- Email: [email protected]
Social capital and violence in poor urban areas of Honduras
Aggression and Violent Behavior 19 (2014) 643–648
Contents lists available at ScienceDirect
Aggression and Violent Behavior
Social capital and violence in poor urban areas of Honduras☆ Nete Sloth Hansen-Nord a,⁎, Mette Skar a, Finn Kjaerulf a, Juan Almendarez b, Sergio Bähr b, Óscar Sosa b, Julio Castro b, Anne-Marie Nybo Andersen c, Jens Modvig a a b c
DIGNITY — Danish Institute Against Torture, Copenhagen, Denmark CPTRT — The Centre for the Prevention and Rehabilitation of Victims of Torture and Their Families, Tegucigalpa, Honduras The University of Copenhagen, Department of Public Health, Copenhagen, Denmark
a r t i c l e
i n f o
Article history: Received 21 August 2014 Received in revised form 27 September 2014 Accepted 27 September 2014 Available online 5 October 2014 Keywords: Social capital Structural Cognitive Violence Victimization Insecurity Trust
a b s t r a c t Honduras has the highest murder rate in the world: the high level of violence threatens the economic and social development of the country as it erodes human and social capital and limits trust among people in poor urban areas. However, neither a detailed consideration of the complex manner in which distinct dimensions of social capital interrelate with violence, nor the potential for double causality has received much attention. Objectives: The study examines the influence of structural social capital (social organization characteristics) and cognitive social capital (social trust and cohesion characteristics) on risk of violence in poor urban areas of Honduras. Methods: The study was carried out in two urban communities of Tegucigalpa experiencing high levels of violence and insecurity. For the quantitative analysis, 1000 individuals older than 18 answered a structured questionnaire. Violence exposure was evaluated based on respondents' self-reporting. Social capital was defined based on the use of the short version of the Adapted Social Capital Assessment Tool. Results: Our results support previous evidence from Guatemala showing that cognitive and structural social capital were inversely related to risk of violence: people with high cognitive social capital had a lower risk of violence (OR 0.46 CI 95: 0.28–0.76) compared to people with low cognitive social capital, whereas people with high structural social capital had a higher risk of violence (OR 1.68 CI 95: 1.04–2.71) compared to people with low structural social capital. Conclusions: Social trust and social activism exhibit significant associations with risk of violence, however, these dimensions are consequences as well as causes of violence. Implications for practice: In an intervention perspective it is important to recognize the difference between social organization and cooperative action for creating change, as these concepts represent very dissimilar levels of collective action toward violence. It is thus important to link the items of social capital, primarily within the structural dimension, to the specific objectives of a given intervention. © 2014 Elsevier Ltd. All rights reserved.
Contents 1.
2.
3. 4.
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . 1.1. Preventing violence through building social capital . . . . . . 1.2. Social capital and violence: a view at the literature . . . . . . Materials and methods . . . . . . . . . . . . . . . . . . . . . . 2.1. Data collection . . . . . . . . . . . . . . . . . . . . . . 2.2. Description of variables . . . . . . . . . . . . . . . . . . 2.3. Statistical analyses . . . . . . . . . . . . . . . . . . . . 2.4. Ethical considerations . . . . . . . . . . . . . . . . . . . Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.1. Cognitive social capital, insecurity and reduced risk of violence 4.2. Structural social capital and increased risk of violence . . . . .
. . . . . . . . . . . .
. . . . . . . . . . . .
. . . . . . . . . . . .
. . . . . . . . . . . .
. . . . . . . . . . . .
. . . . . . . . . . . .
. . . . . . . . . . . .
. . . . . . . . . . . .
. . . . . . . . . . . .
. . . . . . . . . . . .
. . . . . . . . . . . .
. . . . . . . . . . . .
. . . . . . . . . . . .
. . . . . . . . . . . .
. . . . . . . . . . . .
☆ The authors have no conflicts of interest to disclose. ⁎ Corresponding author at: DIGNITY — Danish Institute Against Torture, Bryggervangen 55, DK-2100 Copenhagen, Denmark. E-mail address: [email protected] (N.S. Hansen-Nord).
http://dx.doi.org/10.1016/j.avb.2014.09.013 1359-1789/© 2014 Elsevier Ltd. All rights reserved.
. . . . . . . . . . . .
. . . . . . . . . . . .
. . . . . . . . . . . .
. . . . . . . . . . . .
. . . . . . . . . . . .
. . . . . . . . . . . .
. . . . . . . . . . . .
. . . . . . . . . . . .
. . . . . . . . . . . .
. . . . . . . . . . . .
. . . . . . . . . . . .
. . . . . . . . . . . .
. . . . . . . . . . . .
. . . . . . . . . . . .
. . . . . . . . . . . .
. . . . . . . . . . . .
. . . . . . . . . . . .
. . . . . . . . . . . .
. . . . . . . . . . . .
. . . . . . . . . . . .
644 644 644 644 644 645 645 645 645 646 647 647
644
N.S. Hansen-Nord et al. / Aggression and Violent Behavior 19 (2014) 643–648
4.3. Social organization as a facilitator of violence . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.4. Social organization versus cooperative action: what exactly are we measuring with structural social capital? 4.5. Methodological strengths and limitations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5. Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Acknowledgment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
1. Introduction Crime and violence are key development issues for Latin America with the Central American countries; Honduras, Guatemala, and El Salvador representing the most violent countries in the region (Bank, 2010). Violence is a result of various societal, community and individual factors which interact in a complex manner. Its causes can be assessed within an ecological framework of four levels (Krug, Mercy, Dahlberg, & Zwi, 2002b); at the individual level (first level) personal behavior, history of substance abuse and/or psychological disorders affect the risk of violence, which, in combination with relationship parameters (second level) like family bonds, violent friends and/or low socioeconomic household status can increase the risk of becoming either a victim of violence or a perpetrator (Krug et al., 2002b). Likewise, community factors (level 3), such as income inequality, social acceptability of violence and social capital significantly affect the overall risk of violence (Krug et al., 2002b). Societal factors (level 4) such as ineffectiveness of the justice system and the lack of control of firearms are crucial root causes of violence. The majority of serious crimes in the region are never solved: in Guatemala the impunity reaches around 98% as a result of poor rule of law and political polarization (Matute Rodríguez & Santiago, 2007). High levels of impunity entail low levels of trust in the police and other authorities, and general fear and insecurity in the Latin American populations. Honduras is the world's most violent non-war country with a homicide rate of 85.5 per 100,000 inhabitants representing 7172 homicides in 2012; or an average of 20 murders per day. The murders are primarily committed with firearms (83.4%) and with a high proportion of contract killings (23.5%) (UNAH — IUDPAS, 2013). These threats have increased substantially over the past several years and the government of Honduras lacks sufficient resources to address these issues. The high level of violence threatens the economic and social development of the country in a context of poverty, high level of unemployment and local illicit drug trade, economic and gender inequality, and high firearm availability in combination with cultural norms that support violence (Krug, Dahlberg, Mercy, Zwi, & Lozano, 2002a). Importantly, violence is recognized as a serious public health problem due to violence-related physical and psychological morbidities and disabilities (Matzopoulos, Bowman, Butchart, & Mercy, 2008; Yacoub, Arellano, & Padgett-Moncada, 2006). 1.1. Preventing violence through building social capital Violence prevention efforts in Latin America emphasize the need for launching policies aiming to build social capital due to their ability to provide social control and to engage citizens in partnering with the state to hold the institutions, in particular the police, accountable (Brune & Bossert, 2009; Cuesta & Alda, 2012; Violence in Colombia, 2000). Such policies call for cooperative action from civil society, which requires restoration of social cohesion in local communities and trust between individuals and may yield trust between strangers, providing the basis for peace and development (Bank, 2010; Brune & Bossert, 2009; Cuesta & Alda, 2012). However, if populations do not experience support from both the state and the private sector, neither social stability nor widespread popular support, social capital weakens. Due to the complexity of the problem of violence, prevention programs need a high degree of inter-sectoral involvement where criminal justice
. . . . . .
. . . . . .
. . . . . .
. . . . . .
. . . . . .
. . . . . .
. . . . . .
. . . . . .
. . . . . .
. . . . . .
. . . . . .
. . . . . .
. . . . . .
. . . . . .
. . . . . .
