534254 EUR0010.1177/0969776414534254European Urban and Regional StudiesTemelová et al. research-article2014 European Urban and Regional Studies Article Safe life in the suburbs?Crime and perceptions of safety in new residential developments in Prague’s hinterland, Czech Republic1 European Urban and Regional Studies 1–20 © The Author(s) 2014 Reprints and permissions: sagepub.co.uk/journalsPermissions.nav DOI: 10.1177/0969776414534254 eur.sagepub.com Jana Temelová, Jakub Novák Charles University in Prague, Czech Republic; University of Tartu, Estonia Jana Jíchová Charles University in Prague, Czech Republic Abstract Despite a rich body of literature dealing with suburbanization across Central and East Europe, the issues of crime and safety in residential suburbs have not been addressed. At the same time it is obvious that the existing knowledge on suburban crime derived mainly from Western experience cannot be simply transferred to the post-socialist transition context. This research investigates the issues of crime and safety in new residential neighbourhoods in the hinterland of Prague, the Czech Republic. Suburbanites’ fear of crime and feeling of safety are discussed in the context of registered crime patterns in the Prague metropolitan region. The research draws on data gathered in a questionnaire survey of newcomers to suburban housing. The findings confirm the generally high feelings of safety in low-crime suburban districts. Our analysis further showed that age, previous victimization and length of residency are the main determinants of fear of crime in Prague’s suburban communities. Keywords Czech Republic, fear of crime, new residential developments, Prague, suburbanization Introduction Much scholarly attention in post-socialist urban studies has been paid to the revival of the phenomenon of suburbanization. This is mainly due to the fact that suburbanization represents the most pronounced change in the structure of cities in Central and East Europe (CEE) after 1990. Although there is a rich body of literature dealing with various aspects of the process in the post-socialist context (e.g. Hirt, 2008; Kährik et al., 2012; Krisjane and Berzins, 2012; Leetmaa et al., 2009; Novák and Sýkora, 2007; Ouředníček, 2007; Špačková and Ouředníček, 2012), the issues of crime and safety in newly-built suburban communities have not yet been addressed. New housing construction has attracted mainly younger, well-educated and higher income population groups to the hinterlands of post-socialist cities (Dövényi and Kovács, 2006; Kährik and Tammaru, Corresponding author: Jana Jíchová, Department of Social Geography and Regional Development, Charles University in Prague, Albertov 6, 128 43 Praha 2, Czech Republic. Email: jana.jichova@natur.cuni.cz Downloaded from eur.sagepub.com by guest on July 25, 2014 2 European Urban and Regional Studies 2008; Špačková and Ouředníček, 2012; Tammaru and Leetmaa, 2007). Although the improvement in living conditions dominates the factors leading households to newly-built suburbs, environmental attributes were also considered in decisions to move out of the city (Kährik et al., 2012; Kopečná and Špačková, 2012; Krisjane and Berzins, 2012). As with Western-style suburbanization (Baldassare, 1992; Foster et al., 2010; Katzman, 1980; South and Crowder, 1997), the expectation of a safer, cleaner and more child-friendly living environment compared to core cities contributed, among other factors, to residential preferences of suburbanites in newlybuilt districts on the urban outskirts (Kährik et al., 2012; Kopečná and Špačková, 2012). Although there is no empirical evidence from post-socialist urban regions, Western studies suggest a lower incidence of crime in suburban settlements compared to central cities (Boggs, 1965; Brown, 1982; Fyfe, 2000), but also that wealthier residential suburbs with detached family housing attract more property crime (Hakim, 1980; Hope, 1999). What do these Western-rooted findings imply for crime and safety in new suburban communities of post-socialist cities? This study aims to contribute to the understanding of suburbanization in CEE by addressing the neglected issue of suburban crime and safety. The paper explores fear of crime, feelings of safety, crime-related experiences and prevention measures applied by households who moved to new suburban neighbourhoods located in Prague’s hinterland in the Czech Republic. The following research questions are addressed. (1) How do residents of new suburbs perceive and experience crime and safety in their neighbourhoods? (2) How do suburbanites’ feelings of safety vary according to types of crime and population groups? The subjective views of suburban residents based on a questionnaire survey are supplemented by contextual analysis of registered crime patterns in the Prague metropolitan region. The empirical findings shed light on crime and its perception in new suburban housing developments in CEE urban regions and confront it with existing knowledge derived from Western (mainly the US, the UK and Australia) suburban communities. The study primarily attempts to enrich the existing postsocialist suburbanization research by providing new knowledge of the neglected topic of suburban crime and safety. Bearing in mind the exclusively suburban setting in a specific geographical context, the generalizability of our findings is limited and the study does not aim to contribute to broader conceptualisation of crime and fear of crime that are firmly established in Western criminological literature. Crime and safety in suburban neighbourhoods Despite substantial scholarly interest in suburban crime, most of the literature comes from the context of American suburbs. The urban crisis of the 1960s and the problems experienced by declining central cities (e.g. crime, unemployment, racial tension, growing welfare costs, higher taxes) boosted city-tosuburb movement and suburban growth in US metropolitan areas (Baldassare, 1992; Frey, 1979; Stahura et al., 1980). Katzman (1980) found that families with children and more affluent households are particularly attracted to safer (suburban) neighbourhoods. Already in the 1970s studies on the effects of suburban growth indicated that many of the problems once associated with central cities were also emerging in suburbs, including rising crime and delinquency, traffic congestions and pollution, poverty and unemployment, declining community life or social and political tensions (Baldassare, 1992; Brown, 1982; Stahura and Sloan, 1988). It was in this context that suburban crime received more attention. Existing research on the spatial distribution of urban crime points to variations in crime risks within metropolitan areas. Higher crime rates and high-risk neighbourhoods tend to be located in central and inner cities and on the poorest council estates (in the inner city or on the periphery), while low-risk neighbourhoods include affluent suburbs and retirement areas (Fyfe, 2000). Despite generally lower suburban crime rates, some deviations from this general pattern may appear when various types of crime are investigated separately. For example, Brown (1982) showed that some of the Chicago suburbs displayed higher property crime rates compared to central cities. She related the finding to the location of commercial activities and the high volume of non-resident and day-time populations in these suburbs. The rather Downloaded from eur.sagepub.com by guest on July 25, 2014 3 Temelová et al. exceptional higher-than-city rates of violent crime occurred in suburbs with a high proportion of black residents (Brown, 1982). Analysing data from the Philadelphia metropolitan area, Hakim (1980) pointed to higher rates of residential burglary in wealthy suburbs and suggested a property crime spillover from the adjacent urban centre to wealthier and more accessible suburbs. Generally, literature associated (increased) suburban crime with several phenomena that have been taking place within metropolitan regions: suburbanization of minority and poor populations that generate higher offender rates; suburbanization of retailing and manufacturing activities, which attract property crime; rising mobility of urban offenders and improving accessibility of suburbs; and adaptation of certain aspects of the criminal lifestyle by the suburban middle class (Brown, 1982; Hakim, 1980; Stahura and Sloan, 1988). Several studies, however, pointed to variation in crime victimization at a small geographical scale and emphasized the importance of the micro-environmental dimension in criminological analyses (Bottoms et al., 1987; Morgan, 2001). Brantingham and Brantingham (1995) introduced the term crime attractors to refer to places, areas, neighbourhoods and districts that attract intending offenders because of the known opportunities for particular types of crime. Although they did not refer explicitly to residential suburbs, affluent single-family housing districts can be regarded as attractors for property crime. Explanations of crime occurrence at a local level are built around the fact that various neighbourhoods are exploited for different kinds of crime. The physical properties of the environment and local population characteristics have been generally recognized as significant predictors of crime rates in an