Novel field operations ideas?

by: Adam Bink

Tue Dec 08, 2009 at 16:15


I am sure you guys are sick to death of post-mortem pieces on Maine from me, but I promise just this one more, since I think there are some implications for all campaigns re field operations.

Jasmine Beach-Ferrara, who volunteered in Maine for the No On 1 campaign, has a piece in The Democratic Strategist about the field campaign in Maine. You can read the entire thing here, but there are a few nits I want to pick and two interesting ideas in it, particularly if you're a field geek.

Adam Bink :: Novel field operations ideas?
  • Jasmine, who talked to a straight couple at the door about their feelings on Question 1, writes:

    Walking away, I rated her as a "three," or swing voter, and him as a "four," or likely to vote yes. According to a literal interpretation of the campaign playbook, this conversation had actually been a waste of time in every regard except one: the campaign now knew not to spend time and resources doing further outreach to this couple. At this point in the campaign cycle, an exacting calculus kicks in and attention shifts to turning out identified supporters, "one's" and "two's" on the scale. All other voters are lumped together and categorized as unwinnable. For the next five weeks, this couple and voters like them would not hear directly from the campaign, except in TV ads. This is considered smart organizing, and typically it is.

    To my knowledge, campaigns- including No On 1- don't ignore threes. Threes, by definition, are considered prime undecided voters, are still considered movable and worth spending resources on.

  • On the potential to persuade swing voters, she writes:

    We are losing because we are not persuading swing voters, yes. But that does not mean they cannot be persuaded. People change their views on this issue all the time, as family members and friends respond to a LGBT person's coming out, or as church congregations vote to become open and affirming. Sometimes it takes years and sometimes it takes weeks. Rarely does it happen without a mixture of love, pain and patience. In these more intimate contexts, we call it transformation rather than "persuasion" and it doesn't happen due to canvassing and phonebanking. It happens when the truths of someone's life transcend the doctrine they believe in. Some parts of this process we can map and some remain mysterious to us. It is, after all, the work of the heart we are talking about here.

    I agree with her that sometimes these life events change people's minds. But what's the point here? In a political campaign context, it doesn't seem wise to eternally hold out hope that that will happen and thus keep spending resources on people who tell you straight-up they aren't voting your way. In some contexts, like Washington State, the campaign literally lasts a matter of weeks. When you are perpetually underfunded and without enough staff/volunteers, you just can't try relentlessly to persuade swing voters in the hopes that their gay co-worker or daughter or minister will come out in the midst of your efforts.

  • Jasmine has an interesting proposal regarding changing the way campaigns talk with swing voters- essentially arguing that instead of volunteers, the campaign will recruit special outreach people, particularly ministers who live in-state, and train them in empathic listening skills and engaging with people around faith. The campaign will then assign them a universe of swing voters to track through election day. They will engage them in door-to-door, phone and try and develop a personal rapport, not with scripts, but with intensive personal conversations.

    While this is certainly a good idea, my question is to what extent campaigns already implement it. I am not someone who is a veteran of many field operations in campaigns, but from those I have been on, campaigns to some extent, already utilize people of faith, union members, African-Americans- whatever segment you need to talk to voters, and assign them certain precincts to walk, events to attend, and so forth. What's novel about Jasmine's proposal is the training is the emphasis on special volunteer training and personal rapport development. That's different from assigning an African-American to canvass African-American precincts with just a walk script. I don't know how much time a campaign has to implement this kind of outreach, though, and I don't know what will be different in terms of voter contact. If a specially trained minister volunteer approaches a religious voter and discovers him to be a four or a five, should the minister then proceed as a normal campaign would and decrease or stop the amount of outreach, even after a personal, empathic conversation? Or continue in an effort to "develop personal rapport" with the voter?

  • Encouraging abstention. If a voter replies that they are against your position, ask if they would consider abstaining from that particular election or that column on the ballot. Jasmine's logic is that one less opposition vote is almost as good as vote for your side. And you remind voters that legislators already abstain, so why can't they? This is one idea that's particularly worth exploration.

If anyone has thoughts on these experimentations in field, or if they've already been implemented in your experience, please do share in the comments.


Tags: , , , , , , (All Tags)
Print Friendly View Send As Email
Based on experience (4.00 / 1)
mostly in New Hampshire, Iowa and Vermont and less so in Florida:

A fundemental problem is that we do not do a good job training Canvassers.  This means that the intelligence gained from them is often close to zero.