647 647 647 648 648 648
reforms must be linked to broader reforms that address different sectors including education, health, social protection, and labor markets addressing risk factors at different ecological societal levels (Bank, 2010; Kjaerulf & Barahona, 2010; Moser & McIlwaine, 2006) and thus require long continuous efforts to produce positive results. 1.2. Social capital and violence: a view at the literature Previous research has established that violence is associated with a reduced level of social capital (Dinesen et al., 2013; Hernandez & Grineski, 2012; Putnam, 2001; Sabatini, 2009) and fears related to violence impede social organizing and civic participation (Abom, 2004). Correspondingly, social connection including opportunities for active participation of community members and organizations (both formal and informal) in social and economic life has been shown to be an important protective factor against violent behavior (Bank, 2010; Weiss, 2011). Key elements of social capital represent networks and reciprocated exchange, solidarity, trust, and social control (Portes, 1998). However, operationalizing this complex concept is difficult, and satisfactory measures of social capital are generally hard to find. In this research, we focus on two dominant dimensions in social capital; a “structural” dimension of social capital, consisting of network connections facilitating mutually beneficial collective action through established roles, and a “cognitive” dimension, consisting of attitudes toward trust, shared norms, values, attitudes, and beliefs (Dasgupta & Serageldin, 2000). Such dimensions of social capital, emphasizing the distinction between trust and civic engagement have previously been analyzed in relation to crime rates in the United States with different modeling of the dimensions. These studies include a study by Kennedy, Kawachi, ProthrowStith, Lochner, and Gupta (1998), which found that both dimensions are significantly associated with firearm violence, and two studies showing that high levels of trust are associated with lower homicide rates (Galea, Karpati, & Kennedy, 2002; Rosenfeld, Baumer, & Messner, 2001). The primary results of these studies report that the trust dimension of social capital seems to be associated with reduced homicide rates, whereas less and somewhat irregular support is found for the effects of other indicators of social capital; i.e. civil engagement on victimization (Lederman, Loayza, & Menéndez, 2002; Rosenfeld et al., 2001). This is particularly found in a recent study from Guatemala which emphasizes the importance of separating structural and cognitive social capital in relation to violence (Dinesen et al., 2013). Social trust and social activism do thus exhibit significant associations with homicide rates, however, these dimensions are consequences as well as causes of homicide (Messner, Baumer, & Rosenfeld, 2004). In this study we aim to investigate the characteristics of the association between the structural and cognitive dimensions of social capital and violence in one of the world's most violent settings. 2. Materials and methods 2.1. Data collection This study is based on a survey study undertaken by The Centre for Prevention and Rehabilitation of Victims of Torture and Their Families (CPTRT) in Honduras in 2011. The survey was carried out in two urban areas in Tegucigalpa, Nueva Suyapa (NS) and Villa Nueva (VN). NS and VN represent similar contexts in relation to poverty, levels of
N.S. Hansen-Nord et al. / Aggression and Violent Behavior 19 (2014) 643–648
insecurity and access to basic services. Data collection constituted a random sample of 500 households in NS and 500 households in VN, selected by the use of household mappings provided by the Honduras National Institute of Statistics, representing information on about 5500 individual family members. Data was collected using an intervieweradministered questionnaire facilitated by previously trained interviewers who were young citizens of NS and VN. The survey comprised demographic information, household characteristics, health indicators, social capital, victimization, security and perception of the justice system at an individual level. The first adult (N18 years) member of the household who opened the door was selected as the primary study participant. Interviews were carried out during weekends in order to increase the chance for all household members being home, and to attempt to improve the gender balance in the data. 2.2. Description of variables Social capital was measured based on questions from the short version of the Adapted Social Capital Assessment Tool (SASCAT), which is validated by psychometric techniques in addition to cognitive validation (De Silva et al., 2006). The questions clearly distinguished between structural and cognitive social capital. The cognitive elements of social capital focused on the respondents' values, norms, attitudes and levels of interpersonal trust and perception of the support among neighbors, friends, family and other community members. The structural elements focused on organization through participation in formal or informal organizations or community groups, and active participation in community activities (De Silva et al., 2006). Computing the responses on each question related to either cognitive or structural social capital, we constructed two scores on the degree (low vs. high) of cognitive and structural social capital per individual. As for violence, respondents were asked whether or not they had been victimized within the last 12 months from the time of the interview (dichotomous response). The type of violence included violence in public spaces; in the street, during public transport, and in private vehicles. Exposure to domestic violence was excluded from all analyses. Insecurity was assessed by asking whether the respondent felt secure, relatively secure, insecure or very insecure when leaving one's own home in the community. Occupation of the respondent was categorized into housewife, student, employed, self-employed, paid by the hour, and no job. Household income was assessed as a categorical variable with the following weekly income categories recalculated into the following categories: extreme poverty, poverty, middle income, and high income.
2.4. Ethical considerations Before participating in the study, the respondents were introduced to its aim and promised anonymity. Informed consent was given orally. Necessary precautions were taken to ensure that participation in the study did not expose the respondents to further danger. 3. Results The core analyses of the article examined the effect of structural social capital (social organization characteristics) and cognitive social capital (social trust and cohesion characteristics) on risk of violence in two poor urban areas of Honduras. The mean age of the respondents was 38 years (SD 16.6) and 75.9% were women. Within a time window of 12 months, the prevalence of violence at household level was 21.9%, whereas 10.4% of the respondents had been directly exposed to violence in public spaces. Table 1 shows violence exposure distributed by the explanatory variables representing a set of demographic, socioeconomic, and opinion-based categorical variables. There was an age gradient in violence exposure with the highest violence exposure among the youngest age group (13–23 years). Moreover, victimization was significantly associated with a feeling of insecurity in the community and people who had been victims of violence were generally more insecure in their community. The distribution of high versus low social capital on the explanatory variables is presented in Table 2. Age was significantly associated with cognitive social capital with a negative gradient; younger age correlated to lower cognitive social capital; however this gradient was not found for structural social capital. Out of those who had been exposed to violence about 70% had low cognitive social capital compared to only 30% with high cognitive social capital. For the remaining variables no Table 1 Description of the study population.
Individual level variables Sex
Age
2.3. Statistical analyses Bivariate distributions were tested for statistical significance using the χ2-test for nominal and the γ-test for ordinal variables. Crude analyses between violence and explanatory variables were first performed to determine potential confounders. Multivariate logistic regression analyses were used to estimate odds ratios (OR) for violence exposure given the potential confounders. A backwards model search strategy was applied where the start model included all potential confounders. Insignificant estimates were then removed from the model step by step until all remaining estimates in the final model were statistically significant. Three models were constructed addressing the hypothesis in different ways; model I investigated the association between violence and structural social capital given a set of explanatory variables excluding cognitive social capital; model II investigated the association between violence and cognitive social capital given a set of explanatory variables excluding structural social capital, and model III (the final model) investigated the association between violence and structural social capital controlling for the effect of cognitive social capital and vice versa. All analyses were conducted with a two-sided level of significance of 0.05, and calculated using SPSS 19.0.
645
District
Household income
Occupation
Insecurity
Male Female Missing 13–23 years 24–34 years 35–49 years 50+ years Missing Nueva Suyapa Villa Nueva Missing Extreme poverty Poverty Middle income High income Missing Housewife Student Employed Self employed Paid by the hour No job Missing Secure Relatively secure Insecure Very insecure Missing
Total
Public space violence
N = 1000
N = 94
n
n (%)
239 752 9 246 243 243 260 8 522 477 1 198 237 222 254 89 513 106 137 124 42 41 37 384 103 451 50 12
26 (10.9) 66 (8.8) 36 (14.6) 25 (10.2) 18 (7.4) 15 (5.8) 55 (10.5) 39 (8.2) 15 (7.6) 15 (6.3) 26 (11.7) 30 (11.8) 40 (7.8) 25 (23.6) 11 (8.0) 11 (8.9) 5 (11.9) 2 (4.9) 27 (7.0) 13 (12.0) 46 (10.2) 8 (16.0)
The distribution of nominal variables (sex, district and occupation) was tested using a χ2test. The distribution of ordinal variables (age, income and insecurity) was tested using a γ-test. Statistically significant distributions are depicted in bold (p b 0.05).
646
N.S. Hansen-Nord et al. / Aggression and Violent Behavior 19 (2014) 643–648
Table 2 Distribution of social capital on explanatory variables. Structural social capital
Sex Age
District Insecurity
Household income
Occupation
Public space violence
Male Female 13–23 years 24–34 years 35–49 years 50+ years Villa Nueva Nueva Suyapa Secure Relatively secure Insecure Very insecure Extreme poverty Poverty Middle income High income Housewife Student Employed Self employed Paid by the hour No job Yes No
Cognitive social capital
Low n (%)
High n (%)
Low n (%)
High n (%)
113 (49.6) 383 (53.0) 117 (49.0) 122 (52.1) 122 (53.5) 140 (55.6) 241 (52.3) 261 (52.4) 196 (53.1) 43 (43.9) 233 (53.4) 23 (50.0) 109 (57.1) 111 (49.6) 116 (54.0) 116 (47.4) 256 (51.7) 50 (48.1) 66 (51.2) 64 (54.7) 19 (46.3) 22 (55.0) 37 (41.1) 399 (51.8)
115 (50.4) 340 (47.0) 122 (51.0) 112 (47.9) 106 (46.5) 112 (44.4) 220 (47.7) 237 (47.6) 173 (46.9) 55 (56.1) 203 (46.6) 23 (50.0) 82 (42.9) 113 (50.4) 99 (46.0) 129 (52.7) 239 (48.3) 54 (51.9) 63 (48.8) 53 (45.3) 22 (53.7) 18 (45.0) 53 (58.9) 372 (48.2)
105 (48.6) 350 (51.8) 139 (60.2) 121 (55.3) 107 (49.5) 88 (38.6) 211 (50.0) 247 (51.7) 159 (46.5) 48 (51.6) 221 (54.2) 23 (50.0) 91 (52.0) 112 (50.2) 107 (54.0) 110 (48.7) 235 (50.9) 62 (62.6) 55 (44.4) 59 (52.2) 17 (53.1) 14 (36.8) 61 (70.1) 359 (49.9)
111 (51.4) 326 (48.2) 92 (39.8) 98 (44.7) 109 (50.5) 140 (61.4) 211 (50.0) 231 (48.3) 183 (53.5) 45 (48.4) 187 (45.8) 23 (50.0) 84 (48.0) 111 (49.8) 91 (46.0) 116 (51.3) 227 (49.1) 37 (37.4) 69 (55.6) 54 (47.8) 15 (46.9) 24 (63.2) 26 (29.9) 360 (50.1)
The distribution of nominal variables (sex, district, occupation and violence) was tested using a χ2-test. The distribution of ordinal variables (age, income and insecurity) was tested using a γ-test. Statistically significant distributions are depicted in bold (p b 0.05).
statistically significant differences were seen for the people with high or low social capital. However, there was a tendency that the majority of people feeling secure in their community had a high level of cognitive social capital, and likewise, the majority of people feeling insecure had a low level of cognitive social capital. Students generally had a low level of cognitive social capital and people with no jobs reported higher cognitive social capital, however insignificant. The same tendencies were not seen for structural social capital. Table 3 shows the results of the logistic regression analyses of social capital and public space violence. The crude analyses show that people with high cognitive social capital had less than half the risk of violence compared to people with low cognitive social capital (OR 0.43 CI 95: 0.26–0.69). Students had more than a threefold risk (OR 3.71 CI 95: 2.12–6.48) of being exposed to violence compared to housewives, and people of 50 or more years of age had a 63% significantly lower risk (OR 0.37 CI 95: 0.20–0.70) of being exposed to violence compared to the younger generation (13–23 years). The age gradient in violence Table 3 Logistical regression analysis of social capital and experienced public space violence at individual level.
