Supplemental safety? Exploring experienced safety in relation to other qualities of successful public spaces

ABSTRACT

Combining insights from the urban design and criminological literature, this paper explores the degree to which conviviality, feeling welcome, and feeling at home are related to peoples’ experiences of safety. A questionnaire was distributed on four squares in the city of Utrecht, the Netherlands. While participants valued the positive qualities of the squares differently, feeling safe was connected to a wider, positive evaluation of the quality of space and especially to feeling at home. By investing primarily in public spaces that are ‘homely’ or ‘domesticated’, a ‘supplemental safety’ might therefore be nurtured. This poses important insights for safety management.

Introduction

The degree to which people feel safe in urban public spaces is generally considered a necessary condition for such spaces to function successfully. Perceived lack of safety may negatively impact peoples’ behaviour and result in the avoidance of particular urban areas (Hale 1996; Liska, Sanchirico, and Reed 1988; Rader, May, and Goodrum 2007; San-Juan, Vozmediano, and Vergara 2012; Warr 2000). Avoidance and abandonment of urban areas, and the city centre in particular, may, in turn, be detrimental to the image of the city and its competitive position. More generally, studies have (critically) foregrounded the role of safety and security in the commodification of city spaces and the creation of successful urban (consumer) economies (Atkinson 2003; Carmona 2010a; Hayward 2004; Minton 2012; Oc and Tiesdell 1999). Municipalities’ attempts to stimulate safety has not left the character and appearance of public city spaces untouched. The literature has pointed to how surveillance and policing practices in public and semi-public spaces of cities has changed and increased over time (Crawford et al. 2005; O’Neill and Fyfe 2017). At the same time, urban public spaces are increasingly managed, designed and ‘revitalized’ to attract particular (consumer) groups and, at times, to keep out others (Barker 2017; Carmona 2010a, 2010b; Davis 2006; Minton 2012; Mitchell 1995, 2003; Németh and Schmidt 2007, 2011; Van Stokkom 2009). While studies report evidence that such measures can and do stimulate experienced safety (in particular city spaces, among particular parts of the population), others have suggested that too strong an emphasis on safety and security in, and management and regulation of, urban public spaces may also negatively affect its public character. Indeed, various studies illustrate how excessive control, management, and overdesigned urban public spaces may result in them becoming less and less spontaneous, friendly, accessible, and making them more homogeneous, as well as unfair to particular segments of the population (Barker 2017; Davis 2006; Minton 2012; Mitchell 2003; Németh and Schmidt 2011; Oc and Tiesdell 1999; Whyte 1980; Van Stokkom 2009). Others have also suggested that excessive surveillance and policing may actually make people, at times, feel less safe (Brands, Schwanen, and van Aalst 2015; Hale 1996; Zedner 2003).

While such critiques do not deny the importance of nurturing safety experiences in urban public spaces, they do pose critical questions concerning the (acceptable) costs of (changes in) surveillance and policing practices, and safety management more generally. Such critiques urge scholars to understand and balance safety in relation to other (similarly crucial) qualities of urban public spaces and, therefore, to investigate it in a more holistic manner. The urban planning and design literatures are helpful in that respect: besides positioning experienced safety as one of the crucial qualities of (semi)public city spaces, this literature repeatedly emphasizes the importance of other dimensions such as comfort, pleasurability, animation, sociability, and inclusiveness (Mehta 2019; PPS 2019). Indeed, much like Barker (2017), the current study recognizes a strong need for an approach (in criminology) that ‘decentres crime and (in)security as central organizing concepts’ (848) to urban public spaces. The aim of the current study is to present, elaborate on, and further connect this idea to criminological literature and empirical study on the fear of crime and perceived safety specifically.

