See discussions, stats, and author profiles for this publication at: https://www.researchgate.net/publication/26655762
Correlates of local safety-related concerns in a
Swedish Community: A cross-sectional study
ARTICLE in BMC PUBLIC HEALTH · AUGUST 2009
Impact Factor: 2.26 · DOI: 10.1186/1471-2458-9-221 · Source: PubMed
CITATIONS
READS
18
40
4 AUTHORS, INCLUDING:
Agneta Kullberg
Toomas Timpka
7 PUBLICATIONS 58 CITATIONS
283 PUBLICATIONS 2,759 CITATIONS
Mid Sweden University
SEE PROFILE
Linköping University
SEE PROFILE
Available from: Toomas Timpka
Retrieved on: 09 February 2016
BMC Public Health
BioMed Central
Open Access
Research article
Correlates of local safety-related concerns in a Swedish
Community: a cross-sectional study
Agneta Kullberg*, Nadine Karlsson, Toomas Timpka and Kent Lindqvist
Address: Linköping University, Department of Medical and Health Sciences, SE-581 83 Linköping, Sweden
Email: Agneta Kullberg* - agneta.kullberg@liu.se; Nadine Karlsson - nadine.karlsson@liu.se; Toomas Timpka - tti@ida.liu.se;
Kent Lindqvist - kent.lindqvist@liu.se
* Corresponding author
Published: 8 July 2009
BMC Public Health 2009, 9:221
doi:10.1186/1471-2458-9-221
Received: 5 December 2008
Accepted: 8 July 2009
This article is available from: http://www.biomedcentral.com/1471-2458/9/221
© 2009 Kullberg et al; licensee BioMed Central Ltd.
This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0),
which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
Abstract
Background: Crime in a neighbourhood has been recognized as a key stressor in the residential
environment. Fear of crime is related to risk assessment, which depends on the concentration of
objective risk in time and space, and on the presence of subjective perceived early signs of imminent
hazard. The aim of the study was to examine environmental, socio-demographic, and personal
correlates of safety-related concerns at the local level in urban communities. The specific aim was
to investigate such correlates in contiguous neighbourhoods in a Swedish urban municipality.
Methods: A cross-sectional study design was used to investigate three neighbourhood settings
with two pair-wise conterminous but socially contrasting areas within each setting. Crime data
were retrieved from police records. Study data were collected through a postal questionnaire
distributed to adult residents (n = 2476) (response rate 56%). Composite dimensions of perceived
residential safety were derived through a factor analysis. Logistic regression analysis was used to
examine associations between high-level scores of the three safety-related dimensions and arealevel crime rate, being a victim of crime, area reputation, gender, age, education, country of birth,
household civil status and type of housing.
Results: Three composite dimensions of perceived residential safety were identified: (I) structural
indicators of social disorder; (II) contact with disorderly behavior; and (III) existential insecurity.
We found that area-level crime rates and individual-level variables were associated with the
dimensions structural indicators of social disorder and existential insecurity, but only individuallevel variables were associated with the dimension contact with disorderly behavior. Self-assessed
less favorable area reputation was found to be strongly associated with all three factors. Being
female accorded existential insecurity more than being a victim of crime.
Conclusion: We have identified environmental, socio-demographic, and personal correlates of
safety-related concerns in contiguous neighbourhoods in a Swedish community. The results of this
study suggest that residents' self-assessed area reputation is an important underlying mechanism of
perceived safety. We also found a difference in crime rates and safety-related concerns between
areas with blocks of flats compared with small-scale areas although the neighbourhoods were close
geographically.
Page 1 of 11
(page number not for citation purposes)
BMC Public Health 2009, 9:221
Background
Since the mid-19th century, public health practitioners
have understood that people's residential environment
and housing conditions influence health outcomes [1,2].
Crime is among those factors suggested to be most
strongly related to health in the housing environment,
both by direct exposure and indirectly by residents' perception of feeling unsafe [3,4]. Previous research suggests
that several factors influence residents' perceived safety in
a neighbourhood. These factors can be divided into three
main themes: crime experiences and fear of crime, demographic characteristics, and the design and quality of the
residential environment [5-9].
Crime experiences and fear of crime
Crime in a neighbourhood has been recognized as a key
stressor in the residential environment, when measured in
absolute numbers of reported crimes in the neighbourhood and as residents' personal experience of being a
crime victim. Fear of crime has also been found to have a
significant negative influence on residential quality, and
may have a negative impact on residents' mental health
comparable to crime itself [10,11].
Fear of crime is related to risk assessment, which depends
on the concentration of objective risk in time and space,
and on the presence of subjective perceived early signs of
imminent hazard [12]. Prolonged safety-related worry
and fear of crime is suggested to yield behaviour modification as it potentially decreases physical activity and limits the residents' personal freedom. A consequence of
safety-related worry may therefore be 'time-space inequalities' with implications for physical and mental health
[13,14].
However, crime is an elusive factor with many facets. Violent crime in the immediate environment does not seem
to have a great deal in common with every day offences,
such as car theft or damage, in relation to perceived safety.
Another aspect is the way the local newspaper communicates crime in specific neighbourhoods. Newspaper headlines about local crime may result in attitudes that divide
neighbourhoods into desirable and undesirable areas.
Stories retold by neighbours who have been victims of
crime themselves may result in negative attitudes, and
increase the perception of isolation among residents
[15,16]. There is thus a community-based contagious
process affecting residents' perceived safety in specific
areas through the influence of reports in the media and/or
others around them.
Demographic characteristics
Earlier studies have reported that individual variables
such as gender, age, social and economic resources influ-
http://www.biomedcentral.com/1471-2458/9/221
ence the potential risk of becoming a crime victim and
fear of crime. Even though women are subject to lower
crime victim rates than men, they are found to fear crime
more [17,18]. Similarly, the elderly fear crime to a higher
degree than young adults, even though they have the lowest risk of becoming a victim themselves [19]. Further,
Halpern [20] found that to be unmarried or socially disconnected from conventional society in the area where
you live, the more likely you are to become a victim of
crime.
At neighbourhood level, the proportion of residential heterogeneity, level of education, mobility rate and density of
the neighbourhood were found to be associated with perception of safety of the people living there [21,22].
Design and quality of the residential environment
Newman [23] presented the model of 'defensible space' in
planning and argued for a correlation between design and
function of the built environment and crime rates. He
observed that in small-scale neighbourhoods in which the
residents have good surveillance and control of shared
outdoor space, the possibility of achieving lower factual
crime rates and better maintenance is higher than in
neighbourhoods without the possibility of this informal
control. Newman also observed that residents who had to
share a common entrance and staircase to reach their
dwelling reported more perceived insecurity and a lower
sense of belonging to the neighbourhood.
