After a long process, the voting primary states have nearly ended. California is not only the biggest prize in the calendar on June 7th, it is also the largest with 475 pledged delegates. Senator Sanders is hoping a win here will weaken Secretary Clinton’s argument for the nomination while Secretary Clinton is hoping for not only a win but a decisive win that will show she is still widely popular among the Democratic base (as head to head polls currently show) and that she is the clear choice of the Democratic Party.
This is an update from an earlier analysis posted here. This is part of a series of election predictions. Pennsylvania’s prediction is here and New York’s prediction is here.
California has demographics that strongly favor Secretary Clinton. Despite the strong population of young people with California’s many colleges and universities and the strong progressive streak, Secretary Clinton has the narrow advantage here. The USC poll showed Secretary Clinton leading by 10 but the trusty Field poll showed Secretary Clinton up by 2 (although Secretary Clinton won the CA primary by eight in 2008 and the Field poll showed her up by 2). Polling Latino voters in California has been historically difficult, specifically with those who do not speak English well and the same with AAPI voters, especially older ones who may speak their native country’s language and therefore not get picked up by polls. The results for Arizona among Latinos were also used to predict the results in California (Secretary Clinton overwhelmingly won Arizona Latinos).
To predict Secretary Clinton’s win in California, I looked at the 2008 map where she won California by 8 points. I predict a roughly 6 point win here in 2016 not far from Nate Silver’s +5 prediction and Nate Cohn’s +7 prediction. Secretary Clinton will gain votes with upscale Democrats (West Coast) and with African American and lose votes with rural white voters (the Sierras) and with young Latino voters. Also, this exit poll here of absentee voters shows regions where both Clinton and Sanders will perform well. Clinton won absentees by 12 but as admitted by the pollster, young voters were overrepresented in the poll compared to the actual amount of absentee ballots they cast so it could be underestimating Secretary Clinton by a few points, as we expect with the statewide polls. The poll was also conducted entirely by email which could undersample lower income elderly voters who support Secretary Clinton. Also, Senator Sanders is likely to perform better with voters who vote on Election Day because they lean younger. While exit polls have not always been accurate with regions (although demographics are usually pretty accurate,) they do provide a hint as to how an area will vote. Also, polls historically undercount Latinos and AAPI voters and since many of the polls do not mention that they conducted polls in Korean, Chinese, Vietnamese, Japanese and so forth, it is possible some older AAPI voters who immigrated to the U.S. and primarily speak their native language may be undercounted but are older voters and should be supporting Secretary Clinton. My prediction is based more on demographics.
This is my prediction for the 2016 Democratic primary.
Dark Blue = Solid Clinton
Light Blue = Lean Clinton
Grey = Tossup
Light Green = Lean Sanders
Solid Green = Solid Sanders
2008 Primary Results as Comparison:
Northern California:
Humboldt County:
This county is a quintessential Bernie County. Not only is it heavily white (78%) and rural, it also has a large number of progressives (they are not as upscale as the Marin County progressives) and has Humboldt State University. I expect Sanders to break +20 here.
Mendocino County:
Mendocino County should vote for Senator Sanders as well but it is less favorable to him than Humboldt County. The reason is that due to the large number of wealthy Bay Area transplants here and the growing influence of the wine country (I expect Napa County which has a large number of Latinos and upscale Democrats to support Secretary Clinton) Secretary Clinton will have a base here. What will allow Senator Sanders to carry Mendocino though is the rural working class voters inland in Ukiah and Willits. He will also be helped by longtime residents here as well as baby boomers who settled here in the 1970s. Overall I predict a Sanders +14 here.
Border Counties:
The counties of Del Norte, Siskiyou and Modoc are all heavily white, rural and have large elderly populations (over 20% in many of them). This will allow Secretary Clinton to not face a blowout here (similarly to Oregon’s southern counties) but should not be enough for her to win any of them.
Bay Area:
Alameda County:
This county will narrowly lean Clinton. On one hand, it has heavily African American areas (12% countywide) in Oakland and wealthier areas (Dublin) that will vote strongly for Secretary Clinton. Alameda County also has a 23% Latino and 29% AAPI population. At the same time, Alameda County is also home to Berkeley which is probably one of the strongest Bernie cities in the state. It is home to extremely progressive and young Democrats (and I saw Jill Stein at an event there). The southern part of the county is more AAPI which should favor Secretary Clinton but Berkeley has high Democratic turnout and the hipsters in Oakland may offset the African American voters there supporting Secretary Clinton. Overall I predict Secretary Clinton +2 due to the southern part of Alameda County saving Secretary Clinton.
