A Citizen's Guide to Beating Donald Trump

by David Plouffe

  In retrospect, it really was that simple, I guess. We had gotten all tangled up in analyzing or second-guessing other people’s strategies, and had lost sight of our own North Star for almost three-quarters of an hour. That clarity of mission and purpose is what we need this time. All of us will need it, most importantly the candidate who proclaims the sanctity of funding the grassroots operation, the campaign team who receives it loud and clear and then provides to everyone in the trenches everything you will need to be successful, whether you are an infrequent volunteer or a precinct leader.

  If you are asked to volunteer, asked to give your time and talents, and constantly assured how important it is, the campaign must then honor and respect your commitment by making sure the grassroots campaign is resourced fully and effectively, even when it’s tempting to cut some budgeting corners.

  Many campaigns will ask for money, promising it will help pay for field organizers or local offices or supplies for local volunteers. If you are one of those contributors or fund-raising targets, and you see the grassroots efforts undernourished, give, yes, but also lift your voice. Point it out, even complain.

  If you, a veteran precinct captain in Wisconsin, tweet at the nominee that your team is working hard but with fewer resources than you need, fewer than you had in the Obama campaigns, I promise your warning will get the attention and discussion it deserves.

  * * *

  —

  We have an election to win: one of the most important in our country’s history. A lot that has affected the race has already happened and much more will continue to happen that is beyond our power to control. The state of the economy. Pandemics. Mass attacks, foreign or homegrown. Tensions or outright conflict or war. The quality of the campaign Trump musters.

  But there is a great deal of importance that we can control—we, the volunteers and the official campaign apparatus. The volunteers can control our commitment, our effort, our enthusiastic and infectious joy in the endeavor. The campaign can control the quality of the tools and data and TLC they bring to the party. Let’s hope they also bring some serious joy to this serious task—anything hard is easier when it’s fun too.

  8

  VOTING

  In the 2020 presidential election, the Democratic nominee will enjoy more support among the roughly 250 million eligible voters than will Donald Trump. I would hazard a guess that the number is likely close to 55 percent. Perhaps a bit higher. Among those actually registered, the support levels slip a bit, but the good news is our nominee will enjoy a clear majority of support among the 175 million registered voters.

  Big advantage for the Democrat, no matter who it is. Demographic modeling, polling, and voting history patterns tell us all this. Add the new registrants we need to add before Election Day, and our advantage swells.

  However, actual election results over many campaigns prove that Republicans are “higher propensity voters,” in the lingo. They get to the polls and vote more consistently than Democrats, without much prodding. We can match them, but it takes a lot of time, work, and money.

  In 2016, registered Democrats voted at lower rates than registered Republicans, alarmingly so when you look at younger minority voters. And white voters who did not go to college, a Trump base, saw their percentage of the electorate surge over 2012 levels.

  If we have to deal with the same discrepancy in 2020, big advantage Trump.

  But the respective advantages balance out, right? I’m afraid not. Without any effort at all, Donald Trump starts with an advantage in terms of support levels materializing into actual votes. Add the fact that there are more conservatives than liberals in traditional battleground states, and that advantage gets even greater.

  Trump starts closer to the goal line than we do.

  The solution for Democrats? I’ll quote one of the best campaigners in American history: “Organize the whole state, so that every Whig can be brought to the polls . . . divide the county into small districts and appoint in each a sub-committee . . . make a perfect list of voters and ascertain with certainty for whom they will vote . . . and on election day see that every Whig is brought to the polls.”

  That campaigner was Abraham Lincoln. The year was 1840, when he was a member of the Illinois legislature. Almost 180 years later, such a “perfect list” is the aspiration of all political campaigns, large and small. By decree of the numbers, we must work harder in 2020 to make certain our lists are as close to perfect as possible. Do everything possible to ensure that we are talking to the right voters about the right things at the right time. This is the only way we can win. Otherwise, all the work, money, inspiration, perspiration, rallies, debates, ads, posts, laughs, and tears will come to naught.

  * * *

  —

  GOTV is the acronym for Get out the Vote, which is the mother of all bottom lines. When I first started in politics, our tactics on this vital front were much less precisely targeted than they are today. We made large assumptions based on larger demographic and geographic assessments about the county, district, or state. There was very little individual data, and what we had was crude, maybe a list of voters we had talked to at some point in the campaign, identified as supporters, and intended to contact a second time near Election Day. So not too long ago, certainly through the 1990s, GOTV consisted of sending mail, telephoning, and knocking on the door of everyone in a high-performing Democratic precinct—a college community, say, or an urban center—and encouraging each and every one of them to vote.

  Of course we knew that even the strongest Democratic precinct is, say, 85 percent or even 90 percent loyal, still leaving a chunk of vote for the Republican regardless of who calls them or shows up at the front door. I remember asking about this little problem in the first general election campaign I worked on, for Senator Tom Harkin of Iowa in 1990. If we contact all the registered voters in those strong precincts, aren’t we worried we’ll be turning out supporters of our opponent too?

