Cage Match of the Marks: Senate Polling Miasma


loretta sanchez Kamala HarrisBehind the news:
California A.G. Kamala Harris remains the favorite against Representative and sister Democrat Loretta Sanchez in the race for the U.S. Senate seat being vacated by Senator Babs Boxer.

However, California’s top two pollsters – Mark DiCamillo of the Field Poll and Mark Baldassare of the Public Policy Institute of California (Mark I and Mark II), differ sharply in their findings on how close the race stands.

Field has Harris over Sanchez 42-20%, a gigonda 22-point lead and a big boost from Field’s July survey, which Harris led by 15 points at 39-24%. However, PPIC has it Harris leading 32-25%, a narrower 7-point gap and a tightening from their July survey when Harris led by 18 points, 38-20%.

Amid drastic, previously unannounced changes in survey methodology (Mark I), and the failure to use the ballot designations voters will see (Mark II), the most inexplicable difference between their polls is the huge contrast in their findings among Latino voters – a factor that dramatically affects the MSM narrative about the race (and as we note later, the Vegas odds).

twomarksQue pasa? In a nutshell: Field, who had Sanchez leading Harris 32-21% among Latinos in June and 49-24% in July (which intuitively made sense), now – after changing from its gold-standard practice of calling actual voters to an online experimental method – shows Harris ahead among Latinos 35-34% (which intuitively seems whacko). PPIC, who had Sanchez leading 45-29% among Latinos in July, now shows Sanchez crushing Harris among Latinos 58-16% — a massive 42-point advantage in that important voting bloc.

That, dear Calbuzzers, is a head-snapping difference. If all you care about is who will win the Senate seat, fine and dandy – Harris is your best bet. But if you care to understand why or what affects California voters, this is a disturbing lack of clarity.

One problem Field has with its new methodology was that finding Latino voters able and willing to participate in an on-line survey meant relying on fewer than half as many Latino voters as PPIC reached on the phone. And whether those on-line Latinos were typical of Latino voters generally is a pretty sketchy prospect.

For political professionals, pollsters, analysts and campaign technicians, understanding the dynamics that affect a race, especially among target populations, is important stuff. When top-ranked surveys differ so radically, no one knows what or if anything has influenced voters.

Confusion PortraitAt least somebody’s awake. The alert Matthew Artz at the Mercury News pressed DiCamillo about the shift in the Latino vote:

“You can see that among Latinos, the initial support for Sanchez had a lot to do with the recognition that she had a Latino surname,” DiCamillo said. “That put her in the early lead (with Latinos). Now more voters at least have some inkling of who the candidates are.”


Since not much “on the ground” has happened in the race that seems to make this a logical explanation, beyond a few come-and-go big name endorsements for Dem establishment candidate Harris (starting with President Obama), Calbuzz asked DiCamillo to further explain his analysis:

“It was my judgment that voters, in the absence of lots of specific information about candidates, look for cues to guide their judgments. Usually the strongest cue that they rely on is party.  However, in this race both candidates are of the same party. So, they then look for other cues. Often times gender differences can be a deciding factor.

“But again in this race both candidates are of the same gender. So, it is my belief that one of the more powerful cues in the early polls was the differences in the names of the candidates, especially among Latinos, since they could recognize Sanchez as a Latina,” DiCamillo said in an email.

“However, as the candidates become better known, it is my belief that the influence of the name as a cue recedes and is replaced by other information that voters have learned about the candidates.

“In our latest poll, we see that majorities of voters, including a majority of Latinos, can offer an opinion about both candidates.  And, the very positive image that Latino voters now have about Harris is for many Latinos now overriding the power of their earlier preference for Sanchez based on her being a Latina.  It’s obviously still a factor, just not as big.”

Um, okay.

This is most uncharacteristic conjecture by DiCamillo. We say conjecture because there are no data to back it up and, well, it kinda defies logic – to suggest that Latino voters would abandon a credible Latina congresswoman.

To us, it makes more sense that the Field Poll’s sex change switch to an online survey methodology captured the opinions of atypical Latino voters. Unfortunately, the Field Poll didn’t ask a favorability question about Harris and Sanchez in its June or July polls, which might shed some light on the matter, and PPIC doesn’t ask a favorability question at all.

