Bringing a Survey to a Gun Fight

“Pollingism” Has Failed Democrats and Voters. Here’s Why, and What to Do Instead.

This is a guest post from messaging expert Anat Shenker-Osorio, host of the Words to Win By podcast and Principal of ASO Communications.

Note from Anat: Mike Podhorzer’s ideas, articulations, and moral clarity have informed all elements of this analysis. I owe him and writer, editor, and idea-wrangler extraordinaire Emily Crockett a huge debt. Any errors or omissions are, of course, mine alone.

---

In the now-stale takes about how Democrats lost their electoral way and scuffles around their best present course, leading operatives are brandishing the same compass that led them astray last time. In assessing how to strike back at Republicans and lay tracks for the midterms, what remains unexplored in the data-laden dives are the assumptions about what works to win hearts, minds, and elections in the first place. Not to mention the reason for wanting those wins: to enact the agenda you believe in or, at the very least, blunt the authoritarian assault against Americans now underway.

There are two vastly different ideas about what it takes to achieve political victory, but only one gets real airtime in Democratic circles. This leaves us trapped in circular reasoning, arguing over permutations of a singular strategy. These arguments might look like significant beefs over what Dems should say, to whom, and by what means. But examinations of what went wrong, how to act now, and what to do next fail to even consider the methodology behind these decisions.

The dominant Democratic model views voters as rational individuals who make electoral decisions, including whether to vote at all, based on their conscious preferences about issues. Advocates of this model believe that polling outcomes within controlled survey environments equate to real-world success — that “winning” in testing corresponds to winning in the real world. Although it’s known by other names, the most apt moniker is Pollingism, as it assumes voters’ issue preferences are static, discernible via polling, and hold ultimate, if not exclusive, sway over their voting behaviors and candidate selections. Pollingism does not come with a set agenda for governance; it relies on discerning voters’ registered preferences because it views the political task as winning elections and treats the work of governance as something to be hashed out later.

Pollingism proponents believe that data “shows what voters really think, not what people who work in politics wish they thought.” (Nevermind that the proponents making this claim work in politics.) In their minds, the data have set them free from their biases, including holding fixed stances on right and wrong. The trouble with this is that data aren’t conjured but rather solicited and analyzed according to the assumptions of data collectors. In other words, you only get answers to the questions you ask. And you only get reactions to the ads you produce. And you only assess impacts in the artificial environments you construct. And you only apply findings according to your theory of how humans come to judgments.

An alternative worldview understands voters as inherently social beings, driven by unconscious cues, whose opinions, beliefs, and behaviors are formulated within the context of relationships and dominant discourse. The repeated refrains of friends, family, and trusted messengers, as well as voters’ identities, hold the greatest sway in their actions, electoral and otherwise. This model views political persuasion, even in the sprint of electoral campaigns, as occurring through diverse mechanisms — including offline interactions, persistent media narratives, and social movements. This approach, which I call Magnetism, prioritizes establishing the right conversation — rather than reacting to terms set by the opposition or the moment-in-time polled preferences of voters.

Magnetism is the notion that if you want people to come to your cause, you must be attractive. This, of course, requires having a cause to which to draw people. And, like any magnet, it also means having a polarity that distinguishes you from your opposition. For Magnetism to work, you meet people at the place of their broadly shared values, not their podcast-promoted prejudices, and bring them with you toward your desired policies and the candidates who will enact them.

In short, the first step to winning an election is making it about what you can win on. Strategic campaigns begin with the question: How do I force the issues and conversations that most benefit my side to the fore?

In this post, I’ll explain the biggest problems with Pollingism: the unacknowledged limits of its testing methodology; its failure to reckon with how messages spread; its flawed assumptions about which audiences need to be reached and how; and its blinkered approach to narrative and strategy. Throughout, I’ll contrast Pollingism with Magnetism, with examples of how Magnetism underpins the right’s strategy and has fueled marquee wins on the left. I’ll round out by examining why Pollingism had Democrats stop short of fully exploiting a core weakness for Team Trump uncovered by a dozen independent research projects. And I’ll conclude with what all this means now as we confront the attacks on our lives and our liberties.

A Message Nobody Hears Cannot Move Them

At first glance, Pollingism espouses a so-simple-a-child-gets-it truth: ask people what matters to them and repeat it back. It often gets short-handed as “say popular things” and, by extension, eschew unpopular ones. However, this notion is built off of assumptions that do not bear out in the real world.

Let’s start with an irrefutable claim: a message that nobody hears cannot persuade them. Nor can it get them to act in opposition to the present regime, vote in an election, or select a specific candidate.

For years, we have heard entirely accurate lamentations that Democrats are failing to reach people. To be sure, the right’s far better media system merits every alarm bell, not merely for electoral outcomes but also for our hopes of a functional society, free of lies about everything from vaccines to wind turbines to the dining habits of newcomers. But even setting that imbalance aside, Pollingism’s message creation plus testing model fails to consider, let alone optimize for, how we would get a message to be heard.