area. Moreover, Elffers (2003) and recently Hirschfield et al. (2013) also pointed to the influence of the characteristics of nearby surrounding areas on a neighbourhood’s crime risk, thus suggesting the existence of significant between-area effects. Boggs (1965) identified the familiarity of offenders with their targets as an important crime determinant in high-offender neighbourhoods, while profitability appeared a more relevant component for the types of crime that occur in neighbourhoods of a high social rank. It follows that differences in determinants of suburban violent and property crime exist. While violent crime is more related to the presence of motivated offenders (low-income and minority populations), property crime is more sensitive to opportunities for crime (wealth concentration, commercial land use, housing quality and structure), guardianship and security (police protection, property surveillance, private insurance, neighbourly control) (Baumer, 1985; Brown, 1982; Hakim, 1980; Hope, 1999; Katzman, 1980; Stahura et al., 1980; Stahura and Sloan, 1988). Brown et al. (2004) argued that place attachment may also prevent risk of crime through promoting territorial guardianship of one’s own and neighbouring properties, spending more time in the home area, marking the territory, upkeep of home and yard and reinforcing neighbourly cohesion and watchfulness. Staying with property crime in residential suburbs, which appears the most relevant given the focus of this study, Morgan (2001) explained small-scale variations in burglary patterns in a Perth suburb (Australia) by differences in opportunities for burglary, ease of access to dwellings, expected payoffs from a burglary and the routine activities of residents and visitors to the suburb. Cornish and Clark (1986) depicted some more concrete factors related to opportunity, effort and proximal risks that influence an offender to select a particular middle-class suburb and a particular house for residential burglary: the selected area is easily accessible, with few police patrols, low-security housing, larger gardens and a burgled home is likely to be one where no-one is at home, which appears especially affluent, is detached and on a corner site, has patio doors and bushes (or other) cover. Using British crime data, Hope (1999) found that car ownership and dwelling detachedness is associated with higher risks of property crime, suggesting that suburban types of housing and affluent suburban households are more at risk. Although the connection between crime rates and fear of crime experienced by residents has been often accepted as an evident fact (Lavrakas, 1982; Lewis and Maxfield, 1980), several studies have pointed to the ambiguity of this relationship (e.g. Balkin, 1979; Hale, 1996; Jackson, 2006). Balkin (1979) argued that an inverse relationship between fear of crime and measured victimization rates can occur, as Downloaded from eur.sagepub.com by guest on July 25, 2014 4 European Urban and Regional Studies victimization rates do not take into account people’s exposure to crime, which appears to be negatively related to fear of crime. Besides the inaccurate measurement of real crime risks, Lorenc et al. (2012) suggested that fear of crime may be driven more by factors such as the physical environment or interpersonal relationships than crime itself. Kasl and Harburg (1972) further showed that familiarity with events of crime and violence in a neighbourhood strongly influences residents’ perceptions of safety; thus, people living in high-stress areas (i.e. low socio-economic status, high crime rates, marital and residential instability) were more likely to evaluate their neighbourhoods as unsafe compared to residents of low-stress neighbourhoods. According to Lavrakas (1982), areas located closer to the central city are supposedly associated with a higher fear of crime, while greater security is experienced in suburban areas. He supported the assumption in an empirical study comparing fear of crime and behavioural restrictions in urban and suburban neighbourhoods of the Chicago metropolitan area; residents of city neighbourhoods feel and act as significantly more fearful than their suburban counterparts (Lavrakas, 1982; see also Baumer, 1985). Foster et al. (2010) pointed out that in new suburban housing developments with relatively low crime levels, property crime generates higher fear among residents (new home buyers). Studying the relationship between people’s victimization experience and fear of crime, Skogan (1987)explained an increase in measured effects of property victimization by more property crime taking place in people’s immediate neighbourhood and more frequent occurrence of related victimization. A few existing studies that elaborated on fear of crime in suburban communities in more detail pointed to the important role of individual differences in residents’ perceptions of safety. Wood et al. (2008) demonstrated that gender and income were significant predictors of feelings of safety in suburban areas (see also Foster et al., 2010). Baumer (1985) documented that age differentiates fear of crime strongly in cities, somewhat less in suburbs and very weakly in rural areas, suggesting that in low-crime areas elderly are not more fearful. Similarly, Lavrakas (1982) concluded that being a male, of higher income and younger age is associated with greater feelings of safety both in city and in suburbs, whereas in the suburbs, sex is the primary predictor of the level of fear one experiences in his or her neighbourhood. These findings are generally consistent with the results of numerous research projects conducted in various types of urban environments (e.g. Bannister and Fyfe, 2001; Baumer, 1985; Jíchová and Temelová, 2012a; Skogan, 1987). In addition to individual differences, the role of the suburban residential environment (social and physical characteristics of neighbourhoods) has been recognized in the literature not only in relation to the incidence of crime, but also with regard to perceptions of safety. As for suburb-level physical variables, neighbourhood upkeep, fewer signs of incivility and disorder, better walkability and living in a detached house were found to reduce suburbanites’ fear in their neighbourhoods (Foster et al., 2010, 2011; Lavrakas, 1982; Wood et al., 2008). Foster et al. (2011) illustrated that house design and upkeep, which create opportunities for natural surveillance, reflect residents’ guardianship and discourage expressions of antisocial behaviour, improve perception of safety in the suburban landscape. As for local social attributes, higher social capital, cohesion, informal control and supportiveness of neighbours were shown to be positively associated with feelings of safety in suburbs (Doeksen, 1997; Foster et al., 2010; Lavrakas, 1982; Wood et al., 2008). Another well-established association between personal experience of crime and level of fear was confirmed by Foster et al. (2010) also in research of suburban communities. Several studies suggested that not only were victims more concerned about crime, but indirect (vicarious) victimization, that is, knowing someone else (relative, friend, neighbour) who fell victim to crime or hearing about nearby criminal events from friends or news reports, also sparks fear in people (Baumer, 1985; Lavrakas, 1982; Skogan, 1987). Addressing crime in post-socialist suburbanization research In contrast to the fairly strong interest in suburban crime overseas, there has been little or no research on suburban crime in CEE countries. Although there Downloaded from eur.sagepub.com by guest on July 25, 2014 5 Temelová et al. is a body of scholarly studies exploring various aspects of crime in post-socialist societies, the majority of them lack a geographical dimension. Only a few studies have discussed crime in relation to space and environmental settings in an urban context. Meško et al. (2008) compared differing levels of fear of crime in two post-socialist capitals, Ljubljana (Slovenia) and Sarajevo (Bosnia and Herzegovina). Several studies investigating fear of crime in various residential urban neighbourhoods have also been published in Prague (the Czech Republic) (Jíchová and Temelová, 2012a, 2012b; Temelová et al., 2012). The geography of crime and its explanatory variables were discussed in the case of Baltic capital cities: Tallinn (Estonia) (Ceccato, 2009; Ceccato and Oberwittler, 2008) and Vilnius (Lithuania) (Ceccato and Lukyte, 2011). Although crime and space are the focus of all these studies, they are concerned solely with the urban environment and do not tackle spaces beyond the city boundaries. One of the reasons for the lack of interest in suburban crime is certainly the low magnitude of the problem and the fact that scholars have paid attention to more severe issues of new suburban development in CEE. These include spatial patterns of suburbanization (e.g. Dövényi and Kovács, 2006; Ouředníček, 2007), the role of planning and actors in new residential development (e.g. Leetmaa et al., 2009; Timár and Váradi, 2001), changing population structures in suburban areas (e.g. Kährik and Tammaru, 2008; Tammaru and Leetmaa, 2007), residential decision-making and satisfaction of suburbanites (e.g. Kährik et al., 2012; Kopečná and Špačková, 2012; Krisjane and