Archive / Issue No. 8

Midterm Election Digest

43 days to Election DayDay score 0

01Race developments

Kansas Senate — first competitive numbers logged for this seat

A new Emerson College poll (fielded Sept 15–16, published Sept 18) shows Democrat Adam Hamilton at 45% to Republican incumbent Roger Marshall's 43% — within the margin of error in a seat neither party has treated as competitive this cycle. Same survey has the governor's race at Masterson (R) 51%, Holscher (D) 44%. Both Cook and Sabato still rate the Senate seat Likely/Solid Republican; this is one poll, not a rating change.

Emerson College Polling, Sept 18 · MOE ±3.6, n=750 LV
Texas Governor — real movement, forecasters still skeptical

A ReconMR/Siena University poll (released Sept 17) gave Democrat Gina Hinojosa her first outright lead of the cycle over Gov. Greg Abbott, 49–45%, built on a 65–27 edge among independents. A same-week Emerson poll still had Abbott up 49–46, and other recent surveys are mixed. Prediction markets have moved too: Kalshi and Polymarket both cut Abbott's implied win probability from 87% a month ago to 76% now. Cook and Sabato still rate the race Likely Republican.

Newsweek (reporting ReconMR/Siena, Emerson, Kalshi, Polymarket), Sept 17

Nothing meeting the bar surfaced today in Ohio, Michigan, Georgia, North Carolina, or New Hampshire beyond what was already covered in recent issues — those races are being tracked and will be reported again when something actually moves.

02Polling & forecasts

Silver Bulletin: Democratic Senate-control odds hit a cycle high, 65%

Nate Silver's Deluxe forecast put Democrats' probability of winning Senate control at 65% as of the Sept 20 update — up from 59% on Sept 18 and, per the model, "the highest they've been this cycle." The move tracks a run of strong generic-ballot numbers: NYT/Siena at D+8, YouGov/Economist at D+12 among likely voters.

Silver Bulletin, Sept 20
Cook's Battleground District poll: D+2 on the generic ballot in Trump-won turf

Across the 37 House districts Cook rates as competitive (median Trump+5 in 2024), a Cook/New River Strategies/GS Strategy Group poll (fielded Sept 8–11) found Democrats leading the generic ballot 49–47 — roughly a 7-point swing from those districts' 2024 baseline. Trump's approval within these specific battlegrounds sits at 42% favorable / 58% unfavorable. This is a poll of competitive turf, not a set of individual rating changes.

Cook Political Report, Battleground District Project, Sept 17
No new confirmed rating changes since Sept 20

Cook's Senate and House rating pages remain dated Sept 15 and Sept 11 respectively — no additional move since the Iowa Senate shift reported in Friday's issue. Sabato's Crystal Ball has not posted a new weekly ratings piece since Sept 17's "Seven Weeks to Go" (already reflected in this digest's score log).

03Election administration & security

Vermont wins dismissal of DOJ's voter-roll lawsuit — 23rd of 31 nationwide

A federal judge (Mary Kay Lanthier) dismissed the Justice Department's suit demanding Vermont's full voter file — including home addresses, birthdates, and partial Social Security numbers — ruling the administration exceeded its authority; Vermont's Secretary of State had refused the request on privacy grounds. Per a University of Wisconsin Law School tracker, this is the 23rd of 31 state DOJ voter-data suits dismissed at the district-court level nationwide; no court, at any level, has yet ruled for the DOJ (one appellate court, the 6th Circuit, affirmed a dismissal in Michigan back in June).

Vermont Public, Sept 18 · State Democracy Research Initiative tracker (Univ. of Wisconsin Law School)
SAVE database case — still pending, no ruling

The DHS v. League of Women Voters fight over the SAVE citizenship-check database remains before the Supreme Court with no ruling since it became fully briefed on Sept 15. The lower-court injunction blocking DHS from running bulk citizenship checks against state voter rolls stays in effect in the meantime.

SCOTUSblog, as of Sept 20

04Worth sharing

05Seat-change score

Counts only confirmed individual race-rating moves from a named nonpartisan rater (Cook, Sabato) — flat ±1 per race regardless of how many notches it moved. Polls, fundraising, endorsements, and aggregate probabilities don't count.

DateScoreWhat drove it
Sept 15—predates this feature — no score recorded
Sept 16−1NH Senate: Lean D → Toss-up (Cook)
Sept 170no confirmed rating change
Sept 180Sabato's 8-race House move flagged but not yet confirmed
Sept 19+5Sabato confirmed: NJ-7, MI-4, IA-1, IA-2, CO-5 all moved D-ward (weekly batch, not a daily signal)
Sept 20+1IA Senate: Lean R → Toss-up (Cook)
Sept 210no confirmed rating change today
+5
Net trailing-7-day movement (6 scored days) — all of it D-ward, none R-ward beyond the single Sept 16 NH move. Almost entirely one event: Sabato's Sept 19 weekly batch. Treat this as a lumpy, source-driven signal, not a smooth daily trend.

Score log retroactively reconstructed for days before Sept 20 (when MK requested this feature) from what was actually surfaced in each day's issue — may be incomplete for those earlier days.