Women and Urban Space: Perceptions of fear in relation to Street Harassment and Sexual Offenses

Women and Urban Space: Perceptions of Fear in Relation to Street Harassment and Sexual Offenses

by Ana Sofia Hoch September 2018

Supervisor Dr Kerstin Sailer

A Dissertation submitted in part fulfilment of the Degree of Master of Science Built Environment Space Syntax: Architecture and Cities Bartlett School of Architecture University College London

UCL FACULTY OF THE BUILT ENVIRONMENT BARTLETT SCHOOL OF ARCHITECTURE MSC SPACE SYNTAX: ARCHITECTURE AND CITIES

MSc Space Syntax: Architecture and Cities Dissertation Submission Form A signed and dated copy of the following MUST be inserted after the title page of the hard copy version of your dissertation. If you fail to submit this statement duly signed and dated, your submission will not be accepted for marking.

1. DECLARATION OF AUTHORSHIP I confirm that I have read and understood the guidelines on plagiarism, that I understand the meaning of plagiarism and that I may be penalised for submitting work that has been plagiarised.

I certify that the work submitted is my own and that it has been also submitted electronically and that this can be checked using the JISC detection service, TurnitinÂŽ.

I understand that the work cannot be assessed unless both hard copy and electronic versions of the work are handed in, along with electronic copies of drawings, maps and photographs.

I declare that all material presented in the accompanying work is entirely my own work except where explicitly and individually indicated and that all sources used in its preparation and all quotations are clearly cited.

Should this statement prove to be untrue, I recognise the right of the Board of Examiners to recommend what action should be taken in line with UCL's regulations.

2. COPYRIGHT The copyright of this report remains with me as its author. However, I understand that a copy may be given to my funding body (alongside limited feedback on my academic performance). A copy may also be given to any organisation which has given me access to data and maps (if requested and if appropriate).

I also understand that a copy may be deposited in the UCL E-prints public access repository and copies will be available on the UCL library bookshelves.

Please write your initials in the box if you DO NOT want this report to be made available publicly either electronically or in hard copy.

Name: Ana Sofia Hoch_________________________________________ Signed: _____________________________________________________ Date:

6th September 2018 _____________________________________

2

Abstract Street harassment refers to unwanted remarks and abuses that take place in the public sphere and can turn into violent sexual crime. According to YouGov, in the United Kingdom 64% women have experienced it. In London, most of these actions (31%) have taken place on the streets.

Previous research has studied the phenomenon from a social perspective, but spatial insights remain preliminary. Due to the need of additional research, this study aims to identify the relationship between urban features and the perception of fear of street harassment among women. Therefore, an experimental setup attempts to connect the social research that has been done with the yet exploratory spatial aspects of it.

In order to draw a global picture, an online survey was issued as a platform asking people to indicate which urban situations they feared the most. These results were organised in fear-positive or negative ideas. On this basis, the street configuration of Central London was compared with the Police data. Since there was a high incidence in Soho, this area was analysed in detail. Gate counting and a perceptual survey area were completed.

The findings on a global scale were that people fear the presence of men and solitary spaces most. Also, they fear more unkempt and vandalised buildings as well as low visibility and narrow streets. Segregated spaces and the length of the streets, as potential variables, were tested with statistics in Central London. Results show that crimes correlate with locally integrated streets and inversely with street segment’s length. In Soho, the prevalence of men on streets is evident, especially at night time.

In contrast with the global survey, people asked in the area perceive more danger in streets where not many crimes have been recorded but have characteristics that globally are feared the most.

Since human behaviour is one of the main aspects that space syntax unravels in order to understand its relation to urban space, this study develops a better comprehension of street harassment and sexual crime to improve urban spaces’ design. Moreover, this research will contribute to provide women comfort and enjoyment of public spaces.

Keywords: women, perception of fear, street harassment, sexual crime, urban space, space syntax

3

Table of Contents Abstract .........................................................................................................................................................................................3 List of Illustrations .................................................................................................................................................................6 Glossary ........................................................................................................................................................................................ 8 Chapter 1  Introduction .................................................................................................................................................9 1.1

Research Aims................................................................................................................................................ 12

1.2

Research Questions .................................................................................................................................... 12

1.3

Research Hypotheses................................................................................................................................ 12

Chapter 02  Literature Review............................................................................................................................... 13 2.1 Perception of fear ................................................................................................................................................... 13 Perception of safety ...............................................................................................................................................14 2.2 Public space and gender ................................................................................................................................ 15 2.3 Women’s experience..........................................................................................................................................17 2.4 Street Harassment ...............................................................................................................................................18 2.5 Sexual crime .............................................................................................................................................................. 19 2.6 Summary ................................................................................................................................................................... 20 Chapter 3  Methodology ............................................................................................................................................ 21 3.1 Global scale ................................................................................................................................................................ 21 3.2 City scale: Central London ............................................................................................................................. 23 3.3 Local scale: Soho....................................................................................................................................................23 Chapter 4  Global fear of street harassment ............................................................................................ 25 4.1 Ranking results ....................................................................................................................................................... 27 4.2 Summary ................................................................................................................................................................... 30 Chapter 5  Violent and Sexual Crimes in London .................................................................................. 31 5.1 Data selection criteria ........................................................................................................................................ 31 5.2 Crimes distribution .............................................................................................................................................34 5.3 Spatial analysis ........................................................................................................................................................37 Segment length.......................................................................................................................................................39 4

Multiple Regression Models .......................................................................................................................... 40 5.4 Summary .....................................................................................................................................................................41 Chapter 6  Local scale: Soho.................................................................................................................................. 42 6.1 Social aspects .......................................................................................................................................................... 44 6.2 Visibility........................................................................................................................................................................ 47 6.3 Urban conditions ................................................................................................................................................. 49 6.4 Architectural ............................................................................................................................................................. 51 6.5 Perception of fear of street harassment............................................................................................ 53 6.6 Summary .................................................................................................................................................................... 55 Chapter 7  Discussion ..................................................................................................................................................58 Chapter 8  Conclusions ............................................................................................................................................. 60 8.1 Research Conclusion ......................................................................................................................................... 60 8.2 Limitations .................................................................................................................................................................. 61 8.3 Further analysis .....................................................................................................................................................62 References ...............................................................................................................................................................................63 Appendixes .............................................................................................................................................................................67

5

List of Illustrations Figure 1 European sexual violence. Senthilingam, 2018 ........................................................................9 Figure 2 Street harassment in the United Kingdom. Senthilingam, 2018 ...........................10 Figure 3 Sub-crimes in England and Wales. Ministry of Justice et al, 2013 ........................... 11 Figure 4 Fear’s body map. Nummenmaa, and all (2013) .................................................................... 13 Figure 5 Safety is a second-ranked necessity (Maslow, 1983). Author ......................................14 Figure 6 Routine activity theory (Cohen and Felson, 1974)............................................................... 19 Figure 7 Survey’s interface showed to the participants. All Our Ideas, 2018..................... 22 Figure 8 Global map showing the votes per country. Allourideas.org (2018) ................... 25 Figure 9 Ideas examples in the global survey and scores. Author ...........................................26 Figure 10 Fear- positive (left) and negative (right) ideas results. Author .............................. 27 Figure 11 Social, visual and mobility categories’ results. Author ...................................................28 Figure 12 Global Survey’s results synthesis:. Author ..............................................................................29 Figure 13 Crime types. Police (2018)...................................................................................................................... 31 Figure 14 Crime’s location. Police, 2018 ............................................................................................................ 32 Figure 15 Map with VS offenses in Central London. Police (2018) .............................................. 33 Figure 16 Amount of street segments VS crimes. Police (2018) ..................................................34 Figure 17 Violent and sexual crime counts and their contour. Author ................................... 35 Figure 18 Area with most cases of VS offenses. Author .......................................................................36 Figure 19 VS offenses and NAIN radius N. Author.................................................................................... 37 Figure 20 VS offenses and NAIN RN. Author ...............................................................................................38 Figure 21 Scattergram: segment length in relation to count crimes. Author ................. 40 Figure 22 Boundaries of case of study: Soho and VS offenses. Author ................................. 42 Figure 23 Gate counting streets in Soho. Author .....................................................................................43 Figure 24 Pedestrian distribution by gender by weekday and weekend. Author........45 Figure 25 Boxplot of total crime counts and NAIN R400m in Soho. Author .................. 46 Figure 26 Visibility Integration Analysis with VS offenses. Author ............................................ 47 Figure 27 Boxplot of visibility integration and crime counts. Author ..................................... 48 6

Figure 28 Isovists generated from the area’s boundaries. Author ............................................ 48 Figure 29 Street Width and count crimes boxplot. Author ............................................................ 49 Figure 30 Segment length and crime counts boxplot. Author ................................................... 50 Figure 31 Street width per segment versus offenses. Author ...................................................... 50 Figure 32 Green areas with sexual offenses. Author ............................................................................... 51 Figure 33 Land use distribution in Soho. Author ...................................................................................... 52 Figure 34 Inactive frontage (dotted lines) on gate counting streets. Author ................... 52 Figure 35 Examples of Soho maps coloured by participants. Soho Survey........................ 53 Figure 36 Examples of results Soho’s survey. Author............................................................................54 Figure 37 Fear-positive urban examples: construction passages .............................................. 57 Figure 38 Fear-negative examples in Soho: children and women............................................ 57 Figure 39 Interrelation between the findings of every scale. Author ..................................... 60 Table 1 Methodology overview. Author.............................................................................................................. 21 Table 2 Crime count, NACH and NAIN (RN and R1500) correlation. Author ...................... 39 Table 3 Segment length and crime count correlation. Author .................................................... 39 Table 4 Multiple Regression Models. Author ................................................................................................41 Table 5 Matrix of fear factors recorded in Soho. Author ..................................................................... 55

7

Glossary Street harassment Street harassment refers to unwanted remarks and abuses that take place in the public sphere received from a stranger and include from ogling, whistling, and touching to a violent crime such as assault or rape (Gardner, 1995).

Sexual harassment Sexual harassment is defined as a violation of the physical integrity and/or autonomy which is related to the primary and secondary sexual characteristics of a person.

(Macmillan, Nierobisz and Welsh, 2000). Sexual harassment is often used in private environments such as work or institutional spaces.

Sexual assault Sexual assault includes any unwanted sexual contact. It mostly include sexual touching but any sexual threats can be considered as sexual abuse. (Sexual Offenses Act, 2003).

Sexual violence Sexual violence is "any sexual act, attempt to obtain a sexual act, or other act directed against a person’s sexuality using coercion, by any person regardless of their relationship to the victim, in any setting.� (WHO, 2017)

Rape The Sexual Offenses Act (2003) states that rape is the penetration of the vagina, anus or mouth of another person with a penis when consent has not been given and the offender do not reasonably believe that the victim consents.

8

Chapter 1  Introduction ‘When walking to a friend’s house on a Saturday at about 6:30pm, two drunk men started following me. One grabbed my hair and said- “‘you are too pretty to be out alone” – they had been shouting at me for the last 100m. I felt violated and arrived shaking.’ (Female in Bates, 2014, p.25) Almost all over the world, women are concerned with fear of violence in public spaces. The global statistics are alarming: one out of three women have experienced physical or sexual violence at some point of their lives (United Nations, 2015). Women make up 90% of the victims of sexual crimes and 99% of those imprisoned for these crimes are male (Eurostat, 2015). Consequently, this investigation focuses on the perception of the fear of gender violence in urban spaces among women.

The least severe but most regular kind of gender violence is street harassment and it is defined as unsolicited compliments, noises and physical abuses that take place in the public sphere received from a stranger (Gardner, 1995). Examples of this invasive behaviour are scrutinising, whistling, and touching the victim. The importance of analysing these occurrences is that they can lead to a violent crime such as assault, rape or a murder (Gardner, 1995).

Figure 1 European sexual violence. Senthilingam, 2018

9

In Europe, 83 to 102 million women (45 - 55% of adult women) have experienced sexual harassment since the age of 15 (European Union Agency for Fundamental Rights, 2014) (see Figure 1). In the United Kingdom, 44% of women have experienced physical or sexual violence.

Even though there is a theoretical debate about the geographical distribution of gender violence, and about whether it is in the private or the public sphere, the reality is that it is an ever-present issue in women’s lives. A separate report by Stop Street Harassment (2016) found that in the United Kingdom, 64% of women have experienced street harassment (see Figure 2). In London, in 2017, more than 40% of women had experienced street harassment during the previous year (YouGov, 2016).