Berzins, 2012), their time–space activity and daily mobility (e.g. Krisjane et al., 2012; Novák and Sýkora, 2007), the social contacts and cohesion of suburban communities (e.g. Špačková and Ouředníček, 2012), and experiences of suburban living (e.g. Hirt, 2008). While our research draws on some of the research findings from Western cities, the character of CEE suburbanization and its implications for suburban crime in the post-socialist context needs to be made. The suburbanization of well-educated and affluent households after the collapse of the communist regime in CEE was conditioned by the opportunities that the new economic and political constellation offered to higher income families in search of improved housing conditions (Kährik et al., 2012). Although housing desires (e.g. to own a house with a garden) and life-course-related reasons (e.g. family career) were the main factors, the image of a clean, safe and child-friendly suburban environment also influences households’ decisions to relocate from the central city to new dwellings on the urban fringe (Kährik et al., 2012; Kopečná and Špačková, 2012; Krisjane and Berzins, 2012). New residential suburbs built in the hinterlands of the major cities of CEE are thus middle- or higher income communities dominated by low-density and single-family housing units (Kährik and Tammaru, 2008; Krisjane and Berzins, 2012; Špačková and Ouředníček, 2012). Families with small children, young couples starting a family, and also older couples with grown-up children are among the most common newcomers to suburban zones (Ouředníček, 2003). New housing development has mainly taken the form of detached family dwellings built on the edge of existing rural settlements (Ouředníček, 2003) and it rarely creates suburban gated communities (Brabec and Sýkora, 2009).Unlike in some of the US suburbs, there have so far been no signs of decline or social degradation (e.g. as a result of poverty and the suburbanization of minority groups) and new-built suburbs are preserving their affluent profile. Accordingly we can assume that new residential suburbs in CEE metropolitan areas are characterized by low crime levels and high feelings of personal safety experienced by suburbanites. Further, it can be expected from the (Western) literature that middle- and higher class suburbs with single-family housing attract mainly property-related criminal activity, which implies more concern about property-related offences among suburban residents. Consistent with the existing research findings, a differential perception of safety and crime-specific fear is assumed, particularly in relation to demographic characteristics and victimization experience. Study area The area under investigation in this study covered the suburban zone of Prague. We employed different Downloaded from eur.sagepub.com by guest on July 25, 2014 6 European Urban and Regional Studies Figure 1. Zoning of Prague metropolitan region and the location of the surveyed municipalities. Source: Adapted from Ouředníček et al. (2012, 2013). geographical contexts for different data sources; registered crime patterns were investigated in the whole suburban zone of Prague and a questionnaire survey to examine feelings of safety took place in selected municipalities within the zone. The suburban zone of Prague includes the territory of Prague’s hinterland most intensively hit by residential suburbanization.2 It was delimited by Špačková et al. (2012) on the basis of intensities of housing construction and migration flows (see also Ouředníček et al. (2013). The case study municipalities for the questionnaire survey were selected according to the intensity of housing construction and their location within Prague’s hinterland. The aim was not only to select the municipalities most exposed to new housing development, but also to get a reasonable spatial distribution across Prague metropolitan region. As a result, six municipalities (Chýně, Hovorčovice, Jesenice, Květnice, Velké Přílepy, Zdiby) were selected on the basis of the above criteria (for a deeper analysis of the intensity of new housing construction in the Czech Republic, see Temelová and Ouředníček, 2011). Their location and basic characteristics related to population growth and housing construction are displayed in Figure 1 and Table 1. Data and analysis To discuss crime rates in the Prague metropolitan region the study used official police-recorded statistics provided by the Police Presidium of the Czech Republic. One of the shortcomings of this source is that it covers only registered incidents, while nonreported victimization remains unrecorded. There is a body of literature suggesting that unreported crime forms a major part of criminal activity in many countries (Balkin, 1979; Goudriaan et al., 2006; Skogan, 1977), including the Czech Republic (Marešová and Scheinost, 2001). Yet, the influence of unreported criminal victimization on citizens’ feelings of safety is undeniable (Balkin, 1979; Jíchová and Temelová, 2012b; Marešová and Scheinost, 2001). There are various explanations for non-reporting, ranging from the seriousness of victimization experiences and personal/behavioural features of the victims to the level of confidence in the police and justice system and neighbourhood social cohesion (Goudriaan et al., 2006; Skogan, 1977). The existing research suggests that underreporting concerns mainly less serious incidents, that is, minor property offences or less momentous crime against the person, although some violent acts, such as domestic abuse or sexual offences, often Downloaded from eur.sagepub.com by guest on July 25, 2014 7 Temelová et al. Table 1. The basic characteristics of the surveyed municipalities. Population size (2011) Population growth (1991–2011) (1991 = 100) Net migration ratea (1998–2011) Number of new-built dwellings (1998–2011) Share of new-built dwellings in total housing stock (2011) Intensity of new housing constructionb (1998–2011) Jesenice Velké Přílepy Zdiby Hovorčovice Chýně Květnice 7628 434 3313 389 3013 310 2361 242 2006 368 1672 1454 162 2127 162 577 120 633 74 475 147 597 681 451 80% 53% 68% 60% 81% 79 53 48 34 75 77% 293 Source: Czech statistical office. Note: aAverage annual net migration per 1000 inhabitants; the value for the wider region to which the municipalities administratively belong (Central Bohemia) is 11. bAverage annual number of completed dwellings per 1000 inhabitants; the value for the wider region to which the municipalities administratively belong (Central Bohemia) is 5. go unreported too (Goudriaan et al., 2006; Skogan, 1977). On the contrary, other types of criminal events, such as homicide, car theft or burglary, are relatively fully reported (Marešová and Scheinost, 2001; Skogan, 1977). The analysis of registered crime rates used data for the period 2000–2011 calculated in three-year averages and by 10,000 inhabitants. Apart from total crime rates, attention was also paid to particular types of violent and property crime. The analysis of registered crime patterns was carried out in urban and suburban zones of the Prague metropolitan region (see Figure 1), which were aggregated from local police departments (within Prague) and regional police departments (in Prague’s hinterland). The reason for such rough spatial specification was that police-recorded statistics are not available at the level of municipalities (several municipalities are merged under one police district in Prague’s hinterland). The delimitation of urban zones proposed by Ouředníček et al. (2012) was applied within the administrative borders of Prague. The city was divided into four concentric zones according to the city’s historical growth and the related character of the residentially built-up area: historical core, inner city, outer city and periphery (see Figure 1, for details see also Ouředníček et al., 2012). The data on residents’ perception of crime and safety were collected by way of a questionnaire survey conducted during summer/autumn 2010 and 2011. The survey was carried out with residents living in new suburban housing in six municipalities of Prague’s suburban zone, which had been exposed to intensive housing construction (for details see above). Information was gathered from respondents about perception of safety in their home area, crimespecific fear, crime-related experiences in the neighbourhood, property and personal protection measures, neighbourly relations and personal characteristics (socio-demographics and housing, previous place of residence). The survey sample utilised accessibility sampling. Accessibility sampling was selected following a pilot survey which confirmed that directly approaching randomly selected households does not elicit desirable results due to the high sensitivity of the topic of safety (people refused to participate in the survey if they were approached in their “home” territory). Therefore, a sampling strategy was chosen by which trained interviewers approached the potential survey participants in a “neutral” space, i.e. on the streets of the selected survey municipalities. In this way the anonymity of respondents was enhanced, since no direct connection between a respondent and his/her place (house) of residence could be made. Questioning by trained interviewers and the strategy of accessibility