Land-use mix is another aspect that has been suggested to
influence crime rates. For instance, non-residential property, such as shops, car-parking and public spaces, has
been found to increase the number of crimes [24]. There
is also evidence that visible signs such as litter and graffiti
in an area communicate anti-social behaviour and how
seriously the risk of becoming a crime victim has to be
taken. The perception that one's neighbourhood is unsafe
is likely a constant psychological irritant [25,26]. Several
studies have suggested that perceptions of safety and fear
of crime play an important role in explaining area differences in health through emotional and behavioural
responses on neighbourhood quality [27-29].
Factors such as the appearance of the neighbourhood and
the reputation of the area have thus been found to impact
on residential well-being [30,31] but few studies have
focused on how these characteristics influence safetyrelated concerns. Due to the fact that neighbourhoods are
embedded in a community context in which reputations
are developed, the rumours that evolve may affect residents' sense of belonging in different neighbourhoods in
different ways. If a neighbourhood is held to have a specific reputation, this will influence the way its residents are
perceived by others. Depending on the character of the
Page 2 of 11
(page number not for citation purposes)
BMC Public Health 2009, 9:221
http://www.biomedcentral.com/1471-2458/9/221
reputation, this influence could have either a positive or
negative effect on an individual's self-esteem [32,33].
Study aims
The aim of the study was to examine environmental,
socio-demographic, and personal correlates of safetyrelated concerns at the local level in urban communities.
The specific aim was to investigate such correlates in contiguous neighbourhoods in a Swedish urban municipality.
Methods
Study setting
The study was undertaken at the local level in a Swedish
urban municipality (population 42,000 in 2005). Six
housing areas with a total of 6300 residents (aged 20–79
years) were investigated. The selection of the housing
areas was made after consultation with municipality officials in order to choose areas geographically recognized by
local people. As a basis, the municipal administrative
office provided a map showing the areas geographically
defined for statistical purposes [34].
The six housing areas selected were geographically located
two-by-two, i.e. two areas were spatially contiguous to
each other in three different neighbourhood settings. The
areas were given fictitious names indicating the type of
property. Setting Alpha incorporates area Alpha-house
(detached houses) and Alpha-flat (blocks of flats); setting
Beta incorporates area Beta-mix (mixed tenure and types
of houses) and Beta-flat (blocks of flats); and setting
Gamma incorporates area Gamma-house (detached
houses) and Gamma-flat (blocks of flats).
The proportion of non-residential land-use was higher in
the three areas with blocks of flats. Within each of the
three settings there was a small local community centre
with both public and private services such as a grocery
store, day-care centre, bus transport and a common public
school, all with good access irrespective of the living area
within the setting. Within each setting, the residents also
shared a fair-sized green space for recreation.
The characteristics of the areas based on recorded data are
presented in Table 1. The population in each of the two
areas within the three settings contrasted with the other
demographically and socio-economically. The residents
in the small-scale areas were more socially advantaged
and more residentially stable than the residents in areas
with blocks of flats. Motor-vehicle density was higher in
the small-scale areas (Table 1).
Registered data on crime rates for each housing area were
collected from the local police office (Table 2). The data
on reported crime were grouped into five categories
according to the Swedish penal code: (a) crime against life
and health; (b) crime against freedom and serenity; (c)
theft and robbery; (d) crime of damage; (e) other crimes.
Information for 3 years (2003–2005) was included to
increase area stability for reported crime. In settings Alpha
Table 1: Structural features of the housing areas studieda
Setting Alpha
Setting Beta
Setting Gamma
Variable
Alpha-house
Alpha-flat
Beta-mix
Beta-flat
Gamma-house
Gamma-flat
Total populationb
Residents by type of propertya
Blocks of flats (%)
Detached houses (%)
Period of constructiona
Resident turnover (%)c
Motor-vehicle density (no./1000 inhabitants)d
Mean for whole municipality (index 100):
Gainfully employed 20–64 yearse
Disposable income >20 yearse
Housing allowancef
Social allowanceg
>12 years in school (%)b
High-income residents (%)h
1423
1918
1857
1680
794
1364
10 (0.7)
1413 (99.3)
1966–1975
107 (7.5)
430
1918 (100)
0
1966–1970
363 (18.1)
270
761 (41)
1096 (59)
1951–1960
272 (14.4)
380
1675 (99.7)
5 (0.3)
1966–1970
289 (17.4)
303
0
794 (100)
1971–75
45 (5.8)
464
1262 (93)
97 (7)
1961–1965
262 (18.4)
291
114
131
97
48
264 (26.1)
222 (22)
81
77
113
109
126 (9.5)
44 (3.3)
102
95
79
84
281 (21.5)
202 (15.4)
80
66
89
124
142 (12)
35 (3)
104
122
93
27
98 (16.9)
81 (14)
85
75
100
131
98 (9.7)
39 (4)
aData
source: Statistics Sweden.
30 September 2005.
cResidents' turnover 1 January 2005 to 31 December 2005.
dDate 31 December 2004.
eINKOPAK, 2004.
fHousing allowance for families with children (bostadsbidrag) as well as for pensioners (bostadstillägg).
gThe social allowance should give a reasonable standard of living.
hHigh income defined as ≥ 300,000 Swedish crowns (SEK)/year in 2004. US$1 = SEK6.6; EUR1 = SEK9.0; date 30 December 2004.
bDate
Page 3 of 11
(page number not for citation purposes)
BMC Public Health 2009, 9:221
http://www.biomedcentral.com/1471-2458/9/221
Table 2: Police reported crime in the housing areas studied1
Setting Alpha
Setting Beta
Setting Gamma
Alpha-house Alpha-flat
Beta-mix
Beta-flat
Gamma-house Gamma-flat
0.7 (2.4)
10.6 (5.8)
2.3 (3.1)
10 (9.8)
0.9 (2.5)
7.8 (5.2)
12.4 (12.2)
4.2 (12.7)
21.1 (12.7)
20.4 (60.8)
3.4 (10.1)
4.7 (13.9)
81.2 (48.7)
18.6 (11.2)
36.9 (22.1)
102.1 (100) 33.6 (100)
165.6 (100)
Mean number of police reported crime/1000 inhabitants per
year (%)
(a) Crime against life and health (homicide, manslaughter,
maltreatment)
2.6 (8.9)
23.5 (12.9) 3.3 (6.2)
(b) Crime against freedom and serenity (unlawful threat,
molest)
(c) Theft and robbery
(d) Crime of damage
(e) Other crimes
16.7 (58.5)
5.1 (17.9)
3.5 (12.2)
88.3 (48.5) 36.1 (48.7) 49.3 (48.2)
30.8 (16.9) 13.7 (18.5) 10.8 (10.6)
28.8 (15.9) 17.5 (23.5) 19.6 (19.2)
Total
28.6 (100)
182 (100)
1Data
72.9 (100)
source: Local police office. Based on the years 2003–2005.
