Two months after Scott Wiener beat Connie Chan by 11 points in the June primary, two new surveys put the race to succeed Nancy Pelosi within the margin of error — but both polls were commissioned by sides with a stake in the result.

The tightening, reported first by Mission Local, is the biggest shift in California's 11th Congressional District race since Wiener's primary win. It hands Chan a momentum narrative heading into November. But neither poll is a neutral read: one was paid for by Chan's own campaign, the other by a housing committee backing a ballot measure — and each comes with methodological caveats that cut against taking the top-line numbers at face value.

In June, the shape of the race looked settled. State Sen. Scott Wiener finished the primary with 40.7 percent to Supervisor Connie Chan's 29.7 percent — an 11-point lead that fit months of polling showing Wiener as the frontrunner to replace Speaker Emerita Nancy Pelosi.

Two surveys shared with Mission Local now tell a tighter story. A poll sponsored by Chan's campaign and conducted by Lake Research Partners found Wiener ahead 45 to 40, with 11 percent undecided. A second, sponsored by Fair Housing — a committee supporting the Affordable Housing Guarantee Act on November's ballot — put Wiener at 39 percent and Chan at 38, with 23 percent undecided. On paper, an 11-point gap has collapsed to somewhere between one and five points.

The caveat is in the sponsorship. A campaign that pays for a poll and then releases it is telling you something about the numbers it likes. The Lake Research survey of 500 likely voters, fielded July 14–19 in English and Cantonese, included an oversample of 100 Chinese American voters — a group central to Chan's base — and carried a 4.4 percent margin of error. Lake concluded Chan had gained more than twice as much new support as Wiener, a framing the pollster's client was happy to circulate.

The Fair Housing poll carries its own asterisks, several flagged by Mission Local. It surveyed 509 voters June 26–27 citywide, meaning roughly 10 percent of respondents live outside Pelosi's district and won't vote in this race. It was conducted only in English, leaving out the monolingual Chinese voters most likely to favor Chan. Both distortions pull in opposite directions, which is precisely why a one-point margin from it should be read as noise, not a lead change.

The campaigns are reading the same data very differently. Julie Edwards, Chan's spokesperson, said the polls show Chan "building momentum every day." Joe Arellano, Wiener's spokesperson, dismissed the surge as temporary. "Connie Chan should enjoy this moment," Arellano told Mission Local. "Once San Francisco voters learn that her entire career is built on stopping progress and making housing unaffordable, her poll numbers will sink faster than the Giants' season."

Where the new support is coming from is the real contest. Saikat Chakrabarti, who finished third and endorsed Chan in June, brings a progressive base that overlaps with hers — but also younger urbanists who could just as easily break for Wiener. Jim Stearns, a political consultant who has donated to Chan's campaign, pointed to the large undecided share and split it in two: Chakrabarti voters still choosing between the two finalists, and "the more conservative, Trump-voting Republicans" who backed harder-right candidates in the primary and may sit November out.

"It is a question that's facing both campaigns," Stearns said, "what role do the hard-right voters play in this?" It is an unusual thing to ask in a deep-blue district — and a sign of how thin the margin has become. Buried in the Fair Housing numbers is the reason it commissioned the poll at all: 67 percent of respondents backed the Affordable Housing Guarantee Act, and 58 percent the Housing Trust Fund Expansion Act. The horse race made headlines; the ballot measures were the point.