WASHINGTON (CNN) - A new national poll suggests that a majority of Americans don't think most members of Congress deserve to be re-elected. But most of the anger appears to be directed at Republicans rather than Democrats, who have controlled both houses of Congress for the past two years.
Fifty-eight percent of registered voters questioned in a CNN/Opinion Research Corp. poll released Wednesday said that most members of Congress do not deserve to be re-elected; 37 percent said most members should be returned to office.
"It's a "throw the bums out" sentiment similar to how the public felt in 1994, when Congress switched to GOP control, and 2006, when Congress switched back to Democratic control," said CNN Polling Director Keating Holland.
But which bums should be thrown out?
Fifty percent of those questioned said most Democratic members of Congress deserve to be re-elected, but that number drops to 36 percent for Republican lawmakers.
"This is just not shaping up to be a good year for Republican candidates," Holland said. "It's possible that voters are so angry at George W. Bush that they are thinking about voting against anyone whose name is followed by an 'R' in parentheses."
But while many voters are ready to "throw the bums out," that doesn't mean they think their own member of Congress is a bum. Fifty-five percent of those questioned said their member of Congress deserves re-election; 38 percent said no.
"Americans tend to dislike large groups, such as all bankers or all lawyers, but they also tend to like their own banker or their own lawyer - or their own member of Congress," Holland said. "It's a phenomenon that pollster David Moore calls the 'BIMBY effect' - Americans tend to believe that things are 'better in my back yard.' The BIMBY effect has helped incumbents get re-elected to Congress for years, and 2008 looks like it will follow the same pattern."
Democrats currently control 235 seats in the House of Representatives. Republicans control 199. One seat is vacant.
The CNN/Opinion Research Corp. Poll was conducted October 17-19, with 1,058 adults questioned by telephone. Because each question was asked of only half the respondents, the survey's sampling error is plus or minus 4.5 percentage points.