M8DADOF EC056Mystery deepens. In Field’s latest survey, 98% of Latinos had an opinion about Sanchez and it was 57% favorable and 41% unfavorable — +16%. Among Latinos, 91% had an opinion about Harris and it was 71% favorable and 20% unfavorable — +51%. Again, that’s with the unexpected and unfamiliar internet panel.

The last time we know of that Field asked favorability about Harris and Sanchez was eons ago politically, in January 2016 in live interviews by phone, Sanchez had a 49-13% favorable rating among Latinos (+36%) and Harris had a 37-14% rating (+23%). More Latinos had an opinion about Sanchez (62%) than had an opinion about Harris (51%).

Just for good measure, DiCamillo has pointed at least some California reporters to the fact that PPIC didn’t include ballot designations in its questions, i.e., they didn’t refer to Kamala Harris, Attorney General of California, and Loretta Sanchez, United States Congresswoman, unlike Mark I.

That’s a five-yard penalty in the Calbuzz Official Polling Rule Book, too. But at least PPIC has been consistent in how it refers to the candidates over time and their entire play book hasn’t radically switched in the middle of the game.

sad elephantGOPers vote with their feet. The two Democrats advanced from the June open primary, leaving many Republicans without a candidate they care to support; this might help explain PPIC’s Mark’s most salient issue in his poll: 43% of voters say they won’t vote at all in the Senate race, or they don’t know who they’d support. That compares to just 6% in the presidential race.

As for the notion that Latinos have shifted to Harris because they no longer are persuaded by Sanchez’s surname:

“I don’t have any evidence to show that, and as an observer, there hasn’t been the kind of activity or coverage in the media that would suggest there’d be any significant shift in candidate preference.”

“We’ve done three polls and in each one Harris is ahead and a lot of people don’t know what they’re going to do,” he said. “Latinos are supporting Sanchez by sizable margins in our polls,” he added.

Interestingly, two-thirds (66%) of Latinos say they’re satisfied with their choice in the Senate race, compared to just 42% of whites, according to PPIC.

A brief digression. (As long as we’re ranting about polling, here’s a factoid for all the Democratic bed-wetters about the national polling that shows Hillary Clinton only slightly ahead of Despicable Donald: the average winning percentage of presidential candidates since 1824, when the people started electing presidents directly, is 51.36%. What matters is what happens in the Electoral College, which means Florida, Ohio, Pennsylvania, Michigan, North Carolina, Iowa, etc. Keep your eye on the ball, people.)

bookieCalbuzz – Too Big to Fail? There’s one more, very practical, reason why the DiCamillo-Baldassare cage match matters: their polls shape how we set the betting line for the Senate race.

Sure, the contest as of today looks like a walkover for Queen Kamala, but the divergent surveys suggest there’s still volatility with more than a month to go. So there’s plenty of action to be had laying down on the Over-Under – i.e. wagering on the margin of victory (N.B. for those who like long shots, Vinnie “The Vig,” Capelli d’Angelo, our Las Vegas Bureau Chief and Sporting Life Consultant, is also offering 35-to-1 odds on a Sanchez win).

When last heard from, our bookie was fleecing us on our daft and rash wager favoring the low-energy Jeb Bush to win the Republican presidential nomination.

Full disclosure: Our Bushman bet made us the web site version of Lehman Bros: when a sharp downturn in advertising revenues (btw: Buy Now – Crazy Calbuzz is still offering 2009 rates! Plenty of free parking) left us a little shy of the requisite cash, he started compounding the vigorish on a weekly basis. Let’s just say he’s getting a little impatient for his money.

So we’re looking to make a score on the Senate race Over-Under: our number today is 10 ½ points – if the winner finishes 11 points or more ahead of runner-up, you win; 11 and under, you owe us.

Place your bets.

P.S. Further disclosure — We already owe Mark I dinner for a bet on Clinton’s victory margin over Bernie Sanders which was, in the end 7.1% We were closer, but over. Aagh, more nonstop compounding interest.

subscribe to comments RSS

There are no comments for this post

Please, feel free to post your own comment

You must be logged in to post a comment.