(COR0223)

Written evidence submitted by Prof Tom Kirchmaier and Dr Ria Ivandic, Centre for Economic Performance (CEP) at the London School of Economics and Political Science (LSE) (COR0223)

• 1. The COVID-19 pandemic, and the associated lockdowns, have seen some significant changes in the occurrence and detection of crime across the country.

• 2. CEP’s crime stream has undertaken unique analysis of key crime strands across the past twelve months, using unprecedented access to police data, and with broad focus on policing, domestic abuse, and organised crime.

• 3. Given the Committee’s informed focus, and the Government’s Domestic Abuse Bill, which is currently undergoing Parliamentary scrutiny, this short submission focuses primarily on domestic abuse. We would be keen to discuss our wider work, conclusions and recommendations with the Committee as soon as possible.

• 4. This submission concludes with some notes on our recent research on wide crime trends during and after lockdowns, and suggests where Government should focus policy and analysis to prevent spiralling crime rates in certain areas of the country.

Domestic Abuse

• 5. As Government has worked to make society safer during the pandemic, there have been significant unintended consequences for domestic abuse victims. Our research finds that during lockdown domestic abuse crimes increased on average by 4.5%, while all crimes containing any relationship information remained stable. Abuse by current partners, as well as family members, increased on average by 8.1% and 17.1% respectively, whereas abuse by ex-partners declined by 11.4%.

• 6. Moreover, the increase in domestic abuse calls is driven by third party reporting, which suggests that there is significant under reporting by actual victims. This increase in calls from third parties might point to an increased awareness of noise as neighbours are now at home; increased awareness of domestic abuse since public narratives in the media voiced concern for victims during the lockdown; potential under-reporting by domestic abuse victims; or a combination.

• 7. Our research also shows significant disparity between urban and rural areas, and suggests that domestic abuse is under-reported during lockdown in areas of low-density – such as villages and areas with high proportions of detached homes - where it is much less likely to be overheard, and reported, by neighbours or passers-by. The data suggests that abuse remains present, and undetected, in households where the abuse cannot easily be observed, and therefore reported, by an outsider.

• 8. One over-riding conclusion from this is that there needs to be urgent thought given to how victims can report instances of domestic abuse in lockdown scenarios where it is not possible, or easy, to make a phone call or alert a friend or neighbour.

• 9. Our analysis on five years of crime records and two years of calls-for-service data from London’s Metropolitan Police Service (MPS). This individual-level data allows us to provide a reliable empirical assessment of the changes in volumes of domestic abuse during the lockdown.

Silent Solution and Live Chat

• 10. To alleviate the issue of domestic abuse victims not being able to contact the police, LSE launched a targeted social media campaign with the Met’s Strategic Insights Unit and TVP to promote the Silent Solution, which allows victims to contact the police with minimal verbal communication. In collaboration with Greater Manchester, we launched also launched a social media campaign to promote the use of the novel Live Chat in reporting domestic abuse.

• 11. Our research then looked to determine to what extent technology can help in tackling this under-reporting, and give victims the ability to report abuse in scenarios where it is difficult to make a phone call due to the close proximity of their abuser.

• 12. We looked at how social media could be used to ‘passively’ contact victims – giving them access to information and options to report abuse “silently” that would not easily arouse suspicion from a controlling abuser. 

• 13. Using empowering visuals and accessible information on a website, and designed as a randomized control trial, the experiment used data from Facebook’s location finding to target ads at individuals “living in” the postal sectors assigned to treatment, by advertisements on Facebook and Instagram, in conjunction with three partner forces: GMP, MET, TVP.

• 14. The results showed that the technology works in reaching people: in Greater Manchester 2,200,813 people were reached through the campaign, with 40,537 link clicks made, while in the MET a total of 1,236,367 people were reached and 30,544 link clicks made. And in Thames Valley, a total of 122,311 people were reached and 761 link clicks made.