Structural social capital Cognitive social capital Occupation
Age
Low High Low High Housewife Student Employed Self employed Paid by the hour No job 13–23 years 24–34 years 35–49 years 50+ years
Crude OR OR (95% CI)
Final model OR (95% CI)
1 1.54 (0.99–2.39) 1 0.43 (0.26–0.69) 1 3.71 (2.12–6.48) 1.22 (0.61–2.47) 1.19 (0.59–2.39) 1.53 (0.57–4.11) 0.58 (0.13–2.48) 1 0.73 (0.42–1.26) 0.50 (0.27–0.91) 0.37 (0.20–0.70)
1 1.68 (1.04–2.71) 1 0.46 (0.28–0.76) 1 3.77 (2.06–6.87) 1.55 (0.75–3.21) 1.39 (0.66–2.96) 1.72 (0.56–5.30) 0.78 (0.18–3.44)
The overall Wald-test was statistically significant for all variables (p b 0.05), except for structural social capital in the crude analysis (p b 0.06). Estimates with a significant derivation from 1 are depicted in bold (p b 0.05).
exposure was evident as the older the individual, the lower the risk of being exposed to public space violence. In the model search process, we found that model I showed no statistically significant association between violence and structural social capital while controlling for the explanatory variables. Model II showed that people with high cognitive social capital had a statistically significant 55% reduced risk (OR 0.45 CI 95: 0.28–0.74) compared to people with lower cognitive social capital while controlling for potential confounders (results not shown). In the third and final model, both structural and cognitive social capital were included while controlling for potential confounders. The results indicate that structural and cognitive social capital are two distinct constructs with inverse effects on the risk of violence. Statistically significant OR estimates in the final model suggest that cognitive social capital is a significant protective determinant for victimization (as put forward in the hypothesis) with an increase in cognitive social capital statistically significantly associated with reduction in the risk of violence (OR 0.46 CI 95: 0.28–0.76). In contrast, people with high structural social capital had a higher risk (OR 1.68 CI 95: 1.04–2.71) of being exposed to violence compared to people with low structural social capital. It remained that students were at a much higher risk of being exposed to violence compared to housewives (OR 3.77 CI 95: 2.06–6.87). No sign of confounding was observed between cognitive social capital and violence.
4. Discussion This research contributes to the empirical evidence on the impact of social capital on violence in public spaces providing knowledge on how communities might invest in order to prevent and reduce violence. Social capital is a complex concept in relation to violence and should be analyzed with at least two dimensions distinguishing between social trust indicators and social activism indicators. Our findings show that risk of violent crime may differ for different dimensions of social capital indicating that cognitive components of social capital rather than structural components are related to reduced risk of violence. However, when interpreting these results, it is crucial to keep in mind that due to the cross-sectional design of the study, we cannot distinguish the
N.S. Hansen-Nord et al. / Aggression and Violent Behavior 19 (2014) 643–648
causal route between social capital and risk of violence, and the potential for double causality should not be neglected. 4.1. Cognitive social capital, insecurity and reduced risk of violence Our results support previous evidence showing that people living in low-trust neighborhoods report higher rates of violence than people living in high-trust neighborhoods (Cuesta & Alda, 2012; Galea et al., 2002; Rosenfeld et al., 2001; Vial, Junges, Olinto, Machado, & Pattussi, 2010). Such results strongly indicate that cognitive social capital is a protective factor against violence, probably due to a dynamic of mutual protection in a community characterized by high level of cognitive social capital, and because the sense of solidarity strengthens with high levels of trust among community members. However, the causal relationship could as well be the other way around compared to the initial hypothesis; exposure to violence might affect the level of cognitive social capital rather than cognitive social capital affecting the risk of violence. Furthermore, feeling of insecurity in the community might form part of the causal chain between violence and cognitive social capital. Hence, a potential and somewhat logical explanation of the results could be that people who have been exposed to violence feel less secure, and thus report lower levels of cognitive social capital, and vice versa. Likewise, the fact that elderly people have a higher cognitive social capital and at the same time are less exposed to violence than younger people raises the question: do older people have higher cognitive social capital because they are less exposed to violence and feel more secure in their community? 4.2. Structural social capital and increased risk of violence When social capital is measured by aspects of social organization or cooperative action, like structural social capital, the view is different; people with higher level of structural social capital report higher levels of violence. This finding contradicts with the hypothesis of social organization as a protective factor of violence, prompting various potential explanations. Apart from extremely high crime risks in the study population, one possible explanation might be that people who are more actively engaged in the civic society through e.g. political activities, experience more “time at risk” in urban spaces, and are thus more exposed to violence compared to people who participate less in such community activities. This might also apply for other types of civic engagement such as sports and club participation which has been shown to increase risk of violence (Wright & Fitzpatrick, 2006). Furthermore, our results suggest that occupation status might influence on the relation between structural social capital and violence, as being a student is strongly associated with increased risk of violence compared to other forms of occupation, whether employed or unemployed. Finally, a subcultural perspective within urban criminology theory may play an important role in explaining the paradox that communities with strong network ties and social organization are subject to high crime rates. Browning, Dietz, and Feinberg (2004) argue that communities that cohere around the tolerance of criminal behavior call into question the regulatory role of social organization. To a large extent this might be the case for Honduras representing a historical culture of violence. In addition, the influence of living in a disadvantaged community in terms of poverty and unemployment on risk of violence should not be neglected, as this increases the exposure to street violence (De Coster, Heimer, & Wittrock, 2006). A different explanation model is suggested by Vilalta (2013) who argues that an association exists between violence and community organization against crime, as citizens collectively organize and participate in the fight against crime only if official collective security (e.g. local police) is considered ineffective (Vilalta, 2013); which is the case for NS and VN. Furthermore, repression at the systemic level and the perception of political violence increase certain types of civil society activism in communal groups and in professional associations (Booth &
647
Richard, 2000). This implies a potential reverse causal relationship between social capital and violence; people living in communities with high rates of violence will be more motivated to take collaborative action in violence prevention efforts. 4.3. Social organization as a facilitator of violence Motives for social organization should be recognized when trying to explain the positive relationship between structural social capital and risk of violence, since violence prevention might not always be on the agenda when community members organize. This is because social organization in general may foster certain types of criminal activity in disadvantaged neighborhoods characterized by high levels of poverty and unemployment (Browning et al., 2004). Social networks can thus hamper as well as nurture social cohesion depending on the context in which social networks arise (Sabatini, 2009). In this view, the explanation for the socially organized, high crime neighborhoods in our study can be found in the paradoxically oppositional effects of social organization itself. Structural social capital can thus possibly emerge from or somewhat support criminal networks provided that members of our study population belong to criminal networks. Hence, when interpreting the role of structural social capital in relation to violence, it is important to clearly define the content of the concept. 4.4. Social organization versus cooperative action: what exactly are we measuring with structural social capital? Social organization and cooperative action have been widely used as indicators for structural social capital in the literature, but in a Honduran context, they represent very dissimilar levels of civic participation and collective action. Community organization is very common among community citizens in our study population, particularly among elderly unemployed housewives, whereas cooperative action is less prevalent. Thus, organization is not necessarily linked to active participation in the society, but might indicate being a passive member of community groups e.g. women's groups or religious groups. However, it is not the organization in itself which is interesting in a violence prevention perspective, but more importantly it is the collective action for creating social change facilitated by being organized in the civil society. Hence, in our study the interpretation of structural social capital primarily represents social organization rather than cooperative action, and explanation models referring to cooperative action or collective action might not apply to our study context. In practice, the problem arises from the fact that measuring structural social capital through the sum scale (SASCAT) does not use weights to take the importance of the value/ importance of each specific item of the concept into account, in relation to violence prevention. E.g. “membership of women's groups” has the same value/score as “joined together and talked to authorities”, where the Honduran (CPTRT's) perception is different; joining together and talking to authorities represent a higher level of structural social capital than being a member of a woman's group. The issues of interpreting structural social capital represent a major weakness of the use of the SASCAT instrument for measuring structural social capital in the context of Honduras and other similar Latin American countries. Generally, cooperative action might not be measured merely through organization of the civil society in a Latin American context. Further studies are needed to elaborate the associations between cooperative action and violence. 4.5. Methodological strengths and limitations Strengths of the study include that we did not employ a proxyvictim design, but obtained information directly from victims of violence. Moreover, the fact that we were able to distinguish domestic violence from public space violence represented a distinct advantage for the interpretation of our results. Finally, the present study represents
