Research and analysis

Safety of Women at Night Fund Evaluation

Published 10 April 2024

Applies to England and Wales

Executive summary

Safety of Women at Night Fund

In July 2021, the UK government launched the £5 million Safety of Women at Night (SWaN) Fund. This provided funding for interventions that aimed to improve feelings of safety and prevent violence against women and girls (VAWG) in public spaces at night. Civil society organisations, local authorities, Police and Crime Commissioners (PCCs), and the Chief Constable of the British Transport Police could bid for funding up to £300,000. The Home Office received 82 eligible bids across England and Wales, of which 22 were awarded SWaN funding. Each successful bid delivered multiple interventions targeted at two core beneficiary groups, namely users of public spaces at night and key worker groups in the night-time economy (NTE). Interventions were delivered from November 2021 to the end of March 2022.

Evaluation overview

The Home Office commissioned Verian (formerly known as Kantar Public UK) and Ecorys UK to carry out a process and impact evaluation of the SWaN Fund, employing the following aims and approaches:

1. A Fund-level impact evaluation to determine the net impact of SWaN interventions on target populations across the funded areas. Impact was measured by collecting and analysing survey data on key outcomes, including perceptions of safety, awareness of interventions, and intervention-specific outcomes from target populations in intervention and counterfactual areas. A face-to-face (F2F) survey was conducted of women using public spaces at night (n=1168), and an email survey of NTE workers (n=465). The survey data were supplemented by applying the Qualitative Impact Protocol (QuIP) methodology. This involved conducting in-depth interviews with a sample (n=23) of F2F survey respondents and eliciting open-text responses to the email survey to understand the causal factors attributed to changes in perceptions of safety.
2. A process evaluation to understand the key facilitators and challenges for bid teams at each stage of the SWaN programme, and to identify good practice emerging from SWaN. The process evaluation employed a range of research methods, specifically a document review, proforma survey with bid teams, in-depth interviews, and case studies with 8 funded areas. Across the in-depth interviews and case study research, 53 individuals participated in interviews or focus groups.

Data collection took place from February to May 2022. A 3-week pause in fieldwork occurred from mid-April to early May due to restrictions imposed on government activities prior to local elections.

Impact evaluation findings

Overall, the evaluation found no statistically significant evidence that SWaN interventions improved feelings of safety in public spaces at night. However, there was a range of factors that likely influenced the evaluation’s ability to achieve and detect statistically significant impact. Qualitative findings from the evaluation suggest that achieving an impact on feelings of safety was not attainable within the constraints of the Fund and the evaluation. For example, QuIP analysis highlighted that many complex factors influenced perceptions of safety, including prior experiences of crime and awareness of local or national crime incidents. Furthermore, the short funding period for SWaN interventions (3 to 4 months) meant the time for potential impacts to be realised and measured was limited. Limitations of the evaluation design also reduced the ability to detect potential impact, such as the lack of opportunity to conduct a baseline survey before intervention delivery began.

Although an aim of the Fund was to reduce VAWG, the evaluation did not measure the impact of SWaN interventions on VAWG crimes. Police recorded crime (PRC) data cannot be used as an accurate measure of prevalence for VAWG due to the high level of under-reporting. It is also difficult to explain any changes in the number of police recorded offences, which could be the result of improvements in police recording practices. For these reasons, measuring the impact on VAWG crimes was considered out of scope of this evaluation.

While a significant impact was not detected across users of public spaces at Fund level, the QuIP analysis provided some positive evidence at an individual level. Some respondents identified a positive change in their perceptions of safety at night in some public spaces, and in some instances, the change was directly attributed to the types of intervention funded by SWaN. For example, increased security measures at NTE venues. This suggests that these interventions can be viable mechanisms for improving feelings of safety.

In relation to impact on NTE workers, the evaluation found evidence of some positive effects resulting from SWaN interventions. Analysis found statistically significant evidence that SWaN interventions were associated with improvements in their awareness and understanding of VAWG, confidence in supporting VAWG victims, and capabilities in intervening in VAWG incidents.

Process evaluation findings

In relation to the design and development of SWaN funding bids, stakeholders reported that SWaN provided an opportunity to fill gaps in existing provision to address VAWG, and that the eligibility criteria allowed new organisations to bid and to trial new interventions. However, narrow timescales for the commissioning of the Fund led to challenges in bid development, relating mainly to partnership development and obtaining relevant data to support bids.

There was a wide range of activities conducted across the funded areas. Generally, key learnings from delivery varied according to the type of intervention being delivered; however, some overarching lessons for future delivery were found. For example, effective governance and partnership working was vital for successful delivery in tight timescales, and multi-agency approaches to supporting women in the NTE were important. Funded areas also widely acknowledged the importance of focusing interventions on perpetrator behaviours and attitudes.

Stakeholders believed that users of the NTE were benefiting from SWaN interventions, and that NTE workers were better equipped to recognise and intervene in situations of sexual harassment or VAWG. However, overall, it was felt that more time will be required for longer-term effects to become visible.

Overall recommendations for future funds and evaluations

Implications for future funds and evaluations emerged from the evaluation, some of which are consistent with those arising from the Safer Streets Fund round 1 (Home Office, 2023). Recommendations include:

Future funds

1. Give greater consideration to timescales at each stage of the project. An extended bidding window would help areas to engage key stakeholders effectively and improve bid quality. Although achieving impact will be difficult, increasing the delivery period would increase the likelihood of achieving and detecting significant impacts.
2. Focus on increasing the scale and sustainability of interventions to ensure they have sufficient visibility, participation and longevity to achieve lasting change. Areas suggested that the Home Office could increase the scale of interventions by coordinating activities across funded areas, besides making more investment available overall. In addition, funding guidelines should encourage or require areas to include plans for sustainability in their bid.
3. Increase the continuity of interventions across funded areas and over rounds. Focusing on specific outcomes, interventions or other focus areas could improve the ability to detect potential change. Some stakeholders suggested a national coordination of some interventions to achieve a consistent message and increase reach.
4. Continue to encourage innovative approaches and delivery methods that are not in the scope of other funds. However, this freedom should exist within a more focused approach, with clarity on the core outcomes and intervention categories that the Fund will support so that change is targeted.

Future evaluations

1. Focus on measuring short-term outcomes, which are the direct effects of interventions, particularly when funding periods are short. This will help to demonstrate any short-term effects of SWaN when achieving and measuring longer-term impacts is challenging.
2. Consider the funding period as it relates to the evaluation. Build in more time before the delivery of interventions begins to ensure sufficient time for uncontaminated baseline data collection. Take greater consideration of the post-intervention data collection timeframe to ensure change is measured at the most appropriate time.

Acknowledgements

First, our thanks go to our partners in the research team at Ecorys, particularly Shona Macleod and Katie Smith, to Dr James Copestake, Fiona Remnant and their team at Bath Social & Development Research (BSDR) and to Professor Nicole Westmarland and Dr Kate O’Brien the Centre for Research into Violence and Abuse (CRiVA) at Durham University, for their invaluable contributions on design, fieldwork, analysis and reporting. We are also grateful to Home Office analysts and policy colleagues for their regular guidance and feedback throughout the evaluation. Our thanks also go to the anonymous peer reviewers who provided us with comments during the reporting stage and finally, to the evaluated project staff and their local residents who gave their time generously to the evaluation process.

1. Introduction

This section sets out the background to the Safety of Women at Night (SWaN) Fund and its purpose, as well as a summary of the projects it funded.

1.1 About the Fund

News coverage of the tragic murders of several women in the UK brought to the forefront public concerns about the safety of women and girls and their risk of experiencing violent crime in public spaces. Evidence from the Crime Survey for England and Wales (CSEW), considered the most reliable data source for measuring the prevalence of VAWG, shows that women are more likely to be victims of sexual assault. For example, in the year ending March 2023, 26.5% of women and 6.1% of men reported experiencing sexual assault (including attempts) since the age of 16 (ONS, 2023). Women are also less likely to feel safe walking alone in public settings than men (ONS, 2022).

The UK government’s 2021 Tackling Violence Against Women and Girls Strategy announced several actions to help protect the safety of women and girls (Home Office, 2021a). This included the launch of a new £5 million SWaN Fund for interventions that aimed to improve feelings of safety and prevent VAWG in public spaces at night, including inside night-time economy (NTE) venues.

The Fund was open to civil society organisations (charities, community and voluntary organisations, social enterprises and co-operatives), local authorities, Police and Crime Commissioners (PCCs), and the Chief Constable of the British Transport Police. Each bid could have a maximum value of £300,000.

1.2 Aims of the Fund

The SWaN Fund aimed to:

• reduce incidents of VAWG in public spaces at night, including inside NTE venues
• reduce the fear of VAWG and improve feelings of safety in public spaces at night, including inside NTE venues
• build an evidence base for what works in improving women’s safety in public spaces at night, including inside NTE venues (Home Office, 2021b)

1.3 SWaN-funded projects

The Home Office received 82 eligible bids, of which 22 were awarded SWaN funding. Each bid designed a unique programme comprising multiple intervention types. This section summarises the design of the successful bids, also referred to as ‘funded areas’.

1.3.1 Target audiences and public spaces

Table 1.1 outlines the target audience groups and public spaces of planned interventions, summarising the number and proportion of successful bids targeting each. Nearly all funded areas delivered interventions to multiple audience groups and public spaces. The most common audience groups were members of the general public (95% of bids), NTE workers (91% of bids) and public guardians (82% of bids). Police were also a key audience group, with 68% of bids including police-focused interventions. In relation to public spaces, most bids included interventions targeted at public streets (86% of bids) and NTE venues (68% of bids).

Table 1.1: Intervention target groups and spaces

	Number of bids	% of bids
Target audience groups
Members of the general public	21	95%
NTE workers	20	91%
Public guardians	18	82%
Police	15	68%
Taxi drivers	12	55%
Support workers	7	32%
Other workers	6	27%
Local authorities	4	18%
Target public spaces
Public streets	19	86%
NTE venues	15	68%
Universities	7	32%
Schools	5	23%
On public transports	4	18%
Community venues	3	14%

1.3.2 Interventions

Across the 22 funded areas, projects were commissioned to deliver a range of intervention types. A bid review at the outset of the evaluation identified the following 7 distinct categories of funded interventions, defined based on their target audiences and the grouping of activities aimed at those audience groups:

• interventions aimed at NTE venues and workers
• interventions aimed at other services and workers working at night, including police, transport workers and other emergency services workers
• interventions aimed at changing public perceptions and awareness of VAWG
• guardianship initiatives which increase or employ individuals to promote safety and welfare in the NTE
• personal safety initiatives aimed at supporting individuals’ ability to keep themselves safe
• strategic and systems-wide projects aimed at supporting project delivery, including research and project management personnel
• support services and outreach for victims of crime, or to prevent crimes through early intervention services for young people

The most common intervention categories, each planned by 86% of bids, were interventions aimed at NTE venues and workers, interventions aimed at changing public perceptions and awareness, and guardianship interventions. Over three-quarters of bids (77%) also planned interventions to develop strategies or enable a systemic change of VAWG in NTE.

The 3 most common intervention categories remained dominant when looking at the proportion of planned Fund expenditure on each intervention category. Guardianship interventions received the greatest proportion (27%), followed by public perceptions and awareness interventions (23%), and interventions targeting NTE venues and workers (22%). In some cases, the prevalence of the intervention category was not a direct indication of the proportion of planned expenditure. For example, while 77% of bids featured strategy and systems interventions, only 7% of planned expenditure was allocated to these types of intervention.

The bid review identified a set of intervention sub-types within each of the 7 broader intervention categories. In total, 20 sub-types were identified. Table 1.2 provides a full breakdown of planned interventions by the seven intervention categories and their sub-types. It shows the number and proportion of bids within each planned intervention category and sub-type, and the proportion of total and category Fund spend allocated to each.

Table 1.2: Summary of SWaN interventions and expected spend

	Number of bids	% of bids	% category spend	% Fund spend
Interventions aimed at NTE venues and workers	19	86%		22%
Education and training	18	82%	63%	14%
Communications and key messages	5	23%	23%	5%
Equipment and technology	1	5%	1%	0%
Standards, accreditation, and NTE venue-wide policy changes	10	45%	12%	3%
Interventions aimed at other services/workers working at night	14	59%		11%
Education and training for workers	9	41%	40%	4%
Communications and key messages	1	5%	2%	0%
Equipment and technology	2	9%	42%	4%
Standards, accreditation, and NTE venue-wide policy changes	5	23%	16%	2%
Interventions seeking to change public perceptions and awareness	19	86%		23%
Campaigns focused on public awareness and perceptions of VAWG and NTE	15	68%	59%	13%
Campaigns focused on challenging negative attitudes	7	32%	23%	5%
Campaigns focused on help-seeking and personal safety	5	23%	18%	4%
Guardianship initiatives	19	86%		27%
Safety zones/hubs in NTE spaces	9	41%	34%	9%
Public guardians in NTE spaces (including street pastors and taxi marshals)	15	68%	66%	18%
Personal safety	6	27%		3%
Apps promoting/enabling people to keep themselves safe	3	14%	73%	2%
Equipment (such as alarms and whistles) enabling people to keep themselves safe	5	23%	27%	1%
Strategic and system-wide	17	77%		7%
Project leadership roles	7	32%	43%	3%
Project/admin costs	7	32%	30%	2%
Research and analysis to support increased intelligence and identification of issues	7	32%	23%	2%
Support services	8	36%		6%
Youth work	4	18%	29%	2%
Victim support	4	18%	71%	4%

Five interventions were excluded from the impact evaluation due to the unlikelihood they would have sufficient reach or impact within the funding period to bring about measurable change, although they remained part of the process evaluation. These interventions included apps, personal safety equipment, youth work, most support services, and attitude change programmes taking place in schools. All other interventions, categorised as either public-facing or NTE worker-facing, were considered in scope for impact evaluation.

2. Evaluation design and methodology

This section outlines the aims and research methodology of the SWaN evaluation. More information on the evaluation design is in Appendix A.

2.1 Evaluation aims and research questions

The Home Office commissioned Verian and Ecorys UK to conduct an impact and process evaluation of the SWaN Fund. These aimed to understand the impact of SWaN interventions, and the factors that supported or inhibited implementation at each stage of the SWaN programme.

The overarching aim of the impact evaluation was to measure the impact of SWaN interventions at Fund-level. This means that it sought to understand the average impact across funded areas, rather than for each funded area or for each intervention type. The impact evaluation sought to address the following 4 key research questions:

• what is the impact (if any) of the funded interventions on feelings of safety, with a particular focus on women and girls’ feelings of safety at night?
• does impact differ by type of area (for example, urban versus rural, residential versus commercial, NTE versus residential)?
• what is the best methodology to use to look at and understand the impact of interventions over time?
• do the funded interventions have any other benefits?

As a key aim of the Fund was to reduce incidents of VAWG, the impact evaluation additionally sought to address the following research question:

• what is the impact (if any) of the funded interventions on VAWG crimes?

However, this was later considered out of scope of the evaluation for several reasons. The high level of under-reporting of VAWG offences means that police recorded crime (PRC) data do not provide an accurate measure of prevalence. It is also difficult to identify the reasons behind any change in the number of recorded offences, which are likely driven by several factors, including improvements in police recording practices, and therefore problematic to attribute any changes to SWaN interventions. Appendix A, Section 7.1.2 summarises the 4 research questions addressed by the impact evaluation with the corresponding outcome and impact measures.

