University of São Paulo Medical School, Preventive Medicine Department, Brazil.
Department of Sociology, Utrecht University, The Netherlands.
Soc Sci Med. 2017 Oct;190:92-100. doi: 10.1016/j.socscimed.2017.08.013. Epub 2017 Aug 18.
Since the 1990s, researchers have noted declining trends in crime and violence, particularly homicide, in Western countries. Studies have explored national and sub-national trends using latent trajectory analysis techniques and identified several factors associated with declining and/or increasing trajectories. Social disorganization (SD) has been consistently linked to increases in homicide rates over time, explaining at least some of the spatial and temporal heterogeneity of homicide. Similar studies have not yet been carried out in Latin America's cities. In this paper we use Group Based Trajectory models to study homicide mortality rate [HMR] trajectories in Brazilian municipalities between 1991 and 2010. Then, through binary and multinomial logistic regression we investigated the association between SD in 1991, and the likelihood of an increasing HMR trajectory. We carried out an ecological time series study using all Brazilian municipalities in the period between 1991 and 2010 (n = 4491). Data on homicide deaths were collected from the Mortality Information System of the Ministry of Health and standardized by age to calculate HMR per 100,000 population. Socioeconomic and demographic data for 1991 were used to compose the composite measure of SD. Our results highlight the spatial and temporal heterogeneity of homicide mortality in Brazilian municipalities. While national trends are steadily increasing, disaggregating municipal trajectories shows that this is driven by a small proportion of municipalities in the country. We found that SD is associated with an ascending homicide trajectory. This result generally supports the notion that poor social structural conditions can create 'space' for criminal behavior and groups and, consequently, violent death.
自 20 世纪 90 年代以来,研究人员注意到西方国家的犯罪和暴力行为(尤其是凶杀案)呈下降趋势。研究采用潜在轨迹分析技术探讨了国家和次国家趋势,并确定了与下降和/或上升轨迹相关的几个因素。社会解体(SD)一直与凶杀率的上升有关,它至少解释了凶杀的一些空间和时间异质性。类似的研究尚未在拉丁美洲的城市进行。在本文中,我们使用基于群组的轨迹模型来研究 1991 年至 2010 年巴西各城市的凶杀死亡率[HMR]轨迹。然后,我们通过二元和多项逻辑回归,研究了 1991 年的 SD 与 HMR 上升轨迹可能性之间的关系。我们对 1991 年至 2010 年期间所有巴西城市进行了生态时间序列研究(n=4491)。凶杀死亡数据来自卫生部的死亡率信息系统收集,并按年龄标准化,以计算每 10 万人的 HMR。1991 年的社会经济和人口数据用于组成 SD 的综合衡量标准。我们的研究结果突出了巴西各城市凶杀死亡率的空间和时间异质性。虽然国家趋势稳步上升,但对市政轨迹的细分表明,这是由该国一小部分城市推动的。我们发现,SD 与上升的凶杀轨迹有关。这一结果普遍支持了这样一种观点,即贫困的社会结构条件可以为犯罪行为和群体创造“空间”,从而导致暴力死亡。