Spent a few hours last night swimming through the sea of data available on the political process, examining various sources. It appears a lot of places use the source data at GovTrack.us. They’ve done a great job of maintaining a huge database of congressional data (including historical data!), and it’s mostly in XML and maintained BY HAND (quite a task). Sunlight’s Drumbone API uses this data, and likely, so will I. I also found another fantastic source of bulk data at Transparency Data. This means I potentially will not have to rely on an API at all.
I’ve also realized that apart from financial contributions or earmark costs I’m having trouble finding interesting qualitative data (most is nominal). I might have to integrate some data from other sources. For partisan / ideology scores, I will probably look to the Cook PVI (Partisan Voting Index) or the work done at VoteView. I’ll be keeping my eye out for more in the coming weeks. If I implement the system properly, adding data sources late in the game should not cause too many problems.