The purpose of the process evaluation was to explore the experience of funded areas in developing and implementing SWaN interventions. It aimed to uncover the key factors that supported or inhibited implementation at each stage of the SWaN programme, and to identify good practice emerging from the Fund. The process evaluation was designed to address the following research questions:

• what resources have helped bidders inform their planned interventions and their identification of chosen areas?
• how did consultation with key stakeholders and the community inform project design and delivery?
• what are the factors and barriers that facilitate/inhibit the successful implementation of funded interventions in public spaces at night and in the NTE, in particular focusing on those that have developed innovative interventions and those implementing interventions within premises?
• do the success factors or barriers differ by intervention type and area type?
• is there any evidence that implementing these interventions has changed or freed up police resources at all?
• what were the key factors in areas that worked well with partners?
• what approaches work well to ensure a high level of engagement by the local community?
• how have bidders engaged with relevant VAWG groups, in particular outlining the challenges, successes and lessons learned?
• what (if any) has been the impact on local organisations (women’s groups, schools, universities, local businesses), particularly those involved in the bid development?
• are there any other lessons that would help inform future commissioning of this Fund and best practice?

2.2 A note on terminology

Throughout this report, the phrase ‘violence against women and girls’ (abbreviated to VAWG) is used to describe a range of crimes, including rape and other sexual offences, stalking, ‘upskirting’, and many others. VAWG crimes are known to disproportionately affect women and girls, but men and boys can also be victims. References of VAWG in this report do not include domestic abuse, as this was out of scope for the evaluation. VAWG is the official terminology adopted by the Home Office. In the surveys, the phrase ‘sexual harassment and gender-based violence’ was used instead, to make questions more accessible and digestible for respondents. To remain consistent with the official Home Office terminology, the phrase ‘sexual harassment and gender-based violence’ is referenced as ‘VAWG’ throughout the report.

2.3 Scoping phase methodology and outcomes

The scoping phase aimed to establish a detailed understanding of the funded areas, including their projects, target crimes, target audiences, and the intended outcomes and mechanisms for change. This phase included interviews with 5 stakeholders from the Home Office and external policy experts, a detailed review of the bid documents from the 22 successful bid areas (Section 1.3), and a final stage of analysis and mapping to design an evaluation which was fit for purpose.

The scoping phase highlighted several challenges for measuring the impact of the SWaN Fund. Some challenges related to the design of the Fund itself, predominantly the variability of interventions across funded areas in terms of their inputs, reach and scale. Other challenges related to the availability and collection of data, specifically the problematic use of PRC data (as explained in Section 2.1) to measure crime outcomes, and the lack of time to collect baseline survey data before interventions started. The evaluation was designed to maximise evidence despite these challenges; however, these ultimately affected the ability to detect potential impact (see Section 4.3 for further details of these challenges).

2.4 Impact evaluation methodology

The impact evaluation used a mixed-methods approach. Quantitative surveys were designed to measure the scale of impact for 2 core beneficiary groups. Verian used a face-to-face (F2F) intercept survey to capture responses from women using public spaces at night. An email survey was sent to direct recipients of SWaN interventions (for example, NTE workers). Each beneficiary group was treated distinctly so the analysis measured impact within each group; it did not measure or compare impact between these groups. Qualitative data supplemented survey data to help understand the factors driving positive and negative changes in feelings of safety. This research followed the Qualitative Impact Protocol (QuIP) method, which involved conducting online in-depth interviews with a sample of F2F survey respondents and eliciting open-ended responses to the email survey.

Due to the sensitive nature of the topic, measures were put in place to limit harm to respondents. These included ensuring at least one female interviewer in each F2F survey location and providing all respondents with a list of support services. Sensitive questions around personal experiences of VAWG were presented to F2F survey respondents on showcards, allowing them to indicate a letter or number corresponding to their answer. This helped to protect their confidentiality, as they were not required to verbalise their experiences.

2.4.1 Summary of survey modes and respondents

A. F2F survey

The F2F survey measured the impact of SWaN interventions on a range of key outcome measures. These were: feelings of safety in public spaces at night and inside NTE venues; perceptions of VAWG in the local area; perceptions of the police and other NTE workers; and likelihood to report VAWG to police staff. The scoping phase identified 20 funded areas with interventions in scope for F2F survey data collection and an additional 12 areas for collecting comparison data. Comparison locations were initially identified using recommendations from funded areas, followed by a data-led approach using demographic and PRC data to verify their suitability (see Section 2.4.3 for further details). Within the budget available for the evaluation, Verian aimed to conduct 1,000 F2F interviews.

Interviews were conducted during the evening (between 5pm and 9pm) in high streets, NTE areas, in and around transport hubs, and in university campuses, mirroring the times and public spaces of the SWaN interventions delivered in those areas. Interviewers placed themselves in areas of high footfall, including entrances to venues, and asked women using these public spaces if they would be willing to take part in the survey.

Fieldwork was conducted in 2 phases due to the restrictions imposed on government activities prior to local elections. The first phase ran from 24 March 2022 to 13 April 2022. Fieldwork then resumed for the second phase from 6 to 31 May 2022. It is important to note that some funded activities were likely to have finished before fieldwork started and were no longer visible to respondents at the point of data collection. This could have affected respondents’ recall ability and subsequently survey results. Respondents were selected using a combination of purposive (only females included in the sample) and convenience sampling. The survey achieved 1,168 responses (funded areas: 844 responses; comparison areas: 324 responses). A demographic breakdown of respondents is included in Appendix A, Section 7.1.3.

B. Email survey

The email survey aimed to capture the impact of SWaN interventions delivered to key NTE worker groups. The survey measured impact on the following outcome measures:

• perceptions of personal safety
• perceptions of safety for women and girls using public spaces
• perceptions of VAWG
• self-reported capabilities in recognising and responding to VAWG

This survey was also distributed to a few students who received SWaN training and awareness interventions. However, as this group formed a small proportion of the final sample (less than 1%), email survey respondents are described as NTE workers throughout this report.

Mapping identified 20 funded areas delivering interventions to NTE workers. Funded area leads or intermediaries (for example, training providers) distributed the survey link to target respondents on behalf of Verian. Comparison respondents were also contacted to complete the survey (see Section 2.4.3 below for further details).

Fieldwork was conducted from 17 May 2022 to 6 July 2022, selecting respondents using a combination of purposive (based on occupation) and volunteer sampling. The survey achieved 465 responses, of which 217 were responses in funded areas (51 male and 165 female) and 248 responses in comparison areas (57 male and 188 female). A full demographic breakdown of the email survey respondents is in Appendix A, Section 7.1.3.

2.4.2 Retrospective impact measurement

The co-occurrence of the beginning of the intervention delivery period and the evaluation meant that it was not possible to obtain an uncontaminated baseline survey before interventions launched. Subsequently the evaluation adopted a retrospective approach to impact measurement. This involved conducting a single cross-sectional survey which asked respondents how they felt on the day of the survey, and then the degree to which this had changed for them over the previous 6 months (compared with October 2021). A scale of 0 to 10 was used for the measurement of the change over time, with 0 indicating a significant decrease, to 10 indicating a significant increase. This answer scale would detect any level of impact, even where changes were small.

2.4.3 Impact analysis

Following data collection, the survey data was analysed to assess the impact of SWaN interventions. Impacts were estimated by comparing the scale of change reported by respondents from funded areas with that of respondents in comparison areas, while attempting to account for differences between these 2 groups. The difference in the retrospective change reported between the 2 groups is the impact attributed to SWaN interventions, with this report focusing on differences that are statistically significant at the 5% level. The analysis assumes that the models account for all relevant differences between the bid and comparison areas. Any remaining differences between the 2 groups can then be interpreted as the impact of SWaN interventions. However, there remains a risk that apparent impacts are due to some unobserved differences between the groups, rather than the effects of the Fund. Further information about these models is in Appendix C, Section 7.3.2.

Funded areas provided initial suggestions for counterfactual locations (referenced as ‘comparison areas’ throughout this report) for the F2F survey. Their suitability were verified using a range of data sources, namely: deprivation statistics; Department for Transport journey time statistics; Census population statistics; and PRC data on the number of violent and sexual offences (April to June 2021). This was to confirm that the comparison areas were broadly similar and therefore comparable to the funded areas.

Funded areas distributed the link to the email survey to NTE workers from comparable venues and organisations in similar areas to act as a comparison group. As the Home Office could facilitate contact between Verian and police forces, a more data-led approach to selecting comparison police officers was adopted. This involved identifying similar towns and cities to funded areas using the Office for National Statistics (ONS) analysis on major towns and cities and asking the relevant police force to distribute the survey to officers working in those areas.

2.4.4 Qualitative Impact Protocol (QuIP)

Verian conducted a QuIP study in partnership with Bath Social and Development Research (SDR). QuIP is a qualitative approach to impact evaluation that involves analysing data collected from in-depth interviews. As opposed to quantitative impact analysis which measures the scale of change, QuIP aims to identify the key factors that individuals attribute to the change.

Two QuIP studies were conducted for the SWaN impact evaluation:

1. A full QuIP^{[footnote 1]}, focusing on perceptions of safety and the factors influencing this. In-depth online interviews were conducted with 23 women sampled from funded areas. A sample frame was developed from F2F survey responses based on permission to recontact and reports of change. Survey data was used to identify respondents who had reported changes in their perceptions of safety in public spaces at night, and their perceptions of VAWG, with a final sample frame of 100 cases. From the sample frame, a sample of QuIP participants was randomly selected^{[footnote 2]}. The final sample (n=23) comprised 11 participants who had reported a positive change and 12 who had reported a negative change.
2. A mini QuIP^{[footnote 3]}, focusing on changes in NTE workers’ perceptions of safety for women and girls using public spaces at night, and the factors influencing this. This focused on why they felt that others (specifically women and girls) had become more or less safe in these settings. Email survey respondents who reported a change in their perceptions of safety for women and girls were asked 2 additional ‘open’ questions^{[footnote 4]} that aimed to elicit the factors that had influenced change. In total, 112 NTE professionals provided open-text responses.

A trained QuIP analyst from Bath SDR coded and analysed the data for each QuIP study, using the Causal Map App to identify causal chains and pathways in the data. It is important to note that QuIP findings reflect experiences of change relating to individual participants only. Findings are not generalisable to the wider population.

2.5 Process evaluation methodology

The process evaluation was designed to address the research questions described in Section 2.1 while complementing the impact evaluation to deliver a comprehensive review of the Fund. It used a variety of qualitative primary research methods, with the key elements and phasing of the process evaluation summarised in Figure 2.1.

Figure 2.1: Process evaluation overview

2.5.1 Process evaluation research methods

A. Data collection: qualitative proforma survey

A qualitative proforma survey was emailed to all funded areas in January 2022. This aimed to understand how areas approached bid development and intervention delivery. Funded areas had 2 weeks to complete the proforma, with 21 of the 22 funded areas responding. Responses were analysed and the findings were used to inform the case study selection.

B. Data collection: case studies

Case study research was conducted with stakeholders from 8 funded areas, including strategic, operational and frontline staff. Each case study aimed to explore key intervention themes and identify lessons learned from each stage of the SWaN programme, including intervention delivery and bid development. They also sought to understand how the various interventions delivered in a single funded area interacted or complemented each other.

Case study areas were selected based on key selection criteria, such as geographical location, funding amount and types of interventions included. The research consisted of 2 main elements:

Document review and theory of change (ToC)^{[footnote 5]} development: A review of a range of information sources, including project plans, strategy documents and the proforma survey. Following this, a ToC was developed for each case study area. The theories of change were shown to funded area leads and discussed during interview. This aimed to help researchers understand where deviation from original plans had occurred, and how this might have impacted on intended effects of interventions.

Interviews and focus groups: Each case study involved multiple consultations with 4 main stakeholder groups. Interviews were conducted with the funded area lead, and strategic or operational leads. Interviews or focus groups were conducted with delivery staff and, where appropriate, intervention participants (such as NTE workers and the police). Participants were recruited using a snowball sampling method.

C. Data collection: additional funded area lead interviews

Fourteen interviews were undertaken with bid leads from the remaining funded areas not selected for the case study research to ensure representation across the Fund. These aimed to capture views and experiences at each stage of the SWaN programme, and reflections on lessons learned.

Across the case study research and interviews with additional funded area leads, 53 stakeholders were consulted.

D. Learning workshop

Ecorys UK held an online learning workshop for all funded areas. This enabled projects to share learnings and experiences of delivering SWaN interventions in relation to 2 themes. These themes were identified following initial analysis of early interview data:

1. Consultation and engagement. This discussion theme focused on how projects engaged communities and expert partners in their SWaN interventions before and during delivery.
2. Sustainability. This discussion theme adopted a forward-looking focus on the sustainability of interventions, as well as identifying lessons learned and priorities for future commissioning.

3. Descriptive analysis of crime and safety in SWaN-funded areas

This section presents results from the descriptive analysis of F2F survey responses for a selection of measures. This provides useful context for the experiences, feelings and perceptions of crime and safety in SWaN-funded areas as reported by women using public spaces. Equivalent findings from the email survey are presented in Appendix B.

3.1 Experience of crime

Using data from the F2F survey, Figure 3.1 shows experience of VAWG at night in the last month. Overall, 36% of users of public spaces reported personally experiencing or witnessing VAWG in the last month. A similar proportion was reported in comparison areas (38%)^{[footnote 6]}.

Figure 3.1: Experience or witnessing VAWG

Notes:

1. In the last month in this area, have you personally experienced or witnessed any of these types of VAWG? (Base=844).

Respondents aged 16 to 24 were statistically more likely to report experiencing or witnessing VAWG at night in the last month than almost all older age groups, except the over 60 group. The 16- to 24-year-old group formed more than half (51%) of those who reported experiencing or witnessing VAWG. Analysis also found that individuals with a disability were statistically more likely to report experiencing or witnessing VAWG at night in the last month (30%), compared to those without a disability (19%).

3.2 Worry about crime

As shown in Figure 3.2, more than half (53%) of users of public spaces reported feeling very or fairly worried about VAWG, with a similar proportion reported in the comparison areas (55%). Analysis found that respondents aged 16 to 34 years old, those with a disability, and those who reported experiencing or witnessing VAWG in the last month were statistically more likely to worry about VAWG than other groups.

Figure 3.2: Worry about VAWG in area after dark

Notes:

1. How worried are you about experiencing VAWG in this area after dark/at night? (Base=844).

3.3 Perceptions of safety in public spaces

Figure 3.3 shows perceptions of safety in NTE areas among users of public spaces. Overall, 40% of users of public spaces reported feeling safe in NTE areas. However, feelings of safety inside NTE venues differed by venue type: women were less likely to report feeling safe inside clubs or other age-restricted night-time venues (49% safe), than in bars and pubs (70% felt safe), and in restaurants and cafes and other entertainment venues like cinemas, theatres, and bowling alleys (91% felt safe for both location types).

Figure 3.3: Perceptions of safety in NTE areas and inside NTE venues

Notes:

1. How safe or unsafe would you feel walking on your own after dark/at night where there are pubs, bars and clubs? (Base=844).
2. How safe or unsafe do you feel inside these types (above) of venues in this area…? (Base=311).

Individuals with a disability were statistically significantly more likely to report feeling unsafe in restaurants (17%) compared with respondents without disabilities (5%). Individuals in the 16 to 24 years age group were statistically more likely to report feeling unsafe in pubs and bars and clubs and other age-restricted venues than some of the older age groups. For example, 36% of 16-to 24-year-olds felt unsafe in bars and clubs compared with 16% of 25- to 34-year-olds and 15% of 35-to 44-year-olds.

4. Impact of the Fund

This section presents the results from the impact evaluation. This measured the impact of SWaN interventions on a range of outcomes related to perceptions of crime and safety, and confidence in the police and other NTE workers in addressing VAWG. It also reports the impact on outcome measures related to NTE workers who directly received a SWaN intervention.

4.1 Impact evaluation summary

The summary of findings presented here should be read along with the description of factors which may have mitigated the detection of impact described in Section 4.3 below.

