Recommended
CGD NOTES
This year, the 16 Days of Activism against Gender-Based Violence comes at a time when there is unprecedented global attention on violence against women and children (VAW/C). The theme of “Fund, Respond, Prevent and Collect!” puts a focus on the need for increased financing to support survivors and prevent future violence, as well as the importance of data collection, rigorous analysis and research to guide effective financing and policy decision making.
Seeking to prevent a “return to normal,” since the start of the COVID-19 crisis, we have sought to provide decision-makers with a clearer understanding of the pathways through which violence is likely to increase and potential policy responses to prioritize. As COVID-specific evidence on violence dynamics emerged, we began to synthesize this evidence base through periodic research round ups and an open access evidence tracker of global studies.
In June, we summarized 17 rigorous research studies that had been published since the start of the pandemic, and in September we reviewed an additional 28 studies. With a growing number of diverse research questions and innovative data sources in this third round up, we add an additional 29 studies. Consistent with the approach taken in our previous round ups, we focus on reports, working papers and publications across disciplines and methodologies that move beyond simple month-to-month comparisons from single sources. While we strive to be inclusive, we seek to balance this with rigor, so we do not include papers that do not present a full methodology, samples or VAW/C indicators behind their analysis.
What have we learned? Key Takeaways for this round up
-
Approximately half (14) of the studies we review focus on whether or not VAW/C has increased during the COVID-19 pandemic, with half of these (7) supporting increases. Where studies observe mixed findings or decreases, underreporting, especially by children, may account for results – a cause for concern as victims are less able to report violence and seek help under lockdown conditions.
-
A total of 15 studies explore risk factors or other dynamics around VAW/C in the COVID context, highlighting economic stressors, low social support, lack of employment, substance use, poor mental health, and younger age as salient risk factors associated with intimate partner violence (IPV) and parenting stress, job loss, and lack of support and perceived control as salient risk factors associated with violence against children, in diverse settings. This evidence is consistent with how VAW/C has increased in past pandemic and crisis contexts, although to date, studies primarily rely on small and non-representative samples.
-
A clearer understanding of specific risk factors can facilitate policymakers’ evidence-based decision making to prevent violence and support survivors. However, to date, only a handful of studies analyze “what works” to prevent and respond to VAW/C during COVID-19—rather than simply examining correlates or trends.
- This round up featured a broader geographic spread of studies, including 16 studies from Africa and Asia—as well as a substantive focus on violence against children (13 studies). Studies in the round up also examine broader intersecting vulnerabilities, including dynamics within immigrant communities and for refugee women, as well as conflict-affected settings.
An overview of new studies: Research questions, geographies, and methods
A substantial portion of new research continues to focus on the basic question: Has VAW/C increased or decreased during the pandemic? We highlight 14 new papers—half (7) of which find increases in indicators of VAW/C during COVID-19 or due to associated response measures. The remaining find mixed impacts (4), or decreases (3), the latter all examining reported child maltreatment measures in the United States (and likely reflect under-reporting, as children have less access to education or other service personnel able to observe signs of violence). This accumulated evidence points more strongly to increases in VAW/C measures as compared to earlier studies, including three studies of trauma from hospital clinical records (in the United States, Romania and South Africa).
Some caveats are warranted, as several studies show mixed results depending on the measures or time period. For example, a study examining domestic violence in Los Angeles, United States finds increases in service calls (to police and hotlines), but no changes in arrests and decreases in crimes which vary over time. Another study of domestic violence crimes in London finds overall null impacts, masking increases in current partner violence and violence perpetrated by other family members, and decreases in ex-partner violence. These examples highlight the variability in results across studies due to data sources, time spans and VAW/C measures, all of which require careful contextualization, including the role of under-reporting within VAW/C indicators.
