Some Things that Went Wrong in Wisconsin
I should start by saying the campaign was a beautiful thing. It was incredible that within two short years people pulled together a campaign that spanned the entire state, employed thousands of volunteers, and probably changed more than a few minds. But what interested me most about the campaign was why it failed. Maybe it was never possible to win in Wisconsin, but it is possible to learn from the campaign’s mistakes.
When Fair Wisconsin started two years ago, the goal was to make Wisconsin the first state in the nation to reject a state constitutional amendment about same-sex marriage. They embarked on a grassroots effort that included, I assume, some type of market research and (what I hope was) careful analysis of voter rolls to identify targets for get out the vote efforts. Of course, I don’t really know because as a canvasser I was treated like a “mushroom” (“keep them in the dark and feed them on sh*t”). So everything I say about their campaign is opinion and guesswork gathered from what I could see, and from the (mis)information they gave us lowly campaign peons.
We had a meeting in DC before we went, where our bosses gave us advice on what to expect and how to behave. They told us that campaigns are completely egoless, and that we should do exactly what we are told. Importantly, they told us that we should use the precise language in the scripts we’d get because campaigns use professional services that run experiments with test groups to see which words are the most effective. They told us that we’d be plugging into a magnificently well-oiled machine.
To a certain extent, we were entering a highly organized campaign that was almost astonishing in its ability to slide thousands of volunteers into useful positions without wasting much time. They always had print-outs for the canvassers including little highlighted maps to prevent getting lost or wasting time. They had long lists of people to call and buckets of cell phones constantly charging for the phone banks. They had volunteers smoothly popping into place at all the “staging locations” around the city (and I assume the rest of the state). Despite the administrative organization, I did not see a matching level of organization in control of messaging.
The first strange thing I noticed was that our scripts were very awkwardly written. I doubt that it had been tried in test groups. The unpronounceable sentence structure invited diverging from the script, which volunteers did to a possibly harmful extent. The script we had told us to remind people to vote “No” on “The Constitutional Amendment,” without mentioning same-sex marriage. We were supposed to say that it “outlawed civil unions” and that it was harmful to “loving and committed families” because it denied protections that all families need like “the right to shared health insurance and the right to make medical decisions for a loved one.” I don’t remember the awkward parts offhand because I never said them. Small divergences led to larger ones as people started arguing that “the amendment isn’t just about gay marriage,” or embarked on long conversations with people who were confused about which way to vote.
I didn’t understand why so many voters were confused until I saw the television spots they were running. The ads on our side said “Vote No on the Constitutional Amendment. Voting ‘no’ means NO gay marriage.” I understand that this was meant to counteract the messaging from the other side, which was saying “a No vote means opening the door to gay marriage in our state.” Some supporters of the amendment were going so far as to say that “voting no means gay marriage will be legalized in the state.” Counter-messaging can be effective, but this type had the following problems: 1) I don’t think they did research to see whether the message put out by amendment supporters was confusing the majority of voters, and 2) the counter-message they chose had the effect of confusing everybody. The result was that we often spoke to people who were completely confused about how to vote if they wanted to support gay rights. They were hearing from some people that a yes vote meant enshrining homophobia in the state constitution and from others that voting no meant voting against gay marriage.
Another misstep I noticed was that our list of voters to target was far from accurate. Many people on the list had moved, some up to two years ago. Also, the list was supposed to be carefully crafted to include only voters who would support our position. Their volunteer training was based around the idea that we’d be talking to supporters, just trying to get them to vote. We were not supposed to be convincing anybody which way to vote - supposedly they had already taken care of that with their confusing ad campaign proclaiming that voting no means no gay marriage. Everybody had a different story of how the names had been assembled. Some people said that it was from a list of Democrats. Others said that they took the voters’ history of voting on certain issues and extrapolated which way they would vote based on that (is that even possible???). Others said it was based on their magazine subscriptions (how would they get a list of magazine subscriptions for each voter?). So I have no idea how they assembled the list, but it was wildly inaccurate. As we were canvassing we were skipping over houses that had “A Fair Wisconsin Votes NO” signs on the front lawns, and going directly to houses that had cars with “God Made Adam and Eve, Not Adam and Steve” bumper stickers. Our phone calls were similarly disastrous, and several volunteers quit because they had called 100 houses and been yelled at or hung up on or preached to by the other side on every call.
