Early voting is underway across multiple states with about two weeks left before the 2026 U.S. midterm elections. A longtime election analyst on Zhihu, who published a first comprehensive preview 38 weeks out, has updated it at the 15-day mark, re-sorting the polls that have stabilized over the past ten days.
The most notable shift this cycle is that polling has moved from sharp swings into a relatively calm phase. Most numbers from pollsters—including firms previously tagged with strong partisan labels—fall within 3 percentage points for either side. Exceptions remain: the Senate races in North Carolina and Georgia show clearly wider margins and no longer qualify as tossups.
North Carolina: Both sides have pulled out
North Carolina is one of the states Republicans abandoned early. In RCP’s polling average, the Democratic lead is approaching 10 percentage points. Both campaigns have withdrawn major resources from the state, but Democratic candidate Roy Cooper’s own fundraising capacity still gives him an overwhelming spending advantage in the final stretch.
The Senate race has therefore largely lost its suspense. More worth watching is how far the “top-of-ticket effect” carries down: whether Cooper’s strong advantage can lift downballot candidates for the state legislature. After all, North Carolina’s current congressional map was passed under a Republican majority, and shifts in state-level races will directly affect future maps. The author’s vote-share forecast is Cooper 53% to Michael Whatley 43%.
Georgia: Senate settled, governor not
Georgia’s Senate seat is likewise considered out of reach, but Jon Ossoff carries another task in this election: whether he can pull gubernatorial candidate Keisha Lance Bottoms across the line. The author is blunt about Bottoms, calling her mediocre in fundraising, campaigning, media use and oratory.
Her opponent, Rick Jackson, may not be stronger. He holds a fundraising edge thanks to his own vast wealth, but in a debate he said something as detached from reality as “my plumber drives a Mercedes,” becoming a repeatedly quoted laughingstock.
The real reference point is 2022. That year Georgia’s first-round split-ticket voting was about 8.5%—Senate Democrats ahead by 1 point, governor Republicans ahead by 7.5. The author estimates this year’s split-ticket margin at roughly 5% to 7%, reasoning that Jackson’s candidate quality cannot match Kemp’s that year, while Ossoff’s advantage over Warnock is smaller than Kemp’s gap over Jackson. So Ossoff should win by more than 7 points without trouble, and Bottoms is also expected to squeak by; but a runoff would clearly add variables. The forecast is Ossoff 53% to Mike Collins 44%, Bottoms 50% to Jackson 48%.
New Hampshire: the surname dividend fades
New Hampshire offers another kind of sample. During the primary, state polling briefly narrowed to about 3 percentage points as Republicans fully exploited the advantage of the “Sununu” surname. But after the primary, the surname dividend faded and the baggage of Trump’s second term proved hard to shed, and the polls quickly widened.
The case shows that primary-stage polling is easily distorted by local factors; once the two parties face off directly, the national political environment reasserts its weight over local advantages.
The remaining variables
From this 15-day-out stocktake, the outcomes in several key states on the Senate map are becoming clear, and resources are concentrating in genuinely tight races. But early voting has only just begun, and stabilizing polls do not mean the results are set: turnout structure, event shocks in the final two weeks, and possible runoffs in individual states could still change the final numbers. For those following U.S. politics, what matters more in the next two weeks is not national polling but the composition of early voting by state and the downballot.
0 条评论
还没有评论，写下第一条。