I contributed to this week's print National Journal as part of a special feature titled, "What Will This Election Hinge On?" In a piece called "The Uncertain," now posted online, I focused on the voters in the Diageo-Hotline tracking survey who are uncertain either about how they will vote or whether they will vote.
The pollsters at FD that conduct the Diageo-Hotline poll were kind enough to roll together and share the 2,449 registered voters interviewed as part of their daily tracking survey over the first few weeks of September. I was able to profile the 23% that were either totally undecided or uncertain about their choice (n=973), and the 9% that are only "probable" (not "definite") to vote (n=377), and compare these voters to the larger sample.
The print edition includes a short summary table, but I have put my complete data run in a table below (after the jump).
The results I found most interesting involved the voters that are undecided or uncertain about Obama and McCain:
[T]hese voters harbor doubts about the shortcomings they perceive in Obama and in McCain. By a 34-point margin (52 percent to 18 percent), they see McCain as "more prepared to lead the country" than Obama. And by a nearly opposite 31-point margin (50 percent to 21 percent), they say that Obama "better understands the needs and priorities" of people like them.
The key difference, omitted from the print piece, is that the Obama numbers on the "prepared" question was much lower among the uncertain voters than among all voters. Similarly, the McCain number on the "understands" question was lower among the uncertain voters than among all voters on the full sample.
This sense of uncertainty underscores the point about last Friday's debate that Marc Ambinder made over the weekend:
The first presidential debate was watched by tens of millions of people who were seeing the candidates discuss their views for the first time.
Both campaigns know that the most get-able voters at this point are the ones who are highly engaged with the race but tend to base their views on the highest, loudest levels of information.
The people most likely to move the poll numbers one way or another haven't been tuning into the 30 or so primary debates we've had; low information voters were the most relevant audience Friday night.
The rest of the data follows after the jump.