Similar to previous round ups, the majority of studies still focus on the United States and other high-income countries. However, several studies draw on cross-country data, including from low- and middle-income settings, as well as from China, Romania and South Africa. While geographic diversity is limited, methodologies and data sources continue to expand to give additional insights. For example, the aforementioned cross-country study uses machine learning to identify and analyze hateful and abusive content and cyberbullying on Twitter and Reddit, and analysis from Los Angeles examining child abuse and neglect uses spatial analysis (GIS) to examine the intersection of disease and social vulnerability. Studies still largely rely on either publicly available or service provision administrative data (with two exceptions)—therefore studies identify reported incidence or more severe cases of VAW/C, rather than all incidents (including those which are unreported).
Group A. Papers that measure impacts of COVID-19 or associated response measures on VAW/C
Authors | Location | Data | Methods | Indicator(s) | Finding | |
---|---|---|---|---|---|---|
1. | Babvey et al. 2020 | Cross-country | Twitter (16 countries); Reddit (United States) | Machine Learning algorithm; Temporal analysis | Abusive and hateful language; Cyberbullying | Increase |
Family violence | Increase | |||||
2. | Kovler et al. 2020 | Maryland, United States | Clinical chart review at John Hopkins Children's Center | Annual comparison of cases | Physical child abuse | Increase | 3. | Sanga & McCrary 2020 | United States | Police 911 calls, paired with mobility data | Regression with day-hour fixed effects | Domestic violence | Increase | 4. | Socea et al. 2020 | Bucharest, Romania | Clinical chart review at Surgery Departments | Annual comparison of cases | Domestic violence | Increase |
5. | Takaku & Yokoyama, 2020 | Japan | Survey data (online) | Regression Discontinuity Design | Domestic violence (any, March) | No change |
Domestic violence (any, August) | No change | |||||
Domestic violence (frequently, March) | Increase | |||||
Domestic violence (frequently, August) | No change | |||||
6. | Qin et al. 2020 | Guangdong Province, China | Women's Federation (government) | Hierarchical linear regression | Domestic violence | Increase |
Australia, Canada, United Kingdom, United States | Google search trends data | Hierarchical linear regression | Domestic violence hotline (searches) | Increase | ||
7. | Zsilavecz et al. 2020 | Pietermaritzburg, South Africa | Clinical chart review of trauma patients at Grey's Hospital | Annual comparison of cases | Penetrating trauma and blunt assault (total) | No change |
Penetrating trauma and blunt assault (female proportion) | Increase | |||||
8. | Ivandic et al. 2020 | London, United Kingdom | Police service data (crimes) | Event study analysis | Domestic abuse (partner) | Increase |
Domestic abuse (other family member) | Increase | |||||
Domestic abuse (all relationships) | No change | |||||
Domestic abuse (ex-partner) | Decrease | |||||
9. | McKay et al. 2020 | United States | Gun Violence Archive (firearm incidents) | Poisson pseudo maximum likelihood models with state-time fixed effects | Child-involved shootings in the home | Increase |
Domestic violence shootings | No change | |||||
School shootings | Decrease | |||||
Workplace shootings | Decrease | |||||
Total firearm injuries and deaths | Initial decreases, subsequent increases | |||||
10. | Miller et al. 2020 | Los Angeles, United States | Police 911 calls, crime incidents, arrests and domestic violence hotline calls | Difference-in-difference | Domestic violence (police 911 calls) | Increase |
Domestic violence (hotline calls) | Increase | |||||
Domestic violence (arrests) | No change | |||||
Domestic violence (crimes) | Decrease | |||||
11. | Petrowski et al. 2020 | Cross country | Survey of 48 child helplines; media reports | Before-and-after comparison | Contacts to helplines (overall) | Increase |
Contacts to helplines (violence-related) | Mixed | |||||
Media reporting on changes in violence | Mixed | |||||
12. | Barboza et al. 2020 | Los Angeles, United States | Police reported crime data, paired with mobility data | Negative binomial regression; Spatiotemporal analysis | Child abuse and neglect | Decrease |
13. | Bullinger et al. 2020 | Georgia, United States | County-level referrals to the Division of Family and Child Services | Fixed effects regression | Child abuse or neglect referrals | Decrease |
14. | Rapoport et al. 2020 | New York City, United States | Administration for Children's Services data | SARIMA | Child maltreatment allegations | Decrease |
Investigations warranting preventative services | Decrease |
Table notes: Studies are ordered by direction of impact and alphabetical by first author. Studies routinely examine more than one outcome, including disaggregated outcomes or heterogenous effects (not fully reported in Group A table), however studies are classified according to the main outcome(s). SARIMA = seasonal auto regressive integrated moving average; IPV = intimate partner violence
Papers that measure impacts of COVID-19 or associated measures on VAW/C
Babvey et al. 2020
Using data scraped from Twitter (in 16 countries) and Reddit forums (in the United States) and machine learning to classify content, Babvey and colleagues document increasing trends of hateful and abusive language and cyberbullying (Twitter) and family violence (Reddit) during the lock-down period [Babvey et al. 2020; Child Abuse & Neglect].
