Wang Xu, Macinko James, Porfiri Maurizio, Sipahi Rifat
Department of Mechanical and Industrial Engineering, Northeastern University, Boston, MA, 02115, USA.
Department of Health Policy and Management, Fielding School of Public Health, University of California Los Angeles, Los Angeles, CA, 90095, USA.
SSM Popul Health. 2024 May 23;27:101680. doi: 10.1016/j.ssmph.2024.101680. eCollection 2024 Sep.
Firearm-related crimes and self-inflicted harms pose a significant threat to the safety and well-being of Americans. Investigation of firearm prevalence in the United States (U.S.) has therefore been a center of attention. A critical aspect in this endeavor is to explain whether there are identifiable patterns in firearm acquisition.
We view firearm acquisition patterning as a spatio-temporal dynamical system distributed across U.S. states that co-evolves with crime rates, political ideology, income levels, population, and the legal environment. We leverage transfer entropy and exponential random graph models along with publicly available data, to statistically reveal the formative factors in how each state's temporal patterning of firearm acquisition influences other states.
Results help to explain how and why U.S. states influence each other in their firearm acquisition. We establish that state-to-state influences, or lack thereof, in firearm acquisition patterning are explained by states' percent of gun homicide, firearm law strictness, geographic neighborhood, and citizen ideology. Network-based characteristics, namely, mutuality and transitivity, are also important to explain such influence.
Results suggest that state policies or programs that reduce gun homicides will also help suppress that state's influence on the patterning of firearm acquisition in other states. Furthermore, states with stricter firearm laws are more likely to influence firearm acquisition in other states, but are themselves shielded from the effects of other states' firearm acquisition patterns. These results inform future research in public health, criminology, and policy making.
与枪支相关的犯罪和自我伤害行为对美国人的安全与福祉构成了重大威胁。因此，对美国枪支流行情况的调查一直是关注的焦点。这项工作的一个关键方面是解释枪支获取是否存在可识别的模式。
我们将枪支获取模式视为一个分布在美国各州的时空动态系统，它与犯罪率、政治意识形态、收入水平、人口和法律环境共同演变。我们利用转移熵和指数随机图模型以及公开可用的数据，从统计学角度揭示每个州枪支获取的时间模式如何影响其他州的形成因素。
结果有助于解释美国各州在枪支获取方面如何以及为何相互影响。我们确定，枪支获取模式中各州之间的影响或缺乏影响，可以通过各州的枪支杀人案百分比、枪支法律的严格程度、地理邻域和公民意识形态来解释。基于网络的特征，即相互性和传递性，对于解释这种影响也很重要。
结果表明，减少枪支杀人案的州政策或项目也将有助于抑制该州对其他州枪支获取模式的影响。此外，枪支法律更严格的州更有可能影响其他州的枪支获取，但它们自身却能免受其他州枪支获取模式的影响。这些结果为公共卫生、犯罪学和政策制定方面的未来研究提供了参考。
