Skip to main content
ABC News
Everything You Need To Know About The California Recall Election

On Tuesday, Gov. Gavin Newsom’s future will be decided by California voters in only the fourth recall election of a governor in U.S. history. The vote comes after a wave of criticism over his handling of the COVID-19 pandemic, months of signature gathering by Newsom’s opponents, July’s official pronunciation of a recall election and then a two-and-a-half-month campaign sprint to determine whether Newsom should be recalled and, if so, who should replace him as California’s next governor.

The polls suggest the recall vote is more likely to fail than succeed, but it’s still possible Newsom won’t survive. Whether he continues to hold the state’s top job in Sacramento will affect the state’s governance, its response to COVID-19 and perhaps even the Democrats’ razor-thin majority in the U.S. Senate. Here, then, is a look at the state of the recall polls, who could replace Newsom as governor if he’s recalled and what the fallout might be.

The recall election

According to the FiveThirtyEight polling average of the first question on the recall ballot — whether to remove Newsom from office — 41.6 percent of Californians want to recall Newsom, while 56.2 percent want to keep him in office. 

But it’s important to note that this is not a forecast of what will happen (like those we issue for presidential and midterm elections); it’s just a description of what the polls have said. Usually, a 14.7-percentage-point polling lead is pretty safe, but this election isn’t usual, so some caution is warranted. This is a particularly challenging race to poll accurately because it’s hard to estimate who’s likely to vote. That’s due to two things in particular: the odd timing of the election (September of an odd-numbered year) and the fact that it is being conducted primarily by mail. In other words, don’t be surprised if there’s a larger-than-usual polling error. Indeed, as CNN’s Harry Enten has observed, the polling average of the previous gubernatorial recall election in California (in 2003) missed the final result by 9 points, and the “true” margin of error of polls of recent U.S. House special elections (which, like this recall, have been oddly timed) is around 13 points. At the same time, as Enten wrote over the weekend, there have been only four gubernatorial races since 1998 (out of 243) where the polling average missed by 15 points or more.

Overall, our assessment is that Newsom is a clear favorite to prevail. While an upset wouldn’t be unprecedented, it would qualify as a historically large polling miss.

Just catching up on the California recall election? Start here. | FiveThirtyEight

For much of August, the recall looked like a legitimate toss-up,1 but polls have since found Newsom’s lead expanding in the final few weeks. This could be thanks to the stunning amount of money that Newsom and his allies have poured into the campaign during this time: In August alone, they spent more than $36 million on TV ads, digital ads, canvassing, phone banking and text banking. More broadly, it could just be that California voters are finally starting to tune into the election. Ballots were mailed to every registered voter on Aug. 16, which may have activated more voters and made the electorate more representative of California’s overall (Democratic-leaning) population. 

Indeed, many summer polls that showed a close race did so because Republicans were much likelier than Democrats to say they were certain to vote in the election. How does a Republican win in a solid-blue state like California? In a very low-turnout scenario where most Democrats abstain from voting. But that seems increasingly less likely, based on data on who has voted so far. According to the California firm Political Data, Inc., 7,799,192 mail ballots have already been returned, 52 percent of which were cast by registered Democrats, 25 percent of which were cast by registered Republicans and 23 percent of which were cast by unaffiliated voters. Please, please don’t read too much into these numbers — there’s no guarantee that registered Democrats are voting against the recall (or registered Republicans for it), and in-person voters will likely be disproportionately Republican. Still, it sure doesn’t look like Democrats have an enthusiasm problem.

California’s base partisanship is one reason why this recall looks unlikely to follow the pattern of the one in 2003, when Democratic Gov. Gray Davis was recalled by 11 points. The Golden State is simply much bluer than it was 18 years ago. John Kerry, the Democratic presidential nominee in 2004, won California by just 10 points, but President Biden carried it by 29 points last year. And, crucially, while there has been some grumbling over Newsom’s governorship (even from Democrats), he’s nowhere near as unpopular as Davis was in 2003. As the chart below shows, Newsom’s approval rating has hovered around 50 percent among likely voters in surveys conducted by the Public Policy Institute of California in the leadup to the recall, whereas the same pollster found Davis below 30 percent approval among the likely electorate before the 2003 recall contest.

