Could Newsom Actually Be Recalled?

( Sean Trende) The California recall election against Gov. Gavin Newsom had been a fairly quiet affair, with Republicans considered a long shot at best to oust him. That was before two polls from Emerson College and another for the Los Angeles Times showed support for the recall reaching around 47%. More ominously for Newsom, opposition to the recall has fallen below 50%.

If undecideds break strongly enough for “remove,” he could lose. Against this backdrop is the specter of the 2003 recall, when California voters replaced Gray Davis with Arnold Schwarzenegger. If Republicans could pull off the feat once, they might do so again in 2021.

One thing we should be mindful of is not falling prey to the “unthinkability bias” that plagued the Brexit referendum in 2016. Polls consistently showed a close vote, yet observers almost universally disclaimed the possibility that it would actually happen. If something is genuinely a tossup it should happen about half the time. The fact that almost everyone convinced themselves the coin was going to come up heads meant either something freakish was happening statistically or, more probably, people couldn’t let themselves believe it was actually that close. (

So, for now, the best answer to the question of whether or not the California recall will happen is “maybe.”

That’s something I wouldn’t have said two months ago, which is significant in itself. But to help analyze the situation better, here are three reasons why the recall won’t happen, and three reasons why it might.

Reason It Won’t No. 1: California is bluer than it was in 2003.

For years, I have maintained that elections are fundamentally referenda on the party in power. In other words, voters don’t carefully weigh the arguments of the party in power against those of the challenging party before coming to a cautious conclusion regarding the proper vote to cast. Rather, they examine the track record of the governing party. If they like the job the governing party is doing, they vote for the governing party. If they do not like the job the governing party is doing, they vote against that party.

This tendency has increasingly filtered down to congressional and Senate elections in the past few cycles. As a point of comparison, in 1993, 12 Democrats represented states that George H.W. Bush carried in his losing presidential bid, while 17 Republicans hailed from Clinton states. In 2021, just six senators represent states won by a different presidential candidate than carried the state in the previous election. There is a reasonable chance that by 2025 that number will be reduced to one. Even governor’s races have converged somewhat on national trends (although we should note that Louisiana has a Democratic governor and Vermont has a Republican one). 

The California recall election took place on Oct. 7, 2003. This was in the wake of the then-popular invasion of Iraq and when George W. Bush still enjoyed relatively high national approval ratings in the wake of his handling of 9/11. On Election Day that year, the president’s job approval stood at 53%. 

Today, President Biden’s job approval nationally stands at around 51%. Since Biden is a Democrat, we would expect this to buoy the prospects of Democratic politicians in general much as Bush’s did for Republicans 18 years ago. 

This national job approval must be read in the context of a state’s politics. The following chart shows the partisan lean of California over the course of the last 88 years. The partisan lean is defined as a candidate’s national vote share, subtracted from the share of the vote the state casts for that candidate. Thus, it allows us to control for fluctuations in the national environment.

As you can see, California’s lean was about five points to the left of the country as a whole in 2000. In other words, with President Bush’s national job approval standing at roughly 53%, we might expect his job approval in California to have been in the high 40s. That is enough to win in a state. 

Today, California is over 10 points to the left of the national average. In other words, when a Democratic president is at a 51% job approval nationally, we might expect him to be above 60% in the Golden State. That’s not the type of environment in which Democrats typically lose.

Reason It Won’t No. 2: Gray Davis’s polling was abysmal.

The fact that opposition to Newsom’s recall has fallen below 50% is not great news for the governor, but his position is not nearly as precarious as Davis’s was. In 2003, it was apparent that the recall was going to be successful by early August. “Yes” consistently polled above 50% post-August, ranging between 50% and 69% in the polls. Davis wasn’t just likely to lose the recall election, he was likely to lose it in a landslide. The final RealClearPolitics average had “Yes” leading 55.5%-42%. The final vote was 55%-45%.

This time around, there are still more people who oppose the recall than support it. Moreover, notice that in 2003 the “Yes” forces got exactly what the polling suggested that they should get, while the “Noes” scored a touch higher. The historical tendency in referenda is for undecideds to exhibit status quo bias and vote “no” — so, absent further erosion, Newsom should feel okay about his standing. Of course, this is only a tendency. Sometimes, things break towards “Yea” (for example, Brexit).

