Kudos to Derek Willis and Adrian Holovaty of The Washington Post for the Washingtonpost.com site “U.S. Congress Votes Database.”  One element we find of recent and special interest is the “late night votes”
variables for both the House and Senate.  With a little more
probing and data slicing and dicing, it would make an interesting bit
of visual
statistics/infographics to do a longitudinal comparison of the time of
votes in various congresses.




This site/searchable database is a fine example of how investing in some basic data preparation
can create the potential for a ton of stories.  Why, for example, do
Democrats have such a preponderance (18 out of 20) of Representatives on the “missed
votes” list, but only 9 out of 20 on the similar list for the Senate?



This is
also a fine example of how a newspaper can do good things for itself
while doing good things for the community and readers.  This
database gives the WP reporters and editors a quick look-up of
Congressional activity, the kind of fact and detail that can enrich a
story.  At the same time, citizens can turn to this value-added
form of the public record to answer their own questions.



Derek Willis wrote to the news librarians listserv:



“Folks,



It's not part of a story or series, but the Post today launched a site

that may prove useful to your newsrooms or even as an inspiration to

learn Python: a congressional votes database that covers the

102nd-109th congresses (1991-present). Currently browsable, we're

working on adding a search engine and other features to it. Adrian

Holovaty, who works for washingtonpost.com, and I assembled the data

and he built the web framework to display it. All of the data is

gathered using Python, the database backend is PostgreSQL and the web

framework is Django.”