The replacement election

Should a majority of voters choose to recall Newsom, the ballot’s second question would then come into play: Who should replace Newsom? Out of 46 candidates on the ballot, the winner would simply be the one with the most votes, no matter how small a percentage of the vote they have. And the way things are shaping up, the winner of the replacement vote will fall well short of a majority — and likely get far fewer votes than Newsom does on the first question, even if the recall succeeds. 

Republican Larry Elder, a conservative radio talk-show host, has a clear lead (despite allegations of sexual harassment and his ex-fiancée’s claims that he brandished a gun at her during an argument), but he is polling at just shy of 28 percent in FiveThirtyEight’s tracker, well ahead of Democrat and YouTube celebrity Kevin Paffrath, who is at 7 percent, and two other Republicans, former San Diego Mayor Kevin Faulconer and 2018 GOP gubernatorial nominee John Cox, who are both right around 5 percent.

Now, when push comes to shove, Elder might win more than 28 percent because the pool of voters who cast a replacement vote (the second question) could be notably smaller than the recall electorate (the first question), so Elder’s vote total might represent a bigger slice of that smaller pie. That’s because many Democratic-leaning voters may skip the second question in response to explicit encouragement from Newsom’s campaign to vote “no” on recall and not to cast any vote for a replacement candidate. Additionally, there aren’t any high-profile Democrats running to attract Democratic-leaning voters, because the state party discouraged big names from running so that Newsom’s recall would be less palatable. In a Suffolk University poll released last week, 44 percent of likely voters said they would leave the second question blank, including 61 percent of Democrats compared with only 14 percent of Republicans. As a result, Elder attracted 39 percent of the likely voters who said they would cast a replacement vote, but that worked out to only 18 percent of all likely recall voters. In 2003, there were about 4 percent fewer votes cast for the replacement candidates than on the recall question, but it seems likely the undervote will be much larger this time around.

A cut out illustration of California, with four voter ballot bubbles in different colors laid over it.

related: Why It Was So Easy To Get Gavin Newsom’s Recall On The Ballot In California Read more. »

Still, even if Elder’s vote share does end up being higher than 28 percent, it’s clear there is no 2021 replacement candidate with the kind of broad support Arnold Schwarzenegger attracted in 2003, when he won the vote to replace Davis with about 49 percent in a 135-candidate field. In fact, because of the unusual recall-replacement format, Elder could win with the lowest vote share of any elected governor in modern times. As the chart below shows, over the past half century or so under any electoral format, no governor has won with less than 30 percent of the vote -- a feat Elder could conceivably accomplish if Newsom is recalled.

The lowest vote share for an elected governor dating back to 1968 was 33 percent, which then-Democrat Buddy Roemer garnered to win Louisiana’s all-party primary for governor in 1987, as fellow Democrat Edwin Edwards declined to contest a runoff against Roemer after no candidate won a majority.2 In a regular general election, the lowest win percentage was 35 percent, which Angus King won in Maine’s 1994 gubernatorial race as an independent. And if Elder were to win with 30 percent or less, he’d actually have the lowest winner’s vote share in any gubernatorial race since 1900, surpassing Democrat Ernest Lister’s 31 percent in Washington’s 1912 election.

The potential fallout

We’ve spent so much time looking ahead to the recall election itself that not a lot of ink has been spilled about what the result could mean for the future of California — and national — politics. 

If Newsom is recalled (or even if the recall does better than expected), Republicans will undoubtedly claim they have the momentum going into the 2022 midterms. It’s unclear, though, how predictive a gubernatorial recall election is of future elections. True, if one party consistently overperforms in special elections for Congress, it usually means that party will do well in the next November election. But the result of a single special election should never be overinterpreted; there might have been local issues at play or extenuating circumstances that made that one race a poor bellwether for the nation as a whole. And recall elections are especially prone to extenuating circumstances since something seismic and unusual must have happened to get them on the ballot.

Plus, special elections for Congress are federal elections, meaning partisanship plays a bigger role in how people vote; by contrast, it’s not unheard of for a state to elect a governor from the party opposite to how it usually votes in presidential races. And there is such a small sample of gubernatorial recall elections in U.S. history that it’s impossible to know whether they have any predictive power, anyway. The 2003 recall of Davis possibly foreshadowed a good 2004 election for Republicans — the 51 percent of the national popular vote that then-President George W. Bush received is still the highest for a Republican in over 30 years — but former Gov. Scott Walker’s victory in the June 2012 recall election in Wisconsin didn’t presage then-President Barack Obama’s reelection victory later that year, in which he carried Wisconsin by 7 points.

