Before we move on to 2019, we wanted to take one last look at the 2018 Midterm Election through the lens of Resistbot data. We analyzed usage of the bot’s various voting-related features in the run-up to the election and compared those numbers to election results compiled by the United States Election Project and Harvard Dataverse.

First, the big numbers:

• 583,000 Americans checked their voter registration status using Resistbot
• 241,000 asked the bot to look up their polling location
• 240,000 asked the bot to send them alerts if their voter registration information changed
• Over 77,000 made a pledge to vote
• 12,000 used the bot to register to vote
• Over 125,000 got voter ID information for their states, with over 1,500 asking for help from Spread the Vote
• And lastly, 558,000 told the bot they voted in the election when prompted with a reminder

All together, over 1.1 million Resistbotters either claimed to have voted, or used one of Resistbot’s voting features during the 2018 Midterm Election. If we assume everyone followed through with their vote, we can draw some interesting conclusions.

Voter turnout in this election was historic–the highest for any midterm in a century. According to the United States Election Project, 50.3% of the voting-eligible population (118,532,829 people) turned out to cast a ballot.

As a percentage of voter turnout, Resistbotters tended to concentrate along the West Coast. However, states with rapidly changing political climates–like Colorado, Nevada, and Texas–also had a higher percentage of Resistbotters versus the nation as a whole.

While 0.9% nationally might not seem all that significant, consider this: 1.1 million averages out to roughly 2,500 Resistbotters per House district and dozens of House races were decided by fewer than 10,000 votes!

In fact, in six House races, the candidate’s margin of victory was less than the number of likely Resistbot voters. Below is a chart comparing the two numbers for the 15 closest House races.

Three of these six districts flipped from Republican to Democrat.

• In Utah’s 4th District, Democrat Ben McAdams defeated Republican Mia Love by only 694 votes, compared to roughly 2,300 Resistbotters who likely voted in this district.
• In California’s 21st district, Democrat T.J. Cox defeated Republican David Valadao by only 862 votes, compared to ~1,000 Resistbotters
• In Georgia’s 6th District, Democrat Lucy McBath defeated Republican Karen Handel by 3,264 votes, compared to ~3,700 Resistbotters.

Other interesting races in this chart include a stunning upset in Oklahoma’s 5th District by Democrat Kendra Horn, and the first application of ranked choice voting in a national election, resulting in Democrat Jared Golden’s victory over Republican Bruce Poliquin in Maine’s 2nd District.

If you ever find yourself doubting the effectiveness of get-out-the-vote campaigns, come back to this chart. Just a few hundred votes in either direction could have completely changed the outcome of these races, and in turn, the national picture.

0.9% seems a bit more significant now, doesn’t it?

Now that the election is over, what’s next for Resistbot in 2019?

As the 116th Congress convenes on January 3rd, 2019, Resistbot’s database of elected officials will be updated automatically. Additionally, you will soon be able to message your state legislators and mayors. As always, you can write to your federal officials and state governor by sending the word RESIST to Resistbot on Facebook Messenger, Telegram, or as a Twitter direct message. If none of those work for you, Resistbot also supports old fashioned SMS. Simply text RESIST to 50409 to get started.

From all of of us at Resistbot, we hope you have a wonderful holiday and happy new year! Here’s a few of our favorite tweets from the election:

https://twitter.com/sierra194/status/1055517041434738689