Presently, the dominant approaches to campaign message crafting and evaluation involve (1) randomized-controlled trial (RCT) testing of messages and/or ads, and (2) issue-based polling to figure out which topics voters rate most vital and therefore merit airtime. But these tools cannot tell us what would break through in the cacophony of political noise to which voters are subjected.

What Happens In-Channel, Usually Stays In-Channel

RCTs and surveys have their utility. In my decades of work helping design and win campaigns here and abroad for everything from ballot initiatives, to candidates, to advocacy issues, I use them to improve how we make our arguments. I do not use them to determine which positions to hold. And even in gauging whether, for example, a minimum wage hike is more effectively sold for its practical benefits in a consumer-driven economy or as a fight to ensure people who work for a living can care for our families, it’s critical to understand what we can learn from each of these tools – and what they not only fail to reveal but eclipse from view.

You cannot wield a thermometer to help you change a tire. There are inherent limitations to how experimental results apply in the real world of campaigns, an issue academic experts have long noted for other kinds of RCTs. Further, there are problems with what Pollingism proponents deem worthy of measuring.

Traditional surveys once represented a great advance in selecting political messaging. These replaced the finger-in-the-wind instincts of a small handful of party insiders, an especially critical innovation as the electorate expanded beyond landed white men and communication moved from leaflets to broadcast via radio, then television, and now an ever more fragmented internet that nearly everyone carries around in their pockets.

At the outset of the 2000s, RCTs began to be applied to politics first in field experiments and then for persuasion. In an ideal political persuasion RCT, you have a large control sample that sees some placebo about an apolitical topic, and a series of “treatment” tracks. In each of the latter, viewers see one piece of content — a block of text, slogan, ad, or social media post, for example. Then, all respondents are asked questions that could include anything from vote choice, voting motivation, candidate or party approval, issue preference, and/or willingness to spread the word, depending on what is being tested. Statistically significant differences detected between respondents in a treatment track relative to those in the control are highly likely to be attributable to the content viewed, assuming the samples are decent and the test well administered. In practice, many RCTs fall short of these ideals, cutting costs with small samples in treatment and control, allowing bots and habitual test-takers into samples, and even violating the foundational principle of good science: having transparent methodology so other practitioners can understand the assumptions baked into the set up of tests.

RCTs, like surveys, occur “in channel.” People are recruited to respond (usually through online solicitations), made to pay attention (with financial incentives or simply due to how testing platforms are configured), and register their preferences knowing they’re being observed. Once upon a time, when Americans were captive viewers of limited network television channels, forced to sit through commercials without even a remote control for ready escape, these experiments more closely mirrored real-world conditions. Today, however, political messages reach Americans as they are scrolling or cooking or breaking up skirmishes between their kids. Indeed, they’re mostly doing their best to avoid ads altogether.

In-channel testing gauges the response to a single exposure of a message and pronounces these to be the message’s effect. But messages aren’t single-use plastics. Many don’t work until they are repeated frequently. Some are more rebuttable than others. Some could work if spoken by messengers other than those in the materials tested.

Vote choice tends to be the gold standard in electoral RCTs. Treatments that produce higher numbers of respondents saying they’d vote for the Democratic candidate are considered to have bested the game. While this may seem like the obvious metric — the task come election day is to net more votes — this simplifies how one reaches that goal. You cannot mistake winning a trial heat for winning an election. Often, what you need to measure is what would make the story that favors your side seem credible, what punctures the key strength of your opponent, or what — only upon unrelenting repetition — reframes the electoral stakes to your advantage. Tellingly, when consumer juggernauts use RCTs, their most critical metric is often belief in key attributes such as trust in the brand rather than would you buy this product.

If You Want to Touch a Nerve, You Have to Touch a Nerve

When you recognize that the first task of persuasion is to get heard and also believed, vote choice during election time (and approval ratings outside of campaign season) may not be the most illuminating question. Topics that seem toxic and messages that fail to move at first contact can prove effective under repeat exposure. For example, the initial attempt to attack John Kerry’s military record tested badly; it didn’t make voters appear to favor draft dodger George W. Bush. But there’s a reason why “swiftboating” is now a verb. The noxious repetition of “Swift Boat,” delivered by an astroturf group that allowed Bush to seem blameless, chipped away at Kerry’s credibility and therefore his moral high ground on the Iraq War. Compounded with “flip flopper” and videos of him wind surfing, it sealed Kerry’s fate as an elite, effete, untrustworthy character in contrast to the purportedly straight-talking beer buddy Bush. Bush and his crew had a strategic goal — discredit Kerry’s war record — that stemmed from their candidate’s weakness and their opponent’s perceived strength. And they pursued that in service of winning the election.

“Critical race theory” was a meaningless phrase when Virginia voters first heard it, and wouldn’t have made their list of top political issues. Yet it helped land Glen Youngkin the governor’s mansion in 2021 once Republican operatives didn’t just deploy it via ads and lit drops, but engaged in a surround-sound strategy deployed by angry “Moms for Liberty” in televised school board meetings.

