Legislative Graph

A set of scripts to easily download and import US legislative data into a Neo4j database. This is a work-in-progress, please submit an issue for any errors or feature requests.

Data Model

The data model incorporates a small amount of the data available from GovTrack. Please submit an issue to request any changes / updates. We're really interested in how the commmunity might want to use this data so please let us know!

Also, this file has more detailed information about the data model.

Quickstart

A read-only version of legis-graph with data from the 114th Congress is running in a Neo4j instance at http://52.70.212.93/browser/ Feel free to make use of this instance to explore legis-graph.

For convenience, data from the 114th Congress is available as a Neo4j datastore. Simply download this zip archive, extract and replace the data/graph.db folder in an existing version of Neo4j.

Alternatively, you can use the LazyWebCypher tool with this link. )

Load Data

We're currently working to streamline the data loading process, but for now you can follow these steps to load data.

Download / update data

Sync a particular congress by its number (so, for instance, for the 112th congress, replace <num> with 112.

./sync.sh <num>

Parse data

Use the parse scripts to parse the raw data into CSV files that can be easily loaded into Neo4j.

$ python3 parse_legislators.py
...
$ python3 parse_bills.py
...
$ python3 parse_votes.py
...
$ python3 parse_committees.py
...
$ python3 parse_committee_members.py
...

The scripts require Python 3.

Insert into Neo4j

Use the *.cql scripts to load the data into Neo4j. You need to have the data files available over HTTP, so before you do this step, in another terminal window, just run $ python -m http.server in the repository root.

Be sure Neo4j is running, by default legis-graph assumes Neo4j is running locally.

$ path/to/neo4j/bin/neo4j-shell < load_legislators.cql
...
$ path/to/neo4j/bin/neo4j-shell < load_bills.cql
...
$ path/to/neo4j/bin/neo4j-shell < load_subjects.cql
...
$ path/to/neo4j/bin/neo4j-shell < load_congresses.cql
...
$ path/to/neo4j/bin/neo4j-shell < load_bills_congresses.cql
...
$ path/to/neo4j/bin/neo4j-shell < load_bills_subjects.cql
...
$ path/to/neo4j/bin/neo4j-shell < load_bills_legislators.cql
...
$ path/to/neo4j/bin/neo4j-shell < load_votes.cql
...
$ path/to/neo4j/bin/neo4j-shell < load_committees.cql
...
$ path/to/neo4j/bin/neo4j-shell < load_committee_members.cql
...

Queries

Once the data is loaded in Neo4j we can use queries written in Cypher to discover interesting things about Congress.

General Queries

Find all Legislators:

MATCH (n:Legislator) RETURN n LIMIT 100

Find Steve Daines:

MATCH (n:Legislator {firstName: "Steve", lastName: "Daines"}) RETURN n

What Bills did Steve Daines sponsor?

MATCH (n:Legislator {firstName: "Steve", lastName: "Daines"})<-[:SPONSORED_BY]-(b:Bill) RETURN b

For how many Bills did Steve Daines vote Yea?

MATCH (n:Legislator {firstName: "Steve", lastName: "Daines"})-[v:VOTED_ON]->(b:Bill) 
WHERE v.vote = "Yea"
RETURN b

More advanced queries

Find the number of bills proposed during each congress in the database.

MATCH (c:Congress)<-[:PROPOSED_DURING]-(b:Bill)
RETURN c.number AS congress, count(b) as numProposed

Find the number of bills enacted in each congress in the database and the average number of sponsors bills had during that congress.

MATCH (c:Congress)<-[:PROPOSED_DURING]-(b:Bill)-[:SPONSORED_BY]->(l:Legislator)
WHERE b.enacted = 'True'
WITH c, b, count(l) AS numSponsors
RETURN c.number AS congress, count(b) AS numPassed, avg(numSponsors) AS avgSponsors

Articles

• Introducing legis-graph - US Congressional Data With Govtrack and Neo4j
• Congressional PageRank - Analyzing US Congress With Neo4j and Apache Spark

Authors

• William Lyon
• George Lesica

Terms

The software in this repository is provided "AS-IS" without warranties or guarantees of any kind. Data used by this software is provided by Govtrack.us and should be used under the terms specified by Govtrack.us here.