The main impact measure for this evaluation was changing perceptions of safety among women using public spaces at night, including inside NTE venues and the public spaces around them. Analysis of the F2F survey data showed:

• there was no statistically significant evidence that SWaN interventions improved women’s feelings of safety in public spaces at night
• there was some evidence to suggest that SWaN interventions could have had a counterproductive effect, with a statistically significant reduction in feelings of safety inside pubs and bars detected
• there was no statistically significant evidence that impact on feelings of safety varied between the intervention types planned and by interventions targeting specific public spaces (for example, that feelings of safety on transport networks were influenced by interventions targeting those spaces)
• no statistically significant impacts from SWaN interventions were detected for other outcome measures, specifically: perceptions of the prevalence of VAWG in public spaces at night; confidence in local police and other NTE workers in dealing with VAWG; and likelihood of reporting VAWG incidents to local police and other NTE workers

However, analysis of the email survey data found statistically significant evidence that SWaN interventions were associated with an improvement in NTE workers’:

• awareness and understanding of VAWG
• self-reported confidence in supporting victims of VAWG
• feeling able/equipped to intervene in VAWG at their place of work or study

Despite detecting no significant quantitative evidence that SWaN interventions had their desired impact on users of public spaces, QuIP analysis provides evidence of positive effects for some individuals. Some women linked their improved perceptions of safety in the NTE to the types of interventions funded by SWaN, specifically:

• increased security at NTE venues (including increased surveillance and physical infrastructure in venues)
• increased police presence in NTE areas
• measures adopted by NTE venues (including venue-wide schemes and prevention of drink-spiking measures)

This demonstrated that these types of interventions could be effective mechanisms for improving perceptions of safety in the NTE. QuIP research with NTE workers also revealed examples of positive change from SWaN interventions, such as feeling more able to identify risk and take appropriate action.

4.2 Quantitative impact analysis results

4.2.1 Impact on perceptions of safety in public spaces

Overall, the evaluation found no statistically significant impacts on improved feelings of safety in public spaces at night. All impact estimates on general feelings of safety in NTE areas, on high streets, and on/around public transport are close to zero and not statistically significant. This indicates that any impacts are likely to be small.

However, the impact analysis detected one statistically significant result for feelings of safety inside NTE venues. On average, women in funded areas were slightly more likely than women in the comparison areas to say they thought NTE venues had become less safe in the last 6 months, for safety inside pubs and bars. This suggests that SWaN interventions could have had a counterproductive effect on feelings of safety, although the potential negative impact seems fairly small.

Further analysis explored the extent to which impacts on feelings of safety in NTE areas and inside venues varied between different demographic groups and types of area (see Appendix C, Section 7.3.4. There was some evidence to suggest impacts on feelings of safety in NTE areas differed by age, with a statistically more positive impact on older women than younger women. However, this finding should be interpreted with caution as the confidence intervals are relatively wide (indicating a high level of uncertainty about the difference), and there was no strong prior theory-based reason to predict this difference.

4.2.2 Perceived change in safety of women and girls

NTE workers were also asked the extent to which they thought the safety of women and girls using public spaces at night had changed since October 2021. NTE workers in funded areas reported a statistically significant improvement in their perceptions of safety for women and girls walking on their own in NTE areas after dark. Respondents were also asked about their perceptions of safety for women and girls at night inside NTE venues, at public transportation stations, and while on public transport. The impact estimates for these outcomes were broadly similar, with a perceived increase in safety in funded areas as compared with NTE workers in other areas, although they were not statistically significant. Further analysis on gender was conducted since it was plausible that the effects on perceptions of safety for women and girls would be different for male and female NTE workers. However, analysis showed no statistically significant differences between the estimated impacts by gender.

4.2.3 NTE worker capabilities

The email survey with NTE workers included a set of short-term outcome measures related to their capabilities in recognising and responding to VAWG. These outcome measures captured some of the immediate effects resulting from SWaN-funded interventions, where longer-term impact is difficult to observe.

Findings from the email survey suggest that, overall, SWaN-funded interventions had a positive impact on NTE worker outcomes. In particular, analysis identified statistically significant evidence that the awareness and understanding of VAWG, confidence in supporting VAWG victims, and capabilities in intervening in VAWG incidents had improved for NTE workers in funded areas. Positive effects on the perceptions of their organisation’s capabilities in dealing with VAWG incidents were also found, with statistically significant increases detected for their organisation’s strategy for ensuring the safety of women and its co-operation with other organisations to address VAWG.

4.2.4 Other analyses

Further analysis explored whether impacts on feelings of safety in different public spaces and inside NTE venues varied between the intervention types planned, the public spaces targeted by interventions, and demographic groups. Across all these types of further analysis, no statistically significant evidence was found, except for a small difference in age (see Appendix C, Section 7.3.4).

The F2F survey also asked women about changes in their perceptions of the prevalence of VAWG in public spaces at night, their confidence in the local police dealing with VAWG, and their likelihood of reporting incidents to local police. In each case, there was no statistically significant difference between the views of women in funded areas and women in comparison areas. Further analysis found limited evidence that particular intervention types led to greater impacts on the perceived prevalence of VAWG. Although some intervention types were more strongly associated with the perception that VAWG has decreased (frontline support interventions and personal safety initiatives), estimates were not statistically significant. Similarly, there was no statistically significant evidence that impacts varied by demographic groups or the type of area in which the survey was conducted (see Appendix C, Section 7.3.4 for additional analysis).

Respondents were also asked the extent to which their confidence in NTE venue staff to deal with VAWG and their likelihood to report VAWG to police staff had changed since October 2021. Across both impact measures and NTE venue staff type (pub and bar staff, or club and other age-restricted venue staff), no statistically significant evidence was found.

4.3 Factors influencing the impact of SWaN

Several factors may have limited the ability of the evaluation to identify potential positive impacts on the perceptions and feelings of safety among users of public spaces. These relate to the design of the SWaN Fund, limitations of the evaluation, and broader factors that also influence feelings of safety.

First, it is possible that the relatively low intensity of interventions meant that potential impacts were too small to detect. Funding was spread across a range of activities in each area. This meant that many interventions were being implemented on a relatively small scale. Linked to this, the short delivery period for interventions (around 3 to 4 months) limited the likelihood of achieving and detecting potential positive impacts.

This hypothesis is supported by the low awareness of interventions reported by users of public spaces (see Figure 7.5 in Appendix B, Section 7.2.4). For example, adverts were a common type of intervention (86%) but were only recognised by 9% of respondents in areas where these were funded. Similarly, there was very low awareness of education/training and guardianship initiatives in areas funding these intervention types (both 2%). The low ‘noticeability’ or ‘visibility’ of SWaN interventions is likely to have diluted the impact of the Fund. It appears many individuals had little or no interaction with the interventions to reap their effects.

Second, the variation within the Fund presented substantial challenges for evaluation. There were 22 funded areas, each delivering a different combination of activities, at different scales, and targeted towards different audiences. Due to resource and time constraints, it was not feasible to design evaluation methods for each funded area or intervention. As such, the evaluation was designed to investigate effects at Fund level. However, impacts most likely varied between areas. It is therefore possible that some interventions were effective in particular areas, even if the evaluation could not detect this.

Third, some interventions may need to be delivered for longer for their effects to be detectable among the wider public. For example, although there is evidence of positive effects among NTE workers (see Section 4.2.3), these may not yet have flowed through to observable impacts on the wider public in the area due to the short timeframe. However, it is also possible that these positive effects may not translate to changes in public perceptions of safety. Similarly, the timing of the evaluation fieldwork posed some challenges. Since the start of the evaluation co-occurred with the start of intervention delivery, it was not possible to obtain an uncontaminated baseline survey before interventions launched and some interventions were still being delivered or had only recently been completed when data collection took place. The Fund-level focus of the evaluation, the relatively short timeframe available, and a pause in fieldwork due to local elections meant optimal time data collection for each area was not feasible. As a result, there is a risk that impacts in certain areas were missed or underestimated because the timing of data collection was not aligned with the local implementation of interventions. In addition, as illustrated through the process evaluation (Section 5.5), local stakeholders generally said that more time would be needed for impacts to be observed.

Fourth, SWaN interventions were not being delivered in isolation, which presented challenges in identifying suitable comparison areas and attributing impact to SWaN interventions. Other similar initiatives were being implemented in other parts of the country. In some cases, these were delivered through other Home Office funds with overlapping aims (for example, the Safer Streets Fund (SSF)). In other cases, these may have been delivered through other funds, local schemes, or local voluntary organisations and community groups. The comparison areas for the survey were selected to avoid areas known to be taking part in similar funds (in particular, areas receiving funding through the SSF). However, it cannot be ruled out that other similar activities, or different activities with similar aims, were also taking place in comparison areas over a similar timeframe. This could reduce the apparent impact of SWaN interventions.

Furthermore, there may have been differences between the women surveyed in funded areas and the women surveyed in comparison areas, which could have affected the impact estimates. The comparison areas for the F2F fieldwork were selected on the basis they were similar areas to the places SWaN interventions were implemented (see Section 2.4.3). However, data were collected from women passing through public spaces at particular times and so there may still be important differences between the types of women surveyed in these areas. While the impact analysis tries to account for these differences (see Section 7.3), it is impossible to be sure that there are no unobserved differences affecting the impact estimates.

4.3.1 Factors that influence feelings and perceptions of safety in public spaces

QuIP research (as explained in Section 2.4.4) explored the multi-faceted factors that can influence perceptions of safety and therefore highlighted the complexity in measuring and attributing the impact of SWaN. These factors might have been in effect alongside SWaN interventions, and therefore could have influenced the Fund’s ability to impact perceptions of safety.

QuIP analysis identified three main categories of influencing factors on perceptions of safety – social, personal and environmental. Some of these factors relate to the types of interventions implemented by SWaN; however, these outcomes are not directly attributable to the Fund.

A. Social factors

Relationship to others and/or perceptions of others
Individuals’ relations to “others and/or perceptions of others” was one of the most frequently cited factors contributing to feeling safe. QuIP participants mentioned feeling at ease, especially at NTE venues and on public transport, around “familiar faces” (people known to, considered similar to, or perceived as sociable to the respondent).

Presence of others
The “presence of others” was identified as a factor that can both positively and negatively influence perceptions of safety, depending on whether there are more or less people around and depending on the context, for example, time of day or type of public space.

Many participants shared that the presence of others (more people being around) helped them to feel safe, particularly when accessing the high street and public transport. In line with this, others explicitly stated that less people on the high street and on public transport made them feel unsafe, especially at night. Here, the high street and transport were described as becoming “deserted” and “desolate”. Some participants explained that they perceived these circumstances as increasing the threat of crime or violence and reducing the likelihood that someone could help them if needed.

At the same time, some participants reported that more people around them made them feel unsafe. This was particularly in relation to accessing NTE venues, such as pubs, bars and clubs. Participants described these places as busy or “crowded”, and they perceived this as enabling or allowing VAWG to take place undetected. For example, one participant described avoiding clubs altogether because of the crowds, which made her feel at risk of VAWG perpetrated by people taking advantage of the crowds.

Antisocial behaviour
Anti-social behaviour (ASB) was the most frequently cited factor contributing to feeling unsafe. Examples of ASB observed included people loitering, shouting, fighting and throwing things. These behaviours were often linked to people being under the influence of alcohol or drugs. Some of the women interviewed described feeling “uncomfortable”, “intimidated” and “unsafe” due to this type of behaviour. While there was concern associated with the ASB itself, some women also stated or implied a fear of “what could happen” (the perceived risk that the behaviour could quickly escalate into other forms of crime or violence). A few participants explicitly noted that others’ alcohol consumption or drug use heightened the degree of threat they perceived they faced, such as threat of violence or assault because their behaviour was considered more unpredictable and uncontrollable.

“When people are drunk, I feel like they are more likely to participate in that sort of criminal behaviour [violence/ assault] than if they’re sober.”

B. Personal factors

Personal experiences of VAWG
Personal experiences of VAWG also influenced perceptions of safety. Many women shared that they felt somewhat unsafe in public spaces due to previous (that is, pre-October 2021) personal experiences of VAWG. The most common form of VAWG mentioned in this regard was harassment, and particularly verbal sexual harassment. For example, participants described “cat calling”, being “heckled” and unsolicited “flirting” usually taking place on the streets and in public transport stations.

“If you’re walking down the street with lots of people that are very drunk, you tend to get shouted [at], and in that moment you can feel very small – you’re surrounded by people that find that funny, and you don’t find that funny because the attention is now on you.”

In contrast, some participants explicitly shared that they had not experienced any issues with crime or VAWG themselves, which reinforced their sense of safety.

Perceptions of VAWG
Besides personal experiences, some participants noted how their awareness of VAWG, from word of mouth or the media, negatively affected their perceptions of safety. For example, one participant described hearing stories of other people getting spiked “every other week”, which had a negative influence on their perception of safety.

Perception of other types of crime
Some participants noted how their perception of other crime types (including references to gun or knife crime, theft and vandalism) negatively affected their perceptions of safety.

C. Environmental factors

Some women taking part in the interviews mentioned how environmental factors impacted their feelings of safety in public spaces. Some women felt that streetlighting, or “well-lit” venues contributed to feeling safer, and the absence of streetlighting was a reason why some women felt unsafe at night, particularly in high streets, residential streets and in parks. Streetlighting was not a SWaN-funded intervention and is therefore a wider environmental factor influencing feelings of safety outside of SWaN-funded interventions.

4.3.2 Changes in feelings and perceptions of safety outside of SWaN

Besides understanding the general factors influencing feelings of safety, QuIP also explored the factors participants attributed to positive and negative changes in their feelings of safety since October 2021, but which are not related to SWaN interventions.

For some participants who reported feeling safer, this was because there were more people around compared with October 2021, with a couple stating that this increased footfall was due to COVID-19 lockdown restrictions being lifted. Of the participants who felt less safe, some participants linked to the perception that VAWG was increasing, their increased awareness of VAWG, and their own experiences and observations of VAWG. Spiking was the VAWG crime most frequently reported to be increasing. Some QuIP participants linked the (perceived) increase in other crimes, including gun or knife crime and theft, to a reduced sense of safety. An increased awareness of crime influenced the perception that VAWG and other crimes had increased, through media coverage, social media and by word of mouth. Some participants also raised concerns about police response, for example police not helping when incidents occurred.

The negative causal mechanisms especially help to explain the absence of high-level impact, which can at least partly be attributed to these negative factors that influenced perceptions of safety during the intervention period. These factors could mitigate or slow any positive effects.

4.4 Qualitative evidence of the impact of SWaN interventions

4.4.1 Women in public spaces

Although the quantitative impact analysis suggests that SWaN interventions did not have a significant impact on improving perceptions of safety overall, QuIP research suggested that SWaN interventions can be viable mechanisms for improving perceptions of safety. Specifically, where positive change existed, some QuIP participants directly attributed this to the types of interventions funded by SWaN, including interventions relating to the police and those taking place in NTE venues or among NTE workers.

A. Increased security and procedures at NTE venues

Some QuIP participants linked increased security at NTE venues to improved feelings of safety inside these spaces. These accounts included references to more security guards and more thorough screening on entry (for example, bag checks and the use of metal detectors). When cross-referenced with bid review data collected during the scoping phase, each of these participants were sampled from a funded area with interventions related to security inside NTE venues. As such, it is possible these were direct references to interventions funded through SWaN. Some participants believed that increased security at NTE venues was in direct response to a recent increase in drink-spiking events, with one participant claiming that spiking incidents had already declined because of these interventions.

Furthermore, many of the 23 participants attributed positive changes in their perceptions of safety to an improved environment through measures such as CCTV and improved lighting within a venue. However, since environmental measures were not eligible for SWaN funding, these cannot be attributed to SWaN.

B. Increased police presence

A few QuIP participants linked increased police presence to improvements in their feelings of safety over the intervention period, specifically on the high street at night. Although, just one of those participants was from a funded area that delivered interventions related to police presence through SWaN.