Figure 2 Street harassment in the United Kingdom. Senthilingam, 2018

Although street harassment is considered a crime in some countries such as Portugal and Argentina, in the United Kingdom it is not. According to the Sexual Offenses Act 2013, the only crime related to the activity of street harassment is stalking and following the victim. Exposure, causing sexual activity without consent, sexual assault with or without penetration, voyeurism and rape are categorised as crimes. An example of recorded sub-crimes was published in An Overview of Sexual Offending in England and Wales (Ministry of Justice et al, 2013). As seen in Figure 3, the most frequent crimes were sexual assault and rape on a female and the least ones were the same offenses but with male victims.

10

Recorded sub-crimes 2011-2012 2%

Sexual assault on a female

4%

Rape of a female

11% 37% 18%

Other sexual offenses Sexual activity with minors Sexual assault on a male

28%

Rape of a male

Figure 3 Sub-crimes in England and Wales. Ministry of Justice et al, 2013

In relation to reports of sexual harassment to the police or another authority figure, in the YouGov survey (2017), in London, 11% of women said they had reported at least once and 89% said that they had not. This shows the trend of victims not reporting their experiences in the public sphere because they think that these are socially accepted events, or because of fear of the offender or for other reasons.

Fear of physical and sexual violence weakens women’s autonomy, freedom, and enjoyment in the public sphere and confines their overall development (Day, 1997).

The fact that women change their behaviour and their routines in order to avoid certain areas is a compulsory reason to search for a socio-spatial explanation behind these events. Therefore, this study analyses the general perception of the fear of street harassment in public spaces. More specifically, it aims to understand the relation between the morphology of the streets of Central London and physical and sexual violent threats and crimes. Finally, it examines the area of Soho as a case study of the other scale’s findings.

As Kanes (1992) mentions that street remarks are, to a great degree, situations of opportunity. Even though streets do not cause them, they can play an important role in giving opportunities for them to happen by the way they are configured and related to their users. Therefore, their study and design may protect and reduce the fear women feel while walking in urban spaces in specific conditions. Moreover, they can provide women their right to autonomy, freedom and enjoyment of the urban environment.

11

1.1

Research Aims

Divided by scale of study, this research seeks to: A01

Draw a global picture of urban fear factors in relation to street harassment.

A02

Analyse violent and sexual crime data of Central London in order to understand its relation to the city’s spatial layout.

A03

In Soho, test different urban situations derived from the previous scales to identify if global fears and crime cases occurrence are reflected.

1.2

Research Questions

Based on the aims, the research questions are: Q01

Globally, in relation to street harassment, which urban situations do people fear most?

Q02

In Central London, what is the relationship between street configuration and cases of violent and sexual crimes?

Q03

In Soho, how do the present urban situations explain fear of street harassment and where violent and sexual crimes are likely to happen?

1.3

Research Hypotheses

Finally, the hypotheses corresponding to the questions are: H01

People fear generally the presence of male individuals, low visibility, dense greenery, unkempt urban spaces and inactive building frontages most.

H02

The more segregated the street, the more a violent and sexual crime can happen.

H03

In Soho, there is a set of hypotheses: H03 A

As in the city scale, the more segregated the street, the more a violent and sexual crime can happen.

H03 B

Men are expected to be found more than women moving on the streets.

H03 C

The less visible a street is, the more a violent and sexual crime can happen.

12

Chapter 02  Literature Review The review of the previous literature will begin with perception of safety and fear, overview gender and public space, followed by studies on women’s experiences, street harassment and will conclude with sexual crime.

2.1 Perception of fear In literature, fear is often used as the opposite of security. Technically they are two different concepts. Fear is one of the six basic emotions (Ekman, 1992) and security refers to one of the basic needs of humans. Emotions, in Maslow’s theory, are a primitive need. Therefore, the perception of fear is more instinctive than the realisation of security. Security is socially and culturally variable, whereas fear is universal. This differentiation will be addressed in the following subsection.

Kristen Day (2001) exposed how fear in urban spaces impacts negatively on women’s lives. Even when there is no risk, the idea of them being in danger continues, which is explained by the construction of gender identity for men and women. Her article interviewed male students to determine whether they thought women were vulnerable and the results determined that the majority thought that women were.

Figure 4 Fear’s body map. Nummenmaa, and all (2013)

13

An experiment completed in Finland showed the result of more than ten emotions in the body. These body maps show the areas where subjects feel and react (Figure 4). Fear is often perceived, with elevated heart rate, in the upper chest area and in the head. But what defines the perception of fear among women?

If Warr’s (1984) suggestion that women’s fear of crime is fear of rape is true, then the two components, the social and the environmental, are important. Women fear an offender and the conditions of an offense happening. Their fear may be conscious or unconscious, but is present if the conditions, as learned from other women and past experiences, are present.

Traunmueller (2017) based his investigation on criminological, architectural and urban research that correlated crime and the perception of fear. He proposed methods to quantitatively evaluate theories at a city scale. In his literature review, other researchers (Matei et al, 2001 in Traunmueller, 2017) argued that fear of crime is not limited to actual experience, but even it enters into people’s lives through media. Hence, perception of crime is not restricted to time and space.

Perception of safety Safety, on the other hand, was categorised by Maslow as the second most important stage in his Hierarchy of Needs pyramid (Figure 5). This pyramid was divided in three groups: the basic, psychological and self-fulfilment needs. These needs were hierarchical because in order for the second need to be achieved the first one has to be satisfied.

Figure 5 Safety is a second-ranked necessity (Maslow, 1983). Author

14

Since fear is an emotion, a reaction to a specific environment, safety can be considered as a secondary need that can vary from person to person and culturally between societies. Moreover, its perception can change regarding that person’s gender. Among women, Macmillian, Nierobisz and Welsh (2000) examined the influence of sexual harassment on their perceptions of safety. They argued that these perceptions are related to fears of sexual victimisation and vulnerability. They evaluated the influence on the perception of safety and the results indicated that stranger harassment is more prevalent and extensive than non-stranger harassment.

The role of street design in relation to safety was studied by Rashid et al (2017) by asking: what makes a street to be perceived as unsafe? Their results showed that the street design and the environment are as imperative as the transportation method and the public space quality. Also, building’s façades were important because they provide natural surveillance. Another finding was that women feel unsafe when negative elements of the built environment such as graffiti or used bottles are present. The identification of these elements that women perceive as dangerous are crucial because they affect their patterns of street usage.

The distinction between perceptions of fear and safety is relevant because their reach and consequences are different. Fear, being an emotional reaction extends the physical space. Safety, on the other hand, is more a social construction and is culturally influenced. Even though both perceptions are complex to measure, this investigation focuses on perception of fear.

2.2 Public space and gender According to Ceccato (2016) public spaces are the best environment to reflect the meaning of urban life. To be denominated as ‘public’ a space must complete: being open, accessible to all and should allow self-expression and amusement. They have a geographical bearing, and this means that they extend further their location (Ceccato, 2016). She emphasised that they are not isolated elements and there is a temporary aspect about it, its perception changes according to the time of the day in which it is used.

In relation to its theoretical universal accessibility, the question is whether all users have an equal participation in everyday urban space. As Ceccato mentioned, they reflect a society’s image. Rendell, Penner and Borden (2000) criticised society’s image and how it views most cities through male eyes and how they are designed for them. Any other gender is deeply affected by public space, work environments,

15

transportation systems, neighbourhoods and housing design. These lives are viewed as second class in the access, usage and daily activities in public space (Rendell, Penner and Borden, 2000). But what is the definition of ‘gender’ and how does it differ from the term ‘sex’?

‘Sex –male and female– exemplifies a biological difference between bodies and gender –masculine and feminine– refers to the socially constructed set of differences between men and women.’ (Rendell, Penner and Borden, 2000, p.15) In a broader frame, Rendell explained these terms from feminism, defining it as a form of political practice that unifies action and theory and that is constituted by diverse practices, among them: geography, anthropology, film theory and art history. Feminism being a vast spectrum was narrowed down by Rendell, Penner and Borden (2000) by proposing architectural feminism as a new discourse to challenge the meaning and nature of activities that define gender in different spatial spheres. In the case of this investigation, this ideology covers gender in the feminine perspective and the public sphere.

As Rendell mentioned, space is produced by society and that space is a condition of social production. Therefore, specific places may be ‘sexed’ in correspondence to the biological sex of those who inhabit them, whilst they can also be ‘gendered’ according to the norms, roles and relationships between women and men.

Additionally, feminist geographers, such as Bondi, Massey and Rose, demonstrated that space is produced by gender relations. They maintained that if gender is present in society, the social condition, how people are treated and considered, must influence the spatial environments we use. An illustrative example given by Rendell (2000) is the paradigm of the system of a dominant ‘public male realm of production’ (the city) and a ‘subordinate private female one of reproduction’ (the home). This opposite and patriarchal example is not the complete experience of all urban environments but nevertheless is an association of the general perspective of gender and urban space.

To summarise the arguments in relation to gender and public space, based on the definition of urban space, its theoretical universality is in doubt with regard to the participation of genders other than male, and in how public spaces are sexed and how there is a rational relationship between society and space.

16

2.3 Women’s experience Adding a filter to gender and public space’s frame, there is literature that analyses women’s relation to urban space. This review section draws on the problems of spatial discrimination, male–female power relationships and street harassment.

Kanes (1992), when referring to city streets and their sexual geography, described the interaction among women and men: ‘Armed with society’s tacit approval, men can turn allegedly public city streets into a private male jungle where women are excluded.’ (Kanes, 1992, p. 67) Kanes, Arveda (1991) and other authors, argued how specific male behaviours such as unwanted comments, whistles and obscene gestures (which define part of street harassment) intimidate and violate woman’s self/other boundary and privacy.

Women learn since an early age to be constantly alert, in a conscious or unconscious state, to protect themselves from street remarks. Women, lamentably, become aware that any man can become a potential abuser and any place where men are found alone can threaten their safety. Eventually, Kanes argued that women come to realise that public streets are mostly conceived as men’s property.

Behavioural research (Henley, 1977) had proven that women avoid eye contact, walk more stiffly and restrict their movement in public spaces more than men do. Kanes (1992) explains that if these constant interactions cause stress and fear among women, it is natural for them to change their movement paths, especially at night time.

Mazey and Lee (1983), in a pioneering study on feminist geography, focused on the role of women in space by studying female environmental perception. This field aims to comprehend how humans learn and interact with space (p.37). They discussed the barriers that back then restricted women from walking freely, one of them being street harassment. This problem reflected the power differences in space based on gender. The consequence for environmental perception among women is that public space where men wander freely become hostile and fearful.

In brief, women’s experiences in the public sphere show that the female–male distribution of power is unequal and causes spatial discrimination. This is expressed in multiple ways and street harassment is one of them. These behaviours have proven to affect women negatively, physically and emotionally.

17

2.4 Street Harassment Arveda (1991) analysed the social meanings of men’s harassment of women in public spaces. She proposed multiple social functions of these behaviours and argued how they became a constant environment of sexual violence. Her collection of women’s experiences shows how globally widespread the issue is and how it has become part of their everyday lives. In a discussion of multiple definitions of street harassment she concluded that they violated norms of social behaviour and criticised how the remarks made were often derogatory and insulting.

In relation to the social functions of these unwanted remarks, Arveda (1991) argued that they serve as social controllers, they reinforce spatial boundaries and even make women feel like trespassers in public spaces by intimidating them. Given this social problem, she concluded that more information about the negative effects of this behaviour may reflect a change in wider society.

Mohammed and Van Nes (2017) conducted the first, and so far, study that related Space syntax to street harassment. Space syntax is a set of theories and methodologies developed by Hillier and Hanson (1984–1996) that analyses the relationship between space and social behaviours. The methods have been used widely to explain social phenomena from the spatial perspective. In the author’s case, with an online platform ‘Harassmap’, which allows victims to report sexual abuses, they based their study on the fact that street harassment cases are not equally distributed in Cairo, Egypt.

With a categorization of unwanted remarks and assaults in the area, they identified ‘touching’ reports as the highest case and they ran syntactical analyses and statistics to determine their relation to street segments. Integrated streets, in spaces syntax, are a measure of distance from any street to all others in the system that shows how close the origin is to the rest (Hillier and Hanson, 1984, p.108). In their study, globally integrated streets, in relation to how deep or shallow they were, showed to have a higher touching risk.

