The 2026 midterm election is already upon us and it is shaping up to be a wild ride.

It’s only August, and like Halloween, the 2026 midterms are already upon us. After fabricated polls, phantom PACs, suspicious candidates and the sudden rewriting of political history, voters should prepare for the unexpected.

The 2026 midterm campaign has already produced a polling debacle, an admitted polling hoax and an assortment of political schemes designed to manipulate primaries, ballots and public perceptions. Now some of the Democratic Party’s most popular figures are backing away from positions they championed during the “Woke 1.0” era.

These developments belong to the same story. Modern campaigns are not merely trying to persuade voters. They are trying to control what voters see, which candidates they face and what they remember about the recent past.

Start with Wisconsin, where poll after poll showed democratic socialist Francesca Hong leading the Democratic gubernatorial primary. Marquette University’s final survey placed Hong at 40% after including voters leaning toward a candidate, almost exactly matching her eventual 39.4%. But it gave Milwaukee County Executive David Crowley only 12%.

Crowley won with approximately 39.8%.

A mysterious organization called Median Strategies did even worse, publishing a purported survey showing Hong ahead by 23 points. After the election, Median admitted the poll was fabricated. So was another showing Los Angeles Mayor Karen Bass comfortably ahead of Nithya Raman.

Median called its operation a “short-term social experiment” testing how unverified polling could enter political discourse. Perhaps. It could also have been a prank, an influence operation, an attempt to manipulate prediction markets or a failed effort to establish a polling brand by inventing plausible results.

Whatever the motive, the experiment succeeded. Build a professional-looking website, invent 560 respondents, calculate a plausible margin of error and attach some demographic tables. Campaigns, partisan accounts and news organizations may repeat the result before anyone verifies that the pollster — or the poll — even exists.

Legitimate polling is different from outright fabrication. But legitimate polls have their own credibility crisis.

Republican pollster Patrick Ruffini examined more than 3,000 Senate polls from the past four election cycles and found that 73% overstated the Democratic margin relative to the final result. The average Democratic overstatement was 3.7 points. It approached six points in late July and remained approximately 2.6 points even around Election Day.

Not every difference between a July survey and a November result is polling error. Voters change their minds, and Republican electorates often consolidate late. But a 73–27 split is not random noise.

Polls do not actually sample “all adults” or even “likely voters.” They sample people within those categories who are willing to be polled. Conservatives suspicious of media, universities and political institutions are especially likely to refuse. Pollsters can give additional weight to the few working-class Republicans who participate, but that assumes the cooperative Republican resembles the ten who ignored the call.

If he is more moderate or more trusting of institutions, weighting him more heavily simply amplifies an unrepresentative respondent.

This helps explain how Reuters/Ipsos recently measured President Trump at 33% approval while Rasmussen placed him at 44%. The difference between adults and likely voters accounts for only part of that chasm. Polling averages put Trump near 39%, suggesting Reuters was unusually negative and Rasmussen unusually favorable.

Yet Reuters’s “lowest approval of his presidency” headline traveled around the world. A poll need not be fabricated to function as propaganda. Select the most dramatic result, ignore the average and report a statistically insignificant two-point fluctuation as a national collapse.

Polling is not the only method of manufacturing political reality.

In Pennsylvania’s competitive 7th Congressional District, lawsuits allege that Green Party candidate Andrew Tupone was a registered Republican until February, received Republican help getting onto the ballot and was never properly nominated by the Green Party. Tupone insists he is sincerely running on progressive issues.

If his candidacy is genuine, he has every right to run. If Republicans recruited him to siphon votes from the Democrat, voters deserve to know. In a close House race, a few thousand votes diverted to a supposed Green candidate could determine control of Congress.

Alaska’s Senate primary offered an even stranger case: incumbent Republican Sen. Dan S. Sullivan appeared on the ballot with retired teacher Daniel J. Sullivan Jr. Republicans alleged the second Sullivan entered the race to confuse voters and help Democrat Mary Peltola.

The challenger and Democratic organizations deny coordination, and courts correctly ruled that officials could not invent a new “good faith” qualification to remove an eligible candidate. But the circumstances were suspicious enough to prompt a federal investigation. The challenger changed his party affiliation before filing, initially sought a nearly identical ballot name and used campaign materials resembling the senator’s.

Elsewhere, both parties have manipulated filing rules. Democratic Rep. Chuy García and Republican Sen. Steve Daines waited until the final moments to end their reelection campaigns, leaving preferred successors with little or no meaningful primary competition. Democrats have withdrawn Senate nominees in several red states to clear the field for independents. Republican operatives have funded progressive-sounding “pop-up” super PACs to promote radical or scandal-damaged Democrats.

