Tuesday, November 06, 2012

The technology of election turnout: My experience as a phone bank volunteer

Happy election day, everybody.  Nate Silver shows Barack Obama with a 91.6% chance of winning today, and a projected electoral college of 315-223 votes.  I look forward to the results at the end, so that I can see whether my "not at all close" prediction from early May will give me gloating rights or make me look foolish.  This guy, of course, still believes that Romney is way ahead and has been the whole time.  We'll see, right?

This week I used up some of my remaining Paid Time Off days.  While I was out of the office, I decided to take a break from the obvious regimen of improving at video games, to visit a local phone bank.  I put in six hours on Saturday, and three on Monday.


These events were sponsored by a group called Organizing For America.  They were held in volunteer's houses, and each day 20-30 volunteers would show up, chat about politics, get some free food, and bring their own cell phones, laptops, and chargers to sit around for hours making calls to (in this case) voters in Florida.

I of course love technology and politics, and am delighted to get to see the two working together.  The notion of abstract numbers in a computer database to reach out virtually and produce real, concrete results in another part of the world is exciting to me in a way that counts as both my hobby and career.  Because of this, it was downright heartwarming to see how well the Obama campaign has harnessed the power of crowdsourcing in their Get Out The Vote efforts.

On Saturday I took packets of paper containing 56 names each.  The names were supposed to be from a database of Obama supporters, so there wasn't any need to argue against Mitt Romney much, it was mainly reminding people to get to the polls and telling them how.  Over six hours I went through four packets, which amounts to 224 phone calls.  According to what the data guys were telling me, approximately 1 in 10 of those calls results in an actual phone conversation.  The rest are busy signals, disconnected lines, and people who are simply not home to answer.  (We were instructed not to leave any voice messages.)

Each person on the list had a name, phone number, age, gender, county, and the complete information about their polling location address.  When a live person was on the line, I would read a script that ran something like this:

"Hi, my name is Russell, and I am a volunteer with Organizing For America.  Today is the last day for early voting in Florida.  Florida is a really important swing state that will be very close this year, so I was just calling to see if Barack Obama can count on your vote either early today or on Tuesday?"

I didn't reach any angry Romney supporters.  I did reach a few people who were disgruntled or even angry about "being asked personal questions" or simply being repeatedly subjected to what they considered phone harassment.  On the whole though, a lot of people were fired up about the election.  Many had already voted early, in which case I thanked them and exchanged some pleasant and hopeful words.  Just a few were actually unsure about how to vote and wanted information that I was able to provide (i.e., "Where should I go?"  "What kind of ID do I need to bring?" etc.)  There was one guy who didn't have transportation available and asked if I could help with that.  I said unfortunately I couldn't, being in Texas, but I suggested some alternatives (taxi, asking neighbors).  Any time I talked to somebody who wasn't sure if they could make it, I simply reminded them of the importance of swing states and bumped up their resolve a bit.

After each call I wound up checking a box for the outcome -- no answer, wrong number, don't speak English, or promised to vote, have already voted, promised to vote, did not promise to vote.  After finishing each packet I brought them to a person who was responsible for collating the data and entering it into a computer to submit information about whether the leads were good or not.

On Monday, they had an online program available called a predictive dialer.  I hadn't had it on Saturday since the access has apparently been sporadic.  With the dialer, you enter your phone number in a program and it calls you first.  Then you click a button when you're ready to take calls, and it auto-dials numbers on the list.  It filters out the numbers with no answer, or disconnected tones.  When somebody picks up, you get a beep and have to immediately say "Hello, is so and so there?" (the name pops up on your web browser) or else you risk getting a quick hangup.

This process was a lot more efficient, connecting me with a live person about every minute or so.  But on the other hand, it was also a bit more stressful.  I found that a high proportion of these people had been subject to regular calls and weren't happy about it.  One guy said he's been getting ten calls a day, and it's tiring especially when he already voted.  I wound up apologizing a lot and explaining why the Get Out The Vote effort targets them so much.  Being in the solid red state of Texas, I have not received any election calls myself.  Some were understanding, some hung up pretty pissed off.  Still, there were a fair number of legitimate opportunities to remind people to vote and give them polling information, and some people were genuinely happy that we were doing this.  One guy I called said he was in the middle of calling other states himself.

To wrap up this post, I just want to say that I'm a big believer in crowdsourcing big jobs.  I like wikis, and I like spreading information through Facebook posts and blogs.  Spending some of a long weekend doing election work may not be some people's idea of a good time, but I found it very satisfying.  The other people at the event were friendly and I enjoyed meeting them so we could discuss our shared point of view.

In 2008 I wrote a post about why your vote matters.  I said it's true that your individual vote will not change the outcome of an election, but when you combine the act of voting with encouraging or discouraging other people to vote, that influence can have broad effects of the outcomes of individual states.  Convincing like-minded people to vote with you, or just giving them the extra nudge to get to the polls when it's inconvenient, can multiply your own vote by five or ten or twenty, and if you indirectly influence enough people, then that does have much bigger effects on the practical outcomes of elections.

Furthermore, I think it's great that so many people have caught on to early voting.  One of the biggest concerns in some swing states is that long lines might make it difficult for people to vote when they want to.  I heard a report that some people in Miami Dade county showed up to the polls early in the morning and didn't get out of line until much, much later.  That's a serious logistical problem with elections.  If more people voted early, as Lynnea and I did, then we could get used to the idea of elections that last several weeks instead of one day, and that problem would no longer exist.

Thanks for voting, my friends.

1 comment:

  1. I hate spam phone calls, to the point where if I get too many of them, or even one of them that seems particularly push poll-ish from an identifiable candidate, I won't vote for that candidate.

    I genuinely don't understand why people who otherwise detest annoying marketing, would join a crowd-sourced human "robo"-call campaign unless their family's financial livelihood were directly at stake. That is, if you know you're making people "disgruntled", "angry", and you know your targets consider themselves harassed, and you get hung up on regularly, and if people "weren't happy about" being called this way, and that "it's tiring" to them, and you have to apologize a lot, and others are pretty pissed off ... doesn't that indicate to you that though it may be an efficient and effective means, it's possibly not a respectable or even ethical means of getting your message out there?

    My guess is that apologetics for this kind of phone marketing are analogous to liars for Jesus, who claim the stakes are too high not to use something so efficient and effective, even if it is distasteful.

    My other question is, when the person responsible for collating data submits info about good/bad leads -- is that to target calls better in the future? What kind of response can I give that will possibly reduce the number of future phone calls? If a person calls for Romney, should I respond that I'm a lifetime Democrat, and vice versa?

    Genuinely curious.