When I checked my email a little while ago, I had an invitation to participate in Zogby's online polling, which I did. As a "reward" for responding, I got to read some Zogby reports, the most interesting of which was "Inside the Minds of Undecided Voters: an analysis of the Zogby/Williams Identity Poll"
The poll was conducted Aug 11-16, and beyond that there isn't any info about methodology. The report compares responses of Bush, Kerry and undecided voters, and there's no indication of how many actual undecideds were polled or MOE on the results.
Nevertheless, the poll provided some interesting data which I haven't seen discussed here, including what kind of french fries undecideds prefer (and a lot more relevant data). The answer is below the fold.
Undecideds prefer McDonald's fries over Burger King's by a wide margin. Before getting to more politcally interesting data about undecideds, I want to comment on the data for Bush and Kerry supporters. On virtually every important question, Bush and Kerry supporters disagree. On political questions they are almost diametrically opposite.
Some examples:
Bush favorable/unfavorable:
Bush supporters: 1/99
Kerry supporters: 99/1
Direction of country right/wrong
Bush supporters: 92/4
Kerry supporters: 1/99
Only 7% of Bush supporters thought Iraq was one of Bush's significant failures; 85% of Kerry supporters thought Iraq was a significant failure for Bush.
That's an amazingly polarized electorate, and just more support for the idea that polls like the recent Time or Newsweek polls are deeply flawed and any big swings are pretty unlikely. The race is nearly tied, and seems likely to stay that way unless one candidate scores on an issue that really resonates with the opposing candidates supporters. That hasn't happened yet.
Now the undecideds: I've seen a lot of disdain for undecideds on dKos, but the data indicates while these voters are not as strongly opinionated and probably not as well informed as Bush or Kerry supporters, they are still actively watching the race and are likely to vote. On virtually every issue or question, they fall somewhere between the Bush and Kerry extremes. The number or percentage of undecideds wasn't indicated, but Zogby polls at the time indicated about 6% undecided, with a maximum of 8% undecided in Zogby polls from other time periods.
Here's the one I find most significant - candidates favorable/unfavorable ratings (undecideds only):
Bush 40 fav / 60 unfav
Kerry 22 fav / 72 unfav
and the Bush job performance numbers (again undecideds only):
Bush positive job perf 18 / negative job perf 81
Also:
Satisfied with choice of candidates - both Bush and Kerry supporters are mostly satisfied with the choices, but undecideds say:
Satisfied with choice 9
Wish for other candidates 90
The implication seems pretty clear: undecideds don't like Bush, but they either don't like Kerry or are still waiting to be sold on Kerry.
When will they decide? 32% say after the debates, but 50% say they'll decide somewhere in the last week to just before they leave the voting booth. If you're anxiety prone, don't watch the polls until November 1 or maybe even November 3 (or if it's like 2000, just wait until inauguration day and see who shows up).
Two final comments: I think of the GOP as the party where "values" are pre-dominant over "results", but by about 65% to 43% more Dems want a candidate that shares their values than do GoOPers. Independents more than Dems or GoOPers want a candidate who's more intellectual than compassionate (the metaphor in the report was Tin Man vs. Scarecrow).
Lastly given the rock-hard positions of the Dems and Republicans, the nearly equal splits, and Kerry's relatively high unfavorables among undecideds, it seems pretty easy to see why Kerry has positioned himself the way he has and not gone strongly negative. People critical of the Kerry campaign should reflect on how Kerry might win the middle before offering their advice. The positions that we argue so strenuously for here may not be the ones that win the election.
This is long enough already, but there's a lot more data in the report. I'll be happy to see if answers any questions people might have.