648
N.S. Hansen-Nord et al. / Aggression and Violent Behavior 19 (2014) 643–648
scientifically robust new evidence on important aspects of violence prevention which is the first of its kind carried out in an extremely violent setting in Honduras, adding to the relatively thin knowledge base concerning prevention of violence in poor urban areas in the global South. However, some methodological limitations of the study should be considered; the main one being related to the cross-sectional design which precludes the possibility of concluding on the causal routes of data. Furthermore, the study was characterized by lack of power due to the relatively low number of public space violence cases (N = 94). Selection bias may have occurred if people who had been exposed to violence avoided participating out of fear, which would result in the underreporting of violence exposure. Moreover, underreporting might have occurred due to the sensitive nature of the subject including potential mistrust from the respondents regarding the purpose of the study causing them to refuse to participate. Besides, a natural selection bias was introduced as we could obviously not obtain data on civic participation details on deadly violations which would have been essential to include in the analyses, as violent behavior in Honduras often result in homicides. The possibility of bias should be considered as the interviewers for the study were young citizens of NS and VN and thus the possibility of the respondents knowing the interviewers personally was present. This could have affected the data quality, however, most likely in terms of easier access to and more in-depth answers from the respondents. Since the questions used for measuring social capital in our study were based on the SASCAT, which is considerably shorter than the original A-SCAT questionnaire, it may not measure the concept of social capital comprehensively. There is a risk that the questions were not adapted to the specific cultural setting meaning that some of the concepts were unfamiliar to the target group e.g. community, social support, and trust. For instance, perceptions of trust depend on personal experience and are not a constant as people can be trusted in some things but not others (De Silva et al., 2006). Therefore, the generalized questions of trust in the questionnaire potentially mean that it was difficult to capture the intended interpretation of trust in communities. Additional qualitative investigations of trust including conceptualizations of key concepts are recommended to clarify this limitation further. 5. Conclusions Our findings show that consequences of violent crime may vary for different dimensions of social capital indicating that cognitive components of social capital rather than structural components are related to reduced risk of violence. Our findings highlight the complexity of structural and cognitive social capital in relation to violence, as the different constructs may be simultaneously eroded or fostered by violence. Studying and measuring the association between violence and organized civic activism/cooperative action which intentionally aims at violence prevention need to be prioritized further to obtain more solid policy guidance in violence prevention interventions. Acknowledgment The authors thank the study respondents, for their participation in this study. DIGNITY Danish Institute Against Torture funded this research. References Abom, B. (2004). Social capital, NGOs, and development: A Guatemalan case study. Development in Practice, 14(3), 342–353. http://dx.doi.org/10.1080/09614520 42000191187a. Bank, W. (2010). Crime and violence in Central America: A development challenge — Executive summary. Retrieved from https://openknowledge.worldbank.org/handle/ 10986/2979
Booth, J. A., & Richard, P. B. (2000). The effects of repression, political violence, and pain and loss on social capital in Central America. XXII International Congress of the Latin American Studies Association. Browning, C. R., Dietz, R. D., & Feinberg, S. L. (2004). The paradox of social organization: Networks, collective efficacy, and violent crime in urban neighborhoods. Social Forces, 83(2), 503–534. http://dx.doi.org/10.1353/sof.2005.0006. Brune, N. E., & Bossert, T. (2009). Building social capital in post-conflict communities: Evidence from Nicaragua. Social Science & Medicine, 68(5), 885–893. http://dx.doi.org/ 10.1016/j.socscimed.2008.12.024. Cuesta, J., & Alda, E. (2012). The effects of trust on victimization in Colombia. Journal of Peace Research, 49(6), 833–846. http://dx.doi.org/10.1177/0022343312453063. Dasgupta, P., & Serageldin, I. (2000). Social capital: A multifaceted perspective. World Bank Publications. De Coster, S., Heimer, K., & Wittrock, S. M. (2006). Neighborhood disadvantage, social capital, street context, and youth violence. Sociological Quarterly, 47(4), 723–753. http://dx.doi.org/10.1111/j.1533-8525.2006.00064.x. De Silva, M. J., Harpham, T., Tuan, T., Bartolini, R., Penny, M. E., & Huttly, S. R. (2006). Psychometric and cognitive validation of a social capital measurement tool in Peru and Vietnam. Social Science & Medicine, 62(4), 941–953. http://dx.doi.org/10.1016/j. socscimed.2005.06.050. Dinesen, C., Ronsbo, H., Juárez, C., González, M., Méndez, M.Á. E., & Modvig, J. (2013). Violence and social capital in post-conflict. Guatemala, 43(3), 162–168. Galea, S., Karpati, A., & Kennedy, B. (2002). Social capital and violence in the United States, 1974–1993. Social Science & Medicine, 55(8), 1373–1383. http://dx.doi.org/10.1016/ S0277-9536(01)00274-X. Hernandez, A. A., & Grineski, S. E. (2012). Disrupted by violence: Children's well-being and families' economic, social, and cultural capital in Ciudad Juarez, Mexico. Revista Panamericana de Salud Pública, 31(5), 373–379. http://dx.doi.org/10.1590/S102049892012000500004. Kennedy, B. P., Kawachi, I., Prothrow-Stith, D., Lochner, K., & Gupta, V. (1998). Social capital, income inequality, and firearm violent crime. Social Science & Medicine, 47(1), 7–17. http://dx.doi.org/10.1016/S0277-9536(98)00097-5. Kjaerulf, F., & Barahona, R. (2010). Preventing violence and reinforcing human security: A rights-based framework for top-down and bottom-up action. Revista Panamericana de Salud Pública = Pan American Journal of Public Health, 27(5), 382–395. Krug, E. G., Dahlberg, L. L., Mercy, J. A., Zwi, A. B., & Lozano, R. (2002a). World report on violence and health. World Health Organization. Krug, E. G., Mercy, J. A., Dahlberg, L. L., & Zwi, A. B. (2002b). The world report on violence and health. The Lancet, 360(9339), 1083–1088. http://dx.doi.org/10.1016/S01406736(02)11133-0. Lederman, D., Loayza, N., & Menéndez, A. M. (2002). Violent crime: Does social capital matter? Economic Development and Cultural Change. Vol. 50.2002, 3. (pp. 509–539). Chicago, Ill: Univ. of Chicago Press, 509–539 (0013-0079, ZDB-ID 16883). Matute Rodríguez, A., & Santiago, I. G. (2007). Informe estadístico de la violencia en Guatemala. UNDP Guatemala. Matzopoulos, R., Bowman, B., Butchart, A., & Mercy, J. A. (2008). The impact of violence on health in low- to middle-income countries. International Journal of Injury Control and Safety Promotion, 15(4), 177–187. http://dx.doi.org/10.1080/17457300802396487. Messner, S. F., Baumer, E. P., & Rosenfeld, R. (2004). Dimensions of social capital and rates of criminal homicide. American Sociological Review, 69(6), 882–903. http://dx.doi.org/ 10.1177/000312240406900607. Moser, C. O. N., & McIlwaine, C. (2006). Latin American urban violence as a development concern: Towards a framework for violence reduction. World Development, 34(1), 89–112. http://dx.doi.org/10.1016/j.worlddev.2005.07.012. Portes, A. (1998). Social capital: Its origins and applications in modern sociology. Annual Review of Sociology, 24(1), 1–24. http://dx.doi.org/10.1146/annurev.soc.24.1.1. Putnam, R. D. (2001). Bowling alone: The collapse and revival of American community. Simon & Schuster. Rosenfeld, R., Baumer, E. P., & Messner, S. F. (2001). Social capital and homicide. Social Forces, 80(1), 283–310. http://dx.doi.org/10.1353/sof.2001.0086. Sabatini, F. (2009). Social capital as social networks: A new framework for measurement and an empirical analysis of its determinants and consequences. The Journal of SocioEconomics, 38(3), 429–442. http://dx.doi.org/10.1016/j.socec.2008.06.001. UNAH — IUDPAS (2013). I. U. de D., Paz y Seguridad. Observatiorio de la Violencia. Vial, E. A., Junges, J. R., Olinto, M. T. A., Machado, P. S., & Pattussi, M. P. (2010). Violência urbana e capital social em uma cidade do sul do Brasil: Um estudo quantitativo e qualitativo. Retrieved from http://www.lume.ufrgs.br/handle/10183/66418 Vilalta, C. J. (2013). Towards an understanding of community organization against crime: The case of Ciudad Juarez, Mexico. Stability: International Journal of Security and Development, 2(1). http://dx.doi.org/10.5334/sta.ao. Violence in Colombia (2000). Violence in Colombia: Building sustainable peace and social capital. World Bank Publications. Weiss, H. E. (2011). Adolescents as a source of social control: The utility of adolescent social capital for reducing violent delinquency. Sociological Spectrum, 31(5), 554–578. http://dx.doi.org/10.1080/02732173.2011.589783. Wright, D. R., & Fitzpatrick, K. M. (2006). Social capital and adolescent violent behavior: Correlates of fighting and weapon use among secondary school students. Social Forces, 84(3), 1435–1453. http://dx.doi.org/10.1353/sof.2006.0077. Yacoub, S., Arellano, S., & Padgett-Moncada, D. (2006). Violence related injuries, deaths and disabilities in the capital of Honduras. Injury, 37(5), 428–434. http://dx.doi.org/ 10.1016/j.injury.2005.12.001.
