Legisletters: A Hub for Congressional Correspondence

By Daniel Schuman, cross-posted from Congressional Data Coalition
This week GovLab beta launched a new tool, Legisletters, which automatically gathers congressional correspondence with agencies and publishes it in a searchable, user-friendly interface. The project is the brainchild of Travis Moore and Andrew Miller, with some advice from me, and was magnificently developed by John Krauss at GovLab.
Members of Congress have a hard time tracking their correspondence with federal agencies, in part because of staff turnover and the absence of an inexpensive, easy-to-use tool. It is very hard for an office can be aware of the letters that other offices send. Frequent staff turnover means current staff often have no idea of what was sent in the past.
Fortunately, since members of Congress often publish their correspondence on their websites–often in the less-than-helpful PDF format–it is possible to reconstruct some of the communications.
Legisletters searches House and Senate websites for documents that have the indicia of letters to agencies and pulls them into its database. Once captured, it determines who sent the letter, when it was sent, and allows a full-text search. Using faceted searching, users can fairly easily dig through the database. (Some additional refinements would be welcome.) All the underlying code is open source and published on GitHub.
Legisletters can help address several problems. It can serve as:

• An archive of correspondence by individual members of Congress to agencies, which is very useful for current staff and historians alike.
• A finding aid for other offices interested in partnering on issues, perhaps incorporated into a tool like the nascent “coalition builder.”
• A data source for an alerting tool, like Scout, so journalists and advocates can keep an eye on what a particular office is doing.

In addition, the underlying technology can be repurposed to gather other documents published on the web, such as CRS reports.
Legisletters is still in being developed. Its creators are exploring the following ideas:

• Engaging the offices of congresspeople to help them publish their documents in open formats.
• Asking users to improve the quality of existing letters by allowing them to classify different sections and fix errors.
• Permitting the addition of letters that have not been automatically discovered, both by link and upload.

They are are asking for feedback here. Let them know what you think.