This is important for two reasons. First, doing so may provide more, and valuable, information about potentially undesirable effects of interventions aimed at stimulating safety experiences. Taking such an approach makes it easier to evaluate trade-offs between different experienced qualities of public space. Second, some existing literature suggests that convivial, animated, accessible and inclusive public spaces may actually contribute to people perceiving places as safer (Allen 2006; Barker 2017; Blokland and Nast 2014; Jacobs 1961; Oc and Tiesdell 1999; Shaftoe 2012). Or, as Koch and Latham (2013) have it, while municipalities tend to focus on ‘unruly elements’ and may successfully ‘pacify or tame’ these, what actually ‘troubles many public spaces is that they are lacking in certain domestic qualities. That is, they fail to provide a sense of trust, comfort or amenity that might invite multiple publics to inhabit them’ (9) and feel safe in them. Drawing on these observations, the current study examines how investing in (particular) other (positive) qualities of urban public spaces may work towards, what will be coined here as ‘supplemental safety’.

Perceived safety

Before building on this idea of supplemental safety, it is important to discuss the (traditional) conceptualization and measurement of perceived safety and fear of crime in the criminological literature. Beyond a debate about conceptualization and measurement as such (Farrall, Jackson, and Gray 2009; Ferraro and Grange 1987; Gray, Jackson, and Farrall 2012; Hale 1996; Hinkle 2015), it follows from these literatures that important contextual differences exist in the ways perceived safety is operationalized (Hale 1996; Pleysier 2010; Vanderveen 2006). For instance, in the Dutch context (where the current study was conducted), it is quite common to ask research participants about their ‘veiligheidsbeleving’, what might best be translated as their experienced safety. In Anglo-Saxon literature, however, studies often refer to fear of crime. The latter is more narrowly defined in comparison to ‘veiligheidsbeleving’, even though there remains substantial discussion in the literatures whether or not ‘“fear of crime” [is] simply measuring fear of crime or, perhaps in addition, some other attribute which might be better characterized as “insecurity with modern living”, “quality of life”, “perception of disorder” or “urban unease”?’ (Hale 1996, 84), ‘(…) [therefore] absorbing all sorts of anxieties about related issues of deteriorating moral fabric, from family to community to society (…)’ (Jackson 2006, 261). Studies focusing on the Dutch context generally agree that such attributes and perceptions are also reflected in peoples’ ‘veiligheidsbeleving’, above and beyond a more narrowly defined fear of criminal victimization (Pleysier 2010; Spithoven 2017; Vanderveen 2006).

From the previous, it follows that ‘veiligheidsbeleving’ and fear of crime can be (and most often are) studied as more or less distant feelings, ‘detached’ from a specific space and time. However, they can also be studied as a spatiotemporally complex event, ‘emerging’ and ‘fading’. That is, and according to Spithoven (2017), as ‘(a) time and space-specific response to external stimuli (…)’ (Farrall, Jackson, and Gray 2009, 149–157; Pleysier 2010, 175–179) ‘of crime, crime signals or crime symbols’ (Ferraro 1995, 4; Ferraro and Grange 1987, 72) leading to ‘(…) a sense of immediate threat to one’s security (…)’ (Farrall, Jackson, and Gray 2009, 18 and 245).¹ The current study builds on this second approach to experienced safety as ‘situational’ (see Garofalo 1981; Solymosi, Bowers, and Fujiyama 2015), as this makes it possible to investigate experienced safety as co-existing and/or relating to other (experienced) qualities of particular public spaces.

Qualities of public space

Putting it bluntly, one of the important things the urban planning and design literature teaches us is that making a place as safe and secure as possible does not necessarily provide a good quality public space. But, what then constitutes good quality public space? Following De Magalhães (2010, 562), it ‘makes sense to try to define [good] public space on the basis of what might be its ontological attributes, those essential qualities that give public space its specificity, its publicness’. Many such attributes or qualities have been forwarded by different scholars, often touching upon overlapping issues such as ownership, control, access and/or use (Mehta 2014, 54). For instance, according to The Project of Public Spaces (PPS 2019), four qualities coalesce in, and contribute to, successful public spaces: ‘They are accessible; people are engaged in activities there; the space is comfortable and has a good image; and finally, it is a sociable place: one where people meet each other and take people when they come to visit’ (PPS 2019).

