Finally, we've launched the first big project I've worked on at my job at washingtonpost.com. It's the U.S. Congress Votes Database, which lets you browse every vote in the U.S. Congress since 1991.
You can browse votes in a variety of ways -- both in aggregate and for individual members of Congress. For example, here's the page for Barack Obama.
You can subscribe to an RSS feed for any senator or representative, to keep tabs on how your elected officials are voting. See the RSS page for more info.
It's got several interesting aggregations for each Congress, such as votes that happened after midnight, vote missers and, on a lighter note, vote totals by astrological sign.
The site is updated daily, depending on whether any new data is available from the House and Senate.
I did the Web development and collaborated closely with Congress expert Derek Willis on the content and data. Of course, the site is powered by Django, the best way to build database-driven Web sites.
The ink is far from dry on this project -- a search engine and several other features are still on the way -- and we're very interested in hearing any ideas on how we can expand it. Post a comment, or drop me a line, if you can think of anything we could add.
Posted by Ryan on December 6, 2005, at 12:31 a.m.:
Awesome stuff! You've probably already thought of this, but it would be nice to be able to see how often congressional members vote with/against the general consensus of their party, and to be able to easily highlight votes where they diverged from the party line.
Posted by Ian Bicking on December 6, 2005, at 2:18 a.m.:
Controlled totals would be interesting. For instance, if I could view votes by region, controlled for party. I was looking at this vote: http://projects.washingtonpost.com/congress/109/house/1/votes/224/ -- it was mostly, but not completely, partisan; region and state views are hard to interpret, but it would be interesting to know if the ammendment had particular regional support, i.e., the Republicans that broke were located in a certain place. I could find out by expanding the Republican views and clicking on each person... actually, even if the link title for the members had information about party and state affiliation (preferably what information is not already indicated by their location in the table) I'd be able to figure it out fairly easily. Other ways of noting exceptional behavior would be interesting too... like on member pages you have the party opinion and the member position; noting when they were out of sync (maybe by color) would be interesting.
Better links to the laws would be interesting too. Linking to the text is obvious, of course. Linking, perhaps, to search results would also be quite interesting. But I don't know what the Post wants to link to. Google News searches seem particularly interesting, since they highlight articles about the aspects of a bill that people actually care about. But, as in my first link, I have no idea where to even look to find out what that ammendment was (well, eventually I did find it on the loc.gov site).
Narrow margin not-completely-partisan votes would be an interesting query too. Or simply votes where many people broke with their party.
Posted by Brian Ray on December 6, 2005, at 5:11 a.m.:
I like the clean URLs. They are "Beauty".
Posted by Joe Murphy on December 6, 2005, at 3:28 p.m.:
Sweet work. I love the level of detail you get into. Here's some more ideas for representing the info ... many of these aren't as "informative" as they are "fun." How is django's skill at outputting graphics? Charts would be fun -- ideas here that would be good as graphic charts (in addition to text view) are marked by a [G].
• Party success rate (which party had the most votes on passed bills / failed bills), viewable by session(s) / month(s) / week(s) / day(s). [G]
• Member success rate. The ratio of yes-votes on passed bills / no-votes on failed bills. Viewable by session(s) / month(s) / week(s) / day(s). [G]
• Astrological sign success rate. Same as member, but with the zodiac instead of people.
• House vs. Senate success rate.
• Session success rate.
Now, this gets into more abstract details that may or may not be interesting ... but, can you categorize votes by term lengths? This is only interesting if there's a pattern, but if first-year senators and representatives tend to vote "yes" more often than the average for the year (or the average for their career), that's worth looking into. Are incumbents more likely to vote "no"?
That's all I have for now, but I'm sure there are more fun ways to play with that information out there :).
Posted by Haidong Ji on December 6, 2005, at 6:05 p.m.:
Wow, this is just awesome. Democracy in action, and I am loving it:) It is soooo important to make these kind of information available easily and coneviently for the general public.
I wonder if it will allow ad-hoc queries in the future. Or, better yet, if the data can be distributed, much like the baseball statistics database
A fellow Chicago blogger, assuming you are still in Chicago;)
Posted by cboone on December 7, 2005, at 2:36 a.m.:
This is very cool Adrian. Thanks.
One suggestion: add a description of what was voted on to the feeds. That'd be super useful.
Posted by Adrian on December 7, 2005, at 5:44 p.m.:
I've added the description to the feeds. Thanks for the idea!
Posted by cboone on December 8, 2005, at midnight:
Awesome. Thanks Adrian!
Posted by Steve Duke on December 9, 2005, at 7:55 p.m.:
Great work. Now can you get the Post to promote its existence? Nothing on the home page, and the promo on the politics page is two page scrolls down. Do they take pride in burying their new, innovative stuff?
Posted by Jay Small on December 9, 2005, at 11:47 p.m.:
Does it have a search feature, or is it browse/drill only? Am I missing it?
Posted by steve case on December 11, 2005, at 2:46 p.m.:
I'd like to find out how the vote went on HIPAA I think was H.R. 3323 signed into law in 2001.
Posted by Scott Westerman on January 15, 2006, at 6:37 a.m.:
Way, way cool Adrian. I consult with a political consultancy who will love this resource.
Since you dig 60s music, you gotta show me how to use django to build a front end to play with my complete list of WKNR Music Guides. Being a 60s music junky, you'll dig it.
Thanks for the inspiration!
Posted by anonymous on June 24, 2006, at 7:49 p.m.:
How do you find the records of all of the Washington State (or any state's) Congresspersons, without knowing their names?
Posted by easygi on August 7, 2006, at 4:06 p.m.:
Please add party loyalty, i.e. party line vote totals and percentages, by session. You give it by vote, I'm looking for totals. It's is the best measure of the kind of compromise that solves problems and should be a primary measure of congressional effectivemenss.
Comments have been turned off for this page.