-
September 13, 2007 4:50 PM
Getting Traction from Mobile Action
Most folks reading this blog know that Working Assets has great phones, and also that we run some great political programs. But I'll wager that fewer people know about the ways in which we use our phones and mobile phone know-how as a part of our political work. One example is some work we did with text messaging in the 2006 elections.
When people registered to vote using our online voter registration site, we asked them if they wanted to receive text message reminders about the election on their mobile phones. We had the suspicion that sending a text message would be a good way to remind people it was Election Day, but we really had no actual evidence to back it up. So we teamed up with some researchers from Princeton and the University of Michigan, as well as the New Voters Project -- also a Working Assets donations recipient for this year -- to design an experiment to find out if text messaging actually helped get people to the polls.
The results that we got this week were very encouraging: we found that people who received a text message were about 4.2 percentage points more likely to vote than those who did not receive a message. It was also much more cost effective than other forms of get-out-the-vote outreach. There were plenty of other interesting findings as well, and those who are interested can check out the full report.
It caught the eyes of some of the news media as well. Both the New York Times and Chicago Tribune political blogs picked it up.
We're excited to continue our text messaging reminder program as we ramp up for 2008. Does your mobile phone company make sure you remember it's Election Day? If it doesn't, here's how you can sign up with one that does.
-
Recent CREDO Blog entries
-
This Modern World
-
April 29, 2008 -
April 22, 2008 -
April 18, 2008 -
April 10, 2008 -
April 3, 2008
-
-
Donations Recipients
-
RSS feeds
Discussion
Have you seen that article about those Russian Journalists who cook eggs with the micro waves of cell phones yet..?
Keep up the good work, and I'll take mine over easy..?
I'd be scared, too, if it were true. But the recipe for cell phone scramble is just an internet hoax.
I am surprised by the small +4.2% turnout difference between the control group that got no message and the treatment group that did. This must have been a poltically active group. Few needed the reminder.
Candidate recorded landline robo calls cost less than 10 cents per call. Google "political marketing".
The authors say "the cost per vote was $1.56." What does that mean?
A 4.2 percentage point lift is actually quite significant when it comes to voter mobilization tactics, as measured in field experiments. You can see a chart that summarizes the different effects here. The only tactic more effective in terms of mobilization is door-to-door canvassing, but the costs there are much higher.
Similar experimental studies have shown that robocalls have no statistically significant mobilizing effect. In some cases, the research suggests that they might slightly depress turnout.
The $1.56 figure is the cost per additional vote generated. This means that if you look at the number of additional voters in the treatment group versus the control group -- which we can attribute to the effect of the text message since they were assigned randomly to the groups -- and divide that number into the total cost of sending all of the text messages, we find what it costs to turn out each of those additional voters.
Join the Discussion