Contra Costa County:
If any county in the Bay Area were made for Hillary, it would be this one. Not only does Contra Costa have a 10% African American population (it is home to heavily African American city of Richmond), it also has very upscale suburbs similar to the Philadelphia suburbs and the Upper East Side that Hillary won. There are a few working class areas (Martinez) but they should be outvoted by the wealthier areas (Lafayette, Walnut Creek, Danville). I predict a +13 Secretary Clinton win here.
Marin County:
As a resident of Marin County, I am so excited to finally discuss how my county will vote. At a first glance, Marin County appears to be strongly for Bernie. It is 72% White and very progressive (it was the strongest Obama primary county in California in 2008, backing him by 19 points). Environmentalism is the key issue here, we love hiking in the redwoods on Mt. Tam, we love our salmon, we love protecting our unspoiled coastline and we love our four Whole Foods markets. One town, Fairfax is a proud town with many former hippies very proud of its heritage (and West Marin is very similar with many rural progressives). On the other hand, there are some very wealthy areas in Marin County (Tiburon, Ross) which are still Democratic but many of the Dems there are more moderate and similar to upscale Democrats Secretary Clinton won on the East Coast. The average age is also 44 years old. The bellwether here is Mill Valley which is home to many progressives and many upscale Democrats (there are two Whole Foods within a mile of each other). At the same time, Marin County may appear very pro Secretary Clinton on a data model due to demographics similar to Montgomery County, PA, Montgomery County, MD and Westchester County which overwhelmingly backed Hillary, I expect Marin County to be very close due to the strong progressivism and activism many of the Democrats here have. While I expect the West Marin/Fairfax contingent to post a strong showing for Bernie, Marin will support Hillary thanks to supporters in the southern part of the county but very narrowly with Secretary Clinton winning by +2.
San Mateo County:
The key to determining how San Mateo County votes is by looking at the divide between the areas along the Bay which are full of young tech workers. The upper parts of the county though are more older (Hillsborough, Woodside) and more wealthy which would imply a win for Secretary Clinton. San Mateo County is currently labeled as “too close to call” thanks to the large numbers of tech workers here (and Senator Sanders is also helped by the working class residents of South San Francisco). Secretary Clinton also won here in 2008 by six points despite losses among the young tech workers (although provided there are more now than in 2008). Overall, I give this county Clinton +2 but with the potential closeness it is marked as “too close to call” on the map.
Santa Clara County:
At a first glance, Santa Clara County may appear to be a swing county due to the large number of young people working in Silicon Valley. Once one leaves the valley however and heads to the mountains, the demographics change and become more upscale (Campbell, Saratoga, Mountain View etc), which favors Secretary Clinton. As shown in Oregon, Sanders won wealthy suburban areas in Portland (Clackamas County) by three points (I am using it to compare upscale Democrats on the West Coast) but Clackamas County is not as wealthy as parts of Santa Clara County and it does not have the minority population Santa Clara has. Also, Santa Clara County has a significant Latino population (26%) which is the largest in the Bay Area. The AAPI population here is 35% and while there is little data on how AAPI Democrats vote in the primary (especially Vietnamese due to the large Vietnamese population here), the best data we have is the NY primary results where Secretary Clinton won 60%-65% in heavily AAPI neighborhoods. While Santa Clara County and San Mateo County are similar demographically in some ways, what gives Secretary Clinton a higher percentage here than in San Mateo is the higher Latino population. Also, Secretary Clinton won by 13 points here in 2008 despite President Obama winning among the young tech workers. In terms of polls, the best data here is a Clinton +13 poll here which is a bit out of line with our Secretary Clinton +7 win here.
San Francisco County:
While demographics of many California counties may appear obscure, what is not obscure is the San Francisco demographics. With a large population of progressive voters and young voters, San Francisco is prime territory for Senator Sanders. At the same time here, he will not win a blowout here that he will need to counterbalance Secretary Clinton’s wins in SoCal. The reason is that many of the wealthy voters in the Pacific Heights will favor Secretary Clinton. The AAPI voters in the Sunset District will probably favor Secretary Clinton (but as shown by heavily AAPI areas in NY, probably not by large margins). Overall, I predict a +9 Sanders win here, thanks to margins from young voters and from progressives.
Central Coast:
This region contains Santa Cruz, Monterey and San Luis Obispo Counties.