Scholars have also started to design concrete instruments to assess these essential qualities of public spaces and their ‘publicness’ (Akkar 2005; Evans et al. 2019; Langstraat and van Melik 2013; Mehta 2014; Németh and Schmidt 2011; Varna 2014; Varna and Tiesdell 2010). Using such instruments, the ways attributes or qualities of public spaces come together and assemble part of particular space-times can be studied both through observable aspects (by the researcher) as well as lived through, by visitors, using city spaces (Mehta 2014, 2019). It is in line with the latter that various instruments include peoples’ experienced safety. For instance, when introducing his instrument, Mehta mentions that ‘good public space is accessible and open, is meaningful in its design and the activities it supports, provides a sense of safety, physical and environmental comfort and convenience, a sense of control, and sensory pleasure’ (Mehta 2014, 57).

Safety and perceived qualities of public spaces combined: at home, convivial, and welcome

Resonating some of the ideas addressed in the previous section, various studies that take primary interest in experienced safety advocate that a framework solely investigating peoples’ ‘fear of crime’ or their experienced lack of safety is less informative on the ways people may also have other, more positive experiences related to safety when traversing public spaces (Hutta 2009; Oude Breuil, Schuilenburg, and Van Steden 2014; Schuilenburg and Van Steden 2016). As such, some authors argued that the literature might be furthered by drawing ‘the analytic framework out of its hegemonic order-centred and fear-centred fix’ (Hutta 2009, 258). In fact, a central argument in some prior studies is that positive experiences may (also) be characteristic of experienced safety (for which the concept of the fear of crime in itself leaves little room due to its inherent, negative nature). Combining these insights with the insights from the urban planning and design literatures, at least three positive experiences, allied with (perceived) qualities of public space, can be identified: (I) feeling at home, (II) experiencing a place as convivial, and (III) feeling welcome.

Investigating the literature on feeling at home, for instance Duyvendak (2011) argues that when people ‘(…) feel at home, that they feel “safe”, “secure” and “comfortable”, at “one with their surroundings”’ (27). Hence, Duyvendak argues that both comfortableness and safety are associated with feeling at home. Also Bissell (2008) argues that ‘to be comfortable may be a highly desirable situation. The feeling of comfort may be extremely beneficial, a sensation through which an individual may derive a sense of security’ (1997). Koch and Latham (2013), in turn, argue that efforts to domesticate public spaces should be considered an important alternative to standard safety and security practices, providing ‘a productive way to attend to the processes by which people come to inhabit public places and make a sort of home in the city’ (7). So, feeling at home and experiencing comfort are both theorized in relation to safety as such, but also feature prominently as part of PPS’ (2019) and Mehta’s (2014) broader frameworks of good quality public space, where ‘physical and environmental comfort and convenience’ pertain to one particular dimension (Mehta 2014, 57). Mehta’s instrument, for example, assesses this by the presence of seating and other artefacts, and the perceived physical condition and maintenance of the public space (Mehta 2014, 64).

Scholars have also paid attention to how convivial or lively urban spaces (e.g., Gehl 2011; Shaftoe 2012) may encourage the experience of safety. Convivial urban spaces are considered to be friendly, jovial and lively part of which people ‘co-mingle confidently and harmoniously’ (Barker 2017, 856).² For Gehl (2011), places in which necessary (more or less compulsory activities) but also many optional activities (those activities that people pursue because they wish to do so) take place, are lively urban spaces supportive of many social activities. This serves as a base to share public space with others in a ‘relaxed and undemanding’ (Gehl 2011, 17) way.

