Democratic candidates lead their Republican rivals in five critical House races in California, according to a new set of polls that show President Trump dragging down his own party.
The polls, conducted online by the University of California-Berkeley’s Institute for Governmental Studies for the Los Angeles Times, show Democrats making inroads in districts that Hillary Clinton won over Trump in 2016, seats Democrats must win if they are to reclaim the majority in the House.
“These are all GOP-held seats, so in a normal year you would expect that the Republicans would be comfortably ahead in these districts, and maybe a couple would be close,” said Mark Di Camillo, who directs the Berkeley IGS poll.
“If there happens to be a quote unquote wave election coming up in four weeks, a lot of these seats could be in jeopardy.”
{mosads} In the Central Valley, Rep. Jeff Denham (R) trails his Democratic challenger, venture capitalist Josh Harder (D), by a 50 percent to 45 percent margin.
Trump’s approval rating in the 10th district stands at just 43 percent, and Harder is winning about two-thirds of the Hispanic vote.
Rep. Steve Knight (R), one of only two Republicans to hold a House seat that includes parts of Los Angeles County, trails his Democratic opponent, nonprofit executive Katie Hill, by a 50 percent to 46 percent margin.
Just 45 percent of district residents approve of the job Trump is doing in the White House, and more than half say they have a strongly negative view of his job performance.
The race to replace retiring Rep. Ed Royce (R) is virtually neck and neck. Philanthropist Gil Cisneros (D) leads former Assemblywoman Young Kim (R), who long worked for Royce, by a statistically insignificant 49 percent to 48 percent margin.
Trump’s approval rating stands at 46 percent in the district that covers eastern Los Angeles County and Orange County.
Cisneros, who faced allegations of sexual harassment that have since been withdrawn, has notably high unfavorable ratings, according to the UC Berkeley poll.
In the heart of Orange County, Rep. Mimi Walters (R) trails her Democratic opponent, law professor Katie Porter (D), by a 52 percent to 45 percent margin.
Other surveys have also showed Porter leading Walters in a district where Trump’s approval rating is at just 41 percent.
And in the San Diego-area seat being vacated by Rep. Darrell Issa (R), Democrat Mike Levin holds a 55 percent to 41 percent edge over Diane Harkey, a member of the California Board of Equalization. Trump’s approval rating is a lousy 39 percent in the district.
Trump has better approval ratings, at 49 percent, in an Orange County seat that includes wealthy coastal enclaves like Huntington Beach and Laguna Beach. There, Rep. Dana Rohrabacher (R) is tied with his challenger, businessman Harley Rouda (D), at 48 percent apiece.
“You can definitely see the correlation between Trump’s job ratings and the vulnerability of these districts. It’s a straight-line correlation,” Di Camillo said. “There’s an intensity of feeling that’s on the opposition side in six of these districts that I think is putting them in jeopardy.”
Two Republican incumbents lead their Democratic challengers in districts Trump won in 2016.
In the Inland Empire, Rep. Devin Nunes (R) leads county prosecutor Andrew Janz (D) by a 53 percent to 45 percent margin. Janz has been an impressive fundraiser, but district voters give Trump a 56 percent approval rating. Nunes, the chairman of the House Intelligence Committee, has been a staunch Trump defender.
Farther south, Rep. Duncan Hunter’s (R) struggle with an indictment over campaign finance violations make his bid for re-election difficult even in a district Trump won by 15 points. The Los Angeles Times poll showed Hunter leading former Labor Department official Ammar Campa-Najjar (D) by a 49 percent to 47 percent margin.
Many of the poll’s results match up with other surveys that have shown Porter leading, or Hill trailing Knight within the margin of error, or Nunes with a wide lead over Janz.
But the Berkeley pollsters used an innovative — and controversial — methodology that is likely to earn scrutiny from Republicans unhappy with the results.
While most polls survey voters by telephone, the Berkeley poll surveyed 5,090 likely voters throughout the eight districts tested by reaching them via email.
Those emails were gleaned from the California voter file, and only about a third of state voters have their email addresses on file.
That means the poll was unable to include voters whose email addresses were not included in the voter file.
Voters without emails are disproportionately likely to be older, or of racial minorities, like Hispanic voters.
Older voters tend to back Republican candidates more than Democratic candidates, while Hispanics who were reached in the Berkeley surveys tend to back Democrats by stronger margins.
Pollsters have grown concerned that fewer respondents are answering their phones, as more Americans ditch landlines in favor of cell phones. Many have begun experimenting with other ways to gauge public opinion — including, like the Berkeley poll, by email.
“They know the advantages and limitations of the product,” said Justin Wallin, a Republican pollster in California.
“The actual numbers may not be precise, but I respect their methodological approach and feel that in the absence of a live-interviewer survey using the voter file, no caps on cell phones and speaking in [the right] language, it’s a pretty darned good alternative.”
Di Camillo, the poll director, said using emails to contact voters helps him develop a sample that is more reflective of the actual electorate, without using complex weighting methods that can introduce errors into the response.
And the method helps keep down costs: Pollsters can translate online polls into a voter’s language — in this case, Spanish and Vietnamese — without having to hire callers who speak those languages.
“In California, about six million registered voters have their email addresses on the official voter roles, and you can do random samples off those six million,” he said. “I can certainly look at the segments that are underrepresented in the sample and send more invitations.”