and Gamma, the mean total crime rates in the areas with
blocks of flats were five to six times those in the smallscale areas. In setting Beta only a comparatively small difference in crime rates was observed between Beta-flat and
the mixed tenure area, Beta-mix. Theft and robbery were
the most frequently reported crimes in all areas. The rates
of reported violence and other crimes against personal
safety were ten (setting Alpha) to four (setting Beta) times
as high in the areas with blocks of flats than in the smallscale or mixed areas within those two settings. The residents in the small-scale areas and in the mixed area,
Alpha-house, Beta-mix and Gamma-house, had organized
Neighbourhood Watch programmes against crime and
signposts in those areas were visible.
Study population
A cross-sectional study was conducted among the residents living in the six neighbourhoods. A sample was randomly selected from the census records by Statistics
Sweden and residents aged 20–79 years were invited to
participate in a survey. The questionnaire was sent by post
in October 2005. The questionnaire was followed up by a
reminder 4 weeks later and a further questionnaire was
posted to non-respondents. A total of 1390 residents participated, accounting for 56% of the 2476 individuals
who received the questionnaire.
Measurements
As there is no consensus on the term 'safety-related concerns' [8,35], alternative concepts describing the term are
derived from a set of relevant variables.
Safety-related concerns
The items of perceived neighbourhood disorder were
adapted from the Swedish Living Conditions survey [36]
(items a-f) or original for this study (g), but developed by
the WHO [37]. The participants were asked to assess each
question using a five-point Likert scale: to what extent
they perceived (a) damage and/or graffiti, (b) littering, (c)
car theft, (d) disturbance from neighbours. Participants
were asked to respond using a measurement scale from
'Yes, very annoying' to 'None'. The respondents were also
asked to what extent they perceived (e) tobacco smoking
and (f) consumption of alcohol or use of other drugs to be
an annoyance in their housing area. The respondents were
also asked to estimate (g) how often their sleep was disturbed because of street noise or noisy neighbours.
Questions about general safety and fear of crime were asked
to measure perceived safety based on previously used
questions (with minor modifications). The item, 'Do you
feel safe or insecure in your housing area in the daytime?'
and the item 'Do you feel safe or insecure in your housing
area during the evening and at night?' was intended to
capture time and space aspects of perceived safety in the
residential environment. To measure fear of crime the
question, 'Do you refrain from going outdoors because of
fear of crime?' was used as in previous research [11,3739].
Environmental and personal correlates
The respondents were categorized according to geographic
housing area and neighbourhood setting. A question was
asked on whether the residents had personal experience as
a victim of crime during the last 12 months within the
housing area. The reply option was 'yes' or 'no' and the
item was original but modified from Statistics Sweden
[36]. The participants were also asked to rate their perceived area reputation using a Likert scale: from 'very good'
to 'very bad'. This question is original, but developed from
studies in the west of Scotland [40].
Page 4 of 11
(page number not for citation purposes)
BMC Public Health 2009, 9:221
Area-level crime
The area-level crime rates were represented as the mean
number of police-reported crimes per 1000 inhabitants
per year (year 2003–2005) according to the penal codes
presented in Table 2.
Demographic characteristics
Age (in 2005) was grouped in three categories: 'young' as
20–34 years, 'middle-age' as 35–64 years and 'elderly' as
65–79 years of age. Education was grouped by length of
education. The participants were classified into three
groups: 'low educated' as 1–9 years in school, 'middle
educated' as 10–12 years in school and 'high educated' as
more than 12 years in school. Country of birth was categorized as 'born in Sweden' or 'elsewhere'. Household civil status was categorized as 'family household' when living with
a spouse or significant other and/or children or 'single
household' when living alone. Type of housing was categorized as 'small house' or 'flat'.
Statistical method
A factor analysis was performed to derive simplified
dimensions of safety-related concerns from the set of variables measuring several aspects of perceived safety. The
primary variables included in the analyses were the extent
of graffiti, car theft, litter, disturbance by neighbours, disrupted sleep, disturbance by tobacco smoking and alcohol
consumption, sense of safety during the day, evening and
night, and fear of crime. The extraction method was principal component analysis and an orthogonal rotation was
performed with varimax and Kaiser normalization using
the option of replacement of missing values with the
mean in SPSS [41].
For each dimension of perceived safety identified, the factor scores were computed for each respondent. Due to the
skewness of the distribution of the factor scores (p <
0.001) [42], these were dichotomised by the upper quartile, for use as input variables in subsequent statistical
analyses.
A total of 1097 persons with no missing values on risk
item and condition variables were included in the subsequent analyses. Logistic regressions with robust estimates
of standard errors were performed to examine associations between derived concepts of safety-related concerns
and area- and individual-level variables. Clustering effects
within each neighbourhood were taken into account by
calculating robust estimates of standard errors in Stata.
The multivariate adjusted odds ratios with 95% confidence intervals were estimated for total area crime rate
(area-level characteristic), having been a victim of crime,
area reputation, gender, age, education, country of birth,
household civil status and type of housing (small house vs
flat) (individual characteristics). A level of 5% was considered to be statistically significant.
http://www.biomedcentral.com/1471-2458/9/221
Ethical aspects
The study was approved by the Regional Committee for
Research Ethics at Linköping University.
Results
Non-responder characteristics
The response rate varied between areas and was the highest in area Gamma-house (detached houses), 158/231
(68%), and lowest in area Gamma-flat (blocks of flats),
175/391 (45%). External dropout analysis showed no statistically significant differences in gender among the
responders compared to the study population in five of
the six areas, but in area Alpha-house more females and
fewer males responded than expected. The responders
were older than the study population in three areas
(Alpha-house, Beta-mix, Beta-flat). The group of 'middle
educated' (10–12 years) responders was underrepresented
in three areas (Alpha-house, Alpha-flat, Beta-mix) but
there was no statistically significant difference in the distribution of country of birth between the responders and
the study population (Table 3).
Composite dimensions of deficient residential safety
Three composite dimensions of residential safety-related
concerns were identified in the factor analysis: 'structural
indicators of social disorder', 'contact with disorderly
behaviour' and 'existential insecurity' (Table 4). Each
composite dimension was associated with different sets of
modifying items and conditions.