Santa Cruz County:
Santa Cruz county is similar demographically to Marin County. There are many environmental progressive voters here as well as UC Santa Cruz and Watsonville which is heavily Latino (Santa Cruz County is 30% Latino). While Watsonville should lean Secretary Clinton, it should be outvoted by the rest of Santa Cruz County which I expect to heavily back Bernie. I expect Sanders +15 here.
Monterey County:
Monterey County is 57% Latino, it also has a large upscale white voter population. At the same time, it is important to note that due to the low turnout rates of Latinos, a 57% Latino population does not guarantee a Latino majority in voting, even in a Democratic primary (in a general election a 65% Latino area may have a majority Latino vote) due to the age of many of the voters. Also, many of the white voters progressive similarly to Santa Cruz County. I expect a +4 win here for Senator Sanders.
San Luis Obispo County:
With Cal Poly, a smaller Latino percentage than nearby counties (and the smallest percentage in a coastal county south of the Bay Area) and a large number of rural progressives, San Luis Obispo County should vote strongly for Senator Sanders. I predict Sanders +15 here.
Central Valley:
Sacramento County:
Sacramento County will probably vote for Secretary Clinton but not by a large margin. The reason is that there is a large young population here in Downtown Sacramento. At the same time, this is the state capitol with political people who would be more likely to support Secretary Clinton. Not to mention the African American population here is 11% and the Latino population is 23%. Compared to Marion County in Oregon with Oregon’s state capital which narrowly backed Bernie and had a similar Latino population, Sacramento County should back Secretary Clinton by roughly +8 thanks to the additions of the African American and AAPI populations.
Stanislaus County:
If there is one farming county Senator Sanders wins, it will be this one. As shown in the 2012 election results, President Obama won here despite Stanislaus County having a 44% Latino population here (counties in the southern Central Valley with Latino populations in the 50s voted for Romney). This shows that there is a contingent of white Democrats here that may be much smaller in southern Central Valley counties. Also, many of the white Democrats here are working class, giving Senator Sanders an opening. I still expect a Secretary Clinton by +2.
Merced County:
Despite the 58% Latino population here, this is a county where Senator Sanders may make a close race. The reason is that UC Merced is here and Senator Sanders performs very well with college students. I predict a +4 win for Secretary Clinton with older Latinos offsetting margins from UC Merced. I also see Secretary Clinton performing well with Latinos connected to agriculture due to her strong support from Dolores Huerta and Cesar Chavez’s son in law.
Kern County:
This county is heavily Latino with a large young population as well (30% of the population is under 18). At the same time, many of the Latinos here are involved in the farmworking industry which favors Secretary Clinton due to the strong support she received in heavily Latino rural areas. Also, the white population here is strongly Republican, preventing Senator Sanders from narrowing the margin here with white working class voters and there are no major universities compared to Merced County. Also, Senator Sanders has never won rural Latinos, he has shown an ability to appeal to urban Latinos to some extent but has never won a heavily Latino rural county. I predict a +15 Secretary Clinton win here.
Southern California:
Santa Barbara County:
In California elections, Santa Barbara County is a bellwether. It is a good representation of the state. It has a liberal university, it has wealthy liberal voters, it has heavily Latino communities in an urban area (East Santa Barbara), it has heavily Latino communities in an agricultural area (Lompoc, Santa Maria), it has upscale Republicans (Montecito), and rural Republicans (Santa Ynez). In the primary though, UCSB and the rural Democrats in the northern part of the county will outvote the wealthy Democrats in the Montecito area. My prediction is Senator Sanders by +7.
Ventura county:
As recently as 10 years ago, Ventura County leaned Republican but demographic changes have changed that. Not only has the Latino population grown to 42%, the formerly heavily Republican suburban areas such as Thousand Oaks have become more purple instead of solidly red (although Simi Valley is still solidly red) and while parts of the county are wealthy, the city of Ventura is liberal and should lean Sanders. Oxnard and the Santa Paula area are very Latino and connected to farmworking and should help offset the city of Ventura voters. I predict a win of Secretary Clinton +3 here.
Los Angeles County:
If Secretary Clinton wants a large margin out of California, Los Angeles will be the county where she needs it and she will get it. Besides Silver Lake and other young areas in LA County, there are many younger Latinos would could provide an opportunity for Senator Sanders to narrow the margin among Latino voters for Secretary Clinton. There are upscale wealthy areas but they are unlikely to back Bernie (except maybe Santa Monica) due to not being as “anti establishment” as the progressive areas in Marin and Santa Cruz Counties. Not to mention Los Angeles County has a 48% Latino population and a 9% African American population that is very politically active. In 2008, Secretary Clinton won California by 12, I expect her to gain among African American voters and upscale white voters and lose a little ground among Latino voters. Therefore, I predict a +11 win for Secretary Clinton here.