Contents lists available at ScienceDirect
Aggression and Violent Behavior
Social capital and violence in poor urban areas of Honduras☆ Nete Sloth Hansen-Nord a,⁎, Mette Skar a, Finn Kjaerulf a, Juan Almendarez b, Sergio Bähr b, Óscar Sosa b, Julio Castro b, Anne-Marie Nybo Andersen c, Jens Modvig a a b c
DIGNITY — Danish Institute Against Torture, Copenhagen, Denmark CPTRT — The Centre for the Prevention and Rehabilitation of Victims of Torture and Their Families, Tegucigalpa, Honduras The University of Copenhagen, Department of Public Health, Copenhagen, Denmark
a r t i c l e
i n f o
Article history: Received 21 August 2014 Received in revised form 27 September 2014 Accepted 27 September 2014 Available online 5 October 2014 Keywords: Social capital Structural Cognitive Violence Victimization Insecurity Trust
a b s t r a c t Honduras has the highest murder rate in the world: the high level of violence threatens the economic and social development of the country as it erodes human and social capital and limits trust among people in poor urban areas. However, neither a detailed consideration of the complex manner in which distinct dimensions of social capital interrelate with violence, nor the potential for double causality has received much attention. Objectives: The study examines the influence of structural social capital (social organization characteristics) and cognitive social capital (social trust and cohesion characteristics) on risk of violence in poor urban areas of Honduras. Methods: The study was carried out in two urban communities of Tegucigalpa experiencing high levels of violence and insecurity. For the quantitative analysis, 1000 individuals older than 18 answered a structured questionnaire. Violence exposure was evaluated based on respondents' self-reporting. Social capital was defined based on the use of the short version of the Adapted Social Capital Assessment Tool. Results: Our results support previous evidence from Guatemala showing that cognitive and structural social capital were inversely related to risk of violence: people with high cognitive social capital had a lower risk of violence (OR 0.46 CI 95: 0.28–0.76) compared to people with low cognitive social capital, whereas people with high structural social capital had a higher risk of violence (OR 1.68 CI 95: 1.04–2.71) compared to people with low structural social capital. Conclusions: Social trust and social activism exhibit significant associations with risk of violence, however, these dimensions are consequences as well as causes of violence. Implications for practice: In an intervention perspective it is important to recognize the difference between social organization and cooperative action for creating change, as these concepts represent very dissimilar levels of collective action toward violence. It is thus important to link the items of social capital, primarily within the structural dimension, to the specific objectives of a given intervention. © 2014 Elsevier Ltd. All rights reserved.
Contents 1.
2.
3. 4.
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . 1.1. Preventing violence through building social capital . . . . . . 1.2. Social capital and violence: a view at the literature . . . . . . Materials and methods . . . . . . . . . . . . . . . . . . . . . . 2.1. Data collection . . . . . . . . . . . . . . . . . . . . . . 2.2. Description of variables . . . . . . . . . . . . . . . . . . 2.3. Statistical analyses . . . . . . . . . . . . . . . . . . . . 2.4. Ethical considerations . . . . . . . . . . . . . . . . . . . Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.1. Cognitive social capital, insecurity and reduced risk of violence 4.2. Structural social capital and increased risk of violence . . . . .
. . . . . . . . . . . .
. . . . . . . . . . . .
. . . . . . . . . . . .
. . . . . . . . . . . .
. . . . . . . . . . . .
. . . . . . . . . . . .
. . . . . . . . . . . .
. . . . . . . . . . . .
. . . . . . . . . . . .
. . . . . . . . . . . .
. . . . . . . . . . . .
. . . . . . . . . . . .
. . . . . . . . . . . .
. . . . . . . . . . . .
. . . . . . . . . . . .
☆ The authors have no conflicts of interest to disclose. ⁎ Corresponding author at: DIGNITY — Danish Institute Against Torture, Bryggervangen 55, DK-2100 Copenhagen, Denmark. E-mail address: [email protected] (N.S. Hansen-Nord).
http://dx.doi.org/10.1016/j.avb.2014.09.013 1359-1789/© 2014 Elsevier Ltd. All rights reserved.
. . . . . . . . . . . .
. . . . . . . . . . . .
. . . . . . . . . . . .
. . . . . . . . . . . .
. . . . . . . . . . . .
. . . . . . . . . . . .
. . . . . . . . . . . .
. . . . . . . . . . . .
. . . . . . . . . . . .
. . . . . . . . . . . .
. . . . . . . . . . . .
. . . . . . . . . . . .
. . . . . . . . . . . .
. . . . . . . . . . . .
. . . . . . . . . . . .
. . . . . . . . . . . .
. . . . . . . . . . . .
. . . . . . . . . . . .
. . . . . . . . . . . .
. . . . . . . . . . . .
644 644 644 644 644 645 645 645 645 646 647 647
644
N.S. Hansen-Nord et al. / Aggression and Violent Behavior 19 (2014) 643–648
4.3. Social organization as a facilitator of violence . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.4. Social organization versus cooperative action: what exactly are we measuring with structural social capital? 4.5. Methodological strengths and limitations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5. Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Acknowledgment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
1. Introduction Crime and violence are key development issues for Latin America with the Central American countries; Honduras, Guatemala, and El Salvador representing the most violent countries in the region (Bank, 2010). Violence is a result of various societal, community and individual factors which interact in a complex manner. Its causes can be assessed within an ecological framework of four levels (Krug, Mercy, Dahlberg, & Zwi, 2002b); at the individual level (first level) personal behavior, history of substance abuse and/or psychological disorders affect the risk of violence, which, in combination with relationship parameters (second level) like family bonds, violent friends and/or low socioeconomic household status can increase the risk of becoming either a victim of violence or a perpetrator (Krug et al., 2002b). Likewise, community factors (level 3), such as income inequality, social acceptability of violence and social capital significantly affect the overall risk of violence (Krug et al., 2002b). Societal factors (level 4) such as ineffectiveness of the justice system and the lack of control of firearms are crucial root causes of violence. The majority of serious crimes in the region are never solved: in Guatemala the impunity reaches around 98% as a result of poor rule of law and political polarization (Matute Rodríguez & Santiago, 2007). High levels of impunity entail low levels of trust in the police and other authorities, and general fear and insecurity in the Latin American populations. Honduras is the world's most violent non-war country with a homicide rate of 85.5 per 100,000 inhabitants representing 7172 homicides in 2012; or an average of 20 murders per day. The murders are primarily committed with firearms (83.4%) and with a high proportion of contract killings (23.5%) (UNAH — IUDPAS, 2013). These threats have increased substantially over the past several years and the government of Honduras lacks sufficient resources to address these issues. The high level of violence threatens the economic and social development of the country in a context of poverty, high level of unemployment and local illicit drug trade, economic and gender inequality, and high firearm availability in combination with cultural norms that support violence (Krug, Dahlberg, Mercy, Zwi, & Lozano, 2002a). Importantly, violence is recognized as a serious public health problem due to violence-related physical and psychological morbidities and disabilities (Matzopoulos, Bowman, Butchart, & Mercy, 2008; Yacoub, Arellano, & Padgett-Moncada, 2006). 1.1. Preventing violence through building social capital Violence prevention efforts in Latin America emphasize the need for launching policies aiming to build social capital due to their ability to provide social control and to engage citizens in partnering with the state to hold the institutions, in particular the police, accountable (Brune & Bossert, 2009; Cuesta & Alda, 2012; Violence in Colombia, 2000). Such policies call for cooperative action from civil society, which requires restoration of social cohesion in local communities and trust between individuals and may yield trust between strangers, providing the basis for peace and development (Bank, 2010; Brune & Bossert, 2009; Cuesta & Alda, 2012). However, if populations do not experience support from both the state and the private sector, neither social stability nor widespread popular support, social capital weakens. Due to the complexity of the problem of violence, prevention programs need a high degree of inter-sectoral involvement where criminal justice
. . . . . .
. . . . . .
. . . . . .
. . . . . .
. . . . . .
. . . . . .
. . . . . .
. . . . . .
. . . . . .
. . . . . .
. . . . . .
. . . . . .
. . . . . .
. . . . . .
. . . . . .