• 15. While the geographical targeting had too many spillovers to be able to evaluate the difference, we do find that the social media campaign increased the overall use of the Live Chat. The mean number of live chats per day for the intervention period was 5.9, while that for the pre-intervention period was around 3 live chats per day, showing that live chats doubled during the intervention report. We see increases in initiated live chats for both treatment and control groups.

The Silent Solution and Live Chat - Background

• 16. The silent solution system is manned by a call operator, who is part of police staff. While the actual service is provided on the call/live chat, the intention is to gather enough information to assign an urgency level that will determine how quickly the officer will arrive on scene. Most of the resources will be provided to the victim only when the officer arrives at scene and assigns the risk.

• 17. The Live Chat system is a non-verbal way to communicate to the police. It is also manned by a call operator 24/7 that through textual communication aims to gather the same information that they would on a call. It’s a novel reporting tool, using only in a few forces.

• 18. Our research looked at whether the Silent Solution and Live chat could increase reporting, including during lockdown scenarios, and the Live Chat in particular was shown to succeed in that. We would therefore encourage the Home Affairs Committee to consider how such solutions could be more widely rolled-out, including through the intervention of the Home Office and new Domestic Abuse Commissioner.

• 19. There are a number of related questions to consider, including: how to improve knowledge of Silent Solution and Live Chat; how to make them more easily findable online; how their branding could be improved; and how they could be used not just to report, but also to disseminate advice, i.e. what is domestic abuse and what it means for the victim if an abuser is charged, and what support might be available to them.

Wider Data on Crime and COVID

• 20. Further CEP data analysis has looked at where crime levels have risen and fallen during lockdowns, and gives valuable insight into the areas where focus should be directed over the next few months as lockdown is continued, and then lifted.

• 21. A key finding is that the areas of the country most vulnerable to crime – those parts of the country with a high number of claimants pre-pandemic and with above-median increases in claimants since March 2020 – experienced higher antisocial behaviour, bicycle thefts and increased drug offences after the first lockdown. 

• 22. For example, areas with a higher increase in claimants during lockdown were found to have more anti-social behaviour crimes (25% over the mean compared to the national 16%). Areas which entered the pandemic with above-median claimant rates have higher violent crime rates (30% over the mean).

• 23. The rise in cases of bicycle theft is in line with previous research on the economic value of goods and criminal incentives. During lockdown the value of alternative modes of transport increased, and with it the demand for bicycles. While some of the other crime categories serve as a sign of social unrest, the case of bicycle theft is one that highlights clear underlying economic motives.

• 24. While across the country social distancing measures have been reflected in a decrease in various other acquisitive crimes, this decrease is smaller in areas with higher claimant counts. They are characterised by lower levels of educational attainment, a proportionately higher Black and Asian population, a higher proportion of lone parents and worse health. The clear conclusion is that the economic and crime effects of COVID are hitting the most vulnerable areas the hardest.

• 25. Further Government research and analysis should focus urgently on understanding the spatial distribution of crimes and their correlation to economic outcomes: vital as lockdowns continue. With more businesses potentially ceasing operations, the prediction is that crime figures may rise in the coming months. The economic impact is more intense in certain parts of the country. In particular, the pandemic seems to have affected more strongly areas with lower educational attainment. These areas mostly concentrate in London, and the South and East of England.

• 26. Of particular importance for the Committee will be our conclusion, based on the data, that areas already characterised by deprivation are at risk of entering a difficult loop of poverty and crime. On the individual level, the economic disruption caused by COVID has affected more strongly those from poor families, with critical disparities in educational opportunities. Other explanations of the geographical disparity can also be attributed to gender, including evidence of how women are more likely to lose their jobs in a lockdown than men.

• 27. The Government needs to prioritise efforts to understanding these differences and disparities to target policy effectively and prevent crime spirals that could thwart the recovery and the levelling up agenda. We would be happy to discuss how this should happen with the Committee.

February 2021