Kovler et al. 2020
Reviewing clinical cases of physical child abuse at the John Hopkins Children’s Center, Kovler and colleagues find the proportion of cases during the COVID-19 period (March and April, 2020) were 13 percent of traumatic injuries, as compared to 4 percent and 3 percent in previous years [Kovler et al. 2020; Child Abuse & Neglect].
Sanga & McCrary, 2020
Using police 911 calls from 14 cities in the United States, paired with mobile device data and day-hour fixed effects regressions, Sanga and McCrary show that domestic violence calls increased 12 percent on average and 20 percent during working hours—with likely first-time abuse calls increasing by 16 percent and 23 percent, respectively [Sanga & McCrary, 2020; SSRN Working Paper].
Socea et al. 2020
Reviewing clinical cases of domestic violence among polytraumas in surgical departments in Bucharest, Romania, Socea and colleagues document an average increase of 4.3 times the number of cases during the March to June, 2020 period as compared to the previous three years [Socea et al. 2020; Romanian Journal of Emergency Surgery].
Takaku & Yokoyama, 2020
Using survey data collected online in Japan among mothers whose firstborn child was aged four to ten years old, and a regression discontinuity design, Takaku and Yokoyama show that school closures result in greater frequency of domestic violence (driven by physical violence initiated by the wife) in March 2020, but no change in experiencing any domestic violence; These impacts had dissipated by August 2020 [Takaku & Yokoyama, 2020; SSRN Working Paper].
Qin et al. 2020
Using a combination of data from the government of China’s Guangdong province (representing help-seeking) and Google search data for “domestic violence hotline” in Australia, Canada, the United Kingdom, and the United States, Qin and colleagues run hierarchical linear regressions and find significant relationships between COVID-19 cases and domestic violence measures [Qin et al. 2020; PsyArXiv preprints].
Zsilavecz et al. 2020
Examining cases of trauma presenting at Grey’s Hospital in Pietermaritzburg, South Africa, Zsilavecz and colleagues find the total number of assault victims does not change significantly between March and May 2020, as compared to the median in the previous five years. However, the proportion of female victims increases—driven by blunt assault and penetrating trauma, rather than gunshot wounds or road traffic collisions [Zsilavecz et al. 2020; South African Medical Journal].
Ivandic et al. 2020
Using London Metropolitan Police service data on domestic abuse and an event study estimator, Ivandic and colleagues show that abuse by current partners and family members increased by 8.1 percent and 17.1 percent, respectively, while ex-partner abuse decreased by 11.4 percent—leading to (overall) no change in domestic abuse. The increase in abuse calls comes from third parties, particularly in areas of high density, implying victims may be trapped with perpetrators and not able to report [Ivandic et al. 2020; SSRN Working Paper].
McKay et al. 2020
Using data on firearm incidents from the United States Gun Violence Archive, McKay and colleagues find an overall decrease in firearm injuries and deaths per day (by 30 percent) following emergency declarations and stay at home orders – followed by subsequent increases above predictions as stay at home orders are lifted. While workplace and school shootings decreased during stay at home orders, there were increases in child-involved shootings in the home, though no significant changes in domestic violence [McKay et al. 2020; SSRN Working Paper].