In conclusion, their explanation of the unequal distribution of street harassment reports focused on touching. Their results showed that the concentration of these cases is more likely to be found in the more globally integrated routes, where the territoriality is reduced and hence women are more vulnerable to this kind of harassment. For this investigation, it is the starting point of a wider understanding the physical elements of the built environment and the phenomenon in study.

18

2.5 Sexual crime The fact that most public spaces concentrate people is the main reason why they can be home to unpleasant situations and can lead to crime opportunities (Ceccato, 2016). According to the crime opportunity theory, offenders rationalise and choose targets with high rewards and low risk and effort (Cohen and Felson, 1974). A subfield of this theory is called the routine activity (Figure 6) and stated that the incidence of a crime requires the presence of a motivated offender, a vulnerable victim and environmental conditions, specifically no presence of a guardian, or someone who could protect the victim or witness the crime.

Figure 6 Routine activity theory (Cohen and Felson, 1974).

Stanko (1995) examined women’s fear of crime limited in gendered-structured societies. Her definition of fear of crime was the individuals’ diffuse sense of danger about being physically harmed by criminal violence (p. 48). She also criticised gender differences in relation to fear of a crime. Women report fear at levels that are three times that of men (p. 48). Skogan and Maxfield (1984) argued that women’s fear of crime was raised by greater corporeal vulnerability.

In space syntax, Hillier and Sahbaz (2005) compared residential burglary and street robbery in London. They proposed the method of primary risk band analysis, which was used by Mohammed and Van Nes (2017). One of the most recent studies on crime with space syntax methodology was developed by Tarkhanyan (2015) in relation to drug crime and its urban logic. One of Tarkhanyan’s results is important for this investigation: street permeability and proximity to a high street significantly increased the presence of drug crime.

19

Both studies, Hillier and Sahbaz (2005) and Tarkhanyan (2015), show how the morphology of the street network seemed to affect the distribution of crimes, but since specific crimes have specific urban behaviours, as stated by the authors, no study has analysed sexual crimes with this methodology and hence, the importance of this study.

2.6 Summary The literature review has provided the theoretical background and findings on the adjacent themes of interest. The first important conceptual differentiation was between the perception of fear and safety. As stated, fear is universal and more extensive than safety, which is a second range necessity and is culturally influenced.

Based on this clarification, perception of fear will be analysed in further chapters.

The universal accessibility of Ceccato’s definition of public space was criticised, since not all users have equal participation in its every-day use. Similarly, this mirrors the social relations and distribution of power. Hence, it was argued by Rendell, Penner and Borden that public space is sexed and gendered according to those social norms in the greater context. These gender-based inequities and discriminations are the basis of architectural feminism, the ideology that covers this study in order to challenge and explore the borders of these socio-spatial conditions.

In continuation with spatial discrimination, women experiences studies show how these have caused behavioural adaptations (Henley and all, 1974) that have proven to be negative to women’s health and their movement in the streets. These barriers which are frequently considered trivial, have been shown to make public spaces fearful and hostile for women.

As a conclusion of street harassment, the effects they produce in women are, in general, qualified as negative. The only reference found in space syntax was the touching type, which was most frequently reported in globally integrated streets. The need to continue building up this knowledge is the starting point of this study.

Finally, space and crime studies explained how sexual crimes happen and how women perceive this fear. It is important to note the differences between the dimensions of women’s fear of crime and the reasons behind it, which relates directly to the offenses they receive. In space syntax, even though spatial scientists have studied types of crimes in relation to space, no academic works were found in relation to sexual violence and therefore, this study is relevant.

20

Chapter 3  Methodology In order to understand the relationship between street harassment and the urban environment, a study combining quantitate and qualitative analysis was designed.

Table 1 presents the methodology overview with details of each scale highlighted in subsequent sections.

Table 1 Methodology overview. Author

3.1 Global scale In order to gather information on which urban situations are globally perceived as fearful, an online voting system named All Our Ideas was used. The open-source website www.allourideas.org, developed by Salganik and Levy (2015) offers a combination between quantifiable data collection and the input of unexpected information. The method proposed is called a pairwise wiki survey because it selects one out of two options, but it also allows contributions from the participants.

Therefore, the process is the following: once the researcher posts a question,

21

establishes a list of ideas and shares the survey, participants vote between two of these ideas and they can add new as well. In this study, the wiki-survey was shared through email and social media, hence anyone could participate. The question presented to the participants can be seen in Figure 7. The voting scheme was posted with 81 ideas and every idea had a counterpart, for example: for ‘the presence of a male stranger’ and the counterpart was ‘the presence of a female stranger’.

Figure 7 Survey’s interface showed to the participants. All Our Ideas, 2018

The data was analysed by the website in two main phases. First, the answers were used to estimate the opinion matrix, which included the value that each participant had given to one idea. Then, they were summarised to produce a score for the idea that established the probability that it will beat a randomly selected idea by another participant (Salganik and Levy, 2015).

This methodology allowed to collect broad ideas that are perceived as fearful in relation to the theme in study. It is important to mention that the experiment was open to all genders and therefore not only women’s perceptions were taken into account. The results of this experiment provided the conceptual frame the subsequent studies.

22

3.2 City scale: Central London On the city scale, the relationship between the street configuration and cases of violence and sexual offenses were analysed, guided by the findings of the global scale analysis. The two main methods used were space syntax analysis and statistics. They were limited to a range of 2.5 km radii around Tottenham Court Road Station, embracing Central London.

The cases of violence and sexual offenses were collected from the webpage of the Metropolitan Police Service of London (https://www.police.uk/metropolitan/). The data was narrowed to crimes from June 2017 to May 2018 with Excel.

The narrowed (per time and street case) cases were organised by counts of cases per street segment with Excel. These resulting cases were geo-located in QGIS and joined with space syntax analysis in Depthmap (Turner, 2011) of London’s street network. Normalised choice (NACH RN) radius N and integration (NAIN) radius N, 400, 800 and 1500m were used.

Based on the hypothesis of this scale, three different statistical analysis were done all of them had as dependent variable the crime counts per segment (a sample of 580 cases from the Police data). The crime counts were correlated with the following independent variables: segment length, normalised choice and integration radii n (and 400,800,1500m).

Then, a regression analyses with the variables were generated using IBM SSPS Statistics 25. The results of both processes allowed to form multiple regression models with the same software.

The combination of the qualitative data of the global scale with this quantitative scale added complexity to the study, the perception ideas became more tangible in this stage.

3.3 Local scale: Soho The methodology in the local scale interweaves qualitative and quantitative data: map data (Open Street Maps, 2018), space syntax analysis, gate counting and observations on site, photographical records, police crime data, visual graph analysis, isovists, and descriptive statistics. In a combination of these data types, a personal survey was issued to 50 individuals in Soho in June 2018. Since their sex was important to the study, it was recorded, as well as their age range and occupation (see Appendix 6). The participants were asked if they had experienced street harassment and to select five urban situations they would fear. Finally, they were 23

asked to colour the streets of a map in red if they felt they would be fearful, orange if they felt they would be vigilant and yellow if they felt they would be fearless while walking there.

In the second type of analysis, the two data sets were compared, the police data that are officially registered as ‘violent and sexual offenses’. These were correlated with the street network spatial analysis and statistics were applied in order to test the hypotheses, as stated in the beginning of this chapter.

Other Space Syntax methodologies were used to test other ideas rated in the global scale. As in the city scale, the crime data was tested against the spatial configuration of Soho’s street network with statistics. Different scales of integration and choice were tested against the crime counts with regression models.

For further detail about pedestrian movement, gate counting was realised in 30 streets of the area. Gate counting is an observation technique used in space syntax that record the density of pedestrian movement flows and other characteristics of the sample (Al Sayed et al, 2014). The movement of females and males was counted.

These counts were completed once on a week day (Wednesday) and another time on a weekend (Saturday). The gates were added up every hour for five minutes. For security reasons of the researcher, after 21:00 the counting lasted 3 minutes. These counts were normalised by hour and mapped in QGIS. During the process, observations and snapshots were recorded when needed.

Another methods applied were visual graph analysis (VGA) and isovists in order to test visibility with the crime data. VGA (Turner et al. 2001) allows to connect all the visible points of a space in a human-scaled perspective. Visual integration, the measure used, calculates the mean shortest path for all visible nodes in the graph.

An isovist (Benedikt, 1979) is the area of a spatial environment that is visible directly from a specific point. The results of these analyses were overlapped with the photographic registry of each street throughout the gate counting process. Likewise, the area data such as land use, green areas and other urban elements were mapped to compare the global survey’s ideas.

Finally, the maps of the perception of fear in Soho were associated to the quantitative results of the different methodologies applied in order to determine if the hypotheses at this scale were accepted.

24

Chapter 4  Global fear of street harassment The perception of fear of street harassment can vary between individuals and cultures. Hence, with the intention to frame the urban situations that people perceive as fearful, a survey on the All Our Ideas platform was performed. This chapter dives into the results of this survey. In correspondence with this scale’s research question, the hypothesis proposed is that people fear more the presence of male individuals, low visibility, dense greenery, unkempt urban spaces and not active building frontages.

Since the platform is an open survey and hides the identity of the voters, it shows the votes per location (Figure 8). In order to estimate the sample, a calculation was done (see Appendix 1) and in total approximately 190 people voted from all over the world.

Figure 8 Global map showing the votes per country. Allourideas.org (2018)

The voting lasted two weeks and 5902 votes were registered. The ideas about fear of street harassment were 87 in total and six ideas were added by the participants (see full list in Appendix 1). In order to organise the data, the results were divided in two parts: the ones with a score equal and above 50 were fear-positive (46 ideas) and the scores below 49 were fear-negative (41 ideas). This can be seen in Figure 9, bottom ideas are examples, not the total amount.

25

Presence of a drugged male Hidden walkway Presence of group of males Poor-lit streets Construction passages Constuction site Night-time street Presence of a group of male teenagers Presence of a drugged female Delivery trucks parked Empty parks and open spaces Short lines of sight Quiet street Presence of empty bottles Yourself walking on public stairs Nightclub in a building's first floor Presence of a drunk female Presence of public stairs Buildings with fences Graffiti in walls Noisy street Presence of a male and a female Building's first floor is an office Building with doors and windows Presence of two women Presence of a group of females Well illuminated streets

Score

All Our Ideas Scores

100 90

80

70

30

20

Positive

60

50

40

Negative

10

0

Ideas examples

Figure 9 Ideas examples in the global survey and scores. Author

In relation to the fear-positive ideas division, the top fear-positive ideas were:

1. The presence of a drugged male (score 92)

2. The presence of a drunk male (86)

3. Vandalised areas (81)

4. A male standing in a doorway (81)

5. A hidden walkway (80)

26

Conversely, the five bottom rated ideas were: 1. Presence of children (7) 2. Well illuminated streets (8) 3. Building having a restaurant/cafĂŠ on the first floor (10) 4. Presence of women construction workers (13) 5. A group of females (14) The ranking of the results allowed determining which ideas had equal or more odds of being selected as winner in a vote of the survey (defined as fear-positive ideas) and which had odds of losing (fear-negative ideas). Since not only the five top and bottom ideas are important, the ideas were further categorised as the next section will explain.

4.1 Ranking results The division of the total of the results according to the categorisation of the ideas can be organised into pie-charts (Figure 10). On the one hand, the fear-positive ideas are mainly influenced by the presence and the state of people (social, 37%). This is followed by how the urban elements and conditions (33%) were. Mobility (13%) is relatively important as well as the architectural conditions (9%). The less voted ideas were the visibility conditions (9%).

9%

22% 37%

44%

33% 22% 9%

13%

7% 5%

Figure 10 Fear- positive (left) and negative (right) ideas results. Author

In contrast, the fear-negative ideas were differently arranged. The most influential ideas were again the presence of people (44%). Urban situations (22%) and architectural conditions (22%) were tied. Similar to the other group, visibility (7%) was less voted but the less important aspect was mobility (5%).

27

In relation to the globally most feared idea: the presence of a drugged male (score 92), it is important to point that every situation including ‘male’ was voted as fearinducing, except the presence of a male and a female (30) which was fear-negative.

