partner/family-related homicide were highly heterogeneous during the first year of the COVID-19 pandemic, with some subregions seeing increases in 2020 compared with 2019. Conversely, the rest of the regions of the world showed a decrease in 2020 compared with 2019. The study highlights the need for improved data collection and analysis on gender-related killings, particularly in Africa, Asia, and Oceania, to provide a more comprehensive picture and inform prevention and response strategies. Following the UN’s approach to this subject, Dunaiski and colleagues emphasize the importance of addressing gender-based killings through initiatives such as criminalization of femicide/feminicide, improved risk assessments and protection measures, and the establishment of femicide observatories. They also point out the limitations of their study, which include the inability to make robust causal inferences about the impact of COVID-19 restrictions, reliance on official data, and limited availability of trend data in certain regions.

Complementing this broader comparative analysis, the seventh article of this special issue narrows its geographical scope. Rok Hacin and Gorazd Meško examine the effects of COVID-19 lockdowns on the number of femicides in Croatia, Hungary, Montenegro, North Macedonia, and Slovenia using monthly data from 2017 to 2021. The total number of femicides decreased in all countries between 2017 and 2021, with exceptions in Croatia and Slovenia, where increases were detected in 2020. However, these increases were not consistent with the general trend. The monthly distribution of femicides showed no common peak months among the observed countries, suggesting no clear seasonal distribution of femicides. Analyzing the effects of restriction measures using the Stringency Index (Hale et al., 2021) corroborated that the number of femicides was not dependent on lockdown measures in Croatia, Hungary, and Slovenia. The authors suggest that the lack of a universal definition of femicide may reduce the relevance of the results due to issues with data comparability. They also consider that routine activity theory (Cohen & Felson, 1979) and GST (Agnew, 2012) do not provide a plausible explanation for the observed trends in femicides during the COVID-19 pandemic in the studied countries. The cultural context of the observed countries, with their socialist past and varying degrees of conservatism and gender equality, is discussed as a potential factor influencing the perception and reporting of domestic violence and femicide. Hacin and Meško point out the limitations of their study, which include the small number of countries, a low number of femicide cases that restrict statistical analysis, and issues with data quality and comparability. They conclude that the COVID-19 pandemic and consequential government restrictions did not significantly affect femicide trends in the observed Central and Southeastern European countries, despite increases in domestic violence reported in some studies. The findings challenge the applicability of certain criminological theories in explaining femicide trends during the pandemic in these countries.

With the eighth article of this special issue, we adjust the focus even further to concentrate on specific countries. In it, Limor Yehuda and Irit Ein-Tal examine IPV and femicide in Israel during 2020, the first year of the COVID-19 pandemic, and compare them with data from 2014 to 2019. The findings, based on police data, show a 7% increase in IPV incidents in 2020 compared with 2019. Furthermore, data from a victim-assistance hotline indicate a sharp increase in IPV cases between March and September 2020 compared with the same period in 2019. However, the analysis of femicide rates shows that the number of femicides in 2020 (29 cases) was not significantly different from pre-COVID years, with some years (2015, 2016, and 2018) having higher femicide counts than 2020. A closer examination of femicide cases on a weekly basis during 2020 revealed that the proportion of femicides was higher during lockdown weeks (57%) compared with nonlockdown weeks (28%). March 2020 had an unusually high number of femicides, with three out of five victims murdered during the lockdown period and two others murdered in the days prior to the first lockdown. The study’s findings partially support media claims about increased violence during the pandemic in Israel. Nevertheless, while IPV increased, femicide rates did not show a significant change compared with previous years. Yehuda and Ein-Tal point out the limitations of their study, which include the limited availability of femicide data in Israel, the lack of monthly or weekly breakdowns in official government data, and the inability to compare the impact of lockdowns on femicide in 2020 with years other than 2019 due to missing data.

In the last article of this special issue, Edoardo Cocco, Clara Rigoni, Federico Bolzani, Yuji Z. Hashimoto, and Stefano Caneppele provide a granular analysis of femicide trends in Italy, revealing the complex dynamics underlying the apparent stability in femicide rates during the pandemic. Through an ARIMA time series analysis, the authors examine changes in femicide trends using two data sets—from the Italian Ministry of the Interior and from media reports by Italian associations—spanning from 2012 to 2022. Their findings reveal that overall femicide rates remained stable during lockdowns, but there was a significant shift in victim–perpetrator relationships, with a decline in femicides by former partners and an increase in those by cohabiting partners from March to May 2020. The analysis partially supports the hypothesis that the types of weapons used in femicides during the lockdown period differed from previous years, particularly regarding the use of improvised weapons and edged weapons. It also finds divergent trends in femicide rates between urban and mixed (heterogeneous) regions in Italy during the pandemic, with a decrease in femicides during the spring lockdown followed by an increase in early summer in mixed regions, while urban regions largely followed the average trend for the decade. Cocco and colleagues suggest that the apparent stability in femicide frequency during the pandemic may result from the dual effects of mobility restrictions, providing protection for women who had terminated relationships with violent partners while intensifying the risk for those cohabiting with violent partners or family members. The study highlights the importance of considering regional differences and the dynamics of victim–perpetrator relationships in understanding the impact of the pandemic and public health measures on violence against women. The limitations of the study include the reliance on police statistics and media reports; the relatively low monthly incidence of femicides; and the potential for inaccuracies in the dates, causes of death, and characteristics of the involved parties in the media-sourced data set. The authors emphasize the need for further investigation using microanalytical approaches to disentangle the interactions leading to femicides and to gain a more comprehensive understanding of the phenomenon.

