CA primary: Buckle up, it’s going to be a bumpy night

For the past year, we’ve been conducting tracking polling of the dozens of candidates for the Democratic presidential nomination. 

A consistent thread in those surveys was change: The front runners shifted from former Vice President Joe Biden to Massachusetts Sen.Elizabeth Warren to Vermont Sen. Bernie Sanders.

Several others fluctuated among this top tier, most notably former Mayor Pete Buttigieg of South Bend, Indiana. 

Everything has changed in the last 36 hours, and we are now set for one of the most tumultuous California election nights in recent history.

As Super Tuesday neared we could see a race solidifying as a multi-way contest between several top tier candidates – each candidate having seemingly earned their position over the course of a long campaign. 

Even the addition of former New York City Mayor Mike Bloomberg, a billionaire,  took a natural steady path through the polling. He began in single digits in December, then rose steadily in January and February, reaching a high water mark in our first survey of early by-mail voters, in which he achieved 20% and the second spot in the race in California.

Yet, everything has changed in the last 36 hours, and we are now set for one of the most tumultuous California election nights in recent history.

Our polling examined early voters examined in two waves. 

The initial poll surveyed voters who were flagged by Political Data Inc. as having returned their ballots by Feb. 20.  This was after most voters had their ballots for a couple weeks and were able to consider results in both Iowa and New Hampshire as they made their choice.

A majority of these voters were Californians flagged as “always early,” along with others who fit the profile of early voters – older, more conservative, high income homeowners who have developed the pattern of voting early over successive elections. 

What was missing in these early voters were those voters who have recently converted to vote by mail, young voters and voters from lower socioeconomic backgrounds.  A deeper look at these early voters can be found in a recent Capitol Weekly article unpacking the early voting electorate.

The demographics of the two early waves were similar. 

Both had non-partisans only accounting for 10% of the total Democratic ballots cast. That may surprise many observers, given that nonpartisans are the fastest growing and second largest group of voters in the state.  With these early returns we were not only able to see if an independent voter cast a ballot, but whether the voter cast one with the presidential candidates on it – and only 10% of nonpartisan vote by mail voters actually obtained a ballot with the Democratic presidential candidates. 

We looked at this issue extensively in this recent article looking at these voters and in an earlier deep look at how well these independent voters responded to initial efforts by counties to help them obtain these ballots.

Between the two early surveys there was a small fluctuation in their age, with more young voters turning in their ballots after Feb. 21, increasing the share of voters under 35 from 14% to 21%.  But otherwise, the two early voting subgroups appeared similar. 

The results of these first two waves, which can be seen in more detail in the crosstabs here and here, are summarized in the following table.

The first wave of 5,746 voters, which was covered in an earlier article, gave Sanders a commanding lead, despite the fact that it was drawing from a segment of the population that did not appear to resemble his base of younger and minority voters. 

Other polling suggested that the Sanders vote might be more concentrated in the early voting population simply because his base of supporters were locked-in, with little need to see how the contest played out.

Yet the Sanders support increased in the second wave, likely as more Californians’ began to accept his front-runner status and the release of several polls showing him beating President Trump in head-to-head General Election matchups.

The initial results tonight  could show a strong two-way contest between Sanders and Bloomberg, setting a narrative for early in the evening that might significantly change as later votes are counted.

This growth by three points was amplified by further fracturing of the field of candidates that was increasingly becoming an anti-Sanders lane – ranging from the very ideologically similar Elizabeth Warren, who gained a couple points in polling, to the more moderate candidates Bloomberg and Buttigieg who were appearing to lose steam. 

This result in the second wave of early voting displayed a scenario that could be seen in the recent surveys by the Los Angeles Times and Public Policy Institute of California. These surveys showed a strong potential for Sanders to be consolidating votes among progressive voters, and the splitting of votes among the other candidates sufficient to keep all of them under the critical 15% threshold. 

This would earn Sanders all of the statewide delegates, and in the process likely win him more than a majority of congressional-level delegates, setting him  on a virtually unstoppable course toward the nomination. (For more about the delegate allocation process, see this prior CA120 article.)

