Relative Size of Racial Minority Groups and Residential Preferences
How to cite
Ikeuchi, Rio; Kamada, Takuma. 2025. "Relative Size of Racial Minority Groups and Residential Preferences." AEA RCT Registry. AEARCTR-0015360. https://doi.org/10.1257/rct.15360-1.1
Download: BibTeX · RIS — for reference managers without a browser translator. Zotero and most others pick up this page's citation metadata directly.
* Primary investigator
Abstract
This study examines how the residential preferences of White individuals are influenced by information about the racial composition of non-White groups in U.S. counties. While previous research has focused on the share of a single minority group, this study argues that both the absolute size of a non-White group and its relative size in comparison to other non-White groups influence neighborhood choices. To explore this, respondents are randomly assigned to either a control or treatment group, with the treatment group receiving an additional statement indicating that the focal non-White group has become the largest among non-White groups in the county. This research aims to explore the underexamined role of the relative size of minority groups in shaping residential patterns in the U.S. and to offer insights into the mechanisms driving residential segregation, as well as the broader implications for racial dynamics in neighborhood choices.
Intervention
We set the following two treatment conditions as follows:
A) Control: Respondents read a short news article about demographic change in a hypothetical county, indicating the share of the White population and the increased share of a focal non-White group.
B) Treated: Respondents read the same article as in the control condition, but with an additional concluding sentence stating that the focal non-White group has become the largest non-White group in the county.
Details will be discussed in the experimental design section.
Experimental design
In this experiment, we set control and treated groups as follows:
A) Control: Respondents read a short news article about demographic change in a hypothetical county, indicating the share of the White population and the increased share of a focal non-White group.
B) Treated: Respondents read the same article as in the control condition but with an additional concluding sentence stating that the focal non-White group has become the largest non-White group in the county.
We focus on Hispanic, Black, and Asian groups as the focal non-White groups. Respondents are randomly assigned to one of six treatment arms (2 conditions: treatment vs. control × 3 racial groups). Across all interventions, the shares of the White population and the minority population remain fixed, ensuring that variations arise only from the racial identity of the minority group and whether the information about its rank as the largest non-White group is included.
I conduct a between-group analysis to examine how information on the relative size of a minority group affects White individuals’ hypothetical neighborhood choices. The primary focus is to compare the decision-making patterns between the treated and control groups. Prior analysis using U.S. Census data suggests that an increase in a minority group's rank (in terms of size) leads to greater residential segregation from White individuals (e.g., if the Hispanic population becomes the largest minority group in a county, Hispanic-White segregation increases, even when controlling for group size).
At the beginning of the survey, all respondents, regardless of their assigned intervention, answer questions to gather basic sociodemographic information. They are then asked about their county of residence and their perceptions of neighborhood composition. In the experimental section, participants are randomly assigned to one of six conditions (2: treatment vs. control × 3: racial groups), with each respondent reading only one article. Following the article, respondents complete two choice tasks: one involving a racially segregated county and the other a racially integrated county. The order of these tasks is randomized at the individual level.
Design details
Cited by
No citation data available for this registration yet.