Riverside County:
Riverside County is divided between a heavily Latino area in the western part of the county and a more upscale retirement area with Palm Springs in the eastern part of the county. The Latino votes are hard to predict as shown earlier but should support Secretary Clinton overall. Overall, I predict a +12 point win here for Secretary Clinton and a larger margin than San Bernardino because of the Palm Springs area.
Orange County
Orange County on one hand is home to a large Latino population (34%) and AAPI population (20%) and also has many wealthy coastal areas as well (Newport Beach, Laguna Beach etc.) that should favor Secretary Clinton. Also, one factor that may have been undercounting Secretary Clinton’s support in the polls is her strength with AAPI voters. She won majority AAPI precincts 58%-42% in New York City and won AAPI voters 70%+ in the 2008 California primary. It is possible that Secretary Clinton leads among AAPI voters in CA but many of her supporters who are older and do not speak English well are getting picked up in the polls. There is no data among how the Vietnamese community will vote in regards to Secretary Clinton but it is very likely that older Vietnamese voters who emigrated from Vietnam may not be picked up by polls. Secretary Clinton should be able to at least receive a +10 margin here thanks to margins from the coastal areas. This is another county that has many demographics that are favorable to Secretary Clinton.
San Diego County:
San Diego County has a large college aged population and has a larger white percentage than other SoCal counties (at 47% which also means that not a single county in SoCal is majority non Hispanic White). The 33% Latino population though is very helpful to Secretary Clinton and the 6% African American and 12% AAPI do not hurt either here. While the demographics favor Senator Sanders, the exit poll of absentee voters does not showing Secretary Clinton winning above 60%. That high margin is unlikely thanks to the county’s demographics but still suggests a Clinton win which could be due to upscale voters having a strong presence here. Overall, Secretary Clinton should aim for a +9 win here.
Imperial County:
Predicting how Latinos will vote in California is difficult with Latinos having different voting patterns across the U.S. In Illinois, Sanders won the 75% Latino 4th district (but most of the Latino population was not Mexican) and Secretary Clinton won in heavily Latino parts of Florida and Texas (and probably won in Nevada but others debate that and that is another discussion for another article). I settled on Arizona, due to the similarities in the Latino population and the fact that both states are not caucus states (Nevada is). Yuma County and Santa Cruz County, the two heavily Latino counties in Arizona supported Secretary Clinton +29 and +32 respectively (similar to their 2008 primary results). While Senator Sanders has a chance to make inroads among urban Latinos, he has not shown this ability with rural ones. This is important in a 82% Latino county such as Imperial County. Therefore, I am predicting a +30 margin here for Secretary Clinton but Senator Sanders could perform well with Latino voters statewide by performing better with more liberal urban ones.
Conclusion:
Overall, Secretary Clinton should expect a +6 win statewide. She will play well in SoCal with the best Sanders areas being Santa Barbara and Ventura Counties. She will win the cities in the Central Valley (with the potential exception of Stanislaus County (Modesto) and Merced County) and while she will lose the Sierra rural areas, she may not lose by the large margins she lost by in rural western caucus states thanks to the large elderly population in these areas.
In NorCal, the race will be more split with Secretary Clinton winning the East Bay in the Bay Area, losing San Francisco and keeping it close in Santa Clara, San Mateo and Marin Counties.
In Coastal California, Bernie will win all the coastal counties.
California will be a gripping end to the primary season and while we may not know the final tallies for over a month (California takes about a month to finalize the provisional ballots and counts absentee ballots first which should favor Secretary Clinton in the early count), the absentee ballots and the Election Day votes should be enough to allow Secretary Clinton to declare victory but she may have a larger lead in the early votes and if it is higher than 12, compare it to the exit poll showing her up by 12 among early votes.
Overall, the main reason I believe Secretary Clinton will win by 6 points in California is that in 2008, she won by eight points showing that she has a strong base here and Senator Sanders would need to make inroads among one of her demographics. I have not seen how Senator Sanders can do it in a large enough way to win, especially because there is no strong surge of young voters in the absentee ballots compared to previous years and because there are polls likely undercounting older Latinos and AAPI voters who are likely to support Secretary Clinton.