With just hours until polls close in California, the crucial Democratic presidential contest between Hillary Clinton and Bernie Sanders appears to be tightening. On the Republican side, the unopposed presumptive GOP nominee Donald Trump is trying to show that he can consolidate the Republican electorate behind his candidacy.
Both make for interesting contests, albeit for completely different reasons.
The results of the online exit poll show Hillary leading in absentee votes by 10 points.
As a core part of the Capitol Weekly Absentee Voter Exit Poll, we sought to develop a tool that would allow us to survey enough of the early electorate to track support at the statewide level for each candidate and obtain similar results from each of California’s 53 congressional districts. We seek to give our readers a sense of which districts are solid for a candidate and where the battlegrounds lie, based solely on absentee voters who have returned ballots.
With more than 35,000 respondents overall who have completed the survey, we definitely have the size and depth to provide both a Republican and a Democratic presidential exit poll of absentee voters broken down by congressional district.
Here are our findings, subject to some caveats described below.
Democrats: A Late Surge, or More of the Same?
Both Hillary Clinton and Bernie Sanders can look at the polling and find challenges. See Online Survey Results Here.
The results of the online exit poll show Hillary leading in absentee votes by 10 points. This does not predict that she is going to win by that margin, but it gives us a sense of the current state of the race based on ballots already cast, and the starting point for each campaign as polls open at 7 a.m. on Tuesday morning.
Hillary Clinton 55
Bernie Sanders 45
This result is based on at least 21,554 respondents, weighted by geography, party registration, age, ethnicity and gender to match the voters who have already cast ballots as of June 4, according to Political Data Inc.
The online version provides a congressional-level breakdown, is weighted to age and will continue to collect survey responses through Election Day, accounting for some variations.
Much like traditional exit polls this year, we did find participation rates among groups with stronger support for Sanders (younger, more male) over-represented among our respondent pool relative to the demographics of the actual universe of voters who had returned their ballots.
While topline results can be informative, it is deeper within the data that we see some of the trends that have defined this contest for months. As the following two charts show, income and age are key factors for each campaign. Clinton’s strongest support is older and higher income; Sanders’ younger and less affluent.
Additionally, Clinton is showing a greater lead among women who have returned their absentee ballots, along with a nearly 40% lead among African Americans.
The current strongholds for each candidate reflect these variables. They are areas that can easily be identified as seats where income, age and ethnicity come into play and offer a distinctive character.
San Diego County, thus far among people who have voted, appears to be the greatest stronghold for Clinton.
We also find Clinton getting 55%-to-58% of the early vote in the Bay Area congressional districts and narrower margins in the LA-area. Sanders’ best regions are the far North Coast and the Central Coast. In the Central Valley, we found great disparity on a district-by-district basis.
|Best Performing Hillary Clinton Districts|
|CA36 Palm Desert (Raul Ruiz- D)||71%|
|CA51 San Diego (Juan Vargas- D)||68%|
|CA52 La Jolla (Scott Peters- D)||65%|
|CA43 Los Angeles (Maxine Waters- D)||64%|
|CA53 San Diego (Susan Davis- D)||63%|
|Best Performing Bernie Sanders Districts|
|CA24 Santa Barbara (Lois Capps- D)||59%|
|CA10 Atwater (Jeff Denham- R)||59%|
|CA1 Richvale (Doug LaMalfa- R)||63%|
|CA20 Carmel (Sam Farr- D)||63%|
|CA40 Los Angeles (Lucille Roybal-Allard- D)||67%|
Turnout appears high on the Democratic side, with 2.8 million votes cast and reported to Political Data Inc as of the Saturday before the election.
For comparison, the Saturday prior to the 2014 General Election had 2.5 million, in the 2012 Primary it was 2 million, and in the 2008 Presidential Primary there were only 1.8 million votes cast by that time.
This high turnout comes on the heels of a massive new voter registration with more than 2.3 million new and re-registrations.
Each of these factors could be the kindling and the spark prior to Tuesday’s presidential Primary election.
But, at the same time, the electorate that has turned out so far appears to be very similar to the trends we have seen in the past.
The turnout of independent voters, those with no party preference, appears low, as one would expect in a primary election.
And because of a problem with these voters getting Democratic ballots, some strong Sanders supporters might be getting lost in a paper shuffle. Thus, the independent share of the Democratic presidential primary electorate will be considerably lower than the overall independent primary vote – a factor for which not all recent polls have accounted.
If all independent absentee voters had received Democratic presidential primary ballots automatically, we suspect the Clinton advantage would have been narrowed a bit. Our poll identified around 800 No Party Preference absentee voters who wanted to vote in the Democratic primary but mailed in their ballot without having made the needed request. They favored Sanders by around a 3:2 margin.