On the progressive end, what facilitated the rapid shift from majority opposition to widespread embrace of gay and lesbian marriage is another example of applying strategic thinking to campaigning. From Harvey Milk encouraging people to “come out, come out, wherever you are” to efforts to put beloved gay and lesbian characters on television, advocates recognized they needed a cultural shift before they could successfully pursue legal and policy changes.

Then, in California’s 2008 Proposition 8 ballot initiative, advocates deployed poll-tested messaging with ads that excluded gay and lesbian people, because, according to a senior campaign operative “from all the knowledge that we have and research that we have, [those] are not the best images to move people.” Children were also left out because their inclusion “backfired.” And they lost.

In the aftermath, leaders conducted a complete overhaul, starting with their research approaches, focusing more on qualitative explorations informed by cognitive psychology. They then shifted from “right” to “freedom,” spoke of emotions, showed queer couples, and embraced that this is about families and therefore children. Supporters agreed with the practical rights arguments, but these couldn’t break through nor act as an effective rejoinder to the highly-charged fear mongering coming from the opposition. In contrast, “Love Is Love” and “Love Makes a Family” captured attention, offered a compelling emotional hook against the opposition, and gave everyday people a refrain they wanted to repeat and post about on Facebook. Most importantly, advocates for gay and lesbian equality did not accept the purported unpopularity of their position. They found a new way to persuade people by running on — and not from — their values, despite what initial testing appeared to indicate.

Meanwhile, Pollingism optimizes for what momentarily “moves” an engaged respondent after a one-time read or listen, nearly always without exposure to the opposition messaging that generates an unrelenting din in the real world. These tests tell us what would do best if we had a captive audience to ourselves that actually listened uninterrupted and was handed a ballot immediately, without getting bombarded by the opposition’s counter arguments from all directions first.

Messages that tend to rise to the top “in channel” are those that sound like a reasonable adult giving you a solid argument. This rarely works to get you to stop scrolling long enough to receive a message in the first place. Nor do these tests tend to produce messaging that people can recall, let alone get excited about repeating — an absolutely essential element of success Pollingism doesn’t just ignore, it generally dismisses.

Humans Are Social Creatures

Proponents of Pollingism tend to produce ads and messages that don’t rock any boats (swift or otherwise). The dominant approach from the best-funded campaign operatives is to test messages among the general electorate without extra attention to previous Democratic voters. If you’re scoring by what the greatest number of people in your sample — including those who will never vote for you, despite what they report in an RCT — find palatable, you’re going to come up with something that doesn’t offend in testing but also doesn’t prove memorable once aired.

The Choir – Not the Candidate – Carries the Most Credible Tune

For Pollingism, focusing on what they have decided to label “the base” is “preaching to the choir,” an activity viewed as not just unnecessary – under the theory that progressive Dem voters have nowhere else to go – but detrimental. The presumption is that doing so requires you to say things the rest of the electorate disdains, without considering there may be another way to articulate the case that actually persuades.

Yet, in the real world, it’s precisely said “choir” that makes the joyful noise for the congregation, which goes out and converts new adherents. Unlike ads, what voters’ friends, family, and influencers think, speak, and promote garners their attention and merits credibility.

Further, there is another place that voters who revile the opposition can go: the couch. In every election, there are (at least) three candidates running — yours, theirs, and not voting at all.

In short, Pollingism assumes that what wins elections is persuading the “median voter” (an individual with presumably fixed views despite the fact that both human beings and compositions of the electorate change) that your candidate aligns better with them on their most important issues than the opponent does. Never mind that experimental evidence draws into question whether voters’ stated issue priorities influence their vote choice. And then they proceed to measure which ads about polled top issues move vote choice among swing voters — without considering there may be some other strategy for moving swing voters and, more broadly, winning elections.

In assuming that campaigns are best waged by assessing voters’ issue preferences and making them focal, Pollingism credits voters with knowing candidates’ positions on said issues and/or believing candidates’ claims about them. All of this rests upon the fiction that what people believe about a Democrat running is made out of what that Democrat, or their surrogates and Super PACs, say. Somehow, this fiction also holds that battleground Dems have to overcome the hurdles erected by what their too leftist co-partisans dare to utter, but voters in progressive places aren’t hindered by the Republican-lite messaging crafted for purple districts. And, further still, that America’s notoriously low information electorate can follow the specific ins and outs of the unique “heterodox” positions a down ballot candidate espouses.

So, for example, if your polls tell you the majority of voters want a crackdown on the border, you have your candidate say “I will get tough on the border” in an ad that moved in-survey vote choice. Once again, assuming the candidate doesn’t have their own set views and therefore feels comfortable parroting any survey-tested refrain. But your opponent puts out their own ads, blathers on podcasts, posts on social media and – critically – gets their choir to repeat that your candidate is ready to turn said border into a sieve and lure newcomers with fat checks on “real Americans’” dimes. A message your candidate puts out that gets drowned in rebuttal from the opposition will not seem credible. Meanwhile, you’ve accepted the opposition’s frame that “immigration” equals “border,” without attempting to reframe the electoral conversation on better terms, like attacking MAGA’s monstrous plans for abductions and disappearances, or elevating shared values about freedom and keeping families together.

