COVID-19: Outreach to Domestic Abuse Victims in Times of Quarantine
Lead Research Organisation:
London School of Economics and Political Science
Department Name: Centre for Economic Performance
Abstract
Police reports suggest that domestic abuse (DA) has risen as a result of the pandemic, yet there is concern that the share of DA incidents reported may have fallen. When a victim and abuser are quarantined together, calling the police may jeopardize the victim's safety. As a result, greater numbers of victims are increasingly isolated and at risk.
So how can victims get help? How can authorities let them know what options are available? Direct messaging can be dangerous, since texts from the police may provoke a controlling abuser.
We propose a targeted social media campaign to inform potential high-risk victims about the Silent Solution, a safer option for contacting police. Our approach leverages the wide use of social media, which also poses less risk than direct messaging. Whereas text messages are actively sent, social media adverts are passively received.
The study will identify potential high-risk victims and randomly select half for the media campaign. Analysis of DA calls will show whether the approach is effective. If so, it will provide an approach for reaching isolated DA victims, and for giving them options to get help, that will be of value both during quarantine and beyond.
So how can victims get help? How can authorities let them know what options are available? Direct messaging can be dangerous, since texts from the police may provoke a controlling abuser.
We propose a targeted social media campaign to inform potential high-risk victims about the Silent Solution, a safer option for contacting police. Our approach leverages the wide use of social media, which also poses less risk than direct messaging. Whereas text messages are actively sent, social media adverts are passively received.
The study will identify potential high-risk victims and randomly select half for the media campaign. Analysis of DA calls will show whether the approach is effective. If so, it will provide an approach for reaching isolated DA victims, and for giving them options to get help, that will be of value both during quarantine and beyond.
Publications
Grogger J
(2021)
Comparing Conventional and Machine-Learning Approaches to Risk Assessment in Domestic Abuse Cases
in Journal of Empirical Legal Studies
Ivandic R
(2020)
Changing Patterns of Domestic Abuse during COVID-19 Lockdown
in SSRN Electronic Journal
R. Ivandic, T.Kirchmaier
(2020)
Domestic abuse in times of quarantine
| Description | Domestic abuse (DA) is a social problem marked by high levels of repeat victimization and underreporting. Domestic violence reported to the police has risen since the beginning of quarantine, yet there is concern that actual incidents have risen even more. With the victim and abuser being constantly confined together, calling the police might be difficult for the victim, and may pose great risk to the victim's safety. Our research over the past 8 months dealt with two questions. First, as a precursory, understanding how the stay-at-home orders changed the trends in domestic abuse and whether we can learn anything about the nature of reporting in light of the changes imposed. In such a context non-traditional means of reaching isolated DA victims are of prime importance. Hence, secondly, we launched two social media campaigns to promote two alternative channels of reporting - the Silent Solution and the Live Chat. Our first set of results shed an understanding of the impact of the COVID-19 lockdown on domestic abuse in using detailed administrative data from London's Metropolitan Police Service. This data uniquely integrates calls to the police, crime records, and precise geo-location of each call record. We complement this analysis by exploiting the socio-demographic characteristics of the location of the call to check for heterogeneous differences across density and deprivation levels. Our analysis finds that cumulative numbers failed to show considerable change in abuse levels and masked information about altered patterns of abuse by family, current or ex-partner. Looking at crime data, we observe a significant shift in composition of abuse during the COVID-19 lockdown. Abuse by current partners and family members is found to have increased by 8.5% and 16.4% respectively while ex-partner abuse declined by 9.4%. When we examine changes in violent crimes, we find the same pattern with stronger magnitudes ruling out reporting bias. In sum, this analysis shows that reporting the average change in domestic abuse during lockdown can be misleading when thinking about a policy response. If ex-partner and current partner abuse have the same share in total abuse, despite diverging trends, the total average might remain stable. Yet in this situation of seemingly no change, victims quarantined with their perpetrators are experiencing significantly higher levels of abuse. This insight can also explain the mixed evidence of the effects of lockdowns on domestic abuse worldwide. The lockdown might have affected both incidence of DA and the likelihood of reporting, hence we disentangle these two effects using information of the type of caller and the timing. Our analysis finds a substantial change in the nature of reporting. We show that almost all the increase in domestic abuse calls to the police is driven by third party reporting. This can be explained as the lockdown has trapped many victims with their abuser not giving them the opportunity to report abuse, so the primary channel of reporting is through third