647 647 647 648 648 648
reforms must be linked to broader reforms that address different sectors including education, health, social protection, and labor markets addressing risk factors at different ecological societal levels (Bank, 2010; Kjaerulf & Barahona, 2010; Moser & McIlwaine, 2006) and thus require long continuous efforts to produce positive results. 1.2. Social capital and violence: a view at the literature Previous research has established that violence is associated with a reduced level of social capital (Dinesen et al., 2013; Hernandez & Grineski, 2012; Putnam, 2001; Sabatini, 2009) and fears related to violence impede social organizing and civic participation (Abom, 2004). Correspondingly, social connection including opportunities for active participation of community members and organizations (both formal and informal) in social and economic life has been shown to be an important protective factor against violent behavior (Bank, 2010; Weiss, 2011). Key elements of social capital represent networks and reciprocated exchange, solidarity, trust, and social control (Portes, 1998). However, operationalizing this complex concept is difficult, and satisfactory measures of social capital are generally hard to find. In this research, we focus on two dominant dimensions in social capital; a “structural” dimension of social capital, consisting of network connections facilitating mutually beneficial collective action through established roles, and a “cognitive” dimension, consisting of attitudes toward trust, shared norms, values, attitudes, and beliefs (Dasgupta & Serageldin, 2000). Such dimensions of social capital, emphasizing the distinction between trust and civic engagement have previously been analyzed in relation to crime rates in the United States with different modeling of the dimensions. These studies include a study by Kennedy, Kawachi, ProthrowStith, Lochner, and Gupta (1998), which found that both dimensions are significantly associated with firearm violence, and two studies showing that high levels of trust are associated with lower homicide rates (Galea, Karpati, & Kennedy, 2002; Rosenfeld, Baumer, & Messner, 2001). The primary results of these studies report that the trust dimension of social capital seems to be associated with reduced homicide rates, whereas less and somewhat irregular support is found for the effects of other indicators of social capital; i.e. civil engagement on victimization (Lederman, Loayza, & Menéndez, 2002; Rosenfeld et al., 2001). This is particularly found in a recent study from Guatemala which emphasizes the importance of separating structural and cognitive social capital in relation to violence (Dinesen et al., 2013). Social trust and social activism do thus exhibit significant associations with homicide rates, however, these dimensions are consequences as well as causes of homicide (Messner, Baumer, & Rosenfeld, 2004). In this study we aim to investigate the characteristics of the association between the structural and cognitive dimensions of social capital and violence in one of the world's most violent settings. 2. Materials and methods 2.1. Data collection This study is based on a survey study undertaken by The Centre for Prevention and Rehabilitation of Victims of Torture and Their Families (CPTRT) in Honduras in 2011. The survey was carried out in two urban areas in Tegucigalpa, Nueva Suyapa (NS) and Villa Nueva (VN). NS and VN represent similar contexts in relation to poverty, levels of
N.S. Hansen-Nord et al. / Aggression and Violent Behavior 19 (2014) 643–648
insecurity and access to basic services. Data collection constituted a random sample of 500 households in NS and 500 households in VN, selected by the use of household mappings provided by the Honduras National Institute of Statistics, representing information on about 5500 individual family members. Data was collected using an intervieweradministered questionnaire facilitated by previously trained interviewers who were young citizens of NS and VN. The survey comprised demographic information, household characteristics, health indicators, social capital, victimization, security and perception of the justice system at an individual level. The first adult (N18 years) member of the household who opened the door was selected as the primary study participant. Interviews were carried out during weekends in order to increase the chance for all household members being home, and to attempt to improve the gender balance in the data. 2.2. Description of variables Social capital was measured based on questions from the short version of the Adapted Social Capital Assessment Tool (SASCAT), which is validated by psychometric techniques in addition to cognitive validation (De Silva et al., 2006). The questions clearly distinguished between structural and cognitive social capital. The cognitive elements of social capital focused on the respondents' values, norms, attitudes and levels of interpersonal trust and perception of the support among neighbors, friends, family and other community members. The structural elements focused on organization through participation in formal or informal organizations or community groups, and active participation in community activities (De Silva et al., 2006). Computing the responses on each question related to either cognitive or structural social capital, we constructed two scores on the degree (low vs. high) of cognitive and structural social capital per individual. As for violence, respondents were asked whether or not they had been victimized within the last 12 months from the time of the interview (dichotomous response). The type of violence included violence in public spaces; in the street, during public transport, and in private vehicles. Exposure to domestic violence was excluded from all analyses. Insecurity was assessed by asking whether the respondent felt secure, relatively secure, insecure or very insecure when leaving one's own home in the community. Occupation of the respondent was categorized into housewife, student, employed, self-employed, paid by the hour, and no job. Household income was assessed as a categorical variable with the following weekly income categories recalculated into the following categories: extreme poverty, poverty, middle income, and high income.
2.4. Ethical considerations Before participating in the study, the respondents were introduced to its aim and promised anonymity. Informed consent was given orally. Necessary precautions were taken to ensure that participation in the study did not expose the respondents to further danger. 3. Results The core analyses of the article examined the effect of structural social capital (social organization characteristics) and cognitive social capital (social trust and cohesion characteristics) on risk of violence in two poor urban areas of Honduras. The mean age of the respondents was 38 years (SD 16.6) and 75.9% were women. Within a time window of 12 months, the prevalence of violence at household level was 21.9%, whereas 10.4% of the respondents had been directly exposed to violence in public spaces. Table 1 shows violence exposure distributed by the explanatory variables representing a set of demographic, socioeconomic, and opinion-based categorical variables. There was an age gradient in violence exposure with the highest violence exposure among the youngest age group (13–23 years). Moreover, victimization was significantly associated with a feeling of insecurity in the community and people who had been victims of violence were generally more insecure in their community. The distribution of high versus low social capital on the explanatory variables is presented in Table 2. Age was significantly associated with cognitive social capital with a negative gradient; younger age correlated to lower cognitive social capital; however this gradient was not found for structural social capital. Out of those who had been exposed to violence about 70% had low cognitive social capital compared to only 30% with high cognitive social capital. For the remaining variables no Table 1 Description of the study population.
Individual level variables Sex
Age
2.3. Statistical analyses Bivariate distributions were tested for statistical significance using the χ2-test for nominal and the γ-test for ordinal variables. Crude analyses between violence and explanatory variables were first performed to determine potential confounders. Multivariate logistic regression analyses were used to estimate odds ratios (OR) for violence exposure given the potential confounders. A backwards model search strategy was applied where the start model included all potential confounders. Insignificant estimates were then removed from the model step by step until all remaining estimates in the final model were statistically significant. Three models were constructed addressing the hypothesis in different ways; model I investigated the association between violence and structural social capital given a set of explanatory variables excluding cognitive social capital; model II investigated the association between violence and cognitive social capital given a set of explanatory variables excluding structural social capital, and model III (the final model) investigated the association between violence and structural social capital controlling for the effect of cognitive social capital and vice versa. All analyses were conducted with a two-sided level of significance of 0.05, and calculated using SPSS 19.0.
645
District
Household income
Occupation
Insecurity
Male Female Missing 13–23 years 24–34 years 35–49 years 50+ years Missing Nueva Suyapa Villa Nueva Missing Extreme poverty Poverty Middle income High income Missing Housewife Student Employed Self employed Paid by the hour No job Missing Secure Relatively secure Insecure Very insecure Missing
Total
Public space violence
N = 1000
N = 94
n
n (%)
239 752 9 246 243 243 260 8 522 477 1 198 237 222 254 89 513 106 137 124 42 41 37 384 103 451 50 12
26 (10.9) 66 (8.8) 36 (14.6) 25 (10.2) 18 (7.4) 15 (5.8) 55 (10.5) 39 (8.2) 15 (7.6) 15 (6.3) 26 (11.7) 30 (11.8) 40 (7.8) 25 (23.6) 11 (8.0) 11 (8.9) 5 (11.9) 2 (4.9) 27 (7.0) 13 (12.0) 46 (10.2) 8 (16.0)
The distribution of nominal variables (sex, district and occupation) was tested using a χ2test. The distribution of ordinal variables (age, income and insecurity) was tested using a γ-test. Statistically significant distributions are depicted in bold (p b 0.05).
646
N.S. Hansen-Nord et al. / Aggression and Violent Behavior 19 (2014) 643–648
Table 2 Distribution of social capital on explanatory variables. Structural social capital
Sex Age
District Insecurity
Household income
Occupation
Public space violence
Male Female 13–23 years 24–34 years 35–49 years 50+ years Villa Nueva Nueva Suyapa Secure Relatively secure Insecure Very insecure Extreme poverty Poverty Middle income High income Housewife Student Employed Self employed Paid by the hour No job Yes No
Cognitive social capital
Low n (%)
High n (%)
Low n (%)
High n (%)
113 (49.6) 383 (53.0) 117 (49.0) 122 (52.1) 122 (53.5) 140 (55.6) 241 (52.3) 261 (52.4) 196 (53.1) 43 (43.9) 233 (53.4) 23 (50.0) 109 (57.1) 111 (49.6) 116 (54.0) 116 (47.4) 256 (51.7) 50 (48.1) 66 (51.2) 64 (54.7) 19 (46.3) 22 (55.0) 37 (41.1) 399 (51.8)
115 (50.4) 340 (47.0) 122 (51.0) 112 (47.9) 106 (46.5) 112 (44.4) 220 (47.7) 237 (47.6) 173 (46.9) 55 (56.1) 203 (46.6) 23 (50.0) 82 (42.9) 113 (50.4) 99 (46.0) 129 (52.7) 239 (48.3) 54 (51.9) 63 (48.8) 53 (45.3) 22 (53.7) 18 (45.0) 53 (58.9) 372 (48.2)
105 (48.6) 350 (51.8) 139 (60.2) 121 (55.3) 107 (49.5) 88 (38.6) 211 (50.0) 247 (51.7) 159 (46.5) 48 (51.6) 221 (54.2) 23 (50.0) 91 (52.0) 112 (50.2) 107 (54.0) 110 (48.7) 235 (50.9) 62 (62.6) 55 (44.4) 59 (52.2) 17 (53.1) 14 (36.8) 61 (70.1) 359 (49.9)
111 (51.4) 326 (48.2) 92 (39.8) 98 (44.7) 109 (50.5) 140 (61.4) 211 (50.0) 231 (48.3) 183 (53.5) 45 (48.4) 187 (45.8) 23 (50.0) 84 (48.0) 111 (49.8) 91 (46.0) 116 (51.3) 227 (49.1) 37 (37.4) 69 (55.6) 54 (47.8) 15 (46.9) 24 (63.2) 26 (29.9) 360 (50.1)
The distribution of nominal variables (sex, district, occupation and violence) was tested using a χ2-test. The distribution of ordinal variables (age, income and insecurity) was tested using a γ-test. Statistically significant distributions are depicted in bold (p b 0.05).
statistically significant differences were seen for the people with high or low social capital. However, there was a tendency that the majority of people feeling secure in their community had a high level of cognitive social capital, and likewise, the majority of people feeling insecure had a low level of cognitive social capital. Students generally had a low level of cognitive social capital and people with no jobs reported higher cognitive social capital, however insignificant. The same tendencies were not seen for structural social capital. Table 3 shows the results of the logistic regression analyses of social capital and public space violence. The crude analyses show that people with high cognitive social capital had less than half the risk of violence compared to people with low cognitive social capital (OR 0.43 CI 95: 0.26–0.69). Students had more than a threefold risk (OR 3.71 CI 95: 2.12–6.48) of being exposed to violence compared to housewives, and people of 50 or more years of age had a 63% significantly lower risk (OR 0.37 CI 95: 0.20–0.70) of being exposed to violence compared to the younger generation (13–23 years). The age gradient in violence Table 3 Logistical regression analysis of social capital and experienced public space violence at individual level.