sampling dramatically increased the responses, as only a few people refused to participate. Every Downloaded from eur.sagepub.com by guest on July 25, 2014 8 European Urban and Regional Studies passer-by was approached with an initial question on the place of residence and the type of dwelling they occupied, which filtered out those who live in the area and who moved to new housing construction (built after 1990) as survey participants. Various day and week times and different neighbourhood locations were deliberately chosen in order to contact potential respondents and to obtain a sample that accurately reflected the wider population of migrants to new suburban housing. The survey sample included 615 adult respondents living in families with or without dependent children (i.e. under 15 years old). The sample was divided more or less equally by sex of respondents (54 % females). Table 2 shows the distribution of respondents by age and educational attainment against the distribution gained from the 2011 census for the surveyed municipalities. Unfortunately, the precise calculation of sub-groups in the population of movers to new housing was not possible due to the unavailability of such data. The structure of newcomers could be, however, assessed from the appropriate segmentation of the total municipal population and assumptions based on existing studies of suburbanization in the Prague metropolitan region. Movers to new suburban housing generally belong to higher or middle-income categories and are considerably younger and better educated than the original population of the municipalities (Ouředníček, 2003, 2007; Špačková and Ouředníček, 2012). We believe that a higher share of younger age groups, a lower proportion of persons of late economically active and retirement age and more people with university-level education in our sample is a good reflection of the structure of migrants moving to new housing in the suburban zone of Prague as described in the literature. The analysis of the questionnaire survey was designed to explore the associations between sociodemographic, housing-related and behavioural characteristics of suburban residents and their perceptions of crime and safety in the neighbourhood. The personal/household characteristics included gender, age, education, economic status, family status (children under 15 years living at home), victimization (own or family member), length of residence, dwelling type, security measures and perception of social relations. Table 2. Structure of respondents by age and educational attainment. Age group 15–24 25–39 40–54 55 + Education Basic and lower secondary Higher secondary University Survey samplea (proportion of respondents) Survey municipalitiesb (proportion of population) 18% 45% 23% 14% 12% 40% 26% 23% 17% 31% 47% 43% 36% 27% Source: aQuestionnaire survey 2010/2011; bCzech statistical office (census 2011). As well as the general perception of neighbourhood safety, fear of property crime (burglary, car theft/ break-in) and fear of violent crime (street assault, harassment or conflict with a stranger) were analysed separately. The examination of general feelings of safety in suburbs was included in the empirical section, since no such study has been done either in the Czech Republic or in CEE urban regions. The explanatory variables included in our analysis are established items in research into suburban fear of crime (see e.g. Baumer, 1985; Foster et al., 2010, 2011; Lavrakas, 1982; Wood et al., 2008). Appendix 1 summarizes the variables by response categories and their frequencies. Unfortunately, it was not possible to analyse crime occurrence associated with feelings of safety. The territories of regional police departments (basic units for police-recorded statistics in Prague’s hinterland) cover larger areas consisting of several municipalities, which does not allow the linking of police records on crime with the results of the questionnaire survey. The analysis of the survey data was carried out in two steps using SPSS software. Firstly, exploratory analyses of relations between respondents’ profiles and their perceptions of crime and safety in the neighbourhood were run. Chi-square tests of crosstabulations served to identify variables with a Downloaded from eur.sagepub.com by guest on July 25, 2014 9 Temelová et al. Share of respondents perceiving (in %) Category Burglary Car the/break-in Street assault Harassment or conflict with a stranger No or low fear1 Medium fear1 High fear1 44 27 29 48 74 87 22 23 10 Graphical distribuon in original categorizaon2 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 30 3 3 Figure 2. Crime-specific fear perceived by suburban residents. Source: Questionnaire survey 2010/2011 (n = 615). Note: 1No or low fear = original categories 1 and 2; medium fear = original category 3; high fear = original categories 4 and 5. 25 points in original categorization; 1 = not afraid at all, 5 = very afraid. significant relation to perceptions of crime and safety. Secondly, binary, ordinal and multinomial logistic regression was employed to build a compact model to better understand which factors enhance or reduce fear of crime in the suburban environment and to what extent. Because of the low frequency of responses in some categories and the limited size of our sample, the original categorization obtained from questionnaires was estimated. General perceptions of neighbourhood safety evaluated originally in four categories (safe, rather safe, rather unsafe, unsafe) were simplified into a binomial safe/unsafe variable. In the case of property crime the distribution of responses allowed us to reduce the original fivedegree categorization ranging from “not afraid at all” to “very afraid” into a three-degree categorization (no/low fear, medium fear, high fear) (see Figure 2). The very low frequency of “fear” responses in the case of violent crime made it necessary to recode the variable into two categories only (no fear/at least some fear). Depending on the binomial or multinomial character of the dependent variable, binary or ordinal/ multinomial logistic regression was applied. A manually controlled step-wise backward approach was used to find the final refined models. Logistic models for general perceptions of neighbourhood safety and fear of violent crime did not adequately depict the variability of the dependent variable, therefore only bivariate chi-square tests were employed. However, the modelling was successful for the fear of property-related crime (burglary and car theft/ break-in). Multinomial logistic regression models offered much better results measured by pseudo R2 than the ordinal ones. A weaker performance of ordinal models is consistent with the results obtained in the exploratory analysis. The factors behind “no- or low-fear” perception on one hand and “high-fear” perception on the other hand are not ordinarily related to each other in our case. A comparison of pseudo R2 for the full and the final refined models suggested that the loss of explained variation was small.3 Crime rates in the Prague metropolitan area Crime rates have significantly increased in the Czech Republic since the collapse of the communist regime in the early 1990s. A reduction of crime was evident after the turn of the century, particularly thanks to Downloaded from eur.sagepub.com by guest on July 25, 2014 10 European Urban and Regional Studies Table 3. Crime ratesa in the Prague metropolitan region (2000–2011). Violent crimes Historical core Inner city Outer city Periphery Suburban zone Burglaries Thefts Total crime 2000–2002 2009–2011 2000–2002 2009–2011 2000–2002 2009–2011 2000–2002 2009–2011 103 32 22 18 21 61 20 13 13 14 194 123 94 94 182 209 97 62 69 93 2190 572 359 251 166 1417 374 248 209 162 3138 946 611 501 494 2311 663 433 413 391 Source: Police Presidium of the Czech Republic, Czech Statistical Office. Note: aAverage annual number of offences per 10,000 inhabitants in the periods 2000–2002, 2009–2011 respectively. the effect of prevention measures implemented by police and households, increased patrols and investigative activity by police, and also due to changes in legislation and statistical evidence. Prague, like other capitals, attracts criminal activity. It experiences about one fourth of offences in the country (but only one tenth of people) and the crime rate climbs to double the national average (Jíchová and Temelová, 2012b). The development of the crime rate in the Prague metropolitan region and its particular zones was consistent with the general decreasing trend during the last decade observed in the whole country. Criminal activity decreases from the city centre to the outskirts; the highest level of victimization is thus found in the historical core, while the suburban zone maintains the lowest crime incidences within the metropolitan region (see Table 3). When looking at various types of crime separately, property-related victimizations (thefts and burglaries) dominate, while personal offences are much less frequent in all zones of the Prague metropolitan region (see Table 3). Although the suburban zone maintains the lowest overall crime rate, burglaries occur significantly more often there, reaching the same rate as that of Prague’s inner city. The pattern is consistent with findings in other metropolitan