Whereas most work on residential race relations in US cities is based on the concept of segregation, our approach studies this issue from a who-lives-with-whom perspective. To this end, we study coresidence profiles – percentages of a given racial subpopulation living in different population zones. Population zones are data-driven divisions of a city based on characteristic racial compositions. We used 1990 and 2010 decennial census block-level data for 61 largest US metropolitan areas to calculate coresidence profiles for four major racial subpopulations in each city at both years. Profiles for each race/year combination were clustered into three archetypes. Cities, where given race profiles belong to the same archetype, have similar coresidence patterns with respect to this race. We present the geographic distributions of co-habitation profiles and show how they changed during the 1990–2010 period. Our results revealed that coresidence profiles depend not only on racial preferences but also on the availability of racial groups; cities in the different geographical regions have different coresidence profiles because they have different shares of White, Black, Hispanic, and Asian subpopulations. Temporal changes in coresidence profiles are linked to the increased share of Hispanic and Asian populations.

尽管大多数关于美国城市居住种族关系的工作都是基于种族隔离的概念，但我们的方法是从谁与谁一起生活的角度研究这个问题。为此，我们研究了同居概况——居住在不同人口区域的特定种族亚人口的百分比。人口区是基于特征种族构成的城市数据驱动划分。我们使用 1990 年和 2010 年美国 61 个最大都市区的十年一次人口普查块级数据来计算这两年每个城市四个主要种族亚群的共同居住情况。每个种族/年份组合的概况被分为三个原型。给定种族概况属于同一原型的城市在该种族方面具有相似的同居模式。我们展示了同居概况的地理分布，并展示了它们在 1990-2010 年期间的变化。我们的研究结果表明，同居情况不仅取决于种族偏好，还取决于种族群体的可用性；不同地理区域的城市有不同的核心居住情况，因为它们在白人、黑人、西班牙裔和亚裔亚群中所占的比例不同。同居情况的时间变化与西班牙裔和亚裔人口比例的增加有关。和亚洲亚群。同居情况的时间变化与西班牙裔和亚裔人口比例的增加有关。和亚洲亚群。同居情况的时间变化与西班牙裔和亚裔人口比例的增加有关。