Abstract

This study examines the relationship between the Black American population concentration and electoral support for Black candidates in Houston, Texas, from 2013 to 2023. Drawing on theories of descriptive representation and linked fate, the analysis employs segmented polynomial regression models to identify turning points in the nonlinear relationship between the percentage of Black residents in voting precincts and vote shares received by Black candidates. Using precinct-level data from 25 competitive biracial elections across four election cycles for mayor, city council, at-large positions, city controller, and associated runoffs, the findings demonstrate that Black voters consistently support Black candidates at higher rates as Black population concentration increases, confirming patterns of racial bloc voting. However, the polynomial models reveal that beyond certain population thresholds, marginal support begins to plateau or decline slightly, suggesting increasing heterogeneity in voter preferences within predominantly Black precincts. The inclusion of the 2023 election cycle featuring the high-profile mayoral runoff between U.S. Representative Sheila Jackson Lee and State Senator John Whitmire provides a particularly instructive case, as Jackson Lee’s loss despite strong Black voter support underscores the limits of racial solidarity when confronted by a well-funded opponent with broad cross-demographic appeal. These findings contribute to the literature on racial voting behavior by providing empirical evidence of nonlinear dynamics in Black electoral support and by demonstrating the utility of polynomial regression models for identifying inflection points in racially correlated voting patterns.

This work is licensed under a Creative Commons Attribution 4.0 License.