Replies

A month ago another poll had it like this..

When asked whom they preferred in the 2004 Senate race, likely voters said they preferred Burr over Bowles by a 43 percent to 37 percent margin with 20 percent undecided. They said they liked Burr over Blue by a 45 percent to 33 percent margin, with 22 percent undecided.

The survey had a margin of error of plus or minus 4 percentage points. That means that the Burr-Bowles race could be a toss-up or Burr's edge could be as much as 14 percentage points and that Burr could lead Blue by as few as 4 percentage points or as many as 20 percentage points.

I also think it highly unlikely that Bowles had made up this much ground so quickly. I could understand a shift this large if Bush's numbers were tanking but these seem to have bottomed out lately and even risen slightly. Certainly, his NC numbers must be considerably higher than his national average. I have also not heard of any Burr gaffes or massive Bowles ad buys.

The poll in question could have been of North Carolina Adults or just Registered Voters. These samples tend to give the democrats an advantage over those that survey likely voters only.

If I remember correctly, several 2002 polls had the dems winning or very close in Senate races in North Carolina, New Hampshire, Colorado, South Carolina, Georgia and Minnesota. I'll have to check to see how these turned out!