Structural social capital Cognitive social capital Occupation
Age
Low High Low High Housewife Student Employed Self employed Paid by the hour No job 13–23 years 24–34 years 35–49 years 50+ years
Crude OR OR (95% CI)
Final model OR (95% CI)
1 1.54 (0.99–2.39) 1 0.43 (0.26–0.69) 1 3.71 (2.12–6.48) 1.22 (0.61–2.47) 1.19 (0.59–2.39) 1.53 (0.57–4.11) 0.58 (0.13–2.48) 1 0.73 (0.42–1.26) 0.50 (0.27–0.91) 0.37 (0.20–0.70)
1 1.68 (1.04–2.71) 1 0.46 (0.28–0.76) 1 3.77 (2.06–6.87) 1.55 (0.75–3.21) 1.39 (0.66–2.96) 1.72 (0.56–5.30) 0.78 (0.18–3.44)
The overall Wald-test was statistically significant for all variables (p b 0.05), except for structural social capital in the crude analysis (p b 0.06). Estimates with a significant derivation from 1 are depicted in bold (p b 0.05).
exposure was evident as the older the individual, the lower the risk of being exposed to public space violence. In the model search process, we found that model I showed no statistically significant association between violence and structural social capital while controlling for the explanatory variables. Model II showed that people with high cognitive social capital had a statistically significant 55% reduced risk (OR 0.45 CI 95: 0.28–0.74) compared to people with lower cognitive social capital while controlling for potential confounders (results not shown). In the third and final model, both structural and cognitive social capital were included while controlling for potential confounders. The results indicate that structural and cognitive social capital are two distinct constructs with inverse effects on the risk of violence. Statistically significant OR estimates in the final model suggest that cognitive social capital is a significant protective determinant for victimization (as put forward in the hypothesis) with an increase in cognitive social capital statistically significantly associated with reduction in the risk of violence (OR 0.46 CI 95: 0.28–0.76). In contrast, people with high structural social capital had a higher risk (OR 1.68 CI 95: 1.04–2.71) of being exposed to violence compared to people with low structural social capital. It remained that students were at a much higher risk of being exposed to violence compared to housewives (OR 3.77 CI 95: 2.06–6.87). No sign of confounding was observed between cognitive social capital and violence.
4. Discussion This research contributes to the empirical evidence on the impact of social capital on violence in public spaces providing knowledge on how communities might invest in order to prevent and reduce violence. Social capital is a complex concept in relation to violence and should be analyzed with at least two dimensions distinguishing between social trust indicators and social activism indicators. Our findings show that risk of violent crime may differ for different dimensions of social capital indicating that cognitive components of social capital rather than structural components are related to reduced risk of violence. However, when interpreting these results, it is crucial to keep in mind that due to the cross-sectional design of the study, we cannot distinguish the
N.S. Hansen-Nord et al. / Aggression and Violent Behavior 19 (2014) 643–648
causal route between social capital and risk of violence, and the potential for double causality should not be neglected. 4.1. Cognitive social capital, insecurity and reduced risk of violence Our results support previous evidence showing that people living in low-trust neighborhoods report higher rates of violence than people living in high-trust neighborhoods (Cuesta & Alda, 2012; Galea et al., 2002; Rosenfeld et al., 2001; Vial, Junges, Olinto, Machado, & Pattussi, 2010). Such results strongly indicate that cognitive social capital is a protective factor against violence, probably due to a dynamic of mutual protection in a community characterized by high level of cognitive social capital, and because the sense of solidarity strengthens with high levels of trust among community members. However, the causal relationship could as well be the other way around compared to the initial hypothesis; exposure to violence might affect the level of cognitive social capital rather than cognitive social capital affecting the risk of violence. Furthermore, feeling of insecurity in the community might form part of the causal chain between violence and cognitive social capital. Hence, a potential and somewhat logical explanation of the results could be that people who have been exposed to violence feel less secure, and thus report lower levels of cognitive social capital, and vice versa. Likewise, the fact that elderly people have a higher cognitive social capital and at the same time are less exposed to violence than younger people raises the question: do older people have higher cognitive social capital because they are less exposed to violence and feel more secure in their community? 4.2. Structural social capital and increased risk of violence When social capital is measured by aspects of social organization or cooperative action, like structural social capital, the view is different; people with higher level of structural social capital report higher levels of violence. This finding contradicts with the hypothesis of social organization as a protective factor of violence, prompting various potential explanations. Apart from extremely high crime risks in the study population, one possible explanation might be that people who are more actively engaged in the civic society through e.g. political activities, experience more “time at risk” in urban spaces, and are thus more exposed to violence compared to people who participate less in such community activities. This might also apply for other types of civic engagement such as sports and club participation which has been shown to increase risk of violence (Wright & Fitzpatrick, 2006). Furthermore, our results suggest that occupation status might influence on the relation between structural social capital and violence, as being a student is strongly associated with increased risk of violence compared to other forms of occupation, whether employed or unemployed. Finally, a subcultural perspective within urban criminology theory may play an important role in explaining the paradox that communities with strong network ties and social organization are subject to high crime rates. Browning, Dietz, and Feinberg (2004) argue that communities that cohere around the tolerance of criminal behavior call into question the regulatory role of social organization. To a large extent this might be the case for Honduras representing a historical culture of violence. In addition, the influence of living in a disadvantaged community in terms of poverty and unemployment on risk of violence should not be neglected, as this increases the exposure to street violence (De Coster, Heimer, & Wittrock, 2006). A different explanation model is suggested by Vilalta (2013) who argues that an association exists between violence and community organization against crime, as citizens collectively organize and participate in the fight against crime only if official collective security (e.g. local police) is considered ineffective (Vilalta, 2013); which is the case for NS and VN. Furthermore, repression at the systemic level and the perception of political violence increase certain types of civil society activism in communal groups and in professional associations (Booth &
647
Richard, 2000). This implies a potential reverse causal relationship between social capital and violence; people living in communities with high rates of violence will be more motivated to take collaborative action in violence prevention efforts. 4.3. Social organization as a facilitator of violence Motives for social organization should be recognized when trying to explain the positive relationship between structural social capital and risk of violence, since violence prevention might not always be on the agenda when community members organize. This is because social organization in general may foster certain types of criminal activity in disadvantaged neighborhoods characterized by high levels of poverty and unemployment (Browning et al., 2004). Social networks can thus hamper as well as nurture social cohesion depending on the context in which social networks arise (Sabatini, 2009). In this view, the explanation for the socially organized, high crime neighborhoods in our study can be found in the paradoxically oppositional effects of social organization itself. Structural social capital can thus possibly emerge from or somewhat support criminal networks provided that members of our study population belong to criminal networks. Hence, when interpreting the role of structural social capital in relation to violence, it is important to clearly define the content of the concept. 4.4. Social organization versus cooperative action: what exactly are we measuring with structural social capital? Social organization and cooperative action have been widely used as indicators for structural social capital in the literature, but in a Honduran context, they represent very dissimilar levels of civic participation and collective action. Community organization is very common among community citizens in our study population, particularly among elderly unemployed housewives, whereas cooperative action is less prevalent. Thus, organization is not necessarily linked to active participation in the society, but might indicate being a passive member of community groups e.g. women's groups or religious groups. However, it is not the organization in itself which is interesting in a violence prevention perspective, but more importantly it is the collective action for creating social change facilitated by being organized in the civil society. Hence, in our study the interpretation of structural social capital primarily represents social organization rather than cooperative action, and explanation models referring to cooperative action or collective action might not apply to our study context. In practice, the problem arises from the fact that measuring structural social capital through the sum scale (SASCAT) does not use weights to take the importance of the value/ importance of each specific item of the concept into account, in relation to violence prevention. E.g. “membership of women's groups” has the same value/score as “joined together and talked to authorities”, where the Honduran (CPTRT's) perception is different; joining together and talking to authorities represent a higher level of structural social capital than being a member of a woman's group. The issues of interpreting structural social capital represent a major weakness of the use of the SASCAT instrument for measuring structural social capital in the context of Honduras and other similar Latin American countries. Generally, cooperative action might not be measured merely through organization of the civil society in a Latin American context. Further studies are needed to elaborate the associations between cooperative action and violence. 4.5. Methodological strengths and limitations Strengths of the study include that we did not employ a proxyvictim design, but obtained information directly from victims of violence. Moreover, the fact that we were able to distinguish domestic violence from public space violence represented a distinct advantage for the interpretation of our results. Finally, the present study represents
