Research has repeatedly found that displacement is not more likely in gentrifying neighborhoods. Since the dependent variable—displacement—is difficult to measure, researchers resort to a variety of proxy measures for it. I classify three types of proxies: a population approach that measures compositional changes in neighborhoods over time, an individual approach that measures individual housing mobility, and a motivational approach that traces both individual mobility as well as the reasons why a household moved to determine whether that move was involuntary. I examine the prevalence of these approaches across a sample of the literature. I then test the commensurability of the proxy measures with data from New York City by comparing the rank orderings of neighborhoods with the most and least displacement. I find widely different results across the approaches. I explain these results by examining the underlying mechanisms of displacement that are masked by the other approaches.

中文翻译：

---

测量位移：评估非自愿住宅流动的代理

研究一再发现，在高档社区中，流离失所的可能性不大。由于因变量——位移——难以测量，研究人员对其采取了各种代理措施。我将代理分为三类：衡量社区随时间变化的人口方法，衡量个人住房流动性的个人方法，以及追踪个人流动性以及家庭搬迁原因以确定是否移动是非自愿的。我检查了这些方法在文献样本中的普遍性。然后，我通过比较位移最大和最小的社区的排名顺序，用纽约市的数据测试代理度量的可公度性。我发现这些方法的结果大不相同。我通过检查被其他方法掩盖的位移的潜在机制来解释这些结果。