The Los Angeles Times and the University of Southern California team up to conduct and present a regular survey of voters.  This endeavor by one of California’s major newspapers, partnered with one of the state’s premiere private Universities, is a relative newcomer to this endeavor — with only a few surveys having been released.  Other statewide polls conducted by organizations such as Field and the PPIC have been at this much longer.  Over the weekend and into this week the latest LAT/USC poll results are being released — and to say that the results released thus far have been controversial would be a vast understatement…

Let me hit the "rewind" button just a bit.  Last week the latest results from a Field poll were released.  For the sake of brevity, let’s point out that Field shows the Gubernatorial race at a dead heat, with both Jerry Brown and Meg Whitman currently holding 41% of the vote, with the remained undecided.  In the U.S. Senate race, Field has incumbent Senator Barbara Boxer at 47% of the vote and Carly Fiorina at 41%.

OK, now fast forward to this past weekend, and the results from the LAT/USC survey in those same two races:  Brown 49% to Whitman at 44%; 51% for Boxer and 43% for Fiorina.

Now let me share another important piece of information — the Field Poll was conducted from the 14th to the 21st — and the LAT/USC poll from the 15th to the 22nd.

So the big question — why are the results markedly different between these two major surveys, conducted at the same time?

I always like to point out that there are two major variables that impact the results of surveys.  One of them is what and how questions are asked.  The other is who is being called to answer the questions. 

In this case, the answer to the question of why the LAT/USC poll results are more favorable to the Democrat candidates is this simple — when determining which voters to contact for their survey, they included a significantly higher percentage of Democrat voters, say the Whitman and Fiorina campaigns!  (There are some other questions about the large number of questions that the LAT/USC survey asked before these major races were queried, and how those preceding questions could have influenced the results as well).

Both the Meg Whitman for Governor campaign and the Carly Fiorina for U.S. Senate campaign released memos on this issue — both of which made the case for why the Field results are more credible (the Whitman memo from their pollsters goes into a bit of detail).  I have included both memos below for your perusal.

It is worthy of mention that over at CalBuzz (where proprietors Jerry Roberts and Phil Trounstine — former MSM reporters — now flirt with lawsuits as they attempt to slot every California politico as a character from Tolkien’s LOTR) they heartily disagree with these criticisms of the LAT/USC poll.

I will conclude by saying that it is clear that both of these races are too close to call.  In the case of the Governor’s race, given that Jerry Brown coasted through the primary with no opponent, while Meg Whitman was smacked with tens of millions in negative campaigning from Steve Poizner, the fact that "Moonbeam" doesn’t hold a considerable lead should give his campaign and his supporters cause for pause…  And where the U.S. Senate race is concerned, this race is also way too close to call — bad news for an incumbent Senator running for a fourth term.  It is also significant to note that when these surveys were taken, Boxer was running a statewide ad pummeling Fiorina, and Fiorina’s first ad (which is up now) did not hit the airwaves until after these survey interviews were done.

First the Whitman Memo…  Then the Fiorina Memo…

TO:    Meg Whitman campaign staff
FROM:      David B. Hill, Ph.D., Director, Hill Research Consultants
                   John McLaughlin, CEO,McLaughlin & Associates
DATE:    September 26, 2010
SUBJECT:    LA Times Poll of California Voters

There is strong evidence why the LA Times/USC Poll is out-of-step with the California Field Poll, despite the fact that they polled voters at the same time. The Field Poll conducted interviews from September 14-21 and the LAT/USC Poll was in the field September 15-22. The most significant difference is that the LAT/USC poll under-sampled Republicans. That’s why its results differ so sharply from Field and all the other recently reported public polls, as well as our own internal polls. Field provides a far truer representation of California’s 2010 electorate, in our estimation.

The Times poll reports party identification in two ways, following a pattern used by the Democratic polling firm that computed the results. Sometimes they report all DTS voters separate from partisans (their "2-3-2" party ID) and sometimes they report DTS voters who lean toward a party as partisans (their "3-1-3" party ID). On average, across both methods and for registered and likely voters, the poll has a Democrat advantage of 16.75 percentage points, far too much.



The Field Poll’s party profile for likely voters is 44 percent Democrat and 35 percent Republican, reflecting a Democrat partisan advantage of only 9 percentage points. Brown’s unexpected lead among all voters and likely voters in the outlier LA Times/USC poll may be solely attributable to the fact that its various samples reflect an unrealistic 14- to-20 point Democrat advantage.

This table compares the two polls’ results for party and ballot:



We are also skeptical of the LAT/USC poll because the ballot for governor wasn’t asked until the 19th question in the interviews, allowing for the introduction of partisan bias through the preceding questions. In the voting booth, voters won’t get those 18 additional cues before voting. We also note that questions #4 and #6 have been redacted from the publicly reported results, adding to the difficulty in judging how much bias was introduced.

Bottom line: We do not feel the LA Times poll accurately reflects the current state of the race. We stand by our internal polls, which more closely align with The Field Poll.

MEMORANDUM
 
TO:            Interested Parties
FR:            Martin Wilson
                  Campaign Manager, Carly for California
RE:            Los Angeles Times Poll/CA Senate Race Update
________________________________________________________________________
 
The just-released poll by the Los Angeles Times is neither an accurate nor a reliable reflection of voter interest or likely participation in the upcoming election.
 
Simply put, the Times poll overestimates Democratic voter participation by a wide margin, hence it produced skewed results that are inconsistent with other public and private polling in this race, including the respected Field Poll released on Friday. Specifically, the poll’s sample indicates that self-identified Democrats will hold a 16.75 percent advantage over Republicans, which is at least double what other credible polls indicate. In addition to understating the Republican vote, its results are based on 9 percent participation of independent voters when, again, other pollsters estimate that at least 20 percent of the electorate will be comprised of swing voters. 
 
With such an errant sampling methodology, we are amazed that Boxer’s lead over Carly is not in the double digits as opposed to the eight-point advantage they are giving the incumbent. At this stage and in light of the dynamics of the race during the time this survey was in the field, that is a gap that can, and will, be bridged in the coming weeks.
 
For 10 days, while this survey was being conducted, Barbara Boxer spent several million dollars on unanswered advertising mischaracterizing both her own record and Carly’s record. It is notable that, despite this, Barbara Boxer was unable to increase her standing with voters in the Field Poll. We are now answering back with the launch of our advertising campaign just 72 hours ago, after both the Times and Field polls had been completed. As this campaign moves forward and as voters become aware of the facts about Barbara Boxer, the hyper-partisan career politician, the dynamics in this race will change quickly. Over the course of the next 38 days, the spotlight will be squarely on Barbara Boxer and the low road she has taken toward achieving higher office. Carly will stand in stark contrast to Boxer’s record of failure: a bipartisan problem solver who will use her real-world business experience to find common-sense solutions to solving our state and nation’s problems. As that happens, the numbers in this race will move, so hold on to your hats for an exciting and competitive race to November 2.

This entry was posted on Monday, September 27th, 2010 at 12:00 am and is filed under Blog Posts.