There are differences in the perception of male presence, the rating shows that their state (drugged 92, drunk 86) are the most threatening. Their posture is rated higher if they are standing in a doorway (81), lower if they are sitting on a sidewalk (74) and even lower if they are smoking (69). They are perceived as less fear-inducing if they are a group of teenagers (67), than a group of males (75).

As opposed to the perception of a male presence, situations including ‘females’ were rated as fear-negative, except the presence of a drugged one (65). A drunk female was 20 points less feared, a group of female teenagers was not feared (35), nor was a female stranger (29), which scored 44 points less than its paired idea, a male stranger (73). A group of females (14) are in the bottom of the list, considered as the least fearinducing situations (Figure 11).

Figure 11 Social, visual and mobility categories’ results. Author

The most important midpoint rated ideas were dense greenery (48), presence of a group of people (49) and queuing outside of a bar or nightclub (50) and the presence of empty bottles (50). Even though these ideas do not win or lose the odds of the survey, nightlife environments seem to be shared between multiple fear-inducing

28

ideas and street harassment: drugged (92), drunk (86) and smoking (69) males, nighttime street (68), presence of a drugged female (65), people drinking outside a bar/nightclub (57) and these two last ideas of the mid-point rated ones.

There are substantial differences between paired ideas. For example: absence of people (70) is 20 points fear-inducing whereas crowded sidewalks (45) are only 5 points less feared. In relation to visibility conditions, this is also the case: night-time streets (68) and day-time streets (21) relation shows that night is moderately fearinducing (adding 18 points from the midpoint, whereas daytime streets are 29 points below it). Poor-lit streets (74) compared with well-lit streets (8): insufficient light adds 24 points but adding light reduces fear by 42 points (both calculated from the midpoint). Short lines of sight (57) while long ones (39) are 11 points less fear inducing than the short ones.

In relation to mobility, almost every idea was fear positively rated except parked cars (47) and presence of bicycles (34). The fear-positive leading idea was the presence of motorcycles (68).

Figure 12 Global Survey’s results synthesis:. Author

The pathway type influenced the perception on fear of street harassment. Hidden pathways are rated as fear-positive (80), followed by underground ones (74) and narrow streets (65) are 15 points above the midpoint (Figure 12). These urban conditions relate to low visibility conditions (57-79) mentioned before. In addition,

29

deserted parking areas (77) relates with the fear of absence of people (70) from the social category, as crowded parking (47) relates with crowded sidewalks (45).

Green areas’ ideas majority were not rated as fear-inducing: empty parks (58) was the highest- ranked idea. If they were unkempt (57) they were also fear-positive rated but if they were dense was rated with 48 points. The maintenance of those areas, kempt landscapes (31) adds 19 points of positive relation to street harassment.

Finally, the architectural ideas were also low rated in comparison with the other categories. The most fear-inducing ideas was when a building does not have doors nor windows (64, 14 points above the midpoint), its paired idea had a score of 20 (30 points below) therefore, it can be inferred that adding windows and doors reduces fear perception. The less feared architectural situation is that this floor contains a restaurant or a cafĂŠ (10) which is close to the lowest score, presence of children (7).

4.2 Summary The ranking of the global survey framed the fear factors in relation to street harassment. Social conditions are the most important aspect. The urban and architectural conditions are consecutive important. Mobility and landscape conditions of the urban environment were not perceived as important.

Overall, male presence is rated as the most feared situation, whereas female presence is perceived as fear-negative in almost all their presences. Situations that were expected to be perceived as fearful: dense greenery (49), queuing outside a bar (50) and the presence of empty bottles (50) showed that there is a fear of the unorderly but these do not cause fear of sexual harassment.

Since fear is acknowledged through vision, visibility was top-ranked. People seem to differentiate between elements of low visibility: poor lighting, day-night time and the range of visibility available. In addition, width and length of the streets as well as their use influences the perception of fear of street harassment among its users. The conditions that these have also are related to feelings of fear, vandalised public areas and poor maintenance were top-rated ideas in this category.

Finally, the hypothesis is partially accepted because people fear more the presence of male individuals (67 - 92), low visibility (57 - 79), buildings with no windows nor windows (64) but not dense greenery (48) or unkempt urban spaces (57). This spectrum of ideas will be the comparable basis for other methodologies used in this study.

30

Chapter 5  Violent and Sexual Crimes in London The Police Service offers an overview of crime data in London. With the purpose of identifying if there is a relationship between the city’s street configuration and violent and sexual crime, this chapter analyses space syntax values of the street network in order to overlay these cases and relate them. It is hypothesised that the more segregated the street, the more a violent crime can happen.

5.1 Data selection criteria In Central London, specifically in a radius of 2.5km starting from Tottenham Court Road Station, the crime data from May 2017 - June 2018 was obtained (51 605 crimes, see Appendix 4). This radius is given by the police service and the data was narrowed to the most recent year in order to work with the latest records available. From this total amount, 6 720 (13%) were ‘violence and sexual offenses’ (VS offenses), the third most frequent crime in the area (Figure 13).

Crime Categories in Central London (May 2017- June 2018) Possession of weapons

265

Bicycle theft

1128

Robbery

1891

Vehicle crime

2236

Public order

2331

Burglary

2363

Shoplifting

4617

Violence and sexual offences

6720

Anti-social behaviour

8296

Thefts

21758 0

5000

10000

15000

20000

25000

Figure 13 Crime types. Police (2018)

31

Since this study focuses on street harassment, the location of those crimes was crucial for dividing domestic from public sexual violence to narrow it to street cases.

The leading crime site was on streets (Figure 14).

'Violence and sexual offenses' in Central London (May 2017-June 2018)

Pedestrian Subway

100

Parking area

170

Station

787

Nightclub

974

Building

1064

Street

3992 0

1000

2000

3000

4000

5000

Figure 14 Crime’s location. Police, 2018

The geo-localisation of the violent and sexual offenses data in the study area was visualised in Figure 16, showing the distribution of the 3992 offenses. Some of the offenses are in the same point, therefore 572 cases are represented in the map. As seen, the street cases are spread over the area. Also, the mean amount of cases per month was calculated: 352 cases (see Appendix 3). The months in which more cases are registered are November and December (392 crimes) and the lowest month is April with 313 crimes.

In summary, violence and sexual offenses are the third more frequent cases. From those, the majority happened on streets, in the public sphere. These crime cases are widespread and constant. Hence, an analysis of the street configuration with space syntax measurements may explain the distribution of these cases.

32

Graph 1 Street reports in Central London

Figure 15 Map with VS offenses in Central London. Police (2018)

33

5.2 Crimes distribution In order to understand the spatial behaviour of the phenomena, the fear-positive ideas of the global scale have to be recalled. As seen in Chapter 4, people spatially fear deserted parking lots (score 77), places absent of people (70) and empty parks and open spaces (58). The similarity between these ideas and space syntax is that they often are found in segregated streets. This means that there may be a relationship between the topological measures of segregation (the inverse measure of integration) and the presence of a VS crime in a street segment, as stated in the hypothesis for this scale.

Subsequently, the data shows that there are street segments where multiple cases of sexual crimes have happened. The counts of VS offenses per segment were organised in ranges as seen in figure 20. The majority of streets (more than 300) have one to five cases of VS offenses (Figure 16).

Amount of streets with violent and sexual offenses 350

Crime counts

300 250 200 150 100 50 0

Crime counts in ranges 1<x<5

1<x<10

10<x<20

20<x<30

30<x<40

40<x<50

100<x<150

150<x<350

Figure 16 Amount of street segments VS crimes. Police (2018)

34

Figure 17 Violent and sexual crime counts and their contour. Author

35

As presented in Figure 18, there is an area with concentration of high counts of VS offenses (150-300 cases, last two categories of the area contoursâ&#x20AC;&#x2122; layer) in Leicester Square, Covent Garden and Soho. These higher cases form a triangle between Piccadilly Circus, Leicester Square stations and Shaftesbury Avenue.

Figure 18 Area with most cases of VS offenses. Author

36

5.3 Spatial analysis

Figure 19 VS offenses and NAIN radius N. Author

37

Graphically, these values can be appreciated in Figure 19 and a closer scale in Figure 20, where the categorisation of the crime counts per segment was overlaid with the normalised integration of the area. From the visualisation, it appears that the higher values are located in the more integrated streets whilst the less frequent ones are spread from integrated to segregated street segments. Based on this assumption, this was tested statistically.

Figure 20 VS offenses and NAIN RN. Author

38

In order to determine which syntactic values were associated to the presence of VS offenses, normalized choice (NACH) and integration (NAIN) were correlated with the crime counts. The dependent variable used was crime counts per street segment and the sample size was 580 cases.

NACH

NAIN

NACH

NAIN

RN

RN

R1500

R1500

Crime

Pearson’s R

-.029

.069

.062

.158**

count

Sig. (P-value)

.480

.097

.133

.000

Table 2 Crime count, NACH and NAIN (RN and R1500) correlation. Author

The correlation between normalised angular choice (NACH) and integration (NAIN) radius n and the crime count is not significant. The correlation between NAIN radius 1500m and street crime is significant with an R of 0.158. In addition, a linear regression was modelled with both variables. The results showed that NAIN r1500m has an R Square of 0.007, which means that by one unit change in the amount of crimes, 7% can be considered to be due to the street’s global integration. With these results, the hypothesis is rejected –street segments that are globally more integrated tend to have higher offense occurrence, and the hypothesis was the opposite.

Segment length Another idea that was voted fear-positive in the global scale were short lines of sight (ranking 57). Therefore, street segment’s length was statistically tested to determine if it relates with crime counts. As Table 3 shows, segment length is inversely correlated to the counts per crime, meaning that the shorter the street segment, the more cases of VS offenses.

Segment Length Crime count

Pearson Correlation R

-.207**

Sig. (2-tailed) – P-value

.000

N (sample)

8373

Table 3 Segment length and crime count correlation. Author

The answer can be seen in the Table 3 and Figure 21, street segment length correlates inversely to crime counts in this scale, which means that the shorter the street segment, the more cases of VS offenses can happen.

39

Figure 21 Scattergram: segment length in relation to count crimes. Author

Multiple Regression Models Since normalised integration radius 1500 (NAIN R1500) values correlates with the crime count, three models of multivariate regression with segment length and NAIN (R 400m, 800m and 1500m) were tested (see Table 4).

The correlations show that segment length correlates negatively with crime counts and NAIN R400, 800 and 1500m correlate positively (slightly lower effect). The multiple regression is a small improvement on effects (adjusted R2 explains these).

In model 1 the adjusted R Square is 0.067, so 7% of variation in crime count may be explained by segment length and NAIN400. In comparison with R of 0.207 for segment length, which R2=0.043, hence segment length on its own only explains 4% of variation.

40

Independent Variables

Model 3 0.066

0.062

0.195

-0.196

-0.2

NAIN R400

0.166

Segment Length

Segment Length

-.207**

NAIN R400

.180**

P-value

Model 2

0.067

Beta coefficient

Adjusted R Square

Pearson correlation R

R²

Model 1

R800

0.163

R1500

0.148

R800

-.207**

-.207**

.176**

R1500

.158**

Segment Length

0.000

NAIN R400

0.000

R800

0.000

0.000

0.000

R1500

0.000

Table 4 Multiple Regression Models. Author

As compared, Model 1 is the best improvement of crime counts variance, when short segments and high integration in a local scale is present.

5.4 Summary In synthesis, in the city scale, VS offenses were related with the street network’s configuration. The geo-location of these offenses allowed to detect a concentration of higher counts in Leicester Square, Covent Garden and Soho. Also, a triangle of high crime counts between Piccadilly Circus, Leicester Square and Shaftesbury Avenue was identified. Therefore, the next chapter will focus on this area.

The statistical tests of NACH and NAIN did not explain the global distribution of the crime counts but the local ones (radii 400, 800, 1500m) did. Therefore, the hypothesis is rejected because it is not statistically significant for NACH and it is significant for NAIN R1500 but positively related rather than negatively as hypothesised.

The streets’ segment length was also tested and it is inversely significant correlated with crime counts in the area. Hence, offenses are more likely to happen on shorter segments. Finally, these two variables were tested in a multiple variable regression analysis and three models were tested: the first model performed best, when segment length and NAIN R400 is added. Therefore, crime counts are more likely to happen in a combination of short segments and high integration values in a pedestrian scale.

41

Chapter 6  Local scale: Soho Diving deeper into the perception of fear of sexual offenses, Soho was selected as an in-depth case study because of a high concentration of violent and sexual crimes near the area. How do the present urban situations explain fear of street harassment and where these crimes are likely to happen?