Jane Jacobs explicitly links lively, convivial and animated urban places to (experienced) safety through the mechanism of social control. Indeed, referring to the works of Jacobs (1961) and Felson and Clarke (1998), Shaftoe (2012, 15) notes that ‘well-used convivial places are the alternative to downtown areas abandoned to criminals and the socially rejected’. It is also in the works of Gehl (2011) and Shaftoe (2012) that various urban planning and design principles to ‘influence patterns of activities, to create better or worse conditions for outdoor events, and to create lively or lifeless cities’ (Gehl 2011, 31) are discussed. In fact, part of their work, scant links can be found with the previously discussed instruments to assess the essential qualities of public spaces on aspects of sociability, support of activities, accessibility, and enticing sensory pleasure (see PPS 2019; Mehta 2014).

Finally, Dutch scholars have paid some attention to the degree to which people feel welcome in public spaces. For instance, Boonstra and Hermens (2009) mention that good public squares – in their design – are welcoming, and capable of containing a diverse group of users including strangers and youngsters. They also argue that the development of squares on a basis of mistrust – including (too) much surveillance and security management – may be one reason why people sometimes feel unwelcome on squares in the Dutch city of Rotterdam. Furthermore, while squares may be accessible to people, people may feel less welcome when particular groups have appropriated the square. To feel welcome in and invited to public spaces, according to Van Stokkom (2009, 210), it therefore needs to be clear that visitors can share the space in a ‘unconstrained and care-free way’.

To feel welcome in a particular publicly accessible space is sometimes also referred to in terms of ‘symbolic accessibility’, and is related to accessibility and territorial separation according to Németh and Schmidt (2007, 287). In this, some overlap can be noted in the studied concept of ‘belonging’. Mee and Wright (2009) argue that ‘various practices of boundary making and inhabitation’ (772) signal if particular objects and bodies are considered to belong or are ‘in place’ (see also Cresswell 2015; Boonstra and Hermens 2009). To feel welcome and to belong are therefore also linked to matters (and feelings) of inclusion and exclusion. While studies often investigate feeling welcome, a sense of belonging, and matters of inclusion and exclusion in the light of, for example, gender, age, and disability, the current study approaches it as the degree to which all users encountered within a particular space experience it as inviting and feel ‘in place’. The work of Blokland and Nast (2014) is one example of how such processes can nurture ‘recognizing and being recognized in local spaces’ (1142) based on co-presence, which is considered to affect experienced safety in a positive way. Reflecting on the frameworks from Mehta (2014) and the PPS (2019), feeling welcome can be theorized in relation to processes of in- and exclusion and experiencing the place to be (symbolically) accessible as well as being a sociable place.

Having elaborated on and positioned ‘feeling safe’ amidst these other positive experienced qualities of a public space, the exploratory analysis below examines how feeling at home, experiencing a place as convivial, and feeling welcome relates to experienced safety.

Method

In order to investigate experienced safety, as well as feeling at home, experiencing a place as convivial, and feeling welcome, a first step in this research consisted of selecting various different public spaces. After the selection of these public spaces, on-site street surveys were conducted to investigate these aspects. Below, the selection of the public spaces will firstly be discussed, followed by the details of the survey research, and the analytical approach.

De Neude, Jacobskerkhof, Domplein, and Stadsplateau

It was decided to focus research efforts on four publicly accessible squares in the city centre of Utrecht, the Netherlands: 1) de Neude, 2) Jacobskerkhof, 3) Domplein and 4) Stadsplateau (see Figure 1). City centre squares can be considered a specific type of public space (Shaftoe 2012), but can be expected to differ in terms of how (safe, homely, convivial, and welcome) people perceive them.

