I gave this 90-minute webinar for the Drug Policy Alliance’s Department of Research and Academic Engagement. The premise was simple: journalists use FOIA laws constantly, but academic researchers barely touch them. That’s a missed opportunity, especially for anyone studying drug policy, criminal justice, or public health — fields where government agencies hold enormous amounts of relevant data that never makes it into public datasets.
The session walked through the full arc of using public records in a research workflow. I started with the legal foundations — how federal FOIA and state-level equivalents work, what records you can actually get, and which exemptions agencies tend to claim. From there I moved into the practical side: how to write requests that don’t get ignored, how to negotiate with FOIA officers over scope and fees, and how to appeal denials.
The second half focused on what happens after you get the records. Government data arrives in all kinds of shapes — PDFs, scanned documents, inconsistent spreadsheets, sometimes boxes of paper. I covered data science workflows for cleaning and processing that material using R, Python, OpenRefine, and Tabula. I also spent time on reproducibility: version-controlling your processing scripts, documenting your transformations, and building pipelines that someone else could run and get the same results. That matters both for academic rigor and for defending your work if an agency or reviewer challenges your findings.
I drew on my experience building OPRAmachine, which gave me a front-row seat to how agencies actually handle records requests at scale across New Jersey. That project generated a lot of the practical knowledge I shared about what works and what doesn’t when dealing with government records offices.
For the drug policy audience specifically, I discussed the kinds of data that FOIA can unlock — enforcement records, treatment program data, regulatory compliance files — and how researchers can use that material to study questions about disparities in enforcement, access to treatment, and the gap between policy as written and policy as practiced.
The webinar included live walkthroughs of sample requests and responses, code demos for common data processing tasks, and a Q&A session where participants asked about their own research scenarios.
The full presentation slides are available at foi-webinar.gavinrozzi.com. The complete recording is embedded above.
About the Author