Figure 22 Boundaries of case of study: Soho and VS offenses. Author

42

Figure 23 Gate counting streets in Soho. Author

43

Soho is known for its night life, theatres, pubs and bars. Geographically, it is bounded

by Oxford and Regent Streets, Charing Cross Road and Shaftesbury Avenue (Figure 22). The nearest Underground stations are Tottenham Court Road, Piccadilly and Oxford Circus Stations. Soho is between these areas and therefore is a transition from a working and academic environment to a more touristic and leisure-oriented one.

The violent and sexual crime counts provided by the police are presented in this scale with the aim of demonstrating that the area is part of the high count values identified on the city scale. In fact, the highest count of violent and sexual crimes was located in Shaftsbury Avenue (300 counts per year, as seen in Figure 22).

With the intention of testing the hypotheses of the local scale, the spatial configuration of the streets, gate counts and observations will serve to unravel the riddle.

6.1 Social aspects In order to understand the behaviour of Soho’s socio-spatial context, gate counting and observations were realised in 30 streets. The area included Shaftsbury Avenue to Bateman Streets and Great Windmill Street to Greek Street. Street segments were divided in A, B and C respectively (see Figure 23).

Soho is a very busy area throughout the day. In Figure 24 a comparison between the week and weekend day observations is presented. As seen, on the two days recorded, there were more pedestrians present on a weekday (118 596) than on a weekend day (112 592). The most frequented streets, in order, were Shaftesbury Street (consistent with its NAIN R400 value, 2.02), Wardour Street A (1.89) and Old Compton Street (1.7).

As seen on Figure 24, pedestrian’s gender was recorded. The results show that there is an overall male predominance on Soho’s streets usage. On weekdays, males dominated 28 streets, for example: Winnet St. (69%), Tisbury Court (69%) and Peter St. (67%). On weekends, males overshadow females in 26 streets. On the other hand, male are dominant that doubles that of females in Winnet St (81%), Walkers Court (68%) and Bateman St B (66%).

44

Figure 24 Pedestrian distribution by gender by weekday and weekend. Author

45

Female presence, on the other hand, outnumbered that of males on the weekday sample, on two street segments: Greek Street B (56%), and C (57%). On weekends, females were slightly more frequently counted in Shaftesbury Avenue (52%), and Wardour A (52%).

In general, children were less observed. During the week day, none were counted in streets like Winnet St, Tisbury Court and Rupert St (figure 25). The following streets were used by children the most: Archer St (21%) and Shaftesbury Avenue (15%) as well as Wardour St A (6%). On weekend observations, streets where none was present: Dean St A and Bateman St B. Streets where theatres were near: Wardour St B, Greek St B (8%) and C (8%) were used by many children.

Figure 25 Boxplot of total crime counts and NAIN R400m in Soho. Author

As seen in Figure 25 (based on Appendix 7), when crime counts of Soho are plotted with normalised integration radius 400m (statistically significant in city scale), the values cluster in a medium range of integration (1.33-1.56) and not all in the most integrated as the previous scale.

46

6.2 Visibility Low visibility, derived from the global survey, was rated as a fear-positive factor. The result of a visibility integration analysis of Soho is shown in Figure 26. The most visually integrated streets are Regent (4.29) and Oxford Street (4.63).

The second ones are Charing Cross (3.19) and Wardour Street (3.39), where 51 sexual offenses were reported in total. Streets with medium values (2.61-295) have 21 crime spots, less visible ones (1.60-1.94) have multiple crime cases (25 crime spots). Also, it multiple crime spots are located on street corners.

In order to define which visibility values concentrate more cases of VS offenses, a box plot graph was generated grouping the values in 5 categories (figure 27). In the area, the most visually segregated streets are more likely to present a violent and sexual crime.

Figure 26 Visibility Integration Analysis with VS offenses. Author

47

Figure 27 Boxplot of visibility integration and crime counts. Author

As a next step, isovists were generated from the most visibly integrated streets.

Figure 28 shows how Sohoâ&#x20AC;&#x2122;s visual permeability is reduced: the more visible ones from the outside are Dean Street and Greek Street.

Figure 28 Isovists generated from the areaâ&#x20AC;&#x2122;s boundaries. Author

48

In synthesis, there is a strong relation between the street network’s integration, its visibility and the area’s permeability in relation to sexual crime counts.

6.3 Urban conditions According to the global survey, narrow pathways and short lines of sight were voted as fear-inducing. In relation to the first one, a concentration is found near Regent and Great Windmill Street. In both cases there incidents of sexual violence were reported (52 and 35 cases). However, Shaftesbury Avenue and other main streets are wide and more than 300 counts of crimes were reported in these locations. In order to understand this, a boxplot with Soho’s street width and crime counts was generated.

It results that the streets with more cases are 7.0-10.0m wide. Therefore, no association or logic between street width and offenses can be found.

Figure 29 Street Width and count crimes boxplot. Author

Secondly, short lines of sight depends on street length. Hence, a boxplot with street’s length and crime counts was produced (Figure 29). In congruence with the city scale, the higher count of VS offenses were located in shorter segments (1-30m), as seen in Figure 31.

49

Figure 30 Segment length and crime counts boxplot. Author

Figure 31 Street width per segment versus offenses. Author

50

Soho contains three main green areas: St. Annieâ&#x20AC;&#x2122;s Church Park, Soho and Golden Square. As seen in Figure 32, it is not area with intensive greenery and most offenses were unrelated to these green areas. In synthesis, green areas proximity and segment width are not as important as segment length in relation to crime locations in this scale.

Figure 32 Green areas with sexual offenses. Author

6.4 Architectural A buildingsâ&#x20AC;&#x2122; relation to urban space was rated as relevant in the global survey. In Soho, land uses are mostly restaurants (43%) and retail (35%), as seen in figure 33.

According to the figure, most VS offenses are mostly related to retail buildings and not to bars and nightclubs as stated in the global survey. Due to its leisure-oriented nature, theatres (2%), bars and pubs (10%) are also found frequently in the area.

51

Figure 33 Land use distribution in Soho. Author

Another fear-positive idea was the absence of windows and doors (Figure 34).

Theatres in Soho have long inactive facades, for example: Greek Street A and Wardour Street B. In six of these violent and sexual offenses were registered.

Figure 34 Inactive frontage (dotted lines) on gate counting streets. Author

52

6.5 Perception of fear of street harassment “Harassment arrives in busy streets whereas I am really scared of quiet and small streets.” (Soho’s survey participant, 2018) Fear of street harassment was directly surveyed in Soho, as showed in figure 35. In order to contemplate their perceptions a categorisation with the survey’s colours used was completed (Figure 36 shows six examples, see Appendix 6). The participant’s evaluation of the streets pointed to multiple interpretations of fear of street harassment in the area. Nevertheless, the majority (65%) agreed that most integrated streets, such as Shaftesbury Avenue (70%) and Wardour Street (50%), are perceived as fear-negative. This is contradictory to Chapter 5, since on the city scale, integrated streets showed significant higher cases of sexual offenses.

Figure 35 Examples of Soho maps coloured by participants. Soho Survey

In contrast with this, small alleys like Meard Street, were perceived as fear-inducing (80% orange and 8% red), in line with the global survey. Greek Street, on the other hand, has fewer active frontages this could relate to its negative perception. Another area that was evaluated as fearful multiple times was Soho Square, which may be related to the global perception that empty, closed parks and open spaces that can be perceived as fear-inducing.

Old Compton Street was slightly (2% more) evaluated as fear-inducing, whereas others coloured it as fear-negative (48%). It is important to mention that Soho is also known to be a gay community hub, and therefore gay bars and nightclubs are frequent in these street’s segments.

53

Figure 36 Examples of results Sohoâ&#x20AC;&#x2122;s survey. Author

54

6.6 Summary The main findings of the local scale are that, in Soho, males are more likely to be found on the streets. On the contrary of the city scale, sexual offenses were more likely to occur in medium ranged integrated streets, but in line with that scale, they were found in the shortest segments. Furthermore, visibility graph analysis showed that sexual offenses were more likely to happen in the less visibly integrated streets.

The relation between the global survey and the city networkâ&#x20AC;&#x2122;s findings in Soho showed how a scenario of street harassment and an opportunity of a sexual crime is a combination of more than one variable. Therefore, Table 5 illustrates a summary of the dependent and independent variables tested with the results of Soho survey. The first variable is the crime count per street and the independent ones are organised according to the global survey. The perception of street harassment is provided by the Soho survey.

Table 5 Matrix of fear factors recorded in Soho. Author

55

As seen, the fear-positive values are highlighted, except in variables such as presence of women and children and in binary results: visual permeability and buildingâ&#x20AC;&#x2122;s frontages (See Figure 37). Two examples of streets where multiple crimes are found and the outcome of their variables is as the findings of the scales showed are Shaftesbury Avenue and Wardour Street. Both streets are highly integrated and their street segment width is short. Nevertheless, they present variables that were ranked as fear-negative such as presence of children and high visibility.

Streets that show the ideal scenario for street harassment and that were evaluated as fear-inducing in the Soho survey are Winnet and Romily Streets. They are spatially segregated, male dominated, short in length and width and their visibility is low.

Greek street, on the other hand, shows the best scenario for a fear-negative experience: females and children in the street, visual permeability from the main streets (See Figure 38). Nevertheless, they were perceived as fear-inducing in the Soho survey.

In conclusion, since the set of hypotheses of this scale is the following: H03 A The more segregated the street, the more a VS crime can happen.

H03 B Men are expected to be found more than women moving on the streets.

H03 C Visibility relates to the crime counts, the less visible a street is, the more a VS crime can happen.

In Soho, the first sub-hypothesis is rejected because the streets that presented the higher amount of VS cases were in a middle range of integration. However, the second one is accepted, men are the predominant users of the area (out of 30 streets, 28 were male-dominated on a weekday, and 28 on a weekend sample). Finally, the last one is also accepted because the analysis showed that sexual offenses were more likely to happen in the less visibly integrated streets.

56

Figure 37 Fear-positive urban examples: construction passages

Figure 38 Fear-negative examples in Soho: children and women

57

Chapter 7  Discussion This section brings together results from all scales. In first place, the global hypothesis was partially accepted because it was confirmed that people fear the presence of male individuals, low visibility, and inactive building frontages. However, greenery was not as important as hypothesised.

The global fear-positive perception of “male” containing ideas can be related to Kanes (1992) argument that any male stranger is seen by females as a possible threat. The same perception of the presence of men in various states has a direct relationship with the statistics that women are 90% of the victims of sexual crimes and 99% of the imprisoned are males (Eurostat, 2015). Therefore, the global fear of the presence of male strangers is consistent with official data. On the contrary, children and groups of women were bottom-rated as fear-negative; this can be explained because they are globally perceived as harmless and vulnerable as Day (2001) showed.

As Rashid et al (2017) stated, the role of design influences the perception of its users in relation to street harassment. The quality and conditions of it are imperative, therefore it explains why if it is vandalised (81) or hidden (80), they were rated high in the ranking. Also, the authors found that building’s facades were important because they provide natural surveillance, nevertheless, the ranking of this was not high.

In relation to the tension between fears of the unorderly, vandalized urban spaces, presence of bottles is also a finding of Rashid et al (2017) study, women feel unsafe when negative elements of the urban environment are present. Nevertheless, graffiti was ranked as fear-negative and it was perceived as one of those factors in their study. In a global perspective, the state of the urban environment can cause a negative reaction on their users but it was confirmed that they do not harass women.

Visibility, on the other hand, is a condition of the environment that, as shown in the local scale, affects the perception of fear. Their range of visibility was carefully ranked by the participants: insufficient light rises the perception of fear as double as if they are well-illuminated. Hence, light design is an important suggestion to urban designers. Also, given the fact that women change their routes especially at night (Kanes, 1992) is a direct link to the perception of streets during this times as fear inducing. This was corroborated in Soho, where male are more present at night (See Appendix 8). In addition, the line of sight was also important, short lines of sight were fear-positive.

58

In relation to visibility distance, in Central London the spatial configuration was tested with sexual crime data. It results that crime counts is inversely correlated to the length of the street segment. Furthermore, normalised integration in pedestrian scales is correlated with these offenses. Hence, the hypothesis which predicted that the more segregated the street, the more a violent and sexual crime can happen, was rejected. The correlation with integrated streets follows the findings of other space syntax studies such as Mohammed and Van Nes (2017) and Tarkhanyan (2015), in relation to high streets.

