Model moves Ohio's 7th back toward Miller, with no new polling

Our model now gives Democrat Brian Poindexter a 29 in 100 chance in Ohio's 7th District, against 71% for Republican Max Miller. A week earlier, on August 30, the same model had the district at 50-50. That is a shift of 20.2 points of Democratic win probability in seven days, all of it toward Miller, the incumbent.

The expected margin moved with it. Our model's mean margin was R+0.1 on August 30 and is R+5.2 now. The spread also tightened: the 5th-to-95th percentile range ran from R+16.4 to D+16.5 a week ago and now runs from R+21.1 to D+10.7.

No new district polling explains the change. The three OH-7 surveys in our file all predate the August 30 run. GBAO polled 500 likely voters June 16-20 and had Miller at 44% and Poindexter at 43%. Bedrock Polling, sponsored by Iron Workers International and The Next 50, surveyed 866 likely voters August 7-10 and had Poindexter at 43% to Miller's 35%. Tavern Research polled 635 likely voters on August 4 and had Poindexter at 54% to Miller's 46%. Those results were already in the average that produced the coin flip. Nothing in the payload identifies what moved the estimate since; the model moved, and we are not going to attach a cause to it that we cannot document.

For context, our national average of 24 generic-ballot polls over the past 45 days stands at D+6.2, and our model gives Democrats a 93% chance of reaching the 218 seats needed for House control, with a mean of 239.3 seats. One caveat on that national figure is ours to own: our simulation correlates error nationally, by census division, by state and within each district, but we have no cluster data to correlate demographically similar districts across state lines, so the control probability is marginally firmer than a fuller model would make it.