Discussion

The COVID-19 pandemic’s consequences extended beyond public health, notably affecting IPV and femicide risks. Concerns grew due to lockdowns, economic instability, and disrupted support systems, potentially increasing IPV and femicide by trapping victims and perpetrators together without intervention from guardians. However, most studies in this special issue indicate that femicide rates did not rise during lockdowns or the pandemic’s first year. Yehuda and Ein-Tal (2024) observed a temporary increase in femicides during the initial lockdown month, but the annual figure for Israel (29 in 2020) was too small for conclusive analysis. Similarly, a UNODC regional comparison for 2020 versus 2019 showed femicide rates dropped in four regions but rose in three (Dunaiski et al., 2024). Despite these variations, broader data analysis revealed no pandemic-related increase in femicide, suggesting mobility restrictions had mixed effects. They provided some women protection but increased risks for others, as noted by Treiber (2024) and Cocco et al. (2024), highlighting the complex impact of the pandemic on femicide rates.

Considering the diverse focuses of empirical articles in this special issue, spanning regions to countries, their findings are not contradictory but rather complementary. The overarching message underscores the importance of detailed data for nuanced analysis. The inspiration behind this issue, the work by Aebi et al. (2021), highlights the regret over the lack of detailed data akin to that used by Cocco and colleagues. With more granular data, we might have identified a shift in femicide patterns, from incidents involving former partners to those with current partners, maintaining a consistent overall number, a trend also observed in England and Wales (Abreu Minero & Nivette, 2023). Furthermore, detailed data, similar to those cited by Strang regarding police report analyses, could reveal that IPV increases were partly due to more neighbor reports, as seen in the Netherlands (Coomans et al., 2023). In both scenarios, such insights could bolster an opportunity-based interpretation of IPV and femicide trends during lockdowns, suggesting a reevaluation of the situational hypothesis’s dismissal.

In our view, there are at least five lessons to be learned from this special issue. First, the findings show that there is a complex relationship between the COVID-19 restrictions and IPV/femicide. Several studies (e.g., Cocco et al., 2024; Cunha et al., 2024; Treiber, 2024) underscore the need for nuanced, context-specific analyses that consider regional differences, victim–perpetrator relationships, and situational factors, as well as the dual effects of mobility restrictions on women’s safety. Second, it showcases the essential role of theory in guiding research, incorporating diverse perspectives such as MCT, DLC criminology, SAT, and GST. These frameworks enrich our understanding of the factors contributing to IPV and femicide, underlining the importance of theory-informed research. Third, the issue draws attention to the necessity for enhanced data collection and analysis to improve our grasp of femicide. Several articles (e.g., Dunaiski et al., 2024; Hacin & Meško, 2024; Yehuda & Ein-Tal, 2024) identify the current challenges in data accessibility as well as the importance of investing in robust data collection systems and promoting cross-national collaboration to better track and analyze trends in gender-based violence. Fourth, it identifies the significance of situational and contextual factors in the dynamics of IPV and femicide. It discusses how elements like the relationships between victims and perpetrators, regional variations, and the availability of support services play a crucial role. This highlights the need for prevention and intervention strategies that are tailored to the specific needs of different contexts, especially during crises. Finally, the contributions suggest a variety of strategies for prevention and intervention, including the importance of early childhood interventions, collaborative approaches involving law enforcement and other entities, and the development of plans for future crises. In addition, the consideration of suicide ideation as an indicator of domestic violence risk is mentioned as a practical strategy. This special issue thus provides a foundation for developing policies aimed at reducing femicide and IPV based on empirical evidence and theoretical insights.

The contributions of the authors participating in this special issue extend beyond academic enrichment and can potentially have a real impact on individuals affected by femicide and IPV. Through informing policy decisions, influencing the allocation of resources, and enhancing public awareness, the research shared here plays a critical role in developing more effective strategies for prevention and response, with the ultimate goal of preserving lives and fostering a safer global community. In addition, by integrating a diverse array of perspectives, methodologies, and findings, the articles offer a detailed and nuanced examination of the trends, theoretical underpinnings, and policy considerations related to this urgent social issue. As the world navigates the lingering effects of the pandemic, the insights and recommendations from these studies prove essential for crafting targeted, evidence-based interventions against gender-based violence.