Michigan Senate 2026

Michigan's open Senate seat—Gary Peters is retiring—settled its first half on August 4. Abdul El-Sayed won the Democratic primary by under three points over Haley Stevens, a nomination the outright book had at 98% all summer.

The general has stopped moving. El-Sayed slid from about 63% to 54.5% in the week after the primary; two weeks on he prices at 54.5% against former Representative Mike Rogers at 44.5% on Kalshi's candidate book. The repricing that looked like a trend has flattened into a level. The party book reads 56.5/44.5 and Polymarket's mirror sits three points richer for the Democrats at 59.5/41.5—the same gap it has carried since the primary.

The national picture moved instead. Senate control was 54.5% Republican a week ago; it is 52.5% Republican / 48.5% Democratic now, a two-point drift toward the Democrats on a $374,171 book. Michigan is load-bearing for any Democratic path, and it did not move—so the control repricing came from somewhere else on the map.

The volume is entirely on the underdog. Kalshi's party book turned over $62,686 and $52,700 of it—84%—landed on the Rogers leg, priced at 44.5%. The candidate book does the reverse: $19,193 on El-Sayed against $2,799 on Rogers. Two books on the same seat, opposite flow.

Kalshi and Polymarket both ran the Democratic field into the August 4 close; the pies preserve the final implied distribution each venue carried in.

Beside them, Kalshi's margin-of-victory Categorical, which is doing something odd. It settled on El-Sayed by 0–3 points at 99.5% two weeks ago, and it still turns over $48,910 a day. $47,430 of that—97%—is changing hands on legs that have already settled No, led by $20,476 on El-Sayed, ≥15% at 0.5%, a rung carrying 704,042 contracts of open interest.

A resolved question trading forty-nine thousand dollars a day, almost all of it on outcomes that did not happen, is the most extreme instance of the price-versus-flow inversion running through this catalog today.

The live dominant outcome, priced directly on both major venues: El-Sayed to win the seat on Kalshi's candidate book, and the Democrats to hold it on Polymarket's party book.

The venues still do not agree—Polymarket is three points richer for the Democratic side, 59.5% against Kalshi's 56.5%—and that gap has now persisted for two weeks without closing. On books this liquid a durable three-point spread is a read on venue composition rather than an arbitrage: it is the cross-exchange disagreement about how a progressive nominee travels in a genuine battleground.

The general settles November 3. Kalshi's candidate book prices El-Sayed against Rogers; the party book prices whether the seat stays Democratic or flips—the number that actually feeds Senate control.

The two have converged: 54.5/44.5 on the candidate book, 56.5/44.5 on the party book. Both stopped falling roughly ten days ago. Note the party book's legs sum to 101%, which on a two-way market with $62,686 of turnover is a live spread rather than a rounding artifact.

Zooming out: national Senate control—now 52.5% Republican / 48.5% Democratic, down two points for the Republicans in a week—plus Polymarket's independent read on the same Michigan seat.

The control book is worth reading for its own flow: $366,715 of $374,171, or 98%, traded on the Republican leg. As with Michigan's party book, the money concentrates on one side while the price barely moves.

Michigan voters also decide a governor's race in November; the Republicans nominated John James on August 4.

• The slide stopped. El-Sayed gave back roughly eight points in the week after the primary and has given back nothing in the week since. 54.5% now looks like a level the market has settled on rather than a waypoint.
• National control moved without Michigan moving. 54.5% to 52.5% Republican on the control book while this seat sat still. Whatever repriced the map came from another race—worth cross-reading against Ohio, Georgia and Texas.
• A settled margin book trading $48,910 a day. Ninety-seven percent of it on legs that resolved No. Either the book is being used for something other than the question it asks, or Kalshi's post-settlement quotes are attracting flow they should not.
• El-Sayed's general-election discount. He took the nomination and holds the seat at only 54.5%. That gap is the market's read on how a progressive nominee performs against Rogers, the Republicans' 2024 nominee, in a state Trump carried.
• Candidate book versus party book. They track within two points and their flow runs in opposite directions—$19,193 toward El-Sayed on one, $52,700 toward Rogers on the other. Same seat, same week, two different crowds.
• The three-point cross-exchange spread is now durable. Polymarket 59.5% Democratic against Kalshi 56.5%, unchanged for two weeks. Treat it as venue composition, not opportunity.