tx_lobbying

Very early alpha

About the data

The two main sources of data are:

1. The lists of registered lobbyists, by year: http://www.ethics.state.tx.us/dfs/loblists.htm

2. And the coversheets for the lobbyist activies reports (LA): http://www.ethics.state.tx.us/dfs/search_LOBBY.html

Names come from both sources of data, but only the coversheets have detailed information about names.

The information for lobbying interests come from the registration forms. This information is all entered by hand and will be very hard to use without scrubbing. There should be a way to de-duplicate lobbying interests, and a way to regenerate a lobbyist's interests based on the raw registration data stored.

The forms can, and are, amended often. When you do updates, it's a good idea to go back a few years.

Getting up and running

Set your environment variables. One way is to have a .env file with:

DJANGO_SETTINGS_MODULE=example_project.settings
DATABASE_URL=postgres:///tx_lobbying
HAYSTACK_URL=http://1270.0.01:9200/

# optional if you want to geocode
# https://geoservices.tamu.edu/Services/Geocode/WebService/
TAMU_API_KEY=5f4dcc3b5aa765d61d8327deb882cf99

and then using autoenv or source .env to activate it.

Install deps:

pip install -r requirements.txt
npm install

Get your database up and running with (Postgres instructions):

createdb tx_lobbying  # if you have a simple Postgres setup
phd createdb  # if you have a complicated DATABASE_URL, pip install postdoc

# inital migration
make resetdb

Then load data with:

make scrape

Running tests

Change template1 to be postgis enabled:

psql template1

CREATE EXTENSION postgis;
CREATE EXTENSION postgis_topology;

make test

Development

pip install -r requirement-dev.txt

The project is configured to use sqlite by default, but if don't want to stab yourself, set your DATABASE_URL to real database like Postgres.

Start the development server:

python example_project/manage.py runserver

How the data is modeled

Diagram:

Details:

Models that correspond directly to a piece of raw data are labelled with raw.

• Lobbyist: the main model, represents a lobbyist
• ExpenseDetailReport: Not that useful yet. raw
• Coversheet: Expense coversheets. raw
• LobbyistStats: A summary of expenses for a year
• RegistrationReport: A lobbyist's registration for a year. raw
• LobbyistAnnum: Represent's a year of a lobbyist's existence in the database
• Compensation: How a lobbyist was compensated by their client. raw
• Interest: A lobbyist's clients, aka an interested party
• InterestStats: A summary of an interest's stats for a year
• Address: Any time we need to store an address

The Future

Features not handled in this version but perhaps the next:

1. Handle corporate mergers/splits/renames
2. Link entities to campaign finance (state and federal)

Additional Data Sources

In addition to the data from the TEC, these services are also used:

• Name de-duplication handled by nomenklatura.
• Geocoding handled by TAMU Geoservices