Structural indicators of social disorder
Area-level crime rates were associated with reporting by
the residents of structural indicators of social disorder
(OR 1.010, CI 1.007–1.013) (Table 5). Residents who
estimated their area reputation as less good were more
than twice as likely to report concerns about structural
indicators of social disorder (OR 2.86, CI 2.13–3.84). For
females, the odds ratio of reporting such concerns was
lower (OR 0.78, CI 0.65–0.94). Residents born abroad
reported concerns about structural indicators of social disorder to a lower extent (OR 0.68, CI 0.55–0.84) than residents born in Sweden. Living in a flat was associated with
concerns in this dimension (OR 1.37, CI 1.01–1.84), but
having been a victim of crime, age, and household civil
status were not associated (Table 5).
Contact with disorderly behaviour
The odds ratio for reporting contact with disorderly
behaviour was more than five times higher for residents
living in flats than for those living in detached houses (OR
5.58, CI 3.06–10.17) (Table 5). Personal experience of
crime during the last 12 months increased the likelihood
of being concerned about disorderly behaviour in the residential area to 1.61 (CI 1.03–2.51). Such concerns were
also associated with living in a neighbourhood with estimated less favourable reputation (OR 2.91, CI, 2.12–
Page 5 of 11
(page number not for citation purposes)
BMC Public Health 2009, 9:221
http://www.biomedcentral.com/1471-2458/9/221
Table 3: Demographic characteristics of the survey respondents and study population in the housing areas studied
Setting Alpha
Gender distribution
Respondents
Male
Female
Partial drop-outs: non- respondent gender
Total
Study population
Male
Female
Total
p-value
Age distribution
Respondents
20–34 years
35–64 years
65–79 years
Partial drop-outs: non-respondent age
Total
Study population
20–34 years
35–64 years
65–79 years
Total
p-value
Country of birth
Respondents
Born in Sweden
Born elsewhere
Partial drop-outs: non-respondent country of birth
Total
Study population
Born in Sweden
Born elsewhere
Total
p-value
Years in school
Respondents
1–9 years
10–12 years
≥ 12 years
Partial drop-outs: non-respondent years in school
Total
Study population
1–9 years
10–12 years
≥ 12 years
Total
p-value
No. of questionnaires delivered (N = 2510)
Drop-out on account of deceased or moved (N = 34)
Respondent rate (N = 1390, 56.1%)
Setting Beta
Setting Gamma
Alpha-house
n (%)
Alpha-flat
n (%)
Beta-mix
n (%)
Beta-flat
n (%)
Gamma-house
n (%)
Gamma-flat
n (%)
108 (40.1)
141 (52.4)
20 (7.4)
269
101 (42.1)
125 (52.1)
14 (5.8)
240
139 (44.4)
163 (52.1)
11 (3.5)
313
108 (46)
111 (47.2)
16 (6.8)
235
68 (43.0)
85 (53.8)
5 (3.2)
158
77 (44.0)
82 (46.9)
16 (9.1)
175
514 (50.8)
497 (49.2)
1011
0.035*
646 (48.9)
676 (51.1)
1322
0.246
632 (48.3)
677 (51.7)
1309
0.480
569 (48.2)
611 (51.8)
1180
0.766
280 (48.2)
301 (51.8)
581
0.409
507 (51.6)
476 (48.4)
983
0.461
21 (7.8)
170 (63.2)
74 (27.5)
4 (1.5)
269
56 (23.3)
132 (55)
50 (20.8)
2 (0.8)
240
51 (16.3)
173 (55.3)
84 (26.8)
5 (1.6)
313
59 (25.1)
98 (41.7)
74 (31.5)
4 (1.7)
235
25 (15.8)
87 (55.1)
45 (28.5)
1 (0.6)
158
35 (20)
95 (54.3)
40 (22.9)
5 (2.9)
175
131 (13)
654 (64.7)
226 (22.4)
1011
0.027*
406 (30.7)
695 (52.6)
221 (16.7)
1322
0.051
321 (24.5)
712 (54.4)
276 (21.1)
1309
0.004**
362 (30.7)
529 (44.8)
289 (24.5)
1180
0.044*
97 (16.7)
328 (56.5)
156 (26.9)
581
0.897
280 (28.5)
515 (52.4)
188 (19.1)
983
0.08
253 (94.1)
15 (5.6)
1 (0.4)
269
165 (68.8)
75 (31.3)
0
240
283 (90.4)
30 (9.6)
0
313
186 (79.1)
48 (20.4)
1 (0.4)
235
136 (86.1)
20 (12.7)
2 (1.3)
158
115 (65.7)
56 (32)
4 (2.3)
175
939 (92.9)
72 (7.1)
1011
0.378
870 (65.8)
452 (34.2)
1322
0.375
1188 (90.8)
121 (9.2)
1309
0.852
891 (75.5)
289 (24.5)
1180
0.192
503 (86.6)
78 (13.4)
581
0.843
612 (62.3)
371 (37.7)
983
0.212
82 (30.5)
78 (29)
71 (26.4)
38 (14.1)
269
98 (40.8)
74 (30.8)
26 (10.8)
42 (17.5)
240
88 (28.1)
106 (33.9)
82 (26.2)
37 (11.8)
313
94 (40)
76 (32.3)
31 (13.2)
34 (14.5)
235
55 (34.89)
62 (39.2)
26 (16.5)
15 (9.5)
158
71 (40.6)
63 (36)
19 (10.9)
22 (12.6)
175
266 (26.3)
481 (47.6)
264 (26.1)
1011
0.001**
413
564 (42.7)
632 (47.8)
126 (9.5)
1322
0.017*
518
365 (27.9)
663 (50.6)
281 (21.5)
1309
0.001**
514
505 (42.8)
533 (45.2)
142 (12)
1180
0.116
443
171 (29.4)
312 (53.7)
98 (16.9)
581
0.064
231
438 (44.6)
447 (45.5)
98 (10)
983
0.493
391
269 (65.1)
240 (46.3)
313 (60.9)
235 (53)
158 (68.4)
175 (44.8)
*χ2 test significant at p < 0.05. **χ2 test significant at p < 0.01.
