The Texas Democratic Party released an Election Data Analysis last week looking back at the 2020 election and why Democrats fell short in the state. The report conveys a lot of good information, but it focuses on the field/data side of the campaign and doesn’t look in-depth at anything else. The author was the state party’s data director, so that makes sense, and it is a good summary of that area.
But the Texas Tribune reported on the analysis like this:
Texas Democrats have come to the conclusion that they fell short of their expectations in the 2020 election largely because Republicans beat them in the battle to turn out voters, according to a newly released party report.
Reading the report, it’s easy to see why they and others came to that conclusion. The report doesn’t shy away from fairly definitive “this is why we lost” and “this is how we win” type statements despite only covering a specific part of the campaign. So let’s break down what the report says, what it doesn’t cover, and why it’s so hard to answer the question “why did we lose an election?”
What the report says
The report comes to a few key conclusions, framed around Biden’s 5.5% loss but equally applicable to the disappointing performance at the Senate, House, and state legislative level. First, it shows that Republicans turned out at higher rates than Democrats, both across races and particularly among Latinos. The report also found that while Latino voters in general did not significantly shift from Democrats to Republicans, Latinos in Rio Grande Valley did.
I’m not going to go into detail as to how the report shows this, but if you’re interested I’d encourage you to read the whole report, which is only 29 pages and very digestible. But suffice it to say, the report has good evidence to back up these assertions around turnout and vote shifts.
The report then concludes that the reason this happened was the party’s lack of in-person canvassing. In response to the key question “Why did Republicans win turnout?” The report says:
. . . our turnout work was hampered by two factors: inability to do in-person canvassing and inefficient targeting. As we will show later, inefficient targeting is partially a symptom of our inability to do in-person canvassing, because we were not able to effectively reach large portions of our base for whom we lacked quality contact information.
Essentially, the report says Texas Democrats lost because they lost the turnout battle, and that the turnout battle was lost because they couldn’t do in-person canvassing. It’s easy to read this and think, if only Texas Democrats could have had a big GOTV machine out in the summer and fall, they would have won. This is where I think the report falls short, hampered by its own limitations in scope.
GOTV is not the answer to every problem
Imagine you opened a new food truck selling chicken sandwiches. After a year, your sales are below what you need, so you ask your team to write a report explaining why you missed your targets. The team concludes that you didn’t get the sandwiches in front of enough customers, that you need to select locations with better foot traffic.
At no point does the report address whether the chicken sandwiches taste good. Now, it’s really hard to determine whether or not something tastes good! Taste is highly subjective and there’s no good way to measure it like you can foot traffic. What’s more, you are passionate about these chicken sandwiches, you poured your heart and soul into them, you believe that these are great sandwiches. So your team is wary of bringing up the possibility that the sandwiches just might not taste very good to the average customer.
But the difficulty in measuring taste, the subjectivity of the issue, and the personal importance to all involved doesn’t mean the taste of the chicken sandwich isn’t important. It’s the most important aspect of selling chicken sandwiches!
If you stuck with me through that extended metaphor, the point is that election data analysis is important, but without a real examination of the persuasion and messaging aspects of a campaign, any report is woefully incomplete. Some combination of Trump’s appeal and distaste for the Democrats’ 2020 campaign caused a lot of Latinos in the Rio Grande Valley to move from D to R and I don’t think all the canvassing in the world would have prevented that.
There is a tendency in progressive circles to make the solution to every problem a better and more efficient GOTV program. It’s comfortingly mathematic, where X doors knocked equals Y conversations which equals Z votes gained. If we could just have enough doors knocked, enough phone calls made, enough texts sent, then we win. But all the GOTV in the world won’t matter if the chicken sandwich doesn’t taste good enough.
So persuasion is important; where’s the report on that?
