Standard Categories, Shifting Contexts: Reconsidering Racial Analysis in State-Level Policy Research

• In Event: Beyond Heterogeneity: Methodological Challenges and Opportunities Researching Policy at the Nexus of Race and Place

Abstract

Decades of research show that public programs are not racially neutral; instead, racialized narratives of dependency, work, and deservedness shape policy design and implementation. Yet, this research often relies on fixed racial categories, treating African American and Latinx populations as internally uniform and analytically distinct. While these approaches have illuminated important disparities, they may obscure how racial meaning shifts across geographic contexts and within racialized groups. This paper addresses the methodological implications of these practices. We examine how racialized perceptions of deservingness are produced, vary by place, and shape access to the social safety net and ask: How do states construct racialized perceptions of deservingness, and how do these vary across demographics and racial hierarchies?. 

We focus on Able-Bodied Adults Without Dependents (ABAWDs) in the Supplemental Nutrition Assistance Program (SNAP), a group subjected to work requirements and often treated as race-neutral in policy and research despite being disproportionately Black and Latinx. Using panel data from FNS-583 reports, the SNAP Policy Database, an ABAWD waiver dataset, and CDC WONDER, we constructed outcome measures capturing compliance costs, learning costs, and employment support. All models controlled for political, social, and economic factors and included time-fixed effects. We began with a standard approach, using tripartite racial classification (White, Black, and Latine) and 50-state panel analyses to examine racial disparities in administrative burdens in SNAP and employment support through the Employment and Training (E&T) program from 2004 to 2015. 

Models showed that higher White and Latine composition was associated with reduced administrative burdens and stronger employment support. In contrast, higher Black (non-Latine) composition correlated with increased administrative burdens. As our analysis progressed, a limitation emerged: treating “Latine” as a standalone racial category obscured the experiences of Afro-Latines. We revised our approach to distinguish Afro-Latine and non-Black Latine populations, enabling a clearer view of intra-ethnic stratification in state policy implementation. In this disaggregated approach, the Afro-Latine measure no longer held statistical significance. In contrast, the non-Black Latine measure remained positively and significantly associated with favorable outcomes, indicating the non-Black Latine population measure shaped the positive outcomes.   

Disaggregating the Latine category revealed intra-ethnic stratification, but it did not account for how race is constructed through place. To address this, we adopted a place-based analytic strategy. Instead of estimating relationships uniformly across all 50 states, we stratified states into terciles—low, medium, and high—based on the proportion of African American residents from 2004 to 2015. This approach reflects prior research linking racial composition to restrictive welfare policy and allowed us to examine how racialized state contexts mediate the administrative burdens and employment support. Findings demonstrate that the experiences of Afro-Latine populations become more visible through disaggregated approaches, revealing disparities and benefits in relation to non-Black Latine and African Americans that standard classifications obscure. Our analysis underscores the importance of disaggregating and contextualizing race in studying populations like ABAWDs, who are often treated as administratively neutral. We call for place-aware, reflexive approaches that reveal hidden disparities and prevent the reproduction of inequities in policy research.

Authors

• Frania Mendoza Lua, University of Wisconsin-Madison
  Presenting Author

• Jennifer Renee Daniels, University of Wisconsin-Madison
  Non-Presenting Co-Author