  Yes, came the answer, but it’s the price we pay for making sure we drive as many of your own supporters out to the polls as possible. Think about the inefficiency. Volunteers were out there working hard to get out the vote among supporters who, in many cases, didn’t need the reminder and at the same time encouraging supporters of the opponent to vote. Such were the limitations of the data and technology back in the day. Today we have the computer-based ability to take into account consumer and behavioral data and make detailed person-by-person and house-by-house distinctions to find the needle who may behave electorally different from the haystack.

  Now, I understand that with many people concerned about online privacy—unwanted, unwarranted, and unauthorized tracking, and demographic profiling of all kinds—these tactics in a political campaign may sound a little creepy. But it’s going to happen across all aspects of our life. I am a pragmatist, and given the stakes are we going to disarm?

  Because they surely won’t.

  It’s essential to spend the extra time, care, and resources on those voters we believe are more likely to support us but are also more likely to slip away and not vote at all. Smart, tech-savvy targeting is vitally necessary. If we’re interested in winning this year, or any year in the future, it’s how we have to compete. Fact of life.

  Everyone’s voting history is accessible to campaigns: when you voted and where, though of course not for whom. Campaigns try to focus on voters who could be at risk of not turning out, which is usually defined as someone who has missed voting in some elections, whether they be state, local, or federal. I’d also focus on all new registrants, who have no voting history.

  Ideally, the campaign and its volunteers will have talked to as many individual voters as possible so that in addition to the demographic documentation, we have original, personalized data and information to inform our decisions about how to treat a specific voter. Definitely a supporter? If not, what’s the problem? Definitely going to vote? If not, wha
t’s the problem? Voting early or by mail? I hope it’s clear by now in this little treatise that no information is more accurate and valuable than that gleaned from direct person-to-person contact.

  The campaign puts into the blender as much data as possible: each voter’s past voting history, gender, race (when that can be identified), consumer information (income, car type, likely TV viewing habits, internet habits, news consumption), census information, polling, surveys, and any information gathered by personal interaction a volunteer on the campaign may have had—the most important kernel of them all.

  And out pop the scores—one for “support,” one for “turnout.” Definitely voting for the Democrat? Your support score is 100. Always wearing a MAGA cap? That earns a 0. Legitimately undecided? You’re a 50.

  Regarding turnout, scoring 100 means you’re a lock to vote. The lower that number drops, the better the chance that you will not, push comes to shove, get to the polls, much less vote early.

  These numbers are used to determine what further actions are required by the campaign’s volunteers to ensure your specific vote. You score 100 on support and 100 for turnout? Nothing to worry about there, and I hope the campaign leaves you alone. You score 50 and 50? If you don’t hear from at least one volunteer, something is very, very wrong in the ground campaign.

  The value of those numbers doesn’t end there. They are also used to build “look alike” audiences (which is also how most consumer companies use platforms like Facebook to advertise and reach potential customers). The software puts together voter cohorts of people with scores close to yours. So even though the campaign volunteers will not have talked to all these voters—in fact, it will have talked with only a sample despite all the best efforts—the models’ support and turnout numbers will predict which way every single voter in America will go on November 3.

  Of course the models won’t be 100 percent right, but they were eerily accurate in 2012, with President Obama’s reelection campaign. We did have the advantage of being the incumbent. We had a very good sense of the electorate as it related to Barack Obama. We had a lot of good data about voter opinions over a five-year period. And Mitt Romney was a much more traditional Republican nominee than Trump, so it was easier to model his support and turnout scenarios. And the third-party vote was much smaller than in 2016. In the end, we had some states in which our win margin was within .10 percent of the prediction. Some of our estimates of the early vote breakdown in states like Ohio were off by a scant .01 percent. Eerily accurate.

  But it mattered not a whit that we predicted what was likely going to happen. What really mattered was the fact that accurate models and lists for different voter groups meant that the volunteers were using your time efficiently and effectively—again, talking to the right voters about the right things at the right time.

  Right about now you’re probably wondering about 2016, when even the Trump campaign’s models predicted a convincing Clinton victory, with less than a 20 percent chance of winning. Its leadership held briefings with the national media on Election Day, explaining before the polls closed why they had lost.

  So the models were off, but here’s the surprise: they were not off by much. The killer was that the discrepancies just happened to be (1) on the margins and (2) in the most crucial swing (that is, battleground) states. So (3) pretty much all the mistakes broke against Clinton. Three strikes, and you know what happens next.

  Overall, the Clinton and Trump campaigns had made smart, data- and model-informed decisions about whom to target for persuasion and GOTV. It’s likely that Clinton’s models were too confident about the turnout from some younger voters and African American voters, core members of the Obama coalition that she inherited. And they overestimated support levels in rural and exurban areas, losing counties 70 to 30 that they thought would be 60 to 40 or even 50–50. If they had known this at the time, the Clinton campaign could have increased contact with some of those younger and African American voters in the upper Midwest, and with more urgency. They also could have invested in more persuasion work with rural and exurban voter targets.