Figure 1. Research locations in Utrecht (The Netherlands). 1: De Neude. 2: Jacobskerkhof. 3: Domplein. 4: Stadsplateau (Made with: © QGIS. A Free and Open Source Geographic Information System. Qgis.org. Data: © OpenStreetMap contributors. Openstreetmap.org. Data: Natural Earth. Free vector and raster map data. Naturalearthdata.com)

De Neude (Figure 2) is a very accessible square which hosts a variety of shops, food- and drink outlets, and during the summer months it is especially crowded as a substantial part of the Neude is converted into terraces. De Neude was therefore assumed to support many (optional) activities and sociability, and therefore ought to score high on conviviality. Stadsplateau – which provides an entranceway to the shopping mall ‘Hoog Catharijne’ and Utrecht Central Station – lacks such amenities but is within walking distance of all sorts of facilities for shopping, eating and drinking. While there are some design features aiming to support sensory pleasure, this square was selected because it was estimated to be characterized especially by the presence of necessary activities and be least convivial. Since this area is also characterized by a high amount of surveillance and a high amount of obstacles (stairs; escalators; planters), people might also consider Stadsplateau less welcoming and comfortable or homely. While de Neude also has sizable surveillance and policing, it is a very open square with plenty of seating opportunities and artwork. Based on this, it was presumed to be a comfortable place to spend time. In addition to what were imagined to be two fairly contrasting squares, two less ‘pronounced’ or ‘average’ squares were also selected. Domplein is host to the renown Dom church of Utrecht. In addition to the church, some food and drink outlets, a museum, statues as well as seating options are found on the square, with only limited surveillance, which fits the idea of quite a homely/comfortable and welcoming, and somewhat convivial square. Jacobskerkhof is also characterized by a low amount of surveillance, but also a low amount shopping, food and drink outlets. Hence, people might consider this square as less convivial. While there are some options for seating available, it was in most part expected to find necessary activities taking place on this square.

Figure 2. Impression of the research locations; 1 de Neude, 2 Jacobskerkhof, 3 Domplein, 4 Stadsplateau. Data: © OpenStreetMap contributors. Openstreetmap.org

Participants

In total, 539 passers-by on the four public squares were asked to participate in the current study, 201 people agreed to participate. The mean age of the research participants was 33.97 years (SD = 17.72), about two thirds was female. The majority (self)reported having a Dutch nationality (86.1%).

Procedure

The data were collected during the summer months August and September of the year 2017. In order to maximize comparability, the data were collected only during weekdays, at daytime between 09:18 (earliest) and 16:12 (latest) hours. No sampling took place when events/festivals were hosted on the squares.

While the research participants were asked to fill out the surveys themselves, the researcher remained present to assist and/or clarify when necessary. In order to conduct the research as random as possible, the researcher chose a particular position on the square which was changed every 1–2 hours. Every first passer-by within 5 metres of the researcher was approached and asked to participate in the study. If accompanied by a maximum of three others, these other people were also invited to participate in the study. As evidenced above, the current exploratory study is based on convenience sampling, and should therefore not be considered representative for a larger population.

Measures

The survey started by asking research participants ‘How safe did you feel at the place where you are now, just the moment before I asked you to participate in this study?’. Participants could answer this question on a scale of 1 (not safe) to 10 (very safe). By adding reference to the particular place the participants were in shortly before filling out the survey, this question was used to measure ‘situational’ experienced safety (Garofalo 1981; Solymosi, Bowers, and Fujiyama 2015).

Participants were then asked ‘To what extent do you feel at home at the place where you are now?’. They were also provided with a short and common definition of this construct, in order to make the results more internally valid: ‘By feeling at home in a place we mean that you feel confident and comfortable, and that you have the feeling you can be yourself’. Again, participants could answer this question on a scale of 1 (not at home) to 10 (very at home).

Next, participants were asked ‘To what extent do you think this place is convivial?’, explaining them that with convivial it is meant that a place is friendly and animated/lively. Possible scores again ranged from 1 (not convivial) to 10 (very convivial).

Finally, participants were asked ‘To what extent do you feel welcome at the place where you are now?’. Feeling welcome in a place was defined to the research participants as feeling invited to visit the place and having the feeling that their presence is being accepted. Again, scores between 1 (not welcome) and 10 (very welcome) could be reported.

