RIF Impact Schema

Welcome to the open-source framework for understanding and documenting the impact of government Reductions in Force (RIF). Built with love by graduate students at Carnegie Mellon!

What This Is

A collaborative, public-interest project focused on:

Creating a standardized data schema for tracking workforce reductions across agencies
Helping people understand what government capacity and capabilities have been lost
Making it easy to contribute, understand, and use the framework—whether you’re a policymaker, coder, or curious member of the public

Field Definitions

The schema captures the core elements needed to track and analyze the impact of federal RIF (Reduction in Force) events.

Each field in the CSV is defined with: What it means (semantic purpose) What values are allowed (data types and enums) Formatting rules (e.g., dates in YYYY-MM-DD)

See /schema/rif-schema.yaml for the full YAML specification with examples and validation logic.

Data Provenance

To ensure reliability, consistency, and transparency, data included in this schema should come from sources like: Official agency documents (memos, RIF notices, layoff bulletins) FOIA disclosures and public record requests Credible media or watchdog reports Interviews or statements from agency employees, unions, or stakeholder groups

Citation is strongly encouraged. When applicable, include a link or reference in the data_source field.

We encourage users to annotate data_confidence as: High: Officially verified or directly sourced Medium: Well-supported but may require follow-up Low: Anecdotal, inferred, or partially confirmed

Responsible Use

We’re building this in public—with kindness, transparency, and integrity. Please read our Use Standards to understand how we aim to protect both the people represented in the data and those using it.

Because this project deals with real people’s jobs and livelihoods, we care deeply about:

Privacy: Avoid sharing personally identifiable information or exact employee identifiers
Non-weaponization: This dataset should never be used to exploit vulnerabilities or attack communities
Context-first: Data without interpretation can mislead. Use this as a tool for transparency, not blame

If you’re unsure whether your use case aligns with these values, ask us.

Contributors

We’re a small team of researchers focused on public interest tech, transparency, and cybernetics!

Liv Schaefer – project manager + systems thinker
Amy Kang – the glue + research machine
Jaimie An – data storyteller + design queen
David Fuentes – tech wizard