Feeling unsafe in Italy’s biggest cities
| Author | Riccardo Valente,Sergi Valera Pertegas,Joan Guàrdia Olmos |
| Published date | 01 July 2022 |
| Date | 01 July 2022 |
| DOI | http://doi.org/10.1177/1477370820932075 |
| Subject Matter | Articles |
https://doi.org/10.1177/1477370820932075
European Journal of Criminology
© The Author(s) 2020
Article reuse guidelines:
sagepub.com/journals-permissions
DOI: 10.1177/1477370820932075
journals.sagepub.com/home/euc
Feeling unsafe in Italy’s
biggest cities
Riccardo Valente
Rovira i Virgili University, Spain
Sergi Valera Pertegas
Joan Guàrdia Olmos
University of Barcelona, Spain
Abstract
Crime and fear of crime are key challenges for civic coexistence in contemporary cities,
specifically because of the unequal relationship between the two phenomena. In the case of Italy,
for instance, even though crime has been trending downward over the past few decades, people
are increasingly concerned about their safety. Based on survey data (N = 6002) collected in
Milan, Naples, Rome, and Turin, this research provides a cross-city comparison of the factors that
influence individuals’ feelings of unsafety. The results of a multi-group structural equation model
endorse the prior literature by revealing that being the victim of a crime predicts higher levels
of subjective unsafety. On the other hand, perceived disorder in the neighbourhood, community
disaffection, and objective and subjective measures of social exclusion are also consistent
predictors of residents’ fears in all four cities. All in all, the model outputs indicate that non-
criminal factors have a higher explanatory power than victimization for perceived unsafety. The
implications of these findings for urban safety management are discussed.
Keywords
Urban safety, victimization, perceived disorder, community integration, multi-group SEM
Introduction
Scholars in the field of fear of crime studies are increasingly focusing beyond the nega-
tive emotional state that may result following a given threatening situation. Indeed,
Corresponding author:
Riccardo Valente, Department of Geography, Rovira i Virgili University, Carrer de Joanot Martorell, 15,
43480 Vila-seca, Tarragona, Spain.
Email: riccardo.valente@urv.cat
932075EUC0010.1177/1477370820932075European Journal of CriminologyValente et al.
research-article2020
Article
2022, Vol. 19(4) 849–867
several non-criminal variables have been proven to affect individuals’ fear of crime.
These variables are associated with cognitive processes of threat estimation (Amerio and
Roccato, 2005; Gray et al., 2011), the perceived likelihood of being exposed to actual
risks (Jackson, 2011), and behavioural aspects (Rader, 2004), as well as personal, social
and situational conditions (Carro et al., 2010). It has also been argued that people may
misjudge the causes of the fear of crime. In particular, studies focusing on the urban
dimension of the fear of crime have revealed that crime-related fears could be linked to
the unpredictability surrounding contemporary daily life in urban settings (Bannister and
Flint, 2017; Pratt and Turanovic, 2015).
Three baseline models of fear of crime
Against this background, any model that tries to explain people’s reasons for being con-
cerned or fearful about crime should encompass a comprehensive set of predictors
beyond actual crime and victimization rates. Over recent decades, scholars in the field of
sociology and criminology have made a huge effort in this direction and have tested dif-
ferent theoretical approaches. Franklin, Franklin and Fearn (2008) have synthesized
these efforts and identified three different but complementary models of fear of crime.
In the first place, the vulnerability model focuses on the relationship between direct
or indirect victimization and personal vulnerability (Skogan, 1986). The concept of vul-
nerability is two-fold: on the one hand, authors have stressed the influence of physical
vulnerability related to age or gender, showing that the elderly and women are more
likely to develop high levels of fear of crime (Jackson, 2009; Tulloch, 2000). On the
other hand, social factors (for example, poverty, low educational attainment, unemploy-
ment, or low income) have also been proven to predict higher levels of fear (Rader et al.,
2012; Reichert and Konefal, 2017).
A second approach, labelled the disorder model, deals with the relationship between
perceived signs of environmental disorder and the absence of social control. This combi-
nation has been shown to increase individuals’ perception of risk and their fear of crime
(Acuña-Rivera et al., 2011; LaGrange et al., 1992). This model includes contributions
from the broken windows theory (Wilson and Kelling, 1982) and the literature on incivil-
ity (Brown et al., 2004; Hunter, 1978; Link et al., 2017).
Finally, the third approach is known as the social integration model, and is grounded
on the idea that a sense of belonging, community attachment, social ties, social participa-
tion, and personal investment in the neighbourhood all act as preventive factors that
reduce the fear of crime (Bursik and Grasmick, 1993; McGarrell et al., 1997). However,
although community-based variables might significantly reduce levels of fear (Adams
and Serpe, 2000; Brunton-Smith et al., 2014; Pitner et al., 2012; Scarborough et al.,
2010; Skogan and Maxfield, 1981), some scholars have questioned these effects. For
instance, Kanan and Pruitt (2002) found scant support for this hypothesis in their study
looking at a sample from a neighbourhood in Nashville, Tennessee. Similar conclusions
were reached by Hartnagel (1979) and Maxfield (1984). More recently, the work of
Swatt et al. (2013) provided contrasting evidence on the relationship between fear and
collective efficacy, the latter being the combination of a set of items relating to social
cohesion, willingness to intervene, and capacity for social control. Similarly, Yuan and
850 European Journal of Criminology 19(4)
To continue reading
Request your trial