Page 6 of 11
(page number not for citation purposes)
BMC Public Health 2009, 9:221
http://www.biomedcentral.com/1471-2458/9/221
Table 4: Factor loadings for three dimensions of perceived safety (n = 1390)
Derived factora
Percentage of variation explained
Cronbach alpha
Factor loadings
Graffiti
Car theft
Litter
Disturbing neighbours
Disturbed night rest
Tobacco smoking
Alcohol consumption
Sense of safety in evening and night
Fear of crime
Sense of safety in daytime
aFactor
I. Structural indicators of social disorder
II. Contact with disorderly behaviour
III. Existential insecurity
40.2
0.78
12.1
0.78
10.3
0.65
0.83
0.74
0.73
0.83
0.82
0.58
0.54
0.74
0.73
0.73
loadings < 0.5 are not represented.
4.00). The odds ratio for reporting contact with disorderly
behaviour was significantly lower for the elderly (65–79
years) (OR 0.51, CI 0.38–0.70). There were no associations between area-level crime, gender, education, country of birth and contact with disorderly behaviour, but for
single households the odds ratio was 1.57 times higher
(CI 1.28–1.94) than for family households.
Existential insecurity
The odds ratio of reported existential insecurity (feeling
insecure in the residential area during the day or in the
evening or at night, and fear of crime) in the neighbourhood was associated with area-level crime (OR 1.003, CI
1.001–1.006) but not significantly associated with being
a crime victim during the last 12 months (OR 1.42, CI
0.85–2.37). Residents who thought their area reputation
was less good had more than twice the odds ratio (OR
2.67, CI 1.64–4.35) for experiencing existential insecurity
in the neighbourhood. Existential insecurity in the neighbourhood was reported four times more often by female
residents (OR 4.54, CI 3.21–6.43) compared with males,
and being elderly (65–79 years) increased the odds ratio
for perceived existential insecurity to 1.72 (CI 1.20–2.46)
times that of younger residents. Education, country of
birth, household civil status, and type of housing were not
associated with existential insecurity in the neighbourhood (Table 5).
Discussion
In our study design, an attempt was made to cover several
complementary dimensions of perceived safety in the residential environment. The three derived dimensions captured almost two-thirds of the total item variance. As
could be expected, the dimension 'structural indicators of
social disorder' was associated with the area-level of crime
and to living in a flat. Previous research [24] suggested
that a land-use structure with large non-residential space,
such as in the present areas with blocks of flats, increases
crime rates. A corresponding association between factual
crime rates and perceived structural indicators of social
disorder could be observed in the areas with detached
houses and little non-residential space, where the residents assessed the occurrence of threatening physical
signs as less serious; the crime rates were also relatively
low. It is possible that signs of disorder in the residential
environment give rise to safety-related concerns among
residents, especially in combination with local reports of
violence, assaults or rape communicated through neighbours or the media. In areas where reports of crime and
signs of disorder co-exist, a process of stigmatization
could develop giving an area the reputation of being a
dangerous place [43,44]. In the classic theory of 'broken
windows', Wilson and Kelling [45] suggest that signs of
neighbourhood disorder are perceived as a warning sign
of low reciprocity. More recent research has indicated that
physical signs of social disorder indicate neighbourhood
deterioration, trigger general insecurity among residents
and stimulate outward migration [46-49].
We observed a similar pattern of clear differentiation
between residents in different types of housing for the
dimension 'contact with disorderly behaviour'. Residents
living in flats reported more severe problems with disturbance from neighbours, alcohol consumption, and smoking, and their sleep was disturbed more frequently
because of noise in the residential environment. Contact
with this sort of disorderly behaviour could be interpreted
as a lack of the normal inhibitors of incivilities and crime,
which produces insecurity. If such a pattern lasts for a prolonged in time, residents may suffer from the consequences of safety-related concerns in the living
environment as a source of accumulated stress with negaPage 7 of 11
(page number not for citation purposes)
BMC Public Health 2009, 9:221
http://www.biomedcentral.com/1471-2458/9/221
Table 5: Logistic regression analysis with robust estimates of standard errors for three dimensions of perceived safety
Variables
Structural indicators of social disorder
N (%)
Exposed
cases,
n (%)
Area-level characteristics
Total area
crimeb
Individual characteristics
Experience of crime
No (0)
982 (89.5) 225 (22.9)
Yes (1)
115 (10.5)
Area reputation
Good (0)
416 (37.9)
95% CI
Exposed
cases,
n (%)
Adjusted 95% CI
ORa
Exposed
cases,
n (%)
Adjusted 95% CI
ORa
1.01***
1.01–1.01
-
1.00
-
1.00**
222 (22.6)
1
1
49
(42.6)
1.61*
218
(22.2)
35
(30.4)
39
(9.4)
232 (34.1)
1
2.91***
131 (25.8)
1
0.65–0.94
140 (23.7)
0.93
1
1.36
0.96–1.92
76
(35.5)
153 (24.7)
1.00
0.77–1.32
0.51–1.47
42
(15.9)
0.51***
0.38–0.70
94
(22.0)
129 (29.8)
1
1.27
0.79–2.05
48
(20.3)
1.27
0.69–2.34
209 (23.0)
1
62
(33.0)
1.04
168 (20.2)
1
0.61–1.20
103 (38.7)
1.57***
1
1.01–1.84
35
(7.7)
236 (36.6)
1
51
(44.3)
1.61
1
2.86***
Not very
good (1)
Gender
Male
681 (62.1)
507 (46.2)
139 (27.4)
1
Female
590 (53.8)
137 (23.2)
0.78**
Age (years)
20–34
214 (19.5)
1
35–64
619 (56.4)
64
(29.9)
166 (26.8)
65–79
264 (24.1)
46
(17.4)
0.87
Education
1–9 years
427 (38.9)
97
(22.7)
131 (30.3)
1
433 (39.5)
≥ 12 years
237 (21.6)
Country of birth
909 (82.9)
Born in
Sweden
Born
188 (17.1)
elsewhere
Household civil status
Family
831 (75.8)
household
Single
266 (24.2)
household
Type of housing
Small house 453 (41.3)
Flat
644 (58.7)
Existential insecurity
Adjusted
ORa
41
(9.9)
235 (34.5)
10–12 years
Contact with disorderly behaviour
0.82–3.18
2.13–3.84
1.40*
1.01–1.94
48
(20.3)
1.20
0.83–1.74
218 (24.0)
1
58
(30.9)
0.68***
197 (23.7)
1
79
(29.7)
0.85
52
(11.5)
224 (34.8)
1
1.37*
0.55–0.84
5.58***
0.99–1.00
1.03–2.51
2.12–4.00
0.79–1.10
0.74–1.47
1.28–1.94
3.06–10.17
53
(12.7)
200
(29.4)
54
(10.7)
199
(33.7)
55
(25.7)
117
(18.9)
81
(30.7)
115
(26.9)
95
(21.9)
43
(18.1)
204
(22.4)
49
(26.1)
1.42
1.00–1.01
0.85–2.37
1
2.67***
1.64–4.35
1
4.54***
3.21–6.43
1
0.80
0.58–1.09
1.72**
1.20–2.46
1
0.95
0.68–1.31
0.82
0.53–1.27
1
1.02
184
(22.1)
69
(25.9)
1
79
(17.4)
174
(27.0)
1
0.94
0.85
0.69–1.51
0.61–1.47
0.59–1.23
OR, odds ratio; CI, confidence interval. *p < 0.05;**p < 0.01;***p < 0.001.
aORs are adjusted for all other variables in the table.
bTotal mean number of police-reported crimes/1000 inhabitants per year.