Politics is a shouting match, not a soliloquy — not just in the sense of who gets heard, but also who gets believed. Testing for a delivery mechanism we do not possess (forced viewing or listening by a captive general public), with presidential disapproval (not willingness to take action or opposition to some specific bill) as the metric, Pollingism now advises avoiding language it deems “hyperbolic,” while masked agents disappear people, soldiers occupy our cities, and the president turns what was a Department of Justice into his department of persecution of political enemies.

More confounding still, during election season, Pollingism proponents advise spending hundreds of millions of dollars on end of cycle ads, as if voter eligible Americans make up their minds based on what politicians, or their aligned PACs, pay to say. Instead, what people believe about politicians comes from what is said about them by messengers who are far more trusted. The virtues a politician extols about themself or their political positions are suspect. Anyone other than die-hard partisans discount this as self-interested. Because it is.

On the other hand, the crap flung at politicians, even by other politicians and much more so by outside entities, is more likely to be credited, especially when repeated over and over again. This is because negativity bias is built into our wiring: negative images, as opposed to positive or neutral ones, produce stronger responses in the cerebral cortex. We are more likely to recall, learn from, and retain negative messages than affirmative ones. This extends to how we make judgments: negative stimuli are considered more critical than positive ones in people’s reasoning. Pile onto this how pre-conditioned voters are to believe that politicians are bad, and negative messages fit into an established cognitive groove.

This is not to claim that all negative assertions a candidate or aligned operatives make are instantly deemed accurate. There is still work to be done to ensure any of our claims pass muster with voters. This is especially critical in our disinformation-deluged media environment where people give credence to lies while distrusting truths. Yet Democrats continue to deploy research methods that fail to even consider what would make campaign claims seem more credible, including which messengers would need to utter them, let alone with what frequency. Instead, vote choice is held as the relevant measure. As we’ll explore below, this led to dismissing the potential utility of bringing the harms and horrors of the MAGA agenda more clearly to the fore.

People Believe What They Think People Like Them Believe

Especially higher up on the ballot, direct political communication from a candidate or aligned-organization is a fraction of the information a person receives about an election and is likely the least consequential. Americans mainly get their cues about politics from family, friends, acquaintances, and media (social and traditional) they trust. Direct candidate and PAC communication that does reach them is often filtered through this trusted network, if and when more politically engaged people share coconut tree memes or wave signs calling for mass deportations, for example.

It’s difficult to effectively mobilize credible messengers on our behalf, but it’s possible with long-term, year round, embedded-in-community organizing. One-on-one conversations to develop political consciousness and turn people into leaders who spread the word to others require serious resources and commitment. This is the kind of work that evangelical churches do for the right and unions used to do even more extensively for the left.

Instead, the dominant Democratic campaign model is to mostly use paid advertising and provide some often non-partisan (501c3) funds to community groups to do “field,” usually limited to registering voters and/or getting out the vote (GOTV). This approach uses poll-tested talking points in “off” years and polished content in campaign season, where operatives who also do production and ad buying control each word uttered and image used. And, of course, the aforementioned RCTs “prove” this to be the best course, offering reassuring point estimates and confidence intervals.

To their credit, Democratic operatives have now expanded their testing scope to include social media content. But in this they still ask largely the same questions — does this TikTok video or Insta post cause more people to check the “Democrat” box in an upcoming presidential vote? (Or at other times, does it move the needle on (dis)approval for the sitting leader or party?) What is not measured is whether the claim stated within the content seems true or causes survey respondents to want to spread the message.

Republicans’ approach has long been completely different. They test for the “red meat.” That is, a message that keeps their base engaged and enraged — ready to repeat a refrain until the muddled middle has it cemented in their minds. This is the precursor to the “manosphere,” which existed in the form of aforementioned evangelical churches, supper clubs, NRA affiliates, and so on long before podcast was a word.

Trump took this basic formula and injected it with steroids. He throws “eating the dogs and cats” at the wall, or promises Mexico will build one at the border, and sees what sticks. Trump won over “persuadable” voters in 2016 and 2024 not by discerning their issue preferences and moderating his message to suit them but by motivating and equipping his base to carry his tune. When you are a low-info voter, seeing people you’ve known forever wearing the MAGA hat tells you that this is what people like you believe. And so you do. Trump’s campaigns have been driven by unleashing, not admonishing, his base and recruiting trusted messengers rather than trying to persuade people principally with ads.

Magnetism, in understanding that voters — like all humans — are social creatures who come to judgments on the basis of what their identity group holds as common sense, requires deeper exploration of and attention to people who have previously voted Democratic. When I say “people who have previously voted Democratic,” I mean literally that group, regardless of their race or age, educational degrees or origins, or even stated partisan preferences.

