A new statewide poll of 573 registered voters conducted by KAET-TV/Channel 8 and the Walter Cronkite School of Journalism and Mass Communication at Arizona State University October 19-21, 2004, found that the presidential race in Arizona has tightened and is again a statistical dead heat. Forty-seven percent of those surveyed said they will vote for President George Bush, 42 percent favor Sen. John Kerry, and 1 percent support Libertarian Michael Badnarik. Ten percent were undecided.

When undecided voters who were leaning toward Bush or Kerry were added to those who had already made up their minds, Bush led Kerry 50 percent to 45 percent, with Badnarik receiving 1 percent and 4 percent undecided. Among those most likely to vote on Election Day, Bush led Kerry 49 percent to 42 percent with 1 percent for Badnarik and 8 percent undecided.*

According to the KAET-ASU poll, the race for president in Arizona remains highly polarized. Ninety-five percent of Bush's supporters and 92 percent of Kerry's supporters said they are unlikely to change their minds between now and the election. The survey also found that while women were almost evenly divided in their choice for president (Bush 43 percent, Kerry 48 percent), men were much more likely to support Bush (50 percent) than Kerry (36 percent). Fifty-one percent of those interviewed approved of the job Bush is doing as president, 46 percent disapproved and 3 percent had no opinion. A majority (52 percent) approved of how Bush is handling terrorism, 47 percent approved of the way he is handling the economy and 44 percent approved of the way he is handling the war in Iraq.

Ninety percent of registered voters in Arizona said they had seen at least one of the presidential debates. Sixty-six percent said they saw all three. When asked whom they thought the debates favored, 47 percent said Kerry and 11 percent said Bush. Forty-two percent said they thought the candidates came out about even. Twenty-nine percent of the Republicans interviewed felt Kerry won, 19 percent felt Bush won and 52 percent said the debates were even. Although most voters felt Kerry won the debates, only 7 percent said the debates changed their mind about how they will vote. While the sub-sample of those who did admit to having a change of mind after the debates is too small for statistical examination, it appears the net effect of the debates in Arizona increased Kerry's percentage of the vote by 1 percent.

"Even though this survey shows that the race for president is within the margin of error, it is likely that President Bush leads in Arizona," says KAET-ASU poll director Dr. Bruce Merrill. "Two other polls within the past two weeks also have shown Bush with a lead somewhere between five and seven points. It really is true that the only poll that counts is the one on Election Day. Predicting who will actually go to the polls is very difficult. For that reason, and because of the margin of error, polls should not be used to predict the exact percentage candidates will receive on Election Day. "

The survey also found continued support for Proposition 200, which would deny certain state and local benefits to illegal immigrants and require proof of citizenship when registering to vote, declining since last month's poll when 63 percent of all registered voters said they would support the proposition, 23 percent were opposed and 14 percent were undecided. Fifty-four percent now support Proposition 200, 31 percent are opposed and 15 percent are undecided. Republicans (82 percent) and independents (75 percent) were more supportive of the proposition than Democrats (60 percent). The only other demographic that was related to the vote for Proposition 200 was education. The more formal education a voter possessed, the more likely he or she was to vote against the measure (high school or less, 59 percent for and 27 percent against; some college, 62 percent for and 26 percent against; and college graduates, 48 percent for and 37 percent against). The sample of Hispanics in this poll was not large enough to analyze how they might vote.

The statewide poll has a sampling error of plus or minus 4.1 percent. Forty-nine percent of those interviewed were male; 51 percent were female. Sixty percent of the interviews were conducted in Maricopa County, 17 percent in Pima County, and 23 percent in the less populated counties. The sample consisted of 40 percent Republicans, 35 percent Democrats and 25 percent independents.

For additional analysis of the KAET-ASU poll, watch KAET's Horizon program at 7pm Tuesday, October 26. Dr. Merrill will be a guest on the program to explain the meaning behind the numbers.

*A high-efficacy voter scale was developed based on how people responded to the following items: how important the outcome of the election is to them; how closely they follow politics; whether they read a newspaper on a daily basis; whether they watch a newscast on a daily basis; their level of education; how often they report voting in the past; and whether they believe they have a "say" in what the government does.

QUESTION WORDING

1. Overall, do you approve or disapprove of the job George Bush is doing as president of the United States?

2. Do you approve or disapprove of the way President Bush is handling the U.S. economy?

3. Do you approve or disapprove of the way President Bush is handling the war against terrorism?

4. Do you approve or disapprove of the way President Bush is handling the war in Iraq?

5. In the upcoming presidential election, will you probably vote for:

6. ASK ONLY IF VOTING FOR BUSH OR KERRY:
Would you say you are pretty firm in your support for (candidate chosen) or is there a possibility you might change your mind before election day?

7. Were you able to watch or listen to any of the recent presidential debates between Bush and Kerry?

8. IF SAW DEBATES:
How many of the debates did you see or hear?

9. IF SAW DEBATES:
Regardless of whom you may vote for, do you think the debates favored Bush, Kerry or were pretty much a tie?

10. Proposition 200 will require individuals in Arizona to provide proof of citizenship when registering to vote and show identification when voting. In addition, it will require state and local governments to verify the identity of people when they apply for non-federally mandated state and local benefits and would require state government workers to report people who they suspect of being in the country illegally. It would be a misdemeanor to fail to report such individuals. Will you vote for or against this initiative?