parties. Testing whether proximity of neighbours and their ability to overhear and report ongoing abuse incidents is driving the increase in reporting, we find that in areas of high density, where households live in closer proximity, third party reporting is 33% higher than in areas of low density. We also show while the majority of additional calls in lockdown are during working hours, there is also a 22% increase in third party calls in the evening hours excluding that our findings are purely a reporting effect of present neighbours. Understanding the diverging consequences of the lockdown on victimization can provide evidence for law enforcement agencies to direct interventions to at-risk households and launch information campaigns on the role of community reporting. As the first set of our findings pointed to, there is a significant share of underreporting during lockdown. To mitigate this, we designed, implemented and started to evaluate two social media campaigns to promote two alternative channels of reporting The first step was to design the social media campaign and social media posts in collaboration with a creative agency. This can be viewed on the website (links:https://www.tobeheard.co.uk and https://www.tobeheard.co.uk/gmplivechat), Facebook page (https://www.facebook.com/tobeheard.co.uk) and Instagram page (https://www.instagram.com/tobeheard.co.uk/). Our ads are succinct and empowering. They will publicize the Silent Solution and Live Chat, and how to use it, without the wordiness and lack of clarity of existing public service messages. Moreover, social media targeting is safer than direct messaging via text or WhatsApp. Domestic abusers often monitor the victim's communication with others. Direct messaging by the police runs the risk of putting the victim in danger. Social media ads, in contrast, are indirect, appearing simply as part of the victim's social media feed, with no obvious attribution. Second, we designed our analysis to be well-targeted. We used empirical analysis to identify areas and victims that are high-risk and then used random-assignment design across areas and victims to define the treatment and control groups. Using Facebook advertising manager, we advertised to the target groups with Instagram and Facebook ads. The MET and TVP campaign "To Be Heard" promoting the Silent Solution reaching a total of 1,236,367 people and generating a total of 30,544 link clicks, peaking at a daily reach of over 80,000 unique individuals and generating over 1,200 daily link clicks. In general, for this campaign we see that link clicks follow the trends set by the people reached. On days when a larger number of people are reached, the number of clicks are also higher. The campaign also generated 1,403,939 impressions in total and had an average link click through rate of about 2.5% on average. We see that click through rate is high when the advertisements first reach people, in the initial stages of the campaign, and gradually dip off however, it rises again towards the end of the campaign suggesting that repeated exposure is causing people to engage with the campaign. We also decided to analyse the results by age group and found that 39.38% of the people reached by the campaign belonged to the 18-24 years age group and 33.92% were from the 25-34 year age group. Around 80% of total link clicks also came from these two groups with the 18-24 year age group responsible for 42.8% of total link clicks and 25-34 year age group accounting for 39.1% of the link clicks. The highest link click through rate was 2.54% seen for the 25-34 year age group. The "To Be Heard" campaign in collaboration with the Greater Manchester Police (GMP) force's domestic abuse Live Chat reporting system. The campaign reached a total of 2,200,813 individuals through the campaign and 2,537,333 impressions were made. It reported a total of 40,537 link clicks through to the advertisements. Analysing engagement with these adverts we find that the simpler image text was more effective. The visuals with the line "Instant Messaging for Instant Help", while different graphics, produced the most engagement. This speaks that the copy message has a greater importance in engaging audiences than visuals. The mean link clicks across time seems similar despite varied spending on the campaign. Further, relative to the amount of people reached and the impressions made on any given day, more clicks were observed towards the latter half of the campaign. This suggests greater engagement with the adverts over time implying that repeated advertisement was effective. Next, we graphed the reach and clicks by day of the week so see if any patterns emerged and found that Most of our link clicks were received on Wednesdays. A total of approximately 7,100 link clicks were received on Wednesdays. The graphs suggest that clicks on the advert were higher towards the middle of the week relative to weekends. We also decided to analyse the results by age group and found that 44.5% of the individuals reached by our campaign belonged to the 18-24 age group. This is also the age group that clicked on the advertisement most (49.10%) which is to be expected as they were the ones who viewed the advert the most. In total, this group had a link click through rate of 1.72%. Surprisingly, the age group with the highest click through rate is not the one that was most reached through our campaign. This is the 25-34 age group. Although the group comprised 28.65% of the reached individuals, it was responsible for 35.19% of total clicks at a click through rate of 1.95%. We suspect that this is also the group that is most likely to be