This study examines the clustering of Ellis Act eviction notices in the city of San Francisco over the period 1997 through 2016. The Ellis Act is a California state law that allows landlords to go out of the rental business on a property, withdrawing the property from the market and evicting the tenants as a no-fault eviction. The law has been a source of contention since its enactment in 1986, because tenants argue it is used to displace low-income renters in a broader gentrification process. We address this question with Ripley’s K and Kuldorff’s D measures that assess the degree of “absolute” and “relative” clustering of Ellis Act evictions using all other eviction notices as the control. If the law is used as intended, we should find support for the null hypothesis that Ellis Act evictions do not cluster spatially relative to the background of controls. The alternative hypothesis that they do cluster relative to other types might lend support to the argument that the law serves as a legal mechanism in the displacement of tenants in a process of gentrification. The authors conclude that Ellis Act eviction notices are clustered in San Francisco in both the absolute and relative senses.

本研究调查了 1997 年至 2016 年期间旧金山市的 Ellis Act 驱逐通知集群。Ellis Act 是加利福尼亚州的一项法律，允许房东退出房产租赁业务，将房产从市场和驱逐租户作为无过错驱逐。自 1986 年颁布以来，该法律一直是争论的焦点，因为租户认为它被用来在更广泛的高档化进程中取代低收入租房者。我们用 Ripley 的K和 Kuldorff 的D来解决这个问题使用所有其他驱逐通知作为对照，评估埃利斯法案驱逐的“绝对”和“相对”聚类程度的措施。如果按预期使用该法律，我们应该找到对 Ellis Act 驱逐相对于控制背景在空间上不聚集的零假设的支持。他们确实相对于其他类型集群的替代假设可能会支持这样的论点，即法律在高档化过程中充当租户流离失所的法律机制。作者得出结论，埃利斯法案的驱逐通知在绝对和相对意义上都集中在旧金山。