Participants were also asked about their estimated frequency of visits to the specific square – as a proxy for familiarity – their gender (male/female) and age (in years). The researcher also noted if the research participant was encountered alone, or in company. This is substantiated by the fear of crime literatures, where it is generally agreed upon that these aspects are related to peoples’ sense of vulnerability and should therefore be corrected for (Hale 1996). Finally, halfway through the data collection a terrorist attack took place on the Ramblas in Barcelona, Spain. To control for a possible effect on peoples experienced safety, it was also noted if the survey was conducted before or after the attack took place.

Analytical approach

The analytical approach consisted of three steps. To start with, an important question was whether the four items – experienced safety, feeling at home, experiencing a place as convivial, and feeling welcome – measured separate concepts, and not one single concept. It was expected that these four items would measure separate concepts, but that there could be moderate correlations (>.60) between the four items, given the theoretical overlap of elements of a positive evaluation of publicness. Through investigations of the correlations, Cronbach’s Alpha, and by performing a factor analysis, the possibility that correlations between the items was the result of measuring the same latent concept was checked, and falsified. Therefore, in the second step, relations between these four items were explored with regression analyses. More specifically, the regression analyses measured whether and how feeling at home, experiencing a place as convivial and feeling welcome could be regarded as predictors of experienced safety. This is in line with the idea that safety could be supplemental to (cultivating) other (positive) qualities of urban public spaces. To complement the regression analyses, Structural Equation Modelling (SEM) was performed. This technique ‘(p)rovides an integrative function’ (Bagozzi and Yi 2012, 12) as it ‘(…) includes several traditional multivariate procedures to test data on – theoretically expected – structures for (…) the covariances between the observed variables’ (Hox and Bechger 1998, 354). The key strength of this analytical strategy is ‘(…) to produce a meaningful identification of the correlations between factors’ (Hox and Bechger 1998, 358). So with SEM, ‘(…) we may specify and estimate more complicated path models, with intervening variables between the independent and dependent variables and factors as well’ (Hox and Bechger 1998, 359). In short, such a technique provides a more elaborate model and demonstrates the relations between all measured concepts, with a focus on the outcome of feeling safe. Therefore, SEM was deemed a good fit to test the central importance of multiple dimensions, or qualities, of urban public spaces (safe, but also, homely, convivial and welcome).

Results

Exploring the relations between concepts

Table 1 provides an overview of the regression statistics. The first step of the regression analysis investigated whether feeling at home, experiencing a place as convivial, and feeling welcome predicted feeling safe. There was a significant result for feeling at home (β =.270; p = .001), but not for experiencing conviviality (β = −.060; p = .464) and feeling welcome (β = .011; p = .898). This suggests that the more people feel at home at a certain square, the safer they feel. The proportion of variance explained in this analysis was 6.9%.

Table 1. Linear regression with feeling safe as dependent variable

Step 1							Step 2
	B	95% CI		SE	β	Sig.	B	95% CI		SE	β	Sig.
Constant	7.306	6.248	8.363	.536		.000**	8.073	6.825	9.321	.632		.000**
Feeling at home	.202	.087	.318	.059	.270	.001*	.185	.067	.304	.060	.247	.002*
Conviviality	-.045	-.165	.076	.061	-.060	.464	-.018	-.145	.108	.064	-.025	.774
Feeling welcome	.010	-.145	.165	.078	.011	.898	-.034	-.192	.124	.080	-.038	.673
Gender							-.087	-.508	.335	.213	-.028	.686
Age							-.007	-.019	.005	.006	-.083	.270
In company							-.529	-.97	-.085	.225	-.173	.020*
Ramblas attack							-.544	-1.014	-.075	.238	-.184	.023*
Visiting frequency							.126	-.136	.389	.133	.075	.344
Location (squares)
Jaarbeursplein							REF
Jacobskerkhof							.411	-.162	.985	.291	.126	.159
Neude							.146	-.484	.775	.319	.045	.648
Domplein							.795	.139	1.451	.333	.242	.018*
R^2	.069						.160

* <.05; ** <.001.