  The lesson: we can’t rely just on the models, and we can’t blame the models. We also have to have campaign strategies in place that address all likely scenarios. In 2020, it would be nice if the final models show Trump with a 0 percent chance of winning. That’s not going to be the case. To win, our GOTV operation, utterly dependent on volunteers, will need to start with a deadly accurate sense of which voters the campaign and the volunteers should concentrate on, and then be deadly efficient in carrying out the operation.

  * * *

  —

  Early voting is important. The campaign will work hard to nail down as many of these early votes as possible, thus leaving more resources available for Election Day. Now, I know that some people enjoy the excitement and camaraderie of voting on Election Day. Traveling to your precinct polling location; standing in line with your neighbors and fellow citizens; bringing your children with you into the voting booth; making your choices; turning in your paper ballot or getting a receipt from the machine; getting that important “I voted” merit badge; leaving the school, church, community center, or someone’s garage. That’s right, in San Francisco, where I live now, many official polling stations are in someone’s garage.

  It can give you a great overall feeling that you exercised your franchise in and surrounded by your community. If you feel super strongly about that experience, and your life will be poorer for not having done it, by all means enjoy the Election Day ritual.

  If you feel less strongly about the experience and live in a state that has no-excuse early voting (meaning you want to vote early or absentee), you can just do it, there doesn’t have to be a reason or need. I implore you to make plans to vote early. Apply for an absentee ballot or plan to vote early in person at an early voting site. Some states, like California, allow you to become a permanent vote-by-mail citizen. Like clockwork, your ballot comes in the mail in the weeks prior to Election Day.

  Some states, sadly, require an excuse or reason to vote early by absentee—a health challenge, you will be out of town for work, you have no way of getting to the polls. It’s real retrograde stuff. Check your local rules online.

  For the majority of you who can vote early or by mail unrestrictedly, apply common sense. Something unanticipated could come up—a work crisis, transportation issue, family caregiving, or illness. Why risk losing your own vote? For those of you who plan to be active in the campaign on Election Day, you want to be free and clear of anything that could cannibalize your volunteer time. Driving to your local school to vote, waiting in line, filling in your ballot, then driving back to a phone bank or canvass staging area could take an hour. That’s a bunch of GOTV target voters you won’t be talking to. And if thousands of other volunteers also take time away from their volunteer shifts to vote, all of a sudden we are talking about large numbers of voters who may need a reminder, a prod, or a ride but won’t get it. This would be a tragedy.

  Please, get your vote against Donald Trump in the bank, safely, securely, and on your own schedule, if you can do so legally. This is important no matter where you live. We need your vote to help Democrats up and down the ticket, and while the presidential race in your state may not be close, your local U.S. House race, state senate race, or state rep race could be decided on the margins.

  You should be encouraging everyone in your circle to vote early as well. The same cautionary notes apply for them. In this world, complications abound. So persuade your friends and family and fellow supporters of our nominee to bank their votes. Keep a list so you can check them off and of course ensure that you are signing them up to help with voter turnout efforts in the weeks leading up to and on Election Day.

  Maybe these admonitions sound obsessive, but obsessive is what we need to be in 2020.

  Our nominee’s campaign will be laser focused on maximizing early vote, bot
h in terms of banking votes early and freeing up as many people as possible for Election Day duties, be it GOTV volunteering, canvassing key target neighborhoods in Philly, or calling into Milwaukee from your home in Oklahoma City.

  As I mentioned, the campaigns can track who has voted early in person or mailed in a ballot. The raw numbers—votes in so far—are not that valuable. What is valuable is the makeup of that early voting cohort. In each case, the campaign doesn’t know how these diligent citizens voted, but it does know their respective “support” and “turnout” scores on the 1 to 100 scale, which allow the campaign to estimate what share of the early voters it is probably picking up (that is, if they have high support numbers) and what share it needs to pick up in the rest of the early voting period and on Election Day.

  What it really hopes to see is a lot of voters whose “support” number is good, but whose “turnout” number is not so good—first-time voters, young voters, voters who do not regularly vote, voters who have not voted recently. As the campaign creates and executes early vote plans, and monitors its progress, it will focus closely on these people. It is of the highest imperative to get their votes in the bank early because otherwise they may never be deposited at all.

  If the campaign isn’t happy with the early numbers and believes it is on a trajectory to fall short in a precinct, county, state—or any combination—it may devote more resources to try to goose up the numbers, especially with the critical probably-for-us-but-may-not-vote targets. The Trump campaign will also be laser focused on the early voters. For political insiders, this campaign within the campaign is fascinating. For you—volunteers, supporters of the Democrats’ nominee—it’s more than that. We have to win it. In some jurisdictions, more than half of all votes will come in early.