Miller et al. 2020
Using police sources (911 calls for services, crime incidents, arrests) and domestic violence hotline calls in Los Angeles from January to August 2020 and difference-in-difference modeling, Miller and colleagues find mixed results depending on the time period and the type of data examined: domestic violence calls for services and calls to hotlines increase, but there was no significant change for arrests, and crimes decreased during the initial shutdown. These decreasing trends continue for crimes, as well as arrests, but increases in calls to police and hotlines reverse trend (i.e., also decrease) in post-shutdown periods as restrictions are lifted (May, 2020 onwards) [Miller et al. 2020; NBER Working Paper].
Petrowski et al. 2020
Using data collected from child helplines in 45 countries, Petrowski and colleagues find that overall contacts have increased during the second quarter of 2020 (as compared to the first quarter), with violence-related contacts decreasing in some countries and increasing in others. A similar mixed picture emerges with respect to media reporting on hotline and helpline changes [Petrowski et al. 2020; Child Abuse & Neglect].
Barboza et al. 2020
Using Los Angeles police reported crime data on child abuse and neglect paired with mobility data, analyzed with negative binomial regression, as well as spatiotemporal analysis, Barboza and colleagues show decreases in child abuse and neglect incidence. However, neighborhood factors such as severe housing burden, poverty, school absenteeism and parental unemployment were associated with new and intensifying abuse hotspots during COVID-19 [Barboza et al. 2020; Child Abuse & Neglect].
Bullinger et al. 2020
Using data on child abuse and neglect allegations from January to May 2020 reported to Georgia’s Division of Family and Children Services, Bullinger and colleagues use a fixed effects regression to find emergency declarations resulted in a decrease of approximately 55 percent in total allegations—with larger decreases in emotional, educational and medical neglect, as well as physical and sexual abuse. However, complementary analysis finds time at home is associated with higher allegations, driven by more alleged supervisory neglect cases in metropolitan counties with historically lower allegations and those with higher poverty [Bullinger et al. 2020; SSRN Working Paper].
Rapoport et al. 2020
Using New York City Administration for Children’s Services data and SARIMA modelling, Rapoport and colleagues find fewer allegations of child maltreatment in March, April and May 2020 than expected for all subgroups of reporters (ie. all reporters, nonmandated and mandated reporters) and fewer child protection service investigations warranting preventative services than expected in March 2020 [Rapoport et al. 2020; Child Abuse & Neglect].
Papers exploring the experience of and risk factors for VAW/C during COVID-19
For the first time, the majority of papers (15 in total) explore the experience of VAW/C during COVID-19 with diverse research objectives. These studies all collect primary data from individuals, with two exceptions (analyzing newspaper content). While allowing a rich analysis describing victims’ experiences and the risk factors leading to violence, these studies have the limitation of often relying on small samples, and often not representative (e.g. online snowball sampling, convenience samples). The majority of studies (13) examined violence levels, and/or risk or mediating factors associated with VAW/C—largely supporting the quantitative evidence showing increases in severity and complexity of violence.
Studies find that economic stressors, low social support, lack of employment, substance use, poor mental health and younger age were salient risk factors associated with IPV and that parenting stress, job loss, and lack of support and perceived control were salient risk factors associated with violence against children, in diverse settings. Two additional studies analyzed media content from news articles in the United States and Nigeria, finding that articles both predicted and reported increases in violence in the United States, whereas the coverage in Nigeria was assessed to be low in comparison to the importance of the issue. Analysis of survey data typically utilized simple correlations or cross-sectional regression analysis, with four studies analyzing qualitative data depicting narratives of women or youth. The geographic diversity of studies is much higher in this set of papers, with multiple studies from Africa (Cross country, Egypt, Ethiopia, Nigeria, Tunisia, Uganda) and Asia (Indonesia, Nepal, Singapore).