Added to this is the fact that young voters, those who accounted for a majority of the new voter registration surge, only account for 10% of the votes returned thus far.
At the same time, Latinos, a strong part of the Clinton coalition in other states, also appears to be participating at about half-strength (typical for primaries), and polling suggests that Sanders is cutting into, or taking the lead with these voters – as they are disproportionately younger.
Sanders has maintained that he will win the state even though the demographics are not a strong match for the states in which he’s been strongest. These results indicate it will require a primary election day voter turnout considerably larger than any we’ve seen recently, among groups who historically are the least likely primary voters.
A Republican Attempt to Consolidate
On the Republican side, we can use this tool to better understand Trump’s ability to consolidate different kinds of Republican voters – from rural to urban to suburban districts, and dive into support from different income brackets, ages, gender and ethnicities. See the Online Survey Results here.
Statewide, as expected, the vast majority of Republicans are supporting Trump, with some voters casting protest votes for candidates no longer in the race, or stating that they wrote in another name, with Marco Rubio earning the spot as the top write-in.
Donald Trump 80
Ted Cruz 6
John Kasich 11
Ben Carson 2
Jim Gilmore 1
This result is weighted for age and gender, and based on the results as of Saturday, July 5. The online version is only weighted by age and will continue to update as additional voters complete surveys, accounting for some variation.
The Trump map shows strength in the northern and central portions of the state where the Republicans are white and rural, some urban seats where the Republicans are blue-collar working class whites. His weaknesses are in areas with highly educated white voters, like parts of the Bay Area and areas where there are heavy minority populations, but where working class white voters have left and some high income pockets remain.
A closer look at the districts where Trump is out-performing Mitt Romney reveals that they are some of the districts with the whitest and lowest income residents, areas with high unemployment, and potentially more disaffected Republicans who have felt left out of the state’s economic recovery.
Trump Congressional-Level Results
|CA44 Los Angeles (Janice Hahn- D)||88%|
|CA36 Los Angeles (Tony Cardenas- D)||87%|
|CA43 Santa Ana (Loretta Sanchez- D)||85%|
|CA29 Palm Desert (Raul Ruiz- D)||83%|
|CA 22 Tulare (Devon Nunes- R)||82%|
|CA12 San Francisco (Nancy Pelosi- D)||53%|
|CA37 Los Angeles (Karen Bass- D)||60%|
|CA13 Oakland (Barbara Lee- D)||61%|
|CA11 Concord (Mark DeSaulnier- D)||63%|
|CA19 San Jose (Zoe Lofgren- D)||63%|
It is also possible that the districts where he is doing best are those with the most conservative Republican voters, while those with more centrist or even liberal views are casting protest votes, or choosing to not vote at all. As the survey shows, this ideological divide is the greatest we found in the survey.
The real test for Trump is if he can use this Tuesday’s primary election as proof that he can consolidated voters behind his candidacy.
Looking at past Republican and Democratic presidential primaries, the most recent comparison might be the 2012 Mitt Romney campaign.
In that election, Ron Paul had suspended his campaign while voters in this state were already casting ballots, and at the time of the primary he was making one final push with grassroots supporters attempting to give him a win that could restart a serious candidacy. Despite some opposition, Romney still won 79.5% of the vote.
More similar to the current situation: In 1988 George H. W. Bush had hit the delegate threshold to win the nomination and earned 83% of the vote against Bob Dole and Pat Robertson who had already suspended their campaigns.
Then, four years later, Bush came to California facing another challenge from Pat Robertson, and in this election he only obtained 74% of the vote, despite having already locked up the nomination.
This second election, and his inability to consolidate Republican voters, foreshadowed his general election loss to Bill Clinton.
Currently, we have Trump slightly exceeding the Romney vote percentage, but below the 1988 successful vote percentage of the elder Bush.
Trump’s campaign will have to spin the evening’s result, whatever it is, into a story of a consolidating electorate while the Democrats are likely going to be pointing to the large protest vote, and lack of support in some key districts, as a sign that he cannot unite the GOP going into the fall.
Ed’s Note: Updates to reflect increase in respondents to absentee voter survey to 35,000 overall, 4th graf; updates respondents on Democratic side to 21,554. Pollster Jonathan Brown, a regular contributor to Capitol Weekly’s CA120 column, is the president of Sextant Strategies. Paul Mitchell, vice president of Political Data Inc., and Alan Nigel Yan, an intern from UC Berkeley, assisted with this story.