subject to domestic violence in a lockdown setting, explaining their increased engagement with the adverts. In order to verify the accuracy of the Facebook campaign's location targeting, we use data from the responses of the survey on the GMP "To Be Heard" website. The survey data has a total of 162 observations recorded. These include data on the date of the response, gender, age, information on how they reached the website/survey and postcodes. There are a number of missing observations within each category as this is self reported data. On average in the campaign, 40% of total respondents chose not to provide postcode information. To validate if the postcodes the campaign advertised to were in fact the targeted ones, we filtered out the GMP postcodes from the available postcodes and matched the postcode information shared via the survey to our assignment of postcodes into treatment and control. If targeting were effective and the survey representative, we would expect that more respondents would belong to treatment sectors relative to the control ones. However, this was not the case. We found that Facebook geographical targeting does not work as an equal share of treatment sector areas and control sector areas filled in the survey. Our findings have been actively taken forward by a series of policy makers and police officials. Social media has shown to be affordable in reaching new audiences and promoting changes in and new technology. This work has also been taken forward further with the Greater Manchester Police to launch new campaigns on social media such as to tackle disparities in reporting across ethnicities. and We believe that around 3.5 million women received information on the alternative channels of reporting domestic abuse. |
| Exploitation Route | This research provides important foundations for a deeper understanding of Domestic Abuse during the Covid lockdown, and beyond (in the years to come). We will elaborate on this with the final submission. |
| Sectors | Healthcare,Government, Democracy and Justice |
| Description | The Covid-19 pandemic has put domestic abuse at the centre of attention of both media and policymakers. We very much welcome this and have therefore met with a lot of public policy officials that we discussed in the Engagement reporting section. Moreover, the UK Government has declared tackling domestic violence and abuse a priority, and the Domestic Abuse Bill has been passing through Parliament. The Bill has been welcomed by politicians, campaigners, and police forces, and is a strong first step. We believe it could be improved, though, by making explicit the role of better data and technology in tackling abuse, in line with the concerns uncovered through our research and as part of this grant. So, we have worked with LSE's Public Affairs team to use our work to have impact on the legislation, the political debate around domestic abuse, and the Home Office's longer-term thinking. When the draft legislation was first introduced to Parliament, we sent a briefing to MPs detailing our research and where improvements could be made. Then, in November, we held a series of short and highly-productive individual meetings with influential MPs and members of the House of Lords - including the Chair of the All-Party Parliamentary Group on Domestic Violence and a member of the powerful Home Affairs Committee - to present our work and seek their thoughts. We were also invited to present our findings to the Greater London Authority's Police and Crime Committee to discuss how our research can contribute to change in the capital. These meetings were probing and fascinating, offering excellent insight into the way policymakers think and their questions about academic research. We were also able to present our findings in real-time on the nature of changes in lockdown and the social media campaign efforts. Following those productive sessions, the Public Affairs team helped us secure meetings with the Minister for Safeguarding in the Home Office and the Minister for Justice at the Ministry for Justice, as well as the Prime Minister's advisor on Women and Equalities, all of whom were engaged, expert and expressed a keen interest in taking the conversation forward. Their support for our research, and their advice, inspired us to get even more involved in the policymaking process. When the Domestic Abuse Bill arrived in the House of Lords to be debated just after Christmas, we sent briefings to every politician registered to speak, detailing how our research could improve the legislation. Working with members of the House of Lords and the Public Affairs team, we drafted small but important amendments to the Bill to encourage better use of technology and data, and had them accepted for debate. We then organised a private roundtable, chaired by a former Lord Justice of Appeal, where we presented our findings to 20 members of the House of Lords, took questions on our broad research on domestic abuse, and urged them to support our amendments. The meeting was extremely productive, with the politicians showing both expertise and passion, and genuine interest in our current and future work. The process is ongoing, but there is huge value in the new connections we have made through this influencing work and being able to share our ongoing findings as part of this grant, we were able to clarify and direct policy debates with clear evidence. |
| First Year Of Impact | 2020 |
| Sector | Healthcare,Government, Democracy and Justice |
| Impact Types | Societal,Policy & public services |
| Description | Amendments to Domestic Abuse Bill |
| Geographic Reach | National |