The most frequent scenario in a street segment of Central London is one to five sexual crimes. On the contrary, there are clusters where hundreds of cases have been registered. Why is there a correlation between violent crimes, locally integrated streets and short segments? One explanation is given by the routine activity theory (Cohen and Felson, 1974), if the principal victims are females and offenders are men, they need a location that is highly accessible but whose visibility is low, which leads to short segments as the best scenario.

Soho, as Mort (1998) describes, has been historically an artistic, leisure-oriented and youth-related area. Given the number of tourists that visit the area, they could also be targeted as a victim. As seen on the global scale, nightlife related environments are perceived as fear-inducing, hence the land uses in the area can influence the likelihood of a violent crime happening.

In this scale, NAIN 400m did not explain the crime counts: if the street is less integrated, fewer guardians are present and the likelihood of a crime may be higher.

High integration and crowdedness are related. Therefore, movement flow is predominantly male in the gate-counted area, which can show how the rest of Soho behaves. Hence, if the area is used mainly by males, and they are highly perceived as fear-inducing, it is explained partly why Soho is part of Central Londonâ&#x20AC;&#x2122;s high sexual crime clusters.

In relation to the visibility conditions, fear, as confirmed in the global survey was reflected on a local scale. The analysis showed that the more segregated visually the street more likely it was for a violent and sexual crime to happen. Nevertheless, these scenarios are not perceived in the personal surveys on site.

Finally, this investigation followed the thread of the global survey but each scale revealed different results. There are inconsistencies between scales and the perception of a threat of sexual violence and a sexual crime itself. Though, Sohoâ&#x20AC;&#x2122;s deeper analysis provided more answers and questions to the theme in hand.

59

Chapter 8  Conclusions 8.1 Research Conclusion The starting point of this investigation was the fear of being harmed, psychologically or physically, among women in urban spaces. The first and most important conclusion is that, even though street harassment affects women primordially, men are also responsible and actors of a change in it. In this study, they participated in the global study, which grew into different analyses. Since research and actions on this problem is a social need, together with every other gender, more can be achieved.

Globally, the conclusion is that even though the most voted component of the survey were the social components of it, the spatial layout analysis showed how interrelated they are. Moreover, the methodology of the online survey allowed a richer scope of ideas and their evaluation with other space syntax methods in further scales. The following summary diagram shows the findings and their relation:

Figure 39 Interrelation between the findings of every scale. Author

60

Violent and sexual crimes are the third most frequent crime in Central London.

Hence, this study provides high value to the authorities and to London’s society in general because this methodology can be replicated in different areas of the city.

Moreover, understanding these crimes will help many users and designers to create more welcoming and fear-negative environments for women, based on the conditions that have been found as fear-inducing in this study.

The detailed study of Soho allowed testing the global abstract ideas and their relationships. These results can help policy and urban designers to know the specific factors that need to be addressed when designing. In addition, one important method learned from the study was contact and observe the actual usage of the area’s urban space in order to have an inner and personal perspective of the experience of walking these streets.

8.2 Limitations On the global scale, All Our Ideas, as an open survey, hides voters’ identity and shows only the votes per location. Another limitation was that votes were received from different cultures and the perception of fear of street harassment can be different.

Due to the researcher’s network, most votes were located in America and Europe, nevertheless, every continent provided at least one vote.

In relation to the limitations on the city scale, the police gives an overview of antisocial behaviour and crimes of London, hence the information is provisional. In addition, their data reflect their activity, which is not equal in every street. Also, the location provided is on or near the exact location where the crime happened, so these locations are approximated.

The categories provided (burglary, drugs and others) include a wide range of subcrimes on each category. The violence and sexual offenses category comprises offences against the person such as common assaults, grievous bodily harm and sexual offenses (Home Office and Police.uk, 2018). This means that physical abuses are combined with sexual ones and there is no description of which is which.

Although there is a direct relationship between street harassment and sexual offenses, some types of street harassment, as verbal abuse, are not stated as crimes and therefore, sexual crimes do not reflect street harassment scenarios. A differentiation between these concepts had to be stated in the investigation.

61

The police data, the motivating reasons and the consequences of physical and sexual crimes are different, moreover after this investigation, it can be shown that the spatial configuration and conditions present in an environment can influence in the incidence of a crime and therefore, the division of these crimes should be made in order to continue the study.

On the local scale, there were three main limitations. First, Open Street Maps data are not completely actualised and are sometimes incorrect. Gate counting and observations on site provide only a sample of the activity of the area: even though they were done constantly, they are still only an approximation of the real scenario.

Secondly, the sexual preference of the pedestrians counted was not taken into consideration, since it is not physically evident. Hence, some male dominance may be affected. Thirdly, in relation to the survey, it is not representative of Sohoâ&#x20AC;&#x2122;s population, consequently it is a qualitative approach to the perception of fear in the area.

8.3 Further analysis Different types of analysis that can be developed from this research. One example is the experiment of testing and recording the behaviour and reactions women have in urban spaces. Another analysis that can be followed is the visual perception of the fear of sexual crime, this was part of the pilot study of this investigation. Actually, a comparison between urban situations in London would be interesting to identify specifically which are the situations that women perceive as fearful in relation to street harassment.

In conclusion, this investigation has provided the factors that combined characterise the fear women feel while walking in certain conditions. The understanding of this fear may be used to empower society to change urban physical aspects that give the opportunity for street harassment and sexual crimes. Only then, and with a social agreement of the importance of womenâ&#x20AC;&#x2122;s equal freedom and enjoyment of the urban space, will changes be seen.

62

References Al_Sayed, K., Turner, A., Hillier, B., Iida, S., Penn, A., 2014 (4th Edition), “Space Syntax Methodology”, Bartlett School of Architecture, UCL, London.

All Our Ideas. (2018). In relation to street harassment, which of these urban situations do you fear more? [online] Available at: http://www.allourideas.org/fearsyntax [Accessed 11 Aug. 2018].

Benedikt, M.L., 1979, “To take hold of space: Isovists and isovist fields”, Environment and Planning B, 6: 47 – 65 Bates, L. (2016). Everyday sexism: The project that inspired a worldwide movement.

Macmillan, New York.

Baxter

K.

(2016)

|

Illustrated

Maps.

Soho.

[online]

Available

at:

http://www.katherinebaxter.com/soho/ [Accessed 12 Aug. 2018].

Ceccato, V., 2016. Public Space and the Situational Conditions of Crime and Fear.

International

Criminal

Justice

Review

26,

69–79.

https://doi.org/10.1177/1057567716639099 Cohen, L.E., Felson, M., 1979. Social Change and Crime Rate Trends: A Routine Activity Approach. American Sociological Review 44, 588. https://doi.org/10.2307/2094589 COST Action 1206 - Femicide. (2018). COST Action 1206 - Femicide. [online] Available at: https://www.femicide.net/ [Accessed 15 Aug. 2018].

Day, K 1997, 'Better safe than sorry? Consequences of sexual assault prevention for women in public space' Perspectives on Social Problems, vol 9, pp. 83-101.

Day, K., 2001. Constructing Masculinity and Women’s Fear in Public Space in Irvine, California.

Gender,

Place

&

Culture

8,

109–127.

https://doi.org/10.1080/09663690120050742 Durning L. and Wrigley R. (2000) Gender and Architecture. John Wiley and Sons, LTD. Chichester, New York, Weinheim, Brisbane, Singapore, Toronto.

Ec.europa.eu. (2018). Violent sexual crimes recorded in the EU. [online] Available at: http://ec.europa.eu/eurostat/web/products-eurostat-news/-/EDN-201711231?inheritRedirect=true [Accessed 11 Aug. 2018].

63

European Union Agency for Fundamental Rights. (2014). Violence against women: an EU-wide

survey.

Main

results

report.

[online]

Available

at:

http://fra.europa.eu/en/publication/2014/violence-against-women-eu-wide-surveymain-results-report [Accessed 11 Aug. 2018].

Hillier, B. 1996. Space is the machine: A configurational theory of architecture. London, Press Syndicate of the University of Cambridge.

Hillier, B. & Hanson, J. (1984), The Social Logic of Space, Cambridge University Press: Cambridge. pp. 108-109 Hillier, B. and Sahbaz, O. (2005) High resolution analysis of crime patterns in urban street networks: and initial statistical sketch from an ongoing study of a London boroughâ&#x20AC;&#x2122;.

Hollaback! (2018). About | Hollaback! We Have the Power to End Harassment. [online] Available at: https://www.ihollaback.org/about/ [Accessed 12 Aug. 2018].

Gardner, C.B., 1995. Passing By: Gender and Public Harassment. University of California Press.

Kanes L. (1992) Discrimination by design: a feminist critique of the man-made environment. University of Illinois Press, Urbana and Chicago.

Nummenmaa, L., Glerean, E., Hari, R. and Hietanen, J. (2013). Bodily maps of emotions. Proceedings of the National Academy of Sciences, 111(2), pp.646-651.

Mazey, M. E., & Lee, D. R. (1983). Her space, her place: a geography of women.

Washington, D.C., Association of American Geographers.

Meera Senthilingam, C. (2018). Sexual harassment: How it stands around the globe. [online]

CNN.

Available

at:

https://edition.cnn.com/2017/11/25/health/sexual-

harassment-violence-abuse-global-levels/index.html [Accessed 12 Aug. 2018].

Ministry of Justice, Home Office and the Office for National Statistics (2013) An Overview of Sexual Offending in England and Wales. Statistics Bulletin. Retrieved from https://www.gov.uk/government/statistics/an-overview-of-sexual-offending-inengland-and-wales (accessed 6.13.18).

Mohamed, A.A., van Nes, A., 2017. Street morphology as a starting point for understanding sexual harassment. XI SSS: 11th International Space Syntax Symposium 2017.

64

Mort F. (1998) Cityscapes: Consumption, Masculinities and the Mapping of London since

1950.

Urban

Studies,

[online]

p.

http://journals.sagepub.com/doi/abs/10.1080/0042098984600

Available [Accessed

15

at: Aug.

2018].

Police.uk. (2018). Bloomsbury, Metropolitan Police Service - Police.uk. [online] Available at: https://www.police.uk/metropolitan/E05000129/ [Accessed 12 Aug. 2018].

Rashid, S.A., Wahab, M.H., Rani, Ismail, S., 2017. Safety of street: The role of street design. AIP Conference Proceedings 1891, 020008. https://doi.org/10.1063/1.5005341 Rendell, J., Penner, B., Borden, I., 2000. Gender Space Architecture: An interdisciplinary

introduction.

Taylor

&

Francis,

Abingdon,

UK.

https://doi.org/10.4324/9780203449127 Salganik MJ, Levy KEC (2015) Wiki Surveys: Open and Quantifiable Social Data Collection. PLoS ONE 10(5): e0123483. https://doi.org/10.1371/journal.pone.0123483 Sexual Offences Act 2003 Legislation.gov.uk. (2018). Sexual Offences Act 2003. [online] Available at: http://www.legislation.gov.uk/ukpga/2003/42/part/1 [Accessed 10 Aug. Stanko, E.A., 1995. Women, Crime, and Fear. The Annals of the American Academy of Political and Social Science 539, 46–58.

Stop Street Harassment. (2016). The UK's First National Street Harassment Study | Stop

Street

Harassment.

[online]

Available

at:

http://www.stopstreetharassment.org/2016/03/uknationshstudy/ [Accessed 11 Aug. 2018].

Tandogan, O., Ilhan, B.S., 2016. Fear of Crime in Public Spaces: From the View of Women

Living

in

Cities.

Procedia

Engineering

161,

2011–2018.

https://doi.org/10.1016/j.proeng.2016.08.795 Traunmueller, M.W., 2017. A Quantitative Approach to Evaluate and Develop Theories on (Fear of) Crime in Urban Environments. University College London.

Turner, A., 2011, “UCL Depthmap: Spatial network analysis software”, version 10.14 (London: University College London, VR Centre of the Built Environment, 2011)

65

Turner, A., Doxa, M., O’Sullivan, D., and Penn, A., 2001, “From isovists to visibility graphs: a methodology for the analysis of architectural space”, Environment and

Planning

B

28(1):

103--121.

http://www.vr.ucl.ac.uk/publications/turner2001-000.html UN Women. (2018). Facts and figures: Ending violence against women. [online] Available

at:

http://www.unwomen.org/en/what-we-do/ending-violence-against-

women/facts-and-figures [Accessed 12 Jun. 2018].

