All examples By author By category About

darrenjaworski
Full window Github gist

OK Voting Trends

Voting trends in Oklahoma legislature

This is a two part project:

  1. Retrieve and organize legislator's votes on every bill in a particular session.
  2. Analyze and publish the findings of voting trends amongst the legislators.

This work was originally spawned by @MadiLAlexander. I've gladly contributed.

The overall goal of this work is to illuminate previously unknown trends in the state legislature. There is a wealth of analysis and insight to be gained through this project and further augmentation will be fruitful in more robust and interesting ways.

Part 1: Retrieval

Using the Sunlight Foundation's OpenStates api, we were able to get a full dataset of every legislator, every bill of the current session, and every deciding vote on those bills by the legislators. This data was parsed and organized into csv files allowing manipulation in spreadsheets and availability for visualization and presentation on the web.

All code and datasets are publicly available on GitHub.

Part 2: Analysis

I'm creating a series of visualizations of the data to highlight various dimensions and reveal some trends. To start, I'm creating a matrix of values that represent two legislators voting similarity. This is based on cross reference of each legislator's votes to another legislator's votes. The possible similarity scores range from 0 to 1. 0 representing no voting similarity between two legislators and 1 representing precise complete similarity between the two legislators.

I created a color scale to visually show the difference in the voting similarities. The darker the shade of blue the higher the similarity in votes between the two legislators. Conversely the closer to 0 in similarity score the lighter the shade of blue between the two legislators.

This visualization models the lower chamber voting trends, while another models the upper chamber voting trends. Another page will present the final versions of all visualizations.

This visualization does not analyze the votes based on any qualities of the bills themselves. Ideas I had to add more dimensions to the data are:

I want to find voting trends that aren't obvious to a casual consumer of politics in Oklahoma. Is there a virtual quorum in state government that goes beyond party line? How conservative/liberal are representatives voting trends? How polarized is the state legislature?