Polls are everywhere. The companies pricing the election aren't being examined.
Every outlet in this story is covering the midterms. Only one is covering who profits from them. The Intercept reports that Kalshi and Polymarket are actively dodging state gambling regulations while setting the terms by which political insiders and media read electoral probability. That regulatory arbitrage is shaping how the race is perceived, not just wagered on. The rest of the coverage treats prediction market odds as neutral data. They are not neutral data.
Midterm elections: polls, predictions and political landscape
4 sources · hover a dot to see coverage
What happened
With the November 2026 midterm elections roughly a month out, coverage has consolidated around three tracks: voter sentiment, individual race polling, and post-election scenario planning. The Guardian published a nine-voter feature identifying cost of living and the Iran war as the dominant issues across the political spectrum. The Hill released Iowa polling showing Republican Rep. Ashley Hinson leading Democrat Josh Turek in the open Senate race, while Democratic state auditor Rob Sand holds an edge over Republican Zach Lahn in the governor's race. The Washington Examiner examined Nebraska's 2nd Congressional District, where retiring Rep. Don Bacon's absence gives Democrats their clearest opening in a decade. The Hill also published a forward-looking piece on Democratic Party internal divisions that would follow a House majority win. The Intercept reported separately on Kalshi and Polymarket operating prediction markets on the elections while avoiding state regulatory oversight. That is the wire version. Five outlets have it. What separates the coverage is what each outlet decided counts as the story.
The Intercept found the mechanism everyone else is quoting without examining
The Intercept's piece on Kalshi and Polymarket is the only reporting in this set that treats prediction market odds as a subject rather than a source. The finding is specific: both platforms are sidestepping state gambling regulations while their probability figures circulate freely in political media as if they were polling averages. The piece names the regulatory gap, not just the existence of the markets. Every other outlet in this set either ignores prediction markets entirely or would cite their figures without disclosing that the companies producing them are in active legal gray zones.
The Hill ran both the horse race and the morning-after, which is unusual for one outlet
The Hill published two pieces that sit at opposite ends of the electoral timeline: an Iowa poll with specific candidate names and margins, and a forward-looking analysis of Democratic intra-party conflict that presupposes a Democratic House majority. The Iowa numbers are concrete: Hinson leads Turek, Sand leads Lahn. The post-election piece is more speculative, but its publication before the election signals that Democratic division is already being treated as the story after the story. That framing choice, running the aftermath before the result, is worth noting.
The Guardian's voter feature named the Iran war; no other outlet in this set did
The Guardian's nine-voter feature is the only piece in this set that names the Iran war as a top voter issue. The other four outlets cover Iowa Senate margins, Nebraska district history, prediction market regulation, and post-election party dynamics. None of them mention Iran. Whether that absence reflects editorial bandwidth, geographic focus, or something else is not clear from the coverage alone. What is clear is that a foreign policy conflict significant enough for Guardian voter interviews to surface it does not appear in any domestic political outlet's midterm framing here.
What one side told you that the other didn't
Prediction market regulation is a midterm story. Four outlets missed it.
The Intercept is alone in this set reporting that Kalshi and Polymarket are navigating around state gambling laws while their odds figures become shorthand for electoral probability across political media. The other four outlets treat the midterms as a story about candidates, voters, and margins. None of them ask who is setting the probabilistic frame through which those candidates and margins are being read. When prediction market odds appear in political coverage without disclosure of the regulatory disputes surrounding the platforms producing them, the reader is getting a number without its provenance.
The Iran war appears once, in a voter feature, nowhere in political analysis.
The Guardian's voters name the Iran war alongside cost of living as their top concerns. The Hill's Iowa poll, the Washington Examiner's Nebraska piece, and The Hill's post-election analysis do not mention Iran at all. If a foreign war is shaping voter sentiment enough to surface in qualitative interviews, its absence from quantitative polling frames and district-level analysis is a gap, not a coincidence. The race-by-race coverage is being built on issues that may not match what voters say they are deciding on.
Nebraska's open seat got examined. Dozens of other open seats didn't.
The Washington Examiner's Don Bacon piece is the only district-level race analysis in this set beyond Iowa. Bacon's retirement after holding a Biden-won district for a decade is a legitimate story. But the piece's existence alongside silence on comparable competitive districts elsewhere suggests coverage is gravitating toward familiar names and established narratives rather than mapping the full competitive landscape. One retiring incumbent is not a midterm overview.
What to watch
If Kalshi or Polymarket face a state-level enforcement action or congressional inquiry before Election Day, watch whether the outlets currently citing their odds as neutral data update their framing or continue using the figures without disclosure. The Intercept has already filed the predicate reporting. The question is whether a regulatory action forces the rest of the coverage to catch up.
See how outlets across the political spectrum framed this differently — and what each side left out.