World Health Organization. (2011) Glossary of terms and tools. [online] Available at: http://www.who.int/gender-equity-rights/knowledge/glossary/en/ [Accessed 10 Jun. 2018].

World Health Organization. (2017). Violence against women. [online] Available at: http://www.who.int/news-room/fact-sheets/detail/violence-against-women [Accessed 15 Aug. 2018].

YouGov: What the world thinks. (2012). YouGov | Sexual harassment in the capital . [online]

Available

at:

https://yougov.co.uk/news/2012/05/25/sexual-harassment-

capital/ [Accessed 11 Aug. 2018]

66

Appendixes Tables, graphics, and other supplementary data is provided in this section in order to broaden the understanding of the investigation.

Appendix 1 Source: All Our Ideas, 2018

Categorisation of the votes per country and continent:

394

Canada

185

San Salvador Costa Rica

11

Africa

Country

USA

Votes

Algeria

29

Egypt

38

Kenia

2

2928

69

Colombia

63

Peru

32

Brazil

9

Asia

3518 India

33

Malaysia

35

Indonesia

3

Japan

4

Sweden

52

Denmark

55

Russia

23 9

Germany

39

France

29

Portugal

68

England

521 796

75

Australia

104

Liechtenstein Europe

Votes

Australia

55 55

Approx. amount of voters Total

South America

North America

Country

Amount of votes Average vote per country (except high counts: USA, CA, CR, GB)

5902 31 190

67

Total result of ideas and scores and categories (by the author) id

Idea

Score

Category

1

Presence of a drugged male

92

Social

2

86

Social

3

Presence of a drunk male Vandalized areas (damaged public property)

81

Urban

4

Male standing in a doorway*

81

Social

5

Hidden walkway

80

Urban

6

Deserted train platform*

79

Urban

7

Poor lit ATMs*

79

Visibility

8

Deserted parking lot

77

Urban

9

Presence of group of males

75

Social

Presence of two men Presence of a male sitting on a side of the sidewalk

75

Social

74

Social

74

Visibility

13

Poor-lit streets Exercising outdoors in between bushes or empty lots*

74

Social

14

Underground passageway

74

Urban

15

Presence of a male stranger

73

Social

16

Construction passages

71

Urban

17

Garbage/Rubbish area

71

Urban

18

Presence of construction men

71

Social

19

Construction site

70

Urban

20

Absence of people

70

Social

21

Presence of a male smoking

69

Social

22

Night-time street

68

Visibility

23

Presence of motorcycles passing by

68

Mobility

24

Presence of beggars or homeless people

67

Social

25

Presence of a group of male teenagers

67

Social

26

Low speed cars passing by

67

Mobility

27

Urban space's poor maintenance

67

Urban

28

Presence of a drugged female

65

Social

29

Narrow street pathway

65

Urban

30

64

Architectural

31

Building with no doors nor windows Person standing next to a shuttered building*

62

Social

32

Delivery trucks parked

62

Mobility

33

Public bathrooms usage

60

Architectural

34

Informal commerce

58

Architectural

35

Empty parks and open spaces

58

Urban

10 11 12

68

36

People drinking outside a bar/nightclub

57

Social

37

Unkempt landscape (parks, green areas)

57

Urban

38

Short lines of sight

57

Visibility

39

Service area of a building

55

Architectural

40

Commercial van stopped at traffic light*

55

Mobility

41

Presence of taxi drivers parked

53

Mobility

42

Quiet street

52

Urban

43

Closed parks with fences

52

Urban

44

High speed cars passing by

52

Mobility

45

Presence of empty bottles

50

Urban

46

Yourself queuing outside a bar/nightclub

50

Social

47

Presence of a group of people

49

Social

48

Dense greenery (bushes/trees)

48

Urban

49

Yourself walking on public stairs

48

Social

50

Presence of cars parked

47

Mobility

51

People standing outside tube stations

47

Social

52

Nightclub in a building's first floor

46

Architectural

53

People queuing outside a bar/nightclub

45

Social

54

Crowded sidewalks

45

Social

55

Presence of a drunk female

45

Social

56

Yourself queuing for a bus

44

Social

57

Crowded parking lot

42

Urban

58

Presence of public stairs

41

Urban

59

Wide street pathway

39

Urban

60

Long lines of sight

39

Visibility

61

Buildings with fences

37

Architectural

62

Bar/pub in the first floor

36

Architectural

63

Presence of people queuing for a bus

35

Social

64

Graffiti in walls

35

Urban

65

Presence of a group of female teenagers

35

Social

66

Presence of bicycles passing by

34

Mobility

67

Noisy street

32

Urban

68

Kempt landscape (parks, green areas)

31

Urban

69

Presence of a female smoking

31

Social

70

Presence of a male and a female

30

Social

71

Presence of a female stranger

29

Social

72

Crowded parks and open spaces

28

Urban

73

Building's first floor is an office

27

Architectural

74

Female sitting on a side of the sidewalk

27

Social

69

75

Building with mixed uses (residential/office and commerce)

23

Architectural

76

Day-time street

21

Visibility

77

Building with doors and windows

20

Architectural

78

Building's first floor is commercial

19

Architectural

79

Building's first floor is residential

18

Architectural

80

Presence of two women

16

Social

81

Presence of a couple kissing

15

Social

82

Urban space's good maintenance

14

Urban

83

Presence of a group of females

14

Social

84

Presence of construction women

13

Social

85

Restaurant/cafĂŠ on the first floor

10

Architectural

86

Well illuminated streets

8

Visibility

87

Presence of children

7

Social

*Ideas uploaded by the voters: 1.

2.

3.

4.

5.

6.

Commercial van stopped at traffic light Person standing next to a shuttered building Exercising outdoors in the middle of bushes or empty slots Deserted train platform Poorly lit ATMs Male standing in doorway

Calculation of the pie-charts of the count total Fear-positive count total 46 17

Categories 37%

Social

6

Mobility

13%

4

Visibility

9%

15

Urban

4

Architectural

33% 9% 100% Fear-negative count total

41

Categories

18

Social

44%

2

Mobility

5%

3

Visibility

7%

9

Urban

22%

9

Architectural

22% 100% 70

id

Idea

Score

Category

1

Presence of a drugged male

92

Social

2

Presence of a drunk male

86

Social

4

Male standing in a doorway*

81

Social

9

Presence of group of males

75

Social

10

Presence of two men

75

Social

11

74

Social

13

Presence of a male sitting on a side of the sidewalk Exercising outdoors in between bushes or empty lots*

74

Social

15

Presence of a male stranger

73

Social

18

Presence of construction men

71

Social

20

Absence of people

70

Social

21

Presence of a male smoking

69

Social

24

Presence of beggars or homeless people

67

Social

25

Presence of a group of male teenagers

67

Social

28

Presence of a drugged female

65

Social

31

Person standing next to a shuttered building*

62

Social

36

People drinking outside a bar/nightclub

57

Social

46

Yourself queuing outside a bar/nightclub

50

Social

47

Presence of a group of people

49

Social

49

Yourself walking on public stairs

48

Social

51

People standing outside tube stations

47

Social

53

People queuing outside a bar/nightclub

45

Social

54

Crowded sidewalks

45

Social

55

Presence of a drunk female

45

Social

56

Yourself queuing for a bus

44

Social

63

Presence of people queuing for a bus

35

Social

65

Presence of a group of female teenagers

35

Social

69

Presence of a female smoking

31

Social

70

Presence of a male and a female

30

Social

71

Presence of a female stranger

29

Social

74

Female sitting on a side of the sidewalk

27

Social

80

Presence of two women

16

Social

81

Presence of a couple kissing

15

Social

83

Presence of a group of females

14

Social

84

Presence of construction women

13

Social

87

Presence of children

7

Social

71

id

Idea

Score

23

Presence of motorcycles passing by

68

Mobility

26

Low speed cars passing by

67

Mobility

32

Delivery trucks parked

62

Mobility

40

Commercial van stopped at traffic light*

55

Mobility

41

Presence of taxi drivers parked

53

Mobility

44

High speed cars passing by

52

Mobility

50

Presence of cars parked

47

Mobility

66

Presence of bicycles passing by

34

Mobility

id

Idea

Score

Category

Category

7

Poor lit ATMs*

79

Visibility

12

Poor-lit streets

74

Visibility

22

Night-time street

68

Visibility

38

Short lines of sight

57

Visibility

60

Long lines of sight

39

Visibility

76

Day-time street

21

Visibility

86

Well illuminated streets

8

Visibility

id

Idea

Score

Category

3

Vandalized areas (damaged public property)

81

Urban

5

Hidden walkway

80

Urban

6

Deserted train platform*

79

Urban

8

Deserted parking lot

77

Urban

14

Underground passageway

74

Urban

16

Construction passages

71

Urban

17

Garbage/Rubbish area

71

Urban

19

Construction site

70

Urban

27

Urban space's poor maintenance

67

Urban

29

Narrow street pathway

65

Urban

35

Empty parks and open spaces

58

Urban

37

Unkempt landscape (parks, green areas)

57

Urban

42

Quiet street

52

Urban

43

Closed parks with fences

52

Urban

45

Presence of empty bottles

50

Urban

48

Dense greenery (bushes/trees)

48

Urban

57

Crowded parking lot

42

Urban

58

Presence of public stairs

41

Urban 72

59

Wide street pathway

39

Urban

64

Graffiti in walls

35

Urban

67

Noisy street

32

Urban

68

Kempt landscape (parks, green areas)

31

Urban

72

Crowded parks and open spaces

28

Urban

82

Urban space's good maintenance

14

Urban

id

Idea

30

Building with no doors nor windows

64

Architectural

33

Public bathrooms usage

60

Architectural

34

Informal commerce

58

Architectural

39

Service area of a building

55

Architectural

52

Nightclub in a building's first floor

46

Architectural

61

Buildings with fences

37

Architectural

62

Bar/pub in the first floor

36

Architectural

73

Building's first floor is an office

27

Architectural

75

Building with mixed uses (residential/office and commerce)

23

Architectural

77

Building with doors and windows

20

Architectural

78

Building's first floor is commercial

19

Architectural

79

Building's first floor is residential

18

Architectural

85

Restaurant/cafĂŠ on the first floor

10

Architectural

Score

Category

73

Appendix 2 Source: Ministry of Justice, Home Office and National Office of Statistics 2013

England and Wales Offense group

Number of crimes

2011-2012

Sexual assault on a female

19780

37%

Rape of a female

14767

28%

Other sexual offenses

9793

18%

Sexual activity with minors

5778

11%

Sexual assault on a male

2273

4%

Rape of a male

1274

2%

Total

53665

Recorded Sub-crimes 2011-2012 2% 4%

Sexual assault on a female

11% 37%

Rape of a female Other sexual offenses

18%

Sexual activity with minors Sexual assault on a male 28%

Rape of a male

74

Appendix 3 Source: Police Reports in Central London 450 400 350 300 250 200 150 100 50 0

Street Reports in Central London

Street Reports Mean count: 352 cases Jun-17

340

Jul-17

370

Aug-17

361

Sep-17

326

Oct-17

355

Nov-17

392

Dec-17

392

Jan-18

330

Feb-18

325

Mar-18

365

Apr-18

313

May-18

365

75

Appendix 4 Source: Police UK, 2018 Category

Category

Percentage

Ranking

21758

42%

1

8296

16%

2

Violence and sexual offences

7013

13%

3

Shoplifting

4617

9%

4

Burglary

2363

5%

5

Public order

2331

5%

6

Vehicle crime

2236

4%

7

Robbery

1891

4%

8

Bicycle theft

1128

2%

9

Possession of weapons

265

1%

10

51605

100%

Thefts Anti-social behaviour

Total

Crime Categories (May 2017- June 2018) Possession of weapons Bicycle theft Robbery Vehicle crime Public order Burglary Shoplifting Violence and sexual offences Anti-social behaviour Thefts 0

5000

10000

15000

20000

25000

76

Source: Police UK, 2018

Type

Frequency

Street

3986

Building

1064

Nightclub

974

Station

787

Parking area

170

Pedestrian Subway

106

Strand

16

Total

7103

Violence and Sexual offense type Strand Pedestrian Subway Parking area Station Nightclub Building Street 0

1000

2000

3000

4000

5000

77

Appendix 5 Source: Survey by the researcher in Soho Date: July 2018

Survey in Soho, map by Baxter 2016

Results: SEX

TOTAL

FEMALE

MALE 35

15

50

VICTIMS OF STREET HARASSMENT YES

NO

N/A

33

12

5

50

66%

24%

10%

100%

AGE RANGE 18-40

TOTAL 41-90

45

N/A 4

1

50

78

STATUS RESIDENT 3

TOTAL

WORKER

TOURIST

35

N/A

10

2

Poor Lighting

44

Low visibility

36

Quiet street

31

Absence of people

30

Narrow sidewalks

27

Presence of men

20

Crowded area

9

Dense greenery

8

A busy street

7

Borders and fences

7

No doors/windows

6

Presence of women

0

50

Fear positive situations in Soho Presence of women No doors/windows Borders and fences A busy street Dense greenery Crowded area Presence of men Narrow sidewalks Absence of people Quiet street Low visibility Poor Lighting 0

10

20

30

40

50

Since the sample of the Soho survey was smaller than the global one, the opinions on less ideas was not taken into deeper analysis.