Page 8 of 11
(page number not for citation purposes)
BMC Public Health 2009, 9:221
tive impact on well-being and indirectly on health [50].
The two composite dimension; 'structural indicators of
social disorder' and 'contact with disorderly behaviour'
both reflect the construct collective efficacy introduced by
Sampson [51], and defined as the linkage of social control
and cohesion that reproduces the norms for behaviour in
neighbourhood environments. Several studies demonstrate that where the rules of behaviour are unclear, residents will experience mistrust and safety-related concerns.
Poor social and economic resources and residential instability in a neighbourhood have been found to be strongly
associated with violence mediated by low collective efficacy [52-54]. This pattern could be applicable to the
present study; in the areas with comparatively high violence-related crime rates (blocks of flats), the levels of
education and income were lower, and the density of the
residents and the mobility rate were higher than for areas
with low violence rates (small-scale housing areas). The
occurrence of violence in the immediate environment
may generate a situation where it is harder for residents to
build trustworthy bonds with their neighbours [54]. If
social disturbances are reported and communicated regularly from particular areas in the local media, this negative
information will accumulate over time and have a negative impact on the image of a specific area and partly
mould its identity among its residents.
In accordance with previous studies, we found that female
gender was strongly associated with the composite dimension 'existential insecurity' (to feel unsafe during the day
or in the evening or at night and fear of crime). It is possible that this dimension reflects an individual vulnerability, in contrast to the other concepts, which is derived
from determinants in physical space and the built environment to a higher extent [55]. It is also known that persons who are the object of violence or threats are often
closely related to the criminal [18]. Arguments have been
made on the need for local safety promotion programs to
take different perceptions of males and females into
account [35]. To spend time outdoors in the evening and
at night has been found to be correlated with an increased
risk of being robbed or mugged [56]. Theft and robbery
were the most prevalent types of crime in all areas in our
study. Thus, area-level crime was associated with the existential insecurity dimension of safety-related concerns,
but having been a victim of crime and type of housing
were not. These findings could be interpreted to mean
that existential insecurity is a more complex issue than
area of residence, type of housing and victimization. On
the other hand, in this context it is also interesting to
observe that country of birth, level of education and living
alone were not important for perceived 'existential insecurity'.
We found that area-level crime rate and individual-level
variables were associated with the composite dimensions
http://www.biomedcentral.com/1471-2458/9/221
'structural indicators of social disorder' and 'existential
insecurity', but only individual-level variables were associated with the dimension 'contact with disorderly behaviour'.
We investigated whether there was an association between
the residents' self-estimated area reputation and safetyrelated concerns. Our data show that all three derived
safety-related dimensions of safety-related concerns
('structural indicators of social order', 'contact with disorderly behaviour' and 'existential insecurity ') were strongly
influenced by whether the residents perceived themselves
to live in a residential environment with a less favourable
reputation. This study indicates that self-assessed area reputation is an invisible factor that strongly influences the
sense of insecurity among residents. Why area reputation
is a mechanism that has strong consequences for perceived safety among residents is not investigated here but
it is possible that reputation is related to social dimensions in the residents' immediate environment. In this
regard, clues can be found in studies suggesting that the
occurrence of crime and incivilities is associated with the
development of perceptions that one's neighbourhood is
a 'bad place' [57,58]. Another plausible clue is that if the
estimated reputation is an essential part of the lives of residents, mental images of the area could either strengthen
the attachment to the neighbourhood if you live in a
'good place' or weaken the attachment if you live in a 'bad
place'. The mental images the residents have of their
neighbourhood are likely to have serious consequences
on the community spirit of their neighbourhood, leading
to either a positive or negative situation in which residents
increase or reduce their willingness to come together. A
less favourable reputation develops a process of stigmatization as a 'bad area', leading to an ongoing migration
process in which residents who are able, move away from
the area. Consequently the people with lesser ability and
resources remain. In contrast, the reputation of a desirable
'good area' is attractive and implies a safe and secure place
in which to live.
Living alone was found to be positively associated with
'contact with disorderly behaviour' when compared with
residents living in families. This may be in line with earlier
studies that suggested that living alone could lead to social
isolation and thereby a more vulnerable situation [20],
which likely also generates anxiety.
The three small-scale areas in this study had implemented
Neighbourhood Watch programs. The lowest crime rates
were recorded from these areas and the residents in
detached houses reported the highest perceived safety.
The rationale behind Neighbourhood Watch programs is
to contribute to lower crime rates and to a perception of
safety in a neighbourhood through increased reciprocity,
e.g. looking after strangers and property and improved
Page 9 of 11
(page number not for citation purposes)
BMC Public Health 2009, 9:221
http://www.biomedcentral.com/1471-2458/9/221
informal social control. However, a pre-existing high
social capital in the neighbourhood seems to be a prerequisite for implementation of a successful Neighbourhood
Watch program [18,20,59]. In Sweden, the majority of
Neighbourhood Watch programs are implemented in
wealthy small-scale areas [60]. Therefore, if Neighbourhood Watch programs are to become an effective method
of building safer neighbourhoods, our results suggest that
the design of these programs need to be adjusted to make
them suitable for introduction in areas with blocks of
flats.
Our findings suggest that the implementation of modern
Swedish housing policy targeting social integration has
resulted in inequalities in safety-related concerns among
residents. Future research on whether Neighbourhood
Watch programs are an effective way to promote safer
neighbourhoods is warranted, as well as the development
of methods to implement these programs in areas with
blocks of flats.
Study limitations
There are important limitations that have to be considered
when trying to draw conclusions from this study. A methodological limitation is the cross-sectional design, which
does not allow causal inferences. Given this study design,
and the fact that the variables in the field of study are
strongly interrelated, it is not possible to determine if a
negative area reputation is a cause or a consequence of
low perceived safety. Area reputation appears to be part of
a complex pathway through which the effects of crime on
perceived safety are mediated. However, our results are
only indicative on this matter.