I emphasize this because, generally, Pollingism adherents view voters through the lens of their demographic categories, with race and educational attainment at the fore. They accurately note that key groups are shifting away from Democrats and in this find evidence that politicians aren’t saying popular things. But instead of zeroing in on the topics and tactics that would drive the voters who last brought us to victory (and not just those who match their demographics) to the polls, the same Democratic campaign operatives who lament our lack of message reach keep producing content aimed at the whole electorate.

Telling Sticky Stories

Good stories, and bad ones, have characters and conflict. But what Democrats claim as stories are often more like grocery lists. Affordable housing, quality jobs, accessible healthcare. We have to write down the groceries we need because they’re not memorable, even as the items on them are life-sustaining.

This listmaking fetish goes back to Democrats’ utter confidence in what Pollingism tells them. So, Democrats apply a simple formula: Ask people what they want to hear about, tell people about that thing, then bask in the popularity boost that results. Even when it reliably doesn’t. Voters always care about living costs, yet the last eight months of Democrats’ determined rhetoric on this has coincided with plummeting approval.

If swing voters say they want bipartisanship and hate divisiveness, Democrats respond by singing the praises of working across the aisle. If voters rank the economy number 1 and immigration 11th, Democrats might beef up their talking points about jobs and tone down the talk about a roadmap to citizenship. And political journalists call them savvy for doing so.

This approach might generate more accurate grocery lists, but it doesn’t tell a memorable story. And it doesn’t create a conversation. It’s one thing to agree or disagree with a policy when a survey asks you about it. It’s quite another to feel passionately enough about an issue to tell your friends and family why they should share your views, let alone act upon them.

In fact, Democrats’ allegiance to what’s known as “wins above replacement” or WAR, highly curated calculations of how well each candidate does relative to the partisan score of their district, means they view each downballot race as a rhetorical island. There is no attempt to have some overarching story about what America is, the Democratic agenda, the Republican threat, and so on, and therefore no chance of setting the discursive weather as opposed to having to battle whatever storm Republicans choose to unleash.

Most Republicans are more successful than most Democrats at telling stories that have virality across the board and credibility with their voters. Virality is how far, how fast, and how durably a story spreads. Credibility is whether people believe a story when they hear it — which has little to do with how accurate the story is and a lot to do with whether a story affirms people’s priors, identity, and desires. A story also gains more credibility if the mainstream media treats it like a legitimate controversy with two sides.

The storming of school boards over CRT was dramatic and attention-getting enough to draw coverage, which reinforced the cycle of both virality and credibility. The world defined by Pollingism simply doesn’t include these kinds of real-world consequences and feedback loops.

Taking the Temperature Versus Changing It

The mistake Pollingism demands that we make is to take the temperature of the electorate and calibrate what to say on this basis. But taking the temperature today doesn’t help us to know, much less control, where it will be set tomorrow. It puts us behind in both electoral and movement strategy — failing to dominate daily conversation, and neglecting to steer longer-term public opinion.

While Democrats are brandishing thermometers, Republicans power up flamethrowers. They figure out not merely “where people are,” but rather where they could be capable of going, if they hear some refrain repeated over and again.

MAGA’s stories are potent and memorable, not about the issues people register as most important but rather those that MAGA wants brought to the fore. This is Magnetism 101: make sure you’re fighting on the terrain that benefits you most. This means setting the terms of debate by every possible means by engaging in activities that bring your preferred conversation into people’s ears.

You Only Get Answers to the Questions You Ask

In contrast, Pollingism relies on survey results to determine preferred topics. In these, voter-eligible Americans are given lists of issues and rank their importance. By definition, this brings these ideas into people’s heads. When a working-class white guy in Phoenix is thinking about the water heater part he needs and when he can check on his mom before making it home in time for his wife to leave for her night shift, he doesn’t contemplate trans girls playing volleyball. But if you present questions about this to him, he may well register concerns. As such, merely in formulating certain survey questions, you can get a false sense of what matters to people. This is all the more true given the practice of “pollwashing,” rendering more palatable the horrific plans of the current regime by asking about them in bland terms like “getting tough on the border” or labeling the abduction of people “immigration policy.” It should have come as no surprise to us that Americans are now firmly against the horrors MAGA has unleashed on immigrants; the only reason they appeared to endorse them pre-election was that questions on the topic were worded in the most sanguine ways.

In surveys, “the economy” reliably features as people’s top issue. But this does not translate to meaning that paid ads on this topic are how to get more people to want to vote, let alone for you. Especially if those ads rely upon vague abstractions and not bold breaks from status quo economic policy your candidate is promoting that you repeat over and over. Promising what your opposition labels “price controls” gets conversation going; pledging an “opportunity economy” does not. Although thousands of ads tested may reveal that it shifts vote choice to say your candidate will make things better for working people in contrast to an opponent only out for the rich, this most quotidian of topics is unlikely to break through especially without some novel and concrete policy promise, and thus your ads on it are not noteworthy nor memorable.

