DLP Dispatch #17
Hello, and welcome to the 17th edition of the Data Liberation Project’s newsletter. Inside: New data on boating accidents, new documents on oil spill response costs, a new way to access federal inmate complaint data, a new way to search across Animal Welfare Act citations, and our latest FOIA requests.
New Dataset: Recreational Boating Accidents
The US Coast Guard maintains the Boating Accident Report Database (BARD), which serves as a central, national database of recreational boating accidents. The data come from submissions of CG-3865, a form that boat operators must file to state authorities when they're involved in an incident that results in a death, disappearance, serious injury, or substantial damage.
The reports (and database) include details about each accident, vessel, death, and injury involved.
In January 2023, we filed a FOIA request seeking a full copy of the database, minus fields containing personally identifiable information. The request also sought all relevant database documentation. After a discussion with the Coast Guard’s FOIA office, we narrowed our request to the years since 2009, and also agreed to wait for the release until the 2023 records became available.
On July 12, 2024, the Coast Guard provided the following records, now fully available to the public: three Microsoft Access (.accdb) database files, each covering a distinct time period (2009–2013, 2014–2022, and 2023); a database dictionary; and a “notes” file.
To increase the accessibility of the records, the Data Liberation Project has also converted these files into three additional formats: SQLite, CSV, and Parquet. Other than the format conversion, these files do not modify the information in the records.
In total, the records describe 58,430 boating accidents, as well as 78,316 vessels, 8,935 deaths, and 36,773 injuries connected to those incidents.
To get started with the data, please consult the Data Liberation Project’s introductory documentation.
Many thanks to DLP volunteers Declan Bradley, pdtd, and Kaitlyn Radde for their contributions to the documentation.
New Documents: Oil Spill Response Costs
When oil spills occur in the US and its waters, the US Coast Guard’s National Pollution Funds Center (NPFC) “provides funding for quick response, compensates claimants for cleanup costs and damages, and takes action to recover costs from responsible parties.”
In November 2022, the Data Liberation Project filed a FOIA request to the Coast Guard sought all records collected through several types of NPFC forms and workbooks since January 2000. Through conversations with the Coast Guard, the Data Liberation Project agreed to narrow the request to more recent years — those for which the records are primarily digital.
The Coast Guard has begun providing the responsive records in 150-page increments. The first tranche of records was provided on July 30, 2024. To access the documents and the Data Liberation Project’s descriptions of them, click here.
New Resource: Federal Inmate Complaint Data, De-Coded
One of the challenges of working with the federal inmate complaint data we released last month: Many of the fields use shortened codes to represent prison facilities, complaint subject categories, and filing statuses.
DLP volunteer Declan Bradley has helpfully constructed a version of the dataset that expands those codes into their fuller names and descriptions (e.g., FORT DIX FCI instead of FTD, and CORRESPONDENCE - REJECTION OR CONFISCATION instead of subject 16B), and has paired it with location data for the majority of facilities.
New Resource: Semantic Search for Animal Welfare Act Citations
Since February 2023, the Data Liberation Project — in collaboration with Big Local News — has been publishing a daily-updated dataset of all Animal Welfare Act inspection reports released by APHIS, a branch of the USDA. The dataset includes information about each citation within each report, along with the citation’s statute code, general subject, and narrative findings.
Rowan Copley, Saul Pwanson, and Geoff Corvera have built a new tool that lets you conduct semantic searches across those narratives. That is, you can search by the meaning of a word or phrase, not just via exact matches.
I’ve tested a few phrases (e.g., overdose due to technician error) and found the results — which also locate the relevant facilities on an interactive map — to be useful. Try it here.
New FOIAs
Since the latest dispatch, we’ve filed two new requests. They seek:
- Data on all federal inmate deaths since 2005, in collaboration with Jake J. Smith.
- An updated copy of the Department of Defense’s Excess Defense Articles transfer database, co-requested with Alec Gitelman.
That’s all for now! Thank you for reading, and don’t hesitate to reply.
— Jeremy