648
N.S. Hansen-Nord et al. / Aggression and Violent Behavior 19 (2014) 643–648
scientifically robust new evidence on important aspects of violence prevention which is the first of its kind carried out in an extremely violent setting in Honduras, adding to the relatively thin knowledge base concerning prevention of violence in poor urban areas in the global South. However, some methodological limitations of the study should be considered; the main one being related to the cross-sectional design which precludes the possibility of concluding on the causal routes of data. Furthermore, the study was characterized by lack of power due to the relatively low number of public space violence cases (N = 94). Selection bias may have occurred if people who had been exposed to violence avoided participating out of fear, which would result in the underreporting of violence exposure. Moreover, underreporting might have occurred due to the sensitive nature of the subject including potential mistrust from the respondents regarding the purpose of the study causing them to refuse to participate. Besides, a natural selection bias was introduced as we could obviously not obtain data on civic participation details on deadly violations which would have been essential to include in the analyses, as violent behavior in Honduras often result in homicides. The possibility of bias should be considered as the interviewers for the study were young citizens of NS and VN and thus the possibility of the respondents knowing the interviewers personally was present. This could have affected the data quality, however, most likely in terms of easier access to and more in-depth answers from the respondents. Since the questions used for measuring social capital in our study were based on the SASCAT, which is considerably shorter than the original A-SCAT questionnaire, it may not measure the concept of social capital comprehensively. There is a risk that the questions were not adapted to the specific cultural setting meaning that some of the concepts were unfamiliar to the target group e.g. community, social support, and trust. For instance, perceptions of trust depend on personal experience and are not a constant as people can be trusted in some things but not others (De Silva et al., 2006). Therefore, the generalized questions of trust in the questionnaire potentially mean that it was difficult to capture the intended interpretation of trust in communities. Additional qualitative investigations of trust including conceptualizations of key concepts are recommended to clarify this limitation further. 5. Conclusions Our findings show that consequences of violent crime may vary for different dimensions of social capital indicating that cognitive components of social capital rather than structural components are related to reduced risk of violence. Our findings highlight the complexity of structural and cognitive social capital in relation to violence, as the different constructs may be simultaneously eroded or fostered by violence. Studying and measuring the association between violence and organized civic activism/cooperative action which intentionally aims at violence prevention need to be prioritized further to obtain more solid policy guidance in violence prevention interventions. Acknowledgment The authors thank the study respondents, for their participation in this study. DIGNITY Danish Institute Against Torture funded this research. References Abom, B. (2004). Social capital, NGOs, and development: A Guatemalan case study. Development in Practice, 14(3), 342–353. http://dx.doi.org/10.1080/09614520 42000191187a. Bank, W. (2010). Crime and violence in Central America: A development challenge — Executive summary. Retrieved from https://openknowledge.worldbank.org/handle/ 10986/2979
Booth, J. A., & Richard, P. B. (2000). The effects of repression, political violence, and pain and loss on social capital in Central America. XXII International Congress of the Latin American Studies Association. Browning, C. R., Dietz, R. D., & Feinberg, S. L. (2004). The paradox of social organization: Networks, collective efficacy, and violent crime in urban neighborhoods. Social Forces, 83(2), 503–534. http://dx.doi.org/10.1353/sof.2005.0006. Brune, N. E., & Bossert, T. (2009). Building social capital in post-conflict communities: Evidence from Nicaragua. Social Science & Medicine, 68(5), 885–893. http://dx.doi.org/ 10.1016/j.socscimed.2008.12.024. Cuesta, J., & Alda, E. (2012). The effects of trust on victimization in Colombia. Journal of Peace Research, 49(6), 833–846. http://dx.doi.org/10.1177/0022343312453063. Dasgupta, P., & Serageldin, I. (2000). Social capital: A multifaceted perspective. World Bank Publications. De Coster, S., Heimer, K., & Wittrock, S. M. (2006). Neighborhood disadvantage, social capital, street context, and youth violence. Sociological Quarterly, 47(4), 723–753. http://dx.doi.org/10.1111/j.1533-8525.2006.00064.x. De Silva, M. J., Harpham, T., Tuan, T., Bartolini, R., Penny, M. E., & Huttly, S. R. (2006). Psychometric and cognitive validation of a social capital measurement tool in Peru and Vietnam. Social Science & Medicine, 62(4), 941–953. http://dx.doi.org/10.1016/j. socscimed.2005.06.050. Dinesen, C., Ronsbo, H., Juárez, C., González, M., Méndez, M.Á. E., & Modvig, J. (2013). Violence and social capital in post-conflict. Guatemala, 43(3), 162–168. Galea, S., Karpati, A., & Kennedy, B. (2002). Social capital and violence in the United States, 1974–1993. Social Science & Medicine, 55(8), 1373–1383. http://dx.doi.org/10.1016/ S0277-9536(01)00274-X. Hernandez, A. A., & Grineski, S. E. (2012). Disrupted by violence: Children's well-being and families' economic, social, and cultural capital in Ciudad Juarez, Mexico. Revista Panamericana de Salud Pública, 31(5), 373–379. http://dx.doi.org/10.1590/S102049892012000500004. Kennedy, B. P., Kawachi, I., Prothrow-Stith, D., Lochner, K., & Gupta, V. (1998). Social capital, income inequality, and firearm violent crime. Social Science & Medicine, 47(1), 7–17. http://dx.doi.org/10.1016/S0277-9536(98)00097-5. Kjaerulf, F., & Barahona, R. (2010). Preventing violence and reinforcing human security: A rights-based framework for top-down and bottom-up action. Revista Panamericana de Salud Pública = Pan American Journal of Public Health, 27(5), 382–395. Krug, E. G., Dahlberg, L. L., Mercy, J. A., Zwi, A. B., & Lozano, R. (2002a). World report on violence and health. World Health Organization. Krug, E. G., Mercy, J. A., Dahlberg, L. L., & Zwi, A. B. (2002b). The world report on violence and health. The Lancet, 360(9339), 1083–1088. http://dx.doi.org/10.1016/S01406736(02)11133-0. Lederman, D., Loayza, N., & Menéndez, A. M. (2002). Violent crime: Does social capital matter? Economic Development and Cultural Change. Vol. 50.2002, 3. (pp. 509–539). Chicago, Ill: Univ. of Chicago Press, 509–539 (0013-0079, ZDB-ID 16883). Matute Rodríguez, A., & Santiago, I. G. (2007). Informe estadístico de la violencia en Guatemala. UNDP Guatemala. Matzopoulos, R., Bowman, B., Butchart, A., & Mercy, J. A. (2008). The impact of violence on health in low- to middle-income countries. International Journal of Injury Control and Safety Promotion, 15(4), 177–187. http://dx.doi.org/10.1080/17457300802396487. Messner, S. F., Baumer, E. P., & Rosenfeld, R. (2004). Dimensions of social capital and rates of criminal homicide. American Sociological Review, 69(6), 882–903. http://dx.doi.org/ 10.1177/000312240406900607. Moser, C. O. N., & McIlwaine, C. (2006). Latin American urban violence as a development concern: Towards a framework for violence reduction. World Development, 34(1), 89–112. http://dx.doi.org/10.1016/j.worlddev.2005.07.012. Portes, A. (1998). Social capital: Its origins and applications in modern sociology. Annual Review of Sociology, 24(1), 1–24. http://dx.doi.org/10.1146/annurev.soc.24.1.1. Putnam, R. D. (2001). Bowling alone: The collapse and revival of American community. Simon & Schuster. Rosenfeld, R., Baumer, E. P., & Messner, S. F. (2001). Social capital and homicide. Social Forces, 80(1), 283–310. http://dx.doi.org/10.1353/sof.2001.0086. Sabatini, F. (2009). Social capital as social networks: A new framework for measurement and an empirical analysis of its determinants and consequences. The Journal of SocioEconomics, 38(3), 429–442. http://dx.doi.org/10.1016/j.socec.2008.06.001. UNAH — IUDPAS (2013). I. U. de D., Paz y Seguridad. Observatiorio de la Violencia. Vial, E. A., Junges, J. R., Olinto, M. T. A., Machado, P. S., & Pattussi, M. P. (2010). Violência urbana e capital social em uma cidade do sul do Brasil: Um estudo quantitativo e qualitativo. Retrieved from http://www.lume.ufrgs.br/handle/10183/66418 Vilalta, C. J. (2013). Towards an understanding of community organization against crime: The case of Ciudad Juarez, Mexico. Stability: International Journal of Security and Development, 2(1). http://dx.doi.org/10.5334/sta.ao. Violence in Colombia (2000). Violence in Colombia: Building sustainable peace and social capital. World Bank Publications. Weiss, H. E. (2011). Adolescents as a source of social control: The utility of adolescent social capital for reducing violent delinquency. Sociological Spectrum, 31(5), 554–578. http://dx.doi.org/10.1080/02732173.2011.589783. Wright, D. R., & Fitzpatrick, K. M. (2006). Social capital and adolescent violent behavior: Correlates of fighting and weapon use among secondary school students. Social Forces, 84(3), 1435–1453. http://dx.doi.org/10.1353/sof.2006.0077. Yacoub, S., Arellano, S., & Padgett-Moncada, D. (2006). Violence related injuries, deaths and disabilities in the capital of Honduras. Injury, 37(5), 428–434. http://dx.doi.org/ 10.1016/j.injury.2005.12.001.