California Says Becerra by 25, but History Says That Polls Aren't Election Results

Steve Hilton is staring at a brutal number: 25 points.

The latest UC Berkeley Institute of Governmental Studies poll has Democratic gubernatorial nominee Xavier Becerra leading Republican nominee Hilton 58% to 33% among likely California voters.

The survey included 4,512 likely voters and has a margin of error of about 2 points.

From the New York Post:

The Californians surveyed cited the state’s steep cost of living — with some of the highest housing and fuel costs in the nation — as their top concern.
However, the Dem frontrunner’s lead has grown since a May poll by the same UC Berkeley institute, which found 55% support for Becerra versus Hilton’s 37%.
Hilton has dismissed any worries that his campaign is falling behind.
“Californians are desperate for change and we have momentum,” Hilton told The Post. “Becerra is a career politician and electing him is four more years of the one-party rule of the last 16.”

Nobody should wave away 25 points. A separate September survey from the Public Policy Institute of California put Becerra ahead 60% to 38%. Two surveys showing leads above 20 points deserve attention.

Still, the calendar says Sept. 24. Californians vote on Nov. 3, and American polling has accumulated enough scar tissue over the past decade to make certainty foolish.

The American Association for Public Opinion Research examined 611 polls conducted during the final two weeks of the 2024 campaign. Polls were substantially better than in 2016 or 2020.

They also underestimated Republican vote shares relative to Democrats for the third consecutive presidential election. Democratic margins were overstated by an average of 3.1 points in 2016, 4.6 points in 2020, and 2.7 points in 2024.

The word "always" goes too far. AAPOR found only a tiny Republican overestimate in the 2022 midterms. The recent pattern is nevertheless remarkable: three straight presidential elections in which Republican support finished stronger relative to Democrats than the late polling indicated.

Then, of course, there's 2016, the election that should have retired political certainty permanently. Hillary Clinton's famous 99% chance of winning came from a forecast model, not an individual poll, an important detail. But polling-driven forecasts placed her chances anywhere from 70% to 99%. Donald Trump then won Pennsylvania, Michigan, and Wisconsin.

From the Pew Research Center:

How could the polls have been so wrong about the state of the election?
There is a great deal of speculation but no clear answers as to the cause of the disconnect, but there is one point of agreement: Across the board, polls underestimated Trump’s level of support. With few exceptions, the final round of public polling showed Clinton with a lead of 1 to 7 percentage points in the national popular vote. State-level polling was more variable, but there were few instances where polls overstated Trump’s support.

Wisconsin supplied another lesson in 2022. An August Marquette poll had Democrat Mandela Barnes leading Republican Sen. Ron Johnson 52% to 45% among likely voters. A month later, Johnson led 49% to 48%. He eventually won 50.41% to 49.41%.

None of this turns Hilton's 25-point deficit into three points. The available California polling currently shows Becerra with a considerable lead.

It does tell us something about the temptation to transform polling into prophecy. Polls measure voters at a particular moment using models of who will actually vote. Turnout changes, and opinions move. Some voters are harder to reach. Likely-voter models can miss.

California may eventually deliver something close to today's numbers. Perhaps it won't. Recent election history offers plenty of evidence for watching the votes before treating September polling as November's final answer.

If you enjoy PJ Media's independent reporting and commentary, now's a great time to join us. Use promo code FIGHT and get 60% off your VIP membership.