You had to assume that there would be folks in the media who wouldn't like the recent CBS/New York Times poll that found strong public support for public workers. Sixty percent of those polled oppose stripping public workers of collective bargaining rights; 56 percent opposed cutting pay or benefits of those workers in the name of deficit reduction.
And the New York Times headline today reads "Majority in Poll Back Employees in Public Sector." But the poll is misleading because 20 percent of the responds say they are from union households. If you subtract them, those who favor cutting benefits win the poll. Wow, New York Times.
In case it wasn't clear, by "subtract them" he means exactly that–they shouldn't count in polls of the American public:
O'REILLY: You have a situation where you have 20 percent of the people polled being in the union families and you're telling me you can't throw that out? That that's a legitimate —
COLMES: You're saying — yes, I'm saying that you are conflating numbers, 8 percent of people in the polls say are union members, 20 percent say they are in union families.
O'REILLY: Yes, families.
COLMES: You automatically–but you don't automatically take that 20 percent and move into it the other column.
O'REILLY: I'm not moving it anywhere. I'm just taking it out of the mix.
Following this "logic," the elderly shouldn't be polled on Social Security, blacks shouldn't be polled on civil rights legislation and Democrats shouldn't be polled on Barack Obama's job performance.