Health Freedom, Healthy Food, Peace: Four Real Races in a Three-Seat House — and the Blind Spots Everywhere Else

The starting map: where the margin is actually thin enough to matter

This is the baseline every later section narrows down from. Cook Political Report, Sabato's Crystal Ball, and Inside Elections all publish independent race-by-race ratings; none publishes a numeric win-probability model — only discrete categories (Toss-up, Lean, Likely, Solid). All three have moved in the same direction — toward Democrats — more or less continuously since spring 2026, which matters directly for where focus pays off: a race trending away from you needs defending more urgently than one trending your way.

Three things worth reading carefully rather than skimming past, because they change where energy is actually needed: NC-11 and OH-09 are both genuine, real disagreements between the outlets, not data errors — Cook rates OH-09 more competitive than Sabato or Inside Elections do, the reverse of NC-11's pattern, meaning neither race should be written off or assumed safe on one outlet's word alone. And IA-01's move to Leans Democratic is the single freshest change in this whole set, independently confirmed within the last day of this page being built — a race that just got harder to hold, not easier.

How close these races actually were in 2024 — the real heat map

Ratings categories (Toss-up, Lean, etc.) group races into buckets; actual vote margins show exactly how thin the ice is underneath each one. Sorted tightest to widest, colored by how close the 2024 result actually was — the darker the tile, the smaller the margin, the more a small shift in turnout or persuasion could change the outcome.

All 2024 results here are for the district as it existed in 2024 — most of these lines are unchanged for 2026, but OH-09 was redrawn afterward to favor Republicans by about 4 points, so Kaptur's 2024 margin of 0.64% understates how much tighter (for her) the 2026 rematch against Merrin actually is on the new map. NY-04's exact 2024 margin has conflicting reports (roughly 1,000 to 6,000 votes depending on the source) and is shown here at the midpoint of that range.

The four races where a MAHA endorsement could actually help win the House

MAHA's name stands for the health-freedom and healthy-food pillars this page tracks, backed by two organizations: MAHA PAC, a political committee, and MAHA Action, a larger advocacy organization. Between them they've endorsed candidates for dozens of offices nationwide — but with a House this close, what matters is which of those endorsements sit in a race actually competitive enough to move control of the chamber. Four do. Here's each one, and the specific reason an endorsement could matter there.

Anna Paulina Luna (FL-13) deserves a separate mention even though her own seat is safe: she's the one who sponsored the pesticide-liability amendment covered below, so her MAHA endorsement is already backed by a concrete legislative result, not just a position.

Health freedom: separating a real record from a reputation

This is the health-freedom pillar — COVID-19 institutional accountability. The point of this section is to tell apart members who have actually cast a checkable vote on these issues from members who merely have a reputation for it. The clearest differentiating vote on this subject is from the prior Congress, not the current one — reported here in full because it's real, primary-sourced, and shows genuine daylight between members, which is exactly what a reputation alone can't show.

H.J.Res.7 — ending the COVID-19 national emergency, Feb 1, 2023

Passed 229–197. All 218 voting Republicans supported it; 11 Democrats broke from their party to join them. Members not yet in Congress at the time are marked accordingly rather than left blank. Full bill text and history on Congress.gov.

Seven tracked members (Evans, Gillen, Mackenzie, Bresnahan, Gray, Tran, Barrett) weren't yet in Congress in Feb 2023 — correctly excluded, not counted as "no" votes.

A second, related vote the same week shows the crossover doesn't run deep: H.R.382, the "Pandemic is Over Act" (Jan 31, 2023), targeted the separate Public Health Emergency declaration specifically and passed 220–210 on a strict party-line vote — Don Davis voted Nay with his party on this one, meaning his bipartisan vote on the national-emergency question didn't extend to the public-health-emergency question. It never passed the Senate and never became law.

H.R.185 — ending the CDC's vaccine-proof requirement for foreign travelers, Feb 8, 2023

Passed 227–197. A narrower, distinct vaccine-mandate vote — not a restatement of the emergency-powers votes above, since this targets one specific CDC policy rather than a broad emergency declaration. All 220 voting Republicans supported it; 7 Democrats crossed over.

A House committee (the Select Subcommittee on the Coronavirus Pandemic, chaired by Brad Wenstrup) produced a 520-page final report in December 2024 — but that report only ever received a committee markup vote, never a House floor vote, and the subcommittee was dissolved at the end of that Congress without being reconstituted.

Two other votes sometimes cited as "accountability votes" don't hold up on inspection: a 2023 vote to ban federal funding of gain-of-function research passed by voice vote (no member positions were ever recorded), and a 2023 vote ordering declassification of Wuhan Institute intelligence passed unanimously, 419–0 — before this Congress, and before it could distinguish anyone.

Healthy food: proof the cross-party model already works

This is the healthy-food pillar — pesticide-industry accountability — and the closest thing on this whole page to a demonstrated result, not a hoped-for one: a real bipartisan majority that actually stripped an industry protection, worth understanding in detail before looking anywhere else for where energy might work. Roll Call 148, H.R.7567 (the 2026 Farm Bill), April 30, 2026 — an amendment removing provisions that would have shielded pesticide manufacturers from state lawsuits and blocked local pesticide regulation near schools. Passed 280–142. The amendment carries Rep. Anna Paulina Luna's name on the floor, but multiple outlets describe it as mirroring language originally pushed by Reps. Chellie Pingree and Thomas Massie. Luna is herself a MAHA Action-endorsed member (safe FL-13 seat), a direct link between the endorsement picture above and this specific vote.