These relationships remained intact when controlling for gender, age, being in company, whether the questionnaire was conducted before or after the Ramblas attack, and how often the respondent visited the square. There was no significant result for gender (β = −.028; p = .686), age (β = −.083; p = .270) and visiting frequency (β = .075; p = .344) on feeling safe. There was a significant result for being in company (β = −.173; p = .020),³ the Ramblas attack (β = −.184; p = .023), and the Domplein square (β = .242; p = .018). The proportion of variance explained in this model was 16%, thus slightly higher than the previous model.

SEM was performed as a follow-up on the regression analyses. It was decided to perform the SEM without including the variable ‘location’ for three reasons: (1) the regression analyses showed no significant results for the different squares except for Domplein; (2) not including this variable allows for more statistical power; and (3) the primary aim was to investigate the relationship between the concepts of feeling at home, conviviality, feeling welcome, and feeling safe in general.

The model indicates various direct, but also indirect, relations with experienced safety (Figure 3). First, feeling at home is the most prominent predictor (λ = .25) of feeling safe in the model. This means that participants who reported feeling more at home at a particular square, also tended to report higher levels of experienced safety. While no significant direct relationships were found between conviviality and feeling safe or feeling welcome and feeling safe, both were related to feeling at home (which, as mentioned above, predicted feeling safe). Experiencing the place as convivial was strongly related to feeling welcome (λ = .52), which in turn was the strongest predictor of feeling at home (λ = .45). A negative relationship between gender and feeling welcome was found. This means that men feel less welcome than women in the particular location in which they were surveyed. But still, for both genders equally, feeling welcome was the strongest predictor of feeling at home. Visiting frequency (λ = .26) and age (λ = .17) also showed significant, but less strong, positive relationships with feeling at home. This means that participants who visit the squares more frequently feel more at home, as do older research participants. Also, a negative direct relationship between the Ramblas attack variable and feeling safe was found. Participants who filled out the questionnaire after the attack on the Ramblas in Barcelona, Spain on August the 17^th 2017 felt less safe compared to those who answered the questions before the attack.

Figure 3. Results of Structural Equation Modelling (SEM)

In sum, a key finding from the explorative SEM model is that feeling safe at the study locations seems to be embedded in a wider, positive evaluation of the experienced quality of place and especially to feeling at home, which is connected to feeling welcome as well as experiencing conviviality, even when controlling for various regular variables.

Discussion

The urban planning and design literature emphasizes various attributes or dimensions that contribute to the publicness of urban public spaces, including the degree to which people feel safe in urban public spaces. Drawing on these literatures, a central point of departure for the current study was the idea that experienced safety might be approached (or, indeed stimulated) as ‘supplemental’ or ‘collateral’ to the overall successfulness of urban public spaces (as opposed to the typical central role the criminological literature tends to bestow on the construct). Putting it differently, by investing in various (experienced) qualities of urban public spaces, experienced safety might also be stimulated as a ‘positive side effect’.

The current study focused on three such qualities that, to some extent, are also recognized and highlighted in the literatures primarily interested in safety experience, and that at the same time may be linked to more general frameworks featuring in the urban planning and design literatures, namely: the degree to which people experience urban public spaces as homely, convivial, and welcoming.

Results from the analyses show that feeling safe on the selected squares is connected to a wider, positive evaluation of the quality of space and especially to feeling at home, which in turn is connected to feeling welcome as well as experienced conviviality. These relationships also remain intact when correcting for several demographic and contextual variables.