Overall, however, police presence as an intervention elicited a mixed response regarding its effects on perceptions of safety. For example, one participant had noticed an increase in police presence but stated that this had not improved or reduced their perceptions of safety.

C. Other interventions

QuIP participants also linked other interventions to improved feelings of safety, although only one participant mentioned each one. A visual representation of all the interventions influencing positive changes in perceptions of safety is provided in Appendix D. Except for street lighting, all interventions referenced were reported by participants sampled from areas that funded the relevant activity through SWaN. These interventions included measures to prevent drink spiking (drink toppers and test kits), NTE staff training, and the ‘Ask for Angela’ campaign.

4.4.2 NTE workers’ views on sources of change

As with findings from the F2F survey, QuIP analysis was used to understand the factors NTE workers attributed to their perceived changes in the safety of women and girls over the intervention period. In total, 112 responses were collected from the open-text questions included in the email survey.

QuIP responses linked improved safety for women and girls to SWaN interventions, often in relation to the intervention types where NTE workers were direct beneficiaries of SWaN funding. This included reference to NTE personnel training or capacity building, improved venue procedures, and police resourcing or actions as drivers of change. Specific examples of interventions most relevant to SWaN and mentioned by NTE workers included VAWG training, employment of additional bouncers, and the Ask for Angela initiative. QuIP analysis revealed that some NTE workers felt these interventions had increased awareness of VAWG, confidence in reporting VAWG, and improved the capacity to respond to VAWG within the NTE. For example, these interventions helped staff to recognise potentially dangerous situations and to take necessary action in response to incidents.

“The staff are more alert to possible dangers and are in a better position to identify possible risks and take action should there be need.” (NTE venue – venue management)

Aside from training interventions, some responses also linked an increase in venue security through the employment of bouncers to increased safety inside venues for staff and customers.

“We have increased security. Introduced anti-spiking measures. Have greatly reduced spiking reports.” (NTE venue – venue management)

QuIP responses also linked a range of other factors to positive changes in the perceptions of safety for women and girls. These included environmental factors (like CCTV and lighting), increased police presence, and increased public awareness of VAWG more generally.

References to the perception that the safety of women and girls had decreased over the intervention period were limited in the QuIP research with NTE workers. Where mentioned, sources of negative change were most often linked to wider contextual factors. This included an increase in VAWG incidents, with specific references to harassment, assault and spiking incidents. Other factors mentioned were the presence of predatory groups, and fewer police and other staff members available to help in certain areas.

“An increase in spiking situations in my local area (primarily at night near bars/restaurants) makes me feel it’s less safe for women and girls.” (Woman working in local government)

“There seems to have been an increase in crimes, especially at night within the area in which I work. A lot of these crimes are directed towards woman or lone people. I think this is largely down to the lack of police presence that now exists in the NTE due to sheer lack of numbers. There are far fewer police officers available at any time, especially during nights, therefore there is bound to be an impact on the perception of safety.” (Male, policing)

The respondents’ sector appears to influence these perceptions. For example, police respondents were, in general, much more likely than respondents in other NTE roles to report negative perceptions of safety (regardless of whether a change in perception was identified), whereas virtually all other sectors were more likely to report positive perceptions of safety as compared to police respondents.

5. Process evaluation findings

This section presents the results from the process evaluation, drawing on the various strands of the process methodology (outlined in Section 2.5) to understand the implementation experience at each stage of the SWaN programme. It details perceptions of the bid development process, the key facilitators and barriers to successful delivery of interventions, perceived effects of the interventions, and key learnings around sustainability of interventions and their impact.

5.1 Process evaluation summary

Key findings from the bid design stage are:

• funded area teams felt that SWaN provided the opportunity to trial creative approaches to addressing VAWG and fill gaps in local resources
• teams praised the funding eligibility criteria, which allowed a wider group of organisations to take the lead applicant role compared to other similar Home Office funds
• some teams found the short commissioning timescales a barrier to building new partnerships or gathering relevant data for planning

There are also several overarching and intervention-specific lessons from the delivery stage:

• effective governance and partnership working was essential to support successful delivery in tight timescales
• good collaboration between agencies was important to enable effective delivery; for example, strong links between NTE workers and those delivering guardianship interventions meant vulnerable women could be flagged to receive help
• training on VAWG issues provided through SWaN helped to fill a knowledge gap among NTE workers, meaning they were better equipped to support women
• interviewees expressed that interventions that target perpetrators and potential perpetrators by tackling perceptions of VAWG are important for breaking the cycle of vulnerability in the NTE

Although the impact evaluation detected limited statistically significant effects associated with SWaN interventions, the process evaluation highlighted that:

• bid teams felt that there are likely to be positive effects taking place in their local area but agreed that it will take more time for longer-term effects to become visible
• although the sustainability of many interventions depends on future funding, where possible bid teams designed their interventions so their benefits extend beyond the funding period, for example through investment in technology
• for some teams, the funding increased awareness of VAWG issues, which led to the ability to leverage further funding

These findings are discussed in further detail below.

5.2. Bid development process

For Section 5.2, it is relevant to note that only successful bids were included in the process evaluation research. This means the findings presented below reflect the bid development experiences and characteristics of successful areas only.

5.2.1 Rationale for bidding

Many stakeholders saw SWaN as an opportunity to address specific aspects of existing local VAWG strategies by providing the financial resource required. It also complemented activities being delivered through other funds, including the SSF. In some cases, SWaN offered the opportunity to build upon or add capacity to existing interventions in an area, such as street pastors.

The characteristics of a bid’s area was a key consideration when deciding to apply for funding. Most bid teams described having a large student population or proportion of young NTE visitors, and therefore a disproportionate population of potential VAWG victims. In addition, bid teams perceived urban NTEs had become less safe after the COVID-19 pandemic, reporting, for example, that people were drinking earlier in the day. Bid teams linked this to increased levels of ASB and violence in these areas and felt that SWaN provided timely funding to help address these issues.

The aims of the Fund and the types of interventions that were eligible for funding also influenced decisions to bid. Specific examples given included policing interventions, drink-spiking interventions, and initiatives to support the safety of sex workers. These interventions were outside the scope of other similar Home Office funds, and bid teams valued having the opportunity to fund interventions specific to issues in the NTE.

“We’ve had an increase in reporting of drink spiking, things very specific to the night-time economy that were not covered by Safer Streets 3 … we didn’t have the resources to target the night-time economy.” (Funded area lead)

5.2.2 Determining the focus of bids

Areas primarily used existing data and findings from consultation activities with professional stakeholders (such as the police, local authorities and relevant VCSE organisations) and the public to help determine the focus of their bid.

5.2.3 The use of existing data in SWaN bids

Bid teams used police and local authority data on VAWG crimes, ASB and their hotspots in the NTE to identify the key locations for intervention delivery. In some areas, analysts mapped potential victim and perpetrator profiles to highlight areas of need. Bid teams reported that this data-led approach helped build consensus among the multiple agencies involved in bid design. However, some felt that the tight timeframes negatively impacted their use of data, citing challenges related to data access. This meant they were unable to analyse the data obtained to the extent desired.

5.2.4 Consultation and partnerships with professionals

Bid teams felt that engaging a variety of external stakeholders was an important part of bid design. Networks, including Community Safety Partnerships, provided a forum for organisations to discuss initial ideas and identify potential bid and delivery partnerships. Collecting anecdotal feedback from key stakeholders, such as local police, on their experiences of women’s safety in the NTE also helped to inform bids. Utilising these existing relationships helped some areas to design their bid more efficiently, a key facilitator in bid design considering the time pressures.

Bid teams also felt that the Home Office requirement that bidders consult with VAWG organisations was useful for effective bid development. VAWG organisations helped to fill evidence gaps in PRC data, particularly due to the high under-reporting of VAWG. Participants highlighted a range of reasons for the under-reporting of VAWG, including the perception among women and girls that the police would not take their experience seriously, and victims doubting their own judgement and engaging in self-blame, especially if they were under the influence of alcohol at the time of the incident.

Participants said the “meaningful coproduction” which arose from partnerships added value to proposals. This is highlighted in Box 5.1.

Box 5.1: Working with specialist VAWG partners

One funded area developed an intervention to improve the safety of sex workers. A VAWG organisation supported the intervention design by providing information on the common locations for sex workers in the area and the increased risk of exploitation and VAWG in those locations. The VAWG organisation also shared key survey data which were used to inform intervention design. The bid team felt this was a positive process, as it meant plans reflected the views of those with lived experience of the sex industry.

“Although they are smaller sample sizes, they are voices that would not have been heard through police data.” (Funded area lead)

However, some bid teams found engaging new VAWG organisations in delivery plans challenging when conflicting views arose between VAWG organisations and others, including the police and venue representatives. For example, VAWG organisations in one funded area were concerned that police involvement could discourage engagement by target groups due to their perceived distrust in the police. Bid teams often found it difficult to resolve any conflicts in the short timeframe.

“The conflict between agendas was a challenging tightrope to walk.” (Funded area lead)

5.2.5 Consultation and engagement with the public

Many bid teams used surveys to gather the perspectives of women and VAWG victims. They felt that it was important to capture their perspectives during the bid development stage and survey findings were used to influence bid design, including the types of interventions delivered. Some areas also used surveys to explore hotspots of VAWG incidents, cross-referencing survey responses with police data to ensure interventions were targeted at the most appropriate locations.

These surveys often generated much higher response rates than anticipated. Bid teams felt that this reflected the high degree of concern about VAWG among the public, although often noted that generating a successful response was resource intensive and effective marketing of surveys was important. Drawing upon findings from pre-existing local surveys helped some areas to save time and resources.

5.2.6 Designing interventions to support NTE users

The interventions delivered within a funded area were often designed to offer a cohesive package of support to NTE users, with different strands targeting different spaces and populations. These packages of interventions were premised on the idea of providing ‘wrap-around support’:

“[The idea is to] look at different aspects of the night and economy and get the strands of keeping people safe in venues, keeping people safe out and about in the street, and having a safe place for people to go. And then for professionals, who maybe deal with people in the aftermath of an offence, to know how to ask the questions and respond appropriately.” (Funded area lead)

Guardianship, policing and transport interventions often formed an area’s support package for NTE users. These aimed to improve women’s safety by increasing the presence of both authoritative and non-authoritative figures, helping to reassure women and deter potential perpetrators. Improving perceptions of safety was also a priority, and areas often delivered communications and awareness-raising interventions alongside safe space and guardianship initiatives. These aimed to provide assurance to women that procedures were in place to support and protect them.

Most funded areas also included interventions that aimed to make the legislative and regulatory structures around the NTE more responsible for women’s safety in their bid design. Examples included safe venue accreditation schemes, safety charters, and delivering training to NTE workers. Overall, funded areas wanted to create an NTE culture in which everyone has a duty to recognise and respond to VAWG.

5.2.7 Designing interventions to drive societal change

For many bid teams, an overarching goal of achieving wider long-term cultural and behaviour change also shaped their choice of interventions. Some areas delivered educational interventions outside of the NTE, such as in schools and universities, and perpetrator-focused interventions. These aimed to increase awareness of VAWG, challenge inappropriate behaviour, and create a cultural shift away from victim-blaming.

5.3 Delivery of SWaN interventions

5.3.1 Overarching facilitators

Effective collaboration between multiple agencies was a key facilitator for the delivery of interventions. Adopting this multidisciplinary approach ensured a more cohesive and joined-up response to addressing VAWG in the NTE, with good practice examples including working with local licensing teams and the development of NTE stakeholder groups. One area described how they had used SWaN to “coordinate a community response to VAWG, making all stakeholders part of the solution” (Fund area lead).

Using existing structures and partnerships helped to deliver the interventions efficiently and reduce duplication, although it was important that partner organisations shared the same vision.

“Collaboration has been good, 3 different organisations collaborating together for people that need it. I think that is quite powerful.” (Delivery staff)

Within partnership working, effective governance and project management were important to successful delivery. Having strategic and senior support helped to support quick mobilisation of the funding. Effective partnership working was also facilitated by organised and timely project planning, particularly considering the short timescale. Having regular partnership and project board meetings created a space for bid and delivery teams to review progress and discuss any challenges. Some areas also reported that having staff for specific roles, including for project management and administrative activities, supported effective delivery.

Bid teams felt that press coverage of activity facilitated awareness of interventions and engagement with them. One funded area included a marketing manager within the bid team to ensure full utilisation of the media and reflected that this was one of the most effective aspects of their project.

Several teams used ongoing data analysis and police intelligence to inform their interventions throughout the delivery period. This was particularly relevant to identifying hotspot areas for policing interventions, and high-risk potential perpetrators for interventions with female sex workers.

5.3.2 Overarching challenges

Tight timescales were a key challenge experienced by bid areas at each stage of SWaN, including the delivery stage. Issues related to procurement rules for delivery organisations in the public sector (such as local authorities), and IT security and data protection for interventions using technology, were difficult to overcome in the short delivery timeframe. Areas also found it challenging to secure VAWG organisations as delivery partners, as these organisations were often hesitant due to the short-term nature of the Fund. It was also difficult to implement any learnings that emerged during delivery. Areas, therefore, found it difficult to balance delivering interventions quickly and to a high-quality and felt that a longer delivery period would have been beneficial.

“We were able to deliver but not always in the way that we’d intended.” (Funded area lead)

“The timeline makes it reactive rather than proactive – we wanted to make a prolonged impact.” (Delivery partner)

Some areas experienced difficulties in engaging partners and local businesses in interventions, a key barrier to their delivery. For example, some areas found it difficult to secure participation in interventions by licensed premises due to financial concerns or lack of clarity around the implications of an intervention on their organisation. Resistance was also reported from venues working in niche parts of the NTE sector (for example, venues involved in sex-related work, such as strip clubs).

Some areas found the Home Office requirement of a monthly monitoring return (a form providing details on a bid’s spend and progress in delivery) an administrative burden. This was particularly important considering the tight timescales, and because some delivery teams included unpaid voluntary roles. The preventative nature of interventions also meant areas found it difficult to quantify any early results.

The COVID-19 pandemic also presented a challenge for bid teams as restrictions were still in place during some of the delivery period. In Wales, this included restrictions on indoor activities until the end of January 2022. Areas felt that the combination of COVID-19 restrictions and a winter delivery period meant the NTE did not represent the level of activity typically seen in this space, ultimately limiting the reach and impact of interventions.

In addition, for some interventions, pressures caused by the COVID-19 pandemic meant police and ambulance resources initially planned for SWaN interventions, such as Safe Havens, had to be redeployed to business-as-usual activities. For these reasons, areas felt that interventions may have seen more activity, engagement and greater impact had interventions been delivered in the summer months.

Discussions on the ToCs during the case study research highlighted that these challenges led to funded areas changing their delivery plans. For example, some areas redirected funding from interventions where it was clear that successful delivery could not be achieved within the funding period or due to changes in local need, to increase the scale of other interventions and maximise use of the funding.

5.4 Learning from different categories of SWaN interventions

5.4.1 Interventions targeting NTE venues and workers

SWaN funded 30 interventions focused on NTE venues and workers. Most areas commissioned training interventions to improve the knowledge and skills of NTE workers in recognising and responding to VAWG. Other interventions included venue accreditation schemes, awareness campaigns, and the provision of equipment such as drink-spiking test kits.

A. Educational interventions

Tailoring the content of educational interventions to specific issues faced in a particular venue or locality was especially beneficial. Areas valued the opportunity to create bespoke training packages, and the training was often delivered to a broad range of NTE workers, including DJs and promoters. This helped to support a more holistic understanding of VAWG across NTE workers. However, some areas encountered challenges in engaging NTE venues and workers due to the resource associated with participating in training. Although funding covered the direct costs associated with training provision, attendance required staff to be absent from their normal duties, acting as a barrier in some cases. Despite initial challenges in engagement, positive feedback was received from training participants. In one area, 90 individuals were still on the waiting list for training when funding ended.