Group B. VAW/C experiences and risk factors during COVID-19
Authors | Location | Data | Sample size | Key findings | |
---|---|---|---|---|---|
1. | Brown et al. 2020 | United States | Survey data (phone) | 183 parents | Higher parental support and perceived control during the pandemic are associated with lower stress and child abuse potential. |
2. | Chung et al. 2020 | Singapore | Survey data (online) | 258 parents | Parenting stress mediates the relationship between perceived impact of COVID-19 and harsh parenting. |
3. | Ekweonu 2020 | Nigeria | Newspaper articles | 300 publications | Reviewing 100 days (from March to June) of three daily newspapers, domestic violence is mentioned in 115 news items with dominant content mention of NGO efforts to mitigate violence. |
4. | Fawole et al. 2020 | Nigeria | Case reports from IPV services | 7 survivors | Women described increased severity or new types of IPV during lockdown--for example, the threat of being thrown out of homes--thereby increasing exposure to COVID-19. Linkages were made with economic stressors originating from lockdowns and reduced social support was seen as a barrier to accessing help. |
5. | Forbes Bright et al. 2020 | United States | News articles | 300 articles | Articles predict and report increasing IPV and provide service access information for survivors. |
6. | Gebrewahd et al. 2020 | Aksum, Ethiopia | Survey data (in person)* | 682 women | The prevalence of IPV was 24.6 percent during the lockdown. Risk factors included being a homemaker (versus employed), aged less than 30 years, having an arranged marriage, and partner age between 31 to 40 years. |
7. | Ghimire et al. 2020 | Nepal | Survey data (online) | 556 adults (male and female) | Approximately 18 percent of participants reporting being a victim of interpersonal violence during the lockdown--with an identical percentage reporting perpetration. Victims and perpetrators were more likely to use substances and have lower mental health scores. |
8. | International Rescue Committee, 2020 | Africa (15 countries) | Survey data, paired with qualitative interviews (phone, in-person) | 852 refugees or displaced women; 25 GBV experts | Reported rates of violence against women and girls are increasing, attributed to stress of lockdowns and economic insecurity--as well as threats from security personnel who enforce measures. Additional funding and programming for response is needed in humanitarian settings. |
9. | Jatmiko et al. 2020 | Indonesia | Qualitative interviews (online; phone) | 4 female youth (aged 18 to 23 years) | Technological advances, paired with the high use of social media during COVID-19, provide fertile ground for online sexual violence and harassment from a variety of perpetrators. |
10. | Lawson et al. 2020 | United States | Survey data (online) | 342 parents of 4 to 10 year olds | Technological advances, paired with the high use of social media during COVID-19, provide fertile ground for online sexual violence and harassment from a variety of perpetrators. |
11. | Oguntayo et al. 2020 | Lagos, Nigeria | Survey data (online) | 356 adults (male and female) | Both psychological and socio-contextual factors are correlated with IPV--for example, neuroticism is linked to higher abuse, as are poor living conditions and less stable employment. |
12. | Parkes et al. 2020 | Uganda | Qualitative interviews (phone) | 34 youth (primarily aged 16 to 19 years) | Youth reported multiple stressors during lockdown, resulting in strain of family relationships, including outbreaks of violence and abuse. Community violence, including beatings and misconduct of police and authorities while enforcing lockdowns, was also reported. |
13. | Sabri et al. 2020 | United States | Qualitative interviews (online; phone) | 45 female immigrant survivors; 17 service providers | Immigrant survivors of IPV reported increases in severity and frequency of IPV due to factors such as partners spending more time at home, gun purchases and decreased legal help seeking. Possible mitigation responses include innovative solutions of IPV education and counseling, outreach and service strengthening (safe shelters, virtual services). |
14. | Sediri et al. 2020 | Tunisia | Survey data (online) | 751 women | Women with a history of mental illness and reported abuse during lockdown were found to have more severe symptoms of depression, anxiety and stress. |
15. | Shokair & Hamza, 2020 | Tanta, Egypt | Survey data (in person)* | 160 child victims of family violence | Children in primary school with high rates of reported violence (compared with average or low rates) had increasing mental health problems. |
Table notes: GBV = gender-based violence; IPV = intimate partner violence
* Mode of data collection (whether in person face-to-face or virtual via online platform or phone) was not explicitly mentioned, therefore it is assumed data collection was in person.
