Applications

Public Health Data Systems

Public health surveillance systems rely on accurate and complete data to guide intervention and resource allocation. An SDOD lens reveals that missing or misclassified data often emerge from system-level conditions such as inadequate training, unclear accountability, and technological limitations.

In my analysis of COVID-19 mortality data in North Carolina, missing ethnicity data was not random. It reflected gaps in training for those completing death certificates, lack of prioritization for capturing ethnicity in both health systems and death certification, and limitations in reporting systems.

These conditions resulted in the underrepresentation of Latino deaths—demonstrating how data systems can fail to capture the reality they are meant to describe.

Data Equity and Policy

Efforts to advance health equity increasingly recognize that equity must exist within the data itself.

SDOD helps explain why data systems often fall short: categories may be incomplete, overly aggregated, or shaped by historical and institutional biases. As a result, entire communities can become statistically invisible.

Recent policy efforts—such as revisions to federal race and ethnicity standards—reflect a growing recognition that data systems must be redesigned to better represent diverse populations. These changes are not merely technical updates; they are responses to long-standing structural gaps in how data has been collected and categorized.

Government Data Collection

Government data systems—including the Census, vital statistics, and administrative data—form the foundation for policy and resource distribution.

An SDOD perspective highlights how these systems are influenced by factors such as trust in institutions, language access, technological infrastructure, and historical context. These factors can lead to systematic undercounts or misclassification of marginalized populations.

For example, documented undercounts of Black, Latino, and Indigenous populations in the U.S. Census reflect not just participation issues, but deeper structural conditions that shape who is counted and how.

Data Governance: Training, Accountability, and System Design

Data governance determines how data is collected, managed, and validated. From an SDOD perspective, governance must account for the human and institutional context in which data is produced.

Key factors include:

• Training: Whether data collectors understand the importance of demographic accuracy and how to collect it appropriately.
• Accountability: Whether systems track and respond to missing or incomplete data.
• System design: Whether technologies support or hinder accurate data capture.

In many systems, missing data is not actively monitored or addressed, allowing gaps to persist without intervention. Improving data quality requires not just better tools, but stronger governance structures that prioritize completeness and equity.

Advocacy, Activism, and Community Data

Communities have long understood that data is tied to visibility, resources, and power.

Across the country, advocacy efforts have pushed institutions to improve data collection, release disaggregated data, and acknowledge gaps in representation. At the same time, community-led data initiatives are emerging as powerful alternatives—allowing communities to define, collect, and interpret their own data.

These efforts reflect a core principle of SDOD: data systems are not neutral. They are shaped by power, and they can be reshaped through collective action.

A Common Pattern

Understanding these patterns is the first step toward improving data quality.