Ray Vahey, the man I wrote about in my last post, said that he thought one of the biggest mistakes was failing to properly ally with the communities of color in Milwaukee. The campaign did not reach out to any communities of color until they hired a single organizer to do so with only three months left until the election. She did the best she could but she was just one person stepping into a campaign very late in the game. In the end we were resorting to simply doing “flyer drops” on communities of color, where we’d drive up to a Walmart or an IHOP and put little flyers on every car. Ironically, that may have been even more effective than going door to door, because at least we reached almost every person in those parking lots.
The organization of the campaign broke down somewhat in the days right before the election. The scripts began to change every day as various volunteers complained about the content and the way they were being received (which further reinforced my opinion that they hadn’t been carefully researched to begin with). We went to drop flyers at churches that we were assured were on our side and were expecting us, only to be chased off the property by angry, angry preachers. One of the most embarrassing and unnecessary missteps was when somebody on the campaign decided to go flyering on the day before the election using door hangers specifically printed up for election day. The hangers commanded the reader to “VOTE TODAY.” It was embarrassing because apparently a volunteer from another organization who was doing election protection took our volunteers for troublemakers who were trying to keep the people in these communities away from the polls by getting them to try to vote a day in advance.
In the end, Wisconsin did not become the first state to reject a constitutional amendment about same-sex marriage. But Arizona did. As far as I understand it, that campaign succeeded in a lot of the areas I just wrote about. They actually did hire companies to do smart voter analysis and figured out exactly who to target for maximum impact. They did direct mailings to various sectors of the population, giving each of them the message they needed to be pushed over to our side. Importantly, they stayed on message and did not confuse the voters. I’m so glad the amendment was defeated in Arizona. Maybe all future campaigns against similar amendments will use that model instead of the Wisconsin model.
When Fair Wisconsin started two years ago, the goal was to make Wisconsin the first state in the nation to reject a state constitutional amendment about same-sex marriage. They embarked on a grassroots effort that included, I assume, some type of market research and (what I hope was) careful analysis of voter rolls to identify targets for get out the vote efforts. Of course, I don’t really know because as a canvasser I was treated like a “mushroom” (“keep them in the dark and feed them on sh*t”). So everything I say about their campaign is opinion and guesswork gathered from what I could see, and from the (mis)information they gave us lowly campaign peons.
We had a meeting in DC before we went, where our bosses gave us advice on what to expect and how to behave. They told us that campaigns are completely egoless, and that we should do exactly what we are told. Importantly, they told us that we should use the precise language in the scripts we’d get because campaigns use professional services that run experiments with test groups to see which words are the most effective. They told us that we’d be plugging into a magnificently well-oiled machine.
To a certain extent, we were entering a highly organized campaign that was almost astonishing in its ability to slide thousands of volunteers into useful positions without wasting much time. They always had print-outs for the canvassers including little highlighted maps to prevent getting lost or wasting time. They had long lists of people to call and buckets of cell phones constantly charging for the phone banks. They had volunteers smoothly popping into place at all the “staging locations” around the city (and I assume the rest of the state). Despite the administrative organization, I did not see a matching level of organization in control of messaging.
The first strange thing I noticed was that our scripts were very awkwardly written. I doubt that it had been tried in test groups. The unpronounceable sentence structure invited diverging from the script, which volunteers did to a possibly harmful extent. The script we had told us to remind people to vote “No” on “The Constitutional Amendment,” without mentioning same-sex marriage. We were supposed to say that it “outlawed civil unions” and that it was harmful to “loving and committed families” because it denied protections that all families need like “the right to shared health insurance and the right to make medical decisions for a loved one.” I don’t remember the awkward parts offhand because I never said them. Small divergences led to larger ones as people started arguing that “the amendment isn’t just about gay marriage,” or embarked on long conversations with people who were confused about which way to vote.