Further, in choosing to concentrate your resources on this you are conceding “the economy” as the dominant electoral issue without even trying to bring something else where you hold a far greater advantage to the fore. Economic concerns are critical to voters. But they also fault the party in power for their money woes. There is nothing a candidate can say that is as loud, and as convincing, to a voter as what their experiences “prove” true. Voters may not pay much attention to ads, but they sure do to bills and grocery receipts. It’s worse yet when you’re facing off against an opponent who signed voters’ stimulus checks when he was last in office and has his loyal choir captioning photos of prices at the pump and on the shelves with angry screeds about Democrats’ economic plans.

Let’s Give Them Something to Talk About

A temperature-taking strategy could never have predicted or produced the power or virality of “Willie Horton” and “welfare queen” back in the day or “DEI” and trans panic today. In all of these cases, Republicans noticed a story already being told that struck a chord with a base all too happy to evangelize it. This came with two bonuses: discrediting the racial justice and LGBTQ equality activism that had started to appeal to some white suburban voters while distracting from the real hardships that union-busting, corporate capitulating, program-slashing Republicans had created.

There are also instances where Democrats do break from middle of the road messaging and seize the mic on their own terms. Among these is Lyndon B. Johnson’s 1964 Daisy ad, so controversial it was pulled from air after one viewing, and so potent it was talked about and remembered nevertheless. In the ad a little girl innocently counts as she pulls petals off of a daisy, which become the countdown to a nuclear bomb with an unseen narrator saying, “these are the stakes. To make a world in which all of God’s children can live, or to go into the dark. We must either love each other, or we must die.” Without uttering Republican candidate Barry Goldwater’s name, Johnson set the terms of electoral debate as life or death. Of course, LBJ would have won regardless, but this memorable ad contributed to a landslide for Dems that is the reason we have Medicare (or what’s left of it, at least for now). The Great Society programs were made possible because of the size of the electoral victory and its implied mandate for social progress.

Barack Obama’s 2008 viral web-only “Yes We Can” ad shows that aspirational approaches, not merely negative ones, can also perform this agenda-setting task. Featuring celebrities singing along to Obama’s New Hampshire concession speech from the primary, the ad begins, “It was a creed written into the founding documents that declared the destiny of a nation. Yes we can. It was whispered by slaves and abolitionists as they blazed a trail toward freedom. Yes we can. It was sung by immigrants as they struck out from distant shores and pioneers who pushed westward against an unforgiving wilderness. Yes we can. It was the call of workers who organized; women who reached for the ballots; a President who chose the moon as our new frontier; and a King who took us to the mountaintop and pointed the way to the Promised Land.”

There is a pattern to breakout moments by Democrats: they are unapologetic. Jasmine Crockett’s frequent MAGA takedowns, Zohran Mamdani’s populism, Andy Beshear’s refusal to take the anti-trans bait, Chris Van Hollen’s trip to El Salvador, and J.B. Pritzker’s repudiation of the regime, to name a few. There is a reason why some of these are among the only Democrats voters recall and admire: they are not saying what is poll-tested as “popular” but rather making popular what needs to be said.

Meanwhile, Republicans proudly trumpet their plans to set the conversational terms, secure in the knowledge that this won’t afford Democrats any advantage because they rely on surveys to tell them what to support and say. Bill Stepien, Trump’s 2020 campaign manager boasted in the New York Times, “Will the economy always be top of voters’ minds? Yes. But did voters care as much about trade, immigration, or crime pre-Trump as they do now? No, and that’s because Trump has made people care about things he thinks are important.” Stepien went on, “So don’t assume President Trump will be reacting to any issue next November, because he will be trying to drive the discussion in his direction.” Critically, Trump is bringing crime to the fore by sending troops to cement the story that troops are needed in Democratic run-places.

We witnessed the same rinse and repeat with attacks on DEI. Republicans changed the temperature with a complete narrative strategy. They did not merely air ads tailored in “WAR” fashion to each district come campaign time; they humiliated university presidents in televised Congressional hearings, organized their base to bring these issues up locally, and unleashed media personalities to keep beating the drum, despite the fact that no poll shows voters registering DEI as a top issue. In a Magnetism approach, you draw your supporters in by trumpeting causes you believe in that compel them to sustained action, in order to make them of more mainstream interest. And you revel in repelling the voters who will never support you because this is what forces your political rivals onto hostile terrain — hapless against the conversational terms you have set.

Despite the right being a one-hit divide-to-conquer wonder, the Democrats act like it’s brand new every time Republicans sing this same tired tune. Because, again, Pollingism directs them to speak about “popular” issues, and when others come up to deflect back to the same. This leaves Democrats on the defensive.

Consider the fractured and ineffective response to the demonization of CRT: avoiding the conversation or insisting it wasn’t being taught, thereby implying there’s something wrong with it. When pressed on the topic in 2021, for example, VA gubernatorial candidate Terry McAuliffe made the debate gaffe repeated round the state: “I don’t think parents should be telling schools what they should teach.”Democrats presented no alternative, affirmative story to voters, nor did they clap back at Republicans’ dog whistle politics even though there is a proven approach for contending with this predictable pattern.