On a more general level, the results seem to indicate that a generally positive experience of public space matters for the degree to which people feel safe in it. It aligns with the argument put forth by some scholars that positive experiences may (also) be characteristic of experienced safety, for which the concept ‘fear of crime’ in itself leaves little room. This implies that, instead of applying a solely negative focus in safety management (i.e., removing triggers of fear and unsafety), one could also think about ways to improve safety from a more positive angle (see also Hallsworth and James 2014; Koch and Latham 2013). In doing so, scholars might do well to continue to learn from and integrate ideas that are already available (from the criminological and urban planning and design literatures) and build bridges as the current explorative study has aimed to do. The current analysis provides but one example of how experienced safety can be de-centred and investigated as ‘supplemental’ to (other, positive) experienced qualities of public spaces, such as feeling at home, experiencing a place as convivial, and feeling welcome.

Thinking about safety as part of a broader set of dimensions that characterize the quality or publicness of public space is important because it makes it easier to recognize if, and to what extent, interventions aiming to increase the ways people experience safety in particular city spaces do not disproportionally decrease its (perceived) publicness and/or quality (or specific dimensions thereof). At the same time, it enables us to think about strategies to increase its overall successfulness, including safety. Such a strategy, therefore, provides a way forward in thinking about a ‘toolbox’ for stimulating public spaces experienced as safe and pleasant.

A few limitations of the current research should be noted. First, the current research should be considered exploratory and its results should be interpreted accordingly. That is, elements that contribute to a feeling of safety were explored in one Dutch city, namely Utrecht. While different squares of the city of Utrecht were included, each with their own unique characteristics, it is inevitable that other aspects shape peoples’ safety feelings as well. Moreover, the results are based on a convenience sample of which two thirds of the respondents were female. Still, it can be argued that the current study does modestly add to the current debate about what experienced safety contains and how to make safety management as beneficial as possible.

It should also be taken into consideration that the results reflect the opinion of those using the researched squares for necessary or optional activities. This renders the design quite blind to those for whom safety, feeling at home and welcome (and conviviality) might be most at stake, and possibly forms a crucial reason for not using the research locations altogether. Indeed, as Mehta argues more generally, ‘[p]ublic spaces that already exhibit exclusionary patterns disadvantage the researcher to reach the complete potential audience and users of the space’ (Mehta 2019, 366). The possible absence in the data of those for whom lack of safety, feeling at home and feeling welcome might result in avoiding the research locations, would have implications for the validity of the results and the positive relationships between the constructs reported. This is related to a more general observation that it might, in fact, be very difficult and challenging to balance these experiences between different groups in society (Barker 2017; Duyvendak 2011).

Overall, the results of this explorative study suggest that it might be beneficial to include other, and also more positive qualities of urban public spaces in the investigation of safety feelings. The core finding that people who feel more at home in the investigated public spaces also feel more safe in them can be mobilized as example in that regard, suggesting that the ‘hard work of domesticating public space’ (after Koch and Latham 2013) might be an (additional) beneficial strategy in creating public spaces that are experienced as safe and pleasant.

Acknowledgments

We would like to thank Megan Diederik who assisted in the preparatory- and data collection phase of this study. Also a big thanks to the research participants in the city of Utrecht.

Disclosure statement

No potential conflict of interest was reported by the authors.

Notes

1. Importantly, scholars have highlighted the agency individuals possess in this process (Gabriel and Greve 2003); particular (groups) of individuals can be more receptive to crime, crime signals or crime symbols based on personal dispositions and socialization processes.

2. It is important to mention here that Barker argues that convivial public spaces require some degree of mediation. Her contention is that ‘active management and forms of mediation are required in so much as it facilitates conviviality and enables the safe gatherings of people, but no more’ (Barker 2017, 857).

3. While it was not expected to find a negative relationship between being in the company of someone else and experienced situational safety, it could be hypothesized that being with others amounts to a behavioural response that follows general fear of crime (Liska, Sanchirico, and Reed 1988; Rader, May, and Goodrum 2007). At the same time, this finding should be interpreted with caution, given the unequal distribution (65 persons were encountered alone, while 135 were encountered in company).