Historical Votes (VoteView)

For roll call vote data prior to the 101st Congress (before 1989), this project uses data from VoteView.com, a fantastic resource for historical congressional data maintained by Keith Poole and Howard Rosenthal.

The voteview task downloads, parses, and converts this historical data into the same format used by the modern votes scraper.

Usage

The scripts/voteview.sh script provides a comprehensive example of how to fetch this data. It iterates through all congresses from 1 to 100.

#!/bin/sh
for congress in {1..100}; do
    usc-run voteview --congress=$congress --govtrack $@

    # After the first run, no need to update legislator info.
    export UPDATE_CONGRESS_LEGISLATORS=NO
done

# Also fetch House votes for the 101st Congress, 1989 session
usc-run voteview --govtrack --congress=101 --session=1989 --chamber=h $@

To run for a single historical Congress:

usc-run voteview --congress=50

Options

  • --congress=<number>: (Required) The Congress to fetch data for. There is no default.
  • --chamber=<h|s>: Limit the scrape to either the House (h) or Senate (s).

Data Sources

This task downloads two key file types from VoteView for each chamber and congress:

  • An .ord file containing a matrix of how each member voted on each roll call.
  • A .dtl file containing textual metadata about each roll call vote.

The scraper then correlates this information with the congress-legislators data to associate votes with specific individuals.