B. Schemes and accreditations

Seven areas used funding to develop and implement venue accreditation schemes across their NTEs. These schemes promote safety inside NTE venues by ensuring that participating venues meet certain standards relating to customer safety and staff competency.

The flexibility to use funding to help venues implement policies needed to achieve accreditation was a key facilitator to successful delivery of these interventions. The provision of free equipment, materials and templates also supported engagement from venues and encouraged sign-up to schemes. Although the provision of free resources could be costly to delivery teams, these actions prompted conversations with venues around VAWG. Some areas reported positive feedback from venues on being part of the scheme, and one area suggested that a commitment to women’s safety in venues should be a mandatory licensing condition. Accreditation approaches could help fill the gap in licensing conditions to help ensure that women’s safety is also a primary concern of NTE venues.

“For most venues, women’s safety is the gap that they have got. Like they have all the age verification, drugs policies, things like that.” (Delivery staff)

5.4.2 Interventions to improve public perceptions and awareness of VAWG

Communication campaigns and training programmes were used to improve awareness of VAWG among the public. They also aimed to provide reassurance by “making communities aware that women’s safety is taken seriously” (Funded area lead). The 2 main approaches in this category were interventions working directly with perpetrators to change their behaviour, and universal interventions focused on bringing about attitudinal changes among the general public.

A. Perpetrator-focused interventions

Delivering perpetrator-focused interventions was seen as crucial in creating a cultural change in acceptable behaviour within the NTE. The format of these interventions varied between areas. Some areas adopted mild approaches using short animations to educate potential perpetrators about VAWG. Others adopted more intense perpetrator programmes using behaviour change courses but struggled to engage perpetrators and secure referrals due to the voluntary nature of these interventions. More detail on this is in Box 5.2.

Box 5.2: Preventing VAWG with additional police resources and perpetrator-focused activity

One area funded an intervention that involved deploying additional police officers into the NTE. This intervention comprised deploying 2 plain-clothed officers to identify (potential) perpetrators and alert two uniformed officers, who could then intervene. Although implementing this intervention was difficult due to the impact of the COVID-19 pandemic on police resources, monitoring data showed some promising results. Police officers involved in this intervention intervened in 498 concerning situations during the funding period, resulting in 55 arrests.

This area also implemented a perpetrator-focused intervention, referring potential perpetrators identified through the policing intervention to this behaviour change initiative. The voluntary nature of this intervention and the communication of the intervention’s aims to potential perpetrators meant that this area reported no successful referrals.

“We didn’t get any referrals, if I’m honest, and there were no self-referrals to it at all. I think that probably comes from the way that it’s worded, the way that it is put across to people. I can’t see any man going, ‘oh yes I am potentially a sexual predator.’ They would always think that isn’t them. I don’t know whether the actual scheme itself needs to be reworded or something to get the benefit from it because when we are in the situation where we’re mentioning it the response is always ‘no, I’m not like that. I’m not getting involved in that.’ I think it was always going to be a hard sell.” (Operational lead)

B. Universal approaches to awareness-raising

Awareness-raising interventions, such as communication campaigns and training interventions with the general public, were developed to challenge societal attitudes around acceptable behaviour towards women in the NTE and more widely. Areas reported promising responses to awareness-raising interventions, including from male audiences. Combining innovative approaches to awareness-raising, such as virtual reality-based training and art installations, with media campaigns to deliver the same message helped to generate good discussion and online interaction on VAWG issues.

5.4.3 Guardianship interventions

Areas funded guardianship interventions to improve women’s safety in public spaces outside of NTE venues. Examples included street pastors, the development of safe spaces, and transport marshals. Policing interventions involving additional and dedicated police resource for VAWG were also included in this intervention category.

Areas generally reported high levels of engagement with guardianship interventions, with these services used by women experiencing various levels of vulnerability. The success of these interventions was facilitated by effective collaboration and strong communication between the delivery staff and other NTE workers, such as venue and security staff. This helped to ensure that vulnerable women received the necessary help efficiently, and reassured NTE workers that the women they had referred were safe.

However, resourcing was the primary challenge for guardianship interventions. Paid guardianship roles were costly to bid areas, limiting the sustainability of these interventions as their continuation depended on additional funding. In contrast, areas using voluntary guardianship roles had some distinct resourcing challenges typically experienced by the voluntary sector. These were exacerbated by COVID-19 restrictions, which led to some volunteers choosing to isolate.

Additionally, most policing interventions were resourced by overtime. This was a key challenge to successful delivery, as areas often found it difficult to find the personnel resource needed to cover shifts funded through SWaN. For example, some areas reported a feeling of burnout among police officers resulting from working additional shifts to cover sickness caused by COVID-19. This meant fewer officers were available for overtime. Areas also reported that the negative perceptions associated with policing the NTE, including working unsociable hours, meant some areas struggled to resource officers for overtime.

Box 5.3: Supporting particularly vulnerable women at night

One area created an outreach intervention to support women at high risk of violence and exploitation in the NTE, such as female sex workers and homeless women. Outreach workers engaged in a range of activities, including the provision of physical resources to these women, such as rape alarms and drink testing kits, signposting to support services, and working with police to identify potential perpetrators. They also provided mobile phones with a pre-installed personal safety app, with one participant expressing that “those phones have been a lifeline for those women, who’ve been on the street who’ve had their phones stolen” (Delivery staff).

Ensuring visibility and consistency in the availability of outreach workers helped to build trust with vulnerable women, a key facilitator to the success of this intervention. A multi-agency approach also helped to build strong relationships between outreach workers, local support services and the police.

“We talk about having those kind of ‘hard-to-reach cohorts’, but actually it’s services that are hard to reach. So, we kind of break down those barriers… we [become] known to be trusted, and then they’re happy to engage.” (Funded area lead)

Areas valued the inclusion of specific groups of vulnerable women within the Fund’s scope, noting that specialist funding is often required for work with these groups. The early success of this intervention meant additional funding was secured, ensuring that the relationships built through SWaN could be sustained.

5.4.4 Interventions to improve personal safety

Some areas developed new, or implemented existing, technologies to improve the safety of women, particularly during their travel into or from the NTE. Examples include mobile apps containing safety features, and physical interventions in NTE areas that allowed women to alert authorities when feeling at risk (such as by adding buttons linked to the CCTV control room to street furniture). Although these interventions provide examples of innovative approaches to VAWG prevention, funding was primarily used to develop these technologies. The launch of these technologies after the funding period ended meant the lessons learned on the delivery of these interventions was limited, although areas suggested that early engagement data shows promising results.

Box 5.4: Improved emergency call handling with GoodSAM

One area funded enhanced call handling in their police contact centre through technology called ‘GoodSam’. This technology automatically converts phone conversations to video so call handlers have direct visual access to the caller’s surroundings and uses geo tracking to identify and deploy officers to the caller’s location quickly.

Engaging with other police forces using this technology and learning from their experiences helped to ensure a successful set up and the mitigation of challenges related to legalities and data management.

Effective training was also seen as crucial to ensure staff could maximise the use of this technology, although it was difficult to arrange training around shift patterns. Despite this challenge, the technology was successfully set up, and all control room staff were trained before the end of the funding period (by mid-March 2022).

Feedback from control room staff suggests they found the technology helpful in locating at-risk individuals quickly. Monitoring data also suggests a positive response to the technology with an average of 70 uses per week, facilitated by the range of advertising approaches used to raise awareness and encourage use among the public. This included sharing information in local support groups for women. Based on this success, this area aims to broaden use of GoodSam to include domestic abuse incidents.

5.4.5 Other types of interventions

One area funded the development of new women’s safety champion roles within the police. These roles aimed to ensure women’s safety was a priority in policing the NTE, and to encourage the recruitment of female police officers, particularly into leadership roles. Stakeholders reported that their “high-visibility” approach, comprising a communications campaign and active outreach work by the champions, received positive responses from local women.

“To be able to deploy those [women’s safety champions] specifically around where we know women don’t feel safe, and to have specific conversations about women’s safety is something brand new. And I’m just really proud of that.” (Funded area lead)

5.5 Perceived effects of SWaN interventions

The effects presented in this section are based on the perceptions of the bid teams, rather than representing formal impact measurement. A summary of this evaluation’s assessment of SWaN’s impact is in Section 4.1 of this report.

Areas considered it was too early to see clear evidence of impact, but felt their interventions had positively affected target audiences. Monitoring data and anecdotal evidence referenced by bid teams suggests that there are likely to be positive effects occurring at the local level, but evidence of impact will take time to emerge. These findings align with those of the impact evaluation, which suggests sufficient time will be required to change perceptions of safety (Section 4.3).

Even where the delivery of interventions had deviated from plans related to their ToCs, areas felt they were still on track to achieve their planned effects and, in turn, their longer-term outcomes. This was primarily because they had redirected funds to other interventions with the same intended effects.

5.5.1 Effects on the general public

Anecdotal evidence suggests that early effects of some interventions were starting to be seen amongst the general public. For example, feedback on social media indicated that campaigns were generating discussions around VAWG and was helpful for bid teams in gauging the reach of their campaigns. Areas participating in the case study research also described how SWaN provided an opportunity to start a dialogue in the community and begin to challenge existing attitudes and beliefs.

Some areas described positive effects of interventions focused on behaviour and attitude change in schools and universities. One school participating in awareness-raising activities reported early signs of cultural change in the language used by students, particularly around what constitutes VAWG. As a result, the school introduced compulsory training for new teachers and planned to seek further funding to embed the progress made.

5.5.2 Effects on NTE users

Areas perceived NTE users to be the greatest beneficiaries of SWaN, with the increased presence of officers, street teams or support workers helping to reassure NTE users and reduce the fear of being left alone. Areas also reported an improvement in feelings of safety resulting from these interventions amongst particular groups. For example, capable guardians facilitating journeys home was perceived to have been reassuring for students.

“I suppose, just the general perception of having that guardianship figure, having somebody who actually does care, not necessarily someone in authority, but just somebody to go to if you’ve got a bit of trouble. I think that’s been very valuable.” (Funded area lead)

Some areas reported a decline in the number of sexual assaults and violent incidents occurring over weekends in city centres. Although areas were cautious about directly attributing reductions to SWaN at this stage, some interventions had monitoring data to support this. For example, preliminary data for one preventative policing intervention showed that 163 women had been safeguarded, 3 arrests had been made for sexual assaults, and one attempted rape had been prevented. Police respondents in another area reported arresting known curb-crawlers and preventing a vulnerable woman being removed from a NTE venue by a stranger without her consent. Areas felt the increased presence of police meant officers were available at the right time and place to support vulnerable women.

Anecdotal evidence also suggests wider positive effects on specific groups of vulnerable women. Benefits include signposting to services, such as women’s outreach and health and substance misuse services, and additional support to groups that other similar funds had overlooked. For example, one woman attended a GP appointment for the first time in 3 years, and another restarted engagement with substance misuse services for the first time in a year, following engagement with an outreach intervention.

5.5.3 Effects on NTE workers

For NTE workers, immediate effects stemmed from the training they received through SWaN interventions. Participants reported feeling better equipped to intervene in incidents of VAWG occurring in venues. For example, NTE staff in one area were trained to respond discreetly to Ask for Angela incidents, provide better support to NTE users and challenge certain behaviours before they escalated. In one area, police, CCTV staff and security personnel anecdotally reported a displacement of VAWG incidents from the streets to inside venues. However, positively, venue staff had also received a higher number of reports and felt that victims’ confidence in reporting incidents to them had increased.

Feedback from training participants also highlighted that NTE workers are often victims of harassment themselves and that learning techniques like non-confrontational communication helped them to feel more comfortable in challenging problematic behaviour. These findings reflect those of the impact evaluation, which showed positive impacts on NTE workers’ self-reported confidence in supporting victims of VAWG and dealing with VAWG at their place of work or study (see Section 4.2.3).

Box 5.6: Effects of tackling drink spiking for NTE users and workers

One area developed interventions to prevent drink spiking through the provision of drink-spiking kits to NTE venues and education packs for NTE staff. This area felt that the increased media coverage of drink spiking was having a negative effect on local NTE venues, and wanted to support these businesses. In total, this area reported 164 venues receiving drink-spiking kits and that several arrests had been made for drink spiking since their distribution.

This area also developed an awareness campaign to improve understanding of spiking amongst the public and NTE workers. Having the police lead this work was seen as crucial to ensure consistent messaging on drink spiking between NTE venues, NTE staff, victims and police officers. This also meant learning could be shared with other police forces.

“Through that campaign we had a big success. Many venues took this on and really got involved, and it hit national media. It was a different way of doing that campaign, and doing the posters, and working together with partners – not working in silos.” (Delivery partner)

Participants also found it appealing that the area’s overall approach focused on addressing perpetrator behaviour, rather than focusing on victim behaviour change. The ready availability of test kits within NTE venues allowed for instant testing and specifically helped to place the focus on perpetrators’ actions.

Overall, a positive response from NTE venues was reported, with support for intervention to continue beyond the funding period. Venues also suggested developing an accreditation scheme for venues using test kits to demonstrate their commitment to customer safety.

However, more widely, a key challenge in addressing the issue of spiking was the negative attitudes held by some NTE staff across funded areas in relation to victims of drink spiking and their authenticity. A few NTE workers felt that drink spiking is not as prevalent as the media portrays, and that women may use spiking to excuse drunkenness. This narrative perpetuates victim-blaming and highlights the importance of interventions to change underlying attitudes and behaviours towards women.

5.5.4 Effects on police and other emergency services

Funding enabled police to provide dedicated resource for policing VAWG in the NTE. This helped to generate new intelligence on crime hotspots and incidents.

“On the intelligence gathering side of things, [the intervention] meant that we were coming across stuff that we probably wouldn’t have normally come across just because we were in the right place at the right time.” (Operational lead)

Some police officers were also offered a new plain-clothes training course, which focused specifically on how to recognise and intervene when someone is displaying predatory behaviour. Prior to SWaN, full surveillance training was the only training available to officers working in plain clothes. However, full surveillance training is expensive and not tailored to identifying potential predatory behaviour in the NTE. The new plain-clothes training course was well-received by participants and offered a more tailored and cost-effective approach to improving preventative policing.

“We’ve got nearly 100 officers trained. I’ve had some really good positive feedback, which we did, feedback forms with them [sic] and they have said that it’s really helped cement some of their understanding around predatory behaviour.” (Funded area lead)

Areas reported some interventions helped to reduce pressure on emergency services, including ambulance services and the police. Effective collaboration between interventions and the emergency services enabled the diversion of resources within the NTE. For example, a Safe Place intervention in one area saw 13 referrals, who would have previously been sent to A&E. With support from the Safe Place, only 2 people required further medical attention. Some areas funded additional ambulance services or independent medical organisations, which also helped to reduce pressure on core emergency services.

5.5.5 Effects for local organisations

Local organisations benefitted from the opportunity to work collaboratively and valued the local partnerships that were created.

“There are big projects within this but using local firms and local people really promoted our safer streets. That word of mouth and why we are doing it, really helped people to understand why. Virtually people could see the difference of the town, but it made us capitalise on what worked locally. Partnerships worked well and emphasised good working relationships in the town.” (Operational lead)

Collaboration between local organisations led to greater capacity and a wider reach of services and, for several interventions, joining up local provision was key. For instance, one funded area set up a hub for all volunteers in the NTE area, including mental health triage teams, street pastors and police. This hub facilitated services working together, resulting in a more joined-up approach and more efficient delivery.

One organisation delivering women’s outreach services felt that SWaN increased their visibility in the local area, by offering them the chance to be the lead organisation in their local partnership. Delivery staff in the area also reflected that SWaN improved their relationships with the local council, shelter and accommodation providers, the YMCA (a third-sector organisation), the police, and local businesses near NTE hotspots. One area noted that the partnerships formed through SWaN would be sustained and allow a collaborative effort to achieving longer-term outcomes.