| Policy Influence Type | Contribution to a national consultation/review |
| Impact | We suggested three amendments to the Domestic Abuse Bill that were tabled and supported by various Lords and Baronesses. Based on our academic research, they highlight the importance of the usage of data and technology to find effective solutions to domestic abuse response in policing. |
| URL | https://hansard.parliament.uk/Lords/2021-01-27/debates/33525572-1CFA-44DC-BAF2-A644FBB81EB4/a#contri... |
| Description | Influence on politicians - Evidence Week in Parliament |
| Geographic Reach | National |
| Policy Influence Type | Contribution to a national consultation/review |
| Impact | During a week in November, top researchers including this research team (Kirchmaier and Ivandic) shared their expertise with parliamentarians in a live exhibition. This was an opportunity for us to engage MPs in discussions on the evidence around issues in domestic abuse that affect constituents. Through the online platform, MPs, Peers and parliamentary staff were able to hear 3-minute video briefings and jump into meeting rooms with us to instantly discuss emerging evidence and potential strategies for tackling complex policy problems. During this week, we had one to one meetings with: Simon Fell MP (Con, Member of the Home Affairs Committee) Victoria Atkins MP (Con, Minister for Safeguarding) Rebecca Steen (Office of Durham Policing and Crime Commissioner) Elena Narozanski, Helen Cook (No 10 Women & Equalities Advisers) Baroness Grey-Thompson (Crossbench Peer) Baroness Crawley (Lab) Shaun Bailey MP (Con) Apsana Begum MP (Lab) Laura Beaumont / Abigail Plenty (Ministry of Justice) Nickie Aiken MP (Con) Debbie Abrahams MP (Lab) Baroness Barker (Lib Dem) Baroness Shafik (Crossbench) Baroness Sanderson (Con) |
| URL | https://www.youtube.com/watch?v=X1U3Mx0wLMI&t=6s |
| Description | Roundtable on Domestic Abuse for the House of Lords |
| Geographic Reach | National |
| Policy Influence Type | Contribution to a national consultation/review |
| Impact | Very engaging discussions with the following politicians in the House of Lords on issues around domestic abuse and the Covid-19 lockdowns: Baroness Butler-Sloss Baroness Ros Altmann Baroness Hilary Armstrong of Hill Top Baroness Natalie Bennett Lord Clive Brooke of Alverthorpe Baroness Carlyn Chisholm of Owlpen Lord Byron Davies of Gower Baroness Rita Donaghy Baroness Lynne Featherstone Baroness Ilora Finlay of Llandaff Baroness Sally Hamwee Lord Philip Hunt of Kings Heath Baroness Kennedy of The Shaws Lord Daniel Moylan Baroness Helen Newlove Baroness Dawn Primarolo Lord Simon Russell of Liverpool Baroness Liz Sanderson Baroness Manzila Uddin Lord George Young of Cookham Jessica Agboola (Researcher for Lord Bishop of London). |
| Description | Seminar at the science and evidence ecosystem of the Home Office |
| Geographic Reach | National |
| Policy Influence Type | Influenced training of practitioners or researchers |
| Impact | The audience gained a much depper understanding of the issues surrounding Domestic Abuse, and the limitations in our understanding / knowledge gap. This should lead to better commissioning of future work by the Home Office, as well as better policy making. |
| Description | Talk at the 'Universities at the Met' |
| Geographic Reach | Local/Municipal/Regional |
| Policy Influence Type | Influenced training of practitioners or researchers |
| Impact | The participants gained a much deeper understanding of the issues surrounding Domestic Abuse, in particular during the lock-down period. |
| Description | Work featured in an ONS report on Domestic abuse in lockdown |
| Geographic Reach | National |
| Policy Influence Type | Citation in other policy documents |
| Impact | Real time analysis of the changing patterns in domestic abuse allowed police practitioners to direct resources to those in need. Additional evidence of under-reporting highlighted the importance of expanding different available methods of reporting domestic abuse. |
| URL | https://www.ons.gov.uk/peoplepopulationandcommunity/crimeandjustice/articles/domesticabuseduringthec... |
| Description | Funding from Greater Manchester Police |
| Amount | £6,000 (GBP) |
| Organisation | Greater Manchester Police |
| Sector | Public |
| Country | United Kingdom |
| Start | 02/2021 |
| End | 05/2021 |
| Description | Collaboration with UK Police Forces |
| Organisation | Greater Manchester Police |
| Country | United Kingdom |
| Sector | Public |
| PI Contribution | The team has worked in close collaboration with the three aforementioned forces to launch the social media campaigns on the alternative methods of reporting and provide continuous evidence base suggestions emerging from real time analysis of domestic abuse patterns. The policy deliverables we have provided the forces consist of the social media campaign and a series of short reports. The reports summarize our key findings in a non-technical manner and have been delivered to the forces during quarantine, either directly to domestic abuse leads or by attending meetings and summarizing fidnings in person. |
| Collaborator Contribution | They have continuously provided the data in real time and clarifications on the coding and understanding the data. They have also repeatedly been kind to share their time to brainstorm and improve our social media campaign operationalisation, |
| Impact | All the list of output in the Deliverables is a result of these collaborations. |
| Start Year | 2020 |
| Description | Collaboration with UK Police Forces |
| Organisation | Metropolitan Police Service |
| Country | United Kingdom |
