CBS Poll: Gore Has A Slight Edge

000830 gore bush
On the eve of Election Day, voters are just about evenly divided on which candidate they say they will vote for - with 5 percent of likely voters admitting they won't know for sure what they will do until sometime today.

In interviews conducted over the last three days - Saturday through Monday - 45 percent of likely voters say they are voting, or leaning towards voting, for Vice President Al Gore. A total of 44 percent say they are voting, or leaning towards voting, for Texas Gov. George W. Bush.

CBSNEWS - New York Times Polls







Among committed voters - those who say they won't change their minds, the race is just as tight, with 49 percent for Gore, 48 percent for Bush, and just 2 percent for Green Party candidate Ralph Nader. A total of 1 percent say they will for Reform Party nominee Patrick Buchanan.

Many voting patterns have been consistent throughout the fall. Bush has a 10-point lead among men, and Gore has an eight-point lead with women. Gore has a large lead in the Northeast, while Bush is dominant in the South. Gore has only a narrow edge in the West, while the Midwest - where many of the key battleground states are located - is evenly divided.

In earlier polling, Bush held a slight lead in the CBS News Poll. He still has several real advantages:

  • independent voters, critical to any victory, favor Bush, 42 to 36 percent.
  • Bush's voters remain more enthusiastic than Gore's: 54 percent say they are enthusiastic about voting for Bush, compared with 48 percent of Gore's
  • When asked who WILL win, the advantage is also with Bush: 43 percent of likely voters think Bush will win, while 35 percent say Gore will.
  • The very most likely voters - those who always vote and are paying a lot of attention to the campaign - favor Bush.

    In addition, even at this moment, there is room for change, and anything can happen. A total of 5 percent of all likely voters say they won't really make up their minds what they will do until today.

    This poll was conducted November 4-6, 2000, among a nationwide random sample of 1,527 adults interviewed by telephone. The sample includes 1,268 registered voters, and a proportionately weighted probable electorate of 806 likely voters (1,091 unweighted). The error due to sampling could be plus or minus three percentage points for results based on the samples of both registered and likely voters. The error for subgroups may be larger.