AFOS Weekly · USA-2026 midterms Political Risk Weekly

AFOS Analytics

Real-time political pricing
Prediction market × polls × press
no smoothed averages

Harvard Dataverse · DOI 10.7910/DVN/XRUT8U
United States·Issue №5·Week of 2026-08-27 to 2026-09-03·published thursday 07:00 BRT

What the prediction market did

The eight midterm contracts hold USD 18.58M, up USD 0.40M on the week. Every move was a transfer: in each chamber one side gained exactly what the other lost.

House · Democrats

89.50%

↑ 1.00pp on the week

On USD 5.72M, the deepest book here, and level with the highest the series has been since it opened on July 29.

House · Republicans

11.50%

↓ 1.00pp on the week

On USD 4.33M, at the floor of its own series for seven straight days.

Senate · Democrats

51.50%

↑ 1.00pp on the week

On USD 2.44M. It held at 50.50% for six days, then moved the day after the Republican leader's remark, its highest since August 18.

Senate · Republicans

49.50%

↓ 1.00pp on the week

On USD 1.68M, the thinnest of the four control books.

The fifth published price, on whether the election happens on schedule, reads 96.80%, up 0.50pp, on USD 0.40M. The House series opens on July 29, 2026 and the Senate on April 14, 2026, so superlatives here mean since those dates.

What the polls did

The disagreement comes before the average, because it is the finding. Fifteen rounds from eleven institutes sit inside the thirty-day window, averaging D+6.07. The newest fieldwork closed on August 31, which is fourteen days fresher than what this brief could read a week ago.

Widest

John Zogby Strategies · D+11

Narrowest

AlphaROC · D+4

Spread between institutes

7 points

Zogby surveyed 921 likely voters through August 5; AlphaROC 1,000 adults through August 10. Morning Consult again sits near the bottom with by far the largest samples, 29,719 registered voters in each of two rounds, at D+5 and D+4. The steadiest is The Economist/YouGov, weekly, at D+7, D+6 and D+6. Five houses finished fieldwork on August 17 and split three ways: D+8, D+7, and D+5 from the other three.

What the press reported

What the fixed-list outlets covered, August 28 to September 3

The week's loudest item came from the Republican side of the Senate. The Hill and the Washington Post both reported on September 2 that Majority Leader John Thune, speaking to KELO in Sioux Falls on September 1, said he worries about losing the chamber: "I worry about that, for sure. I'm a realist... I think it's a competitive environment right now." He paired it with praise for his candidates in Maine and Michigan.

The New York Times worked the other party's problems, reporting on September 2 that Democrats are feuding over defections, and asking the same day whether attacks on candidates who inherited a political name will land.

CBS News reported on September 2 a complaint to the Federal Election Commission over dark money in the Texas Senate race; NBC News the same day that progressive Senate candidates keep winning primaries while outspent, and on September 3 that President Trump will campaign in 35 races. The Hill covered pressure over Iran on September 2.

One outlet wrote about the instrument itself: NewsNation, on September 2, on what polling in Michigan's Senate race can and cannot tell readers after primary polls missed.

Our measurement: The two claims point in opposite directions, and the market moved against both. On the House, where the Post sees a narrowing path, the Democratic price rose to 89.50%, matching the top of its series. On the Senate, where the Republican leader sees real risk, it rose to 51.50%, the highest since August 18. Neither is a verdict: a leader warning his own party is doing politics as well as forecasting.

The crossing of the week

Last week's test got an answer, and it came back the other way

Issue No. 4 said a genuine shift would show as one side's price falling while the other rose, and that what had happened instead was a point added to the Republican side with the Democratic side still. This week the point was transferred: in each chamber the Democratic price rose exactly 1.00pp and the Republican price fell exactly 1.00pp. The shape that edition named as the one to watch for appeared seven days later.

But the two legs moved days apart, so this was not one trade

A transfer priced in one move and a transfer assembled over a week are not the same event. In the House the Republican side moved on August 28 and the Democratic side on August 30; in the Senate, August 30 and September 2, three days apart. Both pairs still add to 101.00%, exactly as last Thursday, so the gap between the two books did not close. What changed is which side carries the extra point.

A poll that exists is not the same as a poll we can read

Issue No. 3 declared a poll missing and refused to type it in by hand. This week it happened three times over: The Economist/YouGov published rounds closing August 17, 24 and 31 that the index still has not received. We read them from the pollster's own tables and marked each row with its source. The average moved from D+6.00 to D+6.07, almost nothing, and that is the point: the hole was not distorting the level, it was starving the window.

How to read this number

Added, transferred, and why the difference is the whole signal

Each chamber has two contracts with separate order books, one per party. When one side gets more expensive and the other does not get cheaper, the pair drifts above 100% and nobody has changed their mind about who wins: what changed is the cost of holding both sides. When one side rises by exactly what the other falls, the pair stays put and the money really did move from one answer to the other. That is why this brief reports both prices in a pair, and why we write that the market is paying 89.50% rather than that the probability is 89.50%. The first observes a price. The second is a claim about the world.

Sources

Prediction market

Polling, primary sources

Press, outlets cited in sections 4 and 5

Method

AFOS reads three signals and does not merge them: prediction market prices, published polls, and press coverage. The house average of the generic ballot is a simple mean, unweighted, over a rolling thirty-day window, and when one institute publishes more than one cut of the same round only one enters, by the hierarchy likely voter, registered voter, adults. Rounds that a pollster has published and the index has not yet received are read from the pollster's own tables and carry a marker saying so; when the index catches up, the indexed row replaces ours. Market prices come from a certified capture, two readings eight minutes apart that must agree inside 0.20pp before anything is published. Band markets are published only when the bands sum between 95% and 105%; the popular-vote-margin market is currently above 150% and is collected daily for the series without being published as a reading. This brief never subtracts a market probability from a polling lead, because they are different quantities. Open source, Apache 2.0.

AFOS
AFOS Analytics
Global Political Risk Intelligence · GLOBAL BY DESIGN
www.afos-analytics.com · contact@afos-analytics.com · github.com/AFOS-Analytics

Get it by email

AFOS Weekly follows the US midterms every Thursday, written for voters: where the prediction market, the polls and the press disagree, and what each one measures.

No spam. Unsubscribe anytime.

After signing up you choose whether to receive it in English, Portuguese or Spanish.

AFOS Weekly · Issue No. 5 · Democratic control got a point more expensive in both chambers, and this time the Republican side paid for it | AFOS Analytics