Data-Driven Policymaking: Understanding the Needs and Preferences of Disadvantaged Communities

Paper Title

Seeing Like a Community: Public Perceptions of Data Use in Government

Publication Info

• Topic area: Public perceptions of data governance in government policymaking.
• Keywords: data governance, public administration, equity, transparency, marginalized communities, intersectionality, place-based justice, civic technology, participatory data, algorithmic bias.

Background and Problem

• Problem / challenge: The increasing reliance on data-driven policymaking raises concerns about equity, representation, and the inclusion of marginalized communities. Current systems often perpetuate bias, reinforce systemic inequities, and lack transparency.
• Significance: Understanding public attitudes, especially from disadvantaged communities, is critical for designing equitable and trustworthy data governance systems that align with democratic values.
• Motivation and related work: Prior research has explored algorithmic bias, civic technologies, and participatory governance but has not sufficiently examined how marginalized communities interpret and evaluate data-driven policymaking. This paper addresses this gap by focusing on public perceptions through an intersectional and place-based justice lens.

Solution

• Proposed approach: The study integrates intersectionality and place-based justice with HCI theories to analyze public perceptions of data use in governance, using a mixed-methods approach.
• Novelty:
  1. Empirical analysis of intersectional differences in perceptions of data reliability, transparency, and participation.
  2. Centering the lived experiences of disadvantaged communities in designing equitable data infrastructures.
  3. Actionable design recommendations for participatory civic technologies and co-governance models.
• Procedure and key techniques:
  • Conducted a nationally representative survey (N = 754) using Prolific, with demographic and geographic data collection.
  • Analyzed responses using chi-square tests, Structured Topic Modeling (STM), and content analysis.
  • Investigated themes related to public attitudes, concerns, and suggestions for improving data governance.

Results

• Concrete findings:
  • Disadvantaged and low-income groups expressed heightened concerns about data reliability, bias, transparency, and ethical implications.
  • Black respondents were significantly more likely to prioritize justice and equity in data practices compared to White respondents (OR = 7.91, p = 0.01).
  • Disadvantaged communities emphasized the need for accessible platforms, transparency, and evidence of policy impact.
• Advantage over baselines: Provides national-scale insights into how intersectional and place-based disadvantages shape public perceptions, moving beyond localized or conceptual studies.
• Experiments / evaluation:
  • Survey included Likert-scale questions on attitudes and concerns, and open-ended questions on needs and recommendations.
  • Statistical analysis revealed significant associations between demographic factors (e.g., race, income, community status) and perceptions of data governance.
• Limitations and future work:
  • Sample biases due to Prolific's participant pool, which may overrepresent individuals with higher data literacy.
  • Limited exploration of other intersecting frames (e.g., gender and place).
  • Future research should expand intersectional analyses and explore additional identity dimensions.

Summary

This study examines public perceptions of data use in government, focusing on marginalized communities through an intersectional and place-based justice lens. Key findings reveal that disadvantaged and low-income groups express significant concerns about data reliability, bias, and transparency, emphasizing the need for participatory and equitable governance frameworks. The study provides actionable recommendations for designing civic technologies that foster trust, accountability, and inclusivity. By integrating community voices into data governance, policymakers can address systemic inequities and enhance public trust in data-driven policymaking.