I didn’t understand why so many voters were confused until I saw the television spots they were running. The ads on our side said “Vote No on the Constitutional Amendment. Voting ‘no’ means NO gay marriage.” I understand that this was meant to counteract the messaging from the other side, which was saying “a No vote means opening the door to gay marriage in our state.” Some supporters of the amendment were going so far as to say that “voting no means gay marriage will be legalized in the state.” Counter-messaging can be effective, but this type had the following problems: 1) I don’t think they did research to see whether the message put out by amendment supporters was confusing the majority of voters, and 2) the counter-message they chose had the effect of confusing everybody. The result was that we often spoke to people who were completely confused about how to vote if they wanted to support gay rights. They were hearing from some people that a yes vote meant enshrining homophobia in the state constitution and from others that voting no meant voting against gay marriage.
Another misstep I noticed was that our list of voters to target was far from accurate. Many people on the list had moved, some up to two years ago. Also, the list was supposed to be carefully crafted to include only voters who would support our position. Their volunteer training was based around the idea that we’d be talking to supporters, just trying to get them to vote. We were not supposed to be convincing anybody which way to vote - supposedly they had already taken care of that with their confusing ad campaign proclaiming that voting no means no gay marriage. Everybody had a different story of how the names had been assembled. Some people said that it was from a list of Democrats. Others said that they took the voters’ history of voting on certain issues and extrapolated which way they would vote based on that (is that even possible???). Others said it was based on their magazine subscriptions (how would they get a list of magazine subscriptions for each voter?). So I have no idea how they assembled the list, but it was wildly inaccurate. As we were canvassing we were skipping over houses that had “A Fair Wisconsin Votes NO” signs on the front lawns, and going directly to houses that had cars with “God Made Adam and Eve, Not Adam and Steve” bumper stickers. Our phone calls were similarly disastrous, and several volunteers quit because they had called 100 houses and been yelled at or hung up on or preached to by the other side on every call.
Ray Vahey, the man I wrote about in my last post, said that he thought one of the biggest mistakes was failing to properly ally with the communities of color in Milwaukee. The campaign did not reach out to any communities of color until they hired a single organizer to do so with only three months left until the election. She did the best she could but she was just one person stepping into a campaign very late in the game. In the end we were resorting to simply doing “flyer drops” on communities of color, where we’d drive up to a Walmart or an IHOP and put little flyers on every car. Ironically, that may have been even more effective than going door to door, because at least we reached almost every person in those parking lots.
The organization of the campaign broke down somewhat in the days right before the election. The scripts began to change every day as various volunteers complained about the content and the way they were being received (which further reinforced my opinion that they hadn’t been carefully researched to begin with). We went to drop flyers at churches that we were assured were on our side and were expecting us, only to be chased off the property by angry, angry preachers. One of the most embarrassing and unnecessary missteps was when somebody on the campaign decided to go flyering on the day before the election using door hangers specifically printed up for election day. The hangers commanded the reader to “VOTE TODAY.” It was embarrassing because apparently a volunteer from another organization who was doing election protection took our volunteers for troublemakers who were trying to keep the people in these communities away from the polls by getting them to try to vote a day in advance.
In the end, Wisconsin did not become the first state to reject a constitutional amendment about same-sex marriage. But Arizona did. As far as I understand it, that campaign succeeded in a lot of the areas I just wrote about. They actually did hire companies to do smart voter analysis and figured out exactly who to target for maximum impact. They did direct mailings to various sectors of the population, giving each of them the message they needed to be pushed over to our side. Importantly, they stayed on message and did not confuse the voters. I’m so glad the amendment was defeated in Arizona. Maybe all future campaigns against similar amendments will use that model instead of the Wisconsin model.