Authors' contributions
A number of variables that could contribute to perceived
safety were not included in this study, e.g. health indicators, individual and contextual social capital, length of
residence, and objective neighbourhood characteristics
(concentrated affluence, unemployment rate, outdoor
conditions such as lighting in public areas, etc.)
[33,61,62].
Competing interests
The authors declare that they have no competing interests.
AK, KL, TT planned the study. AK was the main author of
this paper and was involved in all steps to create the manuscript. NK designed the statistical analysis. All authors
read drafts of the manuscript, made comments and suggestions to improve the text before they agreed to submit
it for publication.
Acknowledgements
This study was funded by the Swedish National Rescue Service Agency and
The Building Foundation Platen.
References
1.
2.
3.
4.
5.
In previous housing area studies based on postal questionnaires in Scotland [63] and Australia [64], the participation rates were around 50%, similar to the 56%
achieved in this study. The pattern of non-response was
slightly skewed in the present study according to gender,
age and education for some areas. Some findings may
therefore have been somewhat overestimated, due to the
fact that the participants included a higher proportion of
women (Alpha-house) and older persons (Alpha- house,
Beta-mix, Beta-flat) compared with the total area population.
6.
7.
8.
9.
10.
11.
12.
Conclusion
We have identified environmental, socio-demographic,
and personal correlates of safety-related concerns in contiguous neighbourhoods in a Swedish community. The
results indicate that self-assessed area reputation was
important for all dimensions of safety-related concerns
among the residents. We also found a difference in crime
rates and safety-related concerns between areas with
blocks of flats compared with small-scale areas, although
these neighbourhoods were close geographically with a
shared local centre.
13.
14.
15.
16.
17.
Nightingale F: Florence Nightingale on Society and Politics,
Philosophy, Science, Education and Literature [1820–1910].
In Electronic reproduction Volume 5. Edited by: McDonald L. Palo Alto,
CA; 2003.
Rosen G: A History of Public Health [first published 1958].
New York: John Hopkins; 1993.
Shaw C, McKay H: Juvenile Delinquency and Urban Areas Chicago: University of Chicago Press; 1942.
Kawachi I, Kennedy BP, Wilkinson RG: Crime: social disorganization and relative deprivation. Soc Sci Med 1999, 48(6):719-731.
Fuller-Thomson E, Hulchanski JD, Hwang S: The housing/health
relationship: what do we know? Rev Environ Health 2000, 15(1–
2):109-133.
Jackson RJ: The impact of the built environment on health: an
emerging field. Am J Public Health 2003, 93(9):1382-1384.
van Kamp I, van Loon J, Droomers M, de Hollander A: Residential
environment and health: a review of methodological and
conceptual issues. Rev Environ Health 2004, 19(3–4):381-401.
Bonnefoy X: Inadequate housing and health: an overview. Int J
Environ Pollut 2007, 30(3–4):411-429.
Austin DM, Furr LA, Spine M: The effects of neighborhood conditions on perceptions of safety. J Crim Just 2002, 30(5):417-427.
Baumer TL: Testing a general model of fear of crime: data
from a national sample. J Res Crime Delinq 1985, 22(3):239-255.
Green G, Gilbertson JM, Grimsley MF: Fear of crime and health
in residential tower blocks. A case study in Liverpool, UK.
Eur J Public Health 2002, 12(1):10-15.
Skogan WG, Maxfield MG: Coping with Crime. Individual and Neighborhood Reactions London: Sage Publications; 1981.
Whitley R, Prince M: Fear of crime, mobility and mental health
in inner-city London, UK. Soc Sci Med 2005, 61(8):1678-1688.
Bennett GG, McNeill LH, Wolin KY, Duncan DT, Puleo E, Emmons
KM: Safe to walk? Neighborhood safety and physical activity
among public housing residents.
PLoS Med 2007,
4(10):1599-1606. discussion 1607
Heath L: Impact of newspaper crime reports on fear of crime:
multimethodological investigation. J Pers Soc Psychol 1984,
47(2):263-276.
Baum F, Palmer C: 'Opportunity structures': urban landscape,
social capital and health promotion in Australia. Health Promot
Int 2002, 17(4):351-361.
Evans DJ, Fletcher M: Fear of crime: testing alternative hypotheses. Appl Geog 2000, 20:395-411.
Page 10 of 11
(page number not for citation purposes)
BMC Public Health 2009, 9:221
18.
19.
20.
21.
22.
23.
24.
25.
26.
27.
28.
29.
30.
31.
32.
33.
34.
35.
36.
37.
38.
39.
40.
41.
42.
Pain R: Place, social relations and the fear of crime: a review.
Prog Hum Geog 2000, 24(3):365-387.
Acierno R, Rheingold AA, Resnick HS, Kilpatrick DG: Predictors of
fear of crime in older adults.
J Anxiety Disord 2004,
18(3):385-396.
Halpern D: Social Capital Cambridge: Polity Press; 2005.
Graham H: The challenge of health inequalities. In Understanding Health Inequalities Edited by: Graham H. Berkshire: Open University Press; 2000:3-21.
Matei S, Ball-Rokeach SJ, Qiu JL: Fear and misperception of Los
Angeles urban space – a spatial-statistical study of communication-shaped mental maps. Commun Res 2001, 28(4):429-463.
Newman O: Defensible Space. Crime Prevention Through Urban Design
New York: Collier Books; 1973.
Greenberg SW, William RM: Neighborhood design and crime. A
test of two perspectives. J Am Planning Assoc 1984, 50(1):48-61.
Wandersman A, Nation M: Urban neighborhoods and mental
health – psychological contributions to understanding toxicity, resilience, and interventions.
Am Psychol 1998,
53(6):647-656.
Wood L, Shannon T, Bulsara M, Pikora T, McCormack G, Giles-Corti
B: The anatomy of the safe and social suburb: an exploratory
study of the built environment, social capital and residents'
perceptions of safety. Health Place 2008, 14(1):15-31.
Lindstrom M, Merlo J, Ostergren PO: Social capital and sense of
insecurity in the neighbourhood: a population-based multilevel analysis in Malmo, Sweden.
Soc Sci Med 2003,
56(5):1111-1120.
Chandola T: The fear of crime and area differences in health.
Health Place 2001, 7(2):105-116.
Stafford M, Chandola T, Marmot M: Association between fear of
crime and mental health and physical functioning. Am J Public
Health 2007, 97(11):2076-2081.
Shaw M: Housing and public health. Annu Rev Public Health 2004,
25:397-418.
