« More On That LA Times Poll | Main | Students for Kerry »

Yet More On That LA Times Poll

As a public service, I reproduce the reply (originally in ABC News' The Note) of Susan Pinkus, Director of the Los Angeles Times Poll, to Matthew Dowd's allegations. Note her point about not weighting by party ID, as well as the useful time series data on party ID in the LAT poll. And note that the Democrats' current 13 point lead is not out of line with the previous LAT poll data.

After reading Matthew Dowd's assessment of the Los Angeles Times Poll in ABC News' The Note, I feel that I have to respond to his assertion that the poll is a 'mess.' His negative spin of this poll is, quite truthfully, not unexpected. The Times makes every effort to use sound methodological techniques that are used by most reputable research and polling organizations. The questionnaire and methodology is available for anyone to see and conforms to the guidelines set forth by the National Council on Public Polls and the American Association for Public Opinion Research. Although Dowd does not like the results of the Times poll, I stand behind the poll's results and the sound statistical methods used.

If Dowd doesn't like the Times results, did he have a problem with the latest Gallup and CBS/New York Times polls? The horserace numbers are similar to the results of these two latest polls. Gallup had Kerry ahead by 5 points in the two-way race and CBS had Kerry up by 8 points.

The Times does RDD (random digit dialing) sampling which reaches households with listed and unlisted telephone numbers. The poll weights slightly (for minor corrections) based on census data for sex, race, age and education and does not weight for party ID. Party ID is a moving variable that changes from one election to another, and weighting by party registration makes no sense nationally because many states don't have their voters register by party and some states don't have voters register to vote until the day of the election.

Here is the breakdown of party affiliation in Times polls going back to September 2001:

DATENDEMINDREPSD
06/04438242572
03/044332625102
11/035312526123
04/033381926104
02/034283026104
12/02433282792
08/02335262883
02/026312627112
11/015342825104
09/01538202683

As you can see, the numbers are pretty similar to one another (all within the margins of error).

It is also interesting to me that if the poll is 'a mess,' why is he reporting data from the poll -- results that hopefully make his point? Why doesn't he report that the job approval ratings are very similar to those that other polling firms are finding? For example, a new Fox poll released today shows Bush's job approval 48 percent to 45 percent disapproval. Annenberg's numbers show Bush's approval to disapproval ratings at 48 percent to 49 percent. The Times poll has Bush's positive to negative ratings 51 percent to 47 percent. Annenberg also had Bush handling the situation in Iraq at 40 percent approve and 56 percent disapprove; the Times poll shows 44 percent and 55 percent. Annenberg has Bush handling the nation's economy at 41 percent to 55 percent; the Times had 43 percent to 54 percent.

However, if you look at all the questions, not just the horserace there is uneasiness about what is happening to the country (56 percent think the country would be better off if it moved in a new direction, 58 percent think the country is seriously off on the wrong track -- which most polls are showing) and doubts about President Bush's presidency. On the other hand, Kerry needs to do better than he is about what his proposals are. Which will win stability or change -- we'll know on Nov. 2.

Comments

For what it's worth, it does seem like their party ID numbers are little generous to the Dems. Usually there should be about a five point spread between Dems and Republicans, perhaps even smaller for likely voters. Some recent survey data even showed Republicans pulling ahead in party ID in 2002, which hasn't happened since FDR. I would buy the argument that there is something of a shift back to slight Democratic identification since 2002, but 13 points? I think it's a bit high, and the generic congressional at 19 points is something from an alternate universe. Republican advantage in 1994, when they retook congress with a massive victory, was only in the 7-10 point range in most polls. I think Dowd has a legitimate beef on that question, but all of the other results are totally in line with other polls, so even if the LAT sample is slightly pro-Dem, that's not the reason the results are bad for Bush.

Remember, the FACTS are BIASED.

Other polls are showing that a pretty large majority of Americans believe Bush will win reelection. Given that, it doesn't surprise me that the congressional matchup would be so favorable to Democrats. Americans typically don't care much for one party rule and if they think Bush is going to win then they're going to at least try to make sure we don't have a repeat of the last three years of total Republican control.

Ironically, this scenario may very well result in Democratic one party rule, since it's quite plausible that Kerry will win and a Democratic Congress will come to power. We'll see...

Yo, PhillyGuy. One party rule? The Supremes? Thought so. Thank you.

The LA Times poll is a bit peculiar in that the Dem ID IS high relative to other national polls (even if not so relative to its own previous polls), but Bush's approval rating is actually a few points higher than other national polls.

But this brings up a broader point. One thing that a number of the most recent polls seem to be showing is that Bush's approval ratings seem to have gone up a point or two, but the head-to-head numbers for Bush have gone down a good number of points, from being a virtual dead heat to a real margin for Kerry.

I'm not sure how to explain this. Perhaps there just does reach a point in the public's thinking at which a voter stops dumping on the incumbent and starts looking more optimistically to a challenger to replace him. Given Kerry's low profile these last few weeks, it's hard otherwise to understand the somewhat counterintuitive movement suggested by these numbers.

This poll and other recent polls are showing what I call an unraveling. It is not a tidy, clear-cut instant shift it people's thinking, but as minds change, people do not put all the pieces together at the same time and the same pace. First one thing doesn't make sense, then another. It takes a while for an entire mind-set to reverse itself. And remember, the brainwashing that some people are confronting in some cases has been going on slowly and systematically for over 30 years. My guess is that soon, the neo-cons and the sychophantic followers will run out of fingers, and the dike will collapse in uncontrolled splendor.

You're the one that's brainwashed if you think a lot of people will change after 30 years of going one way. That said, I still think it'll be a solid Kerry win right now...

It can't be a solid win for Kerry if people don't change their minds. All I said, which you misinterpreted, Mimiru, is that there is pent-up frustration with Bush, that will gradually manifest itself, and eventually have some sudden bursts. If that doesn't happen, how can their be a "solid Kerry win right now?" I'm stating no expectation that the far right is going to become suddenly unbrainwashed, but rather that a significant number of middle-of-the-road people are gradually stomaching all they can take and there will be some manifestations of that that appear to be outlyers in the polls that are simply a sign of things about to happen.

I think Eldon is essentially correct but way overstating his case. In other words, the "unraveling" part, absolutely. The "last thirty years of brainwashing" part, not really. Last four is more like it.