The same goes for the border in 2024; Democrats tried to out-tough Republicans, which swing voters didn’t buy and base voters found off putting. And on again for the infamous “she’s for they/them” ad. Despite the fact that the right wing trumpeted their plans to go on all out assault against trans people once their attacks on Marriage Equality stopped yielding fruit, most Democrats failed to deploy effective rejoinders.

Indeed, Democrats often help amplify the opposition’s storyline. Back in the day, for example, they did this by pledging to “end welfare as we know it,” thus crediting a slash-the-government approach associated with the right and demonizing fellow Democrats who had the audacity to support people in need. And, more recently, they tried to out-tough Republicans in ads that repeated “the border” over and again while hammering home the notion of immigrants as villains. Candidates ought to run against their opposition, not the people the opposition scapegoats and reviles.

Democrats are especially likely to get caught off-guard when Republicans repeat the unflattering stories Democrats also tell about themselves. From WelcomeFest to Third Way to Searchlight, Democrats who claim to be concerned that voters reject their party for being too “woke” keep publicizing these accusations, seemingly to ensure voters hear them. Repeating Republican attacks against the party they purport to want to help makes the charges seem more credible and helps them spread farther, especially when the media covers the “controversy.”

In short, reliance on Pollingism has left Democrats in far worse shape than “not having a message;” they’ve failed to adopt a modern communications strategy. They spend so much time dissecting opinion surveys and defending themselves against attacks that they don’t focus on what actually wins elections: making them about the things that you can win on. Let alone what has proven in other times and places effective at ending authoritarian regimes: toppling the pillars of support that prop them up by amassing and sustaining mass non-violent civil resistance.

Consider the 2022 midterms. In the states where MAGA’s threat to our freedoms was front and center on voters’ minds, Democrats defied historic expectations. But where Democrats took Republican bait and let crime and price conversations remain front and center, Democratic losses were perfectly predicted by past patterns.

Yet, throughout 2021 and 2022, Pollingism believers kept insisting that neither emphasizing MAGA nor abortion were good approaches, and Democrats’ only hope was finding better messaging on crime and inflation. Indeed, this same crowd insisted well into 2022 to avoid abortion — often calling it the “a-word” and counseling Dems it was too divisive. All while watching Republicans turn abortion into a core mobilizing issue, and thus ceding all terrain on the topic such that conflicted voters were hearing one drum beat of emotion-fueled condemnation and very little else.

It took the results of the Kansas referendum, rather than polling, to bring them around to temporarily declaring abortion a “popular” issue. Yet, even after abortion ballot initiatives beat the Democratic presidential contender by double digits in Arizona, Florida, Missouri, and Nevada in 2024, pundits and operatives are again deeming it wise to waffle on abortion.

Republicans create the conversations that work to their advantage; Democrats respond to whatever Republicans put forward — or what the survey questions they formulate say voters prioritize. Democrats win when reality breaks through — after the 2008 crash, or in 2020 when the Covid horrors of Trump were fresh — but they don’t do a good job of foregrounding the topics that benefit them.

Indeed, the Pollingism they default to often has them structure their queries so narrowly, they fail to consider the full range of topics, let alone articulations of them, it could be possible to make salient.

Falling Into the ‘Credulity Chasm’

One staple in various deep dives about the 2024 election is an argument over whether Democrats needed to focus on the economy versus making democracy the central issue. This debate seems to occur without recognition there are infinite ways to speak about either of those concepts. “Team economy” believers found voters largely unmoved by “democracy” talk. This is little wonder: Democracy is far too important a concept to try and sell with meaningless abstractions like “democracy.” Messaging on any topic has to be presented in terms of how it will impact the voter.

To most voters, “democracy” means the current system in which they are living. When a white family could afford the picket fence, retirement, and an annual road trip on a single salary, upholding the status quo sounded pretty good to them. But today, when millions are struggling to afford the basics let alone imagine a mortgage, retirement or vacations, “democracy” sounds like accepting the screwed up system voters dislike.

But economy or democracy as concepts, let alone wording choices, are far from the only options as focal points for a campaign. In an election in which an unpopular incumbent party that voters blamed for their economic pain was facing down the return of an even more potent right-wing regime, making clear exactly what MAGA rule would mean was another option.

Pollingism proponents maintain that according to the large volume of ads they tested, vilifying Trump did little to move the vote choice needle. The explanation for this is that he was already so well known, negatives about him were “baked in.” And so they concluded that anti-Trump talk was a failed pursuit instead of considering that perhaps the ways they had tried making these arguments, from messengers to messaging, to means of delivery, was the issue.

My colleagues and I, too, found that efforts to attack Trump’s character failed to move voters in our own experiments. Indeed, we found that “Trump is” statements — i.e. assertions about his beliefs, moral integrity and mental competence — fell short. In other words, focusing on his traits did little, and in qualitative testing before Biden left the race, they invited rejoinders of “yeah, well Biden is…” with unflattering sentence endings. But this ignores the fact that there are other ways to make clear the danger Trump posed. We found that “Trump will do” statements appeared far more effective.