“Collaboration has been good, three different organisations collaborating together for people that need it. I think that is quite powerful.” (Delivery staff)

5.6 Sustainability

Areas designed interventions to support their sustainability beyond the funding period. This section explores how sustainability was embedded in the design of interventions, building on lessons learned from the design and delivery stages.

Most areas expanded existing activities or developed new materials that could be reused to ensure the sustainability of interventions, such as training and awareness campaigns. Areas felt that designing interventions in this way meant that their benefits could extend beyond the funding period.

“We had the ability to create a product that could be sustainable and add value beyond the funding period. So I think it is a more effective investment of money because if we had been asked to deliver training sessions, we would only be able to actually do the training during the funding period, then the money is gone.” (Delivery staff)

In addition, some areas used funding to develop new technologies that, once set up, required minimal ongoing investment and could be used long-term.

However, several areas reflected that securing additional funding was a key factor in ensuring the sustainability of many interventions. Many areas reported that securing funding from the SSF Round 4 was a priority.

“If we get that, it gives us an element of guaranteed funding until September 2023. … as part of our sort of semi-matched funding to that, we still manage the coordination, we still run the training, we’ll still do the comms campaign, but the bigger cost is continuation of the hub obviously because of the staffing and accommodation costs.” (Funded area lead)^

Many areas reported that they planned to continue monitoring the effect of ongoing SWaN activities while applying for funding to support longer-term delivery. Others also reflected that having long-term staffing in place was necessary to sustain any early effects of interventions.

Several area leads described SWaN as providing proof of concept for activities tackling VAWG and safety in the NTE. Some were able to secure additional investment to continue delivery beyond the funding period resulting from increased awareness of issues related to women’s safety through SWaN. For example, one area was using evidence from their SWaN intervention to make a case with their local PCC to mainstream the activities. In another funded area, the police force now has 6 women’s safety champions heading up 72 uniformed officers who are patrolling 24 hours a day, as a result of SWaN. The champions will remain in post beyond the lifetime of SWaN funding, continuing the provision of training and awareness-raising activities for other officers and the general public.

Overall, SWaN provided an opportunity to launch work on improving women’s safety in the NTE and, in some cases, longer-term delivery.

6. Conclusions

This final section brings together the overall findings from all elements of the evaluation and summarises the implications for future funding and evaluation design.

6.1 Lessons learned on the impact of SWaN

Overall, the evaluation found limited evidence of statistically significant impacts from SWaN. While few statistically significant effects were found at Fund level, this does not necessarily mean that the programme and its interventions were unsuccessful. The lack of impact detected could be explained by a variety of factors, including intervention dosage, implementation timeframes, external influences on perceptions of safety, and evaluation reference periods.

This evaluation detected no statistically significant impacts to suggest that SWaN improved women’s perceptions and feelings of safety in public spaces at night (Section 4.2.1). Instead, some of the evaluation evidence indicates that some SWaN interventions inside venues could have had a counterproductive effect on feelings of safety. Although stakeholders in funded areas provided qualitative evidence that SWaN could have had small, positive effects on feelings of safety in the NTE, there was consensus that it was too early to detect significant impacts (Section 5.5).

The evaluation did, however, find evidence to suggest that SWaN interventions led to improvements in NTE worker capabilities. NTE workers who received training and education interventions reported statistically significant improvements in their awareness and understanding of VAWG, and their self-reported confidence in supporting VAWG victims (Section 4.2.3). This was supported by qualitative evidence from NTE workers, collected through the process evaluation, where NTE workers also reported that they felt better supported by colleagues when experiencing VAWG incidents themselves, such as harassment, while working (Section 5.5.3). These findings were reinforced by the mini-QuIP analysis, where NTE workers directly linked SWaN interventions to positive changes in perceptions of the safety of others. Training in particular was felt to improve capability to identify those at risk and take appropriate actions (Section 4.4.2). While these impacts on NTE workers are positive, the benefits of these interventions were not detected among users of public spaces.

The overall lack of impact on feelings and perceptions of safety can be explained through various possible reasons. Evidence suggests that achieving a significant improvement may not have been possible within the Fund’s scope and its delivery timeframe (December 2021 to March 2022). Awareness and use of the types of interventions funded by SWaN among users of public spaces was generally low. The QuIP analysis also highlighted the range of complex, negative factors that can influence individual perceptions of safety, and the factors that respondents directly linked to negative changes in feelings of safety over the intervention period. These factors included personal experiences of crime, including VAWG, and awareness of local or national crime incidents. Together, this evidence suggests that it will be particularly difficult to change perceptions of safety quickly and at a level to detect significant impact or outweigh the effects of negative factors. To achieve this, more time and investment is required.

6.2 Lessons learned on the design and delivery of SWaN projects

The process evaluation highlighted several key learnings from funded areas that are useful for any future programmes designing and delivering similar interventions.

6.2.1 Bidding process

• using data is helpful for determining the focus of bids and building consensus between the multiple agencies often involved in bid design; however, areas should be mindful of challenges in accessing data quickly
• leveraging existing networks helps to engage a range of local stakeholders and design bids more efficiently
• where possible, drawing on findings from pre-existing, local surveys to inform bids helps to save time and resources

6.2.2 Delivery process

• effective partnership working between a range of agencies creates a more cohesive and holistic approach to VAWG prevention
• strong project management helps to overcome challenges quickly to ensure successful delivery
• utilising a range of media tools helps to encourage engagement and awareness of interventions

6.2.3 Selected intervention-specific learnings

• the content of educational interventions for NTE workers should be tailored to address the issues most relevant to that locality or venue
• the time required for staff to attend training and be absent from their normal duties should be carefully considered when implementing training interventions with NTE workers
• engaging with stakeholders who have implemented similar interventions helps to avoid challenges previously experienced, particularly for innovative interventions

6.3. Overall implications and recommendations

The evaluation has highlighted a clear set of implications for future funding rounds and evaluation.

6.3.1 Implications and recommendations for the Home Office for future funds

1. Engage in early advertising to maximise bid preparation time.
2. Increase support to bidders. Stakeholders requested more opportunities to receive information from, and ask questions of, the Home Office, such as through learning events and workshops.
3. Give greater consideration to timescales at each stage of the project. An extended bidding window would help areas to engage key stakeholders effectively and improve bid quality. Although achieving impact will be difficult, increasing the delivery period would increase the likelihood of achieving and detecting significant impacts.
4. Focus on increasing the dosage (that is, the overall investment) and the sustainability of interventions to ensure they have sufficient scale and longevity to achieve lasting change. Some areas suggested that the Home Office could also increase the dosage of interventions by coordinating activities across funded areas. In addition, funding guidelines should encourage or require areas to include plans for sustainability in their bid.
5. Increase the continuity of interventions across funded areas and over rounds. Focusing on specific outcomes, interventions or other focus areas would improve the ability to demonstrate change. Some stakeholders suggested a national coordination of some interventions to achieve a consistent message and increase reach.
6. Future funds should consider striking a careful balance between financial reporting requirements and ensuring projects can focus on delivery, as some local project teams felt the monthly reporting requirements to be a burden and a distraction from delivering their interventions.
7. Continue adopting a flexible funding model that supports a variety of intervention types and also a greater range of eligible delivery organisations (compared to the more restricted eligibility criteria used on other SSF rounds).
8. Continue to encourage innovative approaches and delivery methods that are not in the scope of other funds. However, this freedom should exist within a more focused approach, with clarity on the core outcomes and intervention categories that the Fund will support so that change is targeted.
9. Future funds should consider alternative payment structures and procurement approaches, in light of the challenges that some local teams faced with the practice of claiming in arrears, particularly when working with voluntary sector partners. In addition, the requirement to spend all the budget before the end of the financial year could be made more flexible.
10. Carefully consider how different iterations of SWaN and SSF (and other funds) may impact on the ability of local areas and organisations to effectively develop their bids; and avoid running simultaneous bidding and commissioning processes.

6.3.2 Implications and recommendations for future evaluations

1. Focus on measuring short-term outcomes that are the direct effects of interventions, particularly when funding periods are short. This will help to demonstrate the effectiveness of SWaN when achieving and measuring longer-term impacts is challenging.
2. Consider the funding period as it relates to the evaluation. Build in more time before the delivery of interventions begins to ensure sufficient time for uncontaminated baseline data collection. Apply greater consideration to the post-intervention data collection timeframe to ensure change is measured at the most appropriate time.

7. Appendices

Appendix A: 7.1 Further methods details

7.1.1 Principles for evaluation design

Interviews with key stakeholders in the scoping stage of the evaluation highlighted 5 key principles that were carried forward into the evaluation design:

1. The process evaluation should seek to understand the intended mechanisms and levers of change to build an understanding of ‘what works’ to improve feelings of safety and prevent VAWG in the NTE.
2. NTE workers and other relevant groups working at night should be treated as distinct beneficiary groups due to specific outcomes and data collection considerations that differ from users of NTE spaces.
3. Data collection should consider issues specific to each audience type, including utilising gatekeepers for data collection with NTE venues and workers, and the ethical considerations of collecting data from individuals who are under the influence of alcohol.
4. Outcome measures that could demonstrate any immediate effects from interventions should be included where impact is likely to be limited within the evaluation timeframe.
5. The current climate in relation to VAWG, and the NTE post-COVID-19, could lead to particularly negative feelings regarding safety and comfort that are not related to SWaN interventions but could moderate their impact.

7.1.2 Impact evaluation research questions and outcome measures

The final impact evaluation design aimed to address 4 research questions. This section summarises the measurement approach for each research question.

1. What is the impact (if any) of the funded interventions on feelings of safety, with a particular focus on women and girls’ feelings of safety at night? Verian measured this research question using survey data. Key outcome measures for this research question were respondents’ perceptions of change (since 6 months previously) in: their own level of safety in public spaces during the day and at night; their own level of safety inside NTE venues at night; and the level of safety of women and girls at night.
2. Does impact differ by type of area (for example, urban versus rural, residential versus commercial, NTE versus residential)? This research question was measured using survey data. The analysis was carried out by exploring whether the key outcomes measures for research question 1 were associated with characteristics of individuals, funded areas or SWaN-funded projects. Specifically, the impact analysis measured whether outcomes were associated with: interventions that targeted specific types of public space (for example, public transport); interventions targeting NTE areas; intervention type; and respondent sub-groups (type of area, age and disability status).
3. What is the best methodology to use to look at and understand the impact of interventions over time? This research question was addressed during the scoping phase and evaluation design. The key output from this question was the evaluation design report, which detailed the evaluation methodology outlined in Section 2.
4. Do the funded interventions have any other benefits? This research question was measured using survey data. Key outcome measures for this research question were respondents’ perceptions of change (since six months previously) in: VAWG in public spaces at night; ASB in public spaces at night; levels of confidence in the local police; NTE workers’ capabilities in responding to VAWG in their workplaces (email survey only); and NTE organisations’ capabilities in responding to VAWG (email survey only).

7.1.3 Survey base sizes and demographic breakdowns

A. F2F survey

Table 7.1: Demographic breakdown of face-to-face (F2F) survey respondents

	Total n=844	n	% of respondents
Gender	Men	0	0.0
	Women	844	100.0
Age	16-24	323	38.3
	25-34	210	24.9
	35-49	160	19.0
	50-54	42	5.0
	55-59	45	5.3
	60-64	26	3.1
	65-74	31	3.7
	75 or over	8	0.9
Education	A postgraduate degree	164	19.4
	An undergraduate degree	191	22.6
	A levels, O levels, GCSEs	331	39.2
	Any other qualifications (such as BTEC, diplomas, trade apprenticeships)	108	12.8
	No qualifications	40	4.7
Ethnicity	Asian or Asian British	61	7.2
	Black, black British, Caribbean or African	38	4.5
	Mixed or multiple ethnic groups	42	5.0
	White	682	80.8
	Other ethnic group	16	1.9
Mental or physical health condition lasting longer than 12 months	Yes	193	25
	No	635	73
Of those who have a disability	Yes, affects a lot	53	6.3
	Yes, affects a little	98	11.6
	Not at all	40	4.7

B. Email survey

Table 7.2: Demographic breakdown of email survey respondents

	Total n=217	n	% of respondents
Gender	Men	51	23.5
	Women	165	76.0
	Prefer not to say	1	0.46
Age	16-24	30	13.8
	25-34	60	27.6
	35-49	71	32.7
	50-54	25	11.5
	55-59	16	7.4
	60-64	10	4.6
	65-74	4	1.8
	75 or over	1	0.5
Sector	NTE (for example, bars, pubs, clubs, restaurants)	71	32.7
	Policing	56	25.8
	Transport	2	0.9
	Emergency services (non-police)	2	0.9
	Local government	33	15.2
	Student	2	0.9
	Charity, community and other support work	19	8.8
	Health and social care	7	3.2
	Other	25	11.5
Ethnicity	Asian or Asian British	4	1.8
	Black, black British, Caribbean or African	2	0.9
	Mixed or multiple ethnic groups	3	1.4
	White	203	93.5
	Other ethnic group	1	0.5
Mental or physical health condition lasting longer than 12 months	Yes	43	20
	No	169	78
Of those who have a disability	Yes, affects a lot	6	2.8
	Yes, affects a little	20	9.2
	Not at all	14	6.5

Appendix B: 7.2 Further descriptive statistics on crime and safety (NTE workers)

This section presents results from the descriptive analysis of email survey responses for the same selection of measures presented in Section 3, namely: experience of crime, worry about crime and perceptions of safety in public spaces.

7.2.1 Experience of crime

Experiencing and/or witnessing VAWG was less commonly reported among NTE workers than users of public spaces. As Figure 7.1 shows, a quarter (25%) of respondents had witnessed or experienced VAWG in the last month. However, NTE workers were more likely to have witnessed VAWG in the last month at night than the users of public spaces (20% vs 14%). There were some statistically significant differences in witnessing and experiencing crime between age groups and the industry/sector where NTE workers were employed. Specifically, 20% of those aged 16 to 24 years reported personally experiencing VAWG in the previous month at/near work, which was statistically significantly more than those aged 35 to 49 years (4%), and 39% of workers at NTE venues had experienced or witnessed VAWG, compared with 7% of workers within the police.

Figure 7.1: Experiencing or witnessing VAWG

Notes:

1. In the last month, have you personally experienced or witnessed types of VAWG or gender-based violence at/near your place of work? (Base=217).

7.2.2 Worry about crime

NTE workers were asked how much they worry about others experiencing VAWG (such as their colleagues or customers) in public spaces. Across all types of public spaces, over half (ranging from 53% to 60%) of NTE workers reported feeling very or fairly worried about others experiencing VAWG.

Figure 7.2: Worry about others experiencing VAWG in public spaces

Notes:

1. How worried are you about others, like your colleagues or customers, experiencing VAWG? (Base=215).

Statistically significant gender differences in worry about others experiencing VAWG were identified in 3 of the public space types. Women were more likely than men to report being very worried for others in NTE areas (12% vs 2%) and inside NTE venues (13% vs 2%). Conversely, men were more likely to report not being worried at all about the safety of others on public transportation (8%), than women (2%). Furthermore, respondents who reported having a disability were statistically more likely to worry about others experiencing VAWG. For example, 79% of respondents with a disability were worried about others experiencing VAWG in NTE areas, compared to 54% of respondents without a disability.

7.2.3 Perceptions of safety in public spaces

Overall, more NTE workers felt unsafe (52%) than safe (43%) in NTE areas at night. However, this finding was largely driven by the women in the sample who were statistically significantly more likely than men to feel unsafe (59% vs 27%). Respondents with a disability were also statistically significantly more likely to feel unsafe in NTE areas after dark (67%) than those without a disability (49%), and relatedly, respondents without a disability were statistically significantly more likely to feel safe in these areas (47%) than those with a disability (26%).