Papers exploring experiences of VAW/C during COVID-19
Brown et al. 2020
Using data collected via phone from 183 parents recruited from child- and family-servicing agencies and educational settings in the western United States from April to May 2020, Brown and colleagues find that higher parental support and perceived control are associated with lower stress and child abuse potential (as measured by the Child Abuse Potential Inventory) [Brown et al. 2020; Child Abuse & Neglect].
Chung et al. 2020
Using survey data collected online from 258 parents in Singapore from April to May, 2020, Chung and colleagues find that parental stress mediates the relationship between the perceived impact of COVID-19 and harsh parenting, as well as parent-child closeness [Chung et al. 2020; Journal of Family Violence].
Ekweonu 2020
Reviewing 300 daily news articles (from three newspapers over 100 days from April to June 2020), Ekweonu finds 115 mentions of domestic violence—mostly without additional follow up (80%) and containing mention of intervention by non-governmental organizations [Ekweonu 2020; Journal of Communication and Media Studies].
Fawole et al. 2020
Drawing on 7 case reports from organizations serving women experiencing IPV in Nigeria, as well as media reports, Fawole and colleagues find narratives of increasing severity as well as new types of IPV during lockdown. For example, the threat of being thrown out of homes, thereby increasing exposure to COVID-19, was raised as a form of IPV. Linkages were made with economic stressors originating from lockdowns, and reduced social support was seen as a barrier to accessing informal and formal help [Fawole et al. 2020; Pre-print submission at BMC Women's Health].
Forbes Bright et al. 2020
Using news articles from the first six weeks of the lockdown in the United States, Forbes Bright and colleagues conduct content analysis to show media predications and reports of increasing domestic violence, as well as a focus on information linked to support services for survivors [Forbes Bright et al. 2020; Social Sciences & Humanities Open].'
Gebrewahd et al. 2020
Using data collected from 682 women in Aksum, northern Ethiopia from April to May 2020, Gebrewahd and colleagues find IPV prevalence of 24.6 percent during the lockdown. Factors correlated with IPV include being a homemaker (versus employed), being under 30 years old, having an arranged marriage, and having a partner between the ages of 31 and 40 [Gebrewahd et al. 2020; Reproductive Health].
Ghimire et al. 2020
Using data collected via online survey in Nepal among 556 adults, Ghimire and colleagues describe dynamics from April to July 2020, finding that approximately 18 percent of the sample reports both victimization and perpetration during the lockdown (with higher verbal abuse as compared to physical) from a variety of perpetrators. Higher proportions of victims and perpetrators reported substance use and had lower mental health (though no significance tests were given) [Ghimire et al. 2020; Journal of Nepal Medical Association].''
IRC, 2020
Combining survey data from 852 female refugees or displaced persons and key informant interviews with gender-based violence experts across Africa in 15 humanitarian response settings, the International Rescue Committee documents reported increases in violence against women and girls. Stressors related to lockdowns and economic insecurity, as well as threats from security personnel who enforce measures, are key factors driving increases. Additional funding and programming for response is needed in humanitarian settings [IRC, 2020; Technical Report].
Jamiko et al. 2020
Combining qualitative narratives with online social media content analysis spanning from January to June 2020, Jatmiko and colleagues find that women in Indonesia are at increasing risk of sexual violence and harassment online from a variety of perpetrators [Jamiko et al. 2020; Journal of Society and Media].
Lawson et al. 2020
Using data from an online survey in the United States of 342 parents of children age 4 to 10 from April and May 2020, Lawson and colleagues find parental job loss, depression and previous psychological maltreatment predict maltreatment during the pandemic [Lawson et al. 2020; Child Abuse & Neglect].
Oguntayo et al. 2020
Using data collected via online survey in Nigeria among 356 adults in Lagos, Nigeria, Oguntayo and colleagues show correlations between IPV and both personality traits (neuroticism), as well as socio-contextual factors (poorer living conditions and less stable employment), though no differences by gender [Oguntayo et al. 2020; International Journal of Behavioral Sciences].