In fact, in 2024, organizations across the Democratic and progressive spectrum engaged in one of the most robust and collaborative research efforts the left has undertaken to examine how to speak about Project 2025 most effectively and bring its most noxious elements to the fore. However, Pollingism practitioners, despite their avowed attachment to data, ignored these widely replicated findings and thus left this out of the messaging they deployed in the campaign.

Further, volumes of data before November and since show that what most clearly differentiated the Harris voter from folks who did something different (selecting Trump, third party, or staying home) was whether or not they believed the MAGA Project 2025 agenda would actually come to fruition. Not their desire for said agenda, not their hunger for an “opportunity economy,” or even middle class tax cuts, but whether or not they took Trump at his word that he would unleash the horrors that, now visited upon us, have proven quite unpopular.

We named this phenomenon the credulity chasm. In August of 2024, for example, we found that 58 percent of Democratic voters believed that Republicans would implement the Project 2025 agenda. Only 21 percent of Republicans did. Our post-election survey showed more of the same. As I wrote in Rolling Stone, “Harris voters broke 87/13 on whether it’s likely versus unlikely that Trump will deport millions of immigrants including those here legally, but Trump voters split 60/40 on this. Only 18 percent of Harris voters think a national abortion ban is unlikely, whereas 69 percent of Trump voters give this response. Indeed, on our battery of what is likely to happen under a Trump administration and the attendant outcomes, Harris and Trump voters are almost mirror opposites in their predictions.”

Of course, it’s possible that the campaign itself and affiliated Super PACs couldn’t credibly convey the threats because voters would view their warnings as hyperbolic and self-interested. But this again is where strategy is essential. Instead of asking what are the most RCT-certified “persuasive” ads we can make on the topic voters tell us matter most, we should instead be considering what is the most potent argument in our favor and how we can get it out into the world.

While 2024 may have rendered the credulity chasm the Grand Canyon, it’s very common that effective persuasion requires voters to change their minds about what is true today and what will come to pass next year. And this is precisely what cannot be gleaned if you base most of your intel on measuring vote choice after voters are made to view an ad a single time.

Ironically, Democrats are all too familiar with the potent impacts of voters thinking something is true merely because it is familiar, a cognitive bias known as the illusory truth effect. It can be encapsulated in the phrase “tough on crime.” Republicans make it their business to reshape voters’ realities; it’s talk of crime rather than crime itself that determines how much crime voters believe is happening.

Could Democrats have made voters believe what Trump and his MAGA minions put into 920 published pages would come to pass? To be sure, the Harris campaign made Project 2025 a centerpiece at the convention and attempted to bring it to the fore throughout the campaign. But, again, it’s possible this message couldn’t come directly from the candidate, nor from surrogates like Liz Cheney. Critically, the greatest concentration of outside resources did not focus on pushing this information, nor did Democrats apply the kind of 360 degree actions and activists and ads approach Republicans always use. This would have included, for example, Congressional hearings with Project 2025 architects, press conferences about specific pieces of the agenda, earlier and more robust support for innovative efforts such as having adult film stars convey warnings about the end of porn, and in community organizing to get local folks to bring it up to friends.

Would this have yielded a different electoral outcome? It is impossible to say. But the fact remains we limit what it’s possible to try and are left evaluating which among the very constrained options attempted before merit redo the next time. And the same small group of Pollingism practitioners are still trusted at the helm to solve the problems they didn’t fix with hundreds of millions of dollars last cycle.

Democrats have been overperforming in 2025 elections. And, yes, off-year voters are higher information than their presidential-only counterparts. But we are also seeing much higher turnout — in other words, bringing non-habitual voters into the electorate — than is common. What has changed? It’s certainly not Democratic approval ratings. Instead, the credulity chasm is closing and many voters can now see that MAGA is indeed doing as promised to destroy our freedoms and hand the wealth our work creates to their billionaire backers.

Conclusion

So here we are, with many Democrats responding to each new horror unleashed with a social media post attempting to drive conversation back to what surveys say voters care most about. Instead of speaking about what’s actually occurring, these Democrats have wedged themselves into an impossible corner where what they say barely gets heard and what gets heard can barely be credited. You cannot occasionally tell Americans the truth that this regime is veering into dictatorship while also promising to meet and negotiate in good faith with it. You cannot convey the full extent of the threat before us, let alone marshall public will into needed resistance, when your actions (or lack thereof) routinely undermine your utterances.

We often hear of the need for a “big tent” to combat authoritarianism, a term that seems to mean welcoming in more people but not immigrants and their allies, not LGBTQ people and their loved ones, not women who’d like some reproductive autonomy and their champions. And, yes, it will take a mass movement to bring down this regime. But for a big tent to stand, it requires a central rod — otherwise it’s a tarp that smothers you. To win this battle, we must run on — not from — the broadly-held values most Americans cherish and make clear we are in a battle against fascism, not just inflation.