Notes on the Obama Field Program, Nevada 2008
Nevadans voted for the Republican candidate in each of the Presidential elections of the 80’s, then went for Clinton both times and then went for Bush both times. Over the past 100 years, Nevada has voted for the winner in every election except for 1976 when Ford beat Carter in the state by four and a half points.
The campaign sized up the Nevada race in 2008 this way: Clark County would go blue, the rural areas would go red, and Washoe County could go either way. In 2004, Bush had beaten Kerry by 2.59% in Nevada. Kerry won in Clark County, home of Las Vegas, as expected, by 5.2%, 281,767 votes to 255, 337, a difference of 26,430 votes. Looking at all other non-Clark-or-Washoe counties in Nevada as one group, the “rurals” broke 67% for Bush and 33% for Kerry, 81,808 versus 40,582 votes, a difference of 41,226. Those two results had been expected. The toss-up was what would happen in Washoe County, the home of Reno. After a tough fight, Bush beat Kerry by 4.3%, 81,545 to 74,841, a difference of 6,704 votes, and won the state.
Washoe is by and large a rural county except for Reno, with its large public university and a small but growing young, progressive, outdoorsy-yuppie culture. Indeed, when I was there, the tiny co-op grocery store had finalized plans to expand and move to a proper storefront near downtown and it seemed that every other person my age I met in Reno was a part of some environmental or arts non-profit.
The county was covered in the news as one of the key battleground counties in the nation to watch. NPR wrote that “Washoe County, where Reno is located, is perhaps the swing county that could determine the outcome of this election”. The campaign’s strategy was to run up the score in Las Vegas, fight for the rurals with unprecedented effort for a Democratic candidate, and turn Washoe County blue, meaning increase the number of registered Democratic voters over the level of registered Republicans. If Las Vegas went big for us and the rurals went less big for them and if we could win Washoe County, Barack would win Nevada.
One of the keys to the Nevada strategy was increasing registration of younger voters in Washoe County, who tend to vote with liberal tendencies at a 2 to 1 ratio. If we could increase the youth vote, and if we could activate the sizable Latino population, energize the working poor and middle class and make a strong appeal to the women who had so strongly supported the candidacy of Senator Clinton – she beat Barack by over 5% in the January Democratic primary caucuses – and if enough Republicans and right-leaning Independents were disenchanted by McCain and didn’t turn out or broke our way, we would be okay. That was our task in Washoe County. We had the right candidate; certainly, the base was strong and enthusiastic and had turned out to caucus in record numbers. We just needed to broaden and deepen the support by bringing those at the margins into the fold.
Much has been written about the numbers-and-goal-driven Obama field program that made up the best Presidential campaign in history. I’ll just give you my experience and explain how things seemed to work from my perspective. The folks directing the field program in Nevada (and their bosses in Chicago) set up performance goals for the campaign by working backward from the number of votes that would be needed – based on all available data and the most conservative projections – to beat John McCain. A model would be developed that took into account historical turnout data, current and projected voter registration data, population characteristics, campaign and partner campaign (labor unions, etc.) resource constraints, strengths and weaknesses, etc. to break the state vote total into votes needed from each area, or region, of the state. Achievement of regional vote goals would be the responsibility of the Regional Directors. Their goals would then be broken into vote goals for the various Field Directors in the regions working under the Directors. Washoe County’s Field Director was J, who worked out of the Reno Headquarters. J and the other Field Directors in the state broke their vote goals into vote goals for individual Field Organizers (FO’s), who were each responsible for a certain area and population within that area. The campaign would, from the top-down, define key metrics and performance goals for each of those metrics that would be needed to achieve vote goals. The intense focus, tracking, and measurement of those metrics was what made the campaign run, and run so efficiently.
Picture an upside-down funnel. The small hole at the top is a single metric and single number - the number of votes Barack will need to win Nevada, as generated by the model being used. The wide opening of the funnel at the bottom represents the many metrics and goals at the ground level used to measure the work being done to get to the top number. FOs are responsible for the metrics and numbers at the bottom. Field Organizers are the link between volunteers and the campaign. Organizers are responsible for developing volunteer and community support; building capacity and turning that energy into the work that the campaign had defined would be needed to meet the goals for the Field Directors, which had been defined as the work needed to meet the goals for the Regional Directors, and on up the chain all the way to the number of votes needed to elect Barack.