AIPAC has supported preferred Democratic candidates through benign-sounding organizations such as Chicago Progressive Partnership and Elect Chicago Women, apparently recognizing that its own name has become politically damaging among progressives.

Supporting candidates is legal. Concealing who is doing the supporting transforms persuasion into manipulation.

Then there is the sudden retreat from “Woke 1.0.” This is not merely an attempt to escape electoral consequences. It reflects something more powerful in celebrity politics: a vibe shift.

During the years following 2020, progressive cultural status came from moving left faster than everyone else. Politicians, journalists, corporations and entertainers competed to place themselves on the “right side of history.” Police reform became police defunding. Equal treatment gave way to elaborate racial and identity hierarchies. Skepticism was treated not as disagreement but as evidence of moral failure.

The incentives have changed. Corporate DEI programs are disappearing. The vocabulary of 2020 increasingly sounds dated and faintly ridiculous. Politicians who once promoted police defunding now speak reverently about public safety. The fashionable position is no longer wholehearted participation in Woke 1.0, but amused sophistication about its excesses.

Popular people do not necessarily care about remaining on the “right side of history.” They care about remaining on the right side of prevailing popular opinion. Once the cultural winds change, yesterday’s moral certainty can become today’s awkward old social-media post.

Rep. Alexandria Ocasio-Cortez understands cultural fashion better than almost anyone in Congress. Her influence rests not merely upon votes or legislation but upon attention, popularity and bankability. She can start trends, and she can follow them. What she cannot do for long is buck a cultural shift after popular opinion has moved decisively against her position.

Ocasio-Cortez recently laughed about the period, remarking that “Woke 1 was crazy.” She now treats some of the rhetoric surrounding police defunding and abolition as a product of the pandemic era rather than an enduring political commitment.

New York Mayor Zohran Mamdani has made a similar “journey.” In 2020, he described the NYPD as “racist, anti-queer” and a “major threat to public safety.” He called for defunding and dismantling it. Once running for mayor, he declared that police have a critical role in public safety. He apologized for his earlier language, promised not to reduce police headcount and retained Police Commissioner Jessica Tisch.

Politicians are permitted to change their minds. Experience can produce maturity. Mamdani now bears responsibility for governing a city rather than pleasing an activist audience, and that responsibility may genuinely have changed him.

But neither he nor Ocasio-Cortez should be permitted to treat Woke 1.0 as an embarrassing fashion trend that appeared spontaneously and disappeared without consequence.

People lost jobs for questioning DEI orthodoxy. Parents were denounced for objecting to ideological school policies. Police departments were defunded or demoralized while violent crime rose. Businesses burned during riots that many journalists insisted on describing as mostly peaceful. Ordinary disagreements about biological sex, merit and public safety were treated as evidence of hatred.

Ocasio-Cortez and Mamdani were not helpless observers of that movement. They helped give it political power. Now that the cultural atmosphere has changed, they would prefer voters to judge only their latest formulations.

Yet AOC faces danger from the other direction. The democratic-socialist movement that helped make her famous has already demonstrated its willingness to reject her for insufficient ideological purity. Its newest candidates can accuse yesterday’s revolutionary of becoming today’s establishment politician.

She must remain radical enough for the Democratic Socialists — the new popular kids in parts of the party — while becoming sensible enough for a broader audience that has tired of Woke 1.0. Staking out a position that has become culturally unpopular would erode her influence and bankability. Moving too far toward the center, however, could cost her the movement that created her political identity.

Her “Woke 1 was crazy” formulation performs both tasks. It distances her from the embarrassment without repudiating the movement, admitting responsibility or explaining what she now believes.

That is not quite an apology or even a policy reversal. It is narrative laundering. A political leader who helped establish the old fashion now presents herself as one of the first sophisticated people to notice that it has gone out of style.

This is propaganda in its subtler form. It does not require inventing a poll or making a demonstrably false claim. Yesterday’s radical demand simply becomes today’s misunderstanding, joke or context-dependent remark. The politician claims to have evolved without identifying what was wrong, who was harmed or what principles now govern the revised position.

Voters should welcome genuine reconsideration. But genuine reconsideration requires candor. Otherwise, the transformation is merely rebranding — a way to retain the popularity earned from leading one political fashion while escaping responsibility when that fashion changes.

The 2026 elections may produce a conventional backlash against the party in power. They may produce an unexpected Republican recovery if current polls are again overstating Democratic strength. They may produce a divided Congress. What cannot be trusted is the carefully manufactured political environment surrounding those possibilities.

Unknown pollsters are inventing surveys. Established pollsters disagree by double digits. Operatives are disguising PACs, promoting opposition candidates and exploiting filing deadlines. Politicians are rewriting their own histories before voters can hold them responsible.

The dirty tricks are not waiting for October. Neither is the propaganda. The midterms have already begun.

(Contributing writer, Brooke Bell)