Eureka

Proceed with Caution - Beware of Early Polls

Wednesday, April 9, 2014

Early polls might generate buzz, but reporters, voters, and politicians should view these polls skeptically.  They are unlikely to predict the June primary results very well for three cascading reasons.

Few people are paying attention to the election at this point in the cycle.

In the last five Field and PPIC polls that have asked a head-to-head June primary question, on average 27% of the electorate is undecided (shifting very little between December and the spring).

In PPIC’s January survey, only 28% of likely voters (and just 28% and 21% of likely Republican and Independent voters, respectively) were following election news very or fairly closely.  A statically similar segment of Republicans and Independents were “not at all closely” following the election.  This suggests that roughly half of voters who expressed an opinion in the head-to-head match-ups are following the election “not too closely.” If voters are not really paying attention to the election, do not expect the polls to be an honest portrayal of future outcomes.

And maybe more importantly, based on the PPIC polls from December 2013, January 2014, and March 2014, a quarter of voters are unsatisfied with the choices of candidates.  Don’t expect these voters to pay much attention to the contest; if they don’t like the choices, why would they invest the time?

Because few are paying attention, the candidates, besides Governor Brown, are largely unknown.

Jerry Brown, after a long political career, is a known entity.  The latest Field Poll has just 4% of likely voters offering no opinion of the 3rd term Governor.  However, in the same poll, 50% of likely voters had no opinion of Assemblyman Tim Donnelly, 64% had no opinion of Neel Kashkari and 71% say the same about Andrew Blount.  Even among likely Republican voters, these values are roughly the same.

While Donnelly has seen a bit of bump between December 2013 and April 2014 in the head-to-head matchup, he’s also seen his name recognition increase, while competitor Neel Kashkari’s has largely remained the same.

Early polling is more indicative about baseline name recognition than actual support.

At this point in the election cycle, campaigns have spent little to nothing on candidate advertising, meaning voter impressions are based on little information.

In February 2006, State Treasurer Phil Angelides held an 8 point lead over primary opponent State Controller Steve Westly. Among name recognition, Angelides held a 9 point advantage with likely Democratic voters—statistically the same as his head-to-head margin.  While Angelides would go on to win, his margin (4.8 points) of victory was well below that of his early 2006 lead.

In March 2010, eventual nominee Meg Whitman held an 18 point lead in name recognition over main competitor State Insurance Commissioner Steve Poizner. At that point in time, she yielded a wide margin of victory in the polls—49 points. However, her commanding lead in early polls overestimated her support.  Whitman only won the nomination by a margin of 38 points.

In the other 2010 primary race – for the U.S. Senate – eventual nominee Carly Fiorina actually trailed major primary opponent former Congressman Tom Campbell by 6 points (just outside the poll’s margin of error) as of March 2010.  At this point, both Campbell and Fiorina were equally known entities, helping to explain the close head-to-head.  Another viable challenger, Assemblyman Chuck DeVore, significantly trailed both Campbell and Fiorina in the head-to-head matchups and in name recognition.  Fiorina would go onto to beat Campbell by 35 points in the primary with DeVore coming within 3 points of Campbell.

Overall, the spring polls’ front-runner preformed an average of 19 points worse on primary day than in the early surveys. Why?  Because campaigns don’t spend resources until they are sure people are listening.  Hence, major primary electioneering rarely occurs prior to May.  It is then that voters’ opinions of the candidates move from just recognizing them to supporting them (or not).

So, is Assemblyman Tim Donnelly actually the Republican frontrunner for the #2 spot on the November ballot?

Possibly, but 2 months is a long time in politics.

For one, Donnelly’s support is soft.  His jump in support between December and April is likely explained by his increase in name recognition.  While the March PPIC poll has Donnelly about 15 points ahead among Republicans and 9 points ahead among Independents over Kashkari and Blount, almost 3 times as many Republicans and 4 times as many Independents are undecided than support Donnelly. Even among Republicans, Donnely runs statistically tied with Governor Brown.  In the April Field Poll, just as many Republicans, Independents, Conservatives, and Tea Party supporters are undecided as support Donnelly.

Donnelly could easily see his support stagnate if his challengers can increase their profile positively and hit Donnelly.  Fundraising results suggest that Neel Kashkari should be able to overwhelm Donnelly. Combining both cash-on-hand reported as of mid-March and fundraising since the mid-March filing, Kashkari holds a 16 to 1 fundraising advantage over Donnelly.  In addition, Donnelly continues to make headlines for all the wrong reasons—comparing illegal immigration to war, a possible past larceny conviction, and allegedly violating probation for carrying a loaded gun through airport security—which could prompt backlash and a stronger effort to advance Republican alternatives to the Assemblyman.

Early polls may help candidates create buzz, build momentum, boost fundraising, and gauge early support.  But they offer little insight into what voters are actually thinking and what to expect from them come primary day. All should progress with caution until more people are paying attention and aware of their choices.

Follow Carson Bruno in Twitter: @CarsonJFBruno