Then, think of the ground game, or the work that goes into meeting the vote goal assigned to an area’s Field Director as yet another funnel, this time right side up. The Field Director and his or her FOs begin with a “universe” of people, say, a county’s registered Democrats and independents. This data comes from the Board of Elections and is then refined and cleaned and merged and updated with data mined from past elections – and, crucially, from all the work done by Democrats during the primary. This population is represented by the top, wide opening of the funnel.
In the first phase of the ground game for the general election, we were to expand the width of the funnel; in other words, expand the size of the universe by registering voters. We were also to simultaneously shrink the universe to get rid of non-supporters and folks who had moved or died based on information gathered from calls and canvassing.
So the daily system would be something like this, breaking it down into one cycle of data collection by canvass/phone volunteers and data input by data volunteers. Field Organizers schedule volunteers for canvassing and phonebanking. Those volunteers attempt to – and in about 10-15% of situations do, on average, as far as I’m aware – make contact with a selection of voters from the campaign’s universe as listed in the “walk packets” or “call sheets” that they’re given. This could be 60 doors to knock in a neighborhood or 100 phone numbers to call. The volunteer moves through and the data that is collected for each address or phone number – not home, person moved, person died, line disconnected – and for the voter listed for that address or number – candidate preferences, willingness to volunteer – gets marked down on the voter contact sheets.
The volunteer completes their work and hands it in to the FO. The FO can follow up on any volunteers recruited but otherwise sends it along to the data entry volunteers. These folks punch in the data for the day that will make tomorrow’s work more efficient: those who have been marked as “moved” will be removed from the database. Volunteer time won’t be wasted knocking on that door the next time. Same for those who have died, disconnected numbers, wrong numbers, and those who are firmly voting for our opponent. So the universe shrinks appropriately to help the campaign be more effective and efficient. Our supporters get identified. Maybe we’ll knock on the door five times and make ten calls before we identify that voter as a supporter of ours. But it’s crucial we identify them so that we can a) recruit them to volunteer and b) hound them like crazy until they vote. Undecided voters are identified. Those who are leaning our way are identified. Those people, based on those identifications, can be moved into different sub-lists for the following day’s contact efforts: ask a rock-solid supporter to volunteer, or make sure the undecided voter gets contacted again by a smiling face with a compelling message tomorrow.
We wanted to register people who were likely to vote for Barack – younger, lower income, educated, minorities. By law if you’re registering people to vote, you can’t not register someone based on political preference. So we put ourselves in position to register supporters. The odds that my team would register a Barrack supporter at UNR were higher than that of any other organizer’s population. So part of the strategy for winning Washoe County was for me to run up the voter registration score as much as possible at UNR and then turn those kids out to vote.
The field team would recruit volunteers as a part of that process of expansion and contraction and the work would grow exponentially. More volunteers means greater capacity to reach likely voters, which means more voters registered, more persuasive conversations and otherwise positive interactions with a friendly ambassador from the Obama campaign that might endear the candidate to the voter. Those interactions might provide some motivation for the voter to look into policy, look into the backgrounds and values of the candidates a little more. Enough of those positive interactions, or contacts, over time and the campaign could move people who were undecided or leaning toward another candidate to vote with us. Enough effective contacts – or the right touch at the right time, regardless of frequency – and we could get a progressive-thinking person to go from jaded and obstinate to hopeful and engaged. Or we could get someone to think outside of their parents’ political perspectives (the number one indicator of how someone will vote) and find their own. This is what I was doing on campus, dealing with young people, many of whom voting for the first time and many of whom having only recently moved out of their parents’ house.
In that first phase, the FO works to expand and contract his or her universe to have a shot at meeting goals for voter turnout. Things change when the registration period ends in that the expansion is essentially over except for a small percentage of persuasion opportunities that might expand the support base slightly. From then on, it's about retention and doing whatever is needed to lock down votes while also, always, recruiting volunteers. A few weeks out from Election Day, the ideal situation is that the FO and his or her key volunteers or interns should have hard data on the supporters that need to be turned out. The data is clean – i.e. because of the work of volunteers canvassing and phonebanking, voters still on Democratic voter registration rolls who have moved or not voting with us have been removed from the lists, so the campaign won’t be wasting or resources but instead will efficiently and methodically turn out our voters.
Indeed, in Creston, Iowa, the Obama campaign had been collecting data on, and interacting with, all registered Democrats and Independents in the city for months by the time I got there. I knocked the same doors several times during my week of work. We knew that old Mr. Brown on First Street was caucusing for Barack but his youngest daughter was undecided and she was out of town during the week and his son was voting for Edwards but had been courted by both campaigns for the last several weeks and promised to make up his mind soon, etc. We had that level of detail. I can only imagine how many times those doors had been knocked or how many calls those people received from the time Barack’s campaign set up shop in Iowa in early 2007.