15 of 21 tracked members voted to strip the shield; 6 voted to keep it — evenly split, 3 Republicans and 3 Democrats. Tom Barrett (MI-07) — the same Republican who crossed party lines on the Iran war-powers vote below — also voted here to strip industry protections. That's a second, independent data point on the same person, which is exactly what turns "endorsed" into "proven."

The single clearest blind spot: money that contradicts the framing

This is where "close race + Republican + physician" could be mistaken for "health freedom opportunity" — the data says otherwise. Verified directly against OpenSecrets' campaign-finance data, 23 Sept 2026 (Miller-Meeks, Barrett, Perry). Total "Health" sector money, individual vs. PAC contributions. Note the scale: Miller-Meeks's bar is drawn true-to-size against the other two, not on a separate axis — the difference is the finding.

Miller-Meeks's health-sector money is roughly 8x Barrett's and Perry's combined — a real, verified gap, not an estimate, and the single strongest reason not to treat her tossup race as a natural fit for this coalition's energy despite how close it is. A widely repeated claim that Perry took "$100,000 from drug and insurance industries" does not hold up against the actual data either: his entire health sector, not just pharma, totals under $85,000 — a second reminder that assumptions here need checking before they're acted on.

How much is already being spent to decide these races without us

A race with tens of millions already committed on both sides behaves differently than one that's still lightly contested — this is the context for judging where outside attention could actually shift something versus where the outcome is already being fought over at industrial scale. FEC-direct totals through mid-2026 filings, via api.open.fec.gov. Bars share one dollar scale (max $8.2M) so races are comparable at a glance.

* Partial-cycle figures (committee registered late, or only one quarter's data available) — not a full apples-to-apples comparison.

Peace: the clearest test of who's actually proven

This is the peace pillar — ending the Iran war and the surveillance/civil-liberties questions that ride alongside it — applied as the same test across the whole tracked list, not just the two or three names already singled out above. The point is to see where else a real record exists, not assumed reputation. Primary-sourced from House Clerk roll calls RC85, RC114, RC199, RC221. Yea on the war-powers votes means voting to end/limit U.S. involvement in Iran; Yea on RC221 means voting for the FISA §702 extension. Colors mark position, not right/wrong.

Showing 11 of 22 tracked battleground incumbents. Full vote records and sourcing available on request.

Every other race worth watching, ranked by why

Beyond the core leverage picture above, these are the races where the verified money and vote data still tell a story worth knowing — ranked by how much that story could change how the race is understood, not by who's predicted to win.

Where to focus, in order — and why nowhere else makes the list

This is the answer everything above builds toward. Weighing three things together: how well a candidate's actual record fits the three pillars this page tracks — health freedom (COVID-era institutional accountability), healthy food (pesticide-industry accountability), and peace (the Iran war-powers votes) — how competitive the race is, and whether holding the seat helps keep a majority that includes real allies on both sides of the aisle. This is not a prediction of who wins, and it doesn't take a position on which party should control the House — it's a ranking of where evidence, not assumption, says energy would actually count.

A separate case, not ranked above: NC-11 isn't a Republican-protection question at all — Chuck Edwards was forced off the ballot by an ethics censure recommendation, leaving a financially weak replacement against a well-funded Democratic challenger. A real accountability story in its own right, independent of this ranking.

The House margin is razor-thin: Republicans hold it 220–215, so a net loss of just 3 seats flips control. That reframes the question — it's not about finding a landslide anywhere, it's about a small number of specific seats close enough that focused effort has a real chance of mattering. Barrett, Perry, and Merrin are three genuine Republican-held-or-contested toss-ups; if a Democratic-favoring environment costs those three seats alone, that flips the House. Holding them is disproportionately important given how thin the margin is.

• MAHA's actual capacity is smaller than it looks. MAHA PAC has raised $2.4M against a $100M goal, with ~$287K cash on hand, and zero overlap with any competitive House race. MAHA Action's list is broader but shallow — 14 of 18 endorsements sit in safe seats, non-House races, or races nobody rates competitive.
• Miller-Meeks (IA-01) is the sharpest trap in this whole dataset. Not MAHA-endorsed, and the single largest health-industry PAC total tracked here ($699,954, 72% PAC) — in the tightest race on the list. Treating "any close Republican" as an opportunity would mean backing the one candidate whose own money cuts hardest against the point.
• Merrin and Mitchell are endorsements, not track records. Barrett and Perry both have years of actual votes to point to. Merrin has never cast a federal vote, and his district was just redrawn 4 points more Republican — any result there is at least as attributable to redistricting as to him personally.
• The national environment is a headwind right now, not a tailwind. All three outlets have moved toward Democrats continuously since spring. Any single result needs to be read against that backdrop before crediting anything endorsement-specific.
• The Massie precedent is the caution against treating "hold the majority" as the goal itself. A majority doesn't protect independent-minded members if leadership will spend $32M defeating one in a primary over these exact issues. Seat count and issue-alignment aren't the same thing.

A different question from the ranking above: which individuals matter regardless of which party ends up controlling the chamber. The answer is the same two names from Tier 1 — Barrett and Gluesenkamp Perez — plus Perry as a narrower, COVID-accountability-specific pick. Three more people aren't on any competitive list at all, because their seats are safe, but their continued presence is worth naming as part of what's actually being protected this cycle:

• Ro Khanna — safe seat; the clearest continuing anchor for the war-powers coalition on the Democratic side.
• Warren Davidson — safe seat; the one sitting House Republican who voted against the Iran war and remains in Congress past January.
• Chellie Pingree — safe seat; the lead sponsor, alongside Thomas Massie, of the pesticide-liability fight — living proof the model already works.