Figure 7.3: Perceptions of safety in NTE areas

Notes:

1. How safe or unsafe would you feel walking on your own after dark/at night where there are pubs, bars and clubs? (Base NTE=217).

NTE workers were also asked about how safe they thought women and girls were in public spaces in the NTE. Overall, a majority of NTE workers reported thinking women and girls were unsafe in NTE-related spaces, but with strong differences by location. While nearly half reported thinking they were safe inside NTE venues (44% vs 50% unsafe), only a quarter reported thinking they were safe outside in NTE areas (25% vs 71% unsafe), and slightly more than a quarter reported thinking they were safe at public transportation hubs (27% safe vs 68% unsafe) or on public transportation (30% safe vs 65% unsafe).

Figure 7.4: Perceptions of the safety of women and girls in NTE locations

Notes:

1. Thinking specifically about the safety of women and girls in the area where you work or study, how safe or unsafe do you think it is for your customers/the public/your peers after dark/at night when on their own on public transportation, including buses, trains/trams, and taxis travelling to and from where you work or study (Base=217); at public transportation stations, such as train stations, bus stations, taxi ranks or other transport hubs (Base=217); walking on their own in the NTE area (in other words, areas where the bars, pubs and clubs are clustered) (Base=217); inside NTE venues (for example, inside bars, pubs, clubs) (Base=217)?

Women were statistically significantly more likely than men to consider women and girls unsafe inside NTE venues (54% vs 35%) and at public transport stations (72% vs 55%).

7.2.4 Awareness and use of interventions

Comparing what respondents noticed in public spaces with the kinds of interventions that SWaN funded in those spaces (Figure 7.5) shows that some interventions were not highly visible to users of public spaces and, at the same time, some non-SWaN-funded interventions were noticed. Adverts were a common SWaN-funded intervention but were only recognised by 9% of respondents in areas where this activity was funded. Education and training and guardianship schemes also had low awareness in areas where they were funded. CCTV and lighting were the most likely types of interventions to be noticed, though these were not funded by SWaN (these could be funded locally or via other schemes, such as the SSF).

Figure 7.5: Awareness of interventions versus where SWaN activities took place (F2F survey)

Notes:

1. Which of the following (above) crime prevention activities are you aware of or have you noticed in public spaces? (Base=844)

Appendix C: 7.3 Impact analysis technical note

7.3.1 Overview of impact analysis using survey data

Verian used hierarchical regression models to estimate the impacts of the SWaN Fund on outcome measures collected through the surveys. As described in Section 2.4.2, the outcome measures for each model were measured on a 0 to 10 scale, representing respondents’ perceptions of change over the previous 6 months (the period over which the interventions were implemented).

Each model included a term for whether the respondent was surveyed within a funded area or a comparison area. The estimate for this term represents the estimated difference between an individual in a funded area and a similar individual in a comparison area. Under the assumption that all relevant differences between the funded areas and the comparison areas are accounted for in the model, this can be interpreted as an estimate of the impact attributable to the SWaN Fund.

To support this assumption, each model included a set of variables to account for differences between respondents in SWaN Fund areas and comparison areas. These are described further in Section 7.3.2. The models also included separate intercepts for each area to account for the clustered nature of the sample; in other words, the tendency of responses from individuals in the same area to be more similar than the responses of individuals in different areas. These area-level intercepts will also partially account for differences between the SWaN Fund areas and the comparison areas that are not already directly captured in the model.

However, a key limitation of this approach is that it was not possible to verify that the models fully accounted for all relevant differences between SWaN Fund areas and the comparison samples. There will be some unobserved differences between the groups which are unaccounted for. It is possible that these differences explain some or all of any estimated impact. Similarly, it is possible that these differences mask any true impacts of SWaN Fund interventions.

7.3.2 Model selection

A process of cross-validation was run to select the best fitting model; that is, the model that minimised the likelihood that any observed impact resulted from other unaccounted-for differences between the SWaN Fund areas and the comparison samples.

For each survey mode, the cross-validation involved:

1. Splitting the data into 3 random subsets.
2. Fitting a model using data for 2 of the 3 subsets.
3. Using the output of this model to predict the outcome values of the third subset.
4. Comparing these predicted values against the actual values.

When comparing different model specifications, the stronger specification will generally give predictions which are closer to the actual values.

This process was repeated for each outcome measure within each survey mode, testing 3 separate model specifications of increasing complexity.

Specification 1 included 3 core elements:

1. An indicator for whether the respondent was surveyed as part of a SWaN Fund area or a comparison area.
2. The survey weight (see Section 7.3.3 below).
3. Area-level intercepts.

The SWaN Fund indicator is the key variable of interest, as this represents the estimated impact of SWaN Fund interventions.

Specification 2 added a set of individual-level characteristics (demographic variables), as well as information about the type of area in which the interview was conducted (for the F2F survey) and information about respondents’ jobs (for the email survey).

• the individual-level characteristics were: age, gender (for the email survey, as the F2F survey was women-only), ethnicity, highest education level, disability status, sexual orientation, and marital/relationship status
• the information about types of areas for the F2F survey identified each area as being a residential area or university campus, a high street or shopping area, an NTE area, a transport hub (such as by a railway station), or a park or green space
• the information about respondents’ jobs for the email survey was: sector (for example, staff in NTE venues, police, charity); seniority (whether involved in decision-making about policy/practice in their workplace); and whether they are based in a single location or travel between locations

As described in Section 7.3.3, the propensity score matching method used to derive the survey weight already attempts to account for many of the observed differences between the SWaN Fund areas and the comparison samples. However, it is reasonable to think that including these variables separately in the regression models will better account for these differences than will the survey weight alone.

Specification 3 added a further set of individual-level variables related to respondents’ use of public spaces:

• the respondent’s reasons for visiting the area in which they were surveyed (for example, working/studying, shopping, visiting a pub or bar)
• how often the respondent typically visits the area
• what times of day the respondent typically visits the area

These variables were only relevant for the F2F survey, as they relate to respondents’ use of the area in which they were surveyed. These variables were intended to account for the different ways in which members of the public in the SWaN Fund areas may use these public spaces, in comparison with members of the public in the comparison areas.

The cross-validation suggested there was very little difference in the performance of the 3 model specifications. Specification 2 was used for all the impact models presented in this report, as it most commonly had the best cross-validated performance, although the difference was very marginal. The cross-validation process also demonstrates that the impact results are not sensitive to these choices about how the models were constructed.

7.3.3 Weighting

Verian developed propensity score matched weights for each survey to ensure the samples in funded areas and comparison areas were broadly comparable. The first step in deriving the survey weights was for Verian to fit a logistic regression model predicting whether a respondent was from a funded area or from the comparison sample.

The following predictor variables were used in these models for both the F2F and email surveys:

• age
• ethnicity
• prior highest education level
• disability status
• sexual orientation
• marital/relationship status

In addition, the following variables were used for the F2F survey:

• type of area in which the survey was conducted (residential area, high street or shopping area, NTE area, transport hub, park/green space or footpath/cycleway)
• respondent’s reasons for visiting the area (work or study, shopping, going to a restaurant, going to bars, pubs, clubs or similar venues, or other reason)
• how frequently the respondent typically visits the area
• what times of day the respondent typically visits the area

The following variables were additionally used for the email survey:

• gender (this was not relevant for the F2F survey, in which only women were surveyed)
• job sector
• seniority in their job (whether have primary responsibility for decision-making and setting policy, some responsibility, or no such responsibility)
• whether they work in a single fixed place or multiple locations

These models gave a predicted probability for each respondent of being in a funded area. The matching weighting algorithm then assigned weights to respondents in the comparison areas, giving greater weight to respondents with higher predicted probabilities (those which were most similar to respondents in funded areas). This process accounts for many of the initial differences between the funded areas’ sample and the comparison areas’ sample.

7.3.4 Estimated impacts

A. Main analysis

Table 7.3: Key outputs of SWaN impact models – respondents’ perceptions of a change in their own level of safety in public spaces

Think about a scale where 0 means you feel much less safe and 10 means you feel much more safe. Since October 2021 (in the last 6 months), how have your feelings of safety changed in [public space] in this area?

Public space	Survey	Impact estimate	95% confidence interval (CI): lower bound	95% CI: upper bound	p-value (significant values marked with *)
…on public transport…during the day	F2F	-0.10	-0.24	0.04	0.197
…on a high street or in a shopping area…after dark/at night	F2F	-0.01	-0.19	0.16	0.891
…at public transportation stations…after dark/at night	F2F	0.07	-0.12	0.26	0.487
…where there are pubs, bars and clubs…after dark/at night	F2F	0.02	-0.13	0.17	0.804
…on public transportation…after dark/at night	F2F	-0.08	-0.22	0.06	0.292
…at public transportation stations…after dark/at night	Email	0.22	-0.13	0.57	0.232
…where there are pubs, bars and clubs…after dark/at night	Email	0.32	-0.07	0.71	0.122
…on public transportation…after dark/at night	Email	0.45	0.01	0.88	0.056

Table 7.4: Key outputs of SWaN impact models – respondents’ perceptions of a change in their own level of safety inside NTE venues

Think about a scale where 0 means you feel much less safe and 10 means you feel much more safe. Since October 2021 (in the last six months), how have your feelings of safety inside these venues after dark changed? (F2F survey)

Since October 2021 (in the last 6 months) how, if at all, have your feelings of safety after dark/at night changed when – inside NTE venues (for example, inside bars, pubs, clubs)? (email survey)

Public space	Survey	Impact estimate	95% CI: lower bound	95% CI: upper bound	p-value (significant values marked with *)
…restaurants or cafes	F2F	-0.15	-0.35	0.04	0.139
…bars or pubs	F2F	-0.31	-0.59	-0.04	0.037*
…clubs or other age-restricted night-time entertainment	F2F	-0.24	-0.60	0.12	0.224
…other entertainment facilities, such as bowling, cinema, theatre	F2F	-0.17	-0.37	0.03	0.109
…inside NTE venues (for example, inside bars, pubs, clubs)	Email	0.20	-0.18	0.59	0.330

Table 7.5: Key outputs of SWaN impact models – respondents’ perceptions of a change in the level of safety of women and girls

Thinking about the safety of women and girls, since October 2021 (in the last 6 months) how, if at all, do you think the safety for your customers/the public/your peers has changed after dark/at night when they are…

Public space	Survey	Impact estimate	95% CI: lower bound	95% CI: upper bound	p-value (significant values marked with *)
…walking on their own in the NTE area (that is, areas where bars, pubs and clubs are clustered)	Email	0.48	0.05	0.92	0.042*
…inside NTE venues (for example, inside bars, pubs, clubs)	Email	0.45	-0.02	0.92	0.076
…on their own at public transportation stations, such as a train station, bus station, taxi rank or other transport hub	Email	0.38	-0.06	0.82	0.111
…on their own on public transportation, including buses, trains/trams, and taxis travelling to and from where you work or study	Email	0.45	0.01	0.88	0.056
Since October 2021 (in the last six months) how do you think the safety of women and girls inside your place of work after dark/at night has changed?	Email	0.19	-0.27	0.64	0.435

Table 7.6: Key outputs of SWaN impact models – respondents’ perceptions of a change in VAWG

Since October 2021 (the last 6 months), how much do you think VAWG in public spaces or inside venues at night has changed in this area after dark/at night? (F2F survey)

Thinking back to October 2021, since then, how, if at all, would you say that VAWG in each of these spaces after dark/at night has changed? (email survey)

Outcome	Survey	Impact estimate	95% CI: lower bound	95% CI: upper bound	p-value
…in public spaces or inside venues at night	F2F	-0.26	-0.63	0.11	0.203
…in the NTE area (that is, areas where bars, pubs and clubs are clustered)	Email	-0.13	-0.50	0.19	0.522
…inside NTE venues (for example, inside bars, pubs, clubs)	Email	-0.13	-0.48	0.24	0.521
…at public transportation stations, such as a train station, bus station, taxi rank or other transport hub	Email	-0.08	-0.39	0.23	0.628
…on public transportation, including buses, trains/trams, and taxis travelling to and from where you work or study	Email	-0.07	-0.39	0.26	0.706

Table 7.7: Key outputs of SWaN impact models – respondents’ perceptions of a change in anti-social behaviour

Outcome	Survey	Impact estimate	95% CI: lower bound	95% CI: upper bound	p-value
Since October 2021 (in the last 6 months), how much do you think anti-social behaviour has changed in the area where you work or study after dark/at night?	Email	-0.36	-0.81	0.10	0.147

Table 7.8: Key outputs of SWaN impact models – respondents’ perceptions of a change in the level of confidence in the local police

Since October 2021 (the last 6 months), how has your…changed? (F2F survey)

Thinking back to October 2021, since then, would you say that…has [increased/decreased]? (email survey)

Outcome	Survey	Impact estimate	95% CI: lower bound	95% CI: upper bound	p-value
…confidence in local police dealing with VAWG in public spaces or inside venues at night in this area	F2F	0.13	-0.13	0.40	0.360
…likelihood of reporting VAWG in this area to police	F2F	-0.06	-0.38	0.26	0.738
…your level of confidence in reporting incidents of VAWG you have witnessed in/near the area where you work or study to the police	Email	0.40	-0.05	0.84	0.120
…the police’s level of effectiveness in addressing VAWG after dark/at night in/near the area where you work or study	Email	0.27	-0.24	0.78	0.310

Table 7.9: Key outputs of SWaN impact models – respondents’ perceptions of a change in their capabilities as regards responding to VAWG in their place of work or study

Since October 2021 (in the last 6 months) how, if at all, do you feel that this rating has changed?

Outcome	Survey	Impact estimate	95% CI: lower bound	95% CI: upper bound	p-value (significant values marked with *)
…your awareness and understanding of VAWG	Email	0.59	0.21	0.97	0.003*
…your ability / how equipped you feel to identify when women and girls could be vulnerable or at risk	Email	0.47	-0.14	1.12	0.170
…your ability / how equipped you feel to identify predatory behaviour	Email	0.41	-0.23	1.11	0.230
…your ability / how equipped you feel to intervene if you witness VAWG	Email	0.63	0.11	1.12	0.015*
…your confidence in supporting victims of VAWG	Email	0.92	0.55	1.30	0.000*

Table 7.10: Key outputs of SWaN impact models – respondents’ perceptions of a change in their organisation’s capabilities in regard to responding to VAWG

Thinking back to October 2021, since then, would you say [how you rate the organisation where you work/study] has [increased/decreased]?

Outcome	Survey	Impact estimate	95% CI: lower bound	95% CI: upper bound	p-value (significant values marked with *)
…its overall strategy for ensuring the safety of women and girls in/near its premises/in the areas you work/study/work or study after dark/at night	Email	0.56	0.12	0.98	0.021*
…its overall level of competence in identifying and addressing VAWG inside its premises/the areas you work/study/work or study after dark/at night	Email	0.36	-0.02	0.69	0.101
…their level of co-operation with other organisations to address VAWG in/near its premises/in the areas you work/study/work or study after dark/at night	Email	0.59	0.13	1.04	0.020*

Table 7.11: Key outputs of SWaN impact models – respondents’ perceptions of a change in the level of confidence in NTE staff

Since October 2021 (the last 6 months), how has your confidence in each of these people to deal with sexual harassment or gender-based violence in public spaces or inside venues at night in this area changed? (F2F survey)

Outcome	Survey	Impact estimate	95% CI: lower bound	95% CI: upper bound	p-value
Pub or bar staff	F2F	-0.23	-0.71	0.28	0.402
Club or other age-restricted night-time venue staff	F2F	-0.50	-1.00	0.02	0.064

B. Further analysis 1: SWaN Fund areas targeting particular kinds of public space

Verian conducted further analysis to explore the extent to which certain key impact estimates were associated with characteristics of individuals, funded areas or SWaN Fund projects. Given the small sample size of the email survey, these further analyses were limited to the F2F survey data.