On that point, our Field Director in Northern Nevada, J, told us once that studies showed that it took 73 phone calls from one campaign in one week to make a supporter flip and actually vote for the opponent. It was very helpful for me to have this data at my disposal as an organizer in 2008 and volunteer in 2012. Volunteers can be rattled by upset supporters who are tired of being hounded by all the calls and ads and people coming to the door. Rarely a supporter will threaten to vote for the opposition, but it does happen. Very many times, I’ve comforted volunteers – and once a first-time organizer – with that statistic, and encouraged them to have faith and keep working. Folks in Iowa were getting nightly calls and daily door knocks for weeks leading up to January 3, and sometimes, especially in the last couple weeks, they were getting multiple calls and multiple knocks per day from the same campaign. So you hound and hound and make sure your people get to the polls.
In 2006, Republicans had a voter registration advantage in Washoe County. At the end of the voter registration deadline in the middle of October, 2008, we had achieved our goal of turning the county blue. We had worked to create a registration advantage of 1,300. The progress in Washoe was mirrored in Clark County and even in some of the rural counties. Half of the quarter million new voters who registered in Nevada in 2008 registered as Democrats. Democratic registration outpaced Republican 2 to 1. We had done very well in phase one. This was due to good organizers but mostly to the energy and time and sacrifice and dedication of local volunteers and also largely because of the volunteers who took a weekend or a day off work and stepped in from CA and flew in from other non-battleground states to throw their weight around. This is how you win elections.
The second phase is focused on getting your people – your database that has been expanded by voter registration and some small amount of persuasion conversations and also contracted by continuous identification of voters to remove non-supporters from the universe – to the polls. In Nevada, we would have the benefit of early voting. When a supporter voted early, the Board of Elections removed them from the dynamic database we were working from. So our universe got smaller each day. This is why early voting rules from an organizing perspective. It allows you to bank votes and get people off of the lists so your limited turnout resources stretch further. On Election Day you have a smaller database, a smaller number of people to turn out, because of all of the people who voted early. This election day turnout is the bottom of the funnel. You know who they are, and you're not wasting any time talking to people who have voted or people who are not voting with you. There's no guesswork (as long as the data is good) when a volunteer makes a call or knocks a door; it's just putting in enough contacts per supporter to be certain that he or she will vote. In 2008, 101,604 people, nearly half of the 231,437 registered voters in Washoe, voted early. Early voting was an organizing blessing and hugely important for Barack in the state.
Jonathan Weisman, reporting on the race for the Wall Street Journal, wrote on October 31, the final day of early voting in Nevada:
Today is the final day of early voting in the key swing state of Nevada, and the question must be asked, has Barack Obama already won there?
As of late Thursday night, registered Democrats had cast 225,670 of the 438,129 ballots (51.5%) in the two most populous counties, Las Vegas’s Clark County and Reno’s Washoe. Republicans cast 31.3% with the remainder cast by Independents.
Those two counties account for about 90% of the state’s turnout. Early voting is expected to make up 60% of the Silver State’s 2008 ballots, and with a 20-point margin already banked, assuming most of the registered Democrats voted for Obama, the party goes into Election Day with a big cushion.
Andres Ramirez, vice president of Hispanic programs at NDN, a Democratic organization, says John McCain will need to nab between 75% to 80% of the Independent vote, a tall order given that Nevada polling shows nothing like that level of support. He also would need about 12% to 15% of the Democratic vote, perhaps an easier prospect.
There was record turnout for early voting in Washoe County, and campaign manager David Plouffe reported in the press at the time that 43 percent of the Democrats voting early in Nevada were either new or sporadic Democrats. This was the proof that the universe had expanded and contracted during phase one in the right way. In Washoe, 47.1 percent of early voters were Democrats versus the 35.3 percent who were Republicans.
Barack, of course, won Nevada by a 12.5 percent margin, 55.15 to 42.65. He won Washoe County, mirroring the overall state margin, at 12.7 percent, 55.3 to 42.6. Recall that Kerry had lost Washoe by 4.3 percent in 2004. We had a nearly 17 point swing. McCain won the rurals but by only 19 percent, much better than the 34 percent margin for Bush in 2004. Clark County, which Kerry had won by 5.2%, went for Barack big: he had a 19% margin of victory.
We achieved what we set out to achieve in the state – the rurals went less big for the Republicans, we ran up the score big in Clark County, and we won Washoe County.