Previous Â鶹´«Ã½AV analyses have shown that the 2004 election is quite different from any other in recent memory. For example, more than two in three Americans say they have given quite a lot of thought to the election, a proportion similar to, or even higher than, what Â鶹´«Ã½AV has measured in the past right before the election, when interest is typically at its peak. Additionally, Americans are more polarized in their views of George W. Bush than they have been of any other president seeking re-election -- with a more than 70 percentage-point divide in Republicans' and Democrats' approval ratings of him. Perhaps because of these two factors, Â鶹´«Ã½AV data indicate that Bush and Democratic opponent John Kerry could be vying for a much more limited pool of swing voters than has existed in previous elections.
Â鶹´«Ã½AV has a long history of asking voters about the likelihood of changing their stated vote intentions. However, several different question formats have been used over the years, which makes historical comparisons difficult. But Â鶹´«Ã½AV has consistently asked one version of a "swing voter" question since 1992 -- asking voters if there is "any chance" or "no chance whatsoever" that they will support a different candidate in November than the one they are currently supporting.
This question was asked in a June 21-23 CNN/USA Today/Â鶹´«Ã½AV poll*. Those data indicate that 20% of registered voters are still not fully committed to their preferred candidate, including 18% who say there is some chance they would vote for the candidate they are currently not supporting, and 2% with no opinion. In that poll, 43% of Kerry supporters and 37% of Bush supporters said they would not vote for the opponent. Among likely voters in that poll, the proportion of uncommitted voters was even lower, 13%, with an equal percentage of Bush and Kerry supporters (43% each) expressing an unwavering commitment to their candidate.
|
|
Registered voters |
Likely |
|
|
|
% |
% |
|
|
Some chance could vote for other candidate |
18 |
13 |
|
|
|
|
|
|
|
No chance vote for other candidate |
80 |
86 |
|
|
|
(Bush voter) |
(37) |
(43) |
|
|
(Kerry voter) |
(43) |
(43) |
|
|
|
|
|
|
No opinion |
2 |
1 |
|
Historical Comparisons on Swing Voters
In previous years during the pre-convention phase of the campaign, Â鶹´«Ã½AV has found no fewer than 3 in 10 registered voters in the uncommitted category, putting the current figures at least 10 points below comparable data from previous elections. For example, an April 2000 poll showed 31% of registered voters saying there was some chance they might vote for the other candidate or that they had no preference at the time. At that time, 38% of Bush supporters and 32% of Al Gore supporters were firmly in their candidate's corner.
|
|
Registered voters |
Likely |
|
|
|
% |
% |
|
|
Some chance could vote for other candidate |
28 |
27 |
|
|
|
|
|
|
|
No chance vote for other candidate |
70 |
72 |
|
|
|
(Bush voter) |
(38) |
(40) |
|
|
(Gore voter) |
(32) |
(32) |
|
|
|
|
|
|
No opinion |
3 |
1 |
|
In May 1996, a slightly higher proportion of registered voters, 42%, were still wavering in their vote choice to some degree. At that point in the campaign, 32% of registered voters were solidly backing Bill Clinton, 23% Bob Dole, and 4% Ross Perot.
|
|
Registered voters |
Likely |
|
|
|
% |
% |
|
|
Some chance could vote for other candidate |
39 |
-- |
|
|
|
|
|
|
|
No chance vote for other candidate |
59 |
-- |
|
|
|
(Dole voter) |
(23) |
-- |
|
|
(Clinton voter) |
(32) |
-- |
|
|
(Perot voter) |
(4) |
-- |
|
|
|
|
|
|
No opinion |
3 |
-- |
|
Of the recent campaigns, the 1992 contest swamps all recent elections in terms of the proportion of swing voters. In June of that year, little more than one in three registered voters had fully made up their minds. At that time, the presidential race was quite fluid, with independent candidate Perot leading in the polls over incumbent George H.W. Bush, with Democratic challenger Clinton running a distant third. But the July Democratic convention -- which coincided with Perot suspending his campaign -- helped Clinton surge to a lead he would not relinquish.
|
|
Registered voters |
Likely |
|
|
|
% |
% |
|
|
Some chance could vote for other candidate |
62 |
-- |
|
|
|
|
|
|
|
No chance vote for other candidate |
34 |
-- |
|
|
|
(Bush voter) |
(14) |
-- |
|
|
(Clinton voter) |
(10) |
-- |
|
|
(Perot voter) |
(10) |
-- |
|
|
|
|
|
|
No opinion |
4 |
-- |
|
Though not directly comparable, Â鶹´«Ã½AV did ask a similarly worded question in May 1988 -- "How much of a chance is there you will vote for [the other candidate]?" Those data showed about 4 in 10 registered voters saying there was some chance they would vote for the opposing candidate. Interestingly, in that poll, there were many more committed Michael Dukakis supporters (35%) than George H.W. Bush supporters (23%). However, as the campaign unfolded, that situation reversed.
Bottom Line
Presidential candidates have always devoted a sizable amount of their campaign and advertising budgets to attempting to persuade the swing voters. But with fewer Americans fitting that description this year, the candidates may shift their focus and resources to mobilizing their core supporters to vote than to trying to win over the small pool of swing voters.
*Results are based on telephone interviews with 909 registered voters, aged 18 and older, conducted June 21-23, 2004. For this sample, the maximum margin of error attributable to sampling and other random effects is ±4 percentage points.
For results based on the sample of 722 likely voters, the margin of sampling error is ±4 percentage points.