Macintyre S, Ellaway A, Cummins S: Place effects on health: how
can we conceptualise, operationalise and measure them? Soc
Sci Med 2002, 55(1):125-139.
McLaren L, Perry R, Carruthers L, Hawe P: Introducing a means
of quantifying community reputation: the print media as a
data source. Health Place 2005, 11(2):187-194.
Ellaway A, Macintyre S: Does housing tenure predict health in
the UK because it exposes people to different levels of housing related hazards in the home or its surroundings? Health
Place 1998, 4(2):141-150.
SCB: Statistiska centralbyrån: NYKO-nyckelkodsystemet. Geografisk indelning för statistikredovisning för delområden inom en kommun [Statistics
Sweden. Key-code System. Geographical Classification for Statistical
Account] Stockholm: SCB; 2008.
Whitzman C: Community safety indicators: are we measuring
what counts? Urban Policy Res 2008, 26(2):197-211.
SCB: Statistiska centralbyrån: Undersökningen om levnadsförhållanden,
ULF [Statistics Sweden. Swedish Living Conditions Survey] Stockholm:
SCB; 1999.
Bonnefoy XR, Annesi-Maesano I, Aznar LM, Braubach M, Croxford B,
Davidson M, Ezratty V, Fredouille J, Gonzalez-Gross M, van Kamp I,
Maschke C, Mesbah M, Moissonnier B, Monolbaev K, Moore R, Nicol
S, Niemann H, Nygren C, Ormandy D, Röbbel N, Rudnai P: Review
of evidence on housing and health. Background document.
Fourth Ministerial Conference on Environment and Health: 23–25 June
2004, Budapest, Hungary 2004 [http://www.euro.who.int/document/
hoh/ebackdoc01.pdf]. Geneva: WHO
FHI: The Swedish National Institute of Public Health. Statens
folkhälsoinstitut. Nationella folkhälsoenkäten [National questionnaire on
public health] 2004 [http://www.fhi.se].
Ferraro KF, LaGrange R: The measurement of fear of crime.
Sociol Inq 1987, 57:70-101.
Macintyre S, Maciver S, Sooman A: Area, class and health –
should we be focusing on places or people. J Soc Policy 1993,
22:213-234.
Hair JF, Black B, Babin B, Anderson RE, Tatham RL: Multivariate Data
Analysis Upper Saddler River, NJ: Pearson Prentice Hall; 2006.
D'Agostino RB, Balanger A, D'Agostino RB Jr: A suggestion for
using powerful and informative tests of normality. Am Statist
1990, 44(4):316-321.
http://www.biomedcentral.com/1471-2458/9/221
43.
44.
45.
46.
47.
48.
49.
50.
51.
52.
53.
54.
55.
56.
57.
58.
59.
60.
61.
62.
63.
64.
Goffman E: Stigma. Notes on the Management of Spoiled Identity New
Jersey: Penguin Books; 1963.
Gregory R, Slovic P, Flynn J: Risk perceptions, stigma, and health
policy. Health Place 1996, 2(4):213-220.
Wilson JQ, Kelling GL: Broken windows. Atlantic Monthly 1982,
249(3):29-38.
Perkins DD, Meeks JW, Taylor RB: The physical environment of
street blocks and resident perceptions of crime and disorder: implications for theory and measurement. J Environ Psychol 1992, 12:21-34.
Sampson RJ, Raudenbush SW, Earls F: Neighborhoods and violent
crime: a multilevel study of collective efficacy. Science 1997,
277(5328):918-924.
Sampson RJ, Raudenbush SW: Systematic social observation of
public spaces: a new look at disorder in urban neighborhoods. Am J Sociol 1999, 105(3):603-651.
Kruger DJ, Reischl TM, Gee GC: Neighborhood social conditions
mediate the association between physical deterioration and
mental health. Am J Community Psychol 2007, 40(3–4):261-271.
Steptoe A, Feldman PJ: Neighborhood problems as sources of
chronic stress: development of a measure of neighborhood
problems, and associations with socioeconomic status and
health. Ann Behav Med 2001, 23(3):177-185.
Sampson RJ: Neighborhood-level context and health: lessons
from sociology. In Neighborhood and Health Edited by: Kawachi I,
Berkman LF. New York: Oxford University Press; 2003:132-146.
Morenoff JD, Sampson RJ, Raudenbush SW: Neighborhood inequality, collective efficacy, and the spatial dynamics of urban
violence. Criminology 2001, 39(3):517-559.
Guest AM, Cover JK, Matsueda RL, Kubrin CE: Neighborhood context and neighboring ties. City Community 2006, 5(4):363-385.
Kruger DJ, Hutchison P, Monroe MG, Reischl T, Morrel-Samuels S:
Assault injury rates, social capital, and fear of neighborhood
crime. J Community Psychol 2007, 35(4):483-498.
Koskela H, Pain R: Revisiting fear and place: women's fear of
attack and the built environment.
Geoforum 2000,
31(2):269-280.
LaGrange RL, Ferraro KF, Supancic M: Perceived risk and fear of
crime – role of social and physical incivilities. J Res Crime Delinq
1992, 29(3):311-334.
Ellen IG, Turner MA: Does neighborhood matter? Assessing
recent evidence. Hous Policy Debate 1997, 8(4):833-866.
Haney TJ: "Broken windows" and self-esteem: Subjective
understandings of neighborhood poverty and disorder. Soc
Sci Res 2007, 36(3):968-994.
Cozens PM: Sustainable urban development and crime prevention through environmental design for the British city.
Towards an effective urban environmentalism for the 21st
century. Cities 2002, 19(2):129-137.
Bennett TH, Holloway KR, Farrington DP: Effectiveness of Neighbourhood Watch in Reducing Crime. Report Prepared for BRÅ Stockholm:
Swedish National Council for Crime Prevention; 2008:9.
Massey DS: The age of extremes: Concentrated affluence and
poverty in the twenty-first century.
Demography 1996,
33(4):395-412.
Guite HF, Clark C, Ackrill G: The impact of the physical and
urban environment on mental well-being. Public Health 2006,
120(12):1117-1126.
Ellaway A, McKay L, Macintyre S, Kearns A, Hiscock R: Are social
comparisons of homes and cars related to psychosocial
health? Int J Epidemiol 2004, 33(5):1065-1071.
Ziersch AM: Health implications of access to social capital:
findings from an Australian study.
Soc Sci Med 2005,
61(10):2119-2131.
Pre-publication history
The pre-publication history for this paper can be accessed
here:
http://www.biomedcentral.com/1471-2458/9/221/pre
pub
Page 11 of 11
(page number not for citation purposes)