areas (e.g. Brown, 1982; Hakim, 1980). Interestingly, burglary rates have dropped dramatically in Prague’s suburbs over the last decade; at the beginning of the 2000s they were among the highest within the metropolitan region and even dominated over thefts in this zone (see Table 3). The decrease relates mainly to the reduction of break-ins to weekend houses. On the contrary, burglaries to single-family houses increased in the suburbs during the reported period (contrary to Prague’s urban zones) and today the rate is the highest in the metropolitan region (see Table 4). Obviously, both trends in burglary occurrence are associated with the suburbanization process in Prague’s hinterland. Second homes have been increasingly adapted and used for permanent living (Fialová, 2012), which strengthened local social control, property surveillance and police protection in these areas. At the same time new family housing districts have attracted property crime offenders not only because of the wealthier residential environment found there, but also thanks to the easier accessibility of unoccupied and unfinished dwellings located in these areas. According to Foster et al. (2011), vacant lots and houses under construction represent territorial gaps that are vulnerable to crime due to limited social surveillance. Moore and Shepherd (2007) documented the association between levels of environmental graffiti or property damage with fear of crime. According to Lavrakas (1982), the perception of incivilities and local crime affect suburbanites’ fear in their neighbourhoods, although to a lesser extent than for city dwellers. Interestingly, crimes related to vandalism (breach of the peace, damage to other people’s property) are not negligible in Prague’s suburbs (see Table 4). Although the rate is low, the incidence of breaches of the peace is higher in suburbs than in urban zones (except for the historical core) and together with periphery these are the only zones where the ratio increased during the last decade (between 2000 and 2011). Compared to urban zones of Prague, the suburbs also show higher levels of crime due to intoxication. In view of the higher car dependency of residents in many suburban districts, Downloaded from eur.sagepub.com by guest on July 25, 2014 11 Temelová et al. Table 4. Crime ratesa for selected types of crime in the Prague metropolitan region (2009–2011). Historical core Inner city Outer city Periphery Suburban zone Burglaries into family houses Car thefts Thefts from cars Damage done to other people’s property Breach of the peace Intoxication 0.3 4.2 4.4 11.8 17.4 20.6 35.0 35.0 31.0 24.1 189.2 147.0 90.3 75.0 70.4 65.5 28.7 22.4 19.5 18.0 9.8 2.7 1.6 2.0 3.5 41.3 14.9 8.8 9.4 19.2 Source: Police Presidium of the Czech Republic, Czech Statistical Office. Note: aAverage annual number of offences per 10,000 inhabitants in the period 2009–2011. especially during evening and night hours, it is probably associated with drink driving, which makes the relation to local feelings of safety less straightforward. Perception of crime and safety in suburban communities of Prague A positive perception of safety in low-risk suburban areas was demonstrated in several studies on suburban crime (e.g. Baumer, 1985; Kasl and Harburg, 1972; Lavrakas, 1982). Our findings are consistent with these results, since the majority of our respondents feel safe and secure in their home neighbourhood (87%). There is, however, a differential perception of crime risk between various groups of suburban residents (see Table 5). Those who experienced victimization, personally or within their family, tend to consider their home area less safe than other residents. Our findings confirm the strong relationship between crime-related experiences and feelings of safety documented by many authors (Baumer, 1985; Foster et al., 2010; Lavrakas, 1982; Skogan, 1987). The proportion of people with crime-related experience in their current place of residence reaches 40% among our respondents. Surprisingly, it is comparable with the number found by Jíchová and Temelová (2012a) in inner city neighbourhoods of Prague. Residents’ own personal characteristics, particularly age, also accounted for fear of crime levels in suburban communities of Prague. While older (over 55 years) and retired residents reported lower feelings of safety, young people (15–24 years) and students felt considerably safer in the neighbourhood. Contrary to the findings of previous studies (e.g. Foster et al., 2010; Lavrakas, 1982; Wood et al., 2008), gender did not explain variation in perception of safety in the suburban communities examined in our research. The result is interesting, since women were acknowledged as being more fearful than men in most of the geographical surveys of fear of crime (Pain, 2001). The length of residency in the suburb clearly diminished residents’ perception of safety. While those living in a suburb for less than two years considered their neighbourhood safe and secure, the perception of longer term residents (between two and five years) was noticeably more negative. It is plausible to assume that experience living in the suburban environment, including victimization experience, goes hand in hand with “sobering up” from the initial idealism and with a more realistic evaluation of local risks. Feelings of safety are related also to people’s behavioural reactions to perceived crime risks. Over 90% of respondents claimed to apply some security measures to protect their own property or to enhance personal security. The owners of houses with alarm and safety camera systems tend to perceive higher crime risks in the neighbourhood. Feelings of suburban safety were also considerably lower for residents who said they did not leave the house or only drove by car after dark (the share of such people is, however, low in the sample: 4% and 9%, respectively). The effect of local social relations on fear of crime did not come out in our survey; however, we attribute it to the fact that the social environment was evaluated positively by the majority of our respondents (over 70% consider it as being very good or good). Downloaded from eur.sagepub.com by guest on July 25, 2014 12 European Urban and Regional Studies Table 5. Perception of neighbourhood safety by suburban residents. Profile of respondent Share of respondents perceiving neighbourhood as unsafe Victimization (own or family member) Age group 15–24 years 25–39 years 40–54 years over 55 years Economic status Employee Entrepreneur Maternity leave Retired Student Other Length of residency 2 years or less More than 2 and less than 5 years More than 5 years Security measures Alarm system Camera system After dark – driving by car only After dark – not leaving house All respondents 21.7*** 7.5* 11.8 14.6 20.7** 13.0 16.2 12.9 22.0** 7.1* 4.2 6.2*** 16.5** 14.6 15.2* 26.3*** 24.1*** 28.6** 12.9 Source: Questionnaire survey 2010/2011 (n = 615). Note: Level of significance according to Pearson chi-square tests: 0.01 ***, 0.05 **, 0.1 *. Crime-specific analyses of fear of crime seem to be important in understanding the perception of safety in low-crime neighbourhoods, which is also the case of Prague’s suburban communities. Consistent with the results of studies carried out elsewhere (Foster et al., 2010; Skogan, 1987), property crime generates greater fear among residents of Prague’s suburbs than violent crime (see Figure 2). The investment that households made in building a new home in a suburb logically raises their fear for the acquired property. In the following sections the fear associated with various types of property and violent crime is investigated separately. Fear of property crime Results of the applied statistical model indicate that victimization experience significantly increases a respondent’s likelihood of having a high fear of property crime (burglary and car theft or break-in) and simultaneously decreases the likelihood of falling into the no- or low-fear category (see Tables 6 and 7). Most of our respondents who claimed to have a crime-related experience in the suburbs specified their victimization in one or more of the property crime categories (e.g. burglary into house or garden, car theft, car break-in, theft from a construction site). Fear of residential break-in clearly grows with the age of respondents; while students feel little fear of this kind, older people tend to be much more fearful when it comes to burglary. Similarly, students tend to report low fear of car-related crime. The finding is consistent with the results obtained by Moore and Shepherd (2007), who found that a fear of personal loss was greatest at around 40–60 years and lower at 16–25 years. They attributed this to the varying importance attached to material wealth with age and greater costs of property loss in older age relative to their financial circumstances (Moore and Shepherd, 2007). The attachment to and concern about accumulated family property or car ownership is supposedly much lower in younger age groups. Car theft or break-in is the only type of crime where a differentiated perception of fear by gender came out, since women tend to be less fearful in this case (see Table 7). Interestingly, however, women on maternity leave reported a greater fear of car theft. We can hypothesize that the loss of a