Persuasion (i.e. the taste of the chicken sandwich) is really hard to analyze, far harder than measuring turnout. Voters, particularly swing voters, aren’t good judges of why they vote the way do. Message testing1 is imperfect at best. Everyone tends to believe that the issues they care passionately about are also the ones that will move voters. Experiments testing persuasion messages are much harder to run because how someone voted is not public information the way whether they voted is.
Beyond the comparative difficulty, talking about persuasion means talking about the downsides of issues that activists care about deeply. Let’s use the minimum wage as an example. Raising the minimum wage is clearly popular. It polls well, it passes in referendums, moderate Republicans tend to support some version of it because they worry about opposing it. It also materially benefits millions of Americans, which is a great combination.
What if it wasn’t so popular? The Democratic base still cares about it and it still helps millions of people but swing voters don’t support it. What should a candidate do? Support it and you might lose swing voters (but you also might gain swing voters who respect your principled stance). Oppose it and your base might not turn out (or maybe they will anyway because they hate your opponent). Say nothing and the media will hound you endlessly (unless a Senate candidate has a sex scandal and then you get a free pass). There’s often no clear answer to these conundrums.
And persuading swing voters to vote for a candidate is the easier part of persuasion. The highest difficulty setting in politics isn’t convincing someone to vote for one candidate over the other, it’s changing their mind on an issue, particularly one in a partisan frame. More than any election or candidate or party, changing hearts and minds will get you closer to whatever goal you have and better ensure your long-lasting success. It’s the holy grail of politics, but there’s no instruction manual.
Gay marriage is one of the great success stories of the recent progressive movement and there are a lot of theories as to how it happened, but little of it is transferable to other issues or political campaigns in general. It’s not even clear how much of the change was discrete actions from the LGBT community and activists and how much was just general cultural evolution beyond anyone’s control (probably some of both).
You may have heard about the report the GOP conducted in the wake of their 2012 loss. The central talking point to come out of that report (which did tackle persuasion) was that the GOP needed to become more welcoming to minority and female voters and be more open to immigration reform. At the time it seemed like a smart, even obvious conclusion. Of course, you don’t need me to tell you what happened next.
Persuasion is how you change the world
So persuasion and messaging is hard to test, hard to discuss internally, hard to execute, and provides totally inconsistent results even when done well. You can understand why the Texas Democrats didn’t touch it. But we can’t ignore it because when persuasion messaging works, it can change the world.
In the Georgia Senate runoffs in January, Democrats knew they needed to make up ground from November, particularly in the Ossoff-Perdue race. Ossoff trailed Perdue by nearly 100,000 votes and only made the runoff due to a Libertarian candidate on the ballot. But how do you make up that margin in what was basically a rerun of the election just two months later?
Then Trump decided that the COVID relief checks should be $2,000 instead of $600, which was inexplicably a bridge too far for Congressional Republicans. Out of nowhere, a popular persuasion message with clear separation between the candidates landed in Democrats’ lap. Now (of course) it’s hard to say for sure how many votes were changed due to this issue. But it easy to imagine that it played a key role in Ossoff gaining 150,000 votes on Perdue in the runoff election. Did turnout efforts play a role as well? Absolutely, the pair go hand in hand. Good persuasion messaging makes turnout efforts easier and more effective. That’s why you need both in any good campaign. Neither is sufficient alone.
Next week, we’ll circle back to the report and look more in-depth at the one group where they found significant vote-switching (Latinos in the Rio Grande Valley) and why that may have happened.
Two more pieces of good news:
Neither of these announcements are surprising if you’ve been following the vaccine rollout closely (or reading this newsletter!) but it’s still nice to see these increases and timelines confirmed by the White House.
Message testing is where you poll different arguments (candidate likes cats; candidate likes dogs; candidate likes parakeets) and ask respondents whether each message makes them more or less favorable to the candidate. Then you base your messaging on the fact that 55% of respondents liked the dog message but only 51% liked the cat message. Message testing is absolutely useful but can be over-relied on. This is what people mean when they complain a candidate sounds too “poll-tested.”