Abstract: When Aggregation Obscures Need: Subgroup Disparities In Reported Hardship Among Asian Public Benefit Recipients (Society for Social Work and Research 31st Annual Conference)

Schedule:

Sunday, January 24, 2027

Union Station 15 (Hilton San Francisco Union Square)

* noted as presenting author

Margaret Lee, MSW, Doctoral Student, University of California, Berkeley, Berkeley, CA

Julian Chun-Chung Chow, PhD, Hutto-Patterson Charitable Foundation Professor, University of California, Berkeley, Berkeley, CA

Christine Lou, PhD, Head of Research and Evaluation, San Francisco Human Service Agency, San Francisco, CA

Cheng Ren, PhD, Assistant Professor, State University of New York at Albany, NY

Leyi Zhou, MSW, Doctoral Student, University of California, Berkeley, CA

Background and Purpose: Racial or ethnic aggregation in administrative and research data obscures important variations in need within Asian populations and can distort how need is represented in public benefits systems. This limitation is particularly consequential in public benefits systems, where resource distribution is often based on high level indicators that can mask subgroup differences. Using data from the COVID-19 pandemic, this study examines whether aggregated categories misrepresent patterns of need among Asian public benefit recipients. It focuses on subgroup differences in reported needs, while also assessing household crowding as a condition shaping within-group variation.

Methods: Data from the May 2020 San Francisco Human Services Agency COVID-19 Client Needs Survey (N = 1,228 Asian respondents receiving public benefits) was utilized for analysis. The primary outcome is the total number of self-reported needs across ten areas (e.g., food, housing, employment, legal assistance). Key predictors include Asian subgroup (Chinese [reference], Native Hawaiian/Pacific Islander [NH/PI], and Other Asian, reflecting available administrative categories) and household crowding (people-to-room ratio). Covariates include education (college degree vs. no degree) and employment status. Multiple linear regression with robust standard errors was used to estimate differences in total needs, and logistic regression models were used to estimate the odds of reporting poor or very poor financial health.

Results: Substantial subgroup disparities in reported need were demonstrated. Controlling for covariates, NH/PI respondents reported substantially higher levels of need than Chinese respondents (b = 0.61, p < .001), with Other Asian respondents also reporting higher needs (b = 0.26, p = .004). These differences would otherwise be obscured under aggregated analyses. Household crowding was also associated with increased needs (b = 0.15, p < .001), indicating that structural living conditions contribute to within-group variation in hardship. In logistic models, crowding significantly increased the odds of poor financial health (OR = 1.19, p = .013).

Implications: Aggregated racial categories can misrepresent need, shaping how populations are recognized and resourced within public benefits systems. Subgroup disparities, particularly among NH/PI respondents, would remain undetected in aggregated analyses, contributing to the systematic under-recognition of need among already marginalized populations. While household crowding is associated with greater need and financial insecurity, its primary implication is to highlight how structural conditions shape variation within subgroups, suggesting the value of incorporating household-level indicators into screening and service design. More broadly, disaggregated and structurally informed data practices are critical for ensuring that social welfare systems are more responsive, equitable, and effective for the communities they are intended to serve.