car would have a greater impact on daily life and coordination of activities/journeys in the case of women with small children. The influence of the length of residence on fear of burglary is more complicated. Probably fewer crime-related experiences of short-term residents and their positive expectations of a new suburban environment increase the chance of recent Downloaded from eur.sagepub.com by guest on July 25, 2014 13 Temelová et al. Table 6. Multinomial logistic regression model of fear of burglary (ref. medium fear). Profile of respondent No or low fear Victimization (ref. no victimization) Age group over 55 years (ref. not older) Economic status (ref. rest) Maternity leave Retired Student Length of residency (ref. more than 5 years) 2 years or less More than 2 and less than 5 years Dwelling type (ref. flat) Detached house Semi-detached house Security measures Alarm System Camera System Neighbourly Control Dog High fear Exp(B) 95% CI Exp(B) 95% CI 0.617** 1.583 (0.397–0.960) (0.584–4.294) 1.600** 2.787** (1.001–2.557) (1.051–7.393) 1.253 0.717 1.711* (0.717–2.189) (0.205–2.513) (0.983–2.978) 0.991 1.062 0.386** (0.535–1.837) (0.328–3.438) (0.178–0.836) 2.128*** 1.481 (1.223–3.702) (0.886–2.476) 1.217 2.172*** (0.638–2.322) (1.252–3.769) 0.257*** 0.352** (0.119–0.556) (0.153–0.808) 0.810 0.481 (0.344–1.907) (0.184–1.256) 1.040 2.125* 0.535*** 2.120*** (0.678–1.596) (0.930–4.852) (0.350–0.817) (1.326–3.392) 1.797** 2.180* 0.745 0.853 (1.112–2.902) (0.914–5.204) (0.471–1.179) (0.500–1.455) Source: Questionnaire survey 2010/2011. Note: –2LL = 797.483; df = 26; n = 597; pseudo R2 = 0.223; level of significance: 0.01***, 0.05**, 0.1*. newcomers falling into the no- or low-fear category (see Table 6). This holds true also for car-related property crime (see Table 7). However, those who have lived in the suburbs between two and five years tend to be more fearful of burglary even in comparison with residents who have lived in the neighbourhood for a longer time (reference category). Fear of property crime is also noticeably influenced by dwelling type (see Tables 6 and 7). People living in detached and semi-detached houses have a much lower chance of falling into the no- or low-fear category in the case of burglary. According to Hope (1999), dwelling detachedness is associated with higher risks of property crime (see also Cornish and Clark, 1986). Living in a flat, however, enhances the chance of expressing a high degree of fear of car theft or break-in. This is most likely related to less secure parking on the streets in front of apartment houses. The event model of burglary in a middle-class suburb proposed by Cornish and Clark (1986) suggests that installing an alarm and locks, the presence of a dog and someone at home reduces the likelihood of residential break-in. Our analysis suggests that safety measures have a dual influence on people’s fear of burglary. While some security measures increase feelings of safety (e.g. having a dog), people applying other means of protection (e.g. alarm or a camera system, neighbourly surveillance when going away from home, staying at home after dark) tend to report a greater fear of burglary (see Table 6). People living in houses surrounded by high fences are less likely to fear car thefts (see Table 7). The interviews carried out with Prague’s suburban residents by Hrbková (2012) showed that households installed home security systems (e.g. camera, alarm, locks) to enhance the protection of their own property. Our model shows that home security and behavioural prevention measures do not have any significant effect on greater feelings of safety. Getting a dog was, according to Hrbková (2012), primarily motivated by emotional reasons but having one makes the owners feel safer. Downloaded from eur.sagepub.com by guest on July 25, 2014 14 European Urban and Regional Studies Table 7. Multinomial logistic regression model of fear of car theft/break-in (ref. medium fear). Profile of respondent No or low fear Victimization (ref. no victimization) Gender (ref. male) Economic status (ref. rest) Maternity leave Retired Student Length of residency (ref. more than 5 years) 2 years or less More than 2 and less than 5 years Dwelling type (ref. flat) Detached house Semi-detached house Security measures Alarm system High fence After dark – not leaving house After dark – not going alone High fear Exp(B) 95% CI Exp(B) 95% CI 0.530*** 1.943** (0.334–0.842) (1.163–3.247) 2.069*** 1.446 (1.26–3.398) (0.818–2.558) 0.506** 0.746 1.876** (0.261–0.98) (0.317–1.757) (1.026–3.433) 0.586 0.805 0.350** (0.287–1.196) (0.326–1.984) (0.147–0.836) 1.890** 0.972 (1.04–3.437) (0.561–1.683) 0.784 1.282 (0.392–1.568) (0.713–2.306) 0.763 0.670 (0.316–1.84) (0.262–1.716) 0.305*** 0.272*** (0.130–0.714) (0.108–0.686) 1.641** 1.095 8.943** 0.461* (1.045–2.578) (0.683–1.756) (1.042–76.774) (0.2061.033) 2.556*** 0.633* 9.668** 0.323** (1.528–4.277) (0.370–1.085) (1.069–87.428) (0.122–0.859) Source: Questionnaire survey 2010/2011. Note: –2LL = 716.607; df = 26; n = 592; pseudo R2 = 0.26; level of significance: 0.01***, 0.05**, 0.1*. Fear of violent crime The rates of violent crime in Prague’s suburban zone are the lowest within the metropolitan region (see Table 3). This is accordingly reflected in little fear of personal harm (street assault of various motives, harassment or conflict with a stranger) reported by our respondents. Insufficient frequency of high-fear responses in violent crime categories (see Figure 2) did not allow us to build an adequately meaningful statistical model as presented in the analysis of property crime-related fear. The interpretation of differential fear of violent crime therefore relies on the results provided by a Pearson chi-square test. Age appears to be an important predictor of fear of violent crime in Prague’s suburban communities. Young people (15–24 years) reported a considerably higher fear of street assault, harassment or conflict with a stranger than other age groups (see Table 8). This corresponds with our next finding that students and people with lower secondary education (often students in the middle of their education career) are more fearful when it comes to personal harm. On the contrary, people in the early family stage (25–39 years) and partly also older people (over 55 years) fear violent crime less in their home area. Similarly, Moore and Shepherd (2007) documented that feelings of personal safety tend to increase with age, placing the greatest feelings of harm at around 16– 25 years. Changes in fear with age reflect different exposure to risks, which is largely determined by the lifestyles and routine activities of people in different life cycle stages (e.g. Ferraro, 1995). Young people are likely to have more exposure to, and experience of, aggression in suburbs when going out in the evenings than early stage families or older people with a more settled and home-based lifestyle. In particular, travelling at night from places of entertainment (the city) to suburbs by public transport (Novák and Sýkora, 2007) makes young people more vulnerable to crime risks. The association between previous experience and fear of violent crime did not however emerge, which is probably explained by the dominantly property-related victimization of suburban residents. Yet people who moved to the Prague metropolitan region from other parts of the country, as Downloaded from eur.sagepub.com by guest on July 25, 2014 15 Temelová et al. Table 8. Fear of violent crime perceived by suburban residents. Profile of respondent Gender Male Female Age group 15–24 years 25–39 years 40–54 years Over 55 years Education Lower Secondary Upper Secondary University Economic status Employee Entrepreneur Maternity leave Student Retired Other Previous place of residence Prague Elsewhere in the Czech Republic Length of residency 2 years or less More than 2 and less than 5 years More than 5 years Negative/neutral perception of social relations Security measures Security windows/doors Neighbourly control Tear gas After dark – not leaving house All respondents Share of respondent perceiving fear of Harassment or conflict with a stranger Street assault 11.3 12.5 11.0 16.2* 16.5* 9.3* 16.1* 5.9* 18.8* 10.3** 16.5 13.8 13.6 13.4 9.3 20.2** 14.1 10.5* 12.3 12.5 11.2 17.0* 8.2 12.0 12.4 14.1 10.3 18.0 18.4 12.0 10.6 15.2 11.6 19.4** 9.4 10.8 9.3** 12.1 16.6** 18.2*** 21.8*** 16.9* 14.6** 12.2 22.7** 23.8* 12.8 9.8** 34.8*** 28.6** 11.8 13.9 Source: Questionnaire survey 2010/2011 (n = 615). Note: Level of significance according to Pearson chi-square tests: 0.01***, 0.05**, 0.1*. well as those who have lived there for a longer time (over five years), appear to be more fearful when it comes to violent crime. There is a significant association between fear of violent crime and negative perception of social relations in the neighbourhood (see Table 8). Social cohesion, informal control and helpful neighbouring have been recognized as supportive to feelings of safety in suburbs (Foster et al., 2010; Lavrakas, 1982; Wood et al., 2008). Negative evaluation of local social ties and corresponding fear of crime might be, however, associated with the individuality of those people, who are generally distant and suspicious of their surroundings. Fear of street assault was felt less by respondents who stated that