Updated: Dec 31, 2020
Our mission is to add people-power to the progressive movement. This year, Bluebonnet ran two programs, the Data Fellowship and Tech Fellowship, to recruit, train, and organize tech-savvy volunteers for the 2020 election cycle. Despite the unique challenges that 2020 has presented, we are proud to say that we impacted down-ballot campaigns and progressive causes across the map and brought over 500 new people into the political space. This report provides an early look into our 2020 fellowship and campaign numbers, projects, and general takeaways.
The Data Fellowship was launched immediately following the 2019 election cycle, gaining considerable speed during the 2020 summer. Fellows were recruited and trained to be on volunteer data teams for down-ballot campaigns and progressive partners. 473 fellows were trained and grouped into teams of 3-5 people to be matched with 116 partners over the course of the year. Data fellows provide in-depth data analytics and technical support to their campaign and organizational partners, averaging 5-7 hours per week of volunteer work.
We launched the Tech Fellowship in September 2020 in partnership with Stac Labs, training 112 fellows on four key election technologies (Deck, BallotReady, ActionSprout, and OutVote). Fellows were deployed as volunteers to help state legislative campaigns across seven battleground states, reaching 286 - 368 campaigns. Tech fellows helped campaign partners adopt and implement election technologies and generally assist with voter outreach. All fellows (data and tech) received training on NGP-VAN.
Election Analysis: By The Numbers
*The range denotes the following: the high end is the number of campaigns we assigned fellows to through the Tech Fellowship program, and the low end is the number that can be definitively confirmed as having been helped by tech fellows.
Campaigns Broken Down by Type
Results: You can see how the campaigns did on our interactive Campaign Tracking Dashboard. About 33% of the campaigns across both the Tech and Data fellowships resulted in a win.
In 2020, Bluebonnet worked in 35 states plus Washington, D.C.
Alabama, Arizona*, California, Colorado, Connecticut, Florida*, Georgia*, Iowa, Idaho, Indiana, Kansas, Kentucky, Massachusetts, Maine, Michigan*, Minnesota, Montana, Nebraska, New Jersey, New Mexico, New York, North Carolina*, North Dakota, Ohio, Oklahoma, Oregon, Pennsylvania*, Rhode Island, South Carolina, Tennessee, Texas, Utah, Wisconsin*, West Virginia, Wyoming, and Washington, DC.
* Denotes states that were included in the Tech Fellowship. The Data Fellowship spanned all 35 plus Washington, DC.
Highlighting Fellow Projects
The Ballot chase in Minnesota Senate District 14
The Bluebonnet Data team for Aric Putnam’s Minnesota State Senate campaign used ballot request and ballot return data from the Secretary of State to find voters in the district that had not submitted their absentee ballot. They then merged this information with data from Votebuilder to create ballot chase lists for the campaign. The candidate was down on election night by several hundred votes but won by 300 votes after absentee ballots were counted.
Effective Communication: An Analysis on Postcards in Georgia House District 119
The Bluebonnet Data team for Jonathan Wallace’s campaign for the Georgia State House conducted a controlled experiment of GOTV outreach method effectiveness. They found that volunteer-written postcards were more effective than calls for increasing turnout and that postcards were the only statistically significant method of increasing voter turnout. Furthermore, the team found that postcards were especially effective with black voters, older voters, and female voters.
Building an Automated Tracker for Absentee Ballot Requests in Ohio Congressional District 12
The Bluebonnet Data team for Alaina Shearer’s Ohio Congressional campaign set up an automated reporting of absentee ballot requests in each county in the district so that the campaign could optimize their ballot chase campaign.
Thrutext Sentiment Analyses
Several teams applied sentiment scores to texts received and sent through ThruText. They then determined which texts received the best/worst responses and created word clouds for texts with the best/worst sentiments so that the campaign could quickly diagnose potential issues or positive examples.
Adjusting Polling Data
Teams analyzed raw polling data to determine how the polling provider arrived at their final numbers and provided additional insights into the polling data that the initial service provider didn’t investigate. After this analysis, the teams were available to provide ad-hoc analysis of the polling data if the campaigns had additional questions.
Here are some other ways that fellows helped campaigns (to name a few):
Voter targeting (e.g., building lists in VAN) - 73.6% of data teams did this kind of project
District overview (e.g., mapping census data) - 60.7%
Win number calculation - 42.1%
Early and absentee voter tracking - 32.6%
Software setup - 29.2%
Early Takeaways from the Fellowships
Bluebonnet is influencing how people engage with political campaigns. Before joining the fellowship, 69% had never volunteered for a campaign before; after the fellowship, 90% plan to volunteer for political campaigns in the future.
We’re also impacting career plans. Before the fellowship, only 28% said of fellows said they were somewhat or very likely to pursue a job in the progressive political space or the progressive tech space. After the fellowship, 57% of fellows are somewhat or very likely to pursue a job in the progressive political space or the progressive tech space.
Bluebonnet fellows volunteered an estimated 37,000 hours toward assisting partners and campaigns with data and technical projects in the 2020 election cycle.
Other ways that fellows got involved in the 2020 Election:
Phonebanking for Texas Democrats and Eugene DePasquale’s campaign for PA-10.
Providing tech support for Uptime2020 Election Protection volunteers.
Over 30 fellows supported the Florida Democratic Party’s election night result reporting operation.
We owe a special thank you to the fellows and partners who helped make this year possible. Bluebonnet Data is a people-powered organization, made stronger by the diversity and experiences of its fellows. Fellows mobilized from across the country, from 37 states and Washington, DC, representing over 100 colleges and universities. To get a sense of who some of the fellows are, go to www.bluebonnetdata.org/meet-the-fellows.