Reason It Won’t No. 3: Republicans don’t have a Schwarzenegger.

Perhaps the best thing Republicans had going for them in 2003 was their candidate. As one of the few openly conservative stars in Hollywood, the “Governator,” as he would soon be called, was beloved by California Republicans. This enabled him to survive with a much more moderate public profile, especially on social issues, than national Republicans traditionally support. Schwarzenegger was pro-choice on abortion and favored more liberal immigration policies than the average Republican voter.

At the same time, his status as a pop culture icon and celebrity gave him added cachet with non-ideological voters in the middle. His immigrant status likely helped him with the state’s burgeoning Hispanic population. His “alpha male” persona probably helped him with non-ideological men. His lengthy history supporting Republican candidates and causes made it easier to envision him as a governor; he developed policy credibility with his sponsorship of a successful education ballot initiative in 2002.

Contrast that with the current slate of Republican candidates. Caitlyn Jenner is perhaps the closest thing to Schwarzenegger in the race, given her history of supporting Republican candidates, her celebrity, and the ground-breaking nature of her candidacy, but it isn’t at all clear that her celebrity status works to her benefit in the way that Schwarzenegger’s worked to his. Other candidates, such as talk radio host Larry Elder, John Cox, and Kevin Faulconer, are likely either too conservative, too obscure, or some combination of both, to be counted on to win the state.

On the other hand, there are reasons to think the recall will succeed.

Reason It Will No. 1: The process favors Republicans.

The number one thing that Republicans have going against them in California is their party label. In some ways, the more this is a referendum, the better it is for them.

The California recall is a pure referendum. It proceeds in two steps. The voter first answers “yes” or “no” on the question of whether or not to remove the governor. Only then does the voter proceed to the next question, namely which candidate is preferred if Newsom is recalled. Note that Newsom himself does not appear in the second question. No prominent Democrat appears in the second question. Instead, there are candidates like Holly Baade, a healer and shaman, John Drake, an ice skating coach, and Jackie McGowan, a cannabis business consultant.

This was billed as a selling point for Democrats, but I’m not so sure. It’s true that the entry of Lt. Gov. Cruz Bustamante probably aided the recall of Davis in 2002 by inducing Democrats who disliked Davis to vote “yes,” knowing that they could mark their second ballot for a credible Democrat.

Under these circumstances, though, Democrats are playing a dangerous game. Given the problems that the various GOP candidates have, they would likely have a tough time defeating any credible Democratic candidate. But there isn’t one on the second ballot. The strategy Democrats are employing this time made sense in 2003 (even though it didn’t work) and seems even less suited for this year’s version.

Reason It Will No. 2: Ballot are being mailed to everyone.

The final two reasons the recall effort might succeed are straightforward and can be dealt with briefly. The first is that, under current California law, ballots are to be mailed to everyone. While this might help in a general election, where there is already an atmosphere of excitement and energy, it isn’t clear how it plays here. This could be a situation where, like the infamous Literary Digest poll of 1936, disproportionate energy in one segment of the electorate could skew the results from what we might see in general. In other words, if Democrats are in general apathetic about Newsom’s recall, while Republicans are energized to see him go, it could work to Republicans’ advantage. Having the recall election on Sept. 14, rather than on a normal November Election Day, will not help Democrats.

Reason It Will No. 3: Newsom is vulnerable

Finally, any dispassionate analysis of the recall campaign must acknowledge that Gavin Newsom has genuine problems as a candidate. The Emerson College poll has his job approval at 48%, which, while not radioactive, does put him in something of a danger zone. The state is in the midst of an uptick in COVID cases, and some locales are reinstituting mask mandates. This might combine with stories of his children going maskless at summer camp and him hosting a dinner party at a time when Americans were being urged to skip holiday gatherings to create a plausible narrative of elite indifference to the plight of everyday Americans. The state remains vulnerable to blackouts, which was a hardship that hurt Gray Davis.

In other words, this election does look increasingly like a true tossup. Taken together, I think that Newsom is still favored. Neither he nor President Biden is unpopular enough to lose. But against this analysis, the specter of Brexit looms large. The possibility of Gov. Larry Elder seems so wildly implausible that it is difficult to call that as the likely outcome until all evidence points in that direction. For now, let’s just say that this possibility is far more plausible than it seemed a month ago.