I have never seen more money spent on field operations than was spent in Iowa last cycle.  And yet two of the three major campaigns COMPLELELY missed the size of the turnout.  In my precinct in West Des Moines I had a list of potential voters that I was supposed to check, and then call if they didn't show.

There were about 20 names on the list, about half of whom wound up supporting other people.  But I had 70 total supporters.  This means despite a year of organization and millions of dollars we had failed to identify over 60% of the actual voters.

The reason was that the campaign was sort of a ponzi scheme.  People were identified as 1's or 4's.  If they were 4's in the initial contact, my guess is they weren't canvassed again, or when they were it was assumed that they still did not support our candidate.  Meanwhile 1's were contacted over and over again.  I talked to someone who had someone knock on their door twice in 24 hours and then received 3 phone calls.  I know she was canvassed the next day.  

If more were done to explain how the entire lifeblood of the campaign was based on the intelligence gathered during cavassing I think you would see much better discpline in the initial canvass.  


One thing I felt was that a number of people (0.00 / 0)
were slightly taken aback to find a big probably-gay guy on their doorstep, asking how they were gonna vote.

I don't think people are really persuadable, particularly not on their doorstep. I do think that framing the issue as a matter of equal rights instead of gay rights makes a great deal of sense. Or even of religious freedom. I'd love to see a ballot initiative that said something like, 'Every religious institution shall have the freedom to decide if they recognize same-sex marriage or not; the government shall not dictate matters of faith to churches.'


field (0.00 / 0)
I know the topic is post mortem, but have you listed the thoughts of what you would do if you had more time? Also, Maine seems like as good a prospect for long term, inexpensive, messaging as anyplace. What would it take for continuous persuasive outreach, say $10,000 a month buying radio one month, cable the next, and small town alternatives the third and so forth? Or is there no possibility of long term persuasion, only waiting for the old farts to die?
The down side of what you are describing is it picks on volunteers who are doing something they don't usually do, and doing it to the best of their understanding. In the press of those few weeks, who wants to criticize? It is not a good feeling to receive the three calls if you have been identified as a 1, nor is it good to make the three calls on your list of people to turn out. So, I understand the idea of professionals to replace the volunteers, but I sense two problems. One campaigns are a numbers game. Where do you get the volume of "ministers". If you have ten and they get to 20 people per day, is this going to swing your election? Secondly, who judges which are the good ones? We can't figure out how to identify those individuals who fail at teaching. There seems little prospect to identify who is persuasive to voters by one on one canvassing.
Lastly, proposing abstention seems like something that would be really easy to make look bad. At least I know I don't like republican attacks on Acorn because at root it is to intentionally suppress voting. At the same time, at the group level, not personal level, campaign targeting has worked to minimize voting in areas by reduced campaign presence for a long time. But this isn't the same as saying it to someone's face.

As an experienced field organizer (4.00 / 1)
I think every campaign would love to do this--or at least any smart campaign.  The problem comes down to resources.  Using volunteers is great, but there are never enough, especially if you want 30 minute conversations.  That's only a few conversations per shift, which means talking to enough people requires a helluva lot of volunteers.  Where are all these volunteers going to come from?

I don't know a single campaign that would say no to effective volunteers.  The problem is that except, maybe, for Barack Obama, no campaign has enough volunteers for a normal-sized campaign, never mind something this big.


Trying to microtarget (0.00 / 0)
contact based upon a particular reason why someone feels a particular way on an issue would be very difficult.

First of all, you have to actually accurately break up the different possible reasons why someone opposes or supports same sex marriage. For different people, there will be many different reasons, often too subtle to really be captured by a modeling category. People are also often conflicted, and may have multiple reasons for their opinion. They may also not be able to articulate a reason, and the reason may not be a real reason (which does not necessarily mean that they lie, but that people rationalize and think on their feet).

Second of all, even if you somehow got past that step, you now have perhaps 10+ different categories. Microtargeting (at least as it is currently implemented) works by comparing responses and extrapolating across the broader population on a 2 variable scale, not multiple variables. So you would have to compare each reason for opposing to . If there were any appreciable number of different reasons in your categorization, very few people would have a high score (or even a .51 above average). So for the most part either you would be sending your specialized volunteers (e.g. Catholic ministers) out on what are largely wild goose chases, canvassing people with very low scores, or you would have unwalkable target lists. The only way to get reliable data on particular reasons why individual people have a particular opinion is to actually canvass all of them, which takes lots of resources and sweat.

