Last updated on October 11, 2018 makes every attempt to quantify our impact. Over the past two years, we have run a number of large-scale controlled experiments in partnership with the Analyst Institute, Pantheon Analytics, and the Harvard Kennedy School of Government.  Read more below.

2017 Research

How do you optimize get-out-the-vote SMS programs?

In the lead-up to the November 2017 gubernatorial election in Virginia, ran a cold SMS get-out-the-vote program targeting unmarried women and people of color. After discovering in 2016 research that SMS is a cost-effective method to get people to vote, we began looking at ways to optimize this method. We tested whether it was more effective to send four election-related SMS messages in the month leading up to elections or to send one text message the week before the election. We also looked at which text messages were most impactful – those that provided polling place information, ballot information, or an encouragement to vote based on social norms.

Our results revealed an important learning — sending election reminder text messages in the week before the election can be as effective and cheaper than sending messages over a month-long period leading up to the elections.

We also learned that polling place location messages alone can increase turnout, while messages with ballot information alongside polling place location information may generate further boosts to voter engagement (at least in lower-salience election environments, like an off-year election, where voters may have less information about items on the ballot).


What happens when you use both mail and SMS to get people to vote?

In the lead up to the December 2017 US Senate special election in Alabama,, and Voter Participation Center (VPC) collaborated with Analyst Institute to test the effectiveness of SMS messages and social pressure mail on getting people to go vote. The test targeted African-American voters to determine which tactic most effectively mobilized the group —  SMS alone, mail alone, or SMS and mail combined. The test also considered two different types of SMS messages – informational messages with logistical voting details and rationale messages that prompted recipients to consider reasons to vote.

We learned that social pressure mail – when combined with SMS or sent alone – is the most effective in getting African-American voters to turn out for a special election. Additionally, informational messages performed better than rationalization messages via SMS. Overall, combined mail and SMS messages were the most effective at driving turnout

2016 Research

Can you use one-to-one SMS to register people to vote?

In 2016, ran the country’s first-ever voter registration drive done via text message. We purchased the phone numbers of just under 1 million unregistered voters, and hired actual humans to send them text messages on at a time encouraging them to registered to vote.  We were genuinely curious if this would work, and entered the experiment with no expectations. We were pleasantly surprised to find that this outreach method not only worked, but that it was also cost effective.  This method proved far more cost-effective than door-to-door registration; slightly less expensive than site-based registration, and almost as affordable as mail-based registration.


Can you use one-to-one text messages to increase voter turnout?

Immediately after running the SMS voter registration drive, we turned our attention to GOTV. This time we purchased the cell phone numbers of registered voters, and hired humans to send them SMS reminders that the election was coming.  We also proactively provided polling place location.  We contacted a total of 3.9 million potential voters in this fashion, including just under 1 million potential voters on Election Day alone.  Of this cohort, 1.2 million were part the experiment group.  The experiment showed that texts providing polling locations increased voter turnout by 0.2 percentage points while plan-making texts were ultimately ineffective. Polling location text messages sent cold to young voters targeted from the voter file are an effective turnout tool, generating an effect on par with that observed in an academic meta-analysis of conventional nonpartisan GOTV mail programs.


Will sending text and email reminders to your supporter list increase voter turnout?

Over 2 million people opted in to email election reminders from in 2016, and another 1.1 million people opted in to SMS reminders.  We were curious if these reminders would have any impact.  Our experiment group included 325,000 potential voters who received aggressive SMS reminders, and 510,000 who received a series of email reminders.  The email reminders seemed to slightly decrease turnout, but the result wasn’t statistically significant.  Voters who received SMS reminders, however, were .65% more likely to vote than those who did not.  This effect was three times as large as the effect of sending unsolicited SMS election reminders to registered voters, although both tactics have a statistically significant positive effect on turnout. A cold contact program, moreover, is easier and cheaper to scale given the low acquisition costs associated with purchasing phone numbers compared to building an opted-in list.


Can you increase voter turnout by allowing electronic signatures on absentee ballot applications?

This one is a bit nerdy, so bear with us.  Only a handful of states let you apply for your absentee ballot online. This means that the vast majority of potential absentee voters need to print and mail their forms. Only home printer ownership is so low that Hewlett Packard stopped publishing statistics on it in 2011, when home printer ownership had fallen into the single digits. Moreover post office branches have been closed and hours have been cut, and physical mailboxes are increasingly hard to find. What this means is that printing, stamping, and mailing a physical form is a clear roadblock to getting your absentee ballot. That’s where electronic signatures come in. Most states will let you fax your absentee ballot application. worked with HelloSign to build a version of our absentee ballot tool that let voters “sign” their forms by taking a photograph of a manual signature.  We then faxed the completed form to the appropriate Local Election Official using the HelloFax API.  For the voter, the process was entirely “online” in that they didn’t need to print the form. For the Local Election Official, nothing changed: they still received a signed paper form, only the form was typed instead of handwritten, and therefore easier to read.

Voters were then given a choice of how they would like to submit their form: either they could print and mail their form themselves, or they could sign electronically, and have submit the form for them via fax. Not surprisingly, younger voters were more likely to choose the electronic signature route.  Shockingly, they then went on to vote at higher rates. We like to think of this as election alchemy: turned low propensity voters into high propensity voters simply by making it easier for them to sign and submit their applications.