Articles and Analysis


Party ID: The 2006 Edition

Topics: 2006 , IVR , IVR Polls , Party Weighing , The 2006 Race

We have devoted much attention recently to the flood of new national surveys showing small declines in the Bush job approval rating and modest Democratic gains on the generic House ballot question since mid-September. Until today, I had not looked closely at levels of party identification reported on those surveys. It turns out those have also trended Democratic recently, a finding that may explain some of the apparent "house effect" differences among statewide pollsters over the last few days.

The debate over weighting surveys by party identification has been a focus of this blog since its inception. My posts on the subject from 2004 and beyond are worth reviewing but the gist is this: Pollsters typically ask respondents some variant of a question asking whether they consider themselves "a Democrat, a Republican or an independent?" The so called "Party ID" question has been asked, examined and studied for more than 50 years, and an ongoing debate exists about whether to weight (or statistically adjust) survey results by party.

The crux of the debate is whether party identification is more like a fixed demographic characteristic (such as gender or race) or more like an attitude that can change with the prevailing political winds. For most adults, party identification does appear to be highly stable, changing rarely if ever. The problem is that some small portion of voters (perhaps 10% or 15%) appear willing to jump back and forth -- usually between one of the parties and the independent category -- depending on the wording of the question, its position in the survey, how hard the interviewer pushes for an answer, or, in some cases, what has been happening in the news.

Those who argue for weighting by party say that the real trends tend to be slow and gradual and that party weights can adjust dynamically over time to accommodate these slow moving trends (see also the party weighting page maintained by Prof. Alan Reifman). Those who argue against party weighting (a class that includes most of the national media pollsters) worry that such an approach will suppress real but short-term changes that sometimes occur in reaction to news events (such as the period just after the 9/11 attacks or the period just after the 2004 Republican convention).

A look at the party identification data from the recent surveys suggests we may be in the midst of another such short term change. The table that follows shows party identification results for six national surveys conducted before and after the resignation of Congressman Mark Foley. Five of six show some Democratic gain in party identification:


This change may also explain the wide divergence in results reported by the two automated pollsters in two nearly simultaneous surveys conducted this week in Missouri and Ohio. In both states, SurveyUSA showed the Democratic candidates with significantly greater leads (+14 in Ohio and +8 in Missouri) than Rasmussen (+6 in Ohio and -1 in Missouri). While both pollsters use the automated "interactive voice response" (IVR) methodology, Rasmussen weights by party and SurveyUSA does not. Moreover, the most recent SurveyUSA samples have grown more Democratic since August.

Does this shift in party identification represent a real shift in attitudes among the population of adults or registered voters or does it reflect some short enthusiasm among Democrats to be interviewed? Is the change a momentary spike or will it persist until Election Day? These are the questions that professional pollsters are mulling over right now, and the answers are not obvious. We will just have to wait and see (no pun intended).


Tony v:

A fair number of these concerns would seem to be testable.

Do the party identification numbers before election day match up?

More importantly, I would think a poll that doesn't control for party ID would show a lot of variance in both directions. It's been my impression that such polls tend to show surges a lot, and they altogether show a lot of movement towards one party or the other during changing time.

So is it that hard to do some sort of significance test on changing party ID?

This leads into a lot ofother questions, but the point is you really would imagine that these are scientifically answerable questions.


Alan Abramowitz:

These results almost certainly reflect a temporary spike in Democratic identification in some of the polls. It is highly unlikely, based on what we know about party loyalties in the electorate, that this will last. I would look for the next set of polls to show a return to normal, which may still be a slightly more Democratic electorate than a year or two ago.


Tony v:

Yes but that's the point. What people might vote for today may not reflect what they'll vote on Nov 7, but it still does reflect what they would vote for today. I mean it's not like you should remove the undecided category because no voters will be undecided that day either.

If different polls simultaneously show the same result, it means something. The argument for weighting would be much stronger if non-weighted polls showed more diversity.


Gary Kilbride:

I always place confidence in the PEW surveys in regard to party ID. They sample over a long period of time and the changes, if any, are subtle and sensible compared to previous PEW releases.

PEW was in the field when the Foley scandal broke, polling from September 21 - October 4. Their poll released about a week ago had the largest Democratic edge of the year in party ID at 34-27. I mentioned that on some sites, but perhaps not here. If you look at their prior party ID numbers it was more a case of Republicans detouring to call themselves independent, than respondents switching to Democrat. The 34 was identical to what Democrats had been receiving. In fact five consectuive PEW surveys -- and they've been sampling about once a month -- had Democratic party ID at either 33 or 34. Similarly, until the most recent sample, Republicans were at 29 or 30 for five straight surveys. This time the Republican number dipped to 27 while independent level jumped to 33. It was 30 in the previous sampling.

In the writeup regarding the generic ballot, PEW mentioned about half the poll was conducted before the Foley scandal, and half after the news broke. The numbers were 700+ in each category. PEW unfortunately didn't provide the party ID breakdown, before and after Foley. But since they sample frequently leading up to an election there will be at least one more PEW release, and likely two, before November 7.



Thanks for updating your 2004 piece. Many of us have been puzzled by Rasmussen appearing to be an 'oulier' on races and particularly GWB job approval. Given the historical discrepancy between late election term polling for DEMS vs. final votes, maybe an averaging of Rassmussen and SurveyUSA trends, for example, would offer a more accurate view of reality at a given point in time?

My perennial question- has anyone compared part id trends with unweighted presidential
approval ratings just prior to such a 'seismic' election?




Maybe, instead of a random poll, followed by weighting to reflect unexpected expressed party affiliation, pollsters should get lists of registered voters with party registration, and call people within those lists randomly, with any reweighting only to offset any difference in the response weights between the parties (and independents) as shown by their registration.


Post a comment

Please be patient while your comment posts - sometimes it takes a minute or two. To check your comment, please wait 60 seconds and click your browser's refresh button. Note that comments with three or more hyperlinks will be held for approval.