Counting Asian Americans: Mandates, Compliance, and the Limits of Data Disaggregation In New York City — a Policy Surveillance Study

Background and Purpose: Asian American residents of New York City are a growing population, yet ethnic subgroups remain statistically invisible in public agency data. Federal, state, and municipal mandates require reporting by named subgroup rather than broad racial aggregate. Mandate passage does not automatically produce data availability. National surveillance data indicate that fewer than 40% of COVID-19 cases had complete ascertainment under OMB-compliant race/ethnicity criteria (Erickson et al., 2023), suggesting that reporting requirements alone may not ensure complete subgroup visibility in practice. In a needs assessment of South Asian older adults in NYC, 9.4% of participants showed symptoms of mild or moderate depression, compared with 16% of Caribbean-origin participants and 8% of Indian participants (Ali, 2021). These disparities are not incidental. They reflect administrative systems that do not consistently name specific populations as required (Yi et al., 2022; Young & Cho, 2021). This study asks whether agency documents governing NYC residents reflect the subgroup categories specified across overlapping mandates, and whether patterns of naming vary across document type, agency sector, and jurisdiction.

Methods: This study applies policy surveillance methodology within a legal epidemiology framework (Burris et al., 2016; Wagenaar et al., 2023) to examine statutory and administrative documents produced under six overlapping federal, state, and city instruments governing NYC residents. The corpus includes four layers of policy development and implementation—deliberative, statutory, implementation, and advocacy—across four anchor agencies: NYS Office of Mental Health, NYS OASAS, NYC DOHMH, and NYC HRA. A three-code scheme identifies aggregate terminology (Code0), named statutory subgroup terminology (Code1), and unlisted population terminology, including Indo-Caribbean and Indo-Guyanese references (Code2). Documents are coded in NVivo. Quantitative analysis uses frequency distributions, chi-square tests, and Cramér’s V to assess variation across document layer, agency sector, jurisdiction, and exemption status. Inter-coder reliability is assessed using Cohen’s kappa.

Results: Analysis of the statutory layer across all six instruments shows that four rely on aggregate or illustrative language only, while two mandate named subgroup reporting. Notably, Bhutanese Americans are absent from the statutory instruments examined. Indo-Caribbean and Indo-Guyanese populations do not appear as named required subgroups. Local Law 166 of 2025 creates a write-in pathway for unlisted population visibility, but does not enumerate Indo-Caribbean populations by name. Review of the implementation layer suggests that aggregate terminology remains common in agency documents even where overlapping mandates call for specific subgroup naming.

Conclusions and Implications: These findings document two distinct problems. The first is an implementation gap: statutes may specify subgroup categories not fully reproduced in agency documents. The second is a categorical gap: some populations remain outside the named subgroup framework. No amount of improved implementation addresses a categorical gap. The June 30, 2027 compliance date for the NYS OMH provides a future comparison point for assessing whether statutory change is associated with more specific subgroup naming in administrative documents. For social welfare systems, these distinctions matter because subgroup visibility shapes how inequities are recognized, how populations are described in public-facing materials, and how needs are framed for service planning and resource allocation.