Parkes et al. 2020
Using primary data collected via mobile phones among youth primarily aged 16 to 19 years in Uganda, Parkes and colleagues document multiple stressors facing households under lockdown, resulting in strain of family relationships—including outbreaks of violence and abuse. Within these contexts, family could be a source of psychological burden (particularly with youth out of school), but also a source of both social and economic support. Youth also reported witnessing community violence, including beatings and misconduct by police and authorities while enforcing lockdown measures [Parkes et al. 2020; Technical Report].
Sabri et al. 2020
Analyzing qualitative data from 45 female adult immigrants and 17 IPV service providers conducted online or via telephone, Sabri and colleagues document narratives of increased severity and frequency of IPV due to factors including more time with partners at home, gun purchases, and decreased legal help-seeking. Possible mitigation responses mentioned include innovative solutions to IPV education, counseling and outreach and strengthening services, particularly those via virtual platforms [Sabri et al. 2020; Health Care for Women International].
Sediri et al. 2020
Using data from 751 women collected via a women only Facebook group and snowball sampling during April to May, 2020, Sediri and colleagues find women with a history of mental illness and reported abuse during lockdown reported more severe symptoms of depression, anxiety and stress. Violence during the lockdown also reportedly increased (from 4.4 percent to 14.8 percent), though analysis is descriptive only [Sediri et al. 2020; Archives of Women’s Mental Health].
Shokair & Hamza 2020
What can we learn about mitigation and prevention from this evidence?
While the evidence base continues to grow, in line with the 16 Days of Activism theme, we emphasize a call to shift to more action-oriented studies – those that go beyond identifying trends in VAW/C rates and begin to pinpoint “what works” to effectively prevent and/or respond to violence. To date, we have reviewed 74 studies on VAW/C in the COVID-19 context the majority of which focused on whether VAW/C has increased or decreased during the pandemic. Most of these studies point to either an increase in VAW/C (45 percent) or mixed findings (25 percent), but there are still clear data constraints overall, including those which explain, at least in part, where other researchers find no changes or decreased violence.
In the most recent work, there is also evidence of fluctuations across time, with violence prevalence shifting during initial and post-lockdown periods. While interesting and important for advocacy, it is time to transition away from this question and towards research that is aimed at informing policy efforts to prevent VAW/C and mitigate its impact on survivors. This need is underscored by the fact that – whether or not VAW/C increased in the COVID context – it was already a widespread human rights violation and development issue that warranted additional investment and policy attention. The emergence of studies analyzing risk factors for VAW/C is a step in the right direction: the pinpointing of these risk factors opens up the opportunity to identify interventions with potential to directly address them. Studies examining correlational risk factors must now be complemented by those aimed to unpack causal effects of particular interventions or policies. Across all three round ups, few studies attempt to look at effectiveness of interventions, including one evaluating a media campaign to encourage reporting in Italy, the contribution of firearm policies in the United States, and the effects of alcohol prohibition in Mexico. Qualitative studies, including those documenting lived experiences of vulnerable groups, such as immigrant women, can also aid in prioritizing and developing effective programming. As emphasized in the previous round up, primary studies must be implemented with, and report on, ethical safeguards for collection of remote data. Recent guidance on data collection and evidence generation on violence against women and violence against children during COVID-19 can aid in these efforts.
Encouragingly, new studies have now enabled a more intersectional understanding of how VAW/C is impacting particular vulnerable groups, including those in low- and middle-income countries. Not only has the number of studies focused on children (including adolescents) increased (13 studies in this round up), but recent research also examines VAW/C among refugee populations, as well as those living in conflict-affected areas. Specific pathways through which VAW/C may increase, such as violence perpetrated by security personnel in humanitarian settings or various forms of cyberviolence, are now documented within the COVID context, expanding our collective understanding of how violence may manifest outside the home in addition to within it.
Actionable research will play an essential role in informing policymakers’ decision-making, ensuring that it is evidence-based and therefore likely to move the needle in reducing VAW/C. We support the 16 Days of Activism calls for increased financing, as well as actionable, ethical research for VAW/C prevention—during COVID-19 and beyond.
Rights & Permissions
You may use and disseminate CGD’s publications under these conditions.