they often ask neighbours to keep an eye on their house when away from home. This indirectly indicates that in communities where neighbours are perceived as being willing to help each other, fear of violent crime may be lower. The effect of fear of personal harm on behavioural restrictions is clearly manifested in significantly higher feelings of being unsafe among those who carry tear gas or do not go out after dark (see Table 8). Discussion and conclusions The study investigated perceptions of crime and safety in suburban communities using a survey of residents of newly-built family housing districts in the hinterland of Prague. Although suburbanization has attracted a lot interest on the part of urban scholars in Central and East European countries, there have not been any studies focused on suburban crime and its perception by residents. In this respect our study extends the existing knowledge on various aspects of the suburbanization process in CEE countries. On the one hand, our empirical evidence confirms some of the findings of previous Western studies in a specific and previously uninvestigated suburban residential context of post-socialist cities; on the other hand, it also provides some contrary evidence of differential crime perceptions in Prague’s suburbs. Consistent with the findings from other contexts, mainly the US metropolitan areas (Boggs, 1965; Brown, 1982), Prague’s suburban zone is characterized by considerably lower crime rates than the city Downloaded from eur.sagepub.com by guest on July 25, 2014 16 European Urban and Regional Studies districts. As elsewhere (Hakim, 1980; Hope, 1999), wealthier residential environments with detached family houses attract mainly property-related crime. Our study showed that suburban residents generally perceive their home area as a safe and secure place, but fear of crime tends to strengthen with the years spent in suburbs and with the coming of victimization experiences. A strong association between fear of crime and previous victimization depicted by many researchers (Baumer, 1985; Lavrakas, 1982; Skogan, 1987) was confirmed also in our study of Prague suburban communities. Despite the contrary evidence provided by the literature on suburban crime (e.g. Foster et al., 2010; Lavrakas, 1982), gender did not prove to differentiate fear in new residential communities in Prague’s hinterland. One explanation may relate to gender expectations and different suburban experiences of men and women in relation to crime-specific perception of fear. Suburban women are the main care providers for children while men are the primary household breadwinners (Hirt, 2008). The main safety concern of females is likely to be their children’s and their own personal protection and in this regard suburbs with low violent crime rates are safer places than Prague’s urban zones. Prevailing property crime victimization in suburbs probably makes generally less fearful males more concerned about property losses or damages. Moreover, age proved to be a more important factor differentiating fear of crime under conditions of dominating property victimization. Fear of property crime tends to increase significantly with the age of people. We attribute this to the greater attachment of middle-aged residents to accumulated family assets and simultaneously lower priority of material values in the case of younger age groups. However, further qualitative-based investigation is needed to elucidate the role of gender and age differences in theperception of crime in the suburbs of CEE cities. Considering the identical nature and manifestations of the suburbanization process in other cities across the Czech Republic, we believe that our findings are more or less applicable for most of our suburban settlements. Variations between new residential suburbs as to their physical environment, the structure of newcomers and crime rates are rather low across the country and thus not likely to significantly influence findings in other locations. Certain variations might appear in economically disadvantaged regions marked by above-average unemployment, problems of poverty and higher crime rates. The evident association between victimization and feelings of safety is likely to reinforce the fear of crime perceived by suburban residents in these regions. Making some generalizations on the perceptions of suburban crime in other CEE countries is much more difficult. Locally specific features of the suburbanization process, varied levels of inequalities within societies and different crime rates across post-socialist Europe make any generalization questionable. With the results obtained in this study the discussion cannot go any further without more comparative research. Acknowledgements The authors also wish to thank to anonymous reviewers for valuable comments on an earlier version of this paper. Funding This work was supported by the European Social Fund in Mobilitas (research grants no. MJD338 and no. MJD146) and by the Grant Agency of the Charles University (project no. 6209/2012). Notes 1. 2. 3. It is with great sadness that the Editors of European Urban and Regional Studies report the tragic death of two of the authors of this paper, Jana Temelová and Jakub Novák. Their passing is a major loss to critical urban scholarship in the Czech Republic and to the body of work on urban transformations in Europe. Our thoughts are with the families and colleagues of Jana and Jakub. Residential suburbanization fully hit the Prague metropolitan region in the mid 1990s. Although the southern part of Prague’s hinterland has been most exposed to the process, at the turn of the century suburbanization had penetrated almost all municipalities located within the metropolitan region (Ouředníček, 2007). Pseudo R2 was 0.26 for the full multinomial logistic model of fear of burglary and 0.22 for the final model; in case of fear of car theft/break-in it was 0.30 for the full model and 0.26 for the final model. Downloaded from eur.sagepub.com by guest on July 25, 2014 17 Temelová et al. References Baldassare M (1992) Suburban communities. Annual Review of Sociology 18(1): 475–494. Balkin S (1979) Victimization rates, safety and fear of crime. Social Problems 26(3): 343–358. Bannister J and Fyfe N (2001) Introduction: fear and the city. Urban Studies 38(5–6): 807–813. Baumer TL (1985) Testing a general model of fear of crime: data from a national sample. Journal of Research in Crime and Delinquency 22(3): 239–255. Boggs SL (1965) Urban crime patterns. American Sociological Review 30(6): 899–908. Bottoms AE, Mawby RI and Walker MA (1987) A localised crime survey in two contrasting areas of a city. British Journal of Criminology 27(2): 125–154. Brabec T and Sýkora L (2009) Gated communities in Prague. In: Smigiel C (ed.) Gated and Guarded Housing in Eastern Europe (Forum IFL, Heft 11). Leipzig: Leibnitz-Institut für Länderkunde, pp. 83– 89. Brantingham P and Brantingham P (1995) Criminality of place: crime generators and crime attractors. European Journal on Criminal Policy and Research 3(3): 1–26. Brown BB, Perkins DD and Brown G (2004) Crime, new housing, and housing incivilities in a first-ring suburb: multilevel relationships across time. Housing Policy Debate 15(2): 301–347. Brown MA (1982) Modelling the spatial distribution of suburban crime. Economic Geography 58(3): 247–261. Ceccato V (2009) Crime in a city in transition: the case of Tallinn, Estonia. Urban Studies 46(8): 1611–1638. Ceccato V and Lukyte N (2011) Safety and sustainability in a city in transition: the case of Vilnius, Lithuania. Cities 28(1): 83–94. Ceccato V and Oberwittler D (2008) Comparing spatial patterns of robbery: evidence from a Western and an Eastern European city. Cities 25(4): 185–196. Cornish DB and Clark RV (1986) Introduction. In: Cornish D and Clark R (eds) The Reasoning Criminal: Rational Choice Perspectives on Offending. Secaucus, NJ: Springer-Verlag, pp. 1–16. Doeksen H (1997) Reducing crime and the fear of crime by reclaiming New Zealand’s suburban street. Landscape and Urban Planning 39(2–3): 243–252. Dövényi Z and Kovács Z (2006) Budapest: the postsocialist metropolitan periphery between ‘Catching up’ and individual development path. European Spatial Research and Policy 13(2): 23–41. Elffers H (2003) Analysing neighbourhood influence in criminology. Statistica Neerlandica 57(3): 347–367. Ferraro KF (1995) Fear of Crime: Interpreting Victimization Risk. New York: State University of New York Press. Fialová D (2012) Druhé bydlení v zázemí Prahy. In: Ouředníček M and Temelová J (eds) Sociální proměny pražských čtvrtí. Praha: Academia, pp. 229–249. Foster S, Giles-Corti B and Knuiman M (2010) Neighbourhood design and fear of crime: a socialecological examination of the correlates of residents’ fear in new suburban housing developments. Health & Place 16(6): 1156–1165. Foster S, Giles-Corti B and Knuiman M (2011) Creating safe walkable streetscapes: does house design and upkeep discourage incivilities in suburban neighbourhoods? Journal of Environmental Psychology 31(1): 79–88. Frey WH (1979) Central city white flight: racial and nonracial causes. American Sociological Review 44(3): 425–448. Fyfe NR (2000) The geography of crime. In: Johnston JR, Gregory D, Pratt G and Watts M (eds) The Dictionary of Human Geography. 