79

Examples of the survey completed and the perception of fear of street harassment.

80

Appendix 7 Results of the spatial analysis on Soho Streets segments Source: Author

Location On or near Air Street On or near Archer Street On or near Argyll Street On or near Bateman Street On or near Beak Street On or near Berwick Street On or near Bourchier Street On or near Brewer Street On or near Brewer Street On or near Bridle Lane On or near Broadwick Street On or near Carlisle Street On or near Charing Cross Road On or near Denman Street On or near Duck Lane On or near Falconberg Mews On or near Falconberg Mews On or near Foubert's Place On or near Frith Street On or near Frith Street On or near Ganton Street On or near Glasshouse Street On or near Goslett Yard On or near Great Chapel Street On or near Great Chapel Street On or near Great Chapel Street On or near Great Marlborough Street On or near Great Pulteney Street On or near Great Pulteney Street On or near Greek Street On or near Hollen Street On or near Hopkins Street On or near Ingestre Place On or near Ingestre Place On or near Ingestre Place On or near Ingestre Place On or near Kingly Street On or near Lexington Street On or near Little Argyll Street On or near Little Marlborough Street On or near Livonia Street On or near Lower James Street On or near Lower John Street On or near Manette Street On or near Marshall Street On or near Meard Street On or near Moor Street On or near Newburgh Street On or near Noel Street

Count Crime 7 13 16 16 33 12 23 15 15 4 21 64 18 5 11 22 22 12 24 24 3 5 16 9 9 9 7 2 2 17 18 6 32 32 32 32 39 5 12 2 2 4 12 9 1 3 35 3 68

NAIN R400 1.76 1.35 1.17 1.19 1.29 1.30 1.30 1.19 1.19 1.14 1.40 1.26 2.20 1.59 0.86 1.13 1.02 1.20 1.37 1.37 1.09 2.13 1.26 1.32 1.24 1.25 1.70 1.19 1.19 1.58 1.25 1.17 1.04 1.04 1.20 1.14 1.41 1.63 1.20 1.03 0.92 1.73 1.46 1.11 1.20 1.31 2.05 0.91 1.52

Segment Length 88.09 55.37 87.10 103.64 62.92 36.52 52.76 150.98 150.98 94.60 29.84 42.62 42.61 83.71 61.62 59.06 37.18 39.16 97.74 97.74 32.84 104.08 40.94 57.83 15.63 68.32 70.30 150.98 150.98 85.18 68.32 47.71 58.72 41.56 44.49 32.58 77.04 144.88 62.59 66.77 32.84 50.12 49.93 59.48 59.23 56.97 60.39 19.54 74.13

Visibility INT 2.44 2.3 2.92 2.71 2.9 3.05 2.48 3.01 2.49 2.03 2.54 2.52 3.19 2.38 2 3.29 3.29 2.26 2.45 2.45 2.02 3.44 3.26 3.1 2.02 3.1 3.2 2.33 2.33 2.5 3.32 2.46 2.39 2.39 2.39 2.39 2.46 2.51 2.81 1.89 1.44 2.4 2.45 2.65 2.13 2.21 2.75 2.22 3.08

81

On or near Old Compton Street On or near Old Compton Street On or near Peter Street On or near Piccadilly Circus On or near Piccadilly Circus On or near Poland Street On or near Ramillies Street On or near Regent Street On or near Richmond Buildings On or near Richmond Mews On or near Romilly Street On or near Shaftesbury Avenue On or near Shaftesbury Avenue On or near Shaftesbury Avenue On or near Sheraton Street On or near Soho Square On or near Soho Street On or near Sutton Row On or near Upper James Street On or near Wardour Mews On or near Wardour Street On or near Warwick Street On or near Winnett Street

132 132 4 20 20 4 32 5 2 1 21 300 300 2 2 4 9 1 3 4 17 5 20

1.37 1.37 1.45 1.81 1.89 1.25 1.40 1.67 0.93 0.68 1.77 1.58 1.65 2.13 1.10 1.37 1.61 1.41 1.32 0.88 1.55 1.40 1.38

97.74 97.74 27.20 19.54 18.85 57.00 78.71 141.79 52.81 78.53 58.74 8.53 16.44 78.19 48.59 42.71 47.84 56.54 53.59 83.53 77.30 95.94 47.92

2.96 2.96 2.42 3.25 3.43 2.99 2.56 3.84 2.44 1.63 2.7 2.95 2.97 3.4 2.43 2.54 2.88 2.71 2.4 1.65 3.23 2.46 2.36

82

Appendix 8 Gate counting tables and graphs.

Source: Author

Weekday ID 1 2 3 4 5 6 7

STREET Archer Street Great Windmill Street Shaftesbury Avenue Ruppert Street B Wardour Street B (Church) Winnet Street

F

Weekend F

M

Weekday Weekend

M

C

15%

64%

21%

42%

56%

C 2%

T 1416

T 3544

GT 4960

47%

50%

2%

50%

47%

3%

2472

4944

7416

44%

41%

15%

52%

46%

2%

9044

13084

22128

47%

51%

2%

41%

55%

3%

3072

6828

9900

46%

51%

2%

41%

47%

12%

5052

5272

10324

31%

69%

0%

19%

81%

0%

864

804

1668

31%

69%

0%

31%

59%

10%

1128

468

1596

39%

61%

0%

47%

50%

3%

1356

2616

3972

9

Tisbury Court Ruppert Street A Brewer Street

44%

55%

1%

43%

54%

3%

1092

2916

4008

10

Walkers Court

37%

63%

0%

29%

68%

3%

2032

1200

3232

Peter Street

8

11

31%

67%

2%

41%

53%

6%

3284

2916

6200

40%

54%

6%

52%

45%

3%

7188

9912

17100

13

Wardour Street A (Va Piano) Meard Street

37%

62%

1%

39%

54%

6%

2632

3516

6148

14

Dean Street A

39%

59%

2%

41%

59%

0%

7016

5420

12436

15

Bouchier Street Old Compton Street A Dean Street B

36%

64%

0%

41%

54%

5%

364

772

1136

36%

63%

1%

41%

57%

2%

6896

10488

17384

45%

53%

1%

48%

49%

4%

3456

4868

8324

39%

61%

0%

46%

51%

4%

2064

996

3060

43%

57%

0%

40%

58%

2%

2888

2656

5544

44%

55%

1%

34%

66%

0%

2520

2388

4908

42%

55%

3%

42%

57%

1%

1788

2576

4364

48%

50%

2%

42%

56%

2%

7648

2644

10292

36%

63%

0%

36%

61%

3%

7728

2412

10140

37%

61%

2%

41%

57%

3%

3576

1804

5380

46%

52%

2%

44%

54%

2%

9364

7168

16532

56%

40%

5%

45%

47%

8%

2996

3600

6596

39%

58%

3%

45%

53%

3%

1752

1436

3188

40%

60%

0%

40%

54%

6%

1520

1768

3288

12

16 17 18 19 20 21 22 23 24 25 26 27

Dean Street C Bateman Street A Bateman Street B Frith Street A (Ronnie Scotts) Greek Street A Old Compton Street B Frith Street B (Pret) Old Compton Street C Greek Street B

29

Romily Street A Romily Street B (Pret) Frith Street C

46%

51%

3%

50%

48%

2%

2452

624

3076

30

Greek Street C

57%

37%

6%

38%

55%

8%

4388

636

5024

Total

42%

55%

2%

44%

53%

3%

109048

112592

219324

28

83

Archer Street Great Windmill Street Shaftesbury Avenue Ruppert Street B Wardour Street B (Church) Winnet Street Tisbury Court Ruppert Street A Brewer Street Walkers Court Peter Street Wardour Street A (Va Piano) Meard Street Dean Street A Bouchier Street Old Compton Street A Dean Street B Dean Street C Bateman Street A Bateman Street B Frith Street A (Ronnie Scotts) Greek Street A Old Compton Street B Frith Street B (Pret) Old Compton Street C Greek Street B Romily Street A Romily Street B (Pret) Frith Street C Greek Street C Shaftesbury Avenue B

Gate Counting on Weekday by daytimes

9000

8000

7000

6000

5000

4000

3000

2000

1000

0

Morning Afternoon Night

84

Archer Street Great Windmill Street Shaftesbury Avenue Ruppert Street B Wardour Street B (Church) Winnet Street Tisbury Court Ruppert Street A Brewer Street Walkers Court Peter Street Wardour Street A (Va Piano) Meard Street Dean Street A Bouchier Street Old Compton Street A Dean Street B Dean Street C Bateman Street A Bateman Street B Frith Street A (Ronnie Scotts) Greek Street A Old Compton Street B Frith Street B (Pret) Old Compton Street C Greek Street B Romily Street A Romily Street B (Pret) Frith Street C Greek Street C Shaftesbury Avenue B

Gate Counting on Weekend by daytimes

16000

14000

12000

10000

8000

6000

4000

2000

0

Morning Afternoon Night

85

86

87

Appendix 9 During the personal surveys in Soho, one of the participants shared a poem she had written about street harassment and allowed the author to share it in this investigation

Me Too

By Bethany Fox Me too.

Me too me too me too.

I feel it’s needed for me to address that yes, me too.

I’m sat on the tube and the guy opposite is eyeing me Prying me and all the while denying me of the right to say no Don’t yell obscenities at me calling me “ho” What do you know?

It’s not okay that I can’t walk down the street without a car window been lowered Don’t be offended when I call you a coward Calling me hot from behind a shield In this misogynistic minefield When I stand alone In that moment Entirely divided by the comment you sent Maybe you didn’t mean to put a dent my self confidence Maybe you thought you were enhancing my day Leaning on me and calling me bai Nay. No. Because in that moment The one you try to downplay It doesn’t matter what you say You stay with me Every inch of the way And I didn’t ask for you I didn’t consent for you to comment And who are you?

No seriously who the f. are you?

Looking at me with eyes that undress Are you thinking that they caress?

I feel uncomfortable Naked Alone To name a few words I feel from your tone I’m not saying don’t appreciate.

The beautiful women you tend to berate It’s not too late

88

To change the fate Move away from hate I surround myself with friends that appreciate Me And most of them are men Gentle men My closest and most intimate friends are men.

Because that’s it It's not a question of us and them This can't be something we choose to split This is everyone’s fight And now it's under a public spotlight My experiences are small “Beth don't kick up a fuss over a tiny catcall?”

However If we all close up and keep quiet, then what happens next?

It’s simply a matter of respect The fact that equality is not yet reality is incorrect And I won’t apologise for being direct because I am done With wondering whether I’ll have to run With shaking it off and carrying on With insidious monsters who leave a huge stain With innocent women being honourably slain With a man saying That’s a man’s job What? You can change light bulbs cos you’ve got a nob?

Oh please No seriously, please.

Together we have to end this disease And I’m not pretending I have expertise And no I don’t think I’m Socrates But I have experience I’ve been groped And in case you’re wondering no it wasn’t provoked Even the career I crave is soaked with abuse Rotten to the core And there are girls out there suffering so much more Open your eyes because we all know the score And I’m not being silent Not anymore

89

Appendix 9 This pilot survey was realised to 20 women in June 2018.

90

91

Results

This dissertation is dedicated to every woman who, no matter the obstacles, inspire others to walk with their head held high.

292