Electoral impacts aside, the recall result could really matter in terms of policy. If Elder were to become governor, that would surely have ramifications for the state’s response to COVID-19 and its politics more broadly. Most immediately, directives from the governor’s office on battling the coronavirus could shift sharply away from the status quo. Elder has promised to repeal Newsom’s mask and vaccine mandates if he wins, and he’s also expressed opposition to having young people get vaccinated or wear masks in school. This comes despite a CBS News/YouGov survey last month that found around two-thirds of Californians backed vaccine mandates for health workers and even employees of private businesses.

More broadly, California governors can fill vacancies in state courts and in various executive departments. As such, Elder could try to appoint conservative justices to California’s courts and place conservative officials in state government. He would also have line-item veto power on appropriations bills, which could enable Elder to seek funding cuts for programs he doesn’t support. 

At the same time, a Gov. Elder would not have limitless power to reshape California in his image. Most importantly, Democrats still have supermajorities in both chambers of the state legislature, meaning they would be able to override his vetoes of any bills they want to pass. And any judges Elder appoints would have to be confirmed by the state’s Commission on Judicial Appointments. Finally, an Elder governorship might not last long because the office will be on the ballot again in November 2022. Given California’s blue hue, it’s very possible that, even if Newsom is recalled, the state will have a Democratic governor again in just 16 months.

But there’s one lurking and potentially momentous appointment that Elder could make without being overruled: filling a vacancy in the U.S. Senate. It’s not hard to construct a scenario in which an Elder victory could lead to a power shift in Washington, either, since Democratic Sen. Dianne Feinstein is 88 years old and the chamber is currently divided 50-50, with Vice President Kamala Harris’s tie-breaking vote giving Democrats a narrow hold on power. Earlier this month, Elder pledged to appoint a Republican in Feinstein’s place (should her seat become vacant), which would give Republicans a 51-49 edge in the Senate. How long Elder would even have this power is hard to say, though, as Democrats could use their supermajorities in the state legislature to change the state’s Senate vacancy rules to prevent this scenario from happening.

Finally, one consequence of the recall election — regardless of the result — could be alterations to the recall process itself. It’s easier for recalls to succeed in California than in any other state, and it wouldn’t be surprising if the Democratic legislature tries to amend state law to discourage this from happening again. (It wouldn’t be the first time that Democrats manipulated California’s recall process for their own attempted political benefit.) There is public support for this, too: According to a YouGov/Hoover Institute/Bill Lane Center for the American West poll of the recall from a couple weeks ago, a plurality of Californians favor several reforms to the recall process, including raising the number of signatures required to trigger an election, increasing the filing requirements for replacement candidates and requiring a specific reason to recall a governor.

If you’re interested in seeing how it all shakes out — and really, why wouldn’t you be? — join us for our live blog of the recall election. Just don’t expect to learn the result on Tuesday night: Since the election is being conducted predominantly by mail, it’s going to take several days to count all the votes, just like it did in the 2020 presidential election. Last year, California counted only about two-thirds of its votes on election night, with the rest trickling in over the next several days. The state didn’t even hit 95 percent reporting until 11 days after Election Day.

It likely won’t take as long this time, since the deadline for ballots to be received is one week after the election (in 2020, it was 17 days after), but you should still be prepared to wait multiple days for a winner to be declared if the election is at all close. But don’t worry — we’ll be keeping the live blog live for as long as it takes to project a winner.

Aaron Bycoffe contributed research.

Why a 9/11-era political consensus seems impossible these days

What Texas’ abortion ban could mean for the rest of the country | FiveThirtyEight


  1. This was partly due to one outlier poll from SurveyUSA that showed Californians in favor of recalling Newsom, 51 percent to 40 percent. SurveyUSA has since acknowledged that the wording of their likely-voter question “may have led some respondents opposed to the recall to misidentify themselves as being unlikely to vote.” However, even without that SurveyUSA poll, our polling average would have still shown the recall failing by fewer than 4 points as late as Aug. 21, due to other polls around that time that showed a close race.

  2. Roemer switched parties to become a Republican during his subsequent term in office.

Geoffrey Skelley is a senior elections analyst at FiveThirtyEight.

Nathaniel Rakich is a senior editor and senior elections analyst at FiveThirtyEight.