Acknowledging that the race has changed in the last several days, we can look to these two waves combined. This will serve as a preview of what we will likely see in the early returns, especially in counties that over the weekend his set aside work on signature verification and ballot tabulation in order to focus on the administration of local precincts and voting centers.

In these results we can see a strong pattern that shows the consolidation that the Democratic establishment was hoping for. 

The initial results tonight  could show a strong two-way contest between Sanders and Bloomberg, setting a narrative for early in the evening that might significantly change as later votes are counted.

Overall, this is the picture of expected early California returns if counties post their first results with just those who cast ballots by mail.

This look at the early vote is essentially a look at the vote before South Carolina.

And as we were conducting the work on this survey, everything in the political landscape seemed to shift.  We saw Biden win in South Carolina, then Buttigieg suspend his campaign and Klobuchar abandoning her effort, and then both endorsing Biden, along with Beto O’Rourke and dozens of high-profile Democrats. 

With all these changes we decided to go back into the field and capture more responses from voters who were voting with the benefit of some of the information about what has occurred in the past 36 hours.

In these results we can see a strong pattern that shows the consolidation that the Democratic establishment was hoping for.

As can be seen in more detail with these crosstabs, three things have rapidly occurred in the electorate.

Unlike our surveys of early voters, this poll was done among voters who hadn’t been reported by registrar as having returned a ballot.  If they had voted, their ballots were likely cast over the weekend or first thing on Monday.  Voters who said they had mailed in their ballots were separated from voters who said they were going to vote at the polls or mail their ballot on Election Day.

For independent voters we had data to identify vote-by-mail voters who had the correct partisan Democratic ballot, but for those voting in person we simply asked them if they were going to get a Democratic ballot before including them in the presidential contest survey.

The results, with 930 late by-mail voters and 884 poll voters, shows a striking contrast from our earlier surveys.

As can be seen in more detail with these crosstabs, three things have rapidly occurred in the electorate.

1) Biden has jumped from fifth place with the early voters, all the way to second.  This consolidation comes from a complete collapse of the Buttigieg and Klobuchar support, something that would likely continue into Election Day as even more voters learn about their exit from the race and endorsement of Biden.

2) Warren has gained slightly, although her poll voter support seem to have slipped a bit among the poll voters.  If this kind of result occurs it should allow her to reach the 15% threshold necessary to earn statewide delegates and put her on a path to win many delegates at the congressional district level.

These changes are occurring among the moderate candidates in the contest, but don’t expect Sanders to lose support throughout the night and into Wednesday.

3) Sanders solidifies and gains votes among the more diverse and younger poll voters.  This might not be the result many were expecting, but as we have seen in our survey of voters second choices, some of Buttigieg’s supporters would be expected to move to Sanders, and he could benefit from voters seeing this race as a more binary choice between him and Biden.

4) Mike Bloomberg, seemingly being left out of this Democratic consolidation narrative, finds himself slipping, all the way from 20% and second place among the early vote, to 8% and last among the four finalists.

This sets up what could be a wild election night. 

Early results from the by-mail voters should put Biden under the required 15% necessary to win delegates at the statewide level, and make it appear that Bloomberg will become leader among voters in the moderate wing of the Party. Then, as successive waves of votes come in, we should see Bloomberg fall and, just as rapidly, see Biden rise.

While these changes are occurring among the moderate candidates in the contest, don’t expect Sanders to lose support throughout the night and into Wednesday. 

In fact, there’s reason to expect that Sanders will gain support in the subsequent days and weeks of voting given how later voters are younger and more heavily Latino.  In the end he could reach something around 35%, a result that could have won him all the statewide delegates, and most congressional-level delegates, had Klobuchar and Buttigieg stayed in the race.

Biden will likely increase his support in the following weeks as well, with a final California result that is more reflective of this two-person contest going into the rest of March and remainder of the nomination process.

And some thought in a move to a March primary, California wouldn’t matter.
—
John Howard is the editor of Capitol Weekly