4th ed. Oxford and Malden, MA: Blackwell Publishers, pp. 120–123. Goudriaan H, Wittebrood K and Nieuwbeerta P (2006) Neighbourhood characteristics and reporting crime: effects of social cohesion, confidence in police effectiveness and socio-economic disadvantage. British Journal of Criminology 46(4): 719–742. Hakim S (1980) The attraction of property crimes to suburban localities: a revised economic model. Urban Studies 17(3): 265–276. Hale C (1996) Fear of crime: a review of the literature. International Review of Victimology 4(2): 79–150. Hirschfield A, Birkin M, Brunsdon C, Malleson N and Newton A (2013) How places influence crime: the impact of surrounding areas on neighbourhood burglary rates in a British city. Urban Studies. Epub ahead of print 19 June. DOI:10.1177/ 0042098013492232. Hirt SA (2008) Stuck in the suburbs? Gendered perspectives on living at the edge of the post-communist city. Cities 25(6): 340–354. Hope T (1999) Privatopia on trial? Property guardianship in the suburbs’. In: Painter K and Tilley N (eds) Surveillance of Public Space: CCTV, Street Lighting and Crime Prevention (Crime Prevention Studies, vol. 10). New York: Criminal Justice Press, pp. 15–45. Hrbková J (2012) Kriminalita v suburbiích: případová studie v obci Sulice. Unpublished Bachelor Thesis, Downloaded from eur.sagepub.com by guest on July 25, 2014 18 European Urban and Regional Studies Department of Social Geography and Regional Development, Charles University in Prague, Albertov. Jackson J (2006) Introducing fear of crime to risk research. Risk Analysis 26(1): 253–264. Jíchová J and Temelová J (2012a) Kriminalita a její percepce ve vnitřním městě: případová studie pražského Žižkova a Jarova. Geografie 117(3): 329–348. Jíchová J and Temelová J (2012b) Kriminalita a riziková místa centrálního a vnitřního města: sonda do názorů obyvatel vybraných pražských čtvrtí. In: Temelová J, Pospíšilová L and Ouředníček M (eds) Nové sociálně prostorové nerovnosti, lokální rozvoj a kvalita života. Plzeň: Aleš Čeněk, pp. 46–76. Kährik A and Tammaru T (2008) Population composition in new suburban settlements of the Tallinn metropolitan area. Urban Studies 45(5–6): 1055–1078. Kährik A, Leetmaa K and Tammaru T (2012) Residential decision-making and satisfaction among new suburbanites in the Tallinn urban region, Estonia. Cities 29(1): 49–58. Kasl SV and Harburg E (1972) Perceptions of the neighborhood and the desire to move out. Journal of the American Institute of Planners 38(5): 318–324. Katzman MT (1980) The contribution of crime to urban decline. Urban Studies 17(2): 277–286. Kopečná M and Špačková P (2012) Rezidenční stabilita obyvatel pražského zázemí: případová studie Říčany u Prahy. In: Ouředníček M and Temelová J (eds) Sociální proměny pražských čtvrtí. Praha: Academia, pp. 206–228. Krisjane Z and Berzins M (2012) Post-socialist urban trends: new patterns and motivations for migration in the suburban areas of Rīga, Latvia. Urban Studies 49(2): 289–306. Krisjane Z, Berzins M, Ivlevs A and Bauls A (2012) Who are the typical commuters in the post-socialist metropolis? The case of Riga, Latvia. Cities 29(5): 334–340. Lavrakas PJ (1982) Fear of crime and behavioral restrictions in urban and suburban neighborhoods. Population and Environment 5(4): 242–264. Leetmaa K, Tammaru T and Anniste K (2009) From priority-led to market-led suburbanization in a postcommunist metropolis. Tijdschrift voor economische en sociale geografie 100(4): 436–453. Lewis DA and Maxfield MG (1980) Fear in the neighborhoods: an investigation of the impact of crime. Journal of Research in Crime and Delinquency 17(2): 160–189. Lorenc T, Clayton S, Neary D, Whitehead M, Petticrew M, Thomson H, Cummins S, Sowden A and Renton A (2012) Crime, fear of crime, environment, and mental health and wellbeing: mapping review of theories and causal pathways. Health & Place 18(4): 757–765. Marešová A and Scheinost M (2001) Trendy kriminality v ČR z pohledu roku 2000. Sociologický časopis/Czech Sociological Review 37(1): 23–41. Meško G, Fallshore M, Muratbegovič E and Fields CH (2008) Fear of crime in two post-socialist capital cities: Ljubljana, Slovenia and Sarajevo, Bosnia and Herzegovina. Journal of Criminal Justice 36(6): 546–553. Moore S and Shepherd J (2007) The elements and prevalence of fear. British Journal of Criminology 47(1): 154–162. Morgan F (2001) Repeat burglary in a Perth suburb: indicator of short-term or long-term risk? In: Farrell G and Pease K (eds) Repeat Victimization (Crime Prevention Studies, vol. 12). Monsey, NY: Criminal Justice Press, pp. 83–118. Novák J and Sýkora L (2007) A city in motion: time-space activity and mobility patterns of suburban inhabitants and structuration of spatial organization in Prague metropolitan area. Geografiska Annaler: Series B Human Geography 89(2): 147–168. Ouředníček M (2003) Suburbanizace Prahy. Sociologický časopis/Czech Sociological Review 39(2): 235–253. Ouředníček M (2007) Differential suburban development in the Prague urban region. Geografiska Annaler: Series B Human Geography 89(2): 111–125. Ouředníček M, Pospíšilová L, Špačková P, Temelová J and Novák J (2012) Prostorová typologie a zonace Prahy. In: Ouředníček M and Temelová J (eds) Sociální proměny pražských čtvrtí. Praha: Academia, pp. 268–297. Ouředníček M, Špačková P and Novák J (2013) Metodické problémy výzkumu a vymezení zón rezidenční suburbanizace v České republice. In: Ouředníček M, Špačková P and Novák J (eds) Sub Urbs: krajina, sídla a lidé. Praha: Academia, pp. 309–332. Pain RH (2001) Gender, race, age and fear in the city. Urban Studies 38(5–6): 899–913. Skogan WG (1977) Dimensions of the dark figure of unreported crime. Crime & Delinquency 23(1): 41–50. Skogan WG (1987) The impact of victimization on fear. Crime & Delinquency 33(1): 135–154. South SJ and Crowder KD (1997) Residential mobility between cities and suburbs: race, suburbanization, and back-to-the-city moves. Demography 34(4): 525–538. Špačková P and Ouředníček M (2012) Spinning the web: new social contacts of Prague’s suburbanites. Cities 29(5): 341–349. Downloaded from eur.sagepub.com by guest on July 25, 2014 19 Temelová et al. Špačková P, Ouředníček M and Novák J (2012) Zónyrezidenčnísuburbanizace 2010. Available at: http://www.atlasobyvatelstva.cz/suburbanizace. Stahura JM and Sloan JJ (1988) Urban stratification of places, routine activities and suburban crime rates. Social Forces 66(4): 1102–1118. Stahura JM, Huff CR and Smith BL (1980) Crime in the suburbs: a structural model. Urban Affairs Review 15(3): 291–316. Tammaru T and Leetmaa K (2007) Suburbanisation in relation to education in the Tallinn metropolitan area. Population Space and Place 13(4): 279–292. Temelová J and Ouředníček M (2011) Bytová výstavba [Housing construction]. In: Ouředníček M, Temelová J and Pospíšilová L (eds) Atlas sociálně prostorové diferenciace České republiky. Praha: Nakladatelství Karolinum, pp. 105–109. Temelová J, Čermák D and Jíchová J (2012) Kriminalita a vnímání bezpečnosti v pražských čtvrtích. In: Ouředníček M and Temelová J (eds) Sociální proměny pražských čtvrtí. Praha: Academia, pp. 47–67. Timár J and Váradi D (2001) The uneven development of suburbanisation during transition in Hungary. European Urban and Regional Studies 8(4): 349–360. Wood L, Shannon T, Bulsara M, Pikora T, McCormack G and Giles-Corti B (2008) The anatomy of the safe and social suburb: an exploratory study of the built environment, social capital and residents’ perceptions of safety. Health & Place 14(1): 15–31. Downloaded from eur.sagepub.com by guest on July 25, 2014 20 European Urban and Regional Studies Appendix 1 Appendix 1. (Continued) Description of explanatory variables included in the analysis Explanatory variables Explanatory variables Gender Age group Education Economic status1 Children under 15 years at home Dwelling type Length of residency Previous place of residence Social relations in neighbourhood1 Victimization (own or family member) Property or personal protection Response categories Male Female 15–24 years 25–39 years 40–54 years Over 55 years Lower secondary Upper secondary University Employeea Entrepreneura Maternity leaveb Retiredc Studentd Othera Yes No Detached house Semi-detached house Flat 2 years or less More Than 2 And Less Than 5 Years More Than 5 Years Prague Elsewhere in the Czech Republic Very gooda Gooda Neutralb Badb Cannot say Yes No Yes No Valid responses Abs. In % 284 330 112 273 141 88 104 286 220 218 101 116 51 103 25 361 251 423 122 46.3 53.7 18.2 44.5 23.0 14.3 17.0 46.9 36.1 35.5 16.4 18.9 8.3 16.8 4.1 59.0 41.0 68.8 19.8 70 182 273 11.4 29.6 44.4 158 25.7 443 172 72.0 28.0 152 284 121 15 39 244 369 570 43 24.9 46.5 19.8 2.5 6.4 39.8 60.2 93.0 7.0 Security measures2 Response categories Alarm system Camera system High fence Security windows/ doors Neighbourly control Dog After dark – not leaving house After dark – not going alone After dark – driving by car only Tear gas Valid responses Abs. In % 330 57 213 290 55.0 9.5 35.5 48.3 226 37.7 179 21 29.8 3.5 43 7.2 54 9.0 46 7.7 Source: Questionnaire survey 2010/2011. Note: 1Categories were collapsed in the final models according to the indicated superscript letters. 2Yes/no type of questions; number/proportion of “yes” responses out of 600 valid responses is shown. (Continued) Downloaded from eur.sagepub.com by guest on July 25, 2014