| Sector | Public |
| PI Contribution | The team has worked in close collaboration with the three aforementioned forces to launch the social media campaigns on the alternative methods of reporting and provide continuous evidence base suggestions emerging from real time analysis of domestic abuse patterns. The policy deliverables we have provided the forces consist of the social media campaign and a series of short reports. The reports summarize our key findings in a non-technical manner and have been delivered to the forces during quarantine, either directly to domestic abuse leads or by attending meetings and summarizing fidnings in person. |
| Collaborator Contribution | They have continuously provided the data in real time and clarifications on the coding and understanding the data. They have also repeatedly been kind to share their time to brainstorm and improve our social media campaign operationalisation, |
| Impact | All the list of output in the Deliverables is a result of these collaborations. |
| Start Year | 2020 |
| Description | Collaboration with UK Police Forces |
| Organisation | Thames Valley Police |
| Country | United Kingdom |
| Sector | Public |
| PI Contribution | The team has worked in close collaboration with the three aforementioned forces to launch the social media campaigns on the alternative methods of reporting and provide continuous evidence base suggestions emerging from real time analysis of domestic abuse patterns. The policy deliverables we have provided the forces consist of the social media campaign and a series of short reports. The reports summarize our key findings in a non-technical manner and have been delivered to the forces during quarantine, either directly to domestic abuse leads or by attending meetings and summarizing fidnings in person. |
| Collaborator Contribution | They have continuously provided the data in real time and clarifications on the coding and understanding the data. They have also repeatedly been kind to share their time to brainstorm and improve our social media campaign operationalisation, |
| Impact | All the list of output in the Deliverables is a result of these collaborations. |
| Start Year | 2020 |
| Description | Media engagements |
| Form Of Engagement Activity | A press release, press conference or response to a media enquiry/interview |
| Part Of Official Scheme? | No |
| Geographic Reach | National |
| Primary Audience | Public/other audiences |
| Results and Impact | Articles on our findings reported in the Police Professional, The Guardian, The Independent , The Telegraph, and the Manchester Local News. This contributed to highlighting the wider socio-economic costs of the Covid-19 lockdown on domestic abuse to a wider general audience. |
| Year(s) Of Engagement Activity | 2020 |
| Description | Roundtable The Covid-19 Domestic Abuse Crisis |
| Form Of Engagement Activity | Participation in an activity, workshop or similar |
| Part Of Official Scheme? | No |
| Geographic Reach | National |
| Primary Audience | Professional Practitioners |
| Results and Impact | This webinar brings together three ESRC-funded projects investigating the impact of Covid-19 on domestic abuse. The webinar provides an overview of emergent early research findings, and a discussion on the synergies in our work, with a view to collaboratively advance knowledge, policy and practice in this important area. Each research team will provide a brief and accessible overview of their research projects: a) Hohl and Johnson's analysis of the impact of Covid-19 and related social-distancing measures on the domestic abuse reported to the police, as well as police responses; b) Walklate, Goffrey and Richardson's documentation of how the criminal justice system, including the police, are addressing the domestic abuse 'shadow pandemic' during Covid-19; and c) Kirchmaier, Grogger and Ivandic's research addressing police outreach to victim-survivors during lockdown. These presentations will be followed by a panel session with key domestic abuse and police stakeholders as panellists, where the discussants will reflect on our research findings, address the evolving status of the pandemic, and its implications for victim-survivors of domestic abuse, police forces and other key stakeholders. The webinar aims to raise awareness of key emergent findings, identify shared learning, and establish key next steps for tackling domestic abuse in this current, unprecedented climate. |
| Year(s) Of Engagement Activity | 2020 |
| URL | https://www.youtube.com/watch?v=iSr2x5bJcOo&t=2454s |
| Description | Standing up for victims of Domestic Abuse-pilot meeting |
| Form Of Engagement Activity | A formal working group, expert panel or dialogue |
| Part Of Official Scheme? | No |
| Geographic Reach | National |
| Primary Audience | Policymakers/politicians |
| Results and Impact | In conjunction with Marsha de Cordova MP and Shadow Women and Equalities Secretary, the Labour Party's Community Organising Unit is organising a virtual roundtable meeting in Battersea focusing on supporting victims and survivors of Domestic Abuse. The Covid19 crisis has shown an increase in domestic incidents and has highlighted the already obvious need for additional refuge places and victim support more generally. We want to bring together local communities and people working in the sector to look at the current state of play, identify any gaps in local provision as well as identifying solutions that will make services more sustainable than at present. The idea is that this event will create a local plan that will contribute to national change. |
| Year(s) Of Engagement Activity | 2020 |