johnpfay/WakeVoter

WakeVoter

Analysis of NC census and elections data to identify areas of low minority voter turnout.

Environment

• Python 3.6.6
• Jupyter | Geopandas | GeoPy | requests | spyder | folium

Local Installation

• Clone the repo $ git clone https://github.com/14-cities/WakeVoter.git
• Install requirements with $ pip install -r reqs.txt (NOTE: it is heavily encouraged to do this in a virtual environment)
• Request an (API key)[https://api.census.gov/data/key_signup.html] from the US Census
• When the key arrives in an email, copy and paste it into a file called APIkey.txt at the root of WakeVoter
• To run the applicaiton, start a new Python shell session and import the VoteSqaud.py file
  • $ python
  • >>> import VoteSquad.py

OS Specific troubleshooting

• On Mac OS X, if you get the error OSError: Could not find libspatialindex_c library file you may need to run $ brew install spatialindex to get rtree working properly. See this GitHub issue

• On Ubuntu 18.04, if you get the missing libspatial error, $ apt install libspatialindex-c4v5 libspatialindex-dev libspatialindex4v5 or just apt install python3-rtree and it will pull those in as dependencies.

Data

Voting registration and history data

• https://www.ncsbe.gov/data-stats/other-election-related-data
  • Voter Registration Data By County >
    • Statewide Voter Registration
    • Statewide Voter History

Voting precinct shape files

• https://dl.ncsbe.gov/index.html?prefix=PrecinctMaps/
  • SBE_PRECINCTS_20190507.zip

2010 Census Blocks

• Block features with population & housing data
• Demographic data
  • https://factfinder.census.gov/faces/nav/jsf/pages/searchresults.xhtml?refresh=t
  • Geographies:
    • All geographic types
    • Geographic type: .... .... .... Block - 100
    • Geographic vintage: 2010
    • State: North Carolina
    • County: Wake
    • Census tract:
    • Select one ore more
  • Topics:
    • P1: Race
    • P10: Race for population 18 and over

Address points

• https://s3.amazonaws.com/dl.ncsbe.gov/ShapeFiles/list.html
• https://dl.ncsbe.gov/index.html?prefix=ShapeFiles/
• https://s3.amazonaws.com/dl.ncsbe.gov/ShapeFiles/address_points_sboe.zip

---

Workflows

1. Extract and organize voting data for [Wake] county

• Pull the voter data from the NC SBE servers and subset for the selected county:

  • Manually download the Statewide Voter Registration and Statewide Voter History data sets to a local folder.
    • Will be replaces with code to download files from NCSBE servers...

• Run get_county_voter_registration_file to retrieve/build a county level registration csv file.

  • Run get_county_voter_history_file to retrieve/build a county level registration csv file from the state file.
  • Run get_address_data() to pull NC Address Points data file to a local folder.

• Assemble a feature class of voting points for the county

  • Run get_voter_data() to select voting records for a specific county and assemble them into a shapefile.
    • This uses the address file to attach coordinates to voter registration data

1. Identify census block groups with > 50% black voters

• Obtain block feature dataset and attribute dataset, subset for Wake Co.
• Join attributes to features and isolate block features with > 50 black voters
• Tag block features with voting precincts

2. Compute voting frequency data and geocode it

• Extract voting registration and history data for Wake Co.
• Geocode voter registration data
• Tally the number of elections voted in for each voter registration & join to geocode data

3. Select voting data falling within selected census blocks

• Spatially join voting registration data with census blocks & precinct information
• Compile list of super voters within each block (voters in all elections)
• Compile list of low propensity voters within each block
• Identify contiguous blocks, within precincts, and tally low propensity voters
• Assign super voters for each 100 low propensity voters in contiguous blocks.

---