It is true that the more specialized and personalized a contact is, the more likely it is to be successful, and it's also true that canvassers are often poorly trained (happens most often when there are lots of volunteers that only show up to canvass once or twice right before or on election day). But microtargeting is not a magical silver bullet to solve that problem.

It is also a good idea to try and have volunteers re-canvass the same people as much as possible, but hard to implement in practice.

3s are not normally ignored if the election is still reasonably far away, but they are and should be ignored on election day and the weekend before.


Canvasser Training (4.00 / 2)
I spent a summer helping run a field campaign for the Dennis Shulman Race in NJ-5. I've also organized for some state-house level races. Full disclosure: I'm also Adam's intern.

I think that field directors are under far too much pressure to knock on large numbers of doors, and have little regard for the quality of the contact. Volunteer training is almost always inadequate, and I've never had a lot of trust in paid canvassing.

GOTV canvassing is easy. A previous commenter mentioned how loads of new volunteers show up for only a day or two in the final part of the campaign. Thats fine: have them turn out supporters. Only minimal training is needed.

Persuasion / Identification canvassing, on the other hand, is incredibly difficult, and in my experience almost always neglected by the campaign team. It takes a lot of skill, experience, and knowledge to get a decent persuasion or identification contact at a door. Scripts don't help one bit.

Our best success on the Shulman campaign came, we think, from our interns. We had a core of 13 full to semi-full time interns who were pretty awesome (if I don't say so myself). We had a decent among of success because all of us had the knowledge of the campaign to speak authoritatively on a host of issues, and the experience to go out and connect with a voter. But that only worked because so many of us volunteered 60 hours a week on the campaign for months. How do you instill that in a training session? I don't think its possible. I think that for persuasion canvassing that actually moves votes, you need to spend hours working with volunteers to make them comfortable with the campaign's message, and their ability to deliver it. If you have to send them out with a script, you are wasting a resource that could be spent actually picking up votes somewhere else.

(Obviously, I have no empirical data to back up my claims)

We also had a decent among of success gathering data with less trained volunteers when we were framing the door knock as part of our "listening tour". The purpose of this was to let the voters do the talking. Our script was a slightly longer version of, "Hi, I'm here with Dennis Shulman's Listening Tour. Dennis is preparing for the fall campaign by soliciting ideas from his voters. What do you care about in this election", and people would go on for ten minutes about their pet issues. We'd mark it down and collect the data. It worked reasonable well on the phone too.

I'm not sure if this model can be adapted for a single-issue campaign like we saw in Maine.

I also agree with the above posters on the notion of over-harassing 1s and 2s. This may be an extreme example, but I volunteered for Obama in the New Hampshire primary for 4 weeks, and voters were ready to kill every canvasser that they saw. I think that field campaigns need to be able to put supports down as an absolute, "Do not call/knock" that the campaign will respect. Unfortunately, this would only work with trained canvassers who know which voters to code this way and which not to. I remember one of the Democratic campaigns in New Hampshire (Either the Hillary or Edwards one) distributed, "Please do not knock, I am supporting Hillary/Edwards" door hangers to supports that were probably mildly successful.

To sum it up: contacts absolutely have to be personal, or they might as well not have happened. If a campaign has transient volunteers who can't be thoroughly trained, they shouldn't use them to door knock.


I'd agree with that (0.00 / 0)
My most recent campaign was for a City Council candidate.  The core team did that and we had a lot of success with it, but you can't distill encyclopedic knowledge of the issues to volunteers with just a training.

[ Parent ]
Agree with most of the comments here (0.00 / 0)
I've been doing issue/initiative campaigns intensively for about 20 years in California. Many of those campaigns involved mobilizing new constituencies, often of color, often quite poor, with low levels of experience in electoral politics.

There's a reason that mostly we have gone for identifying strong supporters rather than trying for persuasion. We've never found effective training that gets a significant number of volunteers prepared to have a substantive conversation about our issues -- this is just too much of a stretch even with very committed people. And even if you can drill on the issue substance, getting folks proficient enough to actually hear what someone they meet who disagrees really means by what they say is just too hard.

Heck, half our staff organizers can't usually achieve that level of sophistication.

I did think Jasmine's notion of a pretty sophisticated team of what amount to social service listening professionals could probably change some votes (or win abstentions). But how can any campaign afford the resources to do that level of persuasion? Maybe it could be tested in a small locality. I can't see it working even in a small state.

Can it happen here?


USER MENU

Open Left Campaigns

SEARCH

   

Advanced Search

QUICK HITS
STATE BLOGS
Powered by: SoapBlox