Arrest warrants are an important and pervasive aspect of crime and justice in the United States. There are nearly three million arrest warrants active on any given day, of which several hundred thousand were issued for serious violent crimes (SVCs) such as aggravated assault, robbery, forcible sexual assault, and homicide. In more than a third of those SVC warrants, however, extradition is conditionally waived such that an offender can avoid arrest by leaving town; they can elect banishment over prosecution. Studying extradition limits among these serious offenses presents an opportunity to illuminate both a challenge to public safety and justice, as well as forces underlying discretionary decision making by police. To that end, we study all arrest warrants issued in the United States for SVCs between 2017 and 2019. We model banishment rates at the county-level within a multivariate negative binomial framework. Analyses showed banishment varied as a function of policing capacity, firearm use in crimes, racial composition, and voting behavior during the 2016 presidential election.

中文翻译：

---

驱逐正义：美国的引渡限制

逮捕令是美国犯罪和司法的一个重要而普遍的方面。每天都有近 300 万份逮捕令有效，其中数十万份针对严重暴力犯罪 (SVC)，如严重袭击、抢劫、强迫性侵犯和杀人罪。然而，在超过三分之一的 SVC 逮捕令中，有条件地放弃了引渡，这样罪犯就可以通过离开城镇来避免被捕；他们可以选择驱逐而不是起诉。研究这些严重罪行中的引渡限制提供了一个机会，可以阐明对公共安全和司法的挑战，以及警方酌情决策的潜在力量。为此，我们研究了 2017 年至 2019 年间在美国针对 SVC 发出的所有逮捕令。我们在多元负二项式框架内对县级流放率进行建模。分析显示，在 2016 年总统大选期间，流放率因警务能力、犯罪中使用枪支、种族构成和投票行为而异。