The first form of additional analysis explored the extent to which women’s perceptions of safety in different public spaces were associated with the types of spaces SWaN Fund interventions were targeting. As part of the bid review (see Section 2.3), Verian mapped the planned interventions of each SWaN Fund project against the kinds of public spaces in which those interventions primarily operated. For example, some projects included interventions aimed at transport hubs, while others only focused on NTE areas. The initial impact analysis described above was concerned with overall changes in perceptions of safety in different public spaces. However, it is plausible that effects would have been greater in areas where there was a greater focus on interventions targeting that particular kind of space.

To test this, Verian fit additional hierarchical regression models for the measures related to changes in perceptions of safety. These models were a very similar form to the impact models described previously in Section 7.3.2 but included one additional variable indicating whether a SWaN Fund project had any interventions targeting the relevant type of public space. The estimate for this variable represents the difference in the estimated impacts for areas with SWaN interventions specifically targeting that kind of public space and other SWaN areas (which did not have interventions specifically targeting that kind of public space).

Table 7.12: Differences in estimated impacts associated with interventions targeted at specific types of public space (F2F survey)

Think about a scale where 0 means you feel much less safe and 10 means you feel much more safe. Since October 2021 (in the last 6 months), how have your feelings of safety changed in [public space] in this area?

Outcome	Type of public space targeted	Difference in estimated impact vs. other SWaN Fund areas	95% CI: lower bound	95% CI: upper bound	p-value
…on public transport…during the day	Public transport	-0.07	-0.26	0.11	0.482
…on a high street or in a shopping area…after dark/at night	Quiet streets/high streets	0.08	-0.21	0.36	0.626
…at public transportation stations…after dark/at night	Public transport	0.06	-0.19	0.30	0.682
…where there are pubs, bars and clubs…after dark/at night	NTE venues	-0.01	-0.18	0.15	0.872
…on public transportation…after dark/at night	Public transport	-0.05	-0.21	0.10	0.504

Table 7.13: Differences in estimated impacts inside NTE venues associated with interventions targeted at NTE areas (F2F survey)

Since October 2021 (the last 6 months), how much do you think VAWG in public spaces or inside venues at night has changed in this area after dark/at night?

Outcome	Type of public space targeted	Difference in estimated impact vs. other SWaN Fund areas	95% CI: lower bound	95% CI: upper bound	p-value
Bars or pubs	NTE venues	-0.17	-0.45	0.11	0.289
Clubs or other age-restricted night-time entertainment	NTE venues	-0.24	-0.56	0.04	0.264

C. Further analysis 2: Differences between types of interventions

The second form of further analysis explored the extent to which more or less positive effects were associated with different types of SWaN Fund interventions. Verian fit further hierarchical regression models that had broadly the same form as the impact models described in Section 7.3.2, although instead of including an indicator to identify respondents in SWaN areas, the models included a series of indicators identifying whether a given intervention type was delivered in each area. These models do not account for how different intervention types may interact (for example, the possibility that 2 interventions may be more effective in combination than either one is alone). However, the analysis can still give some indication of the extent to which different interventions are associated with different effects.

Initially, this analysis was focused on one outcome – perceptions of a change in the amount of VAWG in public spaces in the area after dark/at night. However, the analysis was then extended to include perceptions of a change in general feelings of safety after dark, and feelings of safety inside NTE venues.

Table 7.14: Association between intervention types and estimated effects on general perceptions of a change in safety in areas where there are pubs, bars and clubs (F2F survey)

And since October 2021, how have your feelings of safety changed where there are pubs, bars and clubs in this area after dark/at night?

Intervention type	Survey	Impact estimate	95% CI: lower bound	95% CI: upper bound	p-value
Education/training	F2F	0.14	-0.19	0.47	0.479
Personal safety initiatives	F2F	-0.06	-0.27	0.17	0.687
Guardianship initiatives	F2F	-0.01	-0.24	0.21	0.953
Police patrols	F2F	0.04	-0.17	0.25	0.735
Frontline support (such as equipment, or communications aimed at frontline workers)	F2F	0.00	-0.30	0.27	0.987
Community/venue schemes	F2F	-0.23	-0.49	0.07	0.178
Youth work	F2F	-0.05	-0.31	0.20	0.761
Promotion of victim support services	F2F	-0.11	-0.40	0.17	0.551

Table 7.15: Association between intervention types and estimated effects on perceptions of a change in safety inside bars, pubs, clubs or other age-restricted night-time entertainment (F2F survey)

Think about a scale where 0 means you feel much less safe and 10 means you feel much more safe. Since October 2021 (in the last 6 months), how have your feelings of safety inside [bars or pubs/clubs or other age-restricted night-time entertainment] after dark changed?

Intervention type	Survey	Impact estimate	95% CI: lower bound	95% CI: upper bound	p-value
Education/training	F2F	-0.01	-0.45	0.48	0.964
Personal safety initiatives	F2F	0.01	-0.37	0.32	0.966
Guardianship initiatives	F2F	-0.29	-0.65	0.08	0.271
Police patrols	F2F	0.12	-0.16	0.49	0.628
Frontline support (such as equipment, or communications aimed at frontline workers)	F2F	0.02	-0.66	0.60	0.950
Community/venue schemes	F2F	-0.18	-0.69	0.54	0.655
Youth work	F2F	0.24	-0.22	0.71	0.423
Promotion of victim support services	F2F	0.04	-0.36	0.59	0.921

Table 7.16: Association between intervention types and estimated effects on perceptions of a change in VAWG (F2F survey)

Since October 2021 (the last 6 months), how much do you think VAWG in public spaces or inside venues at night has changed in this area after dark/at night?

Intervention type	Survey	Impact estimate	95% CI: lower bound	95% CI: upper bound	p-value
Education/training	F2F	0.37	-0.31	1.02	0.426
Personal safety initiatives	F2F	-0.41	-0.88	0.06	0.227
Guardianship initiatives	F2F	0.37	-0.11	0.88	0.295
Police patrols	F2F	0.01	-0.41	0.45	0.962
Frontline support (such as equipment, or communications aimed at frontline workers)	F2F	-0.47	-1.04	0.12	0.247
Community/venue schemes	F2F	0.51	-0.03	1.08	0.184
Youth work	F2F	0.03	-0.50	0.56	0.939
Promotion of victim support services	F2F	-0.14	-0.74	0.46	0.743

D. Further analysis 3: How effects vary between sub-groups

The third form of further analysis explored the extent to which effects varied between sub-groups of respondents. Verian fit further hierarchical regression models that each had broadly the same form as the impact models described in Section 7.3.4 above, with the addition of an interaction term between the variable identifying SWaN Fund areas and a variable representing the sub-group of interest. These coefficients represent the extent to which estimated effects vary within these sub-groups.

For the F2F survey, this analysis focused on differences related to the age and disability status of respondents, and the type of area in which the survey was conducted. As is conventional for these kinds of model, where a variable has more than 2 categories, a reference category was selected: residential areas and university campuses in the case of the type of area, and those aged 16 to 24 years in the case of the age band. The coefficients then represent the difference between each of the other categories and this reference category. For example, the model shows how the estimated effects for each of the other age groups differs from that of the reference category of those aged 16 to 24 years.

Taking the first example (Table 7.17), the estimated effect of SWaN Fund interventions on feelings of safety after night in areas with pubs, bars and clubs was 0.2 points lower for surveys conducted on high streets or in shopping areas than in residential areas or university campus areas. This difference is not statistically significant. Similarly, there is no statistically significant difference in the estimated effects among respondents with a disability and respondents with no identified disability.

However, SWaN Fund interventions appear to have had a greater impact on feelings of safety in NTE areas among older women than among younger women. The estimated effect increases with age – on average, an increase of 0.01 points for each year of age, or a difference of 0.46 points between women aged 50 or above and women aged 16 to 24 years. In this case, the difference is statistically significant.

Table 7.17: Relationship between key sub-groups (type of area, age, and disability status) and estimated effects on general perceptions of a change in safety in areas where there are pubs, bars and clubs (F2F survey)

And since October 2021, how have your feelings of safety changed where there are pubs, bars and clubs in this area after dark/at night?

Sub-group	Survey	Difference in estimated impact	95% CI: lower bound	95% CI: upper bound	p-value (significant values marked with *)
Type of area: high street/ shopping area vs. residential/ university campus area	F2F	-0.20	-0.60	0.20	0.336
Type of area: NTE area vs. residential/ university campus area	F2F	-0.25	-0.72	0.22	0.304
Age: difference in estimate impact associated with an increase in age of one year	F2F	0.010	0.001	0.019	0.035*
Age: 25 to 34 years vs. 16 to 24 years	F2F	0.17	-0.21	0.54	0.384
Age: 35 to 49 years vs. 16 to 24 years	F2F	-0.03	-0.49	0.44	0.915
Age: 50 years or older vs. 16 to 24 years	F2F	0.46	0.05	0.88	0.030*
Disability status: any disability vs. no identified disability	F2F	0.18	-0.20	0.57	0.356

Table 7.18: Relationship between key sub-groups (type of area, age and disability status) and estimated effects on perceptions of a change in safety inside bars, pubs, clubs or other age-restricted night-time entertainment (F2F survey)

Think about a scale where 0 means you feel much less safe and 10 means you feel much more safe. Since October 2021 (in the last 6 months), how have your feelings of safety inside [bars or pubs/clubs or other age-restricted night-time entertainment] after dark changed?

Sub-group	Survey	Difference in estimated impact	95% CI: lower bound	95% CI: upper bound	p-value
Type of area: high street/ shopping area vs. residential/ university campus area	F2F	-0.82	-2.00	0.41	0.200
Type of area: NTE area vs. residential/ university campus area	F2F	-0.90	-2.10	0.33	0.160
Age: difference in estimate impact associated with an increase in age of one year	F2F	0.006	-0.011	0.024	0.480
Age: 25 to 34 years vs. 16 to 24 years	F2F	0.44	-0.09	0.98	0.119
Age: 35 to 49 years to vs. 16-24 years	F2F	0.48	-0.24	1.18	0.200
Age: 50 years or older vs. 16 to 24 years	F2F	0.05	-0.68	0.81	0.893
Disability status: any disability vs. no identified disability	F2F	0.12	-0.47	0.71	0.712

Table 7.19: Relationship between key sub-groups (type of area, age and disability status) and estimated effects on a perceived change in VAWG (F2F survey)

Since October 2021 (the last 6 months), how much do you think VAWG in public spaces or inside venues at night has changed in this area after dark/at night?

Sub-group	Survey	Difference in estimated impact	95% CI: lower bound	95% CI: upper bound	p-value
Type of area: high street/ shopping area vs. residential/ university campus area	F2F	-0.84	-2.34	0.51	0.252
Type of area: NTE area vs. residential/ university campus area	F2F	-1.01	-2.43	0.39	0.168
Type of area: Transport hub vs. residential/ university campus area	F2F	-0.58	-2.26	0.95	0.488
Type of area: park or green space/ footpath or cycleway vs. residential/ university campus area	F2F	-0.13	-1.90	1.62	0.890
Age: difference in estimated impact associated with an increase in age of one year	F2F	-0.009	-0.026	0.007	0.264
Age: 25 to 34 years vs. 16 to 24 years	F2F	0.29	-0.30	0.91	0.359
Age: 35 to 49 years vs. 16 to 24 years	F2F	0.01	-0.72	0.77	0.976
Age: 50 years or older vs. 16 to 24 years	F2F	-0.53	-1.21	0.17	0.142
Disability status: any disability vs. no identified disability	F2F	-0.27	-0.85	0.35	0.390

Verian conducted similar analyses for the email survey, investigating the extent to which perceptions of how the safety of women and girls in general had changed in the area were related to the gender of the respondent. As the NTE workers covered by the email survey included both men and women, it is reasonable to think that effects on these perceptions may vary by gender.

This analysis involved adding an interaction term between gender and the variable identifying SWaN Fund areas. The coefficients shown in Tables 7.20 and 7.21 then represent the difference in the estimated impacts between women and men.

Table 7.20: Relationship between gender and estimated effects on perceptions of a change in safety of women and girls in general (email survey)

Since October 2021 (in the last 6 months) how do you think the safety of women and girls inside your place of work after dark/at night has changed?

Sub-group	Survey	Difference in estimated impact	95% CI: lower bound	95% CI: upper bound	p-value
Women vs men	Email	-0.09	-0.83	0.63	0.818

Table 7.21: Relationship between gender and estimated effects on perceptions of a change in safety of women and girls in general by gender (email survey)

Thinking about the safety of women and girls, since October 2021 (in the last 6 months) how, if at all, do you think the safety of your customers/the public/your peers has changed after dark/at night when they are…?

Type of area	Sub-group	Difference in estimated impact	95% CI: lower bound	95% CI: upper bound	p-value
…walking on their own in the NTE area (that is, areas where bars, pubs and clubs are clustered)	Women vs men	-0.09	-0.83	0.63	0.818
…inside NTE venues (for example, inside bars, pubs, clubs)	Women vs men	-0.18	-0.94	0.56	0.654
…on their own at public transportation stations, such as a train station, bus station, taxi rank or other transport hub	Women vs men	-0.14	-0.88	0.58	0.719
…on their own on public transportation, including buses, trains/trams, and taxis travelling to and from where you work or study	Women vs men	0.04	-0.69	0.75	0.923

Appendix D: 7.4. Interventions linked to improvements in perceptions of safety among users of public spaces (QuIP)

The QuIP map below (Figure 7.6) presents all interventions that were linked to improvements in perceptions of safety during the SWaN intervention period. The following should be noted when reading the map:

Understanding the arrows: Maps should be read from left to right with the direction of the arrow head on each link reflecting the direction of causation or influence.

Factor labels: The analyst created the factor labels iteratively so they may include additional information relating to the factor category.

Source counts: The number directly above each arrow reflects the number of times a factor was referenced in the data.

Figure 7.6. All interventions influencing positive changes in perceptions of safety during the intervention period (full QuIP)

References

Home Office (2021a) Tackling violence against women and girls strategy. Available online [accessed 20/01/2024]

Home Office (July 2021a) Safety of Women at Night Fund application process. Available online [accessed 20/01/2024]

Home Office (2023) Safer Streets Fund evaluation: Round 1, year ending March 2021. Available online [accessed 20/01/2024]

ONS (2022) Perceptions of personal safety and experiences of harassment, Great Britain: 16 February to 13 March 2022. Office for National Statistics. Available online [accessed 20/01/2024]

ONS (2023) Crime in England and Wales: Annual supplementary tables, March 2023 edition. Office for National Statistics. Available online [accessed 20/01/2024]

1. A ‘full’ QuIP refers to a QuIP study following the full QuIP methodology, with the aim of reaching 24 interview participants. ↩

2. Identified in relation to responses to the following questions: personal safety (Question 11 to Question 19) and VAWG (Question 24). ↩

3. A ‘mini’ QuIP refers to a smaller QuIP study, whereby elements of the QuIP methodology are followed. In this instance, causal chains were explored through responses to survey questions rather than through the traditional method of qualitative interviewing. ↩

4. Question 1: You said the safety of women and girls at your place of work [IF SWGWORKCHG=1-4] got worse [IF SWGWORKCHG=8-11] improved in the last 6 months. What has changed to make you feel this way? For example, has something changed in these areas or venues, or has something else changed? ↩

5. A theory of change is used to illustrate the logic sitting behind an intervention or programme of activity. It sets out the main inputs, activities, outputs, outcomes and impacts for an intervention, to enable the mapping of the causal chains between activities and their desired goals. ↩

6. Descriptive statistics for the comparison areas have been provided as an indication of differences between the 2 groups. They are not the findings from an impact analysis and do not provide a reliable estimate of impact. The results from the impact analysis are presented in Section 4. ↩