Where the Texas polls are.
First, the two races are tightly coupled by a shared statewide error, but a six-point SenateβGovernor gap has held on average across 21 paired polls since June. Our internal model says the most likely single outcome, at 43%, is a split ticket affair. But it's September 23rd, not October 23rd. Let's wait and see. Moving on.
Twenty-four public polls of each race since June 1. The Senate race moved from even in June to a stable Talarico lead of about three points by late July and has not moved since. The Governor race has compressed steadily from Abbott +8 in June to about +2 to 2.5 now, driven by Hinojosa's rising name recognition (a third or more of voters had no opinion of her through August). Adjusted, recency-weighted averages:
Senate: D+3.1
-Talarico 47.4
-Paxton 44.5
Governor: R+2.7
-Abbott 48.1
-Hinojosa 45.5
Across the 21 polls that tested both races in the same sample, Talarico's margin exceeds Hinojosa's by 6.3 points on average (SD 3.1), and the two margins correlate at only 0.49 poll-to-poll. The gap is a real feature of the electorate, and probably not noise. Fox News (July) measured it directly: 12% of Talarico voters back Abbott, 9% of Paxton voters